TAXONOMIC STUDIES AND GENERIC
DELIMITATION IN THE GRASS SUBTRIBE Sorghinae.
Moffat Pinkie Setshogo
A thesis submitted in fulfilment of the requirements for the
degree of Doctor of Philosophy
University of Edinburgh
March 1997
Dedicated to the memory of my father,
Tonkana, and to my mother, Kerileng.
Acknowledgements.
This work was carried out under the supervision of Dr. P.M. Smith. I wish to express
my sincere gratitude to him for the advice and assistance throughout the progress of
the study. I also want to thank Dr. C.E. Jeffree who has been very supportive and
proof read a substantial portion of the thesis.
I am indebted to the University of Botswana for the financial support and for offering
me a study leave to enable me to carry out this study.
The work was carried out at the Department of Botany, University of Edinburgh, as
well as at the Royal Botanic Garden, Edinburgh. I would like to extend my thanks to
the authorities of both institutions, and their staff, who offered help in many ways.
My collection of living material was cared for by Messrs Billy Adams and Bob
Astles. I wish to thank them for their help. My thanks also go to members of the
photographic unit of ICMB, particularly John Anthony, Dave Haswell and Frank
Johnston, for their help. Mr. John Findlay (Botany Department) gave me guidance
with my SEM work, for which I am grateful.
I am indebted to the Directors of various herbaria who loaned me specimens. Helen
Hoy and Marisa Main were in charge of the Edinburgh side of these loans. Also of
help were various seed sources around the world.
I thank Mr. W.M.M. Eddie, who helped me with PANKEY, NTSYS, PAUP and
MacClade computer programmes used in the numeric analysis.
Dr. T. Cope and Dr. G.V. Pope are well remembered for their help and useful
suggestions during my visits to RBGKew. Mrs Zoë Cooke and Mr Bill Eddie, fellow
PhD students deserve special thanks for their useful assistance, fruitful discussions
and for providing the 'buzz' in the lab.
My mother, brothers and sisters have been a source of inspiration for my studies here.
I pay my deep respect and thanks to them.
Lastly, I wish to express a very special thank you to my wife, Dineo, and son, Reba,
who sacrificed a lot of their emotions to enable me to see my studies to the end. Tanki
betsho!
LV
Contents.
Page
Title page.
Declaration.
Acknowledgements.
lv
Abstract.
x
Chapter 1. Introduction.
1
1.1. Project background.
1
1.2. Aims and objectives.
2
1.3. Sampling procedure.
3
Chapter 2. Classification.
9
2.1. Introduction.
9
2.2. Sorghinae and related Andropogoneae subtribes.
9
2.3. Taxonomic History of Sorghinae.
12
2.4. Nomenclatural origins of Sorghinae genera.
18
2.4.1. Hemisorghum C.E. Hubbard.
18
2.4.2. Spathia Ewart.
18
2.4.3. Pseudodichanthium Bor.
18
2.4.4. Cleistachne Benth.
19
2.4.5. Sorghum (L.) Moench.
20
2.4.6. Chrysopogon Trin.
21
2.4.7. Sorghastrum Nash.
22
2.4.8. Vetiveria Bory.
22
2.4.9. Euclasta Franch.
23
2.4.10. Pseudosorghum A..Camus.
24
2.4.11. Asthenochloa BUse.
25
2.4.12. Capillipedium Stapf.
25
2.4.13. Dichanthium Willemet.
25
2.4.14. Bothriochloa Kuntze.
26
2.5. Species of economic importance.
Chapter 3. Concepts of categories and generic delimitation in the Sorghinae.
28
30
3. 1. Introduction.
30
3.2. The subtribe concept.
30
3.3. The genus concept.
31
3.4. The species concept.
33
3.5. Subspecies.
36
3.6. Variety.
38
3.7. Race.
39
3.8. Generic delimitation in the Sorghinae.
40
3.8.1. The Sorghastrae.
45
3.8.2. The Bothriochloastrae.
50
3.9. Conclusion.
53
Chapter 4. Macromorphology.
54
4.1. Introduction.
54
4.2. Sources of plant material.
54
4.3. Evaluation of principal taxonomic characters.
54
4.3.1. Longevity and perennating organs.
54
4.3.2. Culms.
55
4.3.3. Leaves.
.55
4.3.4. Sheath.
56
4.3.5.Ligule.
.
56
4.3.6. Inflorescence.
57
4.3.7. Racemes.
63
4.3.8. Spikelets.
65
4.3.9. Sessile spikelet.
67
4.3.10. Pedicellate spikelet.
67
4.3.11. Homogamous spikelet pairs.
68
4.3.12. Callus.
69
4.3.13.Glumes.
69
4.3.14. Lower lemma.
71
4.3.15. Upper lemma.
72
4.3,16.Palea.
72
4.3.17. Awn.
73
Vi
4.3.18. Lodicules.
73
4.3.19. The androecium.
74
4.3.20. The gynoecium.
74
Chapter 5. Micromorphological studies with the Scanning Electron Microscope. 82
5.1. Introduction.
82
5.2. Materials and Methods.
82
5.3. Glume pits in the genera Bothriochloa and Dichanthium.
83
5.3.1. Introduction.
83
5.3.2. Glume pits.
83
5.3.3. Role of glume pit in cleistogamy.
88
5.3.4. Use of glume pit as a taxonomic character.
89
Chapter 6. Anatomical investigation of the leaf blade.
94
6.1. Introduction.
94
6.2. Materials and Methods.
95
6.2.1. Preparation.
95
6.2.2. Rinsing.
95
6.2.3. Dehydration.
95
6.2.4. Infiltration.
96
6.2.5. Embedding.
97
6.2.6. Sectioning.
97
6.2.7. Mounting ribbons.
97
6,2.8. De-waxing.
97
6.2.9. Staining.
98
6.2.10. Mounting.
98
6.2.11. Cleaning slides.
98
6.3. Evaluation of characters observed.
99
6.3.1. Outline of the leaf lamina in transverse section.
99
6.3.2. Epidermis.
99
6.3.3. Bulliform cells.
100
6.3.4. Vascular bundles.
101
6.4. Discussion.
101
6.5. Conclusion.
104
Vii
118
Chapter 7. Numeric analysis.
7.1. Introduction.
118
7.2. Phenetics.
119
7.2.1. Procedure.
120
7.2.2. Conclusion.
123
127
7.3. Phylogenetics.
7.3.1. Some concepts associated with cladistics.
129
7.3.2. Cladogram interpretation and evaluation.
132
7.3.3. Procedure.
134
7.4. General Discussion.
138
140
Chapter 8. Taxonomic descriptions.
8.1. Introduction.
140
8.2. Sorghinae Nees, Bluff & Schuaer.
141
8.2.1. Cleistachne Benth.
144
8.2.2. Sorghastrum Nash.
146
8.2.3. Asthenochloa Büse.
164
8.2.4. Hemisorghum C.E. Hubbard.
165
8.2.5. Sorghum (L.) Moench.
169
8.2.6. Pseudosorghum A. Camus.
185
8.2.7. Chrysopogon Trin.
187
8.2.8. VetiveriaBory.
210
8.2.9. Pseudodichanthium Bor.
221
8.2.10. Capillipedium Stapf.
223
8.2.11. Spathia Ewart.
233
8.2.12. Bothriochloa Kuntze.
235
8.2.13. Euclasta Franch.
258
8.2.14. Dichanthium Willemet.
263
Chapter 9. Distribution and ecogeography of the Andropogoneae and the Sorghinae.
278
9.1. The Andropogoneae.
278
9.2. The Sorghinae.
281
281
9.2.1. Cosmopolitan genera.
Viii
9.2.2. Discontinuous genera.
292
9.2.3. Endemic genera.
297
Chapter 10. References.
300
Chapter 11. Appendices.
314
ix
Abstract.
Genera and species of the grass subtribe Sorghinae were investigated. This study
covered taxa from the whole range of their morphological variation and geographic
distribution. In order to make the study feasible in the time permitted, a representative
sample was chosen. Parameters used in choosing this sample are outlined in Chapter
1. The objective of the study was to determine and ascertain generic boundaries in the
subtribe and hence reveal relationships between them. Fourteen genera are recognised
in the subtribe. These are Asthenochloa, Bothriochloa, Capillipedium, Chrysopogon,
Cleistachne, Dichanthium, Euclasta, Hemisorghum, Pseudodichanthium,
Pseudosorghum, Sorgahstrum, Sorghum, Spathia and Vetiveria. They were confirmed
to form clearly distinct genera despite suggestions by some authors to the contrary,
some of whom have advocated merging some genera in the subtribe. The genera
divide into two groups. The first group consists of Sorghum and allies, while the
second group is centred around Dichanthium (Chapter 3).
Numeric analysis were used to highlight both phenetic and phylogenetic
relationships between the genera (Chapter 7). The phenetic analysis broadly revealed
a somewhat similar picture to the one I held. Species of the same genus cluster in the
same place. Also those genera that are least similar separate at low phenon levels. The
cladogram produced, however, was totally different from my intuitive classification of
the subtribe.
Formal taxonomic descriptions incorporating keys and specimens seen can be
found in Chapter 8. An attempt was made to document a lot of information about the
taxa studied, e.g., chromosome number(s), even if these were not a direct result of the
research. A new combination in Euclasta, E. oligantha (Hotchst. ex Steud.) M.P.
Setshogo, was recognised (Appendix 11.2).
Three factors consistently affecting the taxonomy of the subtribe are suggested to
be (i) intergeneric hybridisation in the generic complex Bothriochloa-CapillipediumDichanthium; (ii) man induced interspecific hybridisation in Sorghum and (iii)
specific complexes whose member species overlap morphologically (Chapter 3).
x
Several sources of evidence were tapped to address these problems. These were the
use of micro- and macro-morphology and anatomy. The research was based on the use
of herbarium material as well as living plants. Leaf anatomy confirmed the C4
physiological anatomy associated with the grasses (Chapter 6). Scanning Electron
Microscopy (SEM) was used to study glume pits in the genera Bothriochloa and
Dichanthium (Chapter 5).
Distribution and ecogeography revealed that the subtribe Sorghinae occurs in the
tropics and subtropics of the Old World, with a high concentration of species and
genera in the Indian subcontinent (Chapter 9). Except for a few species of
Bothriochloa, Chrysopogon and Sorghum reportedly introduced into the Western
Hemisphere, only Sorghastrum, has a substantial representation there.
As well as producing a comprehensive revision of the subtribe, part of the work
was published in Flora Zambesiaca for those genera that occur in the region. The
manuscript sent for publication appears in Appendix 11.1.
Xi
Chapter 1. Introduction.
1.1. Project background.
It is generally recognised that the Andropogoneae is, in general terms, a well defined
taxonomic entity. This has resulted in most workers favouring the use of the tribal
rank because it is easier to recognise. While taxonomists differ in their views on
generic limits, especially in the larger genera such as Andropogon, there is almost
complete agreement on the limit of the tribe itself (Bews, 1929; Avdulov, 1931;
Hubbard, 1934; Keng, 1939; Pilger, 1940; Hitchcock, 1951). Hackel's work (in De
Candolle, Monographiae Phanerogamarum VI (1889)) forms the first general
account of the Andropogoneae. During Hackel's time, the genus Andropogon was
central to the taxonomy of the tribe. Stapf (in Pram, Flora of Tropical Africa IX
(1917)) started the splitting of this genus, and recognised generic groups in the
Andropogoneae. Taxonomic studies in the 30 years that followed were based on
Stapf s circumscription of the Andropogoneae genera. Following on from Stapf
(1917), the main taxonomic studies on the Andropogoneae were contributions from
floras and monographs (Hitchcock, 1935, 1936, 1951; Bor, 1940, 1960; Blake, 1944;
Backer & Brink, 1968; Clayton, 1972, 1982; Gould, 1975). None of these works,
however, delimited the genera into subtribes, but studied them as members of a wider
taxonomic entity, the tribe. An only attempt at a subtribal study was by Celarier
(1959), in the taxonomic study of his subtribe Sorgheae. Most of the research has
concentrated on those genera considered central to man's economic activities. In the
1950's, Celarier, Harlan, De Wet and their students began investigating the
cytogenetic mechanisms within the generic complex Bothriochloa-CapillipediumDichanthium. These studies concentrated on species complexes in these genera and
their relationships. Similarly, Snowden (1935, 1936, 1955) initiated taxonomic
studies on cultivated and fodder sorghums. Garber (1950) also did some
cytotaxonomic studies in sorghum. The work on sorghums was continued by De Wet,
Harlan and others in the 1960's and 1970's, and more recently, by Lazarides et al
(1991) who revised the sorghums of Australia. Another monographic work in the
Sorghinae is that of the genus Sorghastrum by Davila (1988).
From the above, the realisation dawned upon me that a thorough taxonomic
revision of the subtribe would be worth while. When studying a group as large as the
Sorghinae, information on the group will be found to be available and accessible at
varying levels of detail. Soliciting ideas from the rich literature that exists, and
incorporating my own observations should help strengthen and define the subtribe
and generic concepts applied to the Sorghinae. There are some genera (e.g.
Bothriochloa, Capillipedium, and Dichanthium) whose boundaries are reportedly
rather weak and perhaps arbitrary, and the evidence to keep them apart is not at all
convincing. There was a need to investigate these. Intergeneric hybrids with the
neighbouring subtribe Saccharinae have also been reported. Also apparent is the
morphological overlap between these subtribes.
1.2. Aims and objectives.
This thesis is presented primarily as an investigation intended to provide a taxonomic
revision of the subtribe Sorghinae. This is a world-wide revision of the subtribe. The
main objective of the study was to come up with a generic delimitation based on the
genus concept as applied within the subtribe. The study looked at a series of apparent
anomalies which could be best tackled by looking at the following related studies:
(a) a search for characters consistent enough to support keeping the genera separate
or, otherwise, combining them;
and (b) numerical/cladistic analysis of all data to seek signs of phenetic and
evolutionary relationships in the subtribe.
2
1.3. Sampling procedure.
There are about 152 species reported in the subtribe, so a taxonomic review of the
subtribe had to be representative enough to cover all the genera, the range of their
variability and geographic distribution. In order to make the study feasible in the
time permitted, a sample was chosen. A number of parameters were considered in
choosing this sample. After due consultation with the Flora Zambesiaca (FZ)
Editorial Committee at the Royal Botanic Gardens, Kew, an agreement was reached
that part of the research be a formal taxonomic account of the Flora of this region.
The FZ region covers five southern African countries of Botswana, Malawi,
Mozambique, Zambia and Zimbabwe. Therefore with this in mind, the following
criteria were used in determining the sample:
all Sorghinae taxa in FZ;
representative taxa (i.e., species and genera) of the subtribe that do not otherwise
occur inFZ;
taxa from a representative variety of habitats and ecological niches from which
the subtribe is normally found, and;
the material chosen was to reflect the karyogamy of the subtribe, i.e., if there are
diploids, tetraploids etc., the sample should include members from all these ploidy
levels. Information on this aspect of the sample was obtained from the literature.
Inevitably, availability of plant material (both herbarium and living specimens)
had an influence on the size and composition of the sample. The study has therefore
managed comprehensively to investigate 61 species (Table 1.1).
41
Table 1.1. Species distribution among Sorghinae genera and
representative numbers of species investigated in the study*.
Number
reported
Genus
Asthenochloa Buse
of
species
Number of these species in the
sample
1
1
Bothriochloa Kuntze
±35
10
Capillipedium Stapf
±14
5
Chrysopogon Trin.
±26
11
Cleistachne Benth.
1
1
±20
6
Euclasta Franch.
3
3
Hemisorghum C.E. Hubbard
2
2
Pseudodichanthium Bor
1
1
Pseudosorghum A. Camus
2
1
Sorghastrum Nash
±16
9
Sorghum Moench
±20
6
Spathia Ewart
1
1
Vetiveria Bory
±10
5
±152
61
Dichanthium Willemet
Total
* - The approximate number of species in the different genera follows Clayton &
Renvoize (1986).
The following is a conspectus of the genera and their species in the sample.
The list is arranged alphabetically.
Asthenochloa BUse, in Miq. P1. Jungh.: 367 (1854).
Asthenochloa tenera BUse, in Miq. P1. Jungh.: 368 (1854).
Bothriochloa Kuntze, Rev. Gen. 2: 762 (1891).
4
Bothriochloa biloba S.T. Blake, in Univ. Queensl. Papers, Dept. Biol. ,ii No.3: 27
(1944).
Bothriochloa bladhii (Retz.) S.T. Blake, in Proc. Roy. Soc. Queensl. 80: 62
(1969).
Bothriochloa decipiens (Hackel) C.E. Hubbard, in Kew Bull. 1934: 444 (1934),
Bothriochloa inscuipta (Hochst.) A. Camus, in Ann. Soc. Linn. Lyon 76: 165
(1931).
Bothriochloa ischaemum (L.) Keng, in Contrib. Biol. Lab. Sc. Soc. China, Bot.
Ser. X: 201 (1936).
Bothnochloa kuntzeana (Hackel) Henrard, in Blumea 3: 456 (1940).
Bothriochloa macera (Steud.) S. T. Blake, in Proc. Roy. Soc. Queensl. 80: 64
(1969).
Bothriochloa pertusa (Willd.) A. Camus, in Ann. Soc. Linn. Lyon 76: 164 (1931).
Bothriochloa radicans (Lehm.) A. Camus, in Ann. Soc. Linn. Lyon 76: 164
(1931).
Bothriochloa saccharoides (Swartz) Rydb, in Brittonia 1: 81(1931).
Capillipedium Stapf, in Pram, Fl. Trop. Mr. IX: 169 (1917).
Capillipedium assimile A. Camus, in Lecomte, Fl. Gen. Indo-Chine VII: 314
(1922).
Capillipediumfihiculme Stapf, in Hook. Ic. P1.: t 3085 (1922).
Capillipedium huegelii (Hackel) Stapf, in Hook. Ic. P1.: t 3085 (1922).
Capillipedium parvzjlorum Stapf, in Pram, Fl. Trop. Afri. IX: 169 (1917).
Capillipedium spicigerum S.T. Blake, in Univ. Queensl. Papers, Dept. Biol., ii.
No.3: 43 (1944).
Chrysopogon Trin., Fund. Agrost.: 187 (1820).
5
Chrysopogon aciculatus Trin., Fund. Agrost.: 188 (1820).
Chrysopogon aucheri Stapf, in Kew Bull. 1907: 211(1907).
Chrysopogon fallax S. T. Blake, in J. M. Black, Fl. S. Austral., ed. 2: 60 (1943).
Chrysopogoñ fulvus (Spreng.) Chiov., Fl. Somala 1: 327 (1929).
Chrysopogon gryllus Trin., Fund. Agrost.: 188 (1820).
Chrysopogon latifolius S. T. Blake, in Univ. Queensl. Papers, Dept. Biol., ii.
No.3: 7 (1944).
Chrysopogon orientalis A. Camus, in Lecomte, F!. Gen. Indo-Chine 7: 332
(1922).
Chrysopogon pallidus Trin. ex Steud., Nom. ed. 2(i): 360 (1840).
Chrysopogon plumulosus Hochst., in Jahreshefte WUrtt. 3: 62 (1847).
Chrysopogon serrulatus Inn., in Mem. Acad. Peters. Ser. 6(2): 318 (1832).
Chrysopogon sylvaticus C. E. Hubbard, in Hook. Ic. P1.: t 3365 (1938).
Cleistachne Benth., in Hook. Ic. P1.: t 1379 (1882).
Cleistachne sorghoides Benth., in Hook. Ic. P1.: t 1379 (1882).
Dichanthium Willem., in Usteri, Ann. Bot. 18: 11(1796).
Dichanthium annulatum Stapf, in Pram, F!. Trop. Mr. IX: 178 (1917).
Dichanthium aristatum (Poir.) C. E. Hubbard, in Kew Bull. 1939: 654 (1939).
Dichanthium caricosum (L.) A. Camus, in Bull. Mus. Hist. Nat. Paris 27: 519
(1921).
Dichanthium foveolatum (Delile) Roberty, in Boissiera 9: 170 (1960).
Dichanthium micranthum T. A. Cope, in Publ. Cairo Univ. Herb. 78: 325 (1977).
Dichanthium sericeum (R. Br.) A. Camus, in Bull. Mus. list. Nat. Paris 27: 549
(1921).
Euclasta Franch., in Bull. Soc. list. Nat. Autun 8: 335 (1895).
Euclasta clarkei (Hack.) T. A. Cope, in Kew Bull. 35(3): 704 (1980).
Euclasta condylotricha Stapf, Fl. Trop. Afr. IX: 181 (1917).
Euclasta oligantha (Hochst. ex Steud.) M.P. Setshogo, comb. nov.
Hemisorghum C. E. Hubbard, in Bor, Grasses Burma Ceyl. Ind. & Pakist.: 687
(1960).
Hemisorghum mekongense (A. Camus) C. E. Hubbard, in Bor, Grasses Burma
Ceyl. Ind. & Pakist.: 687 (1960).
Hemisorghum venustum (Thw.) Clayton, in Kew Bull. 27(3): 448 (1972).
Pseudodichanthium Bor, in Indian For. 66: 271 (1940).
Pseudodichanthium serrafalcoides (Cooke et Stapf) Bor, in Indian For. 66: 272
(1940).
PseudosorghumA. Camus, in Bull. Mus. Hist. Nat. Paris 26: 662 (1920).
Pseudosorghum fasciculare A. Camus, in Bull. Mus. Hist. Nat. Paris 26: 662
(1920).
Sorghastrum Nash, in Britton, Man. F!. Northern U. 5.: 71(1901).
Sorghastrum elliottii (Mohr) Nash, North Amer. Flor. 17: 13 (1912).
Sorghastrumfriesii (Pilger) Pilger, in Notizbl. Bot. Gart. Berlin XIV: 96 (1938).
Sorghastrum fuscescens (Pilger) Clayton, in Kew Bull. 30 (3): 509 (1975).
Sorghastrum incompletum (Presi.) Nash var. bipennatum (Hackel) Davila,
Systematic revision of the genus Sorghastrum (1988).
Sorghastrum minarum (Nees) Hitchcock, Contr. U. S. Nat. Herb. 24: 501 (1927).
Sorghastrum nutans Nash, in Small, Fl. Southeast U. S.: 66 (1903).
Sorghustrum pogonostachyum (Stapf) Chippind. & Pole Evans, in Bot. Surv. S.
Afr., Mem.: 247 (1948).
7
Sorghastrum setosum (Griseb.) Hitchcock, Contr. U. S. Nat. Herb. 12(6): 195
(1909).
Sorghastrum stipoides (Kunth) Nash, in N. Amer. Fl. 17: 129 (1912).
Sorghum Moench, Meth.: 207 (1794) nom. conserv.
Sorghum arundinaceum (Desv.) Stapf, in Pram, Fl. Trop. Mr. IX: 113 (1917).
Sorghum halepense Pers., Syn. P1. 1: 101 (1805).
Sorghum nitidum Pers., Syn. P1. 1: 101 (1805).
Sorghum plumosum (R. Br.) Beauv., Agrost.: 132 (1812).
Sorghum stipoideum (Ewart & J. White) C. A. Gardner & C. E. Hubbard, in
Hook. Ic. P1. 34: t 3364 (1938).
Sorghum versicolor Anderss., in Peters, Reise Mossamb.: 563 (1863).
Spathia Ewart, in Ewart & Davies, Fl. N. Terr.: 26 (1917).
Spathia neurosa Ewart & Archer, in Ewart & Davies, F!. N. Terr.: 26 (1917).
Vetiveria Bory, in Lem.-Lisanc., Bull. Sc. Soc. Philom: 42 (1822).
Vetiveria elongata (R. Br.) Stapf ex C. E. Hubbard, in Kew Bull. 1934: 444
(1934).
Vetiveriafihipes (Benth.) C. E. Hubbard, in Kew Bull. 1934: 444 (1934).
Vetiveriafulvibarbis Stapf, in Pram, Fl. Trop. Mr. IX: 158 (1917).
Vetiveria nigritana Stapf, in Pram, F!. Trop. Mr. IX: 157 (1917).
Vetiveria zizanoides (L.) Nash, in Small, Fl. Southeast U. 5.: 67 (1903).
A full synonymy for all taxa in this sample has been incorporated in the taxonomic
descriptions (Chapter 8).
Chapter 2. Classification.
2.1. Introduction.
The grass family, Gramineae, has about 650 genera and 10 000 species, forming one
of the largest families in the flowering plants (Clayton & Renvoize, 1986). The
family is ecologically the most dominant and economically by far the most important
throughout the world (Heywood, 1993). It is widely dispersed in all the regions of the
world wherever vascular plants can survive and sometimes dominates the aspect of
the vegetation. There are about six subfamilies with more than fifty tribes. The
subtribe Sorghinae is in the tribe Andropogoneae of the subfamily Panicoideae. The
Andropogoneae includes about 85 genera and about 960 species (Clayton &
Renvoize, 1986). The tribe is tropical in distribution, particularly inhabiting the
savannah zone, extending into warm temperate regions. Clayton & Renvoize (1986)
divide the tribe into 11 subtribes. The subtribe Sorghinae reportedly consists of 14
genera. These are Asthenochloa BUse, Bothriochloa Kuntze, Capillipedium Stapf,
Chrysopogon Trin., Cleistachne Benth., Dichanthium Willemet, Euclasta Franch.,
Hemisorghum C.E. Hubbard, Pseudodichanthium Bor, Pseudosorghum A. Camus,
Sorghastrum Nash, Sorghum Moench, Spathia Ewart and Vetiveria Bory. Genera are
assembled into subtribes according to overall morphological similarity.
2.2. Sorghinae and other related Andropogoneae subtribes.
An attempt at a detailed discussion on subtribal classification of the Andropogoneae
will not be made here, but a brief outline is essential for a proper understanding of the
evolution and systematic position of the subtribe Sorghinae in relation to other
supposedly related subtribes of the Andropogoneae. The salient differences between
the different subtribes of the tribe are outlined in Table 2.1 below. These character
differences are used in the discussion that follows.
Table 2.1. Sorghinae and related Andropogoneae subtribes.
Subtribe
Descriptive notes
Saccharinae
Inflorescence terminal, of solitary, digitate or paniculate racemes. Spikelets paired,
alike (i.e., both fertile and bisexual); lower floret empty, represented by a lemma,
occasionally male.
Divides into two groups: (i) paniculate inflorescence (Sacchastrae) - have thin glumes
and weak awns (all these thought to be primitive states); plumose spikelets and
spreading callus hairs, and (ii) digitate inflorescence (Eulaliastrae)- variability in the
lower floret.
Germainiinae
Inflorescence terminal, of solitary or digitate racemes. Spikelets paired, dissimilar;
sessile spikelets male or barren, sometimes involucral at the base of the raceme;
pedicellate spikelet bisexual, sometimes suppressed at the base of the raceme.
Sorghinae
Inflorescence terminal or rarely axillary, of single or digitate or paniculate racemes,
the latter often in whorls. Homogamous spikelet pairs occasionally present at the base
of the raceme. Spikelets dissimilar; sessile spikelet bisexual, lower floret empty;
pedicellate spikelet male or neuter.
Andropogoninae
Inflorescence of single, paired or sometimes digitate racemes, these terminal, or
axillary and aggregated into a compound panicle. Homogamous spikelet pairs present
or absent. Spikelets paired, dissimilar; sessile spikelet bisexual; pedicellate spikelet
male or neuter.
According to Clayton & Renvoize (1992), evolution in the tribe
Andropogoneae has followed a course of increasing morphological complexity which
can be summarised as follows:
Modification of the panicle: reduction of a large terminal inflorescence to a short
single raceme, followed by expansion into a compound panicle.
Fragile racemes bearing paired spikelets: initially the spikelets in each pair were
similar, becoming dissimilar, with the pedicellate spikelet reduced in the more
advanced genera.
Emphasis on internode and pedicel evolution: the rachis internodes vary from
simple and slender in the primitive genera to shortened, swollen and fused with the
pedicel in the advanced genera.
10
The Sorghinae is often divided into two generic groups on the basis of
inflorescence type (Clayton, 1972), the Sorghastrae with particulate inflorescence,
and the Bothriochloastrae with racemose inflorescence. The Sorghastrae would
therefore be the primitive group on the basis of the inflorescence type. Using the
same reasoning, the Sorghastrae would therefore be closely related to the generic
group Sacchastrae of the subtribe Saccharinae (see Table 2.1). Both have primitive
paniculate inflorescences, and their common origin is suggested by the close
similarity between Saccharum and Sorghum (Clayton & Renvoize, 1986). Similarly,
the Bothriochloastrae would be closer to the Eulaliastrae of the Saccharinae and
share some features with genera in the Andropogoninae.
On the basis of the arguments advanced above, the subtribe Sorghinae would
therefore seem to be intermediate between the subtribes Saccharinae and
Andropogoninae.
The members of the Sorghinae (except Cleistachne, Asthenochloa and
Sorghastrum), like those of all other subtribes of the Andropogoneae (except
Dimeriinae), are characterised by the presence of the spikelets in pairs, one sessile,
the other pedicellate. It is common practice to use sexuality and the general shape of
these spikelet pairs to determine the evolutionary hierarchy in the tribe. Using this
criterion, the most primitive subtribe becomes the Saccharinae, since both spikelets
of the pair retain their fertility and are very similar. Both spikelets are awned, an
adaptation often associated with dispersal. In the Sorghinae, the sessile and the
pedicellate spikelets differ in sexuality. The pedicellate spikelet is always either male
or neuter and awnless, except in two species of Sorghastrum, S. fuscescens and S.
pogonostachyum, and one of Dichanthium, Dichanthium annulatum var. fecundum,
where the pedicellate spikelet is bisexual and awned. These species share most of
their other characters with members of their respective genera except for the well
developed and awned bisexual pedicellate spikelets. In most of the literature (Blake,
1944; Clayton & Renvoize, 1986), it has been suggested that these species show
affinities with members of the subtribe Saccharinae, especially Saccharum and
11
Eulalia. Evidence from morphological and cytogenetical studies however
overwhelms any suggestions that these species belong to these other genera.
Borgaonkar & De Wet (1964) demonstrated that the gene determining bisexual
pedicellate spikelets is inherited as a simple dominant character over male pedicellate
spikelets.
One other observable trend in the Sorghinae is the loss of fertility of the
pedicellate spikelet. The pedicellate spikelet can be either male or neuter, represented
by barren pedicels or completely suppressed. The genera Sorghum, Hemisorghum,
Pseudosorghum, Pseudodichanthium, Bothriochloa, Dichanthium, Euclasta,
Capillipedium, Spathia, Chrysopogon and Vetiveria, have the pedicellate spikelet
either male or neuter. The genera Sorghastrum and Asthenochloa have the pedicellate
spikelet represented by a barren pedicel. The pedicellate spikelet is completely
suppressed in Cleistachne.
It must be said that, among the subtribes of the Andropogoneae, an aberrant
position is represented by the genera Trachypogon and Germainia of the subtribe
Germainiinae. These genera have adopted the opposite course, reducing the sessile
spikelet and transforming it into a shield behind which the fertile pedicellate spikelet
shelters until maturity (Clayton, 1969). What is actually the case here is that the
sexuality of the spikelets has been reversed. The sessile spikelet is male or neuter
while the pedicellate spikelet is bisexual. There are however occasional specimens
with both spikelets fertile in certain species of Apocopis and Trachypogon, therefore
breaking down the consistency of spikelet sexuality as a character (Clayton, 1969).
2.3. Taxonomic history of subtribe Sorghinae.
A summary of the taxonomic history of the subtribe is given in Table 2.2. The
subtribe as known today, has its foundations in Bluff, Nees & Schauer' s (in Comp.
Fl. Germ. ed. 2(1): 46 (1836)) subtribe Sorgha. This became Sorghinae when the
ending -nae (for the subtribal rank) was incorporated as stipulated in the International
12
Code of Botanical Nomenclature. Bluff, Nees & Schauer followed Linnaeus (1753)
classification system. They recognised classes and orders in the angiosperms. They
classified all the grasses into Class Triandria Order Digynia. This was on the basis of
them having three (tn-) anthers (-andria) and two (di-) stigmas (-gynia). They put
Sorghum, Chrysopogon, Imperata, Heteropogon, Erianthus, and Andropogon in the
tribe Saccharinae. They then subdivided the Saccharinae into subtribes Sorgha and
Andropogones, mainly on the basis of the inflorescence type. Sorghum and
Chrysopogon were put in subtribe Sorgha, because of their paniculate inflorescences,
and the rest of the genera in subtribe Andropogones because of their racemose
inflorescences (These inflorescence types are described in Chapter 4). Nees Von
Esenbeck (1843) put Sorghum and Chrysopogon in the tribe Sacchareae (now a
synonym for the tribe Andropogoneae). These two genera were also put in the tribe
Andropogoneae by Steudel (1855), and Bentham & Hooker (1883). Bentham and
Hooker (1883) also put Cleistachne in tribe Tristigeneae (now a synonym of the tribe
Paniceae). Hackel (In De Candolle, Monographiae Phanerogamarum VI (1889))
subdivided the genus Andropogon of the subtribe Euandropogoneae into 13
subgenera. Members of the subtribe Sorghinae were interspersed in subgenera
Schizachyrium, Amphilophis, Sorghum, Vetiveria, Chrysopogon, Dichanthium, and
Heteropogon. Cleistachne was treated as a separate genus of the same tribe. He
grouped his subgenera into two: Series A, the Isozygi, whose lowermost sessile
spikelets in the raceme are of the same sexuality and form as the upper ones, and
Series B, the Heterozygi, which have the lowermost sessile spikelets differing in
sexuality or form from the upper ones. Hooker (Flora of British India VII (1897))
followed Hackel's subdivision but hastened to add that the groupings of the
subgenera into Isozygi and Heterozygi tended to separate widely subgenera of close
affinity in habit and all other characters, except those adopted to distinguish between
these groups. He writes "Of the best defined I should, following older authors, have
preferred to have treated them as genera, especially Chrysopogon, Cymbopogon and
Heteropogon..." It was after this assertion made by Hooker that the realisation began
that Andropogon was too broad a genus and needed splitting. It was not until the
13
work of Stapf (in Pram, Flora of Tropical Africa IX (1917)) that the different genera
became widely recognised as distinct from Andropogon. Stapf not only recognised
them as distinct genera but demarcated them as the groups Sorghastrae and
Amphilophiatsrae of the tribe Andropogoneae. The Sorghastrae contained Sorghum,
Cleistachne, Vetiveria and Chrysopogon, while the Amphilophiatsrae consisted of
Capillipedium, Bothriochloa (as Amphilophis), Dichanthium, Euclasta, and
Eremopogon. He had in 1896 (In Hooker's Icones Plantarum 25: t 2494) removed the
genus Asthenochloa from the Andropogoneae and placed it in a different tribe
Agrostideae, "for it had affinities with Garnotia"! Keng (1939) proposed that the
Sorghastrae be given a status of a subtribe under the name Sorgheae. Pilger (1940)
independently came to the same conclusion and called his subtribe the Sorgenae. He
then put the genera Capillipedium, Bothriochloa, Euclasta, Pseudosorghum,
Dichanthium and Spathia, together with a number of other genera, in the subtribe
Andropogoninae.
As can be noted from the above historic outline, most genera in the subtribe
came into being due to the genus Andropogon being split up. The genera Sorghum
(1794), Dichanthium (1796), Chrysopogon (1822), Vetiveria (1822), Asthenochloa
(1854), and Cleistachne (1883) were described before Hackel's (1889) monograph of
the tribe Andropogoneae. The remaining members of the classically defined
Sorghinae were described in the late 19th century and early 20th century. These
genera are Bothriochloa Kuntze (1891), Euclasta Franch. (1895), Sorghastrum Nash
(1901), Capillipedium Stapf (1917), Spathia Ewart (1917), Pseudosorghum A. Camus
(1920), Pseudodichanthium Bor (1940) and Hemisorghum C. E. Hubbard (1960).
These genera, together with Asthenochloa and Cleistachne, show affinities to the
original four (Sorghum, Chrysopogon, Dichanthium and Vetiveria) and/or are
segregates of them.
The problem of the composition of the subtribe has persisted at length
resulting in Sorghinae genera being put in different and often unrelated tribes and
genera. Most of the classifications that followed were in agreement with Stapf (1917)
14
that these genera all belonged to the tribe Andropogoneae. Most workers (Bor, 1940,
1960; Celarier, 1960; Clayton, 1972) preferred to group the genera of the
Andropogoneae into informal generic groupings without according them any
taxonomic rank. However, these became so established with repeated usage that they
inevitably have bred a somewhat formalised classification. Celarier et al (1961)
recognised section Bothriochloininae of the subtribe Andropogoninae to include six
apparently closely related genera Bothriochloa, Capillipedium, Dichanthium,
Eremopogon (Hack.) Stapf, Euclasta and Pseudosorghum.
Clayton (1972) recognised three informal generic groups, two of which
contained genera now generally recognised to be in subtribe Sorghinae. These are:
Sorghastrae, containing Sorghum, Hemisorghum, Pseudosorghum, Cleistachne,
Asthenochloa, Vetiveria, and Chrysopogon.
Arthraxonastrae, made up of a single genus, Arthraxon.
Bothriochloastrae,
comprising
Bothriochloa, Capillipedium, Euclasta,
Dichanthium, Pseudodichanthium, Spathia and Eremopogon.
He puts these three generic groups in the subtribe Arthraxoninae which
include two additional genera, Arthraxon and Eremopogon (now a synonym of
Dichanthium).
Clayton & Renvoize (1986) suggest that the Sorghinae consists of two main
groups of genera, centred upon Sorghum and Dichanthium respectively, linked by
Pseudosorghum. Celarier (1958) came to a similar conclusion by making the
observation that Pseudosorghum linked his subtribes Sorgheae and Bothriochloeae to
which Sorghum and Dichanthium belonged respectively.
It should be noted that botanists still prefer to put all Sorghinae genera under
the tribe Andropogoneae. Certainly up to the present, the subtribal classification is far
from being stabilised. Clayton and Renvoize (1986) put all the fourteen genera in
subtribe Sorghinae. Simon (1990) puts eight of them, that occur in Australia, in
subtribe Arthraxoninae. Watson & Dallwitz (1992) also put all the fourteen genera in
15
-
subtribe Andropogoninae. They divide the tribe Andropogoneae into two subtribes,
the subtribe
Andropogoninae
('awned
Andropogoneae')
and the subtribe
Rottboelliinae ('awnless Andropogoneae'). The conclusions of the present study agree
with Clayton & Renvoize (1986) classification of the subtribe. I recognise two
generic groups, the Bothriochloastrae and the Sorghastrae, which I use in my
discussion on morphological trends in the Sorghinae (see Chapter 3).
16
Author (year)
Linnaeus (175 3)
Taxonomic rank
Polygamia
Monoecia
Persoon (1805)
Triandria Digynia
Bluff, Nees & Schauer
(1836)
Triandria Digynia
Tribe Saccharinae
Subtribe Sorgha
Sacchareae
Nees Von Esenbeck
(1843)
Steudel (1855)
Bentham & Hooker (1883)
Andropogoneae
Andropogoneae
Tristegineáe
Stapf (1917)
Sorghastrae
Amphilophiastrae
Keng (1939)
Sorgheae
Pilger (1940)
Sorgenae
Andropogoninae
Celarier (196 1)
Bothriochloininae
Clayton (1972)
Arthraxoninae
Clayton & Renvoize
(1986)
Sorghinae
Simon (1990)
Arthraxoninae
Watson & Dallwitz (1992)
Andropogoninae
('Awned
Andropogoneae')
Genera put in the taxonomic rank
Sorghum, Holcus, Andropogon, Musa,
Celtis, Veratrum, Ischaemum,
Cenchrus, Aegilops, Valantia, Atriplex,
Dalechampia, Acer, Begonia,
Ophioxylon.
Sorghum, and 60 other genera including
Andropogon, Agrostis, Hordeum,
Triticum, etc.
Sorghum, Chrysopogon.
Sorghum, Chrysopogon.
Sorghum, Chrysopogon, and 33 other
genera including Andropogon,
Anthistiria, Zoysia, Apluda, etc.
Sorghhum, Chrysopogon, Imperata,
Miscanthus, Saccharum, Erianthus,
Spodiopogon, Pollinia, Pogonatherum,
Apocopsis, Dimeria, Arthraxon,
Ischaemum, Andropogon, etc.
Cleistachne, Thunbergia, Limnas,
Garnotia, Arundinella, Phaenosperma,
Melinis, etc.
Sorghum, Cleistachne, Chrysopogon,
Vetiveria, Lasiorrhachis, Rhaphis.
Capillipedium, Amphilophis,
Dichanthium, Euclasta, Eremopogon.
Chrysopogon, Cleistachne, Sorghum,
Vetiveria, Lasiorrachis, Rhaphis.
Chrysopogon, Cleistachne, Sorghum,
Vetiveria, Lasiorrachis, Rhaphis.
Capillipedium, Bothriochloa, Euclasta,
Pseudosorghum, Dichanthium, Spathia,
Arthraxon, Hypogynium,
Pseudanthistiria, Eremopogon, etc.
Bothriochloa, Capillipedium,
Dichanthium, Eremopogon, Euclasta,
Pseudo,sorghum.
Bothriochloa, Capillipedium, Euclasta,
Dichanthium, Pseudodichanthium,
Spathia, Eremopogon, Arthraxon,
Sorghum, Hemisorghum,
Pseudosorghum, Cleistachne,
Asthenochloa, Vetiveria, Chrysopogon.
Bothriochloa, Capillipedium, Euclasta,
Dichanthium, Pseudodichanthium,
Spathia, Eremopogon, Sorghum,
Hemisorghum, Pseudosorghum,
Cleistachne, Asthenochloa, Vetiveria,
Chrysopogon.
Bothriochloa, Capillipedium,
Dichanthium, Spathia, Arthraxon,
Sorghum, Vetiveria, Chrysopogon.
Bothriochloa, Capillipedium, Euclasta,
Dichanthium, Pseudodichanthium,
Spathia, Eremopogon, Sorghum,
Hemisorghum, Pseudosorghum,
Cleistachne, Asthenochloa, Vetiveria,
Chrysopogon.
2.4. Nomenclatural origins, and types of Sorghinae genera.
2.4.1. Hemisorghum C.E. Hubbard, in Bor, Grasses of Burma, Ceylon, India &
Pakistan: 686 (1960) - H. mekongense (A. Camus) C.E. Hubbard.
The genus Hemisorghum used to be sunk in Sorghum as Sorghum halepense (L.) Pers
var. mekongense. A. Camus later described this species as Sorghum mekongense in
1922. Later synonyms of S. mekongense came from Rhind (1945), and Schmidt
(1958). Hubbard (1960) put it in a new genus, Hemisorghum. He distinguished it
from the species of Sorghum subgenus Sorghum by the slender, more loosely
spiculate, sessile or subsessile racemes, the awnless upper lemma, glabrous (not
ciliate) lodicules, and by the much reduced pedicellate spikelets'. The type species is
H. mekongense, and is distributed in Indo-China.
2.4.2. Spathia Ewart, in Ewart & Davies, Flora of the Northern Territory: 26 (1917)
S. neurosa Ewart & Archer.
The generic name is derived from Greek spathê, a blade or spatula, referring to the
enlarged leaf sheaths on the upper culm which constitute spathes. Ewart (1917) put
the genus Spathia in subtribe Euandropogoneae, with such genera as Ischaemum,
Pollinia and Andropogon. He suggested that it was intermediate between Pollinia and
Andropogon, but differed from both genera by having sheathing spathes, which are
probably a xerophytic adaptation. The type species is Spathia neurosa Ewart &
Archer, and is distributed south east of Newcastle Waters, Australia.
2.4.3. Pseudodichanthium Bor, in Indian Forester 66: 27 (1940) & in Hooker's
Icones Plantarum 36: t 3598 (1962)- P. serrafalcoides Bor.
the only species of the genus
Pseudodichanthium serrafalcoides is
Pseudodichanthium. The type species description is based on Andropogon
iI
serrafalcoides Cooke et Stapf. This genus was first placed in Andropogon and
Dichanthium presumably because of the imbricate spikelets of which the lower two or
three are homogamous. Apart from these purely superficial similarities the genus is
absolutely distinct from Dichanthium. Rather it can be said to have some close
relationship with the genus Heteropogon. In both Pseudodichanthium and
Heteropogon the raceme is dorsiventral, the spikelets are winged and the lower axis
of the raceme is continuous. However, the sessile hermaphrodite spikelets in
Heteropogon are utterly different from the pedicellate, in Pseudodichanthium they are
very similar and all are broadly winged. In Heteropogon there is a very sharp barbed
callus to the fruit, there is no such callus in Pseudodichanthium (Bor, 1962). The
genus Pseudodichanthium was created by Bor (1954) based on the species
Dichanthium serrafalcoides (Cooke et Stapf) Blatt. et McCann. The main reason for
singling out this species of Dichanthium to form a new genus was because it differed
from other Dichanthium species in appearance, in the texture and disposition of the
glumes, in the pedicellate spikelet being larger than the sessile, and in the winged
glumes (Bor, 1940). The notation of the type species is as follows:
Pseudodichanthium serrafalcoides (Cooke et Stapf) Bor, comb. nov. Syn.
Andropogon cookei Stapf ex Woodrow, in Journ. Bomb. Nat. Hist. Soc. XII: 438
(1893). A. serrafalcoides Cooke et Stapf, in Kew Bull.: 450 (1908). Dichanthium
serrafalcoides Blatt. et McCann, in Journ. Bomb. Nat. Hist. Soc. XXXII: 426 (1928).
P. serrafalcoides is a species confined to Bombay, India, and the monsoon forests of
Oman.
2.4.4. Cleistachne Benth., in Hook. Ic. P1. 14: t 1379 (1882)- C. sorghoides Benth.
The generic name is derived from Greek kleistos, closed, and achne, chaff (husks),
alluding to the spikelets that remain closed even at maturity. This genus was put in
the tribe Tristegineae (now a synonym of the tribe Paniceae) by Bentham (1883). He
states that at first the plant resembles specimens of Sorghum fulvum, but the total
absence of the second spikelet (whether perfect or rudimentary) to each node or notch
19
removes it from the Andropogoneae, and brings it into connection with Arundinella!
It was moved back to the Andropogoneae subtribe Euandropogoneae by Hackel
(1889). The complete suppression of the pedicellate spikelet certainly can be
misleading as it misled Bentham but this genus shows a close similarity with
Sorghum and allies in a majority of its features. The type species is C. sorghoides
Benth., and occurs in tropical Africa and India.
2.4.5. Sorghum Moench, Meth.: 207 (1794) nom. conserv.; Hubbard, in Hook Ic. Fl.
34: t 3364 (1938); Garber, in Univers. California Pubi. Bot. 23: 283-361 (1950);
Snowden, in Journ. Linn. Soc. Bot. 55: 191-260 (1955); Celarier, in Cytologia 23:
395-418 (1959); Ivanyukovich & Doronina, Trudy Priki. Bot. Genet. Selek 69: 18-27
(1980)- S. bicolor (L.) Moench. Blumenbachia Koel., Descr. Gram.: 28 (1802) nom
rej. non Schrad. (1825)- B. halepensis (L.) Koel. Sarga Ewart & White, in Proc. Roy.
Soc. Victoria 23: 296 (1911)- S. stipoidea Ewart & White.
The generic name is derived from the Latin surgum, giant millet. The first taxonomic
description of Sorghum was given by Linnaeus (1753), under the name Holcus. Other
than those species now referred to Sorghum, Linnaeus also included several species
(H lanatus, H laxus etc.) that have since been transferred to another tribe, the
Aveneae (Celarier, 1959). The first use of Sorghum as a generic name was made by
Adanson (1763) but apparently he used it only as a substitute for Holcus of Linnaeus.
It remained for Moench (1794) to distinguish Sorghum from Holcus in those
materials originally considered under Holcus by Linnaeus (Celarier, 1959). Sorghum
was for a long time considered a subgenus of Andropogon but has now gained its
status as a distinct genus. Stapf (1919) in his study of the grasses of tropical Africa,
included in Sorghum the genus Sorghastrum, and his procedure was followed by
Garber (1950) who gave Sorghastrum a subgeneric rank. He put Sorghum in the
subtribe Sorghastrae. He split the subtribe into two main genera, Sorghum and
Cleistachne, and then subdivided the genus Sorghum into six subgenera as follows:
Eu-sorghum, Chaetosorghum, Heterosorghum, Sorghastrum, Para-sorghum, and
20
Stiposorghum. Celarier (1959) accepted most of this classification, but retained
Sorghastrum as a separate genus. Extensive taxonomic studies by Stapf (1919),
Snowden (1935, 1936, 1955), Garber (1950), Vinall et. al (1936) and others have
demonstrated that Sorghum is a genus of considerable variability. Most writers have
always preferred to treat the cultivated sorghums as taxonomically separate from the
wild ones. This has resulted in an ambiguous situation of giving the genus two type
species, one for the cultivated sorghums and the other for the wild ones. Doggett
(1970) suggested that the type species for the cultivated sorghums be S. bicolor (L.)
Moench, and S. arundinaceum Stapf be the type species for the wild ones. This
anomalous situation has been rectified by De Wet (1978). He relegated S.
arundinaceum to subspecies status as S. bicolor subsp. arundinaceum, which
therefore effectively means that the type species is truly S. bicolor.
2.4.6. Chrysopogon Trin., Fund. Agrost.: 187 (1822) nom. conserv.- C. gryllus (L.)
Trin. Rhaphis Lour., Fl. Cochin.: 552 (1790) nom. rejic.- R. trivialis Lour. (= C.
aciculatus). Pollinia Spreng., P1. Pugill. 2: 10 (1815) nom. rejic.- P. gryllus (L.)
Spreng. Centrophorum Trin., Fund. Agrost.: 106 (1822)- C. chinense Trin. (= C.
aciculatus). Trianthum Desv., Opusc.: 69 (1831). Chalcoelytrum Lunell, in Amer.
Midl. Nat. 4: 212 (1915).
The generic name Chrysopogon derives from Greek chrysos, golden, and pogon,
beard, alluding to the hairs on the inflorescence. The genus was described by Trinius
(1822) in his Fundamenta Agrostographiae. This name is a conserved name since
two other people had given it different names before Trinius. These were Louriero
(1790) who called it Rhaphis, and Sprengel (1815) who called it Pollinia. Trinius'
Chrysopogon consisted of two species: C. gryllus (Andropogon gryllus L.), and C.
aciculatus (A. aciculatus Retz.). The type species is C. gryllus (L.) Trin. Trinius'
description of Chrysopogon is so general that Davila (1988) suggested that it maybe
applied to Sorghastrum. She in fact suggested that the genus Chrysopogon is in part a
synonym of Sorghastrum. In 1883, Bentham & Hooker mentioned that the genera
21
Chrysopogon and Sorghum were related. They divided Chrysopogon into two
natural' sections: the typical one, section Chrysopogon, and section Stipoides, which
is exclusively American. Davila (1988) suggests that the latter section represents a
synonym of the genus Sorghastrum. Hackel (1889) put Chrysopogon in Andropogon
subgenus Chrysopogon.
2.4.7. Sorghastrum Nash, in Britton, Man. Fl. North. States: 71 (1901)- S.
avenaceum (Michaux) Nash (=S. nutans). Poranthera Raf., in Bull. Bot. Geneve 1:
221 (1830) non Rudge (1811)-P. nutans (L.) Jacks. Dipogon Steud., Nom. Bot. ed 2,
1: 518 (1840) nom. nud.- S. stipoides.
The first person to recognise Sorghastrum as a separate entity was Rafinesque (1830).
He proposed the name Poranthera for the genus. The name was rejected as it was a
homonym of Poranthera Rudge published in 1811 for a different genus. Fournier
(1881) included the genus Sorghastrum within Andropogon section Sorghastrum.
Fournier's classification has however been disputed by Celarier (1959), and Clayton
and Renvoize (1986) who argue that Andropogon and Sorghastrum represent
probably two different evolutionary lines in the Andropogoneae. Stapf (1917), in his
study of the grasses of tropical Africa, considered Sorghastrum to be only a
subdivision of Sorghum, and this procedure was followed by Garber (1950) who
however gave it a subgeneric rank. Nash (1901) recognised the present genus as a
different taxonomic entity. He gave it a new name Sorghastrum. He separated
Sorghastrum from Sorghum on the basis of the reduction of the pedicellate spikelets
to simple pedicels (i.e. lacking a spikelet) in Sorghastrum. The name is derived from
Sorghum and the Latin suffix -astrum, a poor imitation of, alluding to the
resemblance of the genus to Sorghum. The type species is S. nutans (L.) Nash.
2.4.8. Vetiveria Bory, in Lem-Lisanc, Bull. Sc. Soc. Philom.: 43 (1822)- V.
odoratissima Bory (= V. zizanioides). Lenormandia Steud., in Flora 33: 229 (1850).
Mandelorna Steud., Syn. P1. Glum. 1: 359 (1854)- M insignis Steud. (= V. nigritana).
22
The name Vetiveria is from the Tamil vettivêru, a root dug up. This name was given
to this genus in recognition of the fact that most of its members have aromatic roots.
The taxonomic history of this genus is probably best represented by that of the type
species V. zizanioides. This species, best known as the 'Khas Khas' or 'Vetiver', has
been popular with the peoples of Northern India for a very long time (see Table 2.3Uses). Prior to Bory's description of the genus in 1822, this species had been moved
from genus to genus. Specimens of the grass sent to Linnaeus and Retzius by Koenig
were given differing names by the two authors. Linnaeus (Mant. Alt.: 183 (1771))
described it as something new under the name Phalaris zizanioides, while Retzius
(Observ. Bot. 3: 43 (1783)) described it as Andropogon muricatus (Stapf, in Kew
Bull. (1906)). The latter name was subsequently and so repeatedly used by most
botanists that it gained some popularity. When Bory (1822) described his genus
Vetiveria, he used the epithet 'odoratissima', which alludes to the fragrance of the
roots of this grass. Hackel (In DC., Monogr. Phan. VI (1889)) placed this genus in
Andropogon subgenus Vetiveria. This was repeated by Hooker (Fl. Brit. Ind (1897)).
They both classified it as Andropogon squarrosus Linn. f The name A. squarrosus,
however, turned out to be wrong because it was based on a completely different
specimen from those of Phalaris zizanioides and Andropogon muricatus. R. Brown
(Prodr. Fl. Nov. Holl. (1810)) correctly identified the specimen as Chamaeraphis
spinescens. When the genus Andropogon was split up, and Vetiveria regained its
generic status, it was the earlier epithet, 'zizanioides', first coined by Linaeus (1771),
which was adopted. Therefore V. zizanioides (L.) Nash became the type species.
2.4.9. Euclasta Franch, in Bull. Soc. Hist. Nat. Autun 8: 335 (1895)- E. glumacea
Franch. (= E. condylotricha). Indochloa Bor, in Kew Bull. 9: 75 (1954)- I. clarkei
(Hack.) Bor.
This genus was described by Bor (1954) as a new genus Indochloa Bor. The
description was based on Dichanthium clarkei (Hack.) Haines. Hackel, originally
described the species as Andropogon clarkei Hack. (in Oesterr. Bot. Zeitschr. 41: 49
23
(1891)). Bor (1954) erected a new genus Indochloa, which included Dichanthium
clarkei, and Heteropogon oliganthus (Hochst. ex Steud.) Blatt. et McCann, because
he thought they had been misplaced, for the similarities between the two far
outweighed the differences. H oliganthus had been placed in Dichanthium because it
differed extensively with other Heteropogon species. The name Indochloa Bor, was
rejected however, because the genus had already been described as Euclasta by
Franchet in 1895, who based his description on E. glumacea (=E. condylotricha).
Cope (1980) moved species belonging to Indochoa to Euclasta in order to conform to
recent work on generic limits in the tribe Andropogoneae. He however moved I.
oligantha to Dichanthium. After a thorough examination, I made a new combination
and transferred this species to Euclasta as E. oligantha (Hochst. ex Steud.) M.P.
Setshogo. This species clearly belongs here (see taxonomic description). The type of
the genus is E. condylotricha (Hochst.) Stapf.
2.4.10. Pseudosorghum A. Camus, in Bull. Mus. Hist. Nat. Paris 26: 662 (1920)- P.
fasciculare (Roxb.) A. Camus.
This is an Asian genus described by A. Camus in 1920. Before then, its species were
put by Hackel (1889) in Andropogon subgenus Sorghum. Hackel had noticed that
species A. fascicularis Roxb. did not satisfactorily fit in Andropogon subgenus
Sorghum and hence preferred to list it among 'species dubious' at the end of his
descriptions. Hooker (Fl. Brit. Ind VII: 177 (1897)) followed Hackel's subgeneric
classification but chose to include it in Andropogon subgenus Amphilophis. To settle
this confusion, Camus (1920) erected the new genus, Pseudosorghum, for this
species. It resembles Sorghum but differs in its contracted panicle. From Bothriochloa
(B. bladhii) it differs by its solid rachis internodes and pedicels. The type species of
the genus is P. fasciculare (Roxb.) A. Camus.
24
2.4.11. Asthenochloa BUse, in Miq., P1. Jungh.: 367 (1854)- A. tenera Büse.
Garnotiella Stapf, in Hook Ic. P1. 25: t 2494 (1896)- G. philipinensis Stapf (= A.
tenera).
This genus was described by Büse in 1854. Stapf (1896) placed it, as Garnotiella
Stapf, in the tribe Agrostideae. This genus, like most of the Sorghinae, was placed in
the genus Andropogon by Hackel (in DC., Monogr. Phan. VI, (1889)). He placed it in
Andropogon subgenus Schizachryrium as A. tenera Kunth. The type species is A.
tenera (Kunth.) BUse.
2.4.12. capillipediunz Stapf, in F!. Trop. Aft. 9: 169 (1917)-c. parvflorum (R.Br.)
Stapf. Filipedium Raiz. & Jam, in J. Bombay Nat. Hist. Soc. 49: 682 (1951)-F.
planipedicellatum (Bor) Raiz. & Jam.
The name Capillipedium is from Latin capillus, hair, and pes, foot, alluding to the
fine pedicels. This genus was created by Stapf (1917) to embrace a small number of
closely allied species from the tropical regions of Asia, Australia and Africa,
comprising Hackel's section Capillipedes of Andropogon subgenus Amphilophis
(Blake, 1944). Ohwi (in Acta Phytotax. & Geobot. 11: 165-166 (1942)) referred
species of Capillipedium to Bothriochloa, thus treating the genus as taxonomically
equivalent to Hackel's (1889) circumscription of Andropogon subgenus Amp hilop his.
Intergeneric hybrids have been produced between the genera Bothriochloa,
Capillipedium and Dichanthium through Bothriochloa bladhii (De Wet & Harlan,
1966). This led these authors to suggest that the three genera be combined, with the
generic name Dichanthium taking priority. The type species is C. parvflorum (R.Br.)
Stapf.
2.4.13. Dichanthium Willlemet, in Usteri, Ann. Bot. 18: 11 (1796); De Wet &
Harlan, in Bol. Soc. Arg. Bot. 12: 206-227 (1967)- D. nodosum Willemet (= D.
annulatum) Lepiocercis Trin., Fund. Agrost.: 203 (1822)- L. serrata (Retz.) Trin. (=
25
D. annulatum). Diplasanthum Desv., Opusc.: 66 (183 1)- D. lanosum Desv. (= D.
caricosum). Eremopogon Stapf, Fl. Trop. Aft. 9: 182 (1917)- E. foveolatus (Del.)
Stapf.
The generic name derives from the Greek words dicha meaning "at variance", and
anthos, flower, referring to the lowest spikelet pairs which differ from the upper pairs
in sexuality (Wheeler, Jacobs & Norton, 1982). Bentham's (1878) concept of
Andropogon included Schizachyrium Nees, Cymbopogon Spreng., Hyparrhenia
Anderss. ex Stapf, and Dichanthium Wilim., as well as the group of species later
referred to Amphilophis (Trin.) Nash and Bothriochloa Kuntze (Blake, 1969).
Dichanthium was treated as a subgenus of Andropogon L. by Hackel (1889, 1896).
This classification was later adopted by Hooker and Stapf (1896) and Hitchcock
(1935). However, Stapf (1917) and Camus (1921) suggested that the genera
Andropogon and Dichanthium differed from each other so conspicuously that both
deserved generic rank, and Blake (1944) and Bor (1960) further emphasised the
necessity of recognising both genera. Roberty (in Boissiera 9: 156 (1960)) treated
Dichanthium and Bothriochloa as congeneric with Eremopogon, Euclasta and
Pseudodichanthium, but retained Capillipedium as a distinct genus. The type species
is D. annulatum (Forssk.) Stapf.
2.4.14. Bothriochloa Kuntze, Rev. Gem. P1. 2: 762 (1891)- B. anamatica Kuntze (=
B. bladhii). Andropogon subgen. Gymnandropogon Nees, Fl. Aft. Austr.: 103 (1841)A. radicans Lehm. Gymnandropogon (Nees) Duthie, in Atkinson, Gaz. N. W.Prov. &
Oude 10: 638 (1882). Andropogon subgen. Amphilophis Hack., in DC., Monogr.
Phan. VI (1889) and Hooker, Fl. Brit, Ind VII: 171 (1897). Amphilophis Nash, in
Britton, Man. Fl. North. States: 71(1901)-A. toneyanus Nash.
The generic name is derived from Greek bothros, a spot, and chloa, grass, which
alludes to the pitted glumes of some species of this genus. Bothriochloa was initially
often included in the subgenus Amphilophis of the genus Andropogon. This subgenus
was described by Trinius (1833), and was later adopted by Hackel (1889). It was
26
raised to generic rank by Nash (190 1) and was called Amphilophis. Nash presented a
more detailed description of the genus and described many species. However, Kuntze
(1891) had earlier described the group as a new genus, Bothriochloa, based on a
specimen of Andropogon glaber Roxb. He called the new type species B. anamatica
(= B. bladhii). Bothriochloa, Capillipedium, Eremopogon, and Euclasta were
suggested as a group by Stapf (1917) under the name Amphilophiastrae, and treated
as a 'section' under the name Bothriochloininae by Celarier et al (1961). Ohwi (in
Acta Phytotax. & Geobot. II: 165-166 (1942)) treated Bothriochloa as taxonomically
equivalent to Hackel's circumscription of Andropogon subgenus Amphilophis. He
also combined Bothriochloa and Capillipedium on the basis of morphological
similarities. Gardner (in Fl. W. Aus. 1: 325 (1952)) treated Bothriochloa as a section
of Dichanthium. Roberty (in Boissiera 9: 156 (1960)) likewise treated Dichanthium
and Bothriochloa as congeneric together with Eremopogon, Euclasta and
Pseudodichanthium. Celarier pioneered a cytotaxonomic study of Bothriochloa,
Capillipedium and Dichanthium at the University of Oklahoma in the mid 1950's. In a
series of papers from 1956 onwards, together with Harlan, De Wet, Mehra,
Borgaonkar and others, he made an extensive study of the breeding behaviour and
chromosomes of most of the species of these genera and produced a large number of
interspecific and intergeneric hybrids. The production of intergeneric hybrids led
these workers to suggest that Bothriochloa, Capillipedium and Dichanthium were not
genetically distinct. De Wet & Harlan (1966, 1967) treated Bothriochloa and
Capillipedium as sections of Dichanthium using the generic names as sectional
epithets although there are earlier epithets available in the required rank (Andropogon
section Amphilophis Trin. and section Apananthus Foum. for the first, and
Andropogon subgenus Amphilophis (Trin.) Hack. section Capillipedes Hack. for the
other). Most people who have worked with the genera, however, advocate keeping
them separate, and this I agree with and hence have treated the three genera separately
in the present study. The type species is B. bladhii (Retz.) S.T. Blake.
27
2.5. Species of economic importance.
Most of the Sorghinae grasses are available as fodder. Sorghum is exploited for its
grain. Other members are cultivated for their aromatic oils (used in perfumery), and
as horticultural plants. The economic importance of some members of this group is
summarised in Table 2.3 below:
Table 2.3.: Species of economic importance in the Sorghinae: a
sample.
Species
Use
Bothriochloa bladhii
AF.
Bothriochloa pertusa
AF; LU; WS.
Bothriochloa inscuipta
AF; MI; LU.
Bothriochloa ischaemum
Bothriochloa radicans
LU; WS.
LU.
Bothriochloa saccharoides
AF; MT; WS.
Capil1pedium parvflorum
WS.
Chrysopogon aciculatus
LU; WS.
Chrysopogon gy11us
AF; MT; LU.
Chrysopogon plumulosus
AF; MT; LU.
Dichanthium annulatum
AF; HF; MD; MI; LU; WS.
Dichanthium aristatum
AF; WS.
Dichanthium caricosum
AF; LU; WS.
Dichanthiumfoveolatum
AF.
Dichanthium sericeum
LU.
Sorghastrum nutans
Sorghum arundinaceurn
AF; LU.
LU.
Sorghum bicolor
HF; WS.
Sorghum halepense
AF; WS.
Vetiveria nigritana
MT; LU.
Vetiveria zizanioides
CEO; MI; LU.
AF
animal food
MD medicine
CEO
commercial essential oils
MT materials
HF
human food
WS weed species
LU
land use
29
Chapter 3. Concepts of categories and generic
delimitation in the Sorghinae.
3.1. Introduction.
Botanical classification involves the assembling of plants into taxonomic groups
(taxa) on the basis of their relationships (Davis & Heywood, 1963). These
relationships may be assessed in different terms according to the purpose for which
the classification is made. They may be phenetic, genetic, biochemical, etc. These
groups are assembled into more inclusive groups, and these in turn into still more
inclusive groups. The different levels of groups so produced are recognised as a series
of hierarchical categories. Categories frequently used in the Gramineae are Family,
Subfamily, Tribe, Subtribe, Genus and Species. At the higher levels of the hierarchy,
no single diagnostic character may distinguish the taxa. The present project concerns
itself with the categories Subtribe, Genus and Species. Still, in a few cases, the
categories of Subspecies, Variety and Race are used. The following sections are
therefore an attempt to define these categories.
3.2. The subtribe concept.
Taxonomic delimitation of subtribes is based on facies differences shown by generic
groups. The definition of a subtribe is therefore solely based on the overall
morphological similarity of the genera. A subtribe, ideally, should consist of genera
which share certain characters. This description of the subtribe category makes it
difficult to define and most researchers have always preferred to use the tribe
category which is relatively easy to conceptualise.
Genera of the subtribe Sorghinae are characterised by the occurrence of
spikelets in pairs, one sessile, the other pedicellate. All sessile spikelets are bisexual,
while the pedicellate spikelet can be either male, neuter (sterile) or even absent
30
altogether (as in Asthenochloa, Cleistachne and Sorghastrum). The inflorescence can
be either a panicle of racemes, or consist of digitate or subdigitate racemes, or even a
single raceme. The Sorghinae genera, though sharing a common facies because of
their inflorescence morphology and spikelet configuration, have many other features
whose combinations make them distinct from each other. The overall distribution of
characters among Sorghinae genera is discussed in detail in Chapter 4.
3.3. The genus concept.
The concept of the genus is as old as folk science. Its most important function is to
bring related species together. Davis and Heywood (1963) suggested three primary
parameters that should provide guidelines for the delimitation of generic status,
mainly (i) naturalness, (ii) delimitation from closely related genera, and (iii)
practicability of keeping them distinct or including them in other genera.
One way of defining the naturalness of genera has been by the use of the
concept of morphology. Legendre and Vaillancourt (1969) defined the genus as a
monophyletic group (i.e., all descendants of a single ancestor) of species that
occupies a given adaptive zone. In order to be logical, they also introduced the
concept of phyletic closure, by which all the species descending from a common
ancestor have to be included in the same genus. Thus, the cladistic approach, seeking
monophylesis for generic delimitation, contains a 'valuable logical formation'
(Legendre, 1971). However, fears have been expressed that this approach will lead to
the splitting up of established, easily recognisable genera into many 'splinter groups'.
The goal is to develop a classification that will recognise monophyletic groups, but
disrupt the present classification as little as possible (Funk, 1985a). Pragmatism has
always outweighed other considerations in the establishment and circumscription of
genera. It may be preferable to split also monophyletic genera, however, if their
circumscription, i.e., their artificially chosen size, is too wide for a practicable
definition to be made (Bremer, 1978). The circumscription of genera has remained,
31
by and large, a function of the morphological cohesiveness and discreteness of the
species groups (Sivarajan, 1991).
The concept of the genus is more problematic in 'definable families' than in
'undefinable families' (Sivarajan, 1991). Character discontinuities may be difficult to
find in 'definable families'. Sivarajan (1991) identifies two attitudes governing our
concept of genus, which he calls puristic and pragmatic. The puristic approach
demands nothing short of a definition of genera based strictly on cladistic
relationships (Hennig, 1966; Stevens, 1985). However, the difficulty is that it requires
one to distinguish between primitive and advanced character-states, something which
can seldom be done with any confidence, because it involves too many speculative
conjectures to provide a satisfactory basis for defining the concept (Clayton, 1983).
The more pragmatic among the taxonomists hold the genus to be a 'composite idea'
which must seek to reconcile several parallel lines of thought, a multipurpose system
(Clayton, 1983).
The formal definition of the genus remains substantially the same, i.e., an
abstract and arbitrary level of morphological difference. The key question is where
and how we draw the delimiting lines. Clayton (1983) has discussed this problem in
great detail and has explained a few operational criteria for generic definition
(selection of characters, extent and density of the gap between adjacent clusters, and
predictive value). Cladists still disagree, and hold that genera can, and should, be
based phylogenetically (Sivarajan, 1991). This would be to define a taxon in terms of
the historical process deemed to have produced it - an educated guess, rather than in
terms of tangible phenetic pattern per se.
The problem of the origin of plant genera is discussed much less often than
that of the species and speciation, and most students of plant evolution are inclined to
believe that no processes are required for the origin of genera which are not operating
already in the origin of species (Stebbins, 1956). The Gramineae is one plant family
in which greatly divergent opinions have existed as to the status of genera. To what
extent are genera artificial aggregates of species, grouped together for the
32
convenience of taxonomists, and to what extent are they natural groupings which
reflect the result of evolutionary processes? From a practical standpoint, the genus is
an inclusive category whose species have more in common with each other than with
species of other genera within the same family. Genera, therefore, are aggregates of
closely related species. It is the sum total of all characteristics used to group closely
related species which helps conceptualise the genus as a unit. Genera must therefore
be delimited following a world-wide study of their species.
3.4. Species concepts.
Over the past century or so, many authors have proposed their own definitions of
what a species is. Species concepts in essence fall into three categories: the widely
referred to biological species concept based on reproductive isolation; phylogenetic
species concepts based on basic evolutionary units as revealed by phylogenetic
analysis; and morphological species concepts based on degrees of phenotypic or
genotypic distinctiveness. Some authors, e.g., Simpson (1961) add a fourth one, the
evolutionary species concept.
(i) Biological species concept - The definition of this concept was proposed just over
50 years ago by Mayr (1942), and can be summarised as follows:
A species is a group of interbreeding populations that are genetically
isolated from other groups by reproductive isolating mechanisms such as
hybrid sterility or mate acceptability.
This definition has been the target of much criticism particularly from
phylogeneticists. Donoghue (1985) noted that the biological species concept does not
necessarily yield monophyletic species, whilst Baum (1992) pointed out that the
potential for gene exchange is only loosely coupled to historical relatedness - the
central consideration of systematics. Further, many botanists, because of the general
ease with which plants can hybridise, have specifically rejected the biological species
33
concept. There are reported intergeneric and interspecific hybrids in the genera
Bothriochloa, Capillipedium, Dichanthium and Sorghum. In the strict sense of the
biological species concept, all the species involved would be regarded as belonging
to the same species because they belong to the same gene pool. This would lead to a
taxonomic 'grey area' if adopted where important differences could not be reflected
in classification: hence communication difficulties.
(ii) Phylogenetic species concepts - A number of species concepts have been
discussed in the last decade under the common title of phylogenetic species concepts.
These were mainly a reaction to the biological species concept and were thought of
as alternatives that did not rely on the knowledge of breeding potential. The main
difference between these and the biological species concept is that they do not
consider present biological characteristics directly but rather the acquisition of
defining features during evolution (Donoghue, 1985; Baum, 1992). The role of a
phylogenetic species concept is to reveal the smallest units that are analysable by
cladistic methods and are interpretable as the result of phylogenetic history. One
widely held definition is as follows:
A species is the smallest possible group of sexually reproducing
organisms that possesses at least one diagnostic character which is
present in all group members but is absent from all close relatives of the
group.
This definition has however been criticised for allowing the use of
plesiomorphies (primitive characters) as well as autapomorphies (uniquely derived
characters), and therefore species thus defined by a plesiomorphy could potentially
be paraphyletic. Therefore requirements for monophyly were added to the definition
(Donoghue, 1985; De Queiroz & Donoghue, 1988, 1990). Monophyly requires the
use of only synapomorphic characters (shared derived characters). Wheeler (1990)
points out that the application of the phylogenetic species concept would almost
34
certainly give far larger estimates of the total number of species than the more
traditional biological species concept.
Evolutionary species concept - The evolutionary species concept was first
proposed by Simpson (1951) though as with most other species concepts has
undergone a series of revisions and improvements. Wiley(1981) describes a species
a single lineage of ancestor-descendant populations which maintains its
identity from other such lineages and which has its own evolutionary
tendencies and historical fate.
This lineage concept of species emphasises reproductive continuity through
time. Many authors have pointed out that the division into 'biological' versus
'evolutionary' species concepts is unfortunate because both concepts (Mayr's and
Simpson's) are, in a sense, biological and evolutionary (Minelli, 1993). It is however
sufficient to note that the biological species concept envisages species as
reproductive communities whereas the evolutionary species concept is primarily
ecological. One other criticism is that this evolutionary species concept regards the
species taxon as a class of individuals sharing a common role in the 'economy of
nature' (Minelli, 1993). Changes in classes are restricted to the membership of
individuals within it, whereas a species actually changes or evolves as a whole.
Morphological species concept - Davis & Heywood define the morphological
species as:
assemblages of individuals with morphological features in common and
separable from other such assemblages by correlated morphological
discontinuities in a number of features.
It is intended as a generally applicable concept and takes into account all
available evidence, but insists that the species so recognised must be delimitable by
35
morphological characters. Morphological species may then be subjected to
experimental investigation and be shown to represent some particular evolutionary
situation. As a result of the evidence revealed by such study, the taxonomist may
consider the treatment of the individual case and redraw his specific boundaries,
provided they are still recognisable morphologically, but he may not be unduly
influenced by certain kinds of evidence (Valentine & Love, 1958).
The problem of species definition might be more or less resolved if it were
accepted that no one definition will suffice equally satisfactorily for all organisms
and perhaps for all purposes. It is indeed true that most taxonomists always draw
positive aspects of these species concepts to justify their studies without necessarily
admitting such actions. Definitions of species are concepts, which means that they
are basically pictures of patterns in nature. Interpretations of these pictures differ
from one person to the other.
In the present study, I used the morphological species concept to delimit
species. Species distinctiveness was recorded by recognising discontinuities in
morphology. Representative specimens from the geographical range of distribution
and morphological variation were investigated. It is my feeling that for a project of
this nature, morphological species are an essential first step and the key to further
information. I however did not ignore evidence from other taxonomic sources, but
rather incorporated this into my findings. It is simply helpful to collate as much
information about taxa being studied which will be an organised source of raw
material for further studies.
3.5. Subspecies.
These are groups of plants, more or less separated by a combination of characters,
but not usually isolated genetically (Valentine & LOve, 1958). They have sufficient
attributes in common that they are regional representatives of the one species. An
example is that of species Chrysopogon gryllus Trin. This species has two
subspecies, subspecies gryllus and subspecies echinulatus. Subspecies gryllus has a
remarkably disjunct distribution. In the western part of its range it is centred mainly
on the Mediterranean region and southeast Europe, extending eastwards to northern
Iraq and the Caucasus Mountains. There is then a gap of nearly 5 000km before an
eastern morphologically indistinguishable population is found in Assam, India.
Between these two centres lies subspecies echinulatus. This subspecies occurs from
the extreme northeastern part of Afghanistan eastwards through northern Pakistan
and Kashmir, and along the Himalayas to central Nepal. C. gryllus subspecies
echinulatus has (i) a contracted panicle with rather firm, erect branches; (ii) spikelets
in racemes of 2-3(-5) joints; and (iii) awns rather fine, 12-20mm long. C. gryllus
subspecies gryllus has (i) an open panicle with long capillary branches; (ii) spikelets
strictly in triads; and (iii) awns stout, 25-35mm long.
The subspecies concept also covers cases where taxa differ in chromosome
number or are partly or incompletely intersterile and exhibit some correlated
geographical or ecological differentiation but have an insufficient degree of
morphological differentiation to permit satisfactory treatment as separate species
(Heywood, 1958; Heslop-Harrison, 1957). The subspecies has therefore been widely
accepted as a considerable segment of a species with a distinct area and more or less
distinct morphology, often showing some intergradation, and clearly fulfils a useful
purpose (Davis & Heywood, 1963). Such variants need a name, for communication
purposes.
Subspecies are recognised by some authors as major morphological
subdivisions of the species without particular reference to geographical distribution.
This tendency has been condemned as an artificial means of reducing the number of
species. I am in no doubt that this was particularly the case in delimiting subspecies
in Sorghum bicolor (L.) Moench (the grain sorghums). This species was recognised
as 24 'species' by Snowden (1936), but all these are now considered variants of only
one species, S. bicolor. The species has since been divided into three subspecies on
37
the basis of whether they are cultivated, wild or appear to be a combination of the
two (hybrids). Subspecies bicolor contains all known cultivated grain sorghums.
Subspecies arundinaceum incorporates all wild sorghums associated with the grain
sorghums. Subspecies drummondii incorporates derivatives from hybridisation
between subspecies bicolor and arundinaceum. This highly artificial delimitation of
subspecies is not at all satisfactory and has not helped in resolving classification of
the grain sorghums. I have in the present study, for ease of understanding, retained
the following nomenclature: S. bicolor (strictly for S. bicolor subsp. bicolor) and S.
arundinaceum (for S. bicolor subsp. arundinaceum).
36. Variety.
This is a local facies of species morphologically distinct and occupying a restricted
area. Emphasis is on the small scale, more localised range of the variety, compared
with the large scale, regional basis of the subspecies (Heywood, 1959a). An example
is that of species Bothriochloa decipiens S.T. Blake. This species has two varieties,
var. decipiens and var. cloncurrensis. Both these varieties are endemic to Australia.
Var. decipiens occurs in New South Wales and Queensland, while var. cloncurrensis
is found only in Queensland. In addition, most people would agree that a variety has
few, minor (1 or 2 features) characters distinguishing it, and that it has a random
distribution, i.e., not ecogeographically distinct. Another concept would be of a
'weak' subspecies, i.e., ecogeographically distinct but with very few distinguishing
characters.
The terms subspecies and variety are often used interchangeably or regarded
as synonymous. In the Sorghinae there are varieties which to me would be best
regarded as subspecies. Sorghastrum incompletum has two varieties, var.
incompletum, which is restricted to tropical South America, and var. bipennatum,
which occurs only in the tropics and subtropics of central and southern Africa. In
addition, the main morphological difference between the two is the reduction of
racemes to a triplet of a single sessile spikelet flanked by two pedicels in var.
bipennatum. Var. incompletum has racemes which have 1-3 spikelet pairs. This may
involve several genes. One of the reasons put forward (Grant, 1960) for the
interchangeable use of subspecies and variety, and for the continued use of variety
where subspecies would fit best, is that, Linnaeus used the term variety, and hence
some people prefer to do likewise even in instances where it is inappropriate. This
seems a very unfortunate archaism.
3.7. Race.
A further category, race, is rarely inserted between the subspecies and variety to
cover geographical groupings which are intermediate between the two in a particular
situation (Davis & Heywood, 1963). The problem is that this category is sometimes
translated as meaning subspecies and can lead to serious problems of interpretation
and nomenclature (Heywood, 1958). Harlan & De Wet (1972) divide Sorghum
bicolor subsp. bicolor into five basic races, and ten hybrid races that combine
characteristics of any two or more basic races. The five basic races are Bicolor,
Guinea, Caudatum, Kaffir and Durra. Race Bicolor sorghums are grown across the
range of sorghum cultivation in Africa and Asia. Race Guinea is primarily a sorghum
of West Africa, but is grown in Chad, Sudan, north-eastern Nigeria and Uganda.
Race Kaffir sorghums are important staples across the eastern and southern savannah
from Tanzania to South Africa. Finally, race Durra sorghums are widely grown along
the fringes of the southern Sahara, across and West Africa, the Near East and parts of
India. De Wet (1978) similarly recognises four races for Sorghum bicolor subspecies
arundinaceum: race Aethiopicum of the and African Sahel; race Virgatum of northeastern Africa; race Arundinaceum of the African tropical forest; and race
Verticilliflorum of the African savannah. This therefore effectively means that there
are 19 races between the two subspecies, and is not taxonomically much of a
simplification or clarification (a faction of taxonomy). It would probably be better to
accept such races as groups of 'land races', in the sense of cultivated types, and not
to suggest that they have formal taxonomic status.
39
3.8. Generic delimitation in the Sorghinae.
The main objective of the project was to firstly determine generic boundaries in the
Sorghinae, to find out whether genera exist in the number that they have been
suggested to. Secondly to find out any unifying theme between all the Sorghinae
genera. As mentioned in the preceding section, it obviously becomes a lot more
difficult with the higher taxonomic categories to set up strict character limits within
which members should be found. Genera are relatively easy to conceptualise but very
difficult to define. Cytogeneticists have in many instances discovered that species
placed by all taxonomists in separate genera or even different tribes of the grass
family, show degrees of genetic relationship which in other groups are characteristic
of species belonging to the same genus. The genera Sorghum and Saccharum have
been reported to hybridise (Clayton & Renvoize, 1986).
Each genus includes genetically related and morphologically similar species,
and related genera usually include the extremes of what was once a single genus in
which the intermediary species became extinct (De Wet & Harlan, 1966). It has been
shown through experimentation and from field observations made by Celarier, De
Wet, and Harlan and their students in the 1950's, that intergeneric hybrids are
possible between the genera Bothriochloa, Capillipedium and Dichanthium. Hybrids
have been produced between morphologically similar genera in the Andropogoneae.
A series of hybrids have been produced between Saccharum as the one parent, and
Erianthus, Imperata, Miscanthidum, Miscanthus, Narenga, and Schlerostachya as the
pollen parents (De Wet & Harlan, 1966). The concept of monophyly, therefore, does
not hold in the instances quoted above. If one uses cladistic reasoning, there seems to
be paraphyletic generic groups in the Andropogoneae (see Chapter 7 for cladistic
concepts).
There are basically three situations that present confusion in the taxonomy of
subtribe Sorghinae. These are:
(a) intergeneric natural hybridisation, which has been proven experimentally, in the
generic complex. The species
Bothriochloa-Capillipedium-Dichanthium
40
Bothriochloa bladhii (Retz.) S.T. Blake, is capable of crossing with Capillipedium
assimile A. Camus, C. parvflorum (R. Br.) Stapf, Dichanthium annulatum (Forssk.)
Stapf (De Wet & Harlan, 1970). It is primarily a tetraploid species but 5n, 6n, and 8n
have been obtained (Harlan & De Wet, 1962). Over its complete distribution range,
extending from Southern Africa to Southern Australia, tetraploid B. bladhii
hybridises with tetraploid C. parvflorum (De Wet & Harlan, 1970). These hybrids
are commonly recognised as Bothriochloa glabra (Roxb.) A. Camus. B. glabra is
often sunk under B. bladhii (Faruqi, 1969) following the trend of regarding B. bladhii
as a very variable species. Simon (1989) proposed a rank of subspecies for B. glabra,
as B. bladhii subsp. glabra (Roxb.) Simon, a pragmatic move I totally agree with
because otherwise a lot of potentially valuable information would be lost if the taxon
is put into synonymy with B. bladhii. Also, according to De Wet (1987), in northern
Australia, B. bladhii introgresses with C. parvflorum, producing a series of
intergrades. One of the introgression products (in addition to B. glabra) is called
Capillipedium spicigerum Blake, which is said to be a product of a back-cross
between B. glabra and C. parvflorum (De Wet, 1987). The species B. glabra is
characteristic of African and South east Asian floras (Stapf, 1917; Bor, 1960), while
C. spicigerum is widely spread over the coastal and subcoastal districts of eastern
Australia. Blake (1944) however, questions the wisdom of considering the species C.
spicigerum a hybrid and contends that it is very unusual to find the three at all close
together, and that C. spicigerum is more abundant than either of its presumed parents.
I found the species a morphologically distinct entity and have therefore treated it as a
species in its own right. Also, according to Simon (1989), B. glabra does not occur
naturally in Australia. I have treated it as a morphologically distinct subspecies of B.
bladhii in this study.
Hybrids between B. bladhii and D. annulatum are usually recognisable as B.
grahamii and occur in the Gangentic plains of India, and scattered colonies in Africa
(De Wet & Harlan, 1970), across Southern Asia and in Australia. De Wet and Harlan
(1962) introduced a new term compilospecies, which they defined as a genetically
aggressive species, plundering related species of their heredities, and in some
41
instances completely assimilating a species, causing it to become extinct. They
applied this concept to the Bothriochloa-Dichanthium-Capillipedium complex based
upon the behaviour of Bothriochloa bladhii. Their studies suggested that B. bladhii
was itself of hybrid origin from germplasm assembled from at least 5 species
belonging to the three genera!
The genera Bothriochloa, Capillipedium and Dichanthium are otherwise very
distinct.
interspecific man-induced hybridisation, best exemplified in the genus Sorghum
(subgenus (Eu)Sorghum) which has resulted in a unidirectional improvement or
selection for desired traits. The species Sorghum bicolor (L.) Moench incorporates all
cultivated, semi-wild (subspecies
bicolor)
and wild species (subspecies
arundinaceum) of the genus Sorghum subgenus (Eu)Sorghum. This has resulted in a
very broad S. bicolor. To compound the situation further, nomenclatural subdivisions
occur under each subspecies into races and forms. Hybrids between subspecies
bicolor and arundinaceum, which tend to be intermediate between them are grouped
together under S. bicolor subsp. drummondii. Taxonomy is quite consistent in the
other subgenera of Sorghum,
which are
Stiposorghum, Chaetosorghum,
Heterosorghum and Parasorghum.
species complexes in the different genera, whose members overlap
morphologically and will cross with ease. These are noticeable as a group of species
whose boundaries are arbitrary and are separated by a few characters often correlated
with different geographic regions. Examples include the Dichanthium annulatum
complex which consists of Dichanthium annulatum, D. aristatum and D. caricosum;
the Chrysopogon fulvus complex which consists of Chrysopogon fulvus, C.
serrulatus, C. aucheri and C. plumulosus, etc. A detailed study of these would
constitute numerous other projects.
42
Despite the problems listed above, I found the genera of Sorghinae to be quite
distinct from a morphological standpoint. I think that one has to have operational
concepts to delimit taxa that one is studying. The study was based on
macromorphological investigation of a wide range of species in the Sorghinae. As
mentioned earlier, I admit that no single concept is adequate to wholly define a given
group of plants. It is probably this nature of taxonomic studies that always invokes
further research. It is common practice to divide subtribes into generic groups (Stapf,
1917; Pilger, 1954; Clayton, 1969, 1972). These generic groups end in '-astrae' and
have been applied informally to aid in taxonomic discussions without according them
any formal taxonomic rank. To aid in discussing morphological trends in the subtribe
Sorghinae, I have divided the subtribe into two generic groups, the Sorghastrae and
the Bothriochloastrae (following Clayton, 1972) (Figure 3.1).
43
Ii
led
ted
I
Figure 3.1 Putative relationships between Sorghinae genera (see text for details).
3.8.1. The Sorghastrae.
Members of this group have the following characteristics:
Paniculate inflorescence. Sessile spikelet dorsally or laterally compressed; callus
blunt or pungent; upper lemma bibbed (occasionally entire), awned from the
sinus. No homogamous spikelet pairs.
The genera found in this group are Sorghum, Hem isorghum, Pseudosorghum,
Sorghastrum, Cleistachne, Asthenochloa, Vetiveria and Chrysopogon. Within the
Sorghastrae, four clusters can be recognised: (i) Sorghum and allies (Hem isorghum,
Pseudosorghum) with well developed and dorsally compressed sessile and pedicellate
spikelets. In this group are found paniculate inflorescences which can be either open
(spreading) or compact, and usually with profusely divided branches (see Chapter 4).
Members of this cluster show affinities with those of the subtribe Saccharinae
(generic group Sacchastrae), differing, among other things, in the nature of the
pedicellate spikelet, which is bisexual in the Saccharinae; (ii) the VetiveriaChrysopogon alliance, with laterally compressed spikelets which sometimes display a
pungent callus and predominantly have an entire upper lemma. Inflorescences here,
though still a panicle, have simple pedunculate racemes. Also found here are awned
glumes and aculeolate lower glumes of both the sessile and pedicellate spikelets; (iii)
the Sorghastrum-Asthenochloa cluster with barren pedicels; and (iv) Cleistachne on
its own group with completely suppressed pedicellate spikelets. The following brief
accounts are an attempt to highlight the similarities between the different genera.
Sorghum- A genus with large simple or divided paniculate inflorescences. These bear
short dense racemes. Other characteristic features include convex lower glumes,
which are rounded on the flanks and 2-keeled near the tip, and ciliate lodicules.
Members of the subgenus (Eu) Sorghum are perhaps related to Saccharum with which
45
they will hybridise (Clayton & Renvoize, 1986). S. bicolor, the grain sorghum is a
member of this genus.
Sorghum is the only genus of the Sorghinae for which an extensive infrageneric classification exists, that I think is worth discussing in detail here. The most
comprehensive subgeneric treatment of the genus was initiated by Garber (195 0) who
recognised six subgenera. Garber's subgenus Sorghastrum has now been widely
accorded a generic status, and therefore only five subgenera now remain in the genus.
I agree with this treatment. The five subgenera are:
(Eu)Sorghum (Stapf) Garber - Old World, e.g. S. bicolor, S. halepense (introduced
to the Americas), etc.;
Chaetosorghum Garber - Australian; monotypic, e.g. S. macrospermum;
Heterosorghum Garber - Australian; monotypic, e.g. S. laxflorum;
Parasorghum (Snowden) Garber - Old World; e.g. S. nitidum, S. versicolor;
Stiposorghum Garber - Australian; e.g. S. plumosum, S. stipoideum.
Taxonomy in the subgenera Chaetosorghum, Heterosorghum, Parasorghum,
and Stiposorghum has been comprehensively investigated. As is often the case with
genera which include domesticated taxa, the subgenus (Eu)Sorghum contains many
nomenclatural inconsistencies, which has resulted in the existence of an extensive
synonymy. The subgenus consists of cultivated grain sorghums, a complex of closely
related annual taxa from Africa, and a complex of perennial taxa from Southern
Europe and Asia. The subgenus is usually divided into two groups, the Halepensia
complex, and the Arundinacea complex (De Wet, 1978). The subgenus is now
considered to consist of three species: two rhizomatous taxa, S. halepense, and S.
propinquin (both in the Halepensia), and the large and complex S. bicolor to include
all annual wild, weedy and cultivated taxa (in the Arundinacea). Snowden (1936)
recognised 24 'species' which are all now considered to be variants of S. bicolor.
The grain sorghums are artifacts of human selection and cultivation (De Wet
& Huckabay, 1967; De Wet, Harlan & Price, 1970), and the different cultivated kinds
46
deserve at most racial status (Harlan & De Wet, 1972). The species S. bicolor has
three subspecies:
S. bicolor subsp. bicolor (grain sorghums). Based on S. bicolor (L.) Moench. A
detailed comparative morphological study of grain sorghums led Harlan & De Wet
(1972) to recognise five basic races, and ten hybrid races that combine characteristics
of any two or more basic races. The five basic races are Bicolor, Guinea, Caudatum,
Kaffir, and Durra.
S. bicolor subsp. arundinaceum. Based on S. arundinaceum (Desv.) Stapf, this
subspecies incorporates all wild sorghums associated with the grain sorghums. It
therefore represents the 'wild' component of S. bicolor. De Wet (1978) recognises
four races of this subspecies: Aethiopicum, Virgatum, Arundinaceum and
Verticilliflorum.
S. bicolor subsp. drummondii. Based on S. drummondii (Steud.) Millsp. et Chase,
this subspecies incorporates derivatives from hybridisation between subspecies
bicolor and arundinaceum. It occurs as a weed in Africa wherever cultivated grain
sorghums and their closest wild relatives are sympatric (De Wet, 1978). This
subspecies is extremely variable, a variability undoubtedly enhanced by human
selection of grain races, which then introgress with the wild species (Clayton, 1982).
In the present treatment of the genus, for the sampled taxa belonging to the
subgenus (Eu)Sorghum, the following nomenclature has been retained for ease of
understanding:
S. halepense; S. bicolor (strictly for S. bicolor subsp. bicolor) and S. arundinaceum
(for S. bicolor subsp. arundinaceum).
Hemisorghum- This genus can be easily confused with members of the genus
Sorghum. The main differences lie in the flattened and glabrous lower glume of the
sessile spikelet which is 2-keeled for most of its length. It is because of this character
of the glumes that this genus has been suggested to form a link between Sorghum and
47
Saccharum (Clayton & Renvoize, 1986). Saccharum has flat broadly convex, 2keeled lower glumes while those of Sorghum are convex, rounded on the flanks and
becoming 2-keeled only near the tip. Also the lodicules in Hemisorghum are glabrous
in contrast to the ciliate ones of Sorghum. However, the resemblance is quite striking.
Pseudosorghum- This genus is also very similar to Sorghum. The inflorescence is
contracted or relatively dense, with subdivided but short primary branches, bearing
relatively long narrow racemes (up to 10 or more spikelet pairs - the racemes fairly
long, by contrast to Sorghum). The lodicules are glabrous. Members of this genus,
especially P. fasciculare, are often likened to Bothriochloa bladhii S.T. Blake. I must
say I find this misleading because, the inflorescence of B. bladhii, though paniculate,
has a completely different facies with spikelets which are relatively imbricate. Also
its primary branches are undivided (simple). I, however, agree with Clayton &
Renvoize (1986) that the genus probably forms a link between the two main generic
groups (Bothriochloastrae and Sorghastrae) of the Sorghinae, centred upon Sorghum
and Dichanthium. If the assumption of a change from a paniculate inflorescence to a
racemose one entails the reduction in the length of the inflorescence axis, the change
from divided primary branches to simple ones, and an increase in spikelet pairs (and
hence the raceme length), then the above description of Pseudosorghum
inflorescence would make it evolutionarily intermediate between these two generic
groups of the Sorghinae. Sorghum and allies have paniculate inflorescences with
very short dense racemes (2-3(-5) spikelet pairs) while Dichanthium and allies have
digitate or subdigitate inflorescences with long racemes (> 10 spikelet) pairs.
Sorghastrum- This genus is characterised by sessile spikelets which are accompanied
by conspicuous barren pedicels, very stiff membranous ligules and sheath auricles.
Inflorescence is a panicle with subdivided primary branches bearing short racemes,
these sometimes reduced to triads (S. incompletum var. bipennatum, S. minarum and
S. elliottii). Two species investigated in the study, S. fuscescens and S.
pogonostachyum, have many of the pedicels with bisexual (fertile) pedicellate
spikelets. Bisexual pedicellate spikelets are associated with members of the subtribe
Saccharinae, particularly Saccharum, hence these two species have been suggested to
indicate a link with this genus (Clayton & Renvoize, 1986). In addition to having the
bisexual pedicellate spikelet, S. pogonostachyum has an involucral beard at the base
of the spikelets, a character which again is found in the genus Saccharum. Other than
the sexuality of the pedicellate spikelet, the spikelets are so typical of Sorghastrum
that the species would be misplaced in any other genus. I have therefore retained them
in Sorghastrum.
Asthenochloa- The inflorescence is a panicle with variously divided branches. This is
one other genus of the subtribe with barren pedicels. In this instance, the pedicel has
been reduced to an inconspicuously minute axis hidden by callus hairs. Another
characteristic feature of the genus is the absence of lodicules, a condition often used
as an indication of evolutionary advancement. This genus also has the smallest sessile
spikelets (2mm) in the subtribe. The barren pedicels suggest a relationship with
Sorghastrum.
Cleistachne- The trend towards a suppression of the pedicellate spikelet which started
in Sorghastrum and Asthenochloa is taken further in Cleistachne, with the total
suppression of both the pedicellate spikelet and its pedicel. The complete raceme
actually consists of a single sessile spikelet such that the raceme peduncle appears to
be a pedicel. This contrasts with the situation in the genus Dimeria (subtribe
Dimeriinae), where, though the spikelets are single, in this case are pedicellate and
not sessile. Therefore in this genus the racemes consist of fertile (bisexual) pedicellate
spikelets arranged along a tough rachis. The sessile spikelet has been completely
suppressed. The ciliate lodicules, the convex lower glumes (rounded on the flanks)
and the obtuse callus, all ally Cleistachne to Sorghum.
49
As mentioned earlier, the remaining two genera of the Sorghastrae,
Chrysopogon and Vetiveria, differ from the other members by having paniculate
inflorescences with simple branches. The inflorescence consists of whorls of
pedunculate racemes at regular intervals along a common main axis. Additionally,
these genera have awned glumes. Chrysopogon has racemes which are single-jointed
while Vetiveria has racemes with more than one joint. The species, Chrysopogon
sylvaticus, which has 2-3-spikelet pairs, is often suggested to form a link between the
two genera. Lower glumes of the spikelets are often aculeolate (prickly) in most of
the species of the genera. The growth form of Vetiveria with characteristically
flabellate leaves and short culms distinguishes it from Chrysopogon. Similarly, the
long awned spikelets of Chrysopogon inflorescences give it a very different facies
from those of Vetiveria whose awns tend to be short and often inconspicuous.
3.8.2. The Bothriochloastrae.
Members of this group share the following characteristics:
Racemes solitary, digitate or paniculate; rachis internodes and pedicels often
with a translucent middle line. Sessile spikelet with a blunt callus; lower glume
often pitted; upper lemma entire, awned from the tip. Homogamous spikelets
sometimes present.
The genera found in this group are Bothriochloa, Capillipedium, Euclasta,
Dichanthium, Pseudodichanthium and Spathia. The paniculate inflorescence of
Capillipedium gives it a different facies from the other genera, but the general
similarity of the spikelet structure shows that it belongs to this generic group and not
to the Sorghastrae.
Dichanthium- Dichanthium may be recognised by its homogamous spikelet pairs and
obtuse sessile spikelets, but some species lack these features and approach closely to
Bothriochloa, from which they are separated by the solid pedicels. This genus has a
few species with pitted glumes. These can be distinguished from those of
Bothriochloa by the solid joints and pedicels.
Sessile spikelets of Bothriochloa tend to be acute at the apex compared to the
more obtuse apices of Dichanthium sessile spikelets. Dichanthium has many jointed
sessile / subsessile and digitate racemes. Therefore raceme peduncles are very short in
this genus.
The three genera Bothriochloa, Capillipedium and Dichanthium look very
different as can be seen from the observations made under each. The certainty and
precision of separating them however breaks down when one looks at the species
Bothriochloa bladhii, Capillipedium parvflorum and Dichanthium annulatum. These
species are highly variable, and in some instances, the specimens I have looked at
suggest some germplasm mixing between them. It is the behaviour of these species,
particularly B. bladhii, which prompted De Wet & Harlan (1966) to suggest a
treatment of these genera as one. However, I think other species of the three genera
are so different that they support keeping them separate.
Pseudodichanthium- This genus is presumably related to Dichanthium (Clayton &
Renvoize, 1986). A monotypic genus with imbricate spikelets arranged in short
solitary racemes. It has characteristic scarious lower glumes which are broadly
winged with inflexed margins. The cuims are slender, ascending from a slender base.
It could be misplaced in Dichanthium, but is very different in appearance, in texture
and disposition of the glumes, in the pedicellate spikelet being larger than the sessile,
and in the winged lower glumes. It resembles the genus Dichanthium only in the
imbricate spikelets.
Spathia- The genus has plump Sorghum-like spikelets, but the internode tip and
pedicellate spikelet callus suggest derivation from Dichanthium. This is the only
member of the Sorghinae which has a spatheate inflorescence, one other feature
associated with advancement in the Andropogoneae. These spatheate inflorescences
are both terminal and axillary. Also found at the base of the raceme are 1-3
homogamous spikelet pairs.
Capillipedium- A homogenous genus related to Bothriochloa. B. bladhii probably
lies midway between the two genera. Species of this genus have a very delicate
particulate inflorescence. In Bothriochloa, Capillipedium and Euclasta, the rachis
internodes and pedicels are grooved along the middle of their length, with thin tissue
in the groove. This is commonly referred to as a translucent middle line (furrow). The
absence of glume pits and the capillaceous nature of the inflorescence help in
distinguishing the two genera. It is quite a homogenous genus, and differentiation at
species level proved very demanding due to a close similarity between the species.
Bothriochloa-
The boundaries between
Bothriochloa, Capillipedium,
and
Dichanthium are somewhat blurred, largely due to a complex pattern of hybridisation
created by rapacious introgression of B. bladhii (Clayton & Renvoize, 1986). Some
Bothriochloa species generally have a depressed glandular area, glume pit, in the
middle or upper portion of the lower glume of the spikelets. These can be found on
the glumes of either or both the sessile and the pedicellate spikelets.
The genera Bothriochloa and Capillipedium are characterised by having the
lower sessile spikelets fertile (bisexual), the spikelet pairs on each raceme always
heterogamous on the branches of the inflorescence, and by the pedicels and rachis
internodes having a translucent longitudinal groove. Dichanthium has the lower
spikelet pairs on the racemes similar and unawned (homogamous, i.e., male or
neuter), no groove along the rachis internodes and pedicels, and frequently a subapical fringe of hairs on the lower glume. In addition it differs by the cuneate and
obtuse (rather than acute) glumes and the highly imbricate spikelets.
52
Euclasta- Euclasta has Bothriochloa-like pedicels and rachis internodes, i.e., these
have a translucent longitudinal groove. The inflorescence in Euclasta is a single
terminal raceme (except in E. condylotricha). The cuneate lower glumes of the
spikelets ally it to Dichanthium. It is probably because of this reason that the genus
was once wholly included in Dichanthium.
3.9. Conclusion.
From the observations made from the species sampled (see Chapter 8), it is my
contention that the genera of the Sorghinae are quite distinct. Separate taxonomic
revisions of the genera would undoubtedly reveal detailed evolutionary trends at
specific level, but I am in no doubt that these are all separate genera. There are two
generic groups, the Bothriochloastrae and the Sorghastrae, on the basis of the
characters of the inflorescence and spikelets. These groups are a useful aid to
taxonomic discussions because character differences between them are suggestive of
evolutionary trends in the subtribe. They are therefore not intended as formal
additions to nomenclature. Subtribal delimitation is based mainly on morphological
criteria. This unfortunately results in character overlaps between the Sorghinae and
other subtribes of the Andropogoneae, particularly the Saccharinae and the
Andropogoninae. The Saccharinae are visualised as closest to the ancestral stock.
Evolutionary trends in the Andropogoneae are often explained in terms of the
sexuality of the spikelets, the nature of the panicle, and the thickness of the rachis
internodes. The subtribe Sorghinae probably evolved from the Saccharinae by loss
of sexual function in the pedicellate spikelet (its reduction, loss and total suppression
in some genera), and modification of the inflorescence from a panicle through
subdigitate or digitate racemes to a single raceme. The so-called highly evolved
subtribes of the Andropogoneae tend to have racemose and spatheate inflorescences.
53
Chapter 4. Macromorphology.
4.1. Introduction.
This chapter is an evaluation of characters observed. According to Davis and
Heywood (1963), a character is any attribute (or descriptive phrase) referring to
form, structure and behaviour which the taxonomist separates from the whole
organism for a particular purpose such as comparison or interpretation. Characters
used in the present study were chosen after a thorough investigation of a wide range
of material. Character states cover a wide range of variation. Examples are drawn in
from the sample to put the discussion in context with the study. The same characters
were used in numeric studies (Chapter 7).
4.2. Sources of plant material.
Observations were made mostly from herbarium material. Living material for some
species grown in glasshouses was also used to complement information on characters
and helped enhance appreciation of form and its development. Herbarium material
consulted covered the whole range of the geographical distribution of the Sorghinae.
Specimens from the following herbaria were studied: BM, E, GAB, K, MAL, PRE,
UCGB, SRGH and WAG (codes following Index Herbariorum, 1981).
4.3. Evaluation of principal taxonomic characters.
4.3.1. Longevity and Perennating organs.
Most of the species in the sample are perennials. Annuals are found in the genera
Cleistachne, Euclasta, Pseudodichanthium, Pseudosorghum, Spathia and some
species of Sorghum. The subtribe therefore consists almost exclusively of perennials.
54
These perennials form loose or dense tufts. In some of the grasses, perennating
organs observable are rhizomes and stolons. Stolons are distinguished from rhizomes
in bearing complete green leaves, with sheath and blade, at each node, whereas
rhizomes develop only small, thin, whitish or brownish, scale-like leaves at these
points (Hubbard, 1968). Of interest here are rhizomes of the economically important
species Chrysopogon aciculatus and Sorghum halepense. Chrysopogon aciculatus is
used in lawns and playfields because of its rhizomatous creeping habit. However, the
plant needs constant trimming to keep it in the vegetative state lest it flowers and
produces its long pungent calli which are a menace. Sorghum halepense is very
difficult to eradicate and quite often becomes an obnoxious weed in cultivated fields.
4.3.2. Culms.
Culms of all the species are herbaceous. The orientation of these is variable. They
can be erect or decumbent and spreading. There is some considerable branching at
the base of the cuim that produces lateral shoots. These cuims die back annually and
are replaced by shoots from the basal buds. The perennial grass resulting from
several seasons' growth is thus made up of lateral shoots, hence the tufted habit.
Culms vary in the number of nodes, thus making this character inconsistent.
Cuim nodes can be either glabrous or hairy. Again this was found to be a highly
unreliable character. There is a considerable variation between specimens of the
same species and between nodes on the same plant specimen.
4.3.3. Leaves.
Leaf blades are mostly linear or linear-lanceolate in shape. They also are narrow
(<1cm) except in the genera Hemisorghum, Pseudosorghum and Sorghum, which
have characteristically broad (>1cm) and flat leaves. The leaf characters were found
to be highly variable.
All
4.3.4. Sheaths.
Typically a leaf sheath has the general shape of a hollow cylinder split down one
side. The margins of the sheath commonly overlap, both at the point of attachment
and for all or most of the length of the sheath. Projections of tissue termed auricles
may be developed laterally at the apex of the sheath or base of the blade. The former,
which are called sheath auricles, are present in the genus Sorghastrum. The other
genera do not have either sheath auricles or blade auricles. In two species of
Chrysopogon, C. fallax and C. pallidus, lower leaf sheaths disintegrate into fibres. It
is by this reason that the species are called "woolly-butt" in Queensland (Blake,
1944). These are often wrongly referred to as rootstocks.
The junction of the sheath and blade is typically marked by a band of
mechanical tissue that is different in appearance and texture from either the sheath or
the blade. The abaxial surface of this area is termed the collar. It is at the collar
region that the blade bends away from the sheath, and hence make the two organs
easily distinguishable from each other. The collar is conspicuous in most genera of
the Sorghinae except in Vetiveria. In this genus, both the sheath and blade are
sharply keeled and thus conduplicate. The collar is recognisable as a lateral
indentation, while the ligule is represented by a short unfringed membrane appearing
as a scar.
4.3.5. Ligule.
The ligule is usually a thin, white or brownish membrane, but in some grasses, it may
be a fringe of hairs or be absent. Sorghastrum species have a stiff ligule that is
usually divided into a rounded central lobe and two stiff, pointed lateral projections.
The latter have been referred to as sheath auricles. In the other genera of the subtribe,
variation of the ligule is mainly whether the membrane is fringed or unfringed. The
fringed membrane can be either ciliate or short toothed (lacerate). In none of the
56
material studied was the ligule absent. The type of ligule has been suggested to be
consistent for all species of a genus (Gould & Shaw, 1983).
4.3.6. Inflorescence.
The Sorghinae is characterised by inflorescences of paniculate, digitate or subdigitate
racemes. The subtribe divides into two on the basis of inflorescence type. These are
what I have called the particulate and racemose inflorescence types, and are defined as
follows (see Figures 4.1 - 4.4):
57
C
Figure 4.1 Inflorescence types in the Sorghinae (diagrammatic). A. Spatheate
inflorescence, e.g., Spathia neurosa. B. Paniculate inflorescence, branches simple (all
species of Chrysopogon). C. Racemes solitary, e.g., Euclasta clarkei, DichanIhium
inicranthurn.
IM
Figure 4.2 Inflorescence types in the Sorghinae (diagrammatic). Racemose
inflorescences: A. Subdigitate racemes, e.g., l3oihriochloa insculpia. B. Digitate
racemes, e.g., Dichanthium sericeum. C. Paniculate inflorescence, branches divided,
e.g., Sorghum arundinaceum.
59
IA
Figure 4.3 Inflorescence types in the Sorghinae (diagrammatic). Paniculate
inflorescences: A. Vetiveria species. B. Sorghastrum incompletum var. bipennaturn.
C. Sorghaslrum nutans.
all
Figure 4.4 Inflorescence types in the Sorghinae (diagrammatic). Paniculate
inflorescences: A. Capillipediuni parvifloruni, branches divided. B. Cieisiachne
sorghoides.
61
Paniculate inflorescence - an inflorescence having a central axis with few to
many simple branches or branches that are themselves divided, often profusely. These
are found in the genera Sorghum, Hemisorghum, Pseudosorghum, Sorghastrum,
Asthenochloa, Cleistachne, Capillipedium, Chrysopogon and Vetiveria. Clayton
(1972) refers to these as the Sorghastrae (excluding Capillipedium). In addition to the
paniculate inflorescence these are characterised by having an upper lemma bibbed
(occasionally entire) which is awned from the sinus. Two main clusters can be
recognised within the Sorghastrae: Sorghum and allies with dorsally flattened sessile
spikelets; and the Vetiveria-Chrysopogon alliance with laterally compressed sessile
spikelets, awned glumes, and panicles with verticellate racemes which give the
inflorescence a distinct facies.
Racemose inflorescence - spikelets borne directly on a jointed unbranched axis
(rachis). This inflorescence type is found in Bothriochloa, Dichanthium, Euclasta,
Pseudodichanthium and Spathia. These are the Bothriochloastrae. In addition to the
racemose inflorescence they have an entire upper lemma (barely wider than the awn
base) awned from the tip (except in two Australian species of Bothriochloa, B. biloba
and B. erianthoides, which have bibbed upper lemma). The racemes can be arranged
in one of the following ways:
A single raceme terminal or axillary e.g., Euclasta (except E.
condylotricha) and Pseudodichanthium species.
Two or more racemes arranged in one of the following ways:
radiating from the apex of the cuim, i.e. arranged in a whorl that
terminates the cuim. These are said to be arranged digitately, e.g.,
Spathia neurosa, Bothriochloa pertusa and Dichanthiium sericeum.
borne on a short central axis. When it is conspicuously shorter and
overtopped by the racemes, these are described as subdigitate, e.g.,
Bothriochloa inscuipta and Euclasta condylotricha.
It is, however, not uncommon to find a species with a paniculate inflorescence
in a genus predominantly characterised by species with racemose inflorescences, e.g.
Bothriochloa bladhii. Also a lot of other permutations occur in the group with
racemose inflorescences such that it is not necessarily correct to exclusively associate
digitate or subdigitate racemose inflorescences with a particular genus or indeed a
particular species.
4.3.7. Racemes.
Racemes of the Sorghinae are characterised by the occurrence of spikelets in pairs,
one sessile, the other pedicellate. The raceme terminates in a triad of 2 pedicellate
spikelets and 1 sessile, and those below in sessile/pedicellate pairs. A deviation from
this spikelet arrangement occurs in the genera Cleistachne, Asthenochloa,
Sorghastrum and Chrysopogon. The genus Chrysopogon has spikelets in a triad (1
sessile and 2 pedicellate) forming a raceme. In Cleistachne, the pedicellate spikelet is
completely suppressed, such that the raceme comprises of a single sessile spikelet. In
Asthenochloa, the pedicellate spikelet is represented by a minute axis hidden by
callus hairs. Finally, in Sorghastrum, the pedicellate spikelet is represented by a
barren pedicel, such that the spikelet 'pairs' along the raceme axis consist of a sessile
spikelet and a pedicel, and the terminal triad is a sessile spikelet and two barren
pedicels.
The racemes are mostly more than single-jointed in most of the genera. Joints
here referring to the number of spikelet pairs along the raceme axis (rachis). In
contrast to all the other Sorghinae genera, the genus Chrysopogon, is characterised
by racemes which are single-jointed (Figure 4.1). It is on the bases of this that
Clayton & Renvoize (1986) distinguish this genus from the genus Vetiveria. The
species Chrysopogon sylvaticus has been suggested to form a link between these
genera by having racemes with 2-to-3-joints. There are also some species in
Sorghastrum, e.g., S. incompletum var. bipennatum, and Capillipedium, e.g., C.
parvijiorum, that have single-jointed racemes. It has been a tradition to use the
63
number of spikelet pairs (raceme joints) in distinguishing species complexes in
Bothriochloa (Celarier & Harlan, 1955). A problem encountered in the present study
was incomplete racemes in the herbarium sheets due to the upper spikelet pairs
having disarticulated and fallen off.
One other feature consistently used as a character in genera with racemose
inflorescence, particularly Bothriochloa and Dichanthium, is the number of racemes
per inflorescence. The range, characteristic of a species, is usually present within a
single plant. For this reason, the average number of racemes per inflorescence is a
relatively stable character to use.
The rachis internodes are mostly solid except in Bothriochloa, Capillipedium,
and Euclasta. In these three genera can be found, running the length of both the
rachis internode and pedicel, a translucent longitudinal groove made of a hyaline
tissue in contrast to that of the raised lateral edges. This character has been used
consistently with a combination of other characters to identify species of these
genera.
The most characteristic feature of the Sorghinae is the fragile rachis
disarticulating with each pair of spikelets. The disseminule is thus a composite unit
made up of sessile spikelet, rachis internode, pedicel and pedicellate spikelet. The
raceme either disarticulates at the joints, when there are more than one spikelet pairs,
or falls entire, when it is single-jointed. The overall morphology of this unit is
variable in the Sorghinae. There are however various components which seem to
play a role in dispersal. From the morphological investigations made, the glumes,
awn, lemma of the upper floret, and the callus, appear to be the organs mostly
involved in dispersal. Even though not supported by any experimental proof, features
such as hairiness, size and lightness of these organs are likely to provide an extra aid.
64
4.3.8. Spikelets.
Most classifications of grasses are based on the structure and arrangement of their
spikelets. These organs, by the great diversity in their formation and organisation,
and by the amazing range in shapes and peculiar modifications of their separate parts,
provide a far greater variety of distinguishing characters than do other parts of the
grass plant.
Spikelet configuration in the Sorghinae is as follows (see Figure 4.5):
Sessile spikelet: (1) Glumes (2) (a) lower and (b) upper;
lower floret (empty or barren) represented by (a) a lower
lemma;
upper floret (bisexual) consisting of (a) palea (or absent);
(b) pistil (ovary, 2 styles and 2 feathery stigmas); (c) stamen
(1-3 anthers); (d) 2 fleshy lodicules; (e) upper lemma (may be
bibbed, entire, awned, aristulate, mucronulate, and variable in
breadth).
Pedicellate spikelet (absent in Cleistachne, Asthenochloa, most of
Sorghastrum, and some Sorghum):
Glumes (2) (a) lower (always present) and (b) upper
(sometimes absent);
lower floret (empty or barren or absent) consisting of (a)
lower lemma (maybe absent);
upper floret (male, neuter, or absent) consisting of (a) upper
lemma; (b) palea (absent in most cases); (c) 2 fleshy bodicules
and (d) stamen (1-3 anthers); awn (absent in most cases).
The spikelet configuration of the pedicellate spikelet differs when it is
bisexual, in which case it is similar to the sessile one, i.e., it will have similar organs
present and similar ones absent.
65
b
C
Tst
sg
ug
st
ul
p
a
11
f
ig
0
Htp
Is
a
Ig
Hmp
ug
ul
11
f
p
Is
Figure 4.5 Spikelet configuration in the Sorghinae (Diagrammatic). A. Sessile
spikelet. B. Pedicellate spikelet. C. Location of spikelet pair types along the raceme
axis. a, anther; b, bristle; c, column; Ca, callus; f, filament; Hmp, homogamous
spikelet pairs; Htp, heterogamous spikelet pairs; ig, lower glume; 11, lower lemma; Is,
lodicules; o, ovary; p. palea; sg, stigma; st, style; Tst, terminal spikelet triad; ug, upper
glume; ul, upper lemma.
66
4.3.9. Sessile spikelet.
As mentioned earlier, there are two kinds of spikelets in the Sorghinae, the
pedicellate and the sessile. In all the genera of the Sorghinae, the sessile spikelet is a
specialised female organ due to the loss of function in the lower floret. The lower
floret is represented by a lemma while the upper is always bisexual. The upper
lemma always bears a geniculate awn (when present) from its sinus or tip. The
pedicellate spikelets are nearly always preserved as functioning male organs. Clayton
(1969) suggested that, for sexual dimorphism to be so strongly expressed even in
cases where the spikelet and the disseminule are almost equivalent, might be due to
the fact that the callus forms an essential part of the disseminule, and that this
structure has evolved primarily by modification of the rachis segment. He further
suggested that the callus of the pedicellate spikelet presumably evolved
independently, and opened the way to a reversal of the sexual roles of the sessile and
the pedicellate spikelets such as occurs in Trachypogon Nees and Germainia Bal. &
Poitr. The sexuality of the sessile spikelet is not useful as a diagnostic character since
it is bisexual in all members of the subtribe.
4.3.10. Pedicellate spikelet.
The shape of the pedicellate spikelet and the degree of its development varies within
the Sorghinae. Pilger (1954) assumed that the development of the bisexual pedicellate
spikelet represented a primitive character in the subtribes of the Andropogoneae.
Clayton and Renvoize (1986) however noted that though the sexuality of the
pedicellate spikelet has been regarded as taxonomically important, allowance must be
made for occasional anomalous species. For instance, the pedicellate spikelet was
found to be bisexual in the species Sorghastrum fuscescens, Sorghastrum
pogonostachyum and Dichanthium annulatum var. fecundum. The pedicel and hence
pedicellate spikelet are completely suppressed in Cleistachne. The pedicellate spikelet
is represented by a barren pedicel in Sorghastrum and Asthenochloa. In
Asthenochloa, the pedicel is actually rudimentary, represented only by a minute axis
67
hidden by the callus hairs. It can be represented by a barren pedicel, empty glumes or
well developed in Sorghum. In instances where the pedicellate spikelet is developed,
it is either male or neuter. This is so in the remaining members of the subtribe which
are Capillipedium, Bothriochloa, Dichanthium, Spathia, Pseudodichanthium,
Pseudosorghum, Hemisorghum, Chrysopogon and Vetiveria.
4.3.11. Homogamous spikelet pairs.
One other prevalent feature in the Sorghinae is the occurrence of homogamous
spikelet pairs at the base of the raceme in some genera (Figure 4.5). Steam (1992)
defines homogamous as "with only one kind of flower in the cluster". In the
Sorghinae, sexuality of the homogamous spikelet pairs deviates from the norm of
bisexual sessile spikelet and male or neuter pedicellate spikelet combination by being
either male or neuter. These pairs are usually more or less sessile. In general, they
resemble the pedicellate spikelet in being male or neuter, somewhat larger than the
sessile, but always awnless. These occur in the genera Euclasta (1-3 pairs),
Pseudodichanthium (2-3 pairs), Dichanthium (1-6 pairs) and Spathia (1-2 pairs). In
some Andropogoneae genera these have become involucral and have thus been
suggested to aid as a protective measure against bush fires (Clayton, 1969). This
phenomenon of homogamous spikelet pairs at the base of the raceme prompted
Hackel (1889) to subdivide his genus Andropogon into two principal groupings
(series):
Series A. Isozygi - sessile spikelets of all the pairs alike in sex and form
throughout the spikes;
Series B. Heterozygi - lowest one or more sessile spikelets differing from all
above in sex or form.
The function of the homogamous spikelet pairs has been suggested to be that
of protecting the upper fertile spikelets against fire. The upshot of the modifications
is often to place additional protective envelopes around the flower and fruit,
variously derived from the pedicel, pedicellate spikelet, rachis internode,
homogamous spikelets and subtending leaf sheath (Clayton & Renvoize, 1986). The
authors also suggest that this may "be related to the tribe's (Andropogoneae) success
in open savannah habitats ... where bush fires are an annual event". Having the seed
up in the raceme protected by these sexually non-functional organs below, increases
the seed's chances of survival.
4.3.12. Callus.
The term callus is applied to a thickened hard swelling, rounded or bolster-shaped or
even prolonged downwards as an appendage, often sharp, rarely needle-like or
forked, situated at the base of fertile lemmas or below the glumes (B or, 1960). These
occur at the base of the sessile spikelet and occasionally at that of the pedicellate one.
The callus forms an important part of the disseminule, and has evolved primarily by
the modification of the rachis segment (Clayton, 1969). The needle-like callus is
found in most species of the genus Chrysopogon, where at the base of the glumes it
will be found extending some distance down the rachis internode. A number of
species of this genus have a very sharp callus which is always furnished with a beard
of stiff hairs pointing away from the tip of the callus. This means that if the callus
penetrates the earth, clothing or flesh, the withdrawal of the fruit is made much more
difficult. These are a menace to grazing animals when they gradually work their way
into the flesh and cause abscesses. The callus in the other genera of the subtribe are
mostly blunt and obtuse.
4.3.13. Glumes.
Grasses have their flowers, which consist merely of reproductive organs surrounded,
subtended or more or less enclosed in stiff bracts commonly known as glumes.
Generally speaking, if the protection of the grain is taken over by the lemma and
palea, the glumes tend to be membranous or hyaline. If the glumes themselves are
We
the protecting scales, they tend to be tough, as is the case in the Sorghinae. The
spikelets are generally dorsally or laterally compressed hence the lower glume is
mostly two-keeled while the upper is navicular.
There are various means by which the grain gets released from the dispersed
unit. In many of the smaller grasses the grains become once detached from the
glumes in falling. In some other instances, the grain remains enclosed in the glume
surrounding it, and shut in by the palea. Specialisation often occurs in the spikelet to
facilitate grain release. There are some grasses which have toughened lemmas and
paleas which clasp the grain tightly, while the glumes are reduced in size compared
to the rest of the spikelet and become lax with maturity. In case of the Sorghinae
grasses, the glumes are the same size as the spikelets and toughened, while paleas are
reduced or absent, and lemmas hyaline and membranous. The texture of the glumes
of the sessile spikelet, in most cases, is different from those of the pedicellate
spikelet. It is those of the sessile spikelet, hence those enclosing the grain, that are
toughened while those of the pedicellate spikelet are papery, an adaptation which
might be associated with dispersal for lightening the load. It would seem probable
that the grain is dispersed whilst still between the two glumes and gets dislodged as
the glumes wither away due to the elements. In the cultivated species of Sorghum,
e.g. S. bicolor, the glumes are gaping at maturity, and hence a third to a half of the
grain is exposed. This makes it easy for the grain to be released from the husk. It is
quite probable that this is a result of continuous selection by breeders to improve
grain release. Most of the Sorghinae grasses have the lower glumes of the sessile
spikelets covered with hair, but only in a comparatively small number is the hair
sufficiently long and silky to be of any use in transporting the grain.
Pseudodichanthium serrafalcoides (Cooke et Stapf) Bor, is the only species
with a notably winged lower glume of the sessile spikelet. Ridley (1930) notes that
glumes are boat-like and light, project beyond the sides of the grain and hence act as
wings, so that the grains in them are readily blown away to a considerable distance.
Frequently in other grasses of the Sorghinae, the pedicellate spikelet is barren (and
70
consequently light) and just represented by empty glumes and remains attached to
the dispersal unit. These glumes probably also serve as wings to aid in dispersal. The
glumes do not increase in size nor alter in shape during the fruiting stage, but remain
the same as during the flowering stage.
Glumes of certain species in the genera Bothriochloa and Dichanthium have a
glandular depression, commonly referred to as a glume pit, on the back, at or above
the middle. These are found on the lower glumes of the pedicellate and sessile
spikelet. No particular function has yet been associated with the glume pit, but
Heslop-Harrison (1961) suggested that it might have a role to play in cleistogamy
(see Chapter 5). The character is obviously useful at generic level, but its consistency
is rather wanting at species level. The presence of the glume pit is so variable within
the same inflorescence and between specimens that it can not be used reliably at
species level.
One other noteworthy development on the glumes of the genera Vetiveria and
Chrysopogon, is the presence of awns at the tips of these. The awns are well
developed in Chrysopogon species, and occur in varying frequencies on the glumes
of both spikelets. They are also pubescent to varying degrees, and this state can be
used with some consistency to distinguish different species of the genus. In
Vetiveria, the awns tend to be short and mucronate. The glume awns do not occur in
the other genera of the subtribe.
4.3.14. Lower lemma.
There are two florets, the upper and the lower, in the spikelets of the Sorghinae. The
lower floret is always represented by a lemma, the lower lemma, in the sessile
spikelet. This lemma is always hyaline and ciliate along the margins. The lower
lemma can be present or absent in the pedicellate spikelet. If present, it is usually of
the same texture and length as that of the sessile spikelet.
71
4.3.15. Upper lemma.
As mentioned earlier, both spikelets, i.e., the sessile and pedicellate, have two florets,
the upper and the lower. There is an upper lemma in both spikelets. The upper lemma
in the pedicellate spikelet, if present, is hyaline, about the same size as the upper
glume and awnless (awned in Dichanthium annulatum var. fecundum, Sorghastrum
juscescens and S. pogonostachyum). The upper lemma of the sessile spikelet is
hyaline in texture and is characteristically awned for members of the subtribe. The
apex of this lemma can be either entire or bibbed, and ciliate or glabrous. The
subtribe divides roughly into three groups on the basis of this character. Firstly in the
Sorghum, Asthenochloa, Cleistachne, and Pseudosorghum group, the lemma is
ciliate and bibbed. The second group has a glabrous and entire lemma. Here is found
the genera Bothriochloa, Capillipedium, Dichanthium, and Spathia. In addition to
being entire, the lemma is so reduced in width that it is merely a hyaline base of the
awn (stipiform). The last group consists of Vetiveria, Chrysopogon, Hemisorghum,
and Sorghastrum, in which no state of the character is dominant. Here, however, if
the lemma in entire, it is still wider than the base of the awn. The nature of the
lemma tip, usually given great prominence in artificial keys, seems not to follow any
discernible evolutionary trend in the subtribe.
4.3.16. Palea.
The grass floret is enclosed on the outside by a lemma, and a palea on the inside.
Heywood (1993) likens the spikelet to a branch system in which the glumes and
lemmas represent modified leaf-sheaths and the palea a prophyll (a special bract
replacing the first leaf on the inner side of an axillary branch). The palea is absent in
a majority of the species sampled. The absence of the palea is often suggested to be
an indication of evolutionary advancement in the Gramineae.
72
4.3.17. Awn.
Awns are frequently developed on glumes and lemmas of grasses. In the Sorghinae,
these are developed from the tips or sinuses of the lemma of the fertile (bisexual)
upper floret. Awns have been suggested to act as guiding propellers to the
windblown disseminules. These are variously hairy and in most instances coiled
(twisted) and will easily adhere to any moving (animate) dispersal agent. Awns also
occur at the apex of glumes in the genera Chrysopogon and Vetiveria.
The twisted column of the awn reacts to hygroscopic changes by coiling and
uncoiling, a movement which maybe of some value in freeing the grain from the
tangle of the inflorescence. This function of the awn in conjunction with the
retrorsely bearded callus, has been suggested to facilitate in propelling the grain
along the ground. The awn hairs have also been suggested to have some relation to
the retention of dew, improving efficiency of the hygroscopic mechanism (Clayton,
1969).
4.3.18. Lodicules.
The lodicules are taken to represent vestigial perianth members. They are two
diminutive bodies lying between the lemma and the ovary base in the grass floret,
which by expanding rapidly at the time of anthesis, lever away the rigid lemma
allowing anthers and stigmas to emerge (Heslop-Harrison & Heslop-Harrison, 1996).
The essential feature of functional lodicules is the presence of a cushion of tissue at
the base made of cells which by undergoing enlargement at anthesis, cause the basal
part to swell, so generating a considerable force necessary to prise apart the stiff
outer perianth components. Heslop-Harrison & Heslop-Harrison (1996)
demonstrated that this is driven by the influx of water associated with the
accumulation of potassium ions in the cells of the cushion. This function of the
lodicules is therefore associated with chasmogamous flowers. A balance between
chasmogamy and cleistogamy in grasses has been suggested to be affected by
73
environmental conditions, including day-length (Heslop-Harrison, 1959). The two
phenomena are therefore equally common in most grasses.
Lodicules of most members of the Sorghinae are fleshy, cuneate in shape,
visibly vascularised, and glabrous. An exception occurs in the genera Cleistachne
and Sorghum, which have lodicules ciliate on the apex or apical corners. The
lodicules are absent in the genus Asthenochloa. One would therefore assume that this
genus is necessarily cleistogamous. Also taxa whose florets do not have lodicules are
often considered evolutionarily advanced (Hubbard, 1984).
4.3.19. The androecium.
The spikelets of the grasses of the Sorghinae have 2 or 3 anthers. An unusual
situation occurs in the species Bothriochloa decipiens, which possesses a single
anther. This character, unique in the subtribe, has been suggested to be an adaptation
associated with cleistogamy (Heslop-Harrison, 1961). One observation I made was
that whenever the pedicellate spikelet was male, the anthers were mostly larger than
those of the bisexual sessile spikelet. It was found to be of the same size as those of
the sessile spikelet in a few cases but very rarely smaller in any of the material
studied.
4.3.20. The gynoecium.
The pistil of the grasses of the Sorghinae, like that of most grasses, consists of a
unilocular ovary with a single ovule, and two styles with feathery stigmas. Much
variation is exhibited in the shape of the ovary. It may be ovate, ellipsoid, obovate,
oblong, or more or less flattened. According to Arber (1934), the ovule of the
Gramineae has two integuments, each usually composed of two cell layers. The outer
integument is gradually absorbed after fertilisation, but the inner persists to form the
seed coat. The inner portion of the pericarp (ovary wall) also breaks down and is
absorbed, and the remaining pericarp tissue becomes united with the seed coat. The
74
resulting dry, indehiscent, one-seeded fruit, in which the pericarp and seed are fused,
is called a caryopsis. Some specimens in the sample were found at this fruiting stage.
One observation made was that, for the cultivated sorghums, the caryopsis was found
to be partially exposed between the gaping glumes. This is to facilitate in the release
of the grain for processing. In the others members of the Sorghinae, the grain was
found to be always between the tightly closed glumes, and one always had to prise
very hard to release the grain.
75
/
ID:
11
E12
Figure 4.6 Chrysopogon gryllus subsp. gryllus. 1. Spikelet triad (x2). 2. Ligule (x5).
Sessile spikelet: 3. Lower glume (x 10). 4. Lower lemma (x12). 5. Palea (x 11). 6.
Upper glume (x8). 7, Upper lemma (x13). 8. Ovary (x12). 9. Lodicules (xlO).
Pedicellate spikelet: 10. Lower glume (x7). 11. Upper glume (x7). 12. Pedicel (x7).
76
//\ \
10
1
Ii
\\ki .'/
1
C
\
2
•'\\• \
:1
'
3'-c
Figure 4.7 Bothriochloa kunizeana. Sessile spikelet: 1. Lower glume (x20), Ccallus. Bothriochloa iiisculpta. Pedicellate spikelet: 2. Upper glume (x20). Sessile
spikelet: 3. Lower glume (x20), c-callus. Bothriochloa biloba. Sessile spikelet: 4.
Upper lemma (x45), NB*fi1iform lobes. Cleislachne sorglwides. 5. Awn.
77
4
Figure 4.8 Euclasta oligantha. Pedicellate spikelet: 1. Upper glume (x20). 2. Lower
glume (x20). Sessile spikelet: 3. Lower glume (x40). 4. Upper glume (00). Euclasta
clarkei. 5. Spikelet pair (x20). Pedicellate spikelet: 6. Lower glume (x45). 7. Upper
glume (x45). Sessile spikelet : 8. Lower glume (x45). 9. Upper glume (x45).
Il
( t
il
I
f2d
7
4'Li
5'j/
6
BC
Figure 4.9 Vetiveria nigritana. 1. Spikelet pair (x15). Sessile spikelet: 2. Lower
glume (x15). 3. Upper glume (x20). 4. Upper lemma (x20). 5. Ovary (x20).
Pedicellate spikelet: 6. Lower glume (x20). 7. Upper glume (x20). Vetiveria
zizanioides. Sessile spikelet: 8. Upper lemma (x45).
79
4
A
4
4
II
2
1:
5
(
7
Figure 4.10 Heinisorghum venusturn. Sessile spikelet: 1 Bibbed upper lemma
(05). 2. Lower glume (05). 3. Upper glume (05). 4. Lower lemma (05). 5.
Caryopsis (05). Pedicellate spikelet: 6. Lower glume (05). 7. Upper glume (05). 8.
Pdice1 (05).
ME
U
4
5
Figure 4.11 Dichanthium annulatuni var. papiliosum. 1. Spikelet pair (00). Sessile
spikelet: 2. Lower glume (00). Dichanthium annulatum var. annulatum. Sessile
spikelet: 3. Lower glume (00). Dichanthium aristaturn. Sessile spikelet: 4. Lower
glume (00). 5. Upper glume (00). Dichanthium sericeum. Sessile spikelet: 6.
Lower glume (x13).
Chapter 5. Micromorphological studies by Scanning
Electron Microscope (SEM).
5.1. Introduction.
Features observable only with the electron microscope provide useful characters. The
SEM provides an image of unequalled depth of field which is ideal for comparative
studies. Even though the technique usually uncovers relatively few new characters, it
has however enabled the rapid, recordable and comparative study of a great many
micromorphological features, that have become realistic and practicable as standard
taxonomic characters (Stace, 1989). This technique was used to glean for characters
of the glume.
5.2. Materials and Methods.
Whole glumes from herbarium specimens were mounted on aluminium stubs using a
double sided adhesive tape. These were positioned and duly labelled for
identification purposes. For those specimens sticking out from the surface, a silver
conductive paint (Electrodag 915, Acheson colloids, Plymouth) was used to create an
electrical connection between the specimen and the stub surface. The specimens were
then sputter-coated with gold and viewed using a Stereoscan S250 MK1 Scanning
Electron Microscope (Leica UK Ltd, Cambridge, UK) at an accelerating voltage of
5-25kv. Electron micrographs were recorded on Ilford FP4 and Kodak TMAX 100
120 film.
82
5.3. Glume pit(s) in the grass genera Bothrioch!oa Kuntze
and Dichanthium Willemet.
5.3.1. Introduction.
Some species of the genera Bothriochloa and Dichanthium show a tendency for
pitting. Their species, in common with others of the subtribe, posses digitate or
subdigitate inflorescence with spikelets borne in pairs, one pedicellate and the other
sessile. The inflorescence consists of racemes with the spikelet pairs arranged
distichously along the rachis, terminating in a triplet of two pedicellate spikelets and
one sessile spikelet. Both the pedicellate and the sessile spikelets have two glumes,
the lower and the upper. The pedicellate spikelet is either male or neuter (in which
case it is represented by two empty glumes). The sessile spikelet has two florets: the
lower is represented by an empty lemma; the upper is bisexual and fertile, thus being
the only functional floret in the nodal group. The anther number ranges between 1
and 3.
5.3.2. Glume pits.
The lower glumes of some genera of Andropogoneae are crustaceous and are
sculptured, grooved, wrinkled, pitted or provided with appendages for which it is hard
to find a purpose. Bothriochloa and Dichanthium have some members characterised
by the presence of a glandular depression or pit in the middle or upper portion of the
lower glume of either the sessile spikelet and or the pedicellate one (Table 5.2). The
significance of the glume pit has always been something of an enigma to people who
have worked with these genera. It seems hardly reasonable for a feature so
conspicuous and unusual to be without a function of some sort. Two other genera of
the tribe Andropogoneae said to have glume pits are Kerriochloa C.E. Hubbard
(subtribe Ischaeminae) and Anadeiphia Hack. (subtribe Anthistriinae) (Clayton &
Renvoize, 1986). Five other genera in the Andropogoneae (Table 5.1) have lower
glumes of the sessile spikelet which can be described as lacunose with deep
depressions (Figure 5.1, H grànularis and 0. exaltatus). These are however not in
any way similar to the glume pits of Bothriochloa and Dichanthium. These are just
shallow superficial sculptures which do not have any corresponding markings on the
inner side of the glume.
Table 51. Andropogoneae genera with pitted and lacunose lower
glumes.
Genus
Subtribe
Description of lower glume of sessile spikelet
Bothriochloa
Kuntze
Sorghinae
Two-keeled; convex to flattened on the back; with or without a
conspicuous pit.
Dichanthium
Willemet
Sorghinae
Two keeled; convex to flattened on the back; with or without a
conspicuous pit.
Glyphochloa
D.
W.
Clayton*
Rottboelliinae
Two-keeled; convex on the back; lacunose with deep depressions, or
rugose, or tuberculate or prickly.
Hackelochloa
Kuntze*
Rottboelliinae
Not two-keeled; convex on the back; lacunose with deep depressions.
Heteropholis
E
C.
Hubbard*
Rottboelliinae
Not two-keeled; convex on the back; lacunose with deep depressions
or rugose (transversely).
Kerriochloa
E.
C
Hubbard*
Ischaeminae
Not two-keeled; convex on the back; with a conspicuous pit; relatively
smooth.
Ophiuros
Gaertn.f.
Rottboelliinae
Not two-keeled; convex to flattened on the back; lacunose with deep
depressions, or rugose (with a basal transverse groove).
Anadeiphia
Hack.*
Anthistriinae
Convex to flattened on the back; with a conspicuous pit (at the base, at
least in A. afzeliana and A. hamata); relatively smooth.
Ratzeburgia
Kunth.*
Rottboelliinae
Two-keeled; lacunose with deep depressions (5-6 rows of pits).
* - From Clayton & Renvoize (1986).
84
Table 5.2. Bothriochloa and Dichanthium species with glume pits.
Species
Pit
on
spikelet
single)
sessile Pit
on
(always pedicellate
spikelet (1 to
4 pits)
*Bothriochloa alta (Hitchcock) Henrard
p
a
Bothriochloa macera S. T. Blake
p
a
*Bothriochloa barbinodis (Lag.) Herter
p
a
Bothriochloa bladhii (Retz.) S.T. Blake
p
a
Bothriochloa decipiens (Hack.) C. E. Hubbard
p
a
*Bothriochloa edwardsiana (Gould) L. Parodi
p
a
*Bothriochloa ewartiana (Domin) C. E. Hubbard
p
a
Bothriochloa bladhii (Roxb.) A. Camus subsp. glabra
Simon
p
a
*Bothriochloa hybrida (Gould) Gould
p
a
Bothriochloa insculpta (Hochst.) A. Camus
p
p(l to 4)
Bothriochloa kuntzeana (Hack.) Henrard
p
p
*Bothriochloa palmer! (Nash) Pilger
p
a
Bothriochloapertusa (Willd.) A. Camus
p
a
*Bothriochloa springfield!! (Gould) L. Parodi
p
a
*Bothriochloa wrightii (Hack.) Henrard
p
a
*Dichanthium armatum Blatter & McCann
p
a
Dichanthiumfoveolatum (Delile) Roberty
p
a
Dichanthium micranthum T. A. Cope
p
a
*Dichanthium panchganiense B latter & McCann
p
p
* - The asterisk indicate that these species were not observed in this study but the
information used was obtained from the literature.
p - present; a - absent
Figure 5.1. Glume surface characteristics.
Hackelochloa porfera, outer surface of glume. This shows the outer
surface with lacunose deep depressions.
Hackelochloa porfera, inner surface of glume, smooth.
Ophiuros exaltatus, outer surface of glume with shallow depressions.
Ophiuros exaltatus, inner surface of glume, smooth.
5.3.3. Role of the glume pit in cleistogamy.
Heslop-Harrison (196 1) carried out a study to find out the function of the glume pit in
Bothriochloa decipiens (Hack.) C. E. Hubbard. B. decipiens has a single anther,
which is an adaptation associated with cleistogamy. His main conclusion was that the
glume pit played a role in the mechanism of cleistogamy (fertilisation in an unopened
flower) in this species. Its function is that of an obturator, preventing the emergence
of the single anther from the floret and ensuring its dehiscence in contact with the
stigmas in all florets where the glumes do not open normally at anthesis. In a young
floret, before pollination, the boss of the glume pit (on the inner side of the glume),
depresses the lemma between the two branches of the stigma, lying just above the
level of the tip of the immature anther. Should the glumes fail to open when anthesis
would normally occur, emergence of the stigmas is prevented by their engagement
upon the depressed lemma below the boss. As the filament of the stamen extends, the
anther is then driven hard up against the boss, and internal dehiscence takes place at
the tip in contact with the stigmas. This correlates quite well with the fact that in all
pitted species, there is always a pit on the lower glume of the sessile fertile spikelet,
hence its association with anthesis and pollination. This argument, however, loses
credibility when considering the fact that glume pits do also occur on the lower glume
of the pedicellate male spikelet.
This function of the glume pit depends on the opening of the glumes of the
spikelet, which in turn is influenced by various environmental factors. Therefore
should the glumes open at the maturation of the floret, normal chasmogamous
anthesis is possible, since there is no barrier to the exsertion of either stigma or
anther.
Also of significance is the depth of the glume pit, hence the extension of the
complementary boss on the inner side of the glume. The glume pit varies in size
shape and depth in the species observed. It can be round and small, as in Bothriochloa
pertusa, big and deep, as in B. macera, or slit-like and shallow, as in B. kuntzeana.
One would therefore assume that the involvement of the glume pit in cleistogamy in
the various situations would differ. Heslop-Harrison (1961) concluded that facultative
cleistogamy was a common feature in tropical and subtropical Andropogoneae. There
are some plants that are irregularly pitted on different plants or even different parts of
the same panicle. Detailed studies are needed to determine whether there is any
correlation between pit development, breeding behaviour and environmental control
factors.
5.3.4. Use of the glume pit as a taxonomic character.
Gould (195 9) used the glume pit as a marker character for tracing relationships of Old
World and New World taxa of the genus Bothriochloa, and found it to be a reliable
character when properly used. De Wet and Higgins (1963) however found it highly
unreliable.. Chippindall (1955) found it difficult to separate B. inscuipta from B.
pertusa on the basis of the presence of the glume pit. Both species always have a
single pit on the lower glume of the sessile spikelet. B. inscuipta has, in addition 1 to
4 pits on the pedicellate spikelet (Figure 5.2B). Also the lower glume of the sessile
spikelet in B. inscuipta is glabrous on the back while it is hairy in B. pertusa. There
are however those specimens that have a single pit on both lower glumes (sessile and
pedicellate (making them likely to be B. inscuipta)), and have the lower glumes of the
sessile spikelet hairy (making them B. pertusa). This illustrates the inadequacy of
overweighting a single character. I found the glume pit, as a taxonomic character,
quite limited, except when used at generic level. The occurrence of glume pits is
variable in a single inflorescence and between specimens of the same species.
Figure 5.2. Glume surface characteristics.
Bothriochloa inscuipta, lower glume of sessile spikelet with a single pit
above the centre of the glume.
Bothriochloa inscuipta, lower glume of pedicellate spikelet with four slitlike and shallow pits along the middle of the glume.
Dichanthium micranthum, lower glume of sessile spikelet with a shallow
pit at the centre of the glume.
Dichanthiumfoveolatum, lower glume of sessile spikelet with a deep pit at
the centre of the glume.
WE
Figure 5.3. Glume surface characteristics.
Bothriochloa pertusa, lower glume of sessile spikelet with a deep pit just
above the centre of the glume. Note the long hairs from the glume base and
callus.
Bothriochloapertusa, close up of glume pit.
Bothriochloa edwardsiana, lower glume of sessile spikelet, inner boss.
Bothriochloa bladhii, lower glume of sessile spikelet with a deep pit just
above the centre of the glume.
92
C)
0
bOPfl
?. 0KV
OF
025
S
Chapter 6. Anatomical investigation of the leaf blade.
6.1. Introduction.
Leaf anatomical data has been of considerable use in grass taxonomy. Structurally
the grass leaf is a complex organ, exhibiting a wide range of anatomical features and
providing valuable additional taxonomic information (Ellis, 1986). Leaf characters
rank second to those of flowers and fruits in their use in taxonomic studies. They
however are strictly comparable over a wider range (all vascular plants) than are
those of floral organs. A diversity of cells and tissues, their arrangement and
positions in the leaf blade, provide a pool of characters and character states of high
potential. These cell patterns are characteristic of a species and often diagnostic of
genera and tribes (Stebbins, 1956). These basic differences in leaf anatomy have
facilitated a compilation of anatomical diagnosis (Renvoize, 1981; Watson, et al.,
1985).
It was the pioneering work of Duval-Jouve (1875) that set the pace for the use
of leaf anatomical characters in grass systematics. Later emphasis was mainly geared
towards the structural and functional relationships of the grass leaf, particularly the
correlation of C4 photosynthetic pathway with Kranz anatomy, and the C3
photosynthetic pathway with non-Kranz anatomy. This resulted in several new
systems of higher classification of the Gramineae. Leaf anatomical data from the
study has revealed that the grasses of the subtribe Sorghinae are of C4 physiology.
Taxonomically useful leaf anatomical attributes of the Gramineae have been
comprehensively defined and illustrated by several authors (Ellis, 1976, 1979;
Clifford and Watson, 1977; Watson and Dallwitz, 1980) basing most of their
interpretations on the work of Metcalfe (1960). Ellis (1976, 1979) has made a useful
contribution by attempting to standardise and stabilise grass anatomical terminology
so far employed by all previous authors (and see Rahman, 1988).
94
6.2. Materials and Methods.
Leaf blades from herbarium and living specimens (grown in greenhouses), were used
for the anatomical investigations. Healthy looking leaves from recent collections were
used. An attempt was made to sample at least one species from each genus. This was
however not possible because of the unavailability of suitable representative material.
The sample therefore consisted of the following species: Chrysopogon fulvus, C.
gryllus subsp. echinulatus, C. orientalis, Capillipedium parvy?orum, Dichanthium
micranthum, Sorghastrum friesii, Sorghum halepense, S. versicolor, and Vetiveria
nemoralis.
6.2.1. Preparation.
Leaves were softened by gently boiling them in water in a water bath until they
reverted to their natural shape. This was checked every 5 minutes, and in most
instances took 20 to 30 minutes The boiled leaves were then put in Formalin Acetic
Alcohol (F.A.A.) solution in labelled 20ml vials for at least two days.
6.2.2. Rinsing.
Two to three segments of about 5mm long were cut from the middle part of the
lamina using a pair of scissors. The remainder of the leaf was placed back in the
F.A.A. vials to store indefinitely. The segments were then washed in cold water in
separate labelled vials, with occasional shaking by hand. Several changes of water
were made to remove the F.A.A.
6.2.3. Dehydration.
The dehydration procedure follows that of Johansen (1940). The rinsed segments
were put through a graded series of alcohols as follows:
95
10% alcohol
1 1/2 hours
20% alcohol
1 1/2 hours
30% alcohol
1 1/2 hours
50% alcohol
1 1/2 hours
70% alcohol
2 hours
Johansen solution 1
2 hours
Johansen solution 2
2 hours
Johansen solution 3
2 hours
Johansen solution 4
2 hours
Johansen solution 5
2 hours
Tertiarybutylalcohol (TBA)
2 hours: followed by 2 changes - one overnight.
The time limits are only indications. The material was however not allowed to
stay for too long in any of the alcohol solutions weaker than 50%, otherwise it could
be left overnight at any of the other stages, and the series continued the following
morning.
6.2.4. Infiltration.
The infiltration method used follows that of Purvis, Collier and Wallis (1964). After
the last dehydration stage, the segments were put in a mixture of equal parts liquid
paraffin and tertiarybutylalcohol (TBA) for 6 hours. A specimen bottle was then filled
two thirds full of melted wax and allowed to solidify but not completely harden. The
segments were then covered with the mixture of TBA and liquid paraffin. The bottle
was then placed on the embedding bath and the segments left to infiltrate slowly as
they sank down the wax. They were left for about 1 hour after they had sunk to the
bottom of the bottle. Two changes of pure wax were then made before embedding.
6.2.5. Embedding.
Brass embedding frames and plates were used. The leaf segments were poured into
the embedding frames, arranged properly with a hot needle, and the whole mass
cooled quickly in an ice bath. It is necessary to cool the wax quickly because
otherwise it crystallises: the blocks become full of fluffy white spots and patches, and
this makes the wax difficult to microtome. Finished wax blocks were then labelled,
and ready for sectioning.
6.2.6. Sectioning.
A rotary microtome (3880/A296) was used for cutting the sections. The blade was
inclined at 16°, and sections of 15.tm cut.
6.2.7. Mounting Ribbons.
The ribbons were placed on a drop of water on a slide smeared with glycerine
albumen. More drops of water were added to have the ribbons floating freely. Excess
water was drained off using absorbent tissue paper, leaving the ribbons flat on the
slide, and these were positioned using a needle. The slide was then put on a slide
warmer (20°C) for at least 2 days. Duplicate slides of each species were made.
6.2.8. De-waxing.
Slides were removed from the slide warmer, and taken through a de-waxing series as
follows
Xylene
30 minutes
Xylene
15 minutes
Xylene + absolute alcohol
10 minutes
97
Absolute alcohol
10 minutes
Absolute alcohol
10 minutes
The slides were then transferred to safranin and left for 24 hours
6.2.9. Staining.
The steps in staining follow on from the de-waxing series:
Safranin
24 hours
Absolute alcohol
5 minutes
Absolute alcohol
10 minutes
Absolute alcohol
10 minutes
Fast green
5-20 seconds
Absolute alcohol
10 seconds
Absolute alcohol
10 seconds
Absolute alcohol
5 seconds
6.2.10. Mounting.
Permanent slides were then prepared by mounting the sections in Euparal and
covering them with Chance cover slips. The slides were then put on a slide warmer at
45°C for at least 2 days.
6.2.11. Cleaning slides.
After the Euparal had dried completely, any excess that had oozed out from
underneath the cover slips, was scraped away with a razor blade. The slide was then
wiped with a tissue paper soaked in absolute alcohol.
6.3. Evaluation of characters observed.
6.3.1. Outline of leaf lamina in transverse section.
The outline of the cross section of the leaves of most of the species observed for this
exercise is V-shaped. The two lateral halves of the lamina fold adaxially towards each
other with the midrib or keel forming the bottom of the V(Figures 6.113, 6.2, 6.3 &
6.4A, B, C). There is however one species, Sorghum halepense (Figure 6.1A), which
has an inrolled leaf and does not have a conspicuous midrib. The two leaf surfaces,
abaxial and adaxial, can either have ribs and furrows or none at all. The ribs
correspond to the leaf veins while the furrows are the spaces between the veins. The
furrows are sometimes referred to as the intercostal zones. In instances where these
occur, they are mostly found on the adaxial surface. There is somewhat a correlation
between the presence of ribs and furrows with that of bulliform cells. The distribution
of the bulliform cells is related to the manner in which the leaf is folded. The capacity
for leaves with marked adaxial ribs to roll is a character that seems to have arisen as a
response to ecological conditions, and for this reason leaves of this type are to be
found in grasses between which there are no close taxonomic affinities (Metcalfe,
1960).
6.3.2. Epidermis.
The cells of the two epidermal layers differ. The adaxial epidermis is
characteristically composed of bulliform cells and short cells, which are shallow
above the strands of sub-epidermal sclerenchyma (where present). The abaxial
epidermis consists mostly of short cells (tabular in shape), most of which have
developed protrusions above the surface to form papillae (Figure 6.513). All of the
species observed had papillae. These were mostly found scattered randomly on the
abaxial surface. They do occur on the adaxial surface but with less frequency as on
the abaxial one. Macrohairs were observed on Chrysopogon gryllus subsp.
NZ
echinulatus (Figure 6.41)). Microhairs proved to be rather difficult to observe in these
particular sections.
6.3.3. Bulliform cells.
The distribution and arrangement of bulliform cells in the leaf vary in the species
observed. In ribbed species, e.g., Sorghum versicolor, these cells occupy the furrows
(Figures 6.513 & 6.6) or the intercostal zones. Therefore the bulliform cells are found
on the adaxial surface. Two bulliform cells' arrangements were recognised:
Bulliform cells found only on the adaxial and or abaxial intercostal zones
(furrows) and stopping at the ribs formed by 1st order vascular bundles, e.g.,
Sorghum versicolor (Figure 6.5B & 6.6). These therefore occur over the 2nd and 3rd
order vascular bundles.
Bulliform cells occupying the whole adaxial and or abaxial surface, even over the
1St order vascular bundles over which they are shallow but nonetheless present
(Figures 6.2A & 6.3). On Vetiveria zizanioides, the bulliform cells are found only at
the bottom of the V, in the keel region on the adaxial surface.
The sizes and shapes of the bulliform cells also differ between the species.
They can be large and inflated with short tangential walls than the inner lateral ones,
e.g., Chrysopogon orientalis (Figure 6.5A). They can also be large, but have wavy
walls, e.g. Sorghum versicolor (Figures 6.513 & 6.6). This happens to be the case in
instances where the cells occur in furrows, and when the middle cell of the group is
bigger than the remainder.
The significant function of bulliform cells has been argued over for years
now. According to one view, the bulliform cells are concerned with the unrolling of
developing leaves. Their sudden and rapid expansion during a certain stage of leaf
development is assumed to bring about the unfolding of the blade (Esau, 1965).
Another concept is that, by changes in turgor, these cells play a role in the
hygroscopic opening and closing movements of mature leaves. The presence of these
100
cells in grass leaves would therefore not necessarily indicate a close taxonomic
affinity of the grasses concerned.
6.3.4. Vascular bundles.
There are three orders of vascular bundles. The 1st order vascular bundles form the
ribs on the adaxial surface of the ribbed species. These alternate with groups of 2nd
and 3rd order vascular bundles which are in the intercostal zones. The number and
distribution of these were found to be inconsistent.
For species with V-shaped leaves, many vascular bundles are present in the
keel and are all adaxially arranged. In such cases the median bundle is
indistinguishable from the other 1st order vascular bundles in the rest of the leaf.
Also three 1st order bundles and 3 to 4 other smaller bundles comprise the keel. The
only exception was Sorghum halepense (Figure 6.1A), which has an inrolled lamina,
there is no midrib or keel. The same level of positioning for vascular bundles of all
orders was observed. All bundles are situated towards the abaxial surface of the
blade. The 2nd and 3rd order vascular bundles have a complete bundle sheath, which
is entirely parenchymatic and has chloroplasts (Figure 6.5A), while the sheath is
incomplete for the 1st order vascular bundles. Scierenchyma girders are abaxial.
They occur on both the abaxial and adaxial sides of the vascular bundles less
frequently (Figure 6.313).
6.4. Discussion.
The degree of infolding or inrolling of the leaf varies with environmental conditions
and thus is not of much value diagnostically. However, according to Ellis (1976),
leaves of grasses that are fully expanded under optimum conditions invariably
exhibit a characteristic type of involution in response to environmental stress. He
suggests that it is therefore the nature, and not the degree of the infolding or inrolling
that is of importantance taxonomically and diagnostically. Most of the species
101
observed -in this sample have a V-shaped leaf, which is a resultant appearance of
infolding leaves. This is however not evident enough to support the use of the leaf
outline in this sample as being of specific or generic value. The leaf outline has been
observed to vary within species and genera, and therefore a sampling procedure
should adequately reveal extremes of this variation.
The vascular bundles are usually arranged in a single row embedded at various
positions in the mesophyll. The arrangement found in Sorghinae grasses, is for all
the bundles to be positioned towards the abaxial surface of the blade. Ellis (1976)
suggests that the positioning of the vascular bundles in the blade is a useful
diagnostic character above genus level. The total number of vascular bundles in a
section through a leaf blade is inconsistent. The number varies with the width of the
leaf and hence can not be relied upon. The distribution patterns of bundles of
different orders, and their variation, have been found to be important diagnostic
characters. A formula can be used to describe the pattern of various orders but this
will only be useful if a regular pattern of these is present. No pattern was observed in
the present investigation. Therefore, no discernible conclusion could be reached for
vascular bundle distribution and variation.
All the species observed have a single bundle sheath. Single sheaths are
characteristic of panicoid grasses. Structural variations of bundle sheaths are
important for diagnostic purposes. The typical single or outer sheath cells are either
translucent, without chloroplasts, or they may contain green pigment in plastids
similar to, or differing from. chloroplasts of chiorenchyma of the mesophyll (Ellis,
1976). However, there is a need to standardise leaf sampling for this kind of
character, because older leaves tend to have few or no plastids in bundle sheaths
(Percival, 1929). Bundle sheaths of second order vascular bundles of Chrysopogon
orientalis (Figure 6.5A) were found to contain chloroplasts.
For plants with C4 physiology (Kranz), two types can be distinguished (Brown,
1975,1977):
102
PS (Parenchyma Sheath) type. This type is characterised by chlorenchyma which
is strongly radiate surrounding bundles of approximately equal size. Both bundle
sheaths (inner and outer) are present, but only the outer (parenchyma) sheath forms
starch;
MS (Mestome Sheath) type. This type is characterised by irregular or weakly
radiate chlorenchyma surrounding bundles of differing size. Only one bundle sheath
is present, this forming starch and apparently derived from the mestome sheath
(Clayton & Renvoize, 1986).
The Sorghinae species are of the MS type. C4 physiological set-up is an
advantage in hot climates with some additional advantage if the climate is also dry,
but is ineffective or disadvantageous in cool climates and shady habitats (Clayton &
Renvoize, 1986). Most grass genera are consistently either C3 or C4, with only four
known exceptional genera: Alloteropsis, Eragrostis, Neurachne and Panicum
(Hattersley & Watson, 1992). Furthermore, most C4 genera are consistently of one
biochemical type. There are three known biochemical variants of the C4 acid cycle in
grasses. These are the NADP-ME, NAD-ME, and PCK types. The three variants
differ in part according to the C4 acid predominantly transported from PCA (Primary
Carbon Assimilation, in leaf mesophyll) to PCR (Photosynthetic Carbon Reduction,
in leaf chlorenchyma) cells. In the NAD-ME and PCK types, aspartate is the
predominant C4 acid that diffuses from PCA to PCR tissue, while in NADP-ME
type, it is malate (Hattersley & Watson, 1992). Hattersley & Watson (1992) also
indicate that members of the Sorghinae, and Andropogoneae in general, are of the
NADP-ME biochemical type.
Scierenchyma in the leaf is often associated with vascular bundles, midrib or
keel and leaf margins. The sclerenchyma associated with the vascular bundles is in
the form of subepidermal longitudinal bands following the course of each vascular
bundle. The distribution of these can be correlated with ecological factors. Grasses
from and areas are thus characterised by well developed scierenchyma tissue while
many tropical grasses often have a high proportion of the smaller bundles not
103
accompanied by scierenchyma (Ellis, 1976). No conclusion can be reached
concerning the sclerenchyma of the species in the present sample because of
inconsistent ecological data.
Epidermal cells do not always by themselves exhibit any particularly important
diagnostic characters. Their examination in section is of assistance in the
interpretation of many epidermal structures as seen in surface view. These include
the nature of the papillae, the attachment and structure of other epidermal
appendages, the positioning of the stomata, or the cuticle form and function. These
characters are important taxonomically for comparative studies for a whole range of
species.
6.5. Conclusion.
Though it is appreciated that the transverse section of the leaf blade offers a pool of
potential characters, I had the problem of not having adequate suitable material to
exploit this aspect of taxonomic evidence. The little information that came from the
sections made basically confirms results from past investigations (Clayton &
Renvoize, 1986; Hattersley & Watson, 1992; Watson & Dallwitz, 1992), that the
grasses of the Sorghinae are of C4 physiology. Chlorenchyma cells weakly radiate
from the vascular bundles, and the cells of the bundle sheath have chloroplasts. A
comparative analysis of the leaf blade for a more representative sample of the
subtribe might yield useful taxonomic information. The work that was done on the
available sample did not indicate any striking taxonomic differences in the leaf
anatomy of the different genera. It was decided, reluctantly, not to extend the work
but to concentrate on macromorphology.
104
The following is a legend to abbreviations used in Figures 6.1 - 6.6.
Ab
Abaxial surface
Ad
Adaxial surface
Bc
Bulliform cell(s)
Bs
Bundle sheath
Cc
Chlorenchyma
Cp
Chloroplasts
Cut
Cuticle
Le
Lower epidermal cell(s)
Mac Macrohair
Mes Mesophyll
Met Metaxylem
Pap Papillae
Ph Phloem
Prot Protoxylem
Sb
Silica body
Sc
Scierenchyma
St
Stomata
Ue
Upper epidermal cell(s)
1st vb
First order vascular bundle
2nd vb
Second order vascular bundle
3rd vb
Third order vascular bundle
105
Figure 6.1. Transverse section of the leaf blade (Diagrammatic).
The sections show the leaf outline and the distribution and position(s) of the leaf
tissues.
Sorghum halepense (x60).
Vetiveria zizanioides (x60).
106
Sc.
I
Figure 6.1
107
Figure 6.2. Transverse section of the leaf blade (Diagrammatic).
The sections show the leaf outline and the distribution and position(s) of the leaf
tissues.
Chrysopogonfulvus (x60).
Chrysopogon gryllus (x60).
im
Figure 6.2
109
Figure 6.3. Transverse section of the leaf blade (Diagrammatic).
The sections show the leaf outline and the distribution and position(s) of the leaf
tissues.
Bothriochloa ischaemum (x60).
Pseudosorghumfasciculare (x60).
110
Cut
Ives.
Ind vb.
Figure 6.3
111
Figure 6.4.Transverse section of the leaf blade.
The keel:
Chrysopogon gryllus subsp. gryllus (x 10).
Bothriochloa ischaemum (x 10).
Chrysopogon orientalis (x 10).
A macrohair:
Chrysopogon gryllus subsp. echinulatus (x40).
112
-4-
2nd
v\
\iAd
4
-
Ab
$ Mes
-
.0—BC
-.
-
Ad
Mes
Ab'1i
Cut
1St
vb
&s:
sc
Sc
B
-
A
Mes
1st
vb
\
\
Ab
J
Ad
Iv
Pap
00,
i
:
C
Figure 6.4
Figure 6.5. Transverse section of the leaf blade.
Bulliform cells on adaxial epidermis: Note the different sizes and shapes of the cells.
Chrysopogon orientalis (x40). Also shown are the starch containing second order
vascular bundles.
Sorghum versicolor (x40).
Bothriochloa inscuipta (x40).
Pseudodichanthium serrafalcoides (x40): Note the silica body on top of the
vascular bundle on the adaxial surface.
114
09
-Bc
Ad ,.
2nd
Vb
cp
Ad
C-,---
(_•_'
Pap
Sb
-- — A b
vb
Figure 6.5
Figure 6.6. Sorghum versicolor(xl000). Transverse section of the
leaf blade through a First order vascular bundle showing
anatomical details.
116
3rd
Vb
4.
qJ
Ad
Pap
Ph
Chapter 7. Numeric analysis.
7.1. Introduction.
Modern classifications have their roots in ancient folk taxonomies. These were
basically utilitarian in nature and did not serve any special-purpose classifications.
They permitted communication about organisms among those individuals who
already understood the nature and recognised the features of the organisms being
discussed. Classification systems that developed were mainly building on these as
additional information became available. These classifications were artificial and
established sets of classes on the basis of weighted predetermined characters.
Examples include classifications of Cesalpino, Toumefort, Bauhin, and Linnaeus,
they differ only in the author's choice of differentiating characters. These were
subjective and gave too much weight and hence were biased towards certain
characters. It was in the writings and schemes of the de Jussieus, Adanson, de
Candolle, and Bentham and Hooker, that the principle of using a multitude of
characters for the description of classes gained ground and some credibility. A priori
interpretation of important characters, the essentialism of Linnaeus, was replaced by a
posteriori determination (Estes & Tyrl, 1986). Classifications then became a
reflection of natural phenomena. Darwin's The Origin
of
Species and a more
complete understanding of the theory of natural selection made the most striking
impact of evolutionary thought on plant taxonomy, if little practical change followed
in systematics. The works of Huxley, Anderson, Stebbins, Clausen, and Grant,
provided the intellectual and procedural bases for a new approach (Estes & Tyrl,
1986). The naturalness of taxa could therefore not only be interpreted in terms of
overall similarities (phenetic) made possible by the use of a lot of characters, but also
from their evolutionary history (phylogeny). Sporne (1948, 1949) and Stebbins (1950,
1951) and in a series of other related papers devised and advocated the use of what
Sporne called an advancement index, which was a statistic reflecting the advancement
and primitiveness of angiosperm families. Families could, therefore, on the basis of
118
certain characters, be ordered to reflect an evolutionary hierarchy. Views, however,
differed on what was to be considered primitive and advanced. Mistakes were
undoubtedly made.
The application of experimental techniques to generate and evaluate data
resulted in a significant increase in the information upon which classifications could
be based. Interpretation of variational patterns therefore became more difficult.
Fortunately, there was a corresponding development of techniques for numerical
analysis that allowed for simultaneous examination of large numbers of characters.
Computers provided an entirely new means of information storage and retrieval and
turned the interest of both mathematicians and philosophers towards the problems of
classification. The most influential book was undoubtedly that by Sokal & Sneath
(1963, and also Sneath & Sokal, 1973), which established numerical taxonomy as an
important branch of classification theory and practice. Numerical analysis, though
initially associated with phenetics, has also found widespread use in phylogenetic
systematics. Computer programmes have been designed which can handle both kinds
of data and produce dendograms.
7.2. Phenetics.
The theory of phenetic analysis asserts that natural classifications should be based on
overall similarity. This method is critical of character weighting and emphasis on
phylogeny in classification. The central thrust of phenetic methods in taxonomy is the
objective examination of a suite of characters from an array of individuals in order to
determine patterns of variation for both the characters and the organisms.
Historically, characters were almost exclusively morphological, but the data base has
been expanded to include features from any field of comparative botanical study, such
as anatomy, biochemistry, and physiology. These techniques are therefore based on
the analysis of the phenotype, which provides an estimate of genetic similarity.
119
Two general approaches to deriving phenetic estimate of genetic similarity
have been used, neural analysis and numerical analysis (Estes & Tyrl, 1986). Neural
analysis is the ability to formulate mental groupings on the basis of the totality of
features of the examined specimens. It is to some extent dependent on the experience
and capability of the individual scientist and hence subjective in that sense.
Numerical analysis involves the use of computer-based systems to compare
individuals and groups (Operational taxonomic units, OTU's) on the basis of
numerous characters and to express the results in more easily interpreted diagrams.
Patterns of variation based on scores and even hundreds of characters are compressed
with varying degrees of distortion and loss of information into two dimensional
branching diagrams (phenograms) that serve as simplified models of overall
similarity. In a phenogram, the branching pattern along the vertical axis represents the
degree of phenetic relationship (or similarity). The advantages of numerical analysis
are that the comparative operations are clearly and rigorously defined and are,
therefore, repeatable; complex multidimensional patterns are reduced to easily
interpreted diagrams; classifications based on overall similarity exhibit predictive
potential; and the resultant classifications are based on observations of the organisms
and not on inferred evolutionary lineage. Numerical analysis, however, provides a
phenogram that summarises phenetic relationships, it does not provide a classification
system. It is therefore for the taxonomist still, to make decisions concerning the
circumscription, rank attribution and relationships of any taxa he decides to
recognise.
7.2.1. Procedure.
In this study, phenetic similarities were set out at two levels of OTU' s. Firstly, they
were analysed at species level. This was to see where 'bridging' species between
genera would appear in the dendograms. 'Bridging' species are those species whose
taxonomic positioning could be with either of the genera involved. Secondly, the
similarities were set out at generic level. This was to see whether the genera formed
120
any recognisable clusters within the subtribe. At species level two tests were carried
out. Firstly, a totality of all characters were used, i.e., all those characters which were
observed and have been accumulated from both herbarium specimens and living
plants (Figure 7.1). Secondly, a posteriori weighting of characters was carried out
and only those characters considered morphologically diagnostic were used (Figure
7.2). Similarly, all characters that could be used to delineate the genera of the subtribe
were used to seek phenetic similarities between the genera (Figure 7.3). I did not
perform any a posteriori weighting in the case of the genera because doing that would
reduce the genus definition to something very similar to a species, and hence not be
an all-inclusive category.
Character coding: The coding system used in this exercise was DELTA
(DEscription Language for TAxonomy) (see Appendix 11.3). This is an official
international data standard approved by the Taxonomic Database Working Group. A
PANKEY program package was used to perfect the data matrix and produce a
similarity coefficient matrix. PANKEY is a computer-based package of programmes
that provide specialised tools for diagnosis or identification in biology or medicine
(Pankhurst, 1991). The programme accepts variable or polymorphic characters, which
means that the OTU can be coded for all its possible character states. One other
feature of PANKEY is to generate a matrix of similarity coefficients. It is this matrix
of similarity coefficients that is then used for cluster analysis by other statistical
packages.
Cluster analysis: The statistical package used for producing dendograms was
NTSYS (Numerical Taxonomy and Multivariate Analysis System) Version 1.70. The
matrix of similarity coefficients produced by PANKEY was fed onto NTSYS. The
clustering method used was UPGMA (Unweighted Pair-Groups Method using
Arithmetic averages). This method produced well-resolved phenograms.
Results: The dendograms which were produced are as follows:
121
[NB* - The scales at the top and bottom of the phenograms indicate phenon levels,
the percentages of similarity. Therefore the lower the phenon level, the less similar
the species or genera]
Test 1: The results of this test are shown in Figure 7.1. The species separate into two
main groups at 56.9% similarity. The first group (I) consists of Asthenochloa and
Cleistachne species. This group is characterised by the absence of pedicellate
spikelets. The group then divides again at 63.9% similarity. The basis of this
separation is the total suppression of the pedicel and pedicellate spikelet in
Cleistachne. The second group (II) consists of all those species that have pedicels and
or pedicellate spikelets. This second group then divides into two (III & IV) at 60.4%
similarity on the basis of inflorescence type. Group III consists of all species with
either a branched paniculate inflorescence (V) or those with racemose inflorescence
(VI) (see Chapter 4 for definitions). Group IV consists of species with simple
(unbranched) paniculate inflorescences.
In the whole of the phenogram, species of the same genus cluster at the same
position except those of the genera Bothriochloa, Capillipedium and Dichanthium.
The species of these genera are interspersed amongst each other and do not form any
recognisable clusters. Also the species of the genus Pseudosorghum peculiarly
clusters with Euclasta species.
Test 2: This phenogram (Figure 7.2) was constructed using only those characters that
I considered to be diagnostic of the species of the subtribe. The phenogram looks very
similar to Figure 7.1. The groups are the same. The only differences are in the phenon
levels at which the groups separate and the order of species in the clusters. The
phenon levels are lower here. The whole group begins to resolve at 50.9%. Group I
separates from the rest at 57.9%. Group II divides at 58.8%. This shows that by
removing those characters common to all taxa, makes the taxa less similar. The lower
the phenon level at which taxa separate, the less similar the taxa.
Test 3: A totality of all characters were used to construct a phenogram at generic
level (Figure 7.3). The subtribe primarily divides into two generic groups (I & II).
122
These groups neither correspond to those in Figures 7.1 & 7.2 nor to those I proposed
for the subtribe, the Bothriochloastrae and Sorghastrae (see Chapter 3). Cleistachne
(II) separates from the other genera at the 38% phenon level. This separation is on the
basis of the absence of the pedicellate spikelet. Asthenochloa (III) subdivides from
the rest of group I at 41% similarity, on the basis of the absence of the pedicellate
spikelet and the rudimentary pedicel. Group IV consists of those genera with
pedicellate spikelets. This phenogram shows a clear picture of generic associations.
The following clusters are recognisable: Bothriochloa, Capillipedium, Dichanthium
and
Pseudosorghum; Sorghum-Sorghastrum; Vetiveria, Chrysopogon
and
Hemisorghum; Euclasta, Pseudodichanthium and Spathia. Similarly, here the genus
Pseudosorghum clusters with members of the Bothriochloastrae and not the
Sorghastrae as I suggested in generic delimitation of the subtribe (Chapter 3). The
generic clusters in the phenogram however reflect morphological similarities and
overlaps I referred to in Chapter 3.
7.2.2. Conclusion.
This exercise has been very useful in identifying species complexes (Figures 7.1 &
7.2) and generic groups (Figure 7.3). There are species complexes in the Sorghinae
which consist of morphologically similar species. These always separate at high
phenon levels, e.g., Chrysopogon fallax, C. latfolius and C. pallidus; Dichanthium
micranthum and D. foveolatum, etc.
At generic level, the two main groups do not correspond to my
Bothriochloastrae and Sorghastrae (see Chapter 3). Some of the generic associations
that come out of the clustering however correspond to those I suggested in my
discussion on generic delimitation in the subtribe (Chapter 3), e.g., Bothriochloa,
Capillipedium and Dichanthium; Chiysopogon and Vetiveria, etc.
123
0.560
0.640
0.720
I
0.880
-I
0.800
-
0.960
-I
--
1.040
Level
1.A5tlien 0.639
I.
L--- ---------------------------- -------------------- 2.Cle±st
0.569
--------- -------------------- 3.Euclas
0.822
------------------------- I .Euclae
0.833
--------------------------- 9.Pseudo
0.802
---------- -------------------- 5.luclae
0.774
30.l3otIiu 0.863
L ----------------------- 34.Bothr
0.808
36.Capil 0.914
L----------------38.Capil 0.830
28.Ootihr 0.884
32.Botlir 0.924
---------------
61.Potlir 0.863
35.Bot:hr 0.848
--L-L-I-------------------------- 29.Bothr
0.788
33.Bothr 0.860
37.Cepii
0.753
--------8.Pscudo
0.758
-•..- ........42.Uicha
0.79S
L------------- --
41.Dicha 0.881.
46.Dicha 0.851.
L-----------------------
43.Dicha 0.781
4 4 . Dicha 0.88
L--------------------45.Dicha 0.73:
27.I3otlir
0.149
31.othr 0.777
39.Capii. 0.884
L--------------------40.Capil 0.724
vii
10.Spath 0.699
-----------------------------------55.Sorqh
0.767
54 .Sorgh 0.882
L -------------------- 56.Sorgh 0.796
57.Sorqh 0.859
58.Soiqli 0.878
60.Sorqh 0.822
I
v
.---L-------L---L--L---------------------------- 59.Sorqii 0.665
47.Sorqh 0.914
L ---------------- 49.Sorgli
0.811
50.Sorqh 0.786
51.Sox:qh 0.813
ill
48. Sol cjh
0.851
---------52 .Sorqh
0.920
L -------------- L--L ---- L -------- L ----------------53.Sorqh
0.604
---------------------------------- --11.CIitye
0.771
13.Chrys 0.878
L6.Clirys 0.847
-L ------- L ------------------------- 18.CIiiys
0.761
--------------------- 12.CIn:yS
0.873
-14c11ryS 0.922
-19.Chrys 0.917
L ----- L ------ L ----------20.Clii:ye
0.82!
15.Chrye 0.881. L----------------- ----- 21.Cliiye
0.791.
---17.Chrys 0.681.
----- 6.Ileui.so
0.83.
L -------------------- --------- -7.lieuilso
0.73:
------------22.Vetiv: 0.85
-------------------------- 24.VPt)V.
---23.Vet:iv
0.866
---------------- ------ 25.Vel:iv
0.771
----
Iv
0.821
L ----- L ----- L --------------------------------- 26.Vetiv
level
I ---------I --- - -----I -------------I-------- ----I
0.720
0.800
0.960
1.010
0.880
0.640
# I. Aslhenochloa (enera
#2. aeLctach:e sorghoides
43. Eucla,cta clarkei
#4. Euclasla condylotricha
115. Euclasla oliganiha
#6. Heinisorghum inekongense
V. Ilenzisorghun: venustunt
Pseudodichaislhiuni serrafalcoides
Pseudosorghu,nfasciculare
1110. Spathia neurosa
11 II. Chrysopogon aciculalus
Chrysopogon aucheri
Cl, rysopogon fallax
Chrysopogonfulvus
Chrysopogon gryllus
Ci:rv.wpogon lalifolius
17. Chrysopogon orientalis
#18. Chrvsopogon pallidus
# 19. Chrysopogon plumulosus
Chrysopogon serrulatus
Chrysopogon sylvalicusi
Veliveria elongata
423. Vetiveriafihipes
#24. Vetiveriafulvibarbis
425. Vetiveria nigritana
Vetiveria zizanioides
Bothriochloa biloba
Bothriochloa decipiens
Bothriochloa insculpta
Bothriochloa ischaemum
#3 I. Bothriochloa kui,tzeana
Bothriochloa macera
Bothriocl,loa pertusa
Bothriochloa radica,,s
Bothriochloa saccharoides
Capillipedium a.ssimile
Capillipediumfiliculnse
Capillipediuni huegelii
Capillipediun: parviflorum
Capillipediunt spicigeruna
#4 1. Dichanlhiu,n annulatuns
Dicha,,thium aristatuns
Dlcha,,thium caricosuni
444. Dicha,,tl,iuiufoveolatuni
Dichanll,iu,n n,lcrauthu,n
Dichantl,ium sericeum
1147. Sorghastrum elliottii
Sorghastrumfuscescens
Sorgliaslrum incompletuns
var. bipennaizim
Sorgl,astrunt n,iflarUfl,
Sorghaslriin: nutans
452. Sorglzastruin pogonostachyuni
Sorghastruns setosum
Sorghum arundinaceum
Sorghum I,alepe,tse
Sorghum n,acrospern,u?n
Sorghun, nitidum
Sorghum versicolor
Sorghum p!u,,,osum
Sorghum stipoideunt
Bothriochloa bladhii
0.400
0,600
0.500
1
0.900
0.800
0.100
000
-
0.1'
I
-..Clist
3.505
- 3.LtcLU3
0.844
- 5.Euc].as
0.784
8.Pseuclo
0.815
-9.Pseudo
0.758
.1 .Eucl,i
0.818
31 .I3othr
0.833
33.BoShF
0.875
37 . Cap il
0.787
--------------------28.Bothr
0.925
32.BotIir
0.890
3!.. Bothr
0.878
L-------------- - 9. Bothr
0.861
36. Capil
0.94(
L-- - - 38.CapiI
0.841
31).33o1Iii
0.894
.
0. 8(
-L
-
-L
I
Q)
I
34 . (uI1)!
0.751
-•--- ----5,]
-
O
J
tlJ
. ----------------------- 12.Iiicha
0.80]
.11..Dicha
0.87;
I
. ---L--------------46.Dicha
I
.1----------------------- 43.DicI;a
44 Djulia
--------------------
I
D.uIii
-------------------.19. Cu
.
.
CA
•
105
0
0
706
----------------- -1.Hen,iso
1) 832
54 .Surqti
0.878
56.Solgh
0.789
---55.Suiqli
0.806
.51 .Sorch
0.883
58. Surçjh
0.915
--L --------- 60.SorgI
0.840
0.699
I
V
0 88
0.703
.
,..c
p
0.79!
10.Spat1
------------------------------
o
0.83!
0.844
---------------------40 Capil
V1. I
.----L-----------L--L----------------- 59.Sorgh
0.641
51.SoryI
0.772
47. Sorql;
0.928
. -------- I, --------49 Snrqh
0.841
50. sorcili
0.79!;
.48 .Sorqli
.
0.861
52 . Sorgl;
0 .908
------------ 53.Sotqh
L-------------L-L ------ L---L
0.588
6.Ilewiso
0.744
22 .V't.i.v
0.885
- - - L-----------------24 . Vu I iv
0.85C
I.)
C
'•
HI
-- --- .
•
23 . VC iv
II .95;
-- I ........29.VuLiv
0.821
L --------------------------26.Vutiv
0.121
IJC
y
-------------------------------------------13 , Chtya
--------------
1) . 883
0.91
L,--1-------------1.6.CI;y;
0.83'
.CInys
';.821
--------------------------12.CIil:yu
14.Clriys
(5.911'
0.93
- - .19. I!lu:yv
13.981;
L--L --------20.Chryo
0.812
___.Chrys
0.88(3
----------
.
--------------------------21.Clu.ys
0.711
jy
_____
II_
I
L------- L ------------- L ---- L ----------------------- l8.Cltiys ----Level
I --------- I --------- I --------- I --------- I --------- I ---------- I
0.400
0.500
0.600
0.700
0.800
0.900
1.000
.
#1.AstI,e,,och!oafeiiera
cleisiachne sorghoides
Euclasta clarkes
Euclasta condylolriclta
Euclasta ollgantlsa
Ileinisorglsuntinekongense
V. Ilemisorglsuns venustum
Pseudodiclianthiuin serrafalcoides
Pseudosorgliuinfasciculnre
#10.Spathianeurosa
#II. Chrysopogon aciculotus
Chrysopogon aucheri
Chrysopogonfallax
Cisrysopogonfulvus
chry.sopogongryllus
Chrysopogon latifolius
Chrysopogon orienralls
#I8. Chrysopogonpallidus
# 19. Chrysopogon plusnulosus
#20. chrysopogonserrulatus
#2I. Chrysopogon sylva(icus/
#22. Veti;'eriaelongata
Vetiveriafihipes
Vetiveriafulvibarbis
925.Vetiverianigrilana
#26. Vetiveria zizanioides
#27.BotI,riochloabiloba
#28.Boti:riochloadecipiens
#29. Bollsriochloa insculpia
#30.Bothriochloaischaenuini
#31. Bothriochloa ku,stzeana
#32.BothriochIoan:acera
Bothrioci:loaper(usa
Bolhrioci,loa radicans
BothriochIoasaccharoides
e
436.
,'e
llfp%
Capillipediumltuegelii
Capillipediunsparvflorum
Capillipediumspicigerum
Dicha,,thiun: annum/un:
Dicha,,thium arislatun,
#43.Dichanthiumcaricosun:
#44.Dichanthiumfoveolatuns
#45. Diclioiithsu:n nslcrantlsuni
#46.Dichanthiu,nsericeum
447. Sorghastrun: elliot/u
Sorghastrumfuscescens
Sorghastrum incompletsun
var. bipennatum
#50.Sorghastr::mminaruni
#51.Sorghastrum flu/ails
Sorghastrumpogonostachyuin
Sorgl:astrun:setosuni
Sorghum arundi,wceu,,,
Sorghum halepense
#56.Sorgliumnwcrosperniuni
#57. Sorghum nitidum
#58.Sorgl:unsversicolor
Sorgl::imj,lun,osum
Sorghum stipoideum
Bothriochloa blodhii
LO
C1
4
Figure 7.3. Phenograrn of 14 OTU's based on UPGMA cluster analysis (see
text for details). 1. Asthenochloa. 2. Bothriochloa. 3. Capillipediuin. 4.
Chrysopogon. 5. Cleistachne. 6. Dichanthiurn. 7. Euclasta. 8. Heniisorghum. 9.
Pseudodichanthium. 10. Pseudosurghum. I I . Sorghastrum. 12. Sorghum. 13.
Spathia. 14. Vetiveria.
l'lienoii levels
0.36
0.60
0.72
0.81
1.Asthen
0.41
2.Bothr 1 0.83
6.Dichan 0.73
3.C.pi11 0.61
10. Pseud 0.60
I I.Sorqh 0.68
12.Sorqh 0.58
1. Chi, ijso 0.65
11.Uetiv 0.52
8. Hem so
0.49
'?.[uclas 0.68
8. Pseudo 0.52
13. Spat h 0.38
5. [leisi
0.36
0.18
0.60
0.72
0.81
I'henon levels
7.3. Phylogenetics.
The theory of phylogenetic systematics asserts that evolution by descent is the
causative factor that has produced the diversity among modern plants and permits the
construction of a hierarchical classification. It is therefore the evolutionary histories
of taxa that are of primary interest for systematic investigation and the derived data
and their interpretation form the framework of the classification system. Furthermore,
the taxa in the system must be monophyletic. Evolutionary divergence, lineage, and
branching may be used in the circumscription of taxa, and overall resemblance that
might result from parallel or convergent evolution is not considered. Pheneticists and
phylogeneticists may use the same comparative techniques and the same data sets.
The basic distinction between the two groups rests on a philosophical foundation, that
is, whether in principle similarity or phylogeny should be most significant in the
formulation of a classification.
The core concept of phylogenetic systematics is the use of derived or
apomorphic characters to reconstruct common ancestry relationships and the
grouping of taxa based on common ancestry. It is essential for the phylogenetic
methods that an ancestral type, either a hypothetical ancestor or a relatively putative
primitive taxon, be recognised. Analysis progresses easily if there is fossil evidence,
if the groups are already well known from a phenetic vantage point, and if the
adaptive significance of the various character states is known. Evolution may still be
perceived even when only extant taxa are available to study. Correlation of characters
may provide a reasonable estimate of likely primitive and advanced states.
At about the same time as the birth of numerical taxonomy, Hemiig (1966),
proposed a new approach to determine evolutionary branching patterns, an approach
also readily adaptable to computer technology. This approach is termed cladistics.
Cladistics is basically a methodology that attempts to analyse phylogenetic data
objectively (and hence to produce phylogenetic classification), in a manner parallel to
that in which taxometrics seeks to introduce objectivity into phenetics and phenetic
classification (Stace, 1989). Despite this similarity, the two approaches and their
127
respective proponents are largely in conflict. Cladistics focuses exclusively on
inferred branching patterns of evolutionary lineages. In this approach the
classification must be retrieved directly from the cladogram, a two-dimensional
depiction of the branch points, each taxon must be monophyletic as recognised by
patterns of shared derived characters, and sister groups must be accorded equal rank
in the hierarchical system. In a cladogram the branching pattern along the vertical
axis represents ancestry in time (or evolutionary advancement). A number of
cladograms are possible for the same set of organisms and data. The cladogram with
the fewest character state changes, that is, the most parsimonious, would then have to
be chosen. Cladistic methods are not the only ones to adopt the principle of
parsimony, for the latter has been used by phenetic taxonomists, who attempt to relate
extant OTU's without hypothesising a priori any ancestor or directions of
evolutionary change.
For cladistic classification, the only characters which can be used are those
which are the outcome of evolutionary history, i.e., derived characters. Other
characters are irrelevant. Unfortunately, there is no method, statistical or otherwise,
that can be used to choose the right characters, and these therefore are chosen
subjectively by the taxonomist. The next problem is to decide the polarity of the
characters, that is, which character states are primitive and which are advanced. If
suitable fossils were always available for study, polarity could be decided without any
doubt, but this is practically never the case. Differences have therefore always ensued
between different workers when it comes to this aspect of cladistics.
A lot of debate has always been around since Hennig's proposals on the
cladistic theory. There are different opinions on what cladograms should be
interpreted to represent (see section 7.3.2). Most phylogeneticists argue that cladistic
methods should be used to come up with classifications that supposedly reflect
evolutionary relationships between taxa studied. However, one school of thought, that
of the so called 'pattern cladists', use cladistic methods without assuming that they
should be used to create classifications. These contend that cladistics should be used
im
merely to unravel the pattern of variation rather than detect the true phylogeny. They
believe that the pattern is nearly always close to the genealogy, and sometimes
coincident with it, but there is no certainty that it is the same. It is generally assumed
that when a hypothesis has been put forward about evolution of a group, this will then
be used to create a classification, this is not necessarily so. The onus is, therefore, still
on the taxonomic to impose his own interpretation on the outcome of cladistic
analysis.
Cladistic analyses of the grasses is a recent phenomenon. Linder & Ferguson
(1985) and Campbell & Kellogg (1987) concluded that the grass family is linked with
six small families of the Southern Hemisphere (Anarthriaceae, Centrolepidaceae,
Ecdeiocoleaceae, Flagellariaceae, Joinvilelaceae
and
Restionaceae).
These
conclusions are far from novel.
7.3.1. Some concepts associated with cladistics.
Cladistics has as its basis three concepts: apomorphy, monophyly, and parsimony. An
apomorphy is a derived character. If this character is shared by more than one taxon
it is called a synapomorphy. An apomorphy that is found only in a particular taxon is
called an autapomorphy. Every apomorphous character is paired with the character
from which it is derived, the plesiomorphous character (primitive or ancestral
character). Cladistic methods basically view evolution as an ordered, divergent, stepwise transformation of characters from plesiomorphous to apomorphous states.
If all organisms in a branching genealogical tree have the same single
ancestor, and all the descendants are included, then the group of organisms is said to
be monophyletic. On a cladogram, this translates into any group that includes all taxa
that share at least one synapomorphy. A monophyletic group of taxa is called a dade.
Non-monophyletic groups are of two types: (i) When a group of organisms all have
the same common ancestor, but not all the descendants are included, then the group is
paraphyletic (also called a grade); and (ii) if a group of organisms includes two or
129
more complete branches from different parts of a tree, but does not include the
common ancestor, then the group is said to be polyphyletic. A diagrammatic
representation of monophyly, paraphyly and polyphyly is shown in Figure 7.4. The
horizontal, solid black lines indicate the 'taxa' belonging to the particular group.
The process of assigning status of apomorphy to a character is called
determining polarity. Using an outgroup (or outgroups) is the most common way of
determining which characters are apomorphic (Funk, 1995). Characters found in the
outgroup as well as in some of the taxa in the ingroup (the group being studied) are
considered to be plesiomorphous. Those characters found only in some of the taxa of
the ingroup but are absent in the rest of the ingroup and in the outgroup are
considered apomorphous. An outgroup can be, but is not necessarily, the taxon most
closely related to the ingroup, the sister group.
130
MONOPHYLETIC
1
(ii)
23
4
PARAPHYLETIC
POLYPHYLETIC
1
23
4
Figure 7.4. A diagrammatic representation of the concepts of monophyly, paraphyly
and polyphyly.
131
7.3.2. Cladogram interpretation and evaluation.
Cladograms are statements about the distribution of shared characters. There are
various measures used to evaluate the 'robustness' of a cladogram. The simplest
measure is the tree length. This indicates the number of character state changes
required to explain the character state changes possessed by each terminal taxon on
the tree. The consistency index (CI) measures the amount of homoplasy (extra steps)
in a character. A Cl value of 1 indicates that a character has no homoplasy. The
messier the data on the tree, and the greater the amount of homoplasy, the more CI
approaches 0. The other commonly used statistic is the retention index (RI) which
measures the amount of implied synapomorphy in the data matrix that is retained as
synapomorphy on the tree, i.e., ratio of the apparent synapomorphy to actual
synapomorphy. Similarly, high homoplasy in characters is indicated by values near 0.
The Cl and RI are usually combined to form what is called a resealed consistency
index (RC). The RC excludes characters that do not contribute to the 'fit' of the tree
by excluding autapomorphic characters as well as those showing total homoplasy
(Funk, 1995). As with Cl and RI, RC ranges from 0 to 1, with higher RC values
indicating that characters in the data set are congruent with each other and the tree.
These indices can be used for each individual character as well as for the cladogram
as a whole.
Another approach to estimating the value of a particular cladogram with
respect to the data is by using the technique of bootstrapping. This technique
involves randomly sampling the original data set with replacement to create a series
of 'bootstrap' replicates of the same size as the original data set. New matrices and
their corresponding trees are produced. In each of these trees, percentages of
occurrence of particular monophyletic groups are recorded. These percentages of
occurrences (usually out of 100) that a particular monophyletic group appears among
the trees of the sample data sets can be considered an index of support for that
monophyletic group.
132
Cladistic analysis more often generates many equally parsimonious trees.
Therefore, consensus techniques are usually used. Consensus techniques summarise
sets of trees into a single topology. Two types of consensus techniques are common
in the literature: (I) Strict consensus - depicts only monophyletic groups retained in
all trees; and (ii) Majority rule consensus - shows the branching sequences that are
found in most of the trees and depicts monophyletic groups that are present in a
specified percentage of trees (usually 50%). Both these consensus techniques have the
potential of producing unresolved branching patterns, polytomies. Although
consensus trees are useful in identifying the areas of agreement and conflict among
competing trees, unless a consensus tree is identical to one of the equally
parsimonious trees, it can not be used to study character evolution (Funk, 1995).
As mentioned earlier, cladograms can be interpreted to represent various
assumptions. According to Maddison & Maddison (1992) they can represent:
The hypothesisedphylogenetic history, with lines representing lineages descending
through time and branch points representing speciation events. Two taxa placed next
to one another on separate terminal branches means that they share a common
ancestor unique to them, and neither is the ancestor of the other. If one had been an
ancestor, it would have been placed directly at an internal node of the cladogram.
The relative recency of common ancestry of the observed taxa. Taxa placed
together in a dade share a common ancestor not shared by other taxa outside the
dade. However, though these taxa are placed in terminal position on the cladogram,
the internode leading to the terminal node might or might not have existed in nature.
The hierarchical distribution of shared, derived, homologous character states
(synapomorphies). This interpretation asserts that taxa forming a dade do not
necessarily represent a group of taxa sharing uniquely a common ancestor, but rather
a group of taxa united by concordant synapomorphies. The cladogram is therefore a
summary of evolutionary changes in characters.
The hierarchical distribution of shared characteristics. Placing taxa in the same
dade does not necessarily mean that they share an evolutionary innovation, but rather
133
that they share a similarity. This interpretation makes no claims about evolution. This
interpretation is contradictory because cladograms thus produced are based on the use
of polarised characters, a necessary first step to cladistic analysis. Distribution of
shared characters would best be represented by phenograms.
7.3.3. Procedure.
Character coding: The first thing that was done was to polarise characters.
(Appendix 11.4). Plesiomorphic characters are indicated as 0 while apomorphic ones
have values higher than 0 (1, 2, & 3; see Table 7.1). Unlike the phenetic analysis (see
section 7.2.1), only representative species (type species) of the 14 genera of the
subtribe were used. Also, most characters from phenetic analysis were discarded
because they could not be satisfactorily polarised or polarising them made no sense.
Therefore, only 26 characters were chosen. The genus Saccharum (in subtribe
Saccharinae) is supposedly closely related to the genera of the Sorghinae, hence the
species, S. officinarum, was chosen as an outgroup.
Results and analysis: The statistical packages used for producing cladograms
were MacClade (Maddison & Maddison, 1992) and PAUP (Phylogenetic Analysis
Using Parsimony) (Swofford, 1993). These packages are relatively user-friendly and
(in theory) wholly compatible. MacClade was used for data input and tree
examination/interpretation. PAUP was used for the actual algorithmic analysis that
generates trees and tree summary data. Thus, in order to obtain optimal results, one
must flip backwards and forwards between the two packages.
Four cladograms were produced, each with the following tree statistics: Tree
length: 172; CI: 0.67; RC: 0.44; and RI: 0.65. A Majority rule consensus of these
trees was computed (Figure 7.5). The numbers on branches are the percentages of
occurrence of the particular clades in the original four trees. For instance, on the
Asthenochloa tenera/Cleistachne sorghoides branch, the 100% means that this dade
is present in all the four original trees. There are two unresolved branches,
134
polytomies,. in the tree. The first is represented by the Pseudosorghum fasciculare
branch. The second is Spathia neurosa and the two clades above it containing
Euclasta condylotricha and
Bothriochloa bladhii,
and
Pseudodichanthium
serrafalcoides and Dichanthium annulatum. This cladogram is very different from the
phenogram produced for the same set of genera (see Figure 7.3). I have not used the
cladogram to study character evolution because it does not look like any of the four
original trees.
135
newsorghinae2.1
1
2
3
4
5
6
7
8
9 11 0
11 11 2 11 3
14
1 5 11 6 1 1 7
1 8 11 9 120
2 1 122
23 124 125
26
1
Asthenochloatenera
0
0
0
0
1
1
0
0
?
0
0
2
2
1
?
1
0
0
0
0
0
1
1
2
1
?
0
0
?
3
1
2
0
1
0
0
0
0
0
0
0
i
0
3
2
?
0
0
1
1
0
0
1
0
?
0
1
?
1
0
0
0
3
Cleistachne sorghoides
Euclastacondylotricha
0
1
0
0
0
0
0
0
0
0
0
1
1
4
Pseudodichanthium serrafalcoide
1
?
0
0
1
0
1
?
2
0
0
2
1
0
1
0
0
0
0
1
0
0
1
0
1
1
5
Pseudosorghumfasciculare
1
1
?
0
0
1
0
0
?
0
0
2
1
0
0
0
0
0
0
0
0
0
0
o
1
1
0
0
0
0
1
1
1
1
o
1
6
Spathia neurosa
?
1
0
1
1
2
0
?
1
0
0
2
1
0
1
0
0
0
0
0
7
Chrysopogongryllus
0
0
0
0
1
1
0
1
?
1
0
0
0
0
?
0
1
0
0
0
0
1
0
0
1
1
8
Vetiveria zizanioides
0
0
0
0
1
1
0
1
?
0
0
2
1
0
0
0
0
0
0
0
0
0
1
0
1
1
9
Bothriochloabladhii
0
0
0
0
0
1
1
?
0
.0
1
2
1
?
0
0
0
0
0
0
1
0
1
0
1
i
10
Capillipedium parviflorum
0
0
0
0
1
1
0
0
?
1
1
0
0.
0
?
0
0
0
0
0
0
1
1
0
1
i
11
Dichanthiumannulatum
0
1
0
0
1
0
1
?
1
0
2
1
0
0
0
0
0
0
0
1
0
1
1
Sorghastrum nutans
0
1
0
1
1
0
0
0
?
0
0
0
0
1
1
12
0
1
?
0
0
0
0
0
0
0
1
0
1
1
13
Sorghumarundinaceum
1
?
0
0
1
1
0
0
?
0
0
2
1
0
0
0
0
0
0
0
0
0
0
0
1
-
14
Hemisorghum mekongense
0
0
0
0
0
1
0
1
?
0
2
1
0
0
0
0
0
0
1
0
1
0
0
0
1
0
0
?
0
2
1
0
0
0.0
0
Saccharumofficinarum
0
0
0
15
0
0
0
0
0
0
0
0
0
0
0
0
1
.
Table 7.1. A matrix of species and characters used in cladistic analysis.
(0
CO
Majority rule
Asthenoch/oa ten era
Cleistachne sorghoides
Euclasta condylotricha
Bothrioch!oa bladhii
Pseudodichanthiu,n serrafalco ides
Dichanthium annulazum
Spathia neurosa
Pseudosor-ghutn fasciculare
Sorghum arundinaceu,n
Chrysopogon gryllus
Capillipedium parviJlorutn
Sorghastrum nutans
Veliveria zizanioides
Hemisorghum mekongense
Saccharunt officinaruin
Figure 7.5. A Majority rule consensus cladogram of reperesentative Sorghinae
species.
137
7.4. General Discussion.
The two approaches, phenetics and cladistics, are two basic taxonomic philosophies
at extremes of a continuum. One extreme is that the categories of any classification
should be based on genetic similarity and dissimilarity, and the other contends that
the categories should reflect phylogenetic relationships (Estes & Tyrl, 1986). Even
though plausible arguments can be advanced for the significance of both approaches,
Pankhurst (1991) though advocating experimentation with both methods, notes that
'...the phenetic methods involve few subjective decisions, do not involve any major
approximations, and have effective algorithms. On the other hand, the cladistic
methods are much more ambitious, involve many subjective decisions and several
major approximations, and do not have generally effective algorithms'. In general, a
phylogenetic classification will be better at showing the supposed ancestral
relationships, and a phenetic classification will better represent the characters of
current day organisms. However, if evolution proceeds at a constant rate and in a
strictly divergent fashion, a correct phenogram will indeed appear exactly the same as
a correct cladogram, and a highly competent taxonomist might subjectively produce
the same pattern (Stace, 1989). When 'actual' phylogeny is not known, as in most
angiosperms (and certainly in grasses), phenetic approaches to classification may
often produce as plausible a 'phylogenetic' classification as any, and one that has
excellent predictive value (Hattersley & Watson, 1992).
There are a number of points J could raise from the experience of using
phenetic and cladistic methods in the present study. Cladistic methods are very
sensitive to characters used. The resolution is far much improved if very good few
characters are used, i.e., those characters that truly influence the 'fit' of the
cladogram. If a lot of useless characters are used, they tend to create 'noise' and
cladograms yielded are unresolved and uninformative. Also, reducing the number of
taxa to only a few representative ones seemed to help. In contrast, for phenetic
methods, the more characters one uses, the more resolved are the phenograms. These
methods allow for terminal taxa to have polymorphic characters, i.e., one can code for
138
more than one state of a character in a taxon. This can also be done in cladistic
methods. There is however a practical problem of the algorithms not coping well with
these characters and this leads to the production of erroneous cladograms. I was able
to produce phenograms at generic level because I could code for all the characters in
the species sampled. The same could have been done for cladistic analysis but this
would have resulted in a large portion of the matrix consisting of polymorphic
characters because character variations in other species of the genera would have had
to be included. I however used a sample of representative species. Also, it is almost
impossible to produce a single cladogram that one can use to study character
evolution, which means that one ends up having to compute consensus trees.
Neither of the exercises, phenetic and cladistic, produced dendograms which
exhibited similar generic groups to the ones I proposed for the taxa of the Sorghinae.
Phenograms nevertheless indicated some species complexes and some generic
associations I suggested in my intuitive delimitation of the genera of the Sorghinae
(see Chapter 3). Though worthwhile in terms of generating understanding of method
and philosophy, the exercises did not sway my intuitive understanding of the group
sufficiently to suggest any alternative classifications based on them. I simply do not
trust their outcomes. Between them, I incline to the certainty of similarity analysis
rather than the greater dubiety of phylogenetic assumption, as a basis for
classification.
139
Chapter 8. Taxonomic descriptions.
8.1. Introduction.
The order in which the genera have been arranged is not intended to suggest any
evolutionary sequence. The sequence is a reflection, to some extent, of phenetic
relationships between the different genera. The species are arranged likewise under
each genus. On average, five (5) specimens were closely observed for each species,
but in cases where the specimens showed considerable variation, more specimens
were used to cover this variation. Type specimens, either seen or not, are listed
separately. Sometimes only type localities are indicated due to scanty information,
and only in a few species was I unable to record information pertaining to types. I
was however hesitant to suggest any new types because I felt that the types might
exist somewhere but I was not able to trace them within the time that I had to finish
the research. Typification would have to be done at a later date.
The taxonomic descriptions that follow represent the species throughout the
whole of their morphological variations and ecological ranges. Any outstanding
characteristics are outlined, otherwise it is those characters found in the keys that
should be considered diagnostic. Taxonomic affinities are also discussed briefly.
Additional information (if available), e.g. chromosome numbers, which is not a result
from the present study, is included and sources of origin quoted.
Information on habitat and distribution was obtained mainly from the
herbarium sheets, unless otherwise stated. The specimens studied are arranged
alphabetically by country. Under each country, they are then arranged alphabetically
by author in the order: location, altitude, date of collection, collector, collector(s)
number, and, in parenthesis, the herbarium from which the material consulted came.
It will be obvious in some cases that not all the information was available on the
herbarium sheets.
140
8.2. Sorghinae Bluff, Nees & Schauer, Comp. Fl. Germ. ed. 2, 1: 46 (1836).
Synonym: Bothriochloinae Keng, in Sinensia 10: 282 (1939).
Inflorescence terminal or rarely axillary, of single or digitate or paniculate racemes,
the latter often in whorls; racemes with fragile rachis and slender internodes,
sometimes reduced to triads or single spikelets, occasionally with homogamous pairs.
Spikelets paired, dissimilar, seldom plumose. Sessile spikelet bisexual, usually.
dorsally compressed, the callus usually obtuse with cupuliform or truncate
articulation but the latter sometimes oblique (always so when callus pungent); lower
glume usually firm, more or less convex on the back and abruptly rounded on the
flanks (except Hemisorghum); lower floret reduced to a barren lemma; upper lemma
linear to oblong, entire or bibbed, usually with a glabrous awn. Pedicellate spikelet
male, neuter, or (rarely) bisexual, sometimes much reduced, rarely with a small
callus.
Genera 14.
A key to the genera of the Sorghinae:
Spikelets in racemes which are not interrupted by spathes, or solitary at ends of
branches; the racemes collected into whorled panicles; joints and pedicels terete, i.e.
not furrowed:
Spikelets solitary at the ends of branches; pedicellate spikelets absent, or
represented by barren pedicels:
1. Cleistachne
Pedicels entirely absent.
Pedicels present:
Pedicel well developed.
2. Sorghastrum
Pedicel reduced to a minute axis hidden by callus hairs.
3. Asthenochloa
Spikelets in pairs or threes; pedicellate spikelets present but often more or less
reduced:
141
Spikelets dorsally compressed:
Racemes up to 30cm long, tardily disarticulating, with scabrid
rachis nodes and pedicels; spikelets awnless; lower glume 2keeled for its whole length; lodicules glabrous.
4. Hemisorgh urn
Racemes very much less than 30cm long, readily
disarticulating; rachis nodes and pedicels ciliate; lower glumes
2-keeled near the tip; lodicules glabrous or ciliate:
Lower glume of sessile spikelet convex, rounded on the
5. Sorghum
sides; lodicules hairy.
Lower glume of sessile spikelet flat, with narrowly
inflexed margins; lodicules glabrous.
6. Pseudosorghum
Spikelets laterally compressed:
Spikelets in groups each consisting of a sessile and two more
or less reduced pedicellate spikelets; lower glume smooth, not
7. Chrysopogon
tuberculate.
Spikelets with many pairs of spikelets arranged in whorled
8. Vetiveria
racemes; lower glume tuberculate.
Spikelets in racemes which are interrupted by spathes, or the espatheate racemes
digitate or in pairs or solitary and terminal, sometimes the joints and or pedicels with
a translucent median furrow:
Margins of the lower glumes inturned and rounded at the sides, at the
most keeled upwards; awn usually hairy; lower glume of the sessile
spikelet keeled and broadly winged, glabrous.
9. Pseudodichanthium
142
Margins of the lower glume of the sessile spikelets sharply infolded, 2keeled; awn glabrous; lower glume of sessile spikelet not broadly
winged, hairy or glabrous:
Spikelets arranged in false racemes terminating the branches
and branchiets of a true panicle; pedicels with a median
10. Capillipedium
translucent furrow.
Spikelets arranged in false racemes which are either solitary
and terminal or subsessile, arranged along a central axis but not
at the tips of capillary branches:
Racemes solitary at the ends of branches supported by
spathes, but sessile spikelets not overlapping.
Spathia
Racemes several to many, if solitary then the sessile
spikelets imbricate; spathes absent:
Joints and pedicels with a translucent
longitudinal furrow; all spikelets, both sessile
and pedicellate, in the racemes, dissimilar:
All pairs of spikelets heterogamous,
rarely homogamous; racemes ascending;
glumes sometimes pitted.
Bothriochloa
The lowest 1-3 pairs of spikelets
homogamous; racemes nodding; glumes
never pitted.
13. Euclasta
Joints and pedicels without a translucent furrow;
lowest 1-3 pairs of spikelets homogamous; all
spikelets in the racemes more or less alike, with
obtuse apices.
143
14. Dichanthium
8.2.1. Cleistachne Benth., in Hook. Ic. P1. 14: t 1379 (1882). The type species is
Cleistachne sorghoides Benth.
Coarse annual. Inflorescence a panicle, its primary branches bearing racemes at
regular intervals along their length, each raceme reduced to a single sessile spikelet
(whose apparent pedicel is homologous to the raceme peduncle); spikelets solitary,
sessile; female fertile florets one; upper lemma stipiform; lodicules 2, ciliate;
pedicellate spikelet and pedicel completely suppressed.
Species 1.
Cleistachne sorghoides* Benth., in Hook. Ic. P1. 14: t 1379 (1882).-Hack., In DC.,
Monogr. Phan. 6: 653 (1889).-Stapf, in Pram, Flora of Trop. Africa IX: 154 (1917).Sturgeon, A revised list of grasses of Southern Rhodesia Part IV: 8 (1954).Chipindall, in Meredith, The grasses and pastures of South Africa: 468 (1955).Bogdan, A revised list of Kenya grasses: 56 (1958).-Jackson & Wiehe, An annotated
check-list of Nyasaland grasses: 33 (1958).-Bor, Grasses of Burma Ceylon India &
Pakistan: 119 (1960).-Napper, Grasses of Tanganyika: 98 (1965).-Hood, A guide to
the grasses of Zambia: 58 (1967).-Clayton, in Polhill, Flora of Tropical East Africa,
Gramineae (3): 734 (1982).-Gibbs Russell et al, Grasses of Southern Africa: 87
(1991).
Tall, erect, stilt-rooted annual up to 3m tall. Cuims slender, culm nodes and
internodes glabrous. Leaf blades linear-lanceolate, variable in width but often
expanded to greater than 2cm wide, hairy on both surfaces, but less so on the abaxial
* oldes is a Greek suffix meaning like or resembling, alluding to the resemblance of the species to
those of the genus Sorghum.
144
surface, with prominent midribs; leaf sheaths hairy or glabrous, when hairy, the hairs
just upward of the horizontal; leaf ligule an unfringed membrane.
Inflorescence a very compact panicle; racemes markedly reduced to single sessile
spikelets (the raceme peduncle equivalent to the spikelet pedicel); peduncles and
rachis internodes covered with hair; spikelet colour variable but ranges from straw
coloured to chestnut brown. Spikelets oblong; lower and upper glumes the same size,
slightly coriaceous; lower glume minutely truncate, finely 7-nerved, covered with
uniform short and dense hairs; upper glume glabrous below, hispid towards the apex,
clasped lengthwise by the lower; callus obscure, shortly bearded. Florets 2: lower
floret empty, lemma hyaline, oblong, ciliate, 2-nerved; upper floret bisexual, lemma
hyaline, ciliate on margins, finely 3-nerved, awned; awn 5-8 times the length of the
lemma, geniculate, with a twisted column hairy along the spiral and a serrate distal
bristle; palea shorter than the lodicules; lodicules 2, ciliate; anthers not seen;
caryopsis obovate oblong, truncate, brown. Pedicellate spikelet and pedicel absent.
Chromosome number: 2n = 36 (Garber 1950; Celarier, 1959).
Diagnostic characters and affinities: This is one species in which controversy has
arisen as to whether it is one species or three. Specimens seen showed variation in the
pubescence of the leaf sheath, simplicity (i.e. whether open or compact) of the
inflorescence and its overall colour. In detail (i.e. standard characters of the spikelet)
the specimens are the same. I therefore do not think that it would suffice to create two
or more species based on the variable characters mentioned above. It is quite
unknown as to the degree to which these variations may be partly due to
environmental variation.
Habitat: Grassland or savannah woodland.
Type: Mozambique, Shupanga, Kirk (K!).
Representative specimens:
Malawi: Dedza District, Mau-Livulezi Forest, 09.iv.1964, Adlard 607 (MAL);
Southern Region, Chiradzulu, altitude 1250m, 1970, Brummitt & Banda 9831 (K);
145
Southern region, Zomba District, lower slopes of Zomba Plateau, above Salisbury
bridge, Zomba, altitude 1140m, 26.iv.1980, Brummitt, Masiye & Tawakali 15560
(MAL); Lilongwe District, Agricultural Research Station, 07.iv. 1951, Jackson 457
(MAL).
Mozambique: Niassa District, altitude 1000m, 23. iii. 1964, Torre & Paiva 11359
(K).
Zambia: Dambo, 06.vi.1931, Fraser 14 (K).
Zimbabwe: Salisbury (Harare), Teviotdale, 20.iii.1965, Bingham 1429 (BM);
Melsetter District, Muchira river, Tarka Forest Reserve, altitude 1100m, iii. 1971,
Goldsmith 12/71 (BM); Gokwe District, 19.iii.1984, Mahiangu 981 (K); Salisbury
(Harare), Henderson Research Station, 1 0.iii. 1964, West 4743 (BM); Banket District,
15.iv.1947, Wild 1891 (K).
Distribution: Tropical Africa and India.
8.2.2. Sorghastrum Nash, in Britton, F!. North. United States: 71(1901). The type
species is Sorghastrum nutans (L.) Nash.
Synonyms: Poranthera Raf., in Bull. Bot. Geneve 1: 221 (1830); Dipogon Steud.,
Nom. Bot. ed. 2,1: 518 (1840).
Annuals or perennials. Inflorescence a panicle, the subdivided primary branches
bearing short racemes, these sometimes reduced to triads. Sessile spikelet callus
obtuse or pungent (with involucral hairs in S. pogonostachyum); lower glume
coriaceous, convex, keeled at the tip; upper glume broadly convex; upper lemma
mostly entire, sometimes bibbed, awned; lodicules glabrous. Pedicellate spikelet
usually reduced to a barren pedicel, but present and bisexual in S. fuscescens and S.
pogonostachyum.
Species ± 16.
146
A key to taxa sampled:
Pedicellate spikelet bisexual, some pedicels barren:
Culm nodes glabrous; racemes up to 7-jointed; spikelet surrounded by an
involucre of silvery hairs to 3/4 the length of the sessile spikelet; awn 4mm
1. pogonostachyum
long; upper lemma entire.
Culm nodes hairy; racemes 3-15-jointed; callus with a short beard of hairs
only; awn 25mm long; upper lemma bibbed.
2.fuscescens
Pedicellate spikelet absent, all pedicels barren:
Racemes single-jointed:
Plants annual; sessile spikelet 4-5.5mm long; upper lemma entire; awn
3. incompletum var. bipennatum
25-40mm long, glabrous.
Plants perennial; sessile spikelet 6.5-10mm long; upper lemma
bibbed; awn 25-70mm long, glabrous or hairy:
Sessile spikelet 6.5mm long; lower glume of sessile spikelet
hairy on the back; awn 25-35mm long, glabrous.
4. elliottii
Sessile spikelet 10mm long; lower glume of sessile spikelet
glabrous; awn greater than 35 (-70)mm long, hairy along the
5. minarum
spiral.
Racemes 2-many-jointed:
Culm nodes glabrous; raceme peduncles glabrous; sessile spikelet
4mm long; upper lemma with entire margins, awned or awnless.
6. setosum
Culm nodes hairy; not with the above combination of characters:
Nodal hairs a conspicuous ring; raceme peduncles hairy; sessile
spikelet 6-8mm long.
147
7. nutans
Nodal hairs inconspicuous, short and velvety; raceme
peduncles glabrous; sessile spikelet 4-7(-8)mm long:
Leaf sheaths not produced into auricles; spikelets 5-7(8)mm long; awns 3-8(-10)mm long, straight or bent and
8. friesii
twisted.
Leaf sheaths with auricles; spikelets 4-7(-8)mm long;
awns 8-16mm long, never straight, always bent and
9. stipoides
twisted.
1. Sorghastrum pogonostachyum* (Stapf) Chippind. & Pole Evans, in Bot. Surv. S.
Mr., Mem.: 247 (1948).
Synonyms: Sorghum pogonostachyum Stapf, in Pram, Flor. Trop. Afr. IX: 144
(1917); Mischanthidium gracilius Napper, in Kirkia 3: 120 (1963).
Perennial grass with a creeping rhizome, sending up a small tuft of cuims from base.
Cuims about im tall; culm nodes glabrous. Leaf blades narrow, glabrous, smooth
except at the slightly rough tips; leaf sheaths firm, tight, glabrous, smooth, striate;
ligule short, stout, scarious, pubescent.
Inflorescence a narrow panicle up to 15cm long; racemes loose, containing spikelet
pairs i.e. both sessile and pedicellate spikelets. Sessile spikelet lanceolate; lower
glume chartaceous, sparsely pilose on the back, 5.5mm long; upper glume glabrous;
callus short, round, with hairs 4mm long. Florets 2: lower floret empty, lemma
oblong, ciliate on the margins; upper floret bisexual, lemma entire, 3-nerved, awned
from the tip; awn bristle-like, 4mm long; epaleate; lodicules 2, minute, fleshy; anthers
*
pogon, beard, and stachyum, from stachy, which means 'relating to a spike, the epithet therefore
alludes to the haity spikelets.
148
3, 3mm long; ovary oblong. Pedicellate spikelet lanceolate, also with two florets of
which the upper is bisexual; anthers 3, 3mm long, pedicels pilose, some barren while
others bear the bisexual pedicellate spikelets.
Diagnostic characters and affinities: Many, if not most, of the pedicels bear perfect
spikelets.
Habitat: Wet places, especially along water courses.
Type: Angola, country of the Ganguellas and Ambuellas, Gossweiler 2225 (K!).
Representative specimens:
Malawi: Madzimarera Dambo, Dzalamanyao, 04.xii. 1951, Jackson 693 (K).
Zambia: Chisinga Ranch near Lurungi Upland Dambo system, altitude 4600ft,
17.ix.1961, Astle 901 (K); Luwingu District, Lunte river dambo, 38km east of
Luwingu, 10.ii.1962, Robinson 4939 (K); Central Province, Chakwenga Headwaters,
100-129 km east of Lusaka, 27.x.1963, Robinson 5770 (K; PRE); Mkushi District,
Mkushi river dambo, 27.x.1967, Simon & Williamson 1222 (PRE).
Distribution: South Central and Eastern Africa.
2. Sorghastrumfuscescens (Pilger) Clayton, in Kew Bull. 30(3): 509 (1975).
Synonyms: Miscanthidium fuscescens Pilger, in Notizdbl. Bot. Bart. Berlin II: 806
(1933).
Tufted perennial of up to 2m tall. Culm nodes shortly hairy. Leaf blades linear, 1060cm long, 2-7mm wide, densely and coarsely pubescent; leaf sheaths also
pubescent; ligule a short unfringed membrane.
Inflorescence a panicle of about 20-30cm long; the intemodes of the main
inflorescence axis pilose in the upper third, the rest glabrous, the joints pilose;
racemes loose, 3-10cm long, with 3-15 spikelet pairs; raceme peduncles pilose on the
149
upper third of their length; raceme internodes and pedicels pilose along their whole
lengths; both pedicellate and sessile spikelets alike with a brown developed brown
awn. Sessile spikelet 6mm long, lanceolate; lower glume coriaceous, pilose on the
back, lanceolate, 5.5mm long, brown; upper glume keeled in the middle, pilose on the
keel only, the rest of the body glabrous, slightly longer than the lower glume, 6mm
long, ciliate on the margins. Florets 2: lower floret empty, lemma ovate, ciliate on the
margins; upper floret bisexual, lemma bibbed, ciliate on the margins, awned; awn
25mm long, bigeniculate, glabrous; lodicules 2, opaque, fleshy; epaleate; callus 1mm
long, rounded with short hair; anthers 3, 4mm long; ovary oblong. Pedicellate spikelet
similar to the sessile, also with a bisexual upper floret; glumes coriaceous; anthers 2
or 3, 4mm long; pedicels pilose, some barren while others bear the spikelet.
Diagnostic characters and affinities: Many, if not most, of the pedicels bear perfect
spikelets. This species together with S. pogonostachyum differ from other members of
Sorghastrum in this respect, but conform in most other characters.
Habitat: Grassy hillsides and lake margins.
Type: Tanzania, Njombe District, Msima, Schleiben 1010 (K!).
Representative specimens:
Malawi: Linthipe Dedza river, 1 0.iii. 1951, Jackson 426 (K).
Tanzania: Iringa Province: Njombe, altitude 6000ft, 1931, Emson 129 (K); Mufundi,
Brooke Bond Tea Estate, about 15km south of Mafinga, 5km north of Ngwazi House
in dry marsh, 08°30'S 35°10'E, altitude 1820m, 12.xi.1988, Gereau & Lovett 2454
(K).
Zambia: Mporokoso District, near Muzombwe, western side of Mweru-we-Ntipe,
altitude 3500ft, 15.iv.1961, Phipps & Vessey-Fitzgerald 3218 (BM); Arbecom
District, Kawimbe, altitude 5600ft, 26.iii. 1959, Webster A239 (K).
Distribution: South Central and Eastern Africa.
150
3. Sorghastrum incompletum (Presi.) Nash var. bipennatum (Hackel) Davila,
Systematic revision of the genus Sorghastrum (1988).
Synonyms: Andropogon bipennatus Hackel, Androp. Nov. Fl. 68(8): 142 (1883);
Sorghum bipennatum (Hackel) Kuntze, Re. Gen. P1. 2: 791 (1891), based on
Andropogon bipennatus Hackel; Sorghum bipennatum (Hackel) Stapf, Fl. Trop. Aft.
IX: 144 (1917), based on Andropogon bipennatus Hackel; Sorghastrum bipennatum
(Hackel) Pilger, in Notizdbl. Bot. Gart. Berlin XIV: 96 (1938).
A weak annual up to 1.5m tall (rarely more than 2m), often decumbent and rooting at
the nodes below and with short stilt roots. Cuims unbranched; cuim nodes with short
shiny hairs. Leaf blades up to 30cm long, linear-lanceolate, 3-10mm wide; leaf
sheaths striate; ligule a tough membrane.
Inflorescence a loose, narrowly lanceolate panicle; raceme peduncles capillary,
glabrous; racemes reduced to a single sessile spikelet, always accompanied by 2
barren pedicels; spikelets golden in colour, narrowly ovate, 4-5.5mm long, with a
rounded base; lower glume coriaceous, dark brown, ovate, pilose with white hairs on
the back, 5.5mm long; upper glume smooth on the back, also dark brown, boatshaped. Florets 2: lower floret empty, lemma hyaline, ciliate along margins; upper
floret bisexual, lemma entire, awned; awn 40mm long, smooth, scabrid along the
spiral, bigeniculate; palea present, single-nerved, small; lodicules 2, fleshy; no anthers
seen; ovary ovate. Pedicellate spikelet absent; pedicels 2, 4mm long, barren, pilose
with white hairs.
Chromosome number: 2n = 20 (Olorde, 1975).
Diagnostic characters and affinities: Has a looser and darker panicle compared to
var. incompletum. Also has longer awns.
Habitat: Wooded grasslands.
Type: Suda, Seriba Ghattas, Schweinfurth 2486 (K!).
151
Representative specimens:
Malawi: Dedza District, Mua-Livulezi forest, 09.iv. 1964, Adlard 607 (SRGH);
Lilongwe District, Kampini, 17.iv.1950, Jackson 761 (MAL); Zomba, Naisi road,
altitude 3500ft, 14.v.1949, Wiehe N/106 (K).
Mozambique: Malema District, 09.iv.1962, Lemos & Marrime 330 (PRE, K); Cabo
Delgado Province, Montepuez area, 21 .iv. 1984, Timberlake 3408 (SRGH); Cabo
Delgado, Montepuez, 09.iv.1964, Torre & Paiva 11.833 (K).
Zambia: Chipata District (no exact location given), 17.iii.1970, Abel 182 (SRGH);
Luangwa valley, Mfuwe Camp, altitude 2000ft, 01 .iv. 1966, Astle 4765 (K).
Zimbabwe: Gokwe District (no exact location given), 15.iii.1984, Mahlangu 975
(SRGH, K).
Distribution: African variety of Sorghastrum incompletum: tropical areas of Africa;
Sudano-Zambesian and Guineo-Congo regions.
4. Sorghastrum elliottii (Mohr) Nash, Nor. Amer. Flor. 17: 13 (1912).-Britton &
Brown, Ill. F!. (ed.2)i: 120, fig. 275 (1913).-Silveus, Texas Grasses: 743, 745 (1933).Hitchcock, Man. Grasses U. S. (ed.1): 753, fig. 1672, fig. 1673 (1935).-Blomquist,
Grasses N. Carolina: 205, fig. 229 (1948).-Allen, Grasses of Louisiana: 110 (1975).Davila, Systematic revision of the genus Sorghastrum (1988).
Synonyms: Chrysopogon elliottii Mohr., Bull. Torrey Club 24: 21 (1897);
Andropogon nutans Elliot, Bot. S. C. and Ga. 1: 144 (1816); Sorghum nutans
Chapman, Fl. S. U. S.: 5853 (1883); Sorghum nutans ssp. linneanum Hackel, F!.
Bras. 2: 276 (1883); Andropogon nutans var. linneanus (Hackel) Hackel, in DC.,
Monogr. Phan 6: 531 (1889); Chrysopogon nutans var. linneanus (Hackel) Mohr.,
Bull. Torrey Club 24: 21(1897), based on Sorghum nutans ssp. linneanum Hackel;
Andropogon linneanus (Hackel) Scribner & Kearney, Bull. Us. Dep. Agr. Agrost. 24:
40 (1901), based on Andropogon nutans var. linneanus (Hackel) Hackel;
152
Sorghastrum linneanum (Hackel) Nash, in Small, F!. S. E. U.S.: 66 (1903), based on
Sorghum nutans Chapman; Sorghastrum apalachicolense Hall, Sida 9(4): 302 (1982).
Perennial grass with fibrous roots. Cuims single or tufted, mostly 80-180cm tall,
erect; cuim nodes hidden by leaf sheaths. Leaf blades 40 cm long or less, mostly 38mm wide, striate, glabrous, scabrous (rough); leaf sheaths smooth and glabrous;
ligule firm, membranous, truncate or variously lobed, 2-4mm long, thickened
laterally to form stiff sheath auricles.
Inflorescence a narrow, loosely and sparsely flowered panicle 15-25cm long; racemes
consisting of single sessile spikelets accompanied by 2 hairy pedicels; raceme
peduncles glabrous or sparingly pilose towards the apex. Sessile spikelets deep
chestnut brown, about 6.5mm long; lower glume hirsute with yellowish-brown or
cream white hairs on the back, lanceolate, 9-nerved, 6.5mm long; upper glume
lanceolate, 6nmi long, shortly mucronate, carinate, the back mostly glabrous, 5nerved, sparsely ciliate on the incurved inner margins; callus short and blunt with
yellowish-brown or cream white hairs. Florets 2: lower floret empty, lemma broadly
ovate, 4.5mm long, 2-nerved, ciliate along the margins; upper floret bisexual, lemma
bidentate (bibbed), awned from the sinus; awn 35mm long, dark brown and twisted,
the column geniculate, several times longer than the spikelet; lodicules 2, small,
fleshy, highly vascularised; anthers 3, 2.2mm long; caryopsis obovate; pedicels
4.5mm long, with yellowish brown hairs up to 2mm long.
Chromosome number(s): 2n = 20 (Brown, 1950; Garber, 1950).
Habitat: In sandy soils and in or along margins of woodlands.
Type: Nash (North American Flora 17 (1912)) indicates the type locality to be South
Carolina, USA. Despite an axhaustive search for this type, I was not able to track it.
Representative specimens:
United States of America:
153
Alabama: Lawrence County, 8.8 miles south of Meulton, 23.ix.1970, Kral 41261
(BM).
Georgia: Elbert County, north side of Broad river at Anthony Shoals, 14'/2 miles
south-east of Elberton, 01.xi.1949, Duncan 10574 (BM).
North Carolina: Mecklenburg County, 7 miles west of Huntersville along the
Catwaba river, 02.xi.1958, Ahies & Duke 50115 (E).
South Carolina: McCormick County, 6 miles south-west of McCormick, 1 8.ix. 1949,
Duncan 10394 (BM).
Texas: Smith County, Amigo, 19.ix.1945, Moore 1041 (B, BM).
Virginia: Sussex County, State Game Sanctuary, north-west of Newville, 18 .ix. 1945,
Fernald& Long 14895 (BM).
Distribution: South-eastern and eastern USA.
5. Sorghastrum minarum (Nees) Hitchcock, Contr. U. S. Nat. Herb. 24: 501 (1927).Davila, Systematic revision of the genus Sorghastrum (1988).
Synonyms: Trachypogon minarum Nees, Agrost. Bras.: 349 (1829); Andropogon
minarum (Nees) Kunth, Rev. Gram. 1: 507 (1830), based on Trachypogon minarum
Nees; Stipa penniglumis Trin., Mem. Acad. St. Petersb. 1: 77 (1833); Chrysopogon
minarum (Nees) Bentham, Linn. Soc. 9: 73 (1881), based on Trachypogon minarum
Nees; Sorghum minarum (Nees) Hackel, Fl. Bras. 2: 276: (1883), based on
Trachypogon minarum Nees.
A slender erect perennial to about 1.5m tall. Cuims rooting from the lower nodes;
cuim nodes hidden by sheaths, with short velvety hairs where exposed. Leaf blades
flat, mostly glabrous, infrequently with short tubercle-based hairs on the upper
surface, up to 35cm long and 10mm wide, linear-lanceolate and becoming attenuate;
154
leaf sheaths longer than the internodes, striate, glabrous, drawn into a stiff
membranous ligule of about 10mm long.
Inflorescence a profusely divided panicle 10-25cm long, compact with long awned
spikelets giving it a characteristic facies; main inflorescence axis and raceme
peduncles glabrous. Sessile spikelets 10mm long, narrowly lanceolate; lower glume
glabrous except towards the base and on the side keels, opaque, 7.5mm long,
narrowly lanceolate with a truncate apex; upper glume 7mm long, narrowly
lanceolate, with a single sharp keel on the back and pointed cilia at the apex. Florets
2: lower floret empty, lemma ovate, 3-nerved and ciliate on the margins, hyaline;
upper floret bisexual, lemma awned, bidentate; awn 7cm long, chestnut brown,
bigeniculate, with hairs 2mm long along the spiral, tightly twisted; epaleate; lodicules
2, fleshy; callus sharp, with long whitish hairs, 3mm long; no anthers seen; caryopsis
obovate pedicels 2, hairy, 3.5mm long.
Chromosome number(s): 2n = 20 (Pereira, 1982 & 1986).
Habitat: Plains and open grasslands.
Type: Type not found. Type locality not known.
Representative specimens:
Argentina: Posadas, Loreto, Yabebiriz, 31 .i. 1908, Ekman 563 (BM); Province
Corrientes, 24.ii.1951, Pederson 1033 (E).
Brazil: (no location given), 14.iv.1909, Dusén 7962 (E).
Paraguay: Cordillera de Altos, Cerro Choice, 23.i.1903, Fiebrig 779 (E); Zwischen
Rio Apa und Rio Aquidaban, i.1909, Fiebrig 5089 (BM); Cordillera de Ville-Rica,
i. 1905, Hassler 8774 (BM); in regione lacus Ypacaray, i.1913, Hassler 12480 (B,
BM); Prope Sapucay, ix.1913, Hassler 13025 (E, BM).
155
Distribution: South America, in Argentina, Paraguay, Bolivia and Brazil.
6. Sorghastrum setosum* (Griseb.) Hitchcock, Contr. U. S. Nat. Herb. 12(6): 195
(1909).-Hitchcock, Manual of the grasses of the West Indies: 409 (1936).-Davila,
Systematic revision of the genus Sorghastrum (1988).
Synonyms: Andropogon setosus Griseb., Cat. P1. Cuba: 235 (1866); Sorghum
parvflorum Desv., Hamilton Prodr. P1. Ind. 0cc. 12 (1825); Trachypogon stipoides
(HBK) Nees, F!. Bras.: 351 (1829); Andropogonfrancavillamus Foum., Mex. P1. 2:
56 (1881); Andropogon agrostoides Speg., Anal. Soc. Genc. Argent. 16: 36 (1883);
Sorghum nutans ssp. micranthum var. submuticus Hackel, F!. Bras. 2(3): 275 (1883),
based on Andropogon setosus Griseb; Andropogon nutans var. submuticus (Hackel)
Hackel, in DC., Monogr. Phan. 6: 529 (1889), based on Andropogon setosus Griseb;
Andropogon nutans var. agrostoides (Speg.) Hackel, in DC., Monogr. Phan 6: 529
(1889), based on Andropogon agrostoides Speg.; Sorghastrum francavillanum
(Foum.) Hitchcock, Contr, U. S. Nat. Herb. 12(6): 195 (1909), based on Andropogon
francavillanum Foum; Sorghastrum agrostoides (Speg.) Hitchcock, Bot. Gaz. 51:
300 (1911), based on Andropogon agrostoides Speg.; Sorghastrum parvflorum
(Desv.) Hitchcock & Chase, Contr. U. S. Nat. Herb. 18: 287 (1917), based on
Sorghum parvflorum Desv.; Sorghastrum stipoides (HBK) Nash ssp. agrostoides
(Speg.) Ros., Am & Iz., Gramineas Uruguayas: 201 (1970), based on Andropogon
stipoides HBK and Andropogon agrostoides Speg.
A tall tufted perennial grass. Cuims up to 1.5m tall, glaucous, bluish green; culm
nodes glabrous, upper portion of culm intemodes tinged pinkish-red. Leaf blades
long, up to 50cm long, 7mm wide, involute or flat, striate, glabrous; leaf sheaths
glabrous, striate, rarely hirsute at the apex; ligule a membrane 2mm long.
*
setosu(s)m, setose, bristly, i.e., beset with scattered ascending stiff hairs.
156
Inflorescence a panicle 20cm long, oblong to oblong-linear; racemes 1-5-jointed;
raceme peduncles glabrous; rachis internodes pilose; sessile spikelets accompanied by
pilose pedicels (every sessile spikelet with a pedicel and the terminal with two
pedicels); pedicels about '/2-% the length of the spikelet; spikelet yellowish-brown in
colour; lower glume pilose on the back with very fine and uniformly distributed hairs,
3.5mm long, ovate; upper glume glabrous on the back, shiny, narrowly ovatelanceolate or boat-shaped, 4mm long. Florets 2: lower floret empty, lemma ovate,
same length as lower glume, 2-nerved; upper floret bisexual, lemma entire, 3-nerved,
awnless; lodicules 2, fleshy; anthers 3, 2mm long; ovary obovate, tiny; pedicels
barren, 3mm long.
Chromosome number(s): 2n = 20 (Saura, 1944; Carnahan & Hill, 1961; Davidse &
Pohl, 1972 & 1974; Pereira, 1986).
Diagnostic characters and affinities: This species can be either awned or awnless.
Most specimens seen were however awnless. Pedersen 1003 is the only awned form,
with an imperfect awn about 3mm long, the exserted portion not longer than the
spikelet.
Habitat: In wet seasonally flooded areas.
Type: Type locality, Cuba.
Representative specimens:
Argentina: (both without locality), 07.ii.1951, Pedersen 1003 & 27.iii.1957, 4522
(E).
Brazil: Território Federal de Roraima, Municipio de Boa Vista, Reserva Ecologica
De Maraca, 61°50'W 3°35'N, 28.iii.1987, Lewis 1584 (E); Territory of Roraima,
SEMA Ecological Reserve, liha de Maraca, Roraima, 3°22'N 61°25'W, Santa Rosa
area, about 4km north-east of Station buildings, 3011988, Ratter, Milliken, Ramos &
Mota R6172 (E).
Guatemala: Sacate, 25.viii.1971, Ortiz 1868 (BM).
157
Jamaica: St. Andrew, Greenwich Bridle road, altitude 3000ft, 06.vii.1960, Adams
7496 (BM); Bermuda Mt., Port Royal Mtns., altitude 2300ft, 13.vii.1913, Harris
11635 (BM); Hall's Delight, St. Andrew, altitude 1000ft, 22.vii.1916, Harris 12394
(BM).
Distribution: Tropical and subtropical regions of the Americas.
7. Sorghastrum nutans* (L.) Nash, in Small, Fl. Southeast U. S.: 66 (1903).-Nash, in
North Amer. Fl. 17: 129 (1912).-Britton & Brown, Ill. F!. (ed.2)i: 120 (1913).-Deam,
Grasses of Indiana: 327 (1929).-Silveus, Texas Grasses: 742, 744 (1933).-Hitchcock,
Man. Grasses U. S. (ed.1): 752 (193 5) and (ed.2): 777 (195 1).-Kearney & Peebles, Fl.
P1. & Ferns Ariz.: 153 (1942) and Ariz. P1.: 143 (1951).-Blomquist, Grasses N.
Carol.: 205, 228, 247 (1948).-Fernald, Gray's Man. (ed.8): 234, 333 (1950).-Fasset,
Grasses of Wisconsin: 94, 350 (1951).-Gould, The grasses of Texas: 577 (1975).Davila, Systematic revision of the genus Sorghastrum (1988).
Synonyms: Andropogon nutans L. Sp. P1. ed. 1: 1045 (1753); Andropogon avenaceus
Michx., F!. Bor. Amer. 1: 58 (1803); Andropogon ciliatus Eli., Bot. S. C. & G. 1: 144
(1816); Sorghum nutans (L.) A. Gray, Man.: 617 (1848); Sorghum avenaceum
Chapm., Fl. South U. S.: 583 (1860); Chrysopogon nutans (L.) Benth., in J. Linn.
Soc. Bot. 19: 73 (1881); Chrysopogon avenaceus (Michx.) Benth., in J. Linn. Soc.
Bot. 19: 73 (1881); Sorghum nutans subsp. avenaceum (Michx.) Hack., in Mart., F!.
Bras. 2: 3, 274 (1883); Sorghum nutans subsp. linnaeanum Hack., in Mart., F!. Bras.
2: 276 (1883); Andropogon albescens Fourn., Mex. P1. 2: 56 (1886); Andropogon
confertus Trin. ex Fourn., Mex. P1. 2: 55 (1886); Andropogon nutans var. avenaceus
(Michx.) Hack., in DC., Monogr. Phan. 6: 539 (1889); Chrysopogon nutans var.
avenaceus Coy. et Brann., Rept. Geol. Surv. Ark. 4: 234 (1891); Poranthera nutans
* nutans, nodding (direction), inclining very much from the perpendicular so that the apex is
directed downwards, alluding to the inflorescence which is said to be nodding in the living state.
158
Raf. ex Jacks., Ind. Kew 2: 606 (1894); Poranthera ciliata Raf. ex Jacks., Ind. Kew 2:
606 (1894); Chrysopogon nutans var. linnaeanus Mohr., in Bull. Torn Bot. Club 24:
21(1897); Sorghastrum avenaceum (Michx.) Nash, in Britton, Man. Fl. N. U. States:
71(1901); Andropogon linnaeanus (Hack.) Scribn. et Keam. in Scribn. et Ball., in
Bull. U. S. Dept. Agric., Div. Agrost. 24: 40 (1901); Sorghastrum linnaeanum
(Hack.) Nash, in Small, F!. Southeast U. S.: 66 (1903); Holcus nutans (L.) 0. Kuntze.
ex Stuck., in Ann. Mus. Nac. Buenos Aires 11: 48 (1904); Holcus nutans var.
avenaceus Hack. ex Stuck., in Ann. Mus. Nac. Buenos Aires 11: 48 (1904);
Chalcoelytrum nutans (L.) Lunell, in Amer. Mid!. Nat. 4: 212 (1915).
A perennial grass. Cuims erect from short scaly rhizomes, 80cm to 2m tall; cuim
nodes hispid with stiff short hairs. Leaf blades long, linear, flat, mostly 5-10mm wide
(rarely more than 1 5mm wide), tapering to a narrow base and an attenuate apex,
glabrous, scabrous; leaf sheaths glabrous or the lower ones infrequently slightly
hispid, continued at the apex as a stiff membranous ligule; ligule 2-5mm long, usually
developed marginally as thickened, pointed sheath auricles.
Inflorescence a contracted panicle 15-30cm long; uppermost branchiets, inflorescence
joints, raceme peduncles and pedicels all hispid with silvery hairs; racemes 2-5jointed. Spikelets 6-8mm long, the glümes light brown or straw coloured; lower
glume flat or slightly concave on the back, hirsute, narrowly ovate, 7mm long; upper
glume 7mm long, convex on the back, glabrous. Florets 2: lower floret empty, lemma
broadly ovate, emarginate, ciliate along the margins; upper floret bisexual, lemma
awned, entire, ciliate along margins; awn 15mm long, geniculate, with a tightly
twisted column; epaleate; only filament remains seen; caryopsis narrowly obovate;
pedicel hairy, 4mm long.
Chromosome number(s): 2n = 20 (Carnahan & Hill, 1961); 40 (Church, 1929;
Brown, 1950; Bowden, 1960); 80 (Gould, 1968).
Diagnostic characters and affinities: Inflorescence is said to be nodding in the
living state, hence the name nutans.
159
Habitat: Important range forage species and indicator of range in good condition
(Gould, 1975); prairie grass of North America.
Uses: Native pasture species.
Type: Type locality, Virginia, USA.
Representative specimens:
United States of America:
Indiana: Wells County (no exact location), 15.ix.1901, Deam sn. (E).
Kansas: Douglas County, Lawrence, ix. 1892, Stevens 102 (E).
Missouri: Jefferson County, Seckman, 04.ix.1972, Christ s. n. (E).
New Mexico: Sierra County, south of the Black Range, 24.viii.1904, Metcalfe 1242
(E).
North Carolina: Iredell County, Catwaba river, 24.xi. 1958, Ahies & Haesloop 51874
(E).
Texas: Smith County, Amigo, 07.xi.1945, Moore 1049 (E).
Distribution: Central and East USA, Mexico, introduced to India.
8. Sorghastrumfriesii (Pilg.) Pilg., Notizbl. Bot. Gart. Berlin XIV: 96 (1938).
Synonyms: Andropogon ftiesii Pilger, in R. E. Fries, Erg. Schwed. Rhod.-Kongo
Exped. 1911-1912 I: 195 (1916); Sorghum micratherum Stapf, in Pram, Flor. Trop.
Afr. 9: 142 (1917), based on Andropogon nutans var. angolense Rendle, Cat. Afr. P1.
Weiw. 2(1): 152 (1899); Sorghum friesii (Pilger) Hubbard, in Kew Bull. Misc. Inf.:
109 (1934), based onAndropogonfriesii Pilger.
Tufted, shortly rhizomatous perennial. Culms up to 120cm tall, often rooting from the
lower nodes; cuim nodes many (crowded) in the lower part, with inconspicuous
160
velvety hairs; culm intemodes terete, hollow. Leaf blades glabrous, to 20cm long and
2-6mm wide, linear to linear-lanceolate, usually reflexed, leaf base expanded; leaf
sheaths non-auriculate, glabrous; ligule an unfringed membrane, 0.5-1.5mm long.
Inflorescence a loose open panicle; primary branches divided; raceme peduncles
glabrous, sometimes hairy towards the apex; racemes (2-)4-7-jointed; rachis
internodes and pedicels hairy. Sessile spikelet 5-7(-8)rnrn long, lanceolate; callus
obtuse, bearded; glumes papery; lower glume hirsute on the back, slightly depressed
longitudinally; upper glume glabrous, boat-shaped. Florets 2: lower floret represented
by a hyaline lemma, lemma ovate, 5mm long, finely 2-nerved, ciliate on the upper
margins, apex emarginate; upper floret bisexual, lemma bibbed, 4mm long, lobes
glabrous, awned; awn 3-8(-10)mm long, straight, glabrous; epaleate; lodicules 2, tiny,
fleshy, glabrous; anthers 3, 4-5mm long; ovary oblong; pedicel 5mm long, shorter
than the sessile spikelet, hairy, curved towards the base.
Habitat: Typically favours wet areas such as swamps, road drains and stream banks.
Often on poor sandy soils.
Type: Type not seen. Type locality not known.
Representative specimens:
Botswana: Ngamiland, Boro flood plain, altitude 940m, Biggs M517 (SRGH); 11km
south of Gumare on road to Nokaneng, 18.iii.1976, Ellis 2680 (K); Boro river at
Buffalo fence north of Maun, altitude 920m, 28.ii. 1987, Long & Rae 464 (E); Lediba
at Matsaudi aga Mmatalelo Island, 30.iii.1973, Smith 503 (K); Floodplain near Kwara
Bochai river, 02.vi.1973, Smith 562 (E); Xudum Drainage, Okavango, altitude 3300ft,
15 .iii. 1961, Vessey-Fitzgerald 3241 (BM); Okavango river, 19km north of Shakawe
on Botswana (Bechuanaland) border, 16.iii.1965, Wild & Drummond 7091 (BM);
Kwando, James Camp, 18022'S 23°32'E, 11.iv.1975, Williamson 16(K).
Namibia: Rundu, Dcatsingu, Kavango, altitude 1060m, 13.iii.1986, Potgieter 352
(E).
161
Zambia: Mushwishi Agricultural Station, Broken Hill, altitude 3800ft, 08.xii.1953,
Hinds 175 (K); roadside at Munshiwemba, 1 7.i. 1942, Stohr 744 (PRE); Abercorn
District, Saisi river (near Jericho), 27.ii. 1958, Vesey-Fitzgerald 1547 (K).
Zimbabwe: Gwelo District, G.T.C. sand, altitude 4600ft, 10.iii.1967, Biegel 1989
(BM); Gokwe District, about Y2mile north of Gokwe, 11 .ii. 1963, Bingham 484
(SRGI-I); Victoria District, 30.iii.1973, Chiparawasha 665 (K); Shangani, Gwampa
vlei at Tunke pan, Gwampa Forest Reserve, altitude 3000ft, ii.1956, Goldsmith 3 1/56
(K); Sipolilo, Nyamunyeche Estate, Karoi vlei, 07.ii.1979, Nyariri 673 (PRE);
Matobo District, Mtsheleli Dam, altitude 4400ft, 20.ii.1965, Simon 160 (BM).
Distribution: A southern African species.
9. Sorghastrum stipoides (Kunth) Nash, in N. Amer. Fl. 17: 129 (1912).-Clayton &
Renvoize, in Polhill, F!. Trop. E. Afr. Gramineae Part 3: 732, fig. 169 (1982).
Synonyms: Andropogon stipoides Kunth, in H.B.K., Nov. Gen. Sp. 1: 189 (1816);
Andropogon trichopus Stapf, in Kew Bull. 1897: 287 (1897); Sorghum trichopus
(Stapf) Stapf, in Fl. Trop. Afr. 9: 141 (1917); Sorghum rigidfolium Stapf, in F!. Trop.
Aft. 9: 143 (1917); Sorghastrum trichopus (Stapf) Pug., in Engl. & Pranti,
Pflanzenfam., edn 2, 14e: 142 (1940); Clayton, in Hepper, F!. West Trop. Afr., edn2,
3: 468 (1972); Sorghastrum rigidfolium (Stapf) Chippindall & Pole Evans, in Mem.
Bot. Surv. S. Afr. 22: 247 (1948), Chippindall, in Meredith, Grasses and pastures of
South Africa: 468 (1955).
A robust tufted perennial up to 1.5m tall with a hard creeping rhizome. Cuims erect,
usually unbranched, the internodes glabrous, nodes with a ring of inconspicuous
adpressed white hairs. Leaf blades usually glabrous, 15-45cm long and 3-6mm wide,
usually inrolled, rigid and conspicuously narrowed towards the base; leaf sheaths
auriculate; ligule an unfringed membrane, 1.5-4mm long.
162
Inflorescence a linear-lanceolate panicle, about 20cm long, primary branches
subverticillate, with capillary peduncles; racemes 1-4-jointed; rachis internodes
violet-hirsute. Spikelets all alike, lanceolate, 4-7(-8)mm long, the laterals each
accompanied by a single densely violet-hirsute pedicel, the terminal on each raceme
by two such pedicels; callus rounded, hirsute; lower glume 7mm long, coriaceous,
tawny, green-nerved, pale violet-hirsute, broadly convex across the back, emarginate;
upper glume about same length as lower glume, paler, glabrous proximally, shortly
hirsute distally, emarginate. Florets 2: lower floret reduced to a hyaline lemma; upper
floret bisexual, with a linear bibbed lemma, ciliate on the margins; awn geniculate,
flattened distally, 8-16mm long, glabrous; palea hyaline, minute; lodicules 2, fleshy,
glabrous; anthers 3; caryopsis oblong.
Habitat: Low lying areas subject to flooding.
Type: Colombia, Popayan to Almaguer, near Rio Putes, at the foot of Mt. Socobon,
Humboldt & Bonpland (P, K).
Representative specimens:
Malawi: Zomba, near Tongani, 20.iv. 1950, BW N/492 (MAL); Karonga, Mwemtete
Dambo, 26.vi.1951, Jackson 552 (K); Mposa Dambo, 13.v.1952, Jackson 814 (K);
Nkhotakota District, Katimbira village, T.A. Kanyenda, 13.v.1986, Patel & Kwatha
3194 (MAL).
Mozambique: Sul do Save, Maputo, 20.ii.1952, Mrye & Carvalho 1148 - (K);
Mahotas, Marracuene, xii.1945, Pimenta 52565 (PRE, SRGH); Beira District,
Cheringoma Coastal Area, Zuni Drainage, 5km west of Nyamaruza Camp, v.1973,
Tinley 2842 (K).
South Africa: Transvaal: White river, 'Pull Scar, altitude 3300ft, 05.v.1929, Mogg
s.n. (K); Pilgrimsrest District, 09.ii.1961, Strey 3592 (K).
Zambia: Mbala District, Uninji pans, altitude 5000ft, 16.v.1968, Sanane s.n. (K);
Abercorn District, wet pasture near Lunzuo river bridge, 20 miles from Abercorn,
altitude 5000ft, 5.iv.1959, Webster A283 (K).
163
Distribution: Tropical America and Tropical Africa.
8.2.3. Asthenochloa Buse, in Miq., P1. Jungh.: 367 (1854).-Backer & Brink, Flora of
Java: 602 (1968). The type species is Asthenochloa tenera Buse.
Synonym: Garnotiella Stapf, in Hook. Ic. P1. 25: t 2494 (1896).
Decumbent perennial. Inflorescence a branched panicle; ultimate branchiets pedicellike; spikelets sessile, accompanied by a minute pedicel (rudiment of pedicellate
spikelet) scarcely visible among callus hairs; spikelets very small (2mm long); callus
obtuse; lower glume herbaceous, broadly convex; upper glume laterally compressed,
navicular-keeled, acutely acuminate or with a short awn (mucronate); lower floret
suppressed; lower lemma and palea wanting; upper floret bisexual; upper lemma
shortly bibbed, awned; lodicules absent; anthers 2; caryopsis oblong. Pedicellate
spikelet reduced to a tiny barren pedicel concealed in the callus beard.
Species 1.
Asthenochloa ten era Büse, in Miq., P1. Jungh.: 368 (1854).-Backer & Brink, Flora of
Java: 602 (1968).
Synonym: Garnotiellaphilippinensis Stapf, in Hook. Ic. P1. 25: t 2494 (1896).
A perennial grass. Culms slender, erect or ascending, weak, sparingly branched,
terete, glabrous. Leaf blades narrowly lanceolate, with strongly narrowed base and
acute tip, glabrous on both sides with scabrous margins, the margins pilose
underneath; leaf sheaths glabrous, striate; ligule short, ciliate.
Inflorescence a branched panicle, oblong, 4-10cm long; ultimate branchiets pedicellike, with a cupuliform pibose apex, bearing a minute pedicel (rudiment of pedicellate
spikelet) scarcely visible among callus hairs, 1 sessile spikelet. Spikelet with 2 florets,
164
one empty and the other bisexual, with a geniculate awn; glumes thinly herbaceous,
subequal; lower glume as long as the spikelet or somewhat shorter, with an
emarginate apex, on either side of the back with 2 very slender nerves; upper glume
as long as spikelet, navicular-keeled, with broad hyaline margins embracing floret,
acutely acuminate or with short awn (mucronate), 3-nerved; upper lemma small,
thinly membranous, acutely 2-lobed, with a rather long, geniculate awn from sinus;
awn column twisted, bristle straight and serrated; palea and lodicules absent; anthers
2, linear; ovary glabrous; styles 2, free, rather long; stigma feathery; caryopsis
oblong. Pedicellate spikelet absent.
Diagnostic characters and affinities: The main characteristic feature is the
completely suppressed pedicellate spikelet represented by a minute rudimentary
pedicel hidden by the callus hairs.
Habitat: Shaded places, steep banks of terraces, and of roads, dry river beds.
Type: Philippines, Miagao, Province Ilo-ilo, iii. 1886, Vidal 3994 (K!).
Representative specimens:
Java: altitude 200-600m, 19.iv.1920, Backer 30773 (K); Lombok, altitude 11851265m, 31.v.1909, Elbert 1601 (K); Timor, ii - iv.1929, Walsh 58 (BM).
Philippines: Province of Benguet, Luzon, Buea river, x - xi.1905, Merrill 4322 (K);
Miagao, Province Ilo-ilo, iii.1886, Vidal 3994 (K).
Distribution: Java and The Philippines.
8.2.4. Hemisorghum C. E. Hubbard, in Bor, Grasses Burma Ceylon India & Pak.:
686 (1960). The type species is Hemisorghum mekongense (A. Camus) C. E.
Hubbard.
Tufted perennials. Inflorescence a panicle, its primary branches subdivided, bearing
long loose racemes with glabrous or scabrid rachis internodes. Sessile spikelet callus
165
obtuse; lower glume thinly coriaceous, flattened on the back, glabrous, 2-keeled for
most of its length, the margins becoming sharply inflexed towards the base; upper
lemma awned or awnless; lodicules glabrous. Pedicellate spikelet smaller than the
sessile and neuter.
Species 2.
A key to the species:
Sessile spikelet awnless; leaf blades broad (up to 45mm); panicle axis and raceme
1. mekongense
peduncles relatively thick.
Sessile spikelet awned; leaves narrow (up to 10mm); panicle axis and raceme
2. venustum
peduncles fihiform.
1. Hemisorghum mekongense (A. Camus) C.E. Hubbard, in Bor, Grasses Burma
Ceylon India & Pak.: 687 (1960).
Synonyms: Sorghum halepense (L.) Pers. var. mekongense A. Camus, in Bull. Mus.
Hist. Nat. Paris 25: 497 (1919); Sorghum mekongense (A. Camus) A. Camus, in
Lecomte, Fl. Indo-Chine 7: 323 (1922).-Rhind, Grasses Burma: 65 (1945); Schmidt,
in Agron. Tropicale 13: 212 (1958).
Tufted perennial. Cuims tall, stout, erect, solid, terete, simple or sparingly branched.
Leaf blades lanceolate, broad, 25-45mm wide, flat; ligule very short, truncate, densely
short ciliate.
Inflorescence a large, loose panicle with numerous verticillate branches; racemes fewto-several-noded. Sessile spikelet bisexual, awnless; glumes dissimilar; lower glume
lanceolate, obtuse, flat or slightly convex on the back in the lower part, 2-keeled, with
narrow inflexed margins; upper glume lanceolate, acute, rounded on the back below
the middle, with membranous ciliate margins. Florets 2: lower floret barren, lemma
slightly shorter than the glumes, lanceolate-oblong, palea absent; upper floret
166
bisexual, lemma oblong-lanceolate or narrowly ovate, entire or shortly bibbed, finely
1 -nerved, with the nerve sometimes produced as a minute mucro between the lobes;
palea present, as long as the lemma, similar in texture, linear-oblong, nerveless;
lodicules very small, glabrous; anthers 3, 3mm long; ovary ovate. Pedicellate spikelet
neuter, linear to narrowly lanceolate, smaller than the sessile; lower glume linear to
narrowly lanceolate, 2-keeled, scabrid on the keels; upper glume slightly shorter to
much shorter than the lower, narrowly lanceolate-oblong, keeled on the back towards
the apex, membranous; lower lemma much reduced or absent; epaleate.
Diagnostic characters and affinities: Awnless spikelets, very broad leaves, robust
plant and erect, a large panicle. Both panicle axis and raceme peduncles
comparatively thicker, non-linear and shorter, making the inflorescence much
compact than that of H venustum.
Habitat: Along river banks.
Type: Laos, Mekong, Paklai, Muong mai, Lakône (whereabouts uncertain).
Representative specimens:
Burma: Mergui District, Tenasserui river, 03.v.1932, Maung Po Khant 13417 (4
sheets, K).
Thailand: Palanam Sifaram, altitude 1200m, 06.v.1932, Kerr 21356 (4 sheets, K).
Distribution: South Burma to Indo-China.
2. Hemisorghum venustum (Thw.) W. D. Clayton, in Kew Bull. 27(3): 448 (1972).
Synonyms: Andropogon venustus Thw., Enum. P1. Zeyl.: 367 (1864); Vetiveria
venusta (Thw.) Willis, Rev. Cat. Pb. Ceylon: 110 (1911); Bothriochloa venusta
(Thw.) A. Camus, in Ann. Soc. Linn. Lyon n.s 76: 165 (1931); Capillipedium
venustum (Thw.) Bor, Grasses Burma Ceylon India & Pak.: 113 (1960).
167
Perennial. Tufted grass, 1.2m tall; basal sheaths complicate and imbricate. Culms
unbranched; culm nodes glabrous; cuim internodes terete. Leaf blades mostly folded,
up to 50cm long and 10mm wide; sheaths keeled, glabrous; ligule a ciliate membrane.
Inflorescence paniculate; panicles drooping; pedicels solid; both pedicels and rachis
intemodes are glabrous. Sessile spikelet bisexual, larger than the pedicellate one;
lower glume 4.5mm long, glabrous, ovate-lanceolate, flat on the back, 2-keeled for
most of its length; upper glume 4mm long, with a single keel, spinulose along upper
half of keel, hyaline in most of its body. Florets 2: lower floret empty (barren), lower
lemma hyaline, drawn to a narrow apex; upper floret bisexual, lemma narrowly
elliptic with a bibbed apex, awned; awn about 10mm long, glabrous; anthers 3,
2.5mm long; ovary oblanceolate. Pedicellate spikelet neuter; lower glume 4mm long,
lanceolate, glabrous; upper glume 3.5mm long, strongly carinate; upper lemma
oblong, 2.5mm long; epaleate; pedicel glabrous, non-linear with a cupuliform apex.
Diagnostic characters and affinities: Awned spikelets, awn bristle-like, tussock
forming grass. This species has been moved from genus to genus. It is often placed in
Capillipedium but does not fit there because that genus has pedicels with a translucent
longitudinal groove, and the upper lemma is entire and stipiform. It has also been put
in Vetiveria but this genus has laterally compressed spikelets with lower glume
rounded on the back. Finally it is sometimes placed in Sorghum, but Sorghum itself
has a plump, broadly convex lower glume, 2-keeled only near the tip, ciliate pedicels
and rachis internodes and ciliate lodicules.
Habitat: Roadsides and shallow soils of river banks.
Type: Sri-Lanka, Central Province, Rambodde, altitude 4000ft, Thwaites 2875 (BM!
K!).
Representative specimens:
India: Bababuden Hills District, altitude 6500ft, 29.xii. 1893, Talbot 3267 (K).
Sri-Lanka: road between Hakgala and Nuwara Eluja, altitude 5400ft, 27.xii. 1950,
Ballard 1264 (K); Markelüja, Moray Estate, 27.x.1993, Clayton & Jagosekera 6088
(K); Uva Province, Sita Eluja, lower Patna valley, 29.iii.1968, Tetsuo, Kayama &
Samarakoon 13542 (K); Nuwara Eluja District, between Hakgala Botanical Garden
and Nuwara Eluja, 14.xi.1968, Wirawan 754A (K).
Distribution: Southern India and Sri-Lanka.
8.2.5. Sorghum Moench, Meth.: 207 (1794) nom conserv; Hubbard, in Hook. Ic. P1.
34: t 3364 (1938); Garber, in Univ. California Pubi. Bot. 23: 283-361 (1950);
Snowden, in J. Linn. Soc. Bot. 55: 191-260 (1955); Celarier, in Cytologia 23: 395418 (1959); Ivanyukovich & Doronina, in Trudy Priki. Bot. Genet. Selek. 69: 18-27
(1980). The type species is Sorghum bicolor (L.) Moench.
Synonyms: Blumenchia Koel., Decs. Gram.: 28 (1802); Sarga Ewart & White, in
Proc. Roy. Soc. Victoria 23: 296 (1911).
Annuals or perennials, tufted or sometimes rhizomatous, mostly robust. Inflorescence
a large panicle, its primary branches simple or subdivided, bearing short dense
racemes with hairy internodes. Sessile spikelet dorsally compressed; callus obtuse or
pungent; lower glume cartilaginous, convex, rounded on the flanks but becoming 2keeled near the tip, usually hairy; upper lemma awnless, or bibbed and awned;
lodicules ciliate. Pedicellate spikelet male or neuter, or reduced to a lower glume.
Species ± 20.
A key to taxa sampled:
Nodes glabrous or pubescent:
Racemes tough or tardily disarticulating; grain large, commonly exposed by
the gaping glume; annuals; cultivated.
1. bicolor
Racemes fragile; grain enclosed by the glumes; perennials or annuals; weeds
or wild:
MCI
Perennial with abundant creeping rhizomes; sessile spikelet epaleate;
2. halepense
lodicules ciliate; weed.
Annual, rhizomes absent; palea 2mm long; lodicules glabrous; wild.
arundinaceum
Nodes bearded:
Plants annual:
Racemes 3-7-jointed; spikelets variously coloured; sessile spikelet 57mm long; callus short and blunt; awn 25-40mm long, glabrous.
versicolor
Racemes mostly single-jointed; spikelets a plain colour; sessile
spikelet 10-18mm long; callus long and pungent; awn 60(-155)mm
5. stipoideum
long, hairy on the spiral.
Plants perennial:
Racemes 3-5-jointed; sessile spikelet 4mm long, epaleate; awned or
6. nitidum
awnless.
Racemes 1-6-jointed; sessile spikelet 8mm long; palea present; awn
60(-100)mm long, hairy on the spiral.
7. plumosum
1. Sorghum halepense (L.) Pers., Syn. P1. 1: 101 (1805).-Bor, Flora of Assam 5: 352
(1940).-Sturgeon, A revised list of the grasses of Southern Rhodesia Part IV: 6
(1954).-Chippindal, in Meredith (ed.), The grasses and pastures of South Africa: 460
(1955).-Clayton, in Hepper (ed.), F!. West Trop. Afr. ed.2(3): 467 (1972).-Gould, The
grasses of Texas: 575 (1975).-Wheeler, Jacobs & Norton, Grasses of New South
Wales: 252 (1982).-Gibbs Russell et a!, Grasses of Southern Africa: 320 (1990).Cope & Hosni, A key to Egyptian grasses: 54 (1991).-Simon, A key to Australian
grasses: 158 (1993).
170
Synonyms: Holcus halepensis L., Sp. P1. ed. 1: 1047 (1753); Andropogon
arundinaceus Scop., Fl. Cam. ed. 2(u): 274 (1772); Milium halepense (L.) Cay.,
Descr. P1.: 306 (1802); Blumenbachia halepensis (L.) Koel., Deser. Gram.: 29 (1802);
Andropogon halepensis (L.) Brot., Fl. Lusit. I: 89 (1804); Andropogon Sorghum (L.)
Brot. subsp. halepensis (L.) Hack. var. halepensis (L.) Hack. subvar. genuinus Hack.,
in DC., Monogr. Phan. 6: 502 (1889); Andropogon halepensis (L.) Brot. var.
genuinus (Hack.) Stapf, in Hook., F!. Brit. Ind. 7: 183 (1896); Andropogon halepensis
(L.) Brot. typicus Asch. et Graebn., Syn. Mitteleur. Fl. II Abt. I: 47 (1898); Sorghum
miliaceum (Roxb.) Snowden, in J. Linn. Soc. (Bot.) 55: 205 (1955); Sorghum
miliaceum var. parvispiculum Snowden, in J. Linn. Soc. (Bot.) 55: 209 (1955).
Perennial with abundant creeping rhizomes. Cuims slender to robust, mostly simple
but sometimes branched, 0.5-3.5m tall, up to 2cm wide; cuim nodes exposed,
glabrous. Leaf blades linear, narrowed to a point at the apex, 20-90cm long, 0.5-4cm
wide; leaf sheaths glabrous; ligule an unfringed membrane.
Inflorescence a panicle; lower primary branches naked for about 2-5cm, glabrous;
racemes 1 to 5-jointed; rachis internodes and pedicels slender, densely ciliate with
whitish hairs, hairs up to 1mm long. Sessile spikelet elliptic, 4-6.5mm long, almost
glabrous to densely hairy, awned or awnless; glumes coriaceous; lower glume 2keeled, keel wings ending in minute teeth, forming with the pointed apex a 3-toothed
tip, hairy on the back, dorsally compressed, broadly lanceolate; upper glume sparsely
hairy on the back, boat-shaped; callus short and blunt, shortly bearded. Florets 2:
lower floret empty, lemma lanceolate, ciliate; upper floret bisexual, lemma bibbed,
ciliate, awned or awnless; awn (when present) 12mm long, glabrous, untwisted;
epaleate; lodicules 2, fleshy, ciliate on the apical corners; anthers not seen; ovary
ovate, small. Pedicellate spikelet male, narrower than the sessile, 5.5mm long; glumes
papery; lower glume lanceolate, ciliate on the lateral keels, glabrous, 5.5mm long;
upper glume sharply carinate, glabrous, 5.5mm long; lower lemma broadly
lanceolate, ciliate on the margins, with a truncate apex, 4mm long; upper lemma
171
ovate, ciliate on margins, 3mm long; lodicules 2, fleshy, glabrous; anthers 3, 2.5mm
long; pedicels hairy on the margins.
Chromosome number(s): 2n = 20, 40 (Faruqi, Quraish & Halai, 1979).
Diagnostic characters and affinities: The distinguishing features of this species are
the abundant slender wide-spreading rhizomes, narrow leaf blades, and the somewhat
small contracted panicles. A frequent variation of the spikelets is the presence or
absence of an awn to the upper lemma. The majority of the plants have awned
spikelets, but sometimes in specimens of the same gathering and number, the
spikelets of one panicle are awned and those of the other awnless. It is also not
uncommon to find most of the spikelets in a panicle awnless but accompanied by a
few that are distinctly awned. It is on this basis that some authors have recognised
two forms or varieties, S. halepense subsp. halepense (the typical awned form), and S.
halepense subsp. muticus (the awnless form).
Habitat: In moist areas on river banks, in clay soils and wet sandy soils.
Uses: Cultivated fodder.
Type: Type locality, Syria.
Representative specimens:
Australia: Queensland, Proserpicie, Michael 1415 (E); Q.A.0 - Gatton road,
30.x.1970, TS 5291 (E).
Cyprus: Perapedhi, altitude 3500ft, 13.vii.1940, Davis 1851 (E).
Russia: Caucasus, District Abkhasia, Sukhumi, altitude 3m, 12.vi.1956, Davis 33683
(E).
Saudi Arabia: Jabal Fayja, 100km north-east of Jigan, altitude 5100ft, 02.viii.1982,
Collenette 3711 (E); Saudia City compound, Jiddah, altitude 20ft, 27.ix.1983,
Collenette 4583 (E).
Turkey: Province Tunceli, Ovacik, altitude 1400m, 21.vii.1957, Davis & Hedge
D.31493 (B).
172
USA: North Carolina, Mecklenburg County, between Newell road and R.R tracks, 0.4
mile north of Newell Post Office, Newell road, 21.x.1966, Williams & Matthews s.n.
(E).
Distribution: Tropics of Old and New Worlds.
2. Sorghum arundinaceum (Desv.) Stapf, in Pram, F!. Trop. Afr. IX: 114 (1917).Sturgeon, A revised list of grasses of Southern Rhodesia Part IV: 6 (1954).-Clayton,
in Hepper (ed.), Fl. West Trop. Afr. ed.2(3): 467 (1972).-Clayton, in Polhill (ed.), F!.
Trop. East Afr., Gramineae (3): 727 (1982).-Lowe, Flora of Nigeria, Grasses: 277
(1989).-Cope & Hosni, A key to Egyptian grasses: 54 (1991).
Synonyms: Andropogon arundinaceus Wilid, Sp. P1. 4: 906 (1805); Rhaphis
arundinacea Desv., Opusc.: 69 (1831); Andropogon sorghum var. effusus Hack., in
DC., Monogr. Phan. 6: 503 (1889); Andropogon sorghum effusus Piper, in Proc. Biol.
Soc. Wash. 28: 35 (1915); Sorghum verticillflorum (Steud.) Stapf, in Pram, Flor.
Trop. Afr. IX: 114 (1917); Sorghum stapfii (Hook.f.) Fischer, in Gamble, F.P.I. 10:
1735 (1934); Sorghum pugionfolium Snowden, in J. Linn. Soc. (Bot.) 5: 240 (1955).
Short-lived perennial or an annual. Cuims 0.3-4m tall, often robust, branched; cuim
nodes mostly glabrous, sometimes pubescent. Leaf blades variable, often large, 575cm long, 5-7mm wide, flat, broadly lanceolate, glabrous on both surfaces, with a
prominent whitish midrib, tapering to a fine point; leaf sheaths glabrous, not keeled;
leaf ligule a fringed membrane, edged with a fringe of fine hairs and hairy on the
back.
Inflorescence a broadly spreading panicle, 10-60cm long; main inflorescence axis
angular, glabrous; primary branches divided, pubescent at the nodes; raceme
peduncles glabrous; racemes 2-7-jointed; rachis internodes and pedicels hairy. Sessile
spikelet (4-)7(-9)mm long, lanceolate to narrowly ovate; glumes coriaceous; lower
glume white pubescent, sometimes tomentose or fulvously pubescent, slightly
173
depressed longitudinally on the back; upper glume glabrescent or with sparse hairs on
the back. Florets 2: lower floret empty, lemma lanceolate, 5.5mm long, ciliate on the
margins; epaleate; upper floret bisexual, lemma deeply lobed, ciliate on the lobes and
margins, 3mm long; awn 20mm long, glabrous; palea present, 2mm long, ciliate on
margins and apex; lodicules 2, fleshy, glabrous; no anthers seen; caryopsis obovate,
3mm long. Pedicellate spikelet neuter, linear to lanceolate, 6.5mm long; glumes
papery; lower glume glabrous; upper slightly shorter than the lower, glabrous; lower
lemma glabrous, with a truncate apex; epaleate; pedicel hairy.
Diagnostic characters and affinities: This species is assumed to be the wild
progenitor of the grain sorghum, S. bicolor. The variability in the species can be
attributed to human selection of grain races, and the introgression with the wild
species.
Habitat: Swampy soils, streamsides and black clays; also found in disturbed places
and old farmland.
Type: Ghana, Isert (B).
Representative specimens:
Botswana: North-east District, Tati river, south-east of Francistown, altitude 980m,
04.iii.1985, Long 12242 (E); Northern Division, Chadum valley, South-west African
border, 16km west of Knau Knau, 14.iii.1965, Wild & Drummond 7024 (BM,
SRGH).
Malawi: Central region, Mkhota Kota, altitude 490m, 17.vi.1970, Brummitt 11519
(MAL); Dedza District, Masasa controlled area, 22.v.1989, Chikuni, Pate! &
Nachamba 79 (MAL); Tanga District, Palombe river, 06.v.1952, Jackson 802
(MAL); Kga Boma, 29.i.1943, Leech 12 (PRE); Salmia District, Lifidzi Breeding
Centre, 21.v.1985, Patel & Nachamba 2192 (MAL); Salima District, Chipoka,
25.v.1972, Sa!ubeni 1811 (MAL); South, Njanje District, Mwanalundu Hills,
11.1.1990, Sa!ubeni & Nachamba 5615 (MAL).
174
Mozambique: Baroma Province, Msusa, Zambesi valley, 25.vii.1950, Chase 2805
(BM); Namagoa, Mocuba, (no date), Faulkner 37 (PRE); Gaza Province, Chibute,
Miniquenique, Estacào Experimental do C.I.C.A., 13.vi.1960, Lemos & Balsinhas 98
(PRE); Tete, Cabora Bassa, altitude 215-220m, 09.v.1972, Pereira & Correia 2469
(WAG); Maputo Province, Umbeluzi State farm, near Boane, 19.ii.1985, Timberlake
3386 (SRGH).
Zambia: Mpika District, Mfuwe, 26.iv.1965, Mitchell 2683 (BM, SRGH); (no
location given), 04.i.1963, Rensburg 1161 (WAG).
Zimbabwe: Urungwe District, Mana pools floodplain, 31 .iii. 1981, Dunham 73
(SRGH); Salisbury (Harare), Crowborough Sewage farm, Mufakose township,
06.ii.1974, Lenton s.n. (WAG); Salisbury (Harare) District, Salisbury Research
Station, cultivated, 22.ii.1974, Simon 2386 (PRE).
Distribution: Throughout Africa, extending eastwards to Australia.
3. Sorghum versicolor* Anderss., in Peters, Reise Mossamb.: 563 (1863).-Stapf, F!.
Trop. Afr. IX: 138 (1917).-Sturgeon, A revised list of the grasses of Southern
Rhodesia Part IV: 6 (1954).-Chippindal, in Meredith (ed.), The grasses and pastures
of South Africa: 459 (1955).-Bogdan, A revised list of Kenya grasses: 55 (1958).Jackson & Wiehe, An annotated check-list of Nyasaland grasses: 60 (1958).-Napper,
Grasses of Tanganyika: 96 (1965).-Hood, A guide to the grasses of Zambia: 58
(1967).-Clayton, in Polhill (ed.), F!. Trop. East Aft., Gramineae (3): 729 (1982).Gibbs Russell et al, Grasses of Southern Africa: 302 (1990).
Synonyms: Andropogon serratus var. versicolor Hack., in DC., Monogr. Phan. 6:
522 (1889); Sorghum purpureo-sericeum (A. Rich.) Aschers. & Shweinf. var.
trinervatum Chiov., F!. Somal 2: 439 (1932).
*
versicolor, variously cloured or changing colour, referring to the differently coloured spikelets.
175
An annual of up to 2.5m tall. Culms slender, with a very conspicuous ring of long,
silky white hairs spreading round the culm from each node. Leaf blades hairy on both
surfaces, occasionally glabrous on the upper surface, 10-30cm long, up to 15mm
wide; leaf sheaths glabrous, bearded with long cilia at the mouth; ligule an unfringed
membrane, to 3nmi long.
Inflorescence an open panicle to 25cm long; branches simple; main inflorescence axis
glabrous, smooth, bearded at the nodes; raceme peduncles glabrous, smooth, with
prominent pulvini; racemes 3-7-jointed; rachis intemodes and pedicels long hairy.
Sessile spikelet elliptic-oblong, (5-)5.5(-7)mm long; lower glume coriaceous, glossy,
reddish brown to black, pilose on the back, bearded from the callus with pallid to
reddish hairs; upper glume also coriaceous and same colour as lower, tapers to a fine
apex, pilose above the middle, glabrous and shiny below. Florets 2: lower floret
empty, lemma 4mm long, ciliate on the margins; epaleate; upper floret bisexual,
lemma 3mm long, deeply lobed, ciliate on the lobes and margins, awned; awn (25)40mm long, twisted, scabrid along the spiral; lodicules 2, fleshy, densely ciliate from
the apical corners; anthers 3, 2.5mm long; ovary oblong. Pedicellate spikelet neuter,
smaller than sessile [(3-)4(-5)mm long], lanceolate, greenish; lower glume
membranous, hirsute on the back; upper glume also membranous, glabrous to scabrid,
enclosed in the lower; pedicel hairy.
Diagnostic characters and affinities: This species and S. purpureo-sericeum (A.
Rich.) Aschers. & Schweinf, are very similar that they may just represent a variation
in one species. Differences between them are mainly in their spikelet length (shorter
in S. versicolor), their apparent inability to hybridise (Garber, 1950), and their
differing geographical distributions.
Habitat: In deciduous bushland or wooded grassland, commonly on waterlogged
soils or black clays.
Type: Mozambique, Boror, Peters (whereabouts uncertain).
Representative specimens:
176
Botswana: Northern Division, 6km south-east of Tsau, 18.iii.1965, Wild &
Drummond 7132 (BM, SRGH).
Malawi: Northern Province, Nzimba District, 25 miles south of Rumphi on Ml,
altitude 4000ft, 04.v.1974, Pawek 8571 (PRE, WAG); Zomba District, near
Macheleni hi!!, Tembenu village, 05.iii.1978, Seyani & Patel 800 (MAL).
Mozambique: Malema, Mutuáli, EstacAo Experimental do Instituto do Algodào,
05.iv.1962, Lemos & Marrime 318 (PRE); Sul do Save, 26.xii.1953, Myre & de
Carvalho 1731 (SRGH); Tete, Cabora Bassa, altitude 230-330m, 02.v.1972, Pereira
& Correia 2328 (WAG); Laurenco Marques (Maputo), Macaéne-Magude, 22.i.1948,
Torre 7179 (BM).
Namibia: Grootfontein District, 20km east of Otavi, altitude 1460m, 08.iv.1987,
Long & Rae 712 (E).
Republic of South Africa: Transvaal: 200km south of Geysdorp, altitude 1300m,
22.i.1987, Smook 6283 (WAG); Pretoria, 09.ii.1938, Wit 20 (WAG).
Zambia: Kalomo District, Siatambo, 07.ii.1963, Mitchell 17/63 (BM); Mumbwa
District, Sala Reserve, Cheta river, 26.iii.1963, Vessey-Fitzgerald 4019 (SRGH).
Zimbabwe: Gwelo District, Base, Gwelo kop., altitude 470011, 05.iii.1967, Biegel
1959 (BM); Victoria District, Fort Victoria, Victoria Reserve Zone, 19.iii.1956,
Cleghorn 185 (BM); Melsetter District, Mutambara Tribal Trust Land on main road
near Linsnac!oon farm, altitude 1000m, 26.iii.1969, Crook 850 (PRE); Bubi District,
Gwampa Forest Reserve, ii.1956, Goldsmith 6/56 (PRE); Victoria District, Fort
Victoria area, 02.i.1949, Robinson 336 (SRGH); Melsetter District, Sabi Valley,
Nyanyadzi, 26.iii.1956, Whellan 1009 (WAG).
Distribution: Eastern, central and southern Africa.
4. Sorghum stipoideum (Ewart & White) C. A. Gardner & C. E. Hubbard, in Hook.
Ic. P1. 34: t 3364 (1938).-Lazarides, Hacker & Andrew, Taxonomy, Cytology and
177
Ecology of indigenous Australian sorghums, in Aust. Syst. Bot. 4: 591-635 (1991).Simon, A key to Australian grasses: 158 (1993).-Simon & Latz, A key to the grasses
of the Northern Territory, Australia: 49 (1994).
Synonyms: Sarga stupoidea Ewart & White, in Proc. Roy. Soc. Vict. n.s. XXIII: 297
(1911); Andropogon sargus Ewart, in Proc. Roy. Soc. Vict. XXV: 113 (1912);
Chrysopogon stipoideus (Ewart & White) Domin, in Biblioth. Bot XX: 271 (1915);
Sorghum Mjobergii Cheel, in Kung!. Svensk. Vet.-Akad. Hand!. n.s. LII(10): 3
(1916); Andropogon stipoideus (Ewart & White) C. A. Gardner, Enum. P1. Austri.
Occid.: 5 (1930).
Annual, 1-3m tall, erect. Culms stout, usually branched; culm nodes glabrous, or
pubescent to bearded. Leaf blades up to 30cm long, and 1.3cm wide, broadly linearlanceolate, glabrous or with sparse tubercle-based hairs along the midrib, with
cartilaginous and shortly spinulose margins; leaf sheaths glabrous, tightly clasping the
culm internodes, bearded and constricted at the collar region; ligule an unfringed
membrane, 0.7-3.7mm long.
Inflorescence an open panicle with simple branches; main inflorescence axis striate,
glabrous; raceme peduncles glabrous or occasionally scabrid-ciliate, with prominent
pulvini; racemes mostly 2-jointed (1-3-jointed); rachis intemodes and pedicels hairy.
Sessile spikelet elliptic, 10(-18)mm long (including callus); callus pungent, 3mm
long, slightly curved; lower glume hirsute on the back or partly glabrous and glossy,
narrowed upwards; upper glume with a sharp median keel, scabrid on the back.
Florets 2: lower floret empty, lemma lanceolate, ciliate on the margins, 5mm long;
upper floret bisexual, lemma deeply lobed, lobes and margins ciliate, about 3mm
long, awned from the sinus; awn 60mm(-155nim) long, twisted, glabrous, ciliate
along the column spiral, bristle scabrid-ciliate; lodicules 2, cuneate, ciliate on the
prolonged apical corners; anthers 2(-3), 3.5mm long; ovary minute, obovate.
Pedicellate spikelet male, 8mm(- 18mm) long, subulate; lower glume hirsute on the
back, narrowed upwards into a beak-like apex, ciliate on the margins, aristulate,
178
awniet 2.3nmi long; upper glume scabrid on the back, cuspidulate, with ciliate
margins; lower lemma lanceolate, ciliate on margins, 5mm long; upper lemma
broadly ovate, ciliate on margins, 3.5mm long; lodicules 2, cuneate, ciliate on the
corners; anthers 3, 4mm long; pedicels hairy.
Chromosome number: 2n = 10 (Garber, 1950; Pritchard & Gould, 1964; Hacker,
1991).
Diagnostic characters and affinities: The pungent, extremely long callus of the
sessile spikelet and the awned glumes of the pedicellate spikelet are especially
prominent features. Allied to S. intrans: the glabrous (not bearded) ovary and
caryopsis in S. stipoideum is the only consistent dissimilarity (Lazarides et al, 1991).
Habitat: Occurs widely in deep or shallow clays, sands and barns, in flat or hilly
country, on coastal dunes and beach sands, on the margins of permanent swamps and
lakes or seasonally flooded river terraces, floodouts and levees, and in association
with sandstone, limestone or laterite slopes, ridges or outcrops.
Type: Australia, Western Australia, Northern Province, Noonkanbah, ii. 1911,
Mjoberg 61 (S, K!).
Representative specimens:
Australia:
Northern Territory: on Stuart Highway, 12 miles south-east of Mataranka,
21.iii.1964, Lazarides 7092 (K); about 200 km south of Darwin on road to Pine
Creek. 13°38'S 131°18'E, 01.iv.1977, Pullen 10.580 (E).
Queensland: about 6 km south of Wyndham, on road to Kununurra, 159 FS
127°14'E, 25.iv.1977, Pullen 10.874 (E).
Western Australia: near Derby airport, 8 km south of Derby, 17022'S 123040'E,
18.iv.1985, Alpin et al 44 (K); 41 km east of Derby on Gibb river road at junction
with road to Meda Station, 03.v.1983, Fryxell & Craven 3925 (E); Yamerra Gap,
Napier Range, 17°20'S 124°49'E, 04.v.1983, Fryxell & Craven 3942 (E); 73 km west
179
of Wyndham, about 13 km south of Paradise Pool on Ernest river, north-eastern
Kimberly, 20.iii.1978, Lazarides 8638 (K).
Distribution: Australia: Western Australia and Northern Territory.
5. Sorghum nitidum* (Vahl) Pers., Syn. P1. I: 101 (1805).-Bor, Flora of Assam 5: 351
(1940).-Lazarides, Hacker & Andrew, Taxonomy, Cytology and Ecology of
indigenous Australian sorghums, in Aust. Syst. Bot. 4: 591-635 (1991).-Simon, A key
to Australian grasses: 158 (1993).
Synonyms: Holcus nitidus Vahl, Symb. Bot. II: 102 (1791); Anatherum nitidum
(Vahl) Spreng., Syst. Veg. I: 290 (1825); Andropogon nitidus (Vahl) Kunth, Rev.
Gram. I: 166 (1829); Sorghum tropicum Nees var. muticum Nees, in Hook. Kew
Journ. Bot. II: 99 (1850); Andropogon pedicellatus Steud., Syn. P1. Glum 1: 394
(1854); Andropogon serratus Thunb. var. nitidus Hack., in DC., Monogr. Phan. 6:
521 (1889); Sorghum serratum (Thunb.) 0. Kuntze var. nitidus (Vahi) Domin, in
Biblioth. Bot. XX: 270 (1915); Holcus fulvus R. Br. var. nitidus (Vahl) Honda, in
Tokyo Bot. Mag. XL: 101 (1926); Andropogon amboinicus (L.) Merrill var. nitidus
(Vahl) Backer, Handb. F!. Java 2: 99 (1928).
Perennial grass. Culms 1.2-1.5m tall, branched; culm nodes bearded with white hairs.
Leaf blades finely acuminate, 20-60cm long, up to 12mm wide, with a prominent
white midrib, glabrous or sometimes hirsute with tubercle-based hairs; leaf sheaths
hispid to hirsute with tubercle-based hairs or glabrous, sometimes ciliate on the
margins, bearded at the mouth with stiff hairs to 5.3mm long; ligule a ciliolate
(sometimes glabrous) membrane, 1.5-2mm long.
* nitidus, shining, polished, referring to the shining brown spikelets.
Inflorescence an elongate panicle; main inflorescence axis glabrous, angular; racemes
whorled; raceme peduncles glabrous, with prominent swollen bases (pulvini);
racemes 3-5-jointed; raceme joints hairy; rachis internodes and pedicels with
brownish hairs. Sessile spikelet 4mm long, oblanceolate; glumes crustaceous; lower
glume 4mm long, dorsally compressed, mostly hirsute with tawny to brown hairs on
the keels, polished and shiny; upper glume rounded on the back, 3.5mm long, hirsute
in the upper third, glabrous, smooth and shiny below. Florets 2: lower floret empty,
lemma 3.5mm long, margins inrolled clasping the base of the upper floret, ciliate;
epaleate; upper floret bisexual, lemma 1.3-3mm long, deeply lobed, ciliate on the
margins, awnless or awned; awn (when present) 20mm long, glabrous, column longer
than the bristle; lodicules 2, fleshy (thickened), cuneate, ciliate on corners; anthers 3,
2.5mm long; ovary oblong. Pedicellate spikelet male (well developed) or neuter
(rudimentary), 3.5mm long; glumes membranous or cartilaginous, mostly hairy;
lower glume 3.5mm long, with brownish hairs on the back and keels, glabrous near
the apex; upper glume 3.5mm long, with sparse brownish hairs, ciliate on the hyaline
margins; lower lemma 3.5mm long, ovate, with ciliate margins, finely 2-nerved;
upper lemma 3mm long, ovate with ciliate margins, finely single-nerved; anthers 3,
2mm long; lodicules 2, fleshy, ciliate on corners; pedicel 2.5mm long, hairy on the
margins.
Chromosome number(s): 2n =10 (Celarier, 1958); 20 (Garber, 1950; Celarier,
1958).
Diagnostic characters and affinities: Distinguished by its small, often black
spikelets, whorled simple panicle branches, 3-5-jointed racemes. The lemma of the
upper floret may be awned or awnless. The awned specimens are indicated by an
asterisk (*) in the specimens seen.
Habitat: Occurs on grey and black barns and sands, on hill slopes and in disturbed
habitats.
Type: Australia, Queensland, 1802-5, *Brown 6191 (E!, K!).
Representative specimens:
181
Australia:
Queensland: North Kennedy District, Ayr, 07.vii.1950, *Cowdry 10 (K); between
Ingham and Toobanna, altitude 35ft, 2611931, Hubbard & Winders 6908 (K);
Sarina, 19.i.1931, *Hubbard & Winders 6495 (K); Port Curtis District, 24°22'S
15105 O'E, 40km north-east from Lowmead on cleared pasture, 03 .iv. 1975, * Mcdonald
& Batianoff 1205 (K); Whitsunday Island, 10.vi.1934, * White 10199 (K).
Bhutan: Punakha District, between Chuzomsa and Samtengang, altitude 1300m,
02.x.1987, Wood 5899 (E).
China: Ichang, v.1888, Henry 4234 (K).
Hong Kong: Tai Mo Shan, 10.x.1969, *Hu 8155 (K).
India: North-west Himalaya, Kangra, Plampur, altitude 4000ft, 26.ix. 1896, Gammie
18746 (K); Perola, Tehri, altitude 4000ft, 24.x.1948, *Koelz 22236 (E).
Japan: Province Nagato, 31.x.1976, Furuse 11756 (K).
Nepal: South-west, Kauchaupur District, Royal Sukia Phanta Wildlife Reserve,
altitude 600ft, 01.x.1975, *Schaaf 10 (K).
Papua New Guinea: Eastern Highlands, near Miruma village, Upper Asaro Valley,
Goroka subdistrict, altitude 1650m, 20.vi.1956, Hoogland & Pullen 5406 (K);
Munumu village, District Central, subdistrict Port Moresby, altitude 500m,
18.ix.1973, Isles & Vinas LAE 29077 (E).
Philippines: Tadaya ( Mt. Apo), District of Davao, Island of Mindanao, vi.1909,
Elmer 11025 (E); Batanes Islands, v. - vi. 1907, Fenix 3704 (K); Province of Cagayan,
Luzon, ii. 1912, Ramos s.n. (E).
Thailand: Hui Taleng, Korat, 21.xii.1928, *put 2178 (K).
Distribution: Australia and South-east Asia in the countries of India, Burma, Sri
Lanka, Thailand, China and Japan.
182
6. Sorghum plumosum (R. Br.) Beauv., Agrost.: 132, 165, 178 (1812).-Lazarides,
The grasses of Central Australia: 226 (1970).-Lazarides, Hacker & Andrew,
Taxonomy, Cytology and Ecology of indigenous Australian sorghums, in Aust. Syst.
Bot. 4: 591-635 (1991).-Simon, A key to Australian grasses: 158 (1993).-Simon &
Latz, A key to the grasses of the Northern Territory, Australia: 49 (1994).
Synonyms: Holcus plumosus R. Br., Prodr.: 200 (1810); Andropogon australis
Spreng., Syst. Veg. I: 287 (1825); Andropogon australis Spreng. subsp. plumosus (R.
Br.) Hack. var. genuinus Hack., in DC., Monogr. Phan 6: 523 (1889); Sorghum
plumosum (R. Br.) Beauv. var. typicum Domin, Journ. Linn. Soc. Bot. XLI: 275
(1912); Sorghum plumosum vars. robustissimum Domin et piligerum Domin, in
Biblioth. Bot. XX: 271 (1915); Andropogon plumosus (R. Br.) Backer, Handb. F!.
Java 2: 100 (1928).
Densely tufted perennial. Culms to 6mm wide, 4-5 noded; culm nodes mostly
bearded, occasionally glabrous. Leaf blades to 60cm long and 4-12mm wide, flat,
sometimes loosely folded or terete, long-acuminate, hirsute on the lower surface and
mostly glabrous (sometimes sparsely hairy) on the upper surface; lower leaf sheaths
hirsute with simple and tubercle-based hairs 3-5mm long, upper glabrous, bearded at
the mouth; ligule a 1.5-3mm long, ciliate membrane.
Inflorescence a panicle; main inflorescence axis terete, glabrous and smooth; raceme
peduncles scabrous-hispid, pilose in the axils, with prominent pulvini; racemes 2-4(16)-jointed; rachis internodes and pedicels long hairy. Sessile spikelet 8mm long
(including callus), elliptic; callus 1.3-2.5mm long, shortly pungent, bearded; lower
glume elliptic, cartilaginous, hirsute on the back, brown and shiny, sharply and
shortly ciliate on the keels towards the apex, apex hyaline and truncate; upper glume
smooth and shiny, boat-shaped, ciliate on the inturned margins towards the apex.
Florets 2: lower floret empty, lemma lanceolate, ciliate; upper floret bisexual, lemma
bibbed, ciliate, finely 3-nerved, awned; awn 60 (-100)mm long, column twisted, with
forward facing sharp cilia along the spiral, bristle hispidulous; palea present, small,
183
ciliate; lodicules 2, cuneate, with 2 to 3 cilia on the apex corners; anthers not seen;
ovary oblong, 2.8mm long, flattened on both surfaces. Pedicellate spikelet neuter,
8.5mm long, long-acuminate; lower glume coriaceous, hairy on the back; upper
glume also coriaceous, hairy on the back; lower lemma like that of the sessile
spikelet; upper lemma lanceolate, with ciliate margins; pedicel hairy on margins,
6mm long, those of the terminal spikelets subequal.
Chromosome number(s): 2n = 10 (Hacker, 1991); 20 (Garber, 1950); 30 (Pritchard
& Gould, 1964).
Habitat: Occurs in a wide range of habitats including rocky hills and outcrops,
coastal dunes, plains, swamps, seasonally flooded levees and other low-lying ground.
Type: Australia, Queensland, Burke District, Gulf of Carpentaria, Allen Island,
19.xi.1802, Brown 6192 (BM! K!).
Representative specimens:
Australia:
Northern Territory: 14 miles north-west of Ooratippra Station, 12.v.1955, Lazarides
5272 (K).
Queensland: North Kennedy District, Townsville, on exposed rocky slopes of Castle
Hill, 22.iii.1935, Blake 8165 (K) and Pentland, 02.iv.1935, Blake 8393 (K); Forest
Home Station, Gilbert river, iii. 1931, Brass 1836 (K); Bowen, on upper slopes of hill
north of town, 16.i.1931, altitude 200-300ft, Hubbard & Winders 6547 (BM);
Warrigal, on Great Dividing Range, altitude 1400-1500ft, 02.ii.1931, Hubbard &
Winders 7154 (BM); Chudleigh Park Station, 110 miles north of Hughenden, altitude
2700ft, 14.ii.1931, Hubbard& Winders 7654 (BM).
Indonesia: Sumba, lesser Sunda Island, between Waingapu and Melolo, narrow
coastal plain, 10.vi.1950, Monod 1993 (K).
Distribution: Distributed in Western Australia, the Northern Territory and
Queensland. Also a single specimen seen from the Indonesian island of Sumba.
184
8.2.6. Pseudosorghum A. Camus, in Bull. Mus. Hist. Nat. Paris 26: 662 (1920). The
type species is Pseudosorghumfasciculare (Roxb.) A. Camus.
Decumbent annual. Inflorescence a small dense panicle, with subdivided but very
short primary branches, bearing narrow racemes; rachis intemodes ciliate. Sessile
spikelet callus obtuse; lower glume firmly cartilaginous, glabrous, many-nerved, with
narrowly inflexed margins; upper lemma bibbed, awned; palea present; lodicules
glabrous. Pedicellate spikelet male or neuter, slightly smaller than the sessile; glumes
thin.
Species 2. No material was seen for Pseudosorghum zollingeri.
Pseudosorghum fasciculare (Roxb.) A. Camus, in Bull. Mus. Hist. Nat. Paris 26:
662 (1920).-Bor, Flora of Assam 5: 353 (1940).-Bor, Grasses Burma Ceylon India &
Pakistan: 205 (1960).
Synonyms: Andropogon fascicularis Roxb., Fl. Ind. 1: 269 (1820); Andropogon
gangenticus Hack., in DC., Monogr. Phan. 6: 539 (1889); Andropogon tonkinensis
Balansa in Morot., Journ. de Bot. 4: 112 (1890); Andropogon nitidulus Hook.f., F!.
Brit. Ind. 7: 199 (1896); Sorghum fasciculare (Roxb.) Haines, Bot. Bihar & Orissa:
1034 (1924); Sorghum gangenticus (Hack.) Stapf ex Haines, Bot. Bihar & Orissa:
1034 (1924).
Annual. Culms 60-150cm tall, decumbent at the base and finally erect, rooting from
the decumbent lower nodes, branching at the base, slender, rounded in section,
glabrous, simple or bearing flowering branches. Leaf blades linear-lanceolate, 3050cm long, 4-8mm wide, contracted at the base, acuminate-setaceous at the tip, flat,
rigid, glaucescent, scaberulous, midnerve somewhat thick; leaf sheaths compressed,
tight or somewhat lax, throat hairy; ligule 5mm long, ovate, an unfringed membrane.
185
Inflorescence paniculate, composed of few to many simple or branched racemes with
slender articulate rachis. Sessile spikelet falling with adjacent pedicel and rachis
internode; lower glume with a narrow, truncate apex, and narrowly inflexed margins,
flat, many-nerved; upper glume boat-shaped, keeled with a narrowed apex, 7-nerved.
Florets 2: lower floret empty, lemma slightly shorter, ovate-oblong, with incurved
margins, thinly membranous, 2-nerved, palea absent; upper floret bisexual, lemma
well developed, thinly membranous, long ciliate, bibbed about half way, with stout,
geniculate awn from sinus; palea present; lodicules 2, fleshy; stamens 3, 3mm long;
ovary ovate, tiny. Pedicellate spikelet male or neuter, slender and slightly smaller
than the sessile one; glumes thin.
Chromosome number: 2n =20 (Raman, Chandrasekharan & Krishnaswami, 1959).
Habitat: Fairly common in grassy plains.
Type: India, Chauda Division, Central Province Forestry Dep., vii. 1887, Duthie
10717 (K!).
Representative specimens:
Burma: Bassein District, Thabyu Chaung, 13.xii.1912, Kermode 7350 (K);
Myitkyina District, Kadu, 08.xi.1925, Parkinson 286 (K); South Lamayi Research
Station, 09.xii.1939, Thein Lwin 101 (K).
India: Northern, Debra Dun, x.1890, Duthie 10721 (K); Khuri, Surguja State,
15.x.1947, Koelz 19323 (K); Pierie, 24.xi.1939, Lahiri s.n. (K); Balajhar, Tabher
State, Orissa, altitude 450fi, 06.xii.1940, Mooney 1631 (K).
Philippines: Angat, Province of Bulaeau Luzon, xii.1914, Ramos s. n. (K).
Thailand: Northern Thailand, Province Tak, Lahn Sahng, 27.xii. 1974, Geesnik,
Hiepko & Phengklai 7926 (K); (without exact location), altitude 11 OOft, 03 .xii. 1911,
Kerr 2267 (K, 3 sheets); Loie, Phu Krading, Sam Kokkwak, altitude 700m,
09.xi.1954, Smitinand 2096 (K).
Distribution: Widespread in India, Burma, Philippines and most of South-east Asia.
8.2.7. Chrysopogon Trin., Fund. Agrost.: 187 (1820); nom. conserv. The type species
is Chrysopogon gryllus (L.) Trin.
Synonyms: Rhaphis Lour., F!. Cochin.: 552 (1790); Pollinia Sprend., P1. Pugill. 2: 10
(1815); Centrophorum Trin., Fund. Agrost.: 106 (1822); Trianthium Desv., Opusc.:
69 (1831); Chalcelytrum Lunell, in Amer. Midi. Nat. 4: 212 (1915).
Tufted perennials; ligule a short membrane or a line of hairs. Inflorescence a panicle,
its primary branches whorled, simple, each bearing a raceme reduced to a triplet of
spikelets (rarely racemes with 2 sessile spikelets, e.g. C. sylvaticus); spikelets in
triplets, or in triplets and pairs; the triplet consists of a sessile spikelet and two
pedicellate spikelets; pedicels linear filiform, never longitudinally grooved. Sessile
spikelet laterally compressed, its callus elongated, acute to pungent; glumes subequal,
unequally awned or aristate or awnless; lower glume cartilaginous to coriaceous,
often spinulose on the keels and muricate on the back; upper lemma entire or bibbed,
awned; awn glabrous to pubescent and usually prominent. Pedicellate spikelet male or
neuter; glumes awned or aristate or awnless.
Species 26.
A key to taxa sampled:
Pedicels glabrous:
Cuim internodes terete:
Base of plant creeping, rhizomatous; racemes single-jointed; sessile
spikelet 3-4.5mm long, with a very long (as long as spikelet) callus;
1. aciculatus
upper lemma entire.
Cuims erect, not creeping at base; racemes 2-3-jointed; sessile spikelet
6.5-8mm long; upper lemma bibbed:
187
Racemes consistently 2-jointed; sessile spikelet 8mm long;
lower glume of sessile spikelet smooth; palea 2.6mm long.
2. sylvaticus
Racemes consistently 3-jointed; sessile spikelet 6.5mm long;
lower glume of sessile spikelet echinulate along the lateral
keels; palea 3.5mm long.
gryllus subsp. echinulatus
Culm internodes channelled on one side:
Callus of sessile spikelet 1.5-3.5mm long; awn not overtopping the
pedicellate spikelet:
Leaves much less than 1cm wide; lower glume of sessile
spikelet 5-7-nerved; callus straight or nearly so; lower leaf
3. fallax
sheaths disintegrating into fibres.
Leaves 1-2cm wide; lower glume of sessile spikelet 4-5nerved; callus more or less curved; lower leaf sheaths not
4. latjfolius
disintegrating into fibres.
Callus of sessile spikelet 4-6mm long; column of awn usually
5. pallidus
overtopping the pedicellate spikelet.
Pedicels hairy:
Pedicels half the length of the sessile spikelet or longer:
Sessile spikelet 8-9mm long; awn 60mm long.
6. orientalis
Sessile spikelet 5.5-6mm long; awned or awnless:
Raceme peduncles bearded; sessile spikelet epaleate; both
glumes of the pedicellate spikelet awned.
7. aucheri
Raceme peduncles glabrous; sessile spikelet paleate; upper
glume of pedicellate spikelet awnless.
8. gryllus subsp. gryllus
188
Pedicels shorter than half the length of the sessile spikelet:
Awns of the glumes of the pedicellate spikelet plumose; upper lemma
9. plumulosus
entire.
Awns of the glumes of the pedicellate spikelet glabrous or if plumose
then only at the extreme base; upper lemma bibbed:
Sessile spikelet 5mm long; upper glume of sessile spikelet
keeled, pectinate-ciliate in the lower 3/4 with long, golden
brown rigid hairs.
10.fulvus
Sessile spikelet 6mm long; upper glume of sessile spikelet
rounded on the lower 3/4, keeled and ciliate in upper ¼ only.
11. serrulalus
1. Chrysopogon aciculatus* (Retz.) Trin., Fund. Agrost.: 188 (1820).-Bor, Grasses
Burma Ceylon India & Pakistan: 115 (1960).-Clayton, in Hepper (ed.), Flor. West
Trop. Aft., Gramineae ed.2 (3): 468 (1972).-Lowe, The flora of Nigeria, Grasses: 227
(1989).
Synonyms: Andropogon aciculatus Retz., Obs. Bot. 5: 22 (1789); Rhaphis trivalvis
Lour., Fl. Cochinch.: 553 (1790); Andropogon acicularis Retz. ex Roem. et Schult.,
Syst. Verz. 2: 812 (1817); Centrophorum chinense Trin., Fund Agrost.: 106 (1820);
Rhaphis acicularis (Retz.) Desv., Opusc.: 69 (1831); Chrysopogon trivialis Am. et
Nees, in Nov. Act. Nat. Cur. 19(1): 171 (1843); Andropogon trivialis Steud., Syn. P1.
Glum. 1: 396 (1854).
Rhizomatous creeping perennial. Cuims up to 50cm tall, prostrate at the base, rooting
from the crowded nodes; the upper 2-3 internodes erect and elongated, the one
*
aciculatus, marked with very fine irregular streaks, as if produced by the point of a needle.
bearing the inflorescence the longest. Leaf blades densely crowded in the lower part,
distant in the upper erect part; leaf blades rounded at the base, obtuse or subacute at
the tips, the uppermost short or rudimentary, the others 2-15cm long, 3-5mm wide,
frequently undulate near the margins, glabrous or sparingly pilose at the base; leaf
sheaths terete, tight, smooth, glabrous or sparsely bearded at the mouth or more or
less ciliate on the margins, much longer than the internodes in the lower decumbent
part and shorter in the upper erect part; ligule minutely ciliolate, appearing as a scar.
Inflorescence a prominently exserted panicle of 5-10cm long, said to be open in the
living state but contracted when dry; common axis angular, striate, consists of whorls
of 4-9 racemes, hispid on the upper half, angular; racemes single-jointed. Sessile
spikelet 6-10mm long (including callus), linear, acuminate from about the middle,
purplish; callus acicular, very acute, 3-6mm long, bearded with rusty hairs; lower
glume chartaceous, opaque, awnless, 4mm long, narrowly subtruncate, 2-keeled, the
keels remotely spinulose, the spines somewhat thickened at the base; upper glume
purplish, shortly awned, awnlet 1.5mm long (never exceeds 3mm), acuminate, keel
spinulose on the upper half. Florets 2: lower floret empty, lemma hyaline,
oblanceolate, 2.7mm long; upper floret bisexual, lemma oblong or linear, entire, 2mm
long, glabrous, drawn into a yellowish, untwisted, imperfect awn 5mm long; palea
short, 1.5-2mm long, linear-obtuse, nerveless; lodicules 2, fleshy, glabrous; anthers 3,
1mm long; ovary 2-2.5mm long. Pedicellate spikelet male, 4-6mm long, glabrous,
purple (rarely pallid), subulate-lanceolate; lower glume 5-7-nerved, cuspidate; upper
glume elliptic-lanceolate; lemmas hyaline, ciliate, the lower 3-4.5mm long,
oblanceolate, acute, the upper 2-4mm long, oblanceolate acute; palea 1-1.5mm long,
linear, glabrous; anthers 3, 2mm long; pedicels glabrous.
Chromosome number: 2n = 20 (Larsen, 1963).
Diagnostic characters and affinities: This species has a long (3-6mm), pungent
callus, which is said to be a menace to grazing animals (Bor, 1960).
Habitat: Common in open places.
Uses: Lawns and or playfields (temperate climates).
190
Type: Type locality, India.
Representative specimens:
Cameroon: Victoria (no exact location), 06.iv.1961, Ogu 304 (E).
China: Kochow District (no exact location), 14.v.1929, Tsiang Ying 2281 (E).
Hawaii: Kauai Island (no exact location), 01.vii.1895, Heller 2476 (E).
Nepal: Arum Valley, altitude 5000ft, 27.v.1956, Stainton 463 (E); Agam, altitude
3000ft, 22.vi.1954, Stainton, Sykes & Williams 5862 (E).
Papua New Guinea: Moroba District, Lae, altitude Om, 06ix. 1966, Coode 29519
(E).
Philippines: Butuan SubProvince, Mindanao, iii - v.1911, Weber 1047 (B).
Singapore: Mae Ritchie Reservoir, 04.xii.1948, Sinclair 5364 (E).
Thailand: Kucham, TakBai, altitude Om, 18.ix.1987, Niyomdham & Sriboonma 1630
(E); Krungtep, 12.iii.1958, Sorensen, Larsen & Hansen 2060 (E).
Distribution: Widely distributed in the tropics of Asia, Polynesia and Australia.
2. Chrysopogon sylvaticus* C.E. Hubbard, in Hook. Ic. Plant. 0365 (1938).-Blake,
in Univ. Queensl. Papers 2(3): 5 (1944).-Wheeler, Jacobs & Norton, Grasses of New
South Wales: 144 (1982).-Simon, A key to Australian grasses ed.2: 85 (1993).
A tufted perennial. Culms erect, suberect or oblique, 30-120cm tall, branched from
the middle nodes, 3-5 noded; culm nodes exposed, glabrous. Leaf blades narrowly
linear, to 50cm long, 6mm wide, mostly conduplicate, rarely flat, mostly basal,
margins spinulose-ciliate; leaf sheaths glabrous, keeled, the basal ones flabellate,
*
sylvaticus, growing in the woods; pertaining to the woods.
191
imbricate and compressed, persistent, the upper ones shorter than the internodes;
ligule minutely ciliolate.
Inflorescence a loose panicle, spreading and drooping, deep reddish purple; main
inflorescence axis slender, scabrous, few noded; raceme peduncles verticillate,
hispidulous; racemes consistently 2-jointed, but occasionally 1-jointed (very rarely 3jointed). Sessile spikelet lanceolate, 8mm long (including the callus); callus acute,
bearded with fulvous hairs 3mm long; lower glume narrowly truncate, spinulose on
the lateral keels, scabrid above the middle, smooth below, purplish; upper glume
lanceolate-oblong, spinulose on the keel, running into an awn 3mm long. Florets 2:
lower floret empty, lemma oblong, ciliate on the upper margins, 5mm long; upper
floret bisexual, lemma broadly linear, shortly bibbed, 5mm long, hardened in the
upper half, awned; awn geniculate, twisted, 23mm long, scabrid; palea oblong,
truncate, 2.6mm long; lodicules 2, fleshy; anthers 3, 2.5-3mm long; ovary oblong.
Pedicellate spikelet very reduced, variable in length (3-7mm long), neuter, linear,
muticous, represented by an empty glume; pedicel 5.5mm long, glabrous, flattened.
Diagnostic characters and affinities: This species differs from C. pallidus and its
allies by its flabellate leaf sheaths, the lower ones of which are persistent and do not
break up into fibres; the panicle is, on the whole, much looser, the racemes are
usually 2-jointed, and the pedicellate spikelets are constantly muticous.
Habitat: A characteristic member of the grass flora of Eucalyptus forests of southeast Queensland, particularly on somewhat stony hillsides.
Type: Australia, Queensland, near Moggill, 03.iv.193 1, Hubbard 8587 (K!).
Representative specimens:
Australia:
Queensland: Moreton District, Petrie, 13.ii.1931, Blake 148 (K); Port Curtis District,
mid and upper slopes of Mt. Berserker, 06.iii.1937, Blake 12728 (K); Mt. Edwards
O1.iv.1954, Everist 375 (K); Mt. Petrie, near Brisbane, 13.iv.1930, Hubbard 2156
(K); Moreton District, Enoggera, iii. 1916, White & Bick s. n (K).
192
Distribution: Australia.
3. Chrysopogonfallax* S. T. Blake, in J. M. Black, F!. S. Austral. ed.2: 60 (1943).Burbibge, Australian grasses 3: 44 (1970).-Lazarides, The grasses of Central
Australia: 90 (1970).-Wheeler, Jacobs & Norton, Grasses of New South Wales: 144
(1982).-Simon, A key to Australian grasses ed.2: 85 (1993).-Simon & Latz, A key to
the grasses of the Northern Territory, Australia: 24 (1994).
Tufted erect perennial up to 2m tall, emergent above the foliage. Cuims erect, slender
or moderately robust, somewhat compressed and channelled on one side, with deeply
embedded rootstock; culm nodes glabrous; uppermost cuim internode terete,
sometimes slightly pubescent. Leaf blades narrowly linear, tapering to a very acute
point, 5-45cm long and 4-5mm wide, the upper side smooth, glabrous, the lower side
variably smooth and glabrous or sparsely or densely hairy, with the hairs in all cases
arising from tubercles; leaf sheaths tight, not keeled, with the basal ones somewhat
persistent, at length more or less disintegrating into fibres, the upper ones shorter than
the internodes; ligule ciliolate.
Inflorescence a linear or linear-lanceolate panicle, contracted; main inflorescence axis
scabrous or slightly pubescent; nodes pilose; racemes up to 16 in each whorl; racemes
1-jointed with a spikelet triad of 1 sessile spikelet and 2 pedicellate spikelets; raceme
peduncles scabrous, fihiform. Sessile spikelet narrowly lanceolate, 11mm. long
(including the callus); callus 1.7mm long, bearded with fulvous hairs, straight,
narrowly conical; lower glume 8mm long, cartilaginous, convex on the back, 2-keeled
in the upper 1/3, tip truncate, awnless, more or less smooth in the lower part' lateral
keels muricate and spinulose towards the tip; upper glume narrowly obtuse, sharply
keeled, the keel pectinate and running into a slender scabrous awn 11mm long.
*
fallax, deceptive, fallacious.
193
Florets 2: lower floret represented by a lemma, lemma 6mm long, oblanceolate,
ciliate on margins; upper floret bisexual, lemma oblong-linear, shortly bidentate
(bibbed), 6mm long, membranous, thickened upwards, awned; awn geniculate,
40mm long, column tawny, scabrous, bristle paler, scabrous, straight; palea linear,
nerveless, 3.5mm long, sparsely ciliate on the upper part; lodicules 2, fleshy, cuneate;
no anthers seen; ovary narrowly ovate, minute. Pedicellate spikelet male; lower
glume lanceolate, gradually acute, 9mm long, aculeolate on the lateral keels, scabrous
elsewhere, awned; awn oblique, 6.5mm long, straight; callus minute, smooth; upper
glume lanceolate, aristulate, lower lemma elliptic-linear, 7.5mm long, ciliate on the
margins; upper lemma linear-lanceolate, 6.5mm long, acute; palea lobed, 4mm long,
sparsely ciliate at the tip; anthers 3, 3-5mm long; pedicels subequal, flattened, the
longer 4.5mm long, smooth, slightly broadened near the tip.
Diagnostic characters and affinities: In North-western Queensland it is known as
"woolly-butt" by reason of the fibrous remains of the basal leaf sheaths usually being
prominent (Blake, 1944). It is very similar to C. pallidus from which it may be
distinguished by the rather shorter spikelets, the sessile with a distinctly shorter callus
and shorter more slender awns, the column of the latter not usually overtopping the
pedicellate spikelets.
Habitat: Widely spread in a variety of habitats on both light and heavy soils.
Type: Australia, Queensland, Leichhardt District, Emerald, in open sandy places,
altitude 600ft, 18.iii.1935, Blake 8108 (MEL).
Representative specimens:
Australia:
Northern Territory: about halfway between Anthony Lagoon and Brunnette Downs,
altitude 232m, 16.v.1947, Blake 17805 (K); 'Manbulloo', Katherine, 01.iv.1946,
Hartely 11711(K); near Wavehill Police Station, 25.vi.1949, Perry 2246 (K).
194
Queensland: North Kennedy District, slopes of Great Dividing Range, west of
Alpha, altitude 1 200ft, 22.ii. 1931, Hubbard 7856 (K); Cook District, Chillangoe,
altitude 11 54ft, 22.i. 1931, Hubbard & Winders 6790 (K).
Western Australia: 14 km west of Forest Creek near Ord Hill, Ord River Station,
east Kimberly, 13.iv.1977, Pullen 10.736 (E).
Distribution: Australia: Queensland, Northern Territory and Western Australia.
4. Chrysopogon lat?folius* S.T. Blake, in Univ. Queensl. Papers, Dept. Biol. II No.3:
7 (1944).-Simon, A key to Australian grasses ed.2: 85 (1993).-Simon & Latz, A key
to the grasses of the Northern Territory, Australia: 24 (1994).
Tall, erect, tussocky perennial 2m tall, with a fibrous rootstock. Culms strongly
branched from the upper nodes, somewhat robust, channelled on one side; cuim nodes
exposed, glabrous. Leaf blades up to 60cm long (except for the uppermost which are
short), very broad, 0.7-1.3 (- 2)cm wide, barely distinct from the sheaths, linear, flat
or somewhat conduplicate with many primary nerves, margins with regularly spaced
forward facing spines, midrib prominent, glabrous or softly pubescent; leaf sheaths
prominently striate but not keeled, lower at length gaping, others rather tight, the
lower not splitting into fibres; ligule ciliate with hairs up to 0.5mm long.
Inflorescence a linear panicle, 20-25cm long, rather contracted; main inflorescence
axis stout, pubescent, terete, many-noded, with up to 30 racemes per whorl; raceme
peduncles filiform, straight, hispid; racemes single-jointed. Sessile spikelet linearlanceolate, 7.7mm long (including the callus); callus 2mm long, curved, bearded with
rusty hairs up to 2mm long; lower glume narrowly truncate, incurved, cartilaginous
with very narrowly hyaline margins and apex, smooth on the lower part, the rest
*
latifolius, with broad leaves.
195
muricate or mammillate, becoming spinulose on the keels towards the apex, awnless;
upper glume smooth in the lower part, the rest scabrid, with the keel becoming
aculeolate towards the apex, sparsely ciliate on the hyaline margins, awned, awn
9mm long, scabrid. Florets 2: lower floret empty, lemma obovate, emarginate, ciliate,
4.5mm long; upper floret bisexual, lemma broadly linear, entire, the back
membranous in the upper half with the remainder hyaline, margins glabrous, awned;
awn geniculate, twisted, 40mm long; palea lanceolate, emarginate; anthers not seen;
ovary oblong. Pedicellate spikelet male, lanceolate, 8mm long; lower glume
lanceolate, papery, scabrous, spinulose on the nerves terminating into an awn 6.5mm
long; upper glume as long, acute, awnless or aristate, smooth; lower lemma linearelliptic, ciliate, 6.5mm long; upper lemma narrowly elliptic, ciliate, 5.5mm long;
palea linear, ciliate, 3.2mm long; anthers (only disintegrating remains seen); pedicels
5mm long, scabrous.
Blake (1944) reports of shorter anthers (2.2-3.5mm) in the sessile spikelet than in the
pedicellate spikelet (3.5-5mm).
Chromosome number: 2n = 20 (Celarier, 1959).
Diagnostic characters and affinities: The species is very distinctive by reason of its
stoutness, its broad many-nerved leaves and the numerous panicle-branches (racemes)
to each whorl.
Habitat: Savannah woodland and Eucalyptus forests. Also found on roadsides and
rocky ledges.
Type: Australia, Northern Territory, without definite locality, July-August 1911,
Gilruth & Spencer (MEL).
Representative specimens:
Australia:
Northern Territory: Stuart Hwy, about 5 miles north-west of Katherine, 01 .iv. 1964,
Adams 926 (K); Katherine, altitude 11 Om, 21 .vi. 1946, Blake 16047 (K); 4 miles south
196
of Katherine on Wyndham road, 03.i.1965, Wilson 55 (K);
V2
mile north-west of
Edith River Siding, 30.i.1965, Wilson 226 (E).
Western Australia: 16 miles south-east of Mt. House Station, 22.iv.1955, Lazarides
5156 (K); Inglis Gap, Gibb river road, King Leopold Ranges, 09.iv.1988, Simon 3844
(K); Lennard river, viii. 1905, Staer s. n. (E).
Distribution: Australia.
5. Chrysopogon pallidus* (R. Br.) Trin. ex Steud., Nomenci. Bot. ed. 2(i): 360
(1840).-Blake, in Univ. Queensland Papers 2(3): 14 (1944).-Lazarides, The grasses of
Central Australia: 93 (1970).-Simon, A key to Australian grasses ed.2: 85 (1993).Simon & Latz, A key to the grasses of the Northern Territory, Australia: 24 (1994).
Synonyms: Holcus pallidus R. Br., Prodr.: 199 (1810); Polliniapallida (R. Br.) R. &
S., Syst. II: 829 (1817); Andropogon palidus (R. Br.) Kunth, Revis. Gram. i: 165
(1829); Chrysopogon gryllus (L.) Trin. var. pallidus (R. Br.) Benth., Fl. Austr. VII:
537 (1878); Andropogon gryllus (L.) Trin. subsp. pallidus (R. Br.) Hack., in DC.,
Monogr. Phan. 6: 552 (1889); Andropogon gryllus(L.) Trin. subsp. calcaratus Hack.,
in DC., Monogr. Phan. 6: 553 (1889); Chrysopogon gryllus (L.) Trin. subsp. pallidus
(R. Br.) Domin, in Biblioth. Bot. Y.X. Heft 85: 272 (1915).
Tussocky perennial with well developed stolons, to lm tall, erect, the base covered
with the fibrous remains of old leaf sheaths. Cuims sparingly branched, channelled on
one side, 3-5 noded. Leaf blades numerous, folded, up to 40cm long and 6mm wide
(when flattened), pubescent in the ligule area; leaf sheaths bearded at the mouth, those
of the basal leaves disintegrating into long persistent fibres; ligule densely ciliate with
hairs up to 0.7mm long.
*
pallidus, pale.
197
Inflorescence a panicle; main inflorescence axis bearded at the nodes; racemes singlejointed, verticillate, up to 15 per whorl; raceme peduncles capillary, up to 7cm long
(never more than 10cm long), scabrid, slightly thickened near the tip and there
obliquely truncate. Sessile spikelet lanceolate, pallid, 1 1mm long (including the
callus); callus very slender, curved, pungent, bearded with tawny hairs of 2.5mm
long; lower glume muricate on the lateral keels, becoming spinulose towards the
apex, tip muticous or sometimes the keel nerves running into awnlets, cartilaginous;
upper glume smooth, ciliate on the hyaline margins, carinate, scabrous on this keel,
awned, awn straight and 14mm long (up to 18mm long). Florets 2: lower floret
empty, lemma oblanceolate, 6.5mm long, ciliate on margins; upper floret bisexual,
lemma broadly linear, shortly bibbed, thickened in the upper half in the centre,
margins sparsely ciliate in the upper half, 6.3mm long, awned; awn 54mm long (up to
70 mm long), sharply geniculate, column densely scaberulous, about '/2 the length of
the awn, strongly twisted, dark brown, bristle paler, scabrid; palea linear, obtuse,
sparsely ciliate, 2.5mm long; lodicules 2, fleshy, opaque, cuneate and highly
vascularised; anthers not seen (just filament remains); ovary ovate to lanceolate.
Pedicellate spikelet male, lanceolate, 11-14mm long; lower glume 7-nerved, more or
less spinulose on the keels, papery, scabrid or glabrous elsewhere, awned, awn 9mm
long (up to 12mm long); upper glume as long as the lower, elliptic-lanceolate,
membranous, awnless but sometimes aristulate. Florets 2: lower floret empty, lemma
linear-elliptic, thinly membranous with hyaline margins, 2-nerved, ciliate to the base;
upper floret male, lemma linear, acute with broad hyaline ciliate margins; palea
linear, sparsely ciliate on the upper part; anthers 3, 5.5-5.8mm long; pedicels
glabrous, thickened upwards, strongly convex on the back, 6.5mm long (up to 7mm
long).
Blake (1944) reports that the anthers of the sessile spikelet are 3-5mm long, which
means that they are shorter than those of the pedicellate spikelet.
Diagnostic characters and affinities: Distinguished particularly by the long pungent
callus to the sessile spikelet and the long stout awn with the column overtopping the
198
pedicellate spikelet. The basal leaf sheaths readily disintegrate into a mass of fibres
(as in C. fallax). The fibres, as far as could be observed, seemed more numerous in
this species than in C. fallax.
Habitat: Found in association with other grasses in Eucalyptus forests.
Type: Australia, Queensland, Burke District, Islands of the Gulf of Carpentaria, 1802,
Brown 6190 (E!).
Representative specimens:
Australia:
Northern Territory: on Stuart Highway, 16 miles south-east of Katherine,
18.xii.1963, Lazarides 6997 (K).
Western Australia: 1.5km south-east of mining Campsite, Mitchell Plateau, North
Kimberly, 21.i.1982, Kenneally 7874 (K); 6km north of mining Campsite on track to
Port Warrender, Mitchell Plateau, North Kimberly, 05.xii.1982, Kenneally 8626/A
(K).
Distribution: Australia.
6. Chrysopogon orientalis* (Desv.) A. Camus, in Lecomte, Fl. Gen. de Indo-Chine 7:
332 (1922).-Bor, Flora of Assam 5: 359 (1940).-Bor, Grasses Burma Ceylon India &
Pakistan: 118 (1960).
Synonyms: Rhaphis orientalis Desv., Opusc.: 69 (1831); Andropogon wightianus
Nees ex Steud., Syn. P1. Glum. 1: 395 (1854); Andropogon aristulatus Hochst. ex
Steud., Syn. P1. Glum. 1:397 (1854); Andropogon breviaristatus Steud., Syn. P1.
Glum. 1: 396 (1854); Chrysopogon wightianus (Nees ex Steud.) Thw., Enum P1.
Zeyl.: 366 (1864).
*
orientalis, eastern.
199
Perennial. Cuims 50-90cm tall, caespitose and erect, slender, branching above. Leaf
blades collected mostly at the base, narrow, linear, 3-10cm long, 3-4mm wide, flat or
rolled, glabrous; leaf sheaths carinate, smooth or scabrid, tight, shorter than the
internodes; ligule a rim of short hairs.
Inflorescence a panicle, with 4-12 branches/peduncles per whorl, somewhat thickened
at the top. Sessile spikelet linear, laterally compressed, 8-9mm long; lower glume
glabrous, 8-8.5mm long, flat towards the base, compressed above and hispid,
coriaceous, margins glabrous and involute; upper glume 8-9mm long, clasping at the
base, margins hyaline, coriaceous, awned, awn 12-15mm long, covered with stiff
spreading hairs, the rest of the body scabrid. Florets 2: lower floret empty, lemma
hyaline, linear-obtuse, ciliate; upper floret bisexual, lemma entire, 8-9mm long,
terminates into a subulate awn 5-6cm long; lodicules 2, small, glabrous; anthers 3,
3mm long; ovary oblanceolate. Pedicellate spikelet male, linear-lanceolate, 10-12mm
long, pale violet, dorsally compressed; lower glume with 7 equidistant nerves, keel
somewhat ciliate, awned, awn about the same length as glume or longer, hispid; upper
glume shortly aristate, ciliate on the margins. Florets 2: lower floret empty, lemma
linear-obtuse, hyaline, ciliate, 2-nerved; upper floret male, lemma hyaline, shorter
than the upper glume, ciliate, 1-nerved; anthers 3, 3.5mm long; pedicel 7-11.5mm
long, with rigid reddish cilia on margins.
Chromosome number: 2n = 20 (Gould & Soderstrom, 1974).
Diagnostic characters and affinities: The vegetative shoots have a characteristic
appearance due to the equitant arrangement of the leaf sheaths.
Habitat: Common grass on sandy soil, meadow and roadsides.
Type: India, Mont. Nilagiri, Hohenacker 1285 (K!).
Representative specimens:
China: Hainan, 18-30/vii-1932, Fung 20258 (E).
India: Madras (no exact location), xi. 1899, Bourne 3124 (K); Madras, altitude 2000m
(no exact location), vii. 1889, Gamble 21442 (K).
200
Singapore: Kampong Padang, 04.vii.1953, Sinclair & Salleh 7530 (E).
Sri-Lanka: Northern Province, Jaffna District, Pallai, altitude 4m, 14.i.1970, Clayton
5250 (K); Irat, 18.iii.1927, Gaiston 1441 (K); about 28 miles east of Kandy on
Malinyangana road near Madugoda, altitude 800m, 12.iv.1970, Gould 13383 (K).
Thailand: Kok Kradunk Nuu, TakBai, altitude Om, 02.ix.1988, Niyomdham &
Ueachirakan 1925 (K).
Distribution: Indian subcontinent.
7. Chrysopogon aucheri (Boiss.) Stapf, in Kew Bull. 1907: 211 (1907).-Bor, Grasses
Burma Ceylon India & Pakistan: 116 (1960).-Clayton, in Hepper (ed.), Flora West
Tropical Africa, Gramineae ed.2 (3): 468 (1968).
Synonyms: Andropogon aucheri Boiss., Diagn. ser. 1(5): 77 (1844); Chrysopogon
ciliolatus var. aucheri Boiss., F!. Or. 5: 458 (1884); Andropogon aucheri var.
genuinus Hack., in DC., Monogr. Phan. 6: 560 (1889).
Perennial tussock forming grass. Culms 10-60cm tall, only the flowering culms
issuing from the dense tufts; culm nodes glabrous or rarely with short velvety hairs.
Leaf blades with tubercle-based hairs, densely puberulous (very rarely glabrous),
linear-lanceolate, 1.5-3mm wide, up to 10cm long, equitantly arranged giving the
plant a distinct appearance; leaf sheaths also densely puberulous and hairy at their
mouths; ligule a ciliolate rim.
Inflorescence a terminal panicle; inflorescence axis with up to 10 joints, with each
joint supporting a whorl of a few to many racemes; racemes single-jointed; raceme
peduncles fttivously or pallidly bearded. Sessile spikelet laterally compressed; lower
glume glabrous except for the pilose apex, 5.5-6mm long, awnless; upper glume,
6.5mm long, with flimsy inner ciliate margins and a pilose apex, awned, awn
puberulous and 7.5mm long. Florets 2: lower floret empty, lemma hyaline, elliptic
201
and entire; upper floret bisexual, lemma oblong and bidentate (bibbed), with ciliate
margins near place of attachment to the lower ovary region, with a minutely
hispidulous awn 2.3cm long; bodicules 2, fleshy, with a truncate apex and cuneate
base; anthers 3, 3.5mm long; ovary ovate. Pedicellate spikelet dorsally compressed,
male; lower glume coriaceous, glabrous with a pilose apex, 7-nerved, ovate, awned,
awn obscurely hispid, 5mm long; upper glume shortly awned, 3-nerved, glabrous
with ciliolate margins. Florets 2: lower floret empty, lemma ovate, ciliolate; upper
floret male, lemma entire, awnless; anthers 3, 4mm long (i.e. slightly longer than
those of the bisexual flower of the sessile spikelet); pedicels unilaterally long
plumose, 2mm long.
Chromosome number(s): 2n = 20 (Faruqi, Quiraish & Halai, 1979); 40 (Celarier,
1959).
Diagnostic characters and affinities: Some specimens seen having culm nodes with
short velvety hairs, a character not at all associated with Chrysopogon.
Habitat: Common on roadsides and rocky slopes.
Type: Type not seen. Type locality unknown.
Representative specimens:
Iran: (no location given), altitude 400m, 31.iii.1974, Davis & Bokhari 56217 (E);
Kerman Province, altitude 1400-1700m, 0l.v.1975, Parris 75.583 (E); Baluchestan,
altitude 200m, 08.iii.1977, Runemark, Assdi & Sardabi 22497 (E).
Pakistan: Baluchistan, Quetta, altitude 1500-1600m, 14.v.1965, Lammond 1234 (E);
Baluchistan, Kalat, 10.iv.1965, Lammond 333 (E); Baluchistan, Quetta, 21.viii.1942,
Sinclair 2722 (E).
Distribution: Arabia, Tropical Africa, and the Indian subcontinent.
8. Chrysopogon gryllus (L.) Trin., Fund. Agrost.: 188 (1820).-Bor, Grasses Burma
Ceylon India & Pakistan: 117 (1960).
202
Synonyms: Andropogon gryllus L., Cent. P1. 2: 33 (1756); Holcus gryllus (L.) R. Br.,
Prodr.: 199 (1810); Pollinia gryllus (L.) Spreng., Pugill. 2: 10 (1815); Apluda gryllus
(L.) Presi, Cyp. et Gram. Sci.: 55 (1820); Chrysopogon glabratus Trin., in Mem.
Acad. Sci. Petersb. ser. 6(2): 318 (1832); Andropogon glabratus (Trin.) Steud., Syn.
P1. Glum. 1: 395 (1854); Andropogon voyleanus Steud., Syn. P1. Glum. 1: 397
(1854).
C. gryllus subsp. gryllus:
A large coarse tufted perennial grass of up to 2m tall. Cuims somewhat robust, a
number of them coming out of the same clump with a common root system; cuim
nodes glabrous, exposed. Leaf blades up to 30cm long, up to 6mm wide, linear or
linear-lanceolate; leaf hairiness variable, most with sparse tubercle-based hairs
(especially those of vegetative plants), more especially towards the leaf base and
ligule vicinity but others only with echinulate margins, and still others glabrous; leaf
sheaths striate, glabrous or with a few tubercle-based hairs towards the mouth; ligule
a short ciliolate rim.
Inflorescence a large open panicle, to 30cm long; raceme peduncles with fulvous tips,
capillary, glabrous with a tint of purple; racemes are single-jointed. Sessile spikelet
glumes very firm; lower glume glabrous, scabrid, echinulate on the keels, 6.5mm
long, awnless; upper glume 7.5mm long, awned, minutely hispid, glabrous and
glossy, margins ciliate. Florets 2: lower floret empty, lemma 5mm long, ovate, with
ciliate margins; upper floret bisexual, lemma bidentate (bibbed), 6mm long, oblong,
with a puberubous awn of 4cm long; palea present, 3.5mm long, entire; lodicules 2,
fleshy; anthers 3, 3.5mm long; ovary ovate, 3mm long. Pedicellate spikelet neuter;
glumes with a tinge of purple; lower glume 7mm long, ovate, shortly awned,
glabrous, hispid along the keels; upper glume 7mm long, boat-shaped, glabrous,
awnless, ciliate along the margins; pedicels 5.5mm long, fulvously hairy along one
margin, the rest of its body glabrous.
203
Chromosome number(s): 2n = 20 (Mehra, Subramanyam & Swaminanthan, 1962);
40 (Celarier, 1959).
Habitat: Found in dry open slopes.
Uses: A pasture plant up to the flowering stage.
Type: Described from southern Europe ("In Phaetia, Helvetia, Verone sauvages"
(Tsvelev, 1976)). Type not seen.
Representative specimens:
Afghanistan: Province Nangarhar, Torkham, altitude 700m, 30.iv.1969, Hedge,
Wendelbo & Ekberg W7448 (E).
Nepal: Mugu Karnali Valley, altitude 7000ft, 15.viii.1952, Polunin, Sykes &
Williams 5230 (E); Tumar Valley, altitude 6500ft, 02.viii.1956, Stainton 1194 (E);
Kusma, altitude 3000ft, 05.ix.1954, Stainton, Sykes & Williams 7047 (E).
Turkey: Usak, altitude 1000m, 13.vi.1965, Goode & Jones 2367 (E); Mardin,
altitude 400m, 10.v.1966, Davis 42730 (E); Nigole/Katseri, altitude 1200-1300m,
22.vi.1952, Davis 19146 (E); Province Tunceli, Pertek-Hozat, altitude 1100m,
13.vii.1957, Davis & Hedge D31015 (E).
Distribution: Arabia, Temperate Himalaya.
C. giyllus subsp. echinulatus*:
Inflorescence smaller and narrower than that of subsp. gryllus, brownish purple in
colour; racemes consistently 3-jointed. Sessile spikelet: lower glume 6.5mm long,
awnless, echinulate along the keels; upper glume 6.5mm long, awned, base of
awn/apex of glume echinulate. Florets 2: lower floret empty, lemma 5mm long,
ciliate, palea entire, 3.5mm long; upper floret bisexual, lemma bidentate (bibbed),
*
echinulatus, echinulate, with very small prickles.
204
ciliate towards the apex, 3-nerved, elliptic; awn 2cm long, glabrous or hispid; anthers
3, 3.5mm long; caryopsis oblong-obovate. Pedicellate spikelet male or neuter; lower
glume glabrous, awned, purplish, hispid at the apex and along the keels; lemmas (i.e.
both upper and lower) same as lower lemma of sessile spikelet; upper glume awnless,
with involute ciliate margins; anthers 3, 3.5mm long; pedicels glabrous, gradually
thickened upwards, laterally flattened.
Type: Northwest India, Royle 226 (LIV).
Representative specimens:
India: Bashahr State (no exact location), 09.vii.1890, Lace 376 (E); Punjab, Baspa
Valley, altitude 12 500ft, 12.vii.1939, Ludlow & Sherrff 7424 (E).
Pakistan: Harara District, altitude 2000ft, 18.ix.1958, Burt & Kazmi B1288 (E);
Kashmir, Sonamarg, altitude 8500ft, 02.ix. 1956, Polunin 56/641(E).
Distribution: Northwest Himalaya.
9. Chrysopogon plumulosus Hochst., in Jahreshefte Wurtt. 3: 62 (1847).-Clayton &
Renvoize, in Polhill (ed.), Flor. Trop. East Mr., Gramineae (3): 737 (1982).Chaudhary, Grasses of Saudi Arabia: 413 (1989).-Cope & Hosni, A key to Egyptian
grasses: 54 (1989).
Synonyms: Aristida chrysopila Steud., Nom. Bot., ed. 2, 1: 131 (1840); Chrysopogon
quinqueplumis A. Rich., Tent. F!. Abyss. 2: 450 (1851); Andropogon aristidoides
Steud., Syn. P1. Glum. 1: 397 (1854); Andropogon quinqueplumis (A. Rich.) Steud.,
Syn. P1. Glum. 1: 398 (1854); Andropogon aucheri Boiss. var. quinqueplumis (A.
Rich.) Hack., in DC., Monogr. Phan. 6: 561 (1889); Chrysopogon aucheri (Boiss.)
Stapf var. quinqueplumis (A. Rich.) Stapf, in Kew Bull. 1907: 211 (1907);
Chrysopogon aucheri (Boiss.) Stapf var. pulvinatus Stapf, in Pram, Flora Tropical
Africa IX: 161 (1917).
205
Perennial with wiry cuims, 10-90 cm tall; cuim nodes exposed and glabrous. Leaf
blades cauline, sometimes forming cushions at the bottom, 1-15cm long, 1-3mm
wide, linear-lanceolate, glabrous or sometimes with sparse tubercle-based hairs
underneath (on adaxial surface) and around the ligule area, bluntly acute and
becoming acuminate; leaf sheaths glabrous, very tight, the uppermost clasping the
culm, ending in a small flag leaf; leaf ligule a short ciliolate rim.
Inflorescence a terminal panicle of about 3-7cm long, with delicate capillary branches
with a fulvous beard at the tip; the inflorescence axis slender, jointed and with whorls
of 3-to-many pedunculate racemes coming off at the joints. Sessile spikelet narrowly
oblong; lower glume laterally compressed to a rounded keel, 6mm long, slightly
pilose at the tip; upper glume sparsely ciliate on the keel, bears a plumose awn 12mm
long. Florets 2: lower floret empty, lemma hyaline and ciliate towards the base; upper
floret bisexual, lemma entire, bears a puberulous awn 3cm long; ovary ovate;
lodicules 2, fleshy; callus short and blunt, 1.2mm long. Pedicellate spikelet 6.5mm
long, pallid or purplish, male or neuter, both the upper and lower glumes with
plumose awns; upper glume awn is 6mm long and lower glume awn is 10mm long.
The texture of the glumes of the pedicellate spikelet is different from those of the
sessile spikelet. Those of the sessile spikelet are tough and shiny while those of the
pedicellate spikelet are papery; pedicels taper downward, unilaterally villous and
about half the length of the sessile spikelets.
Chromosome number: 2n = 40 (Fedorov, 1974).
Diagnostic characters and affinities: This species is characterised by the long
plumose awns of the glumes.
Habitat: Found in fixed sand dunes, dry rocky slopes and dry wadi beds.
Type: Saudi Arabia, Jedda to Mecca, Schimper (whereabouts uncertain; based on
same gathering as Aristida chrysopila Steud.).
Representative specimens:
206
Oman: Salalah, altitude lOm, 15.x.1984, Mcleish 139 (E); Dhofar, altitude 810m,
22.ix.1985, Miller 7754 (E).
Saudi Arabia: Dawadimi Camp, altitude 2900ft, 02.iii.1983, Collenette 4050 (E);
Ulayyah, altitude 570m, 17.iv.1982, Podzorski 969 (B).
Somalia: Upper Sheekh, altitude 1500m, 09.x.1978, Allen & Elmi 454 (E);
Shabeellaha Dhexe, altitude 50m, 30.v.1989, Thulin & Dahir 6740 (E).
Yemen Arab Republic: Marib, altitude 1300m, 04.iv. 1981, Miller & Long 3418 (E).
Distribution: Arabia, Tropical East Africa.
10. Chrysopogon fulvus* (Spreng.) Chiov., Fl. Somala. 1: 327 (1 929).-Bor, Grasses
Burma Ceylon India & Pakistan: 116 (1960).
Synonyms: Pollinia fulva Spreng,. Pugill. 2: 10 (1815); Andropogon monticola
Roem. et Schult., Syst. Verz. 2: 665 (1827); Andropogon sprengelii Kunth, Rev.
Gram. 1: 166 (1829); Chrysopogon monticola (Roem et Schult.) Haines, in Indian
For. 40: 495 (1914); Andropogon montanus Koen. ex Trin., in Spreng., Neue
Entdeck. 2: 93 (1921) Chrysopogon montanus Trin. ex Spreng., in Spreng., Neue
Entdeck. 2: 93 (1921).
Perennial grass. Culms densely tufted, erect and branching, 5 0-70cm tall. Leaf blades
linear-acuminate, glabrous, 4mm wide; leaf sheaths keeled, glabrous, lower ones
overlapping; ligule a hairy rim.
Inflorescence consists of racemes with or filiform peduncles; main inflorescence axis
with a whorl of racemes; racemes single-jointed (rarely 2-jointed), with fulvously
*
fulvus, tawny, dull yellow with a mixture of grey and brown, yellowish-brown, this is the
predominant colour of the inflorescence parts.
207
villous tips; the spikelets purplish in colour and this together with the yellowishbrown colour of the peduncles give the inflorescence a distinct facies. Sessile spikelet
5nmi long, laterally compressed; lower glume linear, hispid towards the apex on the
keels, awnless; upper glume pectinately ciliate (fulvus) on the keel, 3-nerved, awned,
awn obscurely hispidulous, 8.5mm long. Florets 2: lower floret empty, lemma hyaline
with ciliate margins; upper floret bisexual, epaleate, lemma bidentate (bibbed),
hyaline, with an awn 2cm long, glabrous; callus short and blunt, 0.5mm long, shortly
bearded; lodicules 2, fleshy; anthers 3, 3mm long; ovary ovate. Pedicellate spikelet
male or neuter, both glumes awned; lower glume ovate, sparsely hairy on the keel,
awn 6mm long, obscurely hispid, lemma oblong-ovate, hyaline, ciliate on margins;
upper glume hyaline, glabrous, awn 5.5mm long, obscurely hispid; pedicel 2mm
long, unilaterally villous.
Chromosome number(s): 2n = 20, 80 (Mehra & Gill, 1968).
Habitat: Shrubby or grassy hill slopes.
Uses: Cultivated fodder.
Type: Type locality, India.
Representative specimens:
India: (no location given), ix. 1920, Anderson s. n. (E); Khundura District (no exact
location), 10.xii.1888, Duthie 8489 (E); Tehri, altitude 400ft, 14.iv.1948, Koelz
21766 (E); Bashahr State, Taughin Dogn - Upper Kauawar, altitude 6500ft,
26.ix.1890, Lace 628 (E); Bombay Presidency, Belguam, 28.xi.1944, Sinclair 3781
(E); Nagarjuna Kouda Valley, altitude 300ft, 14.vii.1961, Thothathrii 9666 (E).
Distribution: Tropics of Asia and East Africa.
11. Chrysopogon serrulatus* Trin., in Mem. Acad. Sci. Petersb. Ser. 6(2): 318
(1832).-Bor, Grasses Burma Ceylon India & Pakistan: 118 (1960).-Clayton &
Renvoize, in Polhill (ed.), Flor. Trop. East Africa: 736 (1982).-Gibbs Russell et al,
Grasses of Southern Africa: 85 (1991).
Synonyms: Andropogon trinii Steud., Syn. P1. Glum. 1: 395. (1854); Andropogon
ciliolatus Steud., Syn. P1. Glum. 1: 396 (1854); Andropogon caerulens Steud., Syn.
P1. Glum. 1: 395 (1854); Chrysopogon ciliolatus (Steud.) Boiss., F!. Orient. 5: 458
(1884); Andropogon trinii Steud. var. increscens Hack., in DC., Monogr. Phan. 6:
558 (1889); Andropogon monticola Schult. var. trinii Hook.f., F!. Brit. Ind. 7: 193
(1896); Chrysopogon montanus Trin. var. serrulatus (Trin.) Stapf, in Pram, F!. Trop.
Afr. IX: 160 (1917); Chrysopogon montanus Inn var. tremulus (Hack.) Stapf, in
Pram, Flora of Tropical Africa IX: 160 (1917); Chrysopogon fulvus (Spreng.) Chiov.
var. tremulus (Hack.) Chiov., Rac. Bot. Miss. Consol. Kenya: 325 (1935);
Chrysopogon fulvus (Spreng.) Chiov. var. serrulatus (Trin.) R. B. Stewart, in
Brittonia 5: 446 (1945).
Perennial. Cuims up to 150cm tall, sometimes robust. Leaf blades glabrous except in
the ligule vicinity where there are sparse tubercle-based hairs on the margins on both
surfaces, linear-lanceolate, 2-10mm wide; leaf sheath glabrous, rarely hairy at the top;
leaf ligule a ciliolate rim.
Inflorescence paniculate, up to 15cm long; consists of racemes with peduncles
fulvously bearded at the tips. Sessile spikelet narrowly oblong, 6mm long; lower
glume awnless, slightly hispid at the tip, 6mm long; upper glume with few white hairs
on the back, with a glabrous awn about 12mm long, obscurely hispidulous. Florets 2:
lower floret empty, lemma hyaline, single-nerved with sparsely ciliolate margins;
upper floret bisexual, lemma hyaline, bibbed (bidentate), awn 2.5cm long, slightly
*
serrulatus, finely serrate, i.e., saw edged with sharp teeth pointing forwards.
209
puberulous, ovary about 1mm long, completely enclosed by 2 fleshy lodicules,
anthers 3, 3mm long. Pedicellate spikelet male or neuter, glumes awned; lower glume
6mm long, with white short hairs on the back, awn 5mm long, hispidulous; upper
glume glabrous, 6.5mm long, awn 2.5mm long, obscurely hispidulous, lemma
hyaline, sparsely ciliolate on the margins; anthers 3, 3mm long; pedicel 2.5mm long,
fulvously villous unilaterally.
Diagnostic characters and affinities: This species is very similar to C. plumulosus
except that it has a longer callus (up to 1.5mm long), longer upper glume of sessile
spikelet (4-7mm long) and the glabrous or obscurely hispidulous (not plumose) awns
of the glumes of both the sessile and pedicellate spikelets. Also the upper lemma apex
is bibbed in C. serrulatus and entire in C. plumulosus.
Habitat: Common in shallow sandy soils and rocky landscape. Also in dry
overgrazed soils in deciduous bushland.
Type: Nepal, Wallich (LE).
Representative specimens:
Botswana: Gaborone, Aedume Park, altitude 1050m, 09.xi.1977, Hansen 3276 (K).
Zimbabwe: Nuanetsi District, Sengwe Tribal Trust Land, 15.xi.1973, Cleghorn 2910
(K); Gwanda District, Tuli Pasture Research Sub-Station, altitude 2500ft, 5.i.1955,
Oates 1742 (E); Urungwe District, altitude 1700ft, 16.x.1957, Phipps 797 (K);
Nyamandhlovu District (no exact location), 21.ix.1956, Plowes 1880 (K).
Distribution: Tropical Africa, Northwest India.
8.2.8. Vetiveria Bory in Lem.-Lisanc., in Bull. Sci. Soc. Philom: 42 (1822). Type
species is Vetiveria zizanioides (L.) Nash.
Synonyms: Lernomandia Steud., in Flora 33: 229 (1850); Mandelorna Steud., Syn.
Pt. Glum. 1: 359 (1854).
210
Tufted perennials with tough rhizomes; lower leaf sheaths much compressed,
flabellate-imbricate; conduplicate leaf blades which do not bend away from the cuim
at the collar region but just a lateral indentation marks the end of the sheath and the
beginning of the blade; leaf sheaths flattened. Inflorescence a panicle, its primary
branches whorled, simple or unbranched, each bearing a raceme; racemes (2-)3 to
many-jointed, pedunculate; spikelets paired, compressed laterally, differing in sex,
one sessile, the other pedicellate; rachis internodes and pedicels slender and gradually
thickened upwards. Sessile spikelet laterally compressed, its callus obtuse to pungent,
often large and conical; lower glume papery to coriaceous, muticus, spinulose on the
keels and muricate or mammilate on the back; upper glume aristate; upper lemma
entire or bibbed, awned or awnless. Pedicellate spikelet male or neuter.
Species 10.
A key to taxa sampled:
Sessile spikelet awnless.
1. zizanioides
Sessile spikelet awned:
Lower glume of sessile spikelet 8mm long.
2.fi lip es
Lower glume of sessile spikelet 5-7mm long:
Racemes 2 to 3-jointed; leaf blades long-attenuate. 3. elongata
Racemes many-jointed:
Lower glume of sessile spikelet 5mm long, smooth; upper
lemma awn 10mm long.
4.fulvibarbis
Lower glume of sessile spikelet 7mm long, muricate on the
back; upper lemma awn 5mm long.
5. nigritana
1. Vetiveria zizanioides (L.) Nash, in Small, F!. Southeast U. S.: 67 (1903).-Stapf, in
Kew Bull. 1906: 346-349, 362 (1906), & in Pram, Flor. Trop. Africa IX: 157 (1917).Sturgeon, A revised list of the grasses of Southern Rhodesia Part IV: 8 (1954).-
211
Chippindall, in Meredith (ed.), The grasses and pastures of South Africa: 470 (1955).Jackson & Wiehe, An annotated check-list of Nyasaland grasses: 70 (1958).-Bor,
Grasses Burma Ceylon India & Pakistan: 258 (1960).-Napper, Grasses of
Tanganyika: 99 (1965).-Clayton, in Hepper (ed.), Flor. West Trop. Africa ed.2 (3),
Gramineae: 470 (1972).-Lowe, The flora of Nigeria, Grasses: 291 (1989).
Synonyms: Phalaris zizanioides L., Mant. P1. 2: 183 (1771); Andropogon muricatus
Retz., Obs. Bot. 3: 43 (1783); Agrostis verticillata Lamk., Ecyci. Meth. Bot. 1: 59
(1783); Anatherum muricatum (Retz.) P. Beauv., Ess. Agrost.: 150 (1812); Vetiveria
odoratissima Lem.-Lisanc., in Bull. Soc. Philom. (Paris): 43 (1822); Vetiveria
odorata Vivey, J. Pharm. 13: 501 (1827); Andropogon festuco ides J. S. Presi ex C. B.
Presl, Rel. Haenk. 1: 340 (1830); Vetiveria muricata (Retz.) Griseb., Fl. Brit. West
Ind.: 560 (1864); Vetiveria arundinacea Griseb., F!. Brit. West. Ind.: 559 (1864);
Sorghum zizanioides (L.) 0. Ktze., Rev. Gen. P1. 2: 791 (1891); Andropogon
zizanioides (L.) Urban, Symb. Antill. 4: 79 (1903); Holcus zizanioides (L.) 0. Ktze.
ex Stuck., in Aim. Mus. Nac. Buenos Aires 11: 48 (1904); Anatherum zizanioides (L.)
Hitchc. et Chase, in U.S. Natl. Herb. Contrib. 18: 285 (1917).
Perennial grass of up to or greater than 2m tall; cuims erect, usually sheathed all
along. Leaf blades linear, acute, up to 90cm long, 4-20mm wide, erect, rigid, firm or
somewhat spongy, striate; leaf sheaths compressed, particularly the lower ones,
sharply keeled, fan shaped, imbricate, very smooth; ligule a scarious rim.
Inflorescence a long oblong panicle of up to and over 30cm long, contracted; the
main inflorescence axis stout, glabrous, smooth; whorls 6-10 with up to 20
pedunculate racemes; peduncles filiform, slightly scabrous; racemes very slender;
rachis internodes about the same length as spikelets or slightly longer, glabrous;
pedicels shorter than the sessile spikelets, glabrous. Sessile spikelet linear-lanceolate,
variable in colour (yellowish, olive, violet brown, purplish); callus obtuse, 1mm long
or less, glabrous; lower glume muricate on the back, spinulose on the keels; upper
glume spinulose on the keel. Florets 2: lower floret empty, lower lemma as long as
212
the glumes, acute, retrorsely ciliate; upper floret bisexual, lemma up to 3mm long,
narrow, oblong-lanceolate, mucronate, glabrous; anthers 3, 2-3mm long; ovary ovate,
3.5mm long. Pedicellate spikelet sparingly aculeolate or almost smooth, male; lower
lemma same as for sessile spikelet; upper lemma acute, entire; anthers 3, 2.5mm long;
pedicel glabrous.
Chromosome number: 2n = 20 (Celarier, 1959).
Diagnostic characters and affinities: The awnless spikelets give the inflorescence a
different facies from other species of Vetiveria. Raceme peduncles are much shorter,
making the inflorescence much contracted and compact. The lower glumes of both
the sessile and pedicellate spikelets are muricate, giving the inflorescence a 'prickly'
appearance. Spikelets much smaller, inflorescence characteristically long.
Habitat: Mostly in cultivated fields and basically in places associated with the
function it is intended for, e.g., it might be grown as edging to roads.
Uses: Principally used for thatching; not grazed except when young and tender. Its
root is the scented khas khas which is used for tattis or grass screens which are kept
wetted to cool rooms. Commercial oils are extracted from its roots and used in
perfumery.
Type: India, Koenig (LINN).
Representative specimens:
Ghana: Achimota, v.1931, Irvine 1638 (E).
India: Najafgarh, 29.vii.1956, Bilandi s.n. (E); Kissar, Autumn 1884, Coldstream s.n.
(E).
Malay Peninsula: near Kampong Sungei Kluang, Penang, 14.xi.1950, Sinclair 6668
(E).
Mauritius: Curepipe, altitude 1840ft, 27.iv.1889, Johnston s.n. (E).
Nepal: Butwal, altitude SOOft, 07.x.1954, Stainton, Sykes & Williams 8823 (E).
213
Pakistan: Shekhupung District, near Sadhoke, about 25 miles north of Lahore,
20.vii.1958, Burt Bi 166 (E).
Thailand: altitude 300m (no location given), 05.ix.191 1, Kerr 2007 (E); Maharat,
altitude'400m, 02.iii.1958, Sorensen, Larsen & Hansen 1820 (E).
Distribution: Southeast Asia to Tropical Africa; introduced to most parts of the
world.
2. Vetiveriafihipes (Benth.) C.E. Hubbard, in Kew Bull. 1934: 444 (1934).-Blake, in
Univ. Queensland Papers 2(3): 23 (1944).-Burbidge, Australian grasses 3: 46 (1970).Wheeler, Jacobs & Norton, Grasses of New South Wales: 268 (1982).-Simon, A key
to Australian grasses ed. 2: 175 (1993).-Simon & Latz, A key to the grasses of the
Northern Territory, Australia: 54 (1994).
Synonyms: Chrysopogon elongatus (R. Br.) Benth. var. fihipes Benth., Fl. Austral.
VII: 539 (1878); Andropogon elongatus (R. Br.) Spreng. var. JIlipes (Benth.) Hack.,
in DC., Monogr. Phan. VI: 565 (1889); Chrysopogon gryllus (L.) Trin. var. spicigera
Maid. & Betche, in Proc. Linn. Soc. N. S. W.: 15 (1916).
A tufted perennial. Culms relatively slender, 70-100cm tall, unbranched or sparingly
branched from the middle nodes, closely sheathed in the lower part. Leaf blades
rather obtuse, 15-45cm long, 3-6mm wide, with a scabrous adaxial surface, smooth
abaxial surface; leaf sheaths smooth.
Inflorescence a whorl of paniculate racemes, narrow and with a few (up to 6) joints;
racemes are longer than their peduncles, purple; peduncles filiform, puberulous near
the base, the rest of their length scabrid, with an obliquely truncate tip, slightly
thickened upwards; rachis internodes glabrous. Sessile spikelet grey or purple, linearlanceolate, narrowed from about the middle, 8mm long, acute or acuminate; callus
subulate-conical, about 2.5mm long with fulvous hairs on the sides; lower glume
coriaceous, with an indistinct median keel near the tip which is sometimes spinulose,
214
scabrid, muricate along the keel nerves from the middle towards the base, spinulose
along the keels towards the apex; upper glume mucronate, spinulose along the keel.
Florets 2: lower floret barren, lower lemma elliptic, ciliate along the margins, nearly
as long as the glumes; upper floret bisexual, lemma entire or shortly bibbed, elliptic
or becoming ovate, about same length as lower, awned; awn well developed, about
16mm long, slightly twisted but not geniculate, scabrous; anthers 3, 2.5mm long;
ovary elliptic, 4mm long. Pedicellate spikelet neuter, narrowly lanceolate, 5.5mm
long; lower glume smooth, lower lemma very much like that of the sessile spikelet
but much shorter; upper glume oblong-elliptic; upper lemma linear, sparsely ciliate in
the upper part, very short; pedicels rarely exceeding the top of the sessile spikelet,
flat, glabrous.
Chromosome number(s): 2n = 40 (Celarier, 1959).
Diagnostic characters and affinities: This species is less robust in form relative to
other species of Vetiveria. Blake (1944) mentions that the peduncles sometimes are
subdivided and the racemes coming from these single-jointed. The specimens
-
observed did not show this.
Habitat: Common on sandy river banks or alluvium soils.
Type: Australia, Queensland, Banks of the Baloon, 05.iv.1946, Michell s. n. (K!).
Representative specimens:
Australia:
Queensland: Cook District, Mareeba, altitude 1300ft, 18.vi.1935, Blake 9475 (E);
Cook District, near Cooktown, 14.v.1970, Blake 23230 (E); North Kennedy District,
Magnetic Island, near Townsville, altitude Om, 18J. 1931, Hubbard & Winders 6643
(E); Leichhardt District, Mackenzie river, 90 miles south-east of Nebo Township,
13.vii.1962, Story & Yapp 163 (E).
Distribution: Australia, mainly eastern Queensland.
215
3. Vetiveria elongata (R. Br.) Stapf ex C. E. Hubbard, in Kew Bull.: 444 (1934).Blake, in Univ. Queensland Papers 2(3): 19 (1944).-Simon, A key to Australian
grasses ed.2: 175 (1993).-Simon & Latz, A key to the grasses of the Northern
Territory, Australia: 54 (1994).
Synonyms: Holcus elongatus R. Br., Prodr.: 200 (1810); Sorghum elongatum Beauv.,
Agrost.: 178 (1812); Andropogon elongatus (R. Br.) Spreng., Syst. Veg. 1: 287
(1825).-F. Muell., Fragm. VII: 121 (1873), First Census: 132 (1882), Second Census:
222 (1889).-Hack, in DC., Monogr. Phan VI: 564 (1889).-Ewart & Davies, F!. N.
Territ.: 30 (1917); Chrysopogon elongatus Benth., F!. Austral. 7: 538 (1878).-F.M.
Bail., Synops. Queensland Fl.: 645 (1883), Catal. Queensl. P1.: 620 (1913); Rhaphis
elongata (R. Br.) Chase, in Contrib. U. S. Nat. Herb. XXIV: 205 (1925).
Perennial tussock forming grass, tussocks of up to im in diameter. Culms up to 2.5m
tall, smooth, sheathed nearly all along. Leaf blades long-attenuate, mostly
conduplicate for most of their length, up to 50cm long and 8mm wide, spinulose on
the margins and keel, abaxial surface glabrous and smooth, upper with scattered
rather short tubercle-based hairs.
Inflorescence a long (up to 30cm) and contracted panicle; consists of more than 10
whorls of pedunculate racemes; peduncles up to 4.5cm long, glabrous and smooth.
Racemes mostly 2-3-jointed, purplish to pallid; radius intemodes 3.2-4.2mm long,
glabrous; pedicels shorter, 2-3mm long, also glabrous. Sessile spikelet bisexual,
linear-lanceolate, 5-7mm long; callus narrow-cuneate, 2mm long, bearded on the
margins with white hairs up to 4mm long; lower glume muticus, coriaceous, 5.5mm
long, 2-keeled, spinulose along the keels; upper glume coriaceous, aristate, carinate,
the keel spinulose upwards. Florets 2: lower floret empty, lower lemma elliptic
oblong, slightly shorter than the glumes, ciliate; upper floret bisexual, lemma oblong,
bidentate (bibbed), awned; awn 7mm long, scabrid, weakly twisted; palea present,
small, hyaline; lodicules 2, fleshy; anthers 3, 4mm long; caryopsis oblong. Pedicellate
spikelet male, linear-lanceolate, 5.5mm long; lower glume with a median keel below
216
the tip in the middle of the two lateral ones, smooth; upper glume elliptic-obtuse;
both lemmas ciliate along the margins, the lower about the same length as the glumes,
elliptic-oblong, the upper narrowly oblong and slightly shorter; anthers 3, 4mm long;
pedicel glabrous.
Diagnostic characters and affinities: This species is easily distinguished by its tall
culms with narrow and dense inflorescences, and the long attenuate leaves, usually
permanently folded. Blake (1944) used the number of raceme joints as another
character separating this species from V. paucflora. Even though I did not see
specimens of V paucifiora, I found this character very unreliable because the V.
elongata specimens I observed had mostly 2-3-jointed racemes, a character associated
with V. paucflora.
Habitat: Prefers moist habitats such as beach ridges, ri'er and stream banks or sea
facing slopes.
Type: Australia, 1802-1805, Brown s.n. (K!).
Representative specimens:
Australia:
Northern Territory: Leila Creak and McArthur river, altitude 112-178m, 14.v.1947,
Blake 17766 (K); Yirrkala, 07.viii.1948, Specht 839 (K); 250km east-northeast of
Darwin, altitude 60m, 15.vi.1978, Story 8351 (K); 66 miles east of Borroloole,
05.vi.1967, Symon 5068 (K).
Queensland: Cook District, 7km east-southeast of Aurukun Mission, 05.xii.1981,
Clarkson 4112 (K).
Distribution: Australia.
217
4. Vetiveria fulvibarbis* (Trin.) Stapf, in Pram, F!. Trop. Afr. IX: 158 (1917).Clayton, in Hepper (ed.), Flor. West Trop. Africa, Gramineae ed.2 (3): 470 (1972).Lowe, The flora of Nigeria, Grasses: 289 (1989).
Synonyms: Andropogon verticillatus Schumach., in Schumach. & Thonn., Beskr.
Guin. P1.: 50 (1820); Andropogon fulvibarbis Trin., in Mem. Acad. Petrsb. 6mer
ser.II: 287 (1832).-Hack., in DC., Monogr. Phan. 6: 544 (1889).
A tufted perennial. Culms up to 2m tall, less robust, sparingly branched; most of the
middle and upper cuim nodes of the flowering culms exposed and glabrous. Leaves
folded, linear with compressed leaf bases, glabrous or sparsely hairy on the abaxial
surface in the ligule vicinity; leaf sheaths sharply keeled, separated from the leaf
blades by an indentation at the collar area; ligule a ciliolate rim.
Inflorescence consists of raceme whorls (6-8 nodes) with up to and over 12
pedunculate racemes per whorl (joint); raceme joints with fulvous hairs; rachis joints
oblique, minutely ciliolate; rachis internodes glabrous, flat on the sessile spikelet side
and rounded on the back, gradually thickened upwards; pedicels similar but shorter.
Sessile spikelet subulate-linear; lower glume smooth on the back except for the
rigidly ciliolate margins, 5mm long; upper glume sharply keeled and also rigidly
ciliolate along the keel, aristulate from the tip, the margins flexible and ciliate. Florets
2: lower floret empty, lower lemma oblong, about the same length as glumes, ciliate;
upper floret bisexual, lemma short, bibbed, awned; palea linear, glabrous; awn
slender, about 1cm long; ovary tiny, oblong; lodicules 2, fleshy, completely enclosing
the ovary; anthers 3, 3.5 mm long. Pedicellate spikelet male, as long as the sessile,
glumes smooth, papery; lower glume acutely acuminate, purplish, smooth, with a
short bristle; upper glume opaque, ciliate along margins; both lower lemma and upper
*
fulvibarbis, with yellowish-brown (fulvus), hairs (barbis), referring probably to the rachis joints.
218
lemma narrowly oblong, hyaline and ciliate along margins; palea present, ciliate on
margins; anthers 3, 3.5mm long.
Diagnostic characters and affinities: The perfect awns give it a distinct facies from
V. nigritana and V zizanioides. The glumes are papery and relatively smooth.
Habitat: Common in wet and swampy waterlogged grass plains.
Type: Type not seen. Type locality indicated as West Africa.
Representative specimens:
Chad: Méma, 300km north of Segou, x.1955, Wintrebert 9/1957 (K).
Ghana: Legon Hill, 05.x.1955, Adams 3343 (K); Accra Plains, 25.ii.1927, Daiziel
8429 (E); Accra Plains, 15 miles south of Akosombo, 28.xii.1969, Easterly 1287 (E);
Achimota, vi.1961, Irvine 4503 (E); on Akuse to Kpong road, 18.x.1953, Morton s.n.
(K); between Nyamalege and Fuu villages, Tamale-Salaga road, 02.viii.1963, Rose
Innes 31873 (K).
Distribution: Mainly a West African species.
5. Vetiveria nigritana Stapf, in Pram, F!. Trop. Afr. IX: 157 (1917).-Sturgeon, A
revised list of the grasses of Southern Rhodesia Part IV: 8 (1954).-Chippindall, in
Meredith (ed.), The grasses and pastures of South Africa: 469 (1955).-Napper,
Grasses of Tanganyika: 99 (1965).-Hood, A guide to the grasses of Zambia: 58
(1967).-Clayton, in Hepper (ed.), Flor. West Trop. Africa ed.2 (3), Gramineae: 470
(1972).-Clayton, in Polhill (ed.), Flor. Trop. East Africa, Gramineae (3): 739 (1982).Lowe, The flora of Nigeria, Grasses: 291 (1989).-Gibbs Russell et al, Grasses of
Southern Africa: 353 (1991).
Synonyms: Andropogon nigritanus Benth., in Hook., Niger Fl.: 573 ();Mandelorna
insignis Steud., Syn. P1. Glum. 1: 359 (1854); Andropogon squarrosus var. nigritanus
Hack., in DC., Monogr. Phan. 6: 544 (1889); Anatherum muricatum Rendle, in Cat.
Aft. P1. Welw. II: 153 (1899).
219
Tufted perennial. Cuims 1.5-3m tall, unbranched: upper cuim nodes exposed,
glabrous, lower ones hidden by leaf sheaths. Leaf blades narrow, up to 90cm long and
7mm wide, cauline; leaf sheaths sharply keeled; ligule a scarious rim.
Inflorescence a terminal panicle of whorled, pedunculate racemes; racemes manyjointed; the main inflorescence axis and raceme peduncles minutely ciliolate towards
the apex (upward facing cilia); raceme peduncle length reduces upwards, giving the
panicle a lanceolate appearance; whorls 8-10 and with up to 15 racemes; rachis
internodes and pedicels gradually thickened upwards. Sessile spikelet: lower glume
compressed laterally, leathery, spinulose on the back, narrowly linear-lanceolate,
7mm long; upper glume sharply keeled in the middle, spinulose along the keel, with
an inflexed margin, drawn to a short aristate tip, 7mm long. Florets 2: lower floret
represented by a hyaline lemma, lemma ovate; upper floret bisexual, lemma hyaline,
with a bibbed apex; awn 5mm long, slightly exserted from the glumes or enclosed,
bristle-like, glabrous on the lower half (column) and minutely toothed on the bristle;
lodicules 2, fleshy; no anthers seen; ovary oblong-lanceolate. Pedicellate spikelet
neuter, shorter than the sessile one; both glumes less coriaceous, and less spinulose;
lower glume 5mm long, sparingly aculeolate along the keel towards the apex or
almost smooth; upper glume smooth with flexible ciliate margins; lower lemma
hyaline with ciliate margins; pedicel solid, glabrous, thickened upwards.
Diagnostic characters and affinities: Very much like V zizanioides except for the
fact that it has longer spikelets and is awned. Awn bristle-like, enclosed or more or
less exserted from the sessile spikelet.
Habitat: Occurs mainly on river banks, occasionally on roadsides and forested areas.
Type: Nigeria, Nun R., Vogel (K!).
Representative specimens:
Botswana: Northwest District, Okavango river, Old Mohembo, 16km north of
Shakawe on Botswana border, 19.iii.1965, Wild & Drummond 7074 (BM).
220
Malawi: Port Herald, Tengam, 28.xi.1950, Jackson 310 (MAL); Khonjeni Estate,
Thyolo, altitude 2400ft, 16.xii.1974, Schwarz 14 (MAL).
Mozambique: Boroma Province, Ulere station, Zambezi river, 1 0.vii. 1956, Chase
2671 (BM); Chitengo, 13km to Acampamento, 02.v.1978, Diniz 163 (WAG).
Ghana: near Tamale, viii. 1964, Irvine 4552 (E).
Zambia: Mankoya District, near resthouse, Mankoya, 20.xi.1959, Drummond &
Cookson 6644 (E, BM); Kafue District (no exact location), 09.ii.1963, Van Rensberg
1359 (BM).
Zimbabwe: Gokwe District, Copper Queen N.R.A, near Morowa river, 21.xii.1963,
Bingham 901b (BM).
Distribution: Tropical Africa.
8.2.9. Pseudodichanthium Bor, in Indian For. 66: 271 (1940).-Bor, in Hook. Ic. P1.
t3598: 1-3 (1962). The type species is Pseudodichanthium serrafalcoides (Cooke et
Stapf) Bor.
Annual. Inflorescence a terminal solitary raceme; lowest 2-3 rachis internodes short
and glabrous, the upper ones longer and ciliate on the edges; spikelets much
imbricate, distichous; lowest 2-3 spikelet pairs homogamous. Sessile spikelets callus
obtuse; lower glume cartilaginous, broadly convex, the margins expanded into wings;
upper glume papery, shorter than the lower. Pedicellate spikelets larger than the
sessile
Species 1.
Pseudodichanthium serrafalcoides (Cooke et Stapf) Bor, in Indian For. 66: 272
(1940).-Bor, Grasses Burma Ceylon India & Pakistan: 204 (1960).-Bor, in Hook. Ic.
P1. t3598: 1-3 (1962).
221
Synonyms: Andropogon cookei Stapf ex Woodrow, in J. Bombay Nat. list. Soc. 13:
438 (1893); Andropogon serrafalcoides Cooke et Stapf Blatt. et McCann, in Kew
Bull. 1908: 450 (1908); Dichanthium serrafalcoides (Cooke et Stapf) Blatt. et
McCann, in J. Bombay Nat. Hist. Soc. 32: 426 (1928).
Annual grass. Culms slender, ascending from a slender base, 60cm tall; cuim nodes
glabrous. Leaf blades narrowly linear, up to 7.5mm wide, more or less hirsute;
sheaths glabrous; ligule an unfringed membrane, 1mm long.
Inflorescence a solitary raceme at the end of cuims and branches; raceme with 4-10
spikelet pairs; lowest 2-3 rachis internodes short, glabrous, the upper ones ciliate on
the edges; lowest 2-3 spikelet pairs homogamous, neuter: Spikelets much imbricate,
glabrous, pedicellate spikelet larger than the sessile. Sessile spikelet bisexual; lower
glume broadly winged, margins narrowly inflexed; upper glume papery, shorter than
the lower. Florets 2: lower floret empty, lemma oblong, obtuse, hyaline, epaleate;
upper floret bisexual, lemma stipiform; lodicules, 2, fleshy; anthers 3, 1.5mm long;
ovary obovate, tiny. Pedicellate spikelet similar to the sessile but longer, neuter;
lower glume with inflexed margins; lowest pedicels very short, glabrous, the
remainder longer, densely ciliate on one edge.
Diagnostic characters and affinities: The lower glumes of both the sessile and
pedicellate spikelets are broadly winged. Bor (1940) notes the following peculiarity
about this species: "Above the attachment of the leaf sheath is a broad band of tissue
consisting of large pith-like cells. In many grasses the cuims are swollen at the nodes,
but this swelling is not on the stem but at the base of the leaf sheath. In
Pseudodichanthium, however it is the base of the internode, above the attachment of
the leaf, which is enlarged. It is from the 'supranodal' cushions that the adventitious
roots are produced and it is here, presumably, that cell division takes place so that the
internode can increase in length. In contrast to the hard and woody shaft of the
internode, the supranodal cushion is soft and spongy, even in dried specimens".
Habitat: Occurs on limestone escarpments under dense monsoon forests.
222
Type: India, Western Ghats, Sakarpattar, near Lanauli, x.1898, Woodrow sm. (K!).
Representative specimens:
India: Khandala, near Echo point, x.1918, McCann 9403 (K); Bombay State,
Mahableshwar, x.1923, McCann s.n. (K); Bombay, Purandhar Fort top, 10.x.1950,
Santapau s. n. (K); Bombay, Merwar, x.1918, Talbot s. n. (K).
Oman: Dhofar, above Dhalqut, 16°43'N 53°14'E, altitude 790m, 04.xi.1991, Cope
504 (E, K); Dhofar, on the inland road to Dhalqut from Sarfait, 16043'05"N
53012'16"E, altitude 1100m, 02.x.1993, Cope 683 (K); Dhofar, near Tobruk Fort,
between Salalah and Taqah, altitude 600m, 28.ix.1993, Cope 673 (K); Dhofar, cliff
overlooking Dhalqut, 03.x.1993, McLeish 2870 (K).
Distribution: India and Oman.
8.2.10. Capillipediuin Stapf, in Pram, F!. Trop. Afr. IX: 169 (1917). The type species
is Capillipedium parvflorum (R. Br.) Stapf.
Synonym: Filipedium Raiz. & Jam, in J. Bombay Nat. Hist. Soc. 49: 682 (1951).
Annuals or perennials; mostly rambling. Inflorescence a delicate, loose panicle,
bearing short 1-5(-8)-jointed racemes at the ends of capillary primary and secondary
branches; rachis internodes and pedicels finely filiform, with a translucent
longitudinal groove. Sessile spikelet callus obtuse; lower glume cartilaginous, broadly
convex to slightly concave, acute or obtuse; upper lemma hyaline base of the awn,
entire; awn glabrous. Pedicellate spikelet male or neuter.
Species ± 14.
A key to taxa sampled:
Annual grass; culms weak, decumbent and trailing, markedly geniculately branched.
1. fihiculme
223
Perennial grasses; cuims robust, erect, simple or branched:
Cuim nodes glabrous; lower glume of sessile spikelet slightly depressed along
the middle:
Culms woody; lower glume of sessile spikelet glabrous; racemes 2-32. assimile
jointed.
Culms herbaceous; lower glume of sessile spikelet hairy; racemes 3-8jointed.
3. spicigerum
1
Cuim nodes bearded; lower glume of sessile spikelet not slightly depressed
along the middle:
Rachis internodes hairy; lower glume of sessile spikelet flat, glabrous;
4. huegelii
upper glume of sessile spikelet emarginate.
Rachis internodes glabrous; lower glume of sessile spikelet concave,
hairy; upper glume of sessile spikelet mucronate.
5. parvflorum
1. Capillipedium fihiculme* Stapf, in Hook. Ic. P1.: t 3085 (1922).-Bor, Grasses
Burma Ceylon India & Pakistan: 111(1960).
Synonym: Andropogonfihiculmis Hook.f., Fl. Brit. hid. 7: 181 (1896).
A weak and trailing annual, 30-60 cm tall. Culms slender, branched from the
geniculate nodes; culm nodes bearded. Leaf blades lanceolate, 5-20cm long, 0.81.2cm wide, constricted at base, glabrous or hairy, apex acute to acuminate or
produced to a setaceous point, margins scabrid; leaf sheaths terete, glabrous, bearded
at the mouth; ligule an unfringed membrane.
*
fihiculme, with thread-like stem (cuim).
224
Inflorescence a panicle, slender; primary branches and raceme peduncles capillary;
racemes 3-4-jointed; rachis internodes with a median translucent longitudinal groove,
margins densely hairy. Sessile spikelet elliptic-lanceolate, hairy, acute; callus short,
bearded; lower glume oblong-lanceolate, glabrous, shortly ciliate on the keels towards
the apex; upper glume boat-shaped, single-keeled, emarginate, glabrous. Florets 2:
lower floret empty, lemma ovate, 1.5mm long, hyaline; upper floret bisexual, lemma
stipiform; awn 15-20mm long, hispid; epaleate; lodicules 2, fleshy; anthers 3, 1.3mm
long; ovary tiny, ovate. Pedicellate spikelet male, elliptic-lanceolate; lower glume
glabrous, 2.5mm long, lanceolate; upper glume lanceolate, 2.5mm long (or slightly
shorter), smooth; lower lemma oblong, 2mm long, obtuse; anthers 3, 1.5mm long;
pedicel with a median translucent longitudinal groove, hairy along the raised margins.
Diagnostic characters and affinities: Has very slender cuims. The epithet fihiculme
alludes to these.
Habitat: In low lying areas of river flood plains.
Type: India, Dhondsee, 08.xii.1894, (without a collector's name) (K!).
Representative specimens:
India: Madras, Tellicherry, 29.xii.1938, Ammal 1240 (K); Tellicherry, North
Malaher, 03.i.1969, Ammal 1832 (K); Chodan, 2 miles east of Sasan, altitude 600ft,
19.x.1970, Hodd 275 (K); Bombay, Unai, south to S.B. hills, 02.xi.1953, Santapau
17201 (K).
Distribution: India (endemic).
2.Capillipedium assimile A. Camus, in Lecomte, Fl. Gen. Indo-Chine VII: 314
(1922).-Bor, Flora of Assam 5: 363 (1940).-Bor, Grasses Burma Ceylon India &
Pakistan: 110 (1960).
Synonyms: Andropogon assimilis Steud., in Zoll., Syst. Verz.: 58 (1854);
Andropogon subrepens Steud., Syn. P1. Glum. 1: 397 (1854); Andropogon glaucopsis
225
Steud., Syn. P1. Glum. 1: 397 (1854); Capillipedium glaucopsis (Steud.) Stapf, in
Hook. Ic. P1.: t 3085 (1922); Capillipedium subrepens (Steud.) Henr., in Blumea III:
463 (1940).
Perennial suffrutescent grass. Cuims slightly woody, more or less round in section,
but somewhat flattened on one side, highly branched from the nodes, branches
spreading, decumbent, 1.8-3.5m tall; internodes glabrous; culm nodes bearded or
glabrous. Leaf blades flat, linear-lanceolate, narrowed at the base, 3-10mm wide, up
to 25cm long, upper surface hirsute, lower glabrous or hispid, margins spinous
toothed, midrib thick, broad and white above; leaf sheaths compressed, the lower
somewhat lax, shortly bearded in the throat, with tubercle-based hairs along most of
their length; ligule a short, truncate, ciliolate membrane.
Inflorescence a panicle; main inflorescence axis, primary branches and raceme
peduncles all glabrous; bearded at the nodes; racemes solitary or paired, 2-3-jointed;
rachis internodes and pedicels with a linear-elliptic translucent longitudinal groove,
hairy along the margins. Sessile spikelet linear-oblong, 2-3.5mm long; lower glume
linear-oblong, glabrous on the back, slightly depressed, ciliate along the keels, with a
narrowly truncate apex; upper glume broadly lanceolate, acute, single-keeled,
depressed longitudinally along both sides of the keel, glabrous. Florets 2: lower floret
empty, lemma ovate, obtuse, nerveless; upper floret bisexual, lemma stipiform,
narrow, glabrous; awn 6-12mm long, very slender; epaleate; lodicules 2, fleshy,
opaque, cuneate; anthers not seen; ovary ovate. Pedicellate spikelet neuter, linearlanceolate, smaller than the sessile; lower glume linear-lanceolate, acute, glabrous,
keels ciliate; upper glume acute, ciliate.
Chromosome number(s): 2n = 20 (De Wet, Borgaonkar & Richardson, 1963); 40 (
Larsen, 1963).
Diagnostic characters and affinities: Has highly branched woody culms.
Habitat: Found mainly in wooded valleys and forests.
226
Uses: Has a very hard solid stem used as a pipe cleaner by smokers (Bor, 1960).
Type: Type locality, India.
Representative specimens:
China: Yunnan, western flank of the Shweli-Salwin divide, latitude 25°20'N, altitude
6000-700011, viii.1912, Forrest 9135 (E); Tuhshan, altitude 400m, 20.viii.1930,
Tsiang 6513 (E).
India: Mussoorie, altitude 600011, ix. 1920, Anderson s. n. (E); Chutia Magpur, Tundi
hills, altitude 1 SOOft, xi. 1886, Campbell 7743 (E); Western Himalayas, Punjab, Larji,
Kulu, altitude 4500ft, 04.xi. 1931, Koelz 3096; Bashr, North-western Himalayas,
Manglad valley, altitude 6000ft, 29.ix.1891, Lace 1051 (E); Ghogar Dhar, Mandi,
Himachal Pradesh, altitude 1800m, 16.x.1985, Mcbeath 1672 (E); Bihar, near the
summit of Perisnath, altitude 430011, xi. 1884, Watt 9904 (E).
Thailand: Chiang Mai, lower elevation of Doi Suthep, altitude 350-700m,
08.ix. 1967, Iwatsuki, Koyama & Fukuoka T9434 (E); 12km south-east of Fang along
the Fang-Chiengrai trail, 19'56'N 99'18'E, altitude 550m, 26.vii. 1968, Larsen,
Santisuk & Warncke 2734 (E).
Tibet: Kougbo Province (no collecting details), (no date), Ludlow, Sherrff& Taylor
s. n. (E).
Distribution: South-east Asia.
3. Capillipedium spicigerum S. T. Blake, in Univ. Queensl. Papers, Dept. Biol. II
No.3: 43 (1944).-Burbidge, Australian grasses 3: 42 (1970).-Wheeler, Jacobs &
Norton, Grasses of New South Wales: 136 (1982).-Simon, A key to Australian
grasses (ed.2): 82 (1993).-Simon & Latz, A key to the grasses of the Northern
Territory, Australia: 23 (1994).
227
Stout and tall perennial grass. Culms 1-1.5m tall, glabrous; cuim nodes with a ring of
short velvety hairs or glabrous. Leaf blades narrowly linear, narrowed towards the
base, tapering to a fine point, flat, up to 40cm long, 5-8mm wide, hirsute behind the
ligule, lower surface sparsely hairy; leaf sheaths longer or shorter than the internodes,
glabrous or hairy with tubercle-based hairs; ligule a short scarious fringed membrane,
ciliolate.
Inflorescence a panicle; main inflorescence axis terete, shortly villous; primary
branches semiverticillate, opposite or the upper ones alternate, glabrous or pubescent;
raceme peduncles glabrous. Racemes 3-8-jointed; rachis internodes and pedicels
subequal, with a median translucent longitudinal groove, ciliate on the sides. Sessile
spikelet oblong-lanceolate, 3-4.2mm long; lower glume truncate, slightly depressed
along the middle, more or less hairy on the back, keels scabrous-ciliate upwards;
upper glume glabrous, scabrous on the keel upwards. Florets 2: lower floret empty,
lemma ovate, acute, 2.8mm long; upper floret bisexual, lemma stipiform; awn with
stipe 12-18mm long; lodicules 2, fleshy, opaque, cuneate; anthers 3, 1.3mm long;
ovary oblong. Pedicellate spikelet male, 3mm long; lower glume acute, ciliate on the
keels; upper glume glabrous; lower lemma ovate, obtuse; anthers 3, 1mm long.
Chromosome number: 2n = 40 (De Wet, Borgaonkar & Richardson, 1963).
Diagnostic characters and affinities: This species is often thought to be a hybrid
between Bothriochloa bladhii and Capillipedium parvflorum. It certainly shows
affinities with C. parvflorum. They differ in the sense that C. parvflorum has a
prominently concave lower glume, while in C. spicigerum, the lower glume is slightly
depressed. They also differ in the number of raceme joints, which are 1-2 (rarely 3) in
C. parvflorum, and 3-8 in C. spicigerum. These inflorescence characteristics make
these species easily distinguishable at a glance.
Habitat: Widely spread over the coastal and subcoastal districts of eastern Australia
in Eucalyptus forests.
Type: Australia, Queensland, Moreton District, Northgate, Brisbane, 02.vi. 1940,
Blake 14267 (K!).
228
Representative specimens:
Australia:
New South Wales: 6.4km by road south-west of Singleton on the Windsor road,
altitude 90m, 12.vi.1975, Coveny & Powell 6562 (K).
Queensland: Moreton District, Brisbane, in railway enclosure between Northgate and
Virginia, altitude 25ft, 01.i.1941, Blake 14307 (K).
New Caledonia: Noumea, Montravel, altitude 120m, 20.xii.1964, Mackee 196 (K);
Noumea, Montravel, altitude 50-100m, 20.iii.1969, Mckee 20305 (K).
Distribution: Mainly an Australian species but also found in the Pacific Islands.
4. Capillipedium huegelii (Hack.) Stapf, in Hook. Ic. Pt.: t 3085 (1922).-Bor, Grasses
Burma Ceylon India & Pakistan: 111(1960).
Synonyms: Andropogon huegelii Hack., in DC., Monogr. Phan. 6: 49 (1889);
Andropogon foetidus Donna Lisboa ex Lisboa, in J. Bombay Nat. Hist. Soc. 6: 205
(1891); Andropogon schmidii Hook.f., F!. Brit. Ind. 7: 180 (1896); Capillipedium
huegelii (Hack.) Blatt. et McCann, in J. Bombay Nat. Hist. Soc. 32: 420 (1928);
Capillipedium foetidum (Lisboa) Raitz. et Jam, in Indian For. 77: 752 (1951).
A perennial grass. Culms 30-80cm tall, erect or branching from base, shortly
creeping; cuim nodes hairy. Leaf blades 10-25cm long and 8mm wide, linear,
constricted at the base, glabrous, margins scabrid; leaf sheaths compressed; ligule a
fringed membrane, lacerate.
Inflorescence a somewhat contracted panicle; main inflorescence axis hirsute or
glabrous, with long cilia at the joints; primary branches and raceme peduncles
glabrous; number of raceme joints variable, but mostly 2-4-jointed; rachis internodes
with a median longitudinal groove, hairy along the margins. Sessile spikelet oblonglanceolate, 3.5mm long; lower glume flat on the back, glabrous, shortly ciliate along
229
the keels towards the apex, truncate at the apex; upper glume smooth, single-keeled,
scabrous upwards along the keel, ciliate on the inturned margins. Florets 2: lower
floret empty, lemma oblong with an incised apex, 2mm long; upper floret bisexual,
lemma stipiform, single-nerved; awn 15mm long; lodicules 2, fleshy, cuneate; anthers
not seen; caryopsis obovate. Pedicellate spikelet neuter, 4-4.5mm long, lãnceolate;
lower glume lanceolate, dorsally flat, glabrous, ciliate on the keels; upper glume
smooth, single-keeled, ciliate on the margins, lanceolate; lower lemma lanceolate,
emarginate; pedicel with a median longitudinal groove, hairy on the margins.
Chromosome number(s): 2n = 20 (De Wet, Borgaonkar & Richardson, 1963).
Habitat: Found on grassy plains.
Type: India, Nilagiri, 1818-1835, Schmids.n. (K!).
Representative specimens:
India: Central Province, Jubbulpur District, Marble Rocks, near Jubbulpur,
25.xi.1889, Duthie 9898 (K); Bombay, Khandala, Western Ghats, xii.1922,
*sedgwick 7970 (K); Bombay, 20.xi.1907, Talbot s. n. (K).
* - with a pilose inflorescence axis, and hispid primary branches and raceme
peduncles.
Distribution: Central to Western India (endemic).
5. Capillipediumparvflorum* (R. Br.) Stapf, in Pram, Fl. Trop. Afr. IX: 169 (19 17).Bor, Flora of Assam 5: 362 (1940).-Sturgeon, A revised list of grasses of Southern
Rhodesia Part IV: 8 (1954).-Bogdan, A revised list of Kenya grasses: 56 (1958).Jackson & Wiehe, An annotated check-list of Nyasaland grasses: 32 (195 8).-Bor,
Grasses Burma Ceylon India & Pakistan: 112 (1960).-Napper, Grasses of
* parviflorum, parvus, small, tiny, and forum, flowers, alluding to the small spikelets.
230
Tanganyika: 99 (1965).-Clayton, in Polhill (ed.), Fl. Trop. East Mr., Gramineae (3):
718 (1982).
Synonyms: Holcus parvflorus R. Br., Prodr.: 199 (1810); Sorghum parvfiorum
Beauv., Agros: 132 (1812); Anatherum parvy7orum Spreng., Syst. 1: 290 (1825);
Holcus caerulescens Gaud., in Freycin. Voy. Bot.: 411 (1829); Andropogon
micranthus Kunth, Rev. Gram. I: 165 (1829); Steud., Syn. P1. Glum. 1: 396 (1854);
Hack., in DC., Monogr. Phan. 6: 488 (1889); Hook.f., F!. Brit. Ind. VII: 178 (1896);
Andropogon alternans J. S. Presi., in C. B. Presi., Reliq. Haenk. 1: 342 (1830);
Rhaphis caerulescens Desv., Opusc.: 69 (1831); Chrysopogon violascens Trin., in
Mem. Ac. Petersb. 6 me ser. II: 319 (1832); Andropogon quartianianus A. Rich.,
Tent. F!. Abyss. II: 469 (1851); Andropogon violascens Nees ex Steud., Syn. P1.
Glum. 1: 396 (1854); Andropogon parvispicus Steud., Syn. P1. Glum. 1: 397 (1854);
Andropogon capilliflorus Steud., Syn. P1. Glum. 1: 397 (1854); Rhaphis villosula
Nees ex Steud., Syn. P1. Glum. 1: 397 (1854); Rhaphis microstachya Nees ex Steud.,
Syn. P1. Glum. 1: 397 (1854); Andropogon serratus Miq., Ann. Mus. Lugd.-Bat. II:
290 (1866); Chrysopogon parvijiorus Benth., Fl. Austr. VII: 537 (1878);
Chrysopogon parvispicus (Steud.) Wats., in Atkins., Gaz. N. W. Ind.: 392 (1882).
Tufted perennial, mostly 75-100cm tall. Cuims with bearded nodes. Leaf blades
glaucous, the blades expanded, erect and straight, linear from an often narrowed and
slightly contracted base, 30cm long, 2.5mm wide, finely pointed, margins scabrid;
leaf sheaths terete, tight or at length slipping from the cuims, the lowermost ones
longer, the others usually shorter than the internodes, more or less with tubercle-based
hairs and frequently villous on the collar; ligule a very short, truncate, ciliolate
membrane.
Inflorescence a profusely divided delicate panicle; main inflorescence axis terete;
primary branches alternate, opposite or semiverticillate, divided from low down and
up to the third degree; raceme peduncles filiform, glabrous. Racemes 1-jointed,
reduced to 1 sessile and two pedicellate spikelets, rarely 2-jointed with 2 sessile and 3
231
pedicellate spikelets; rachis internodes very similar to the raceme peduncles and
glabrous. Sessile spikelet narrowly oblong, 2.8-5mm long; lower glume with short
hairs on the back, shallowly concave in the middle, ciliate upwards on the keels,
truncate, with 2 intracarinal nerves; upper glume single-keeled, scabrid on the keel,
the median nerve drawn to a mucro. Florets 2: lower floret empty, lemma 2-2.5mm
long, ovate; upper floret bisexual, lemma stipiform, linear, flattened, entire, singlenerved; awn 20nmi long (including the upper lemma), minutely hispidulous; anthers
3, 1.5mm long; ovary oblong, 2.1mm long. Pedicellate spikelet lanceolate, male or
neuter, 2.5-3.5mm long; lower glume shortly hairy on the keels, truncate, with 3
intracarinal nerves, acute; upper glume glabrous, lanceolate, with 2 intracarinal
nerves; lower lemma oblong, truncate; anthers 3, 1.5mm long; pedicels with a median
translucent longitudinal groove, subequal, hairy on the raised margins.
Chromosome number(s): 2n = 20, 40 (De Wet, Borgaonkar & Chheda, 1961; De
Wet, Borgaonkar & Richardson, 1963).
Diagnostic characters and affinities: The racemes consist of a single triad of
spikelets at the ends of branches. Also the lower glume of the sessile spikelet is
prominently concave along the back between the nerves.
Habitat: Occurs on grasslands.
Type: Australia, 14.i.1803, Brown 6188 (BM! K!).
Representative specimens:
Malawi: Dowa District, Uperere Mission, Chankalamu Dambo, 02.xi.1950, Jackson
249 (MAL); Agricultural Research Station near Lilongwe, 30.ix.1952, Kantikana 4
(BM); Chipata mountain, altitude 6000ft, 04.v.1963, * Verboom 978 (SRGH).
Zambia: Fort Jameson, 25.v. 1963, * Verboom 761 (SRGH, BM).
Zimbabwe: Gwelo District, 6 miles south from Gwelo, altitude 4600ft, 18 .iii. 1967,
Biegel 2000 (BM); Irelawney, 26.iv. 1931, Brian 3757 (SRGH); Salisbury (Harare)
District, Ballantyne Park, altitude 4950ft, 26.iii.1965, Crook 730 (BM); Salisbury
(Harare), university grounds near main entrance gate, altitude 1500m, 01.i.1968,
232
Crook P55 (PRE); Salisbury (Harare), Marlborough, ii.1954, Kerr 6 (PRE);
Irelawney District, v.1931, Litt 162 (BM); Salisbury (Harare), wet ground bordering
Gwebi river, Mazoe road, 22.ii.1974, Simon 2385 (SRGH, PRE).
* - racemes 2-jointed (i.e. with 2 sessile and 3 pedicellate spikelets).
Distribution: Widely distributed throughout the tropics of the Old World, China and
Japan.
8.2.11. Spathia Ewart, in Ewart & Davies, Fl. N. Tern: 26 (1917). The type species is
Spathia neurosa Ewart & Archer.
Tufted annual. Inflorescence terminal and axillary, of digitate racemes enclosed in a
large membranous spathe; raceme with one homogamous pair; internodes and
pedicels solid, oblique at the tip, densely ciliate on edges; sessile spikelet callus
obtuse; lower glume coriaceous, convex, acute, covered with brown hair on the back;
upper lemma entire, awned. Pedicellate spikelet well developed, neuter, with a short
oblong callus.
Species 1.
Spathia neurosa Ewart & Archer, in Ewart & Davies, Fl. N. Tern: 26 (1917).-Blake,
in Univers. Queensl. Papers 2(3): 47 (1944).-Lazarides, The grasses of Central
Australia: 228 (1970).-Simon & Latz, A key to the grasses of the Northern Territory,
Australia: 49 (1994).
Tufted, rather coarse, reddish or straw-coloured annual, up to 75cm tall. Culms erect
or somewhat bent at the lower nodes, simple or sparingly branched, flattened or
shallowly channelled on one side, hairless, smooth, powdery close below the nodes or
shiny, with 2-5 densely hairy nodes. Leaf blades up to 16cm long and 4mm wide,
233
usually flat, tapered to a long fine point with thickened sharp edges; leaf sheaths
loose, hairless, more or less smooth, strongly curved, the lower ones much shorter
than the internodes, the upper ones longer than the internodes, broad, inflated, with
short blades; ligule thin, papery, transparent, fringed with minute hairs.
Inflorescence of 3 (sometimes 4 or 5) racemes, spatheate, terminating the culm and
branches; spathe membranous, broad, up to 25cm long; racemes digitate, sessile to
subsessile, villous with brown hairs, each with the basal pair of spikelets
homogamous (male or neuter) and persistent; rachis internodes and pedicels linearfihiform, curved, solid, densely ciliate on both margins. Sessile spikelet lanceolate,
5.5-6.2mm long, dorsally compressed; callus very small, with dense fulvous hairs;
lower glume 2-keeled, narrowed upwards, densely hairy on the back and with a
subapical beard of longer, tubercle-based, rich brown hairs about 6mm long
extending downwards close to the keels; upper glume slightly longer, lanceolate,
acute, keeled upwards, 3-nerved, villous on the back in the lower part, ciliate on the
keel with tubercle-based hairs. Florets 2: lower floret empty, lower lemma hyaline,
nerveless, lanceolate, acute, ciliate upwards, 2.5-2.7mm long; upper floret bisexual,
upper lemma stipiform, 1.8-2.7mm long, passing into a geniculate scabrous awn; awn
19-35mm long, with a villous twisted column and a yellowish weakly twisted bristle;
palea present, small; anthers 3, 0.5mm long; ovary oblong, 2-2.8mm long. Pedicellate
spikelet neuter, awnless, dorsally compressed, reduced to very unequal glumes, about
5mm long, oblanceolate; lower glume oblong-elliptic, membranous, with narrowly
inflexed hyaline ciliate margins, pubescent like that of the sessile spikelet; upper
glume much shorter, pubescent upwards on the midvein; pedicel about half as long as
the pedicellate spikelet.
Diagnostic characters and affinities: Racemes protected by large sheathing bracts,
spathes, which are actually enlarged bases of foliage leaves. The inflorescence is only
partly exserted at maturity. Spathes are large and have prominent veins. This is the
only member of the subtribe that has spathes. In contrast to other spatheate
234
inflorescence, the spathe in S. neurosa does not interrupt the inflorescence, hence
producing partial inflorescence, but rather encloses the inflorescence.
Habitat: Grasslands with cracking clay soils.
Type: Australia, Northern Territory, 30 miles southeast of Newcastle Waters,
09.viii. 1911, Hill 504 (K!).
Representative specimens:
Australia:
Northern Territory: Sondan Station, 14.vii.1947, Ardut (K); 26 miles northnorthwest of Brunette Downs, altitude 222m, 16.v.1947, Blake 17818 (K, BM);
between Sondan Station and Rankine Store, altitude 253m, 21.v.1947, Blake 17911
(K); north of Austral Downs, altitude 200m, 30.v.1947, Blake 17973 (K); Austral
Downs Station, 27.vii. 1971, Latz 1681(K); 25 miles west-northwest of Rockhampton
Downs Station, 02.vii.1948, Perry 1590 (K).
Queensland: about 35 miles north of Camooweal, altitude 302m, 25.v.1947, Blake
17951(K); about 50miles south-southeast of Camooweal, altitude 204m, 28.v.1947,
Blake 17955 (K).
Distribution: Australia.
8.2.12. Bothriochloa Kuntze, Rev. Gen. P1. 2: 762 (1891). The type species is
Bothriochloa bladhii (Retz.) S. T. Blake.
Synonyms: Andropogon subgen. Gymnoandropogon Nees, Fl. Aft. Austr.: 103
(1841); Gymnandropogon (Nees) Duthie, in Atkinson, Gaz. N.W. Prov. & Oude 10:
638 (1882); Amphilophis Nash, in Britton, Man. F!. North. States: 71(1901).
Perennials. Inflorescence of few to many-jointed pedunculate digitate or subdigitate
racemes, the branches undivided or the lower ones sparingly branched; racemes
without homogamous spikelet pairs, always terminate in a triad; spikelets paired, one
235
sessile, the other pedicellate, similar in shape or the pedicellate reduced and smaller,
always differing in sex; rachis internodes and pedicels linear-fihiform, with a
translucent longitudinal groove. Sessile spikelet callus obtuse; lower glume mostly
cartilaginous, broadly convex to slightly concave, sometimes with 1-3 circular pits on
the back, acute; upper lemma hyaline base of the awn, entire (bibbed in B. biloba and
B. erianthoides). Pedicellate spikelet much like the sessile or smaller, male or neuter.
Species ± 35.
A key to taxa sampled:
Cuim nodes glabrous:
Culm intemodes terete:
Lower glume of the sessile spikelet hairy on the lower '/2, upper '/2
glabrous, slightly depressed longitudinally.
1. ischaemum
Lower glume of the sessile spikelet glabrous, convex on the back.
2. saccharoides
Cuim internodes channelled on one side:
Sessile spikelet 5.5mm long; anther 1.
3. decipiens
Sessile spikelet 7mm long; anthers 3:
Lower glume with a single pit above the middle; palea absent;
upper lemma stipiform.
4. macera
Lower glume without a pit above the middle; palea present;
upper lemma bibbed.
5. biloba
Cuim nodes with a ring of hairs:
Lower glume of the sessile spikelet not pitted.
6. radicans
Lower glume of the sessile spikelet pitted:
7. bladhii
Inflorescence paniculate.
236
Inflorescence subdigitate or digitate:
8. pertusa
Inflorescence digitate.
Inflorescence subdigitate:
Cuim internodes terete; sessile spikelet 5.5mm long;
lower glume of pedicellate spikelet with a single slitlike pit.
9. kuntzeana
Culm intemodes channelled on one side; sessile
spikelet 4.5mm long; lower glume of pedicellate
spikelet with 0-4 slit-like pits.
10. insculpta
1. Bothriochloa ischaemum (L.) Keng, in Contr. Biol. Lab. Sc. Soc. China, Bot. Ser.
X: 201 (1936).-Bor, Grasses Burma Ceylon India & Pakistan: 108 (1960).
Synonym: Andropogon ischaemum L., Sp. P1.: 1047 (1753).
A tufted perennial of up to 70cm tall; shortly rhizomatous. Culms branched or
unbranched above; culm nodes exposed, either hairy or glabrous; lower culm nodes
(most) with prophylls; culm intemnodes solid and glabrous. Leaf blades linear and
narrow, 1-3.5mm wide, mostly basal; young leaves with tubercle-based hairs, mature
ones glabrous or at most with a few scattered hairs towards the leaf base; leaf sheaths
keeled, glabrous and about half the length of internodes; leaf ligule a fringed ciliate
membrane.
Inflorescence terminal and of 3-10 subdigitate racemes with a characteristic purple
tinge; the racemes are either solitary or clustered, and the lower longer than the
inflorescence axis, 15-20-jointed; rachis internodes and pedicels linear and have a
longitudinal translucent groove; rachis nodes somewhat hairy. Sessile spikelet oblong
to elliptic; callus bearded with hairs; glumes membranous; lower glume lightly haired
on the lower '/2, with a slight depression longitudinally; upper glume glabrous, boat-
237
shaped, ciliate on the margins towards the apex. Florets 2: lower floret empty, lemma
oblong to obtuse, sparsely ciliate; upper floret bisexual, lemma stipiform, awned; awn
slightly twisted, glabrous; lodicules 2, fleshy, glabrous; anthers not seen; ovary ovate.
Pedicellate spikelet subequal to the sessile.
Chromosome number(s): 2n = 40, 60 (De Wet, Borgaonkar & Richardson, 1963).
Diagnostic characters and affinities: There seems to be no reliable way of
separating this species from B. radicans (Lehm.) A. Camus. The glabrous black nodes
are a noticeable feature, but they are found occasionally among tropical specimens.
Nevertheless, the two species, though intergrading, appear to be cytogenetically
distinct (De Wet & Higgins, 1963).
Habitat: Dry stony places, borders of fields, and slopes.
Uses: Native pasture species.
Type: Described from southern Europe ("In Europae australiorus" (Tsvelev, 1976)).
Type not seen.
Representative specimens:
China: Chihh Province (no exact location), 26.vii.1912, Mary Strong Clemens 1580A
(E); Kweichow Province (no exact location), altitude 600m, 1 4.xi. 1931, Steward,
Chiao & Cheo 921 (E).
Greece: Mt. Peristeria, altitude 3700ft, 20.vii.1937, Balls & Balfour Gourlay B3668
(E).
Nepal: Yara, altitude 12000ft, 02.viii.1954, Stainton, Sykes & Williams 2133 (E).
Pakistan: Harara District, Abbottabad, 15.xi.1958, Burt 1547 (E); Chitral, altitude
8000ft, 27.vi.1958, Stainton 2769 (E).
Saudi Arabia: Abba-Jabal Sawdah, altitude 5500ft, 17.viii.1983, Collenette 4450
(B); Dalaghran National Park, altitude 7000ft, 21 .viii. 1983, Collenette 4480 (E).
Tibet: Lhasa, altitude 12500ft, 1 1.vii.1943, Ludlow & Sherrff 9476 (E).
238
Turkey: B6 Yozgat, Melikli, altitude 1300-1400m, 10.vii.1982, Davis & Ekim 69043
(E).
USSR: Lennisky District, altitude 1500-2000m, 01.viii.1985, Elias, Murray &
Newcombe 10027 (E).
Yugoslavia: (no exact location), 20.vii. 1960, Webster 4074 (E).
Distribution: South-east Asia, Southern Europe, USSR, USA, and North Africa.
2. Bothriochloa saccharoides (Swartz) Rydb, in Brittonia 1: 81 (193 1).-Gould, The
grasses of Texas: 592 (1975).
Synonyms: Andropogon saccharoides Swartz, Prodr. Veg. Ind. 0cc. 26 (1788);
Andropogon saccharoides ssp. leucopogon subvar. paucirameus Hack., in DC.,
Monogr. Phan. 6: 407 (1889); Sorghum saccharoides Kuntze, Rev. Gen. P1. 2: 792
(1891); Holcus saccharoides Kuntze, in Stuckert, An. Mus. Nac. Buenos Aires 11: 48
(1904); Amphilophis saccharoides Nash, N. Amer. Fl. 17: 125 (1912); Andropogon
saccharoides var. surius Krause, Bot. Centbl. Beuhefte 32: 334 (1914).
Caespitose perennial with cuims erect or somewhat geniculate; cuim nodes exposed,
shortly bearded. Leaf blades glabrous except for a few long hairs in the ligule area,
linear, firm, flat or irregularly folded, mostly 3-6(-8)mm wide, tapering to a long
attenuate tip; ligule a hyaline unfringed membrane, 1 -3mm or more long.
Inflorescence of densely-flowered subdigitate racemes; racemes numerous, erect or
loosely spreading in age, 6-10-jointed, silvery hairy; rachis internodes and pedicels
with a translucent longitudinal groove, both long plumose. Sessile spikelet: lower
glume 4.5mm long, ovate, glabrous, with a short hairy callus, minutely spinulose on
the keels towards the apex, convex on the back; upper glume 4-4.5mm long, glabrous,
boat-shaped, sharply keeled, the keel minutely spinulose upwards. Florets 2: lower
floret empty, lemma hyaline, ovate, slightly ciliate on the upper margins; upper floret
239
bisexual, lemma stipiform, awned; awn 13mm long; epaleate; lodicules 2, fleshy;
anthers not seen; caryopsis oblanceolate. Pedicellate spikelet neuter, somewhat
shrivelled and smaller than the sessile spikelet; lower glume narrowly ovate,
glabrous.
Chromosome number(s): 2n = 60, 120 (Gould, 1975).
Diagnostic characters and affinities: Has a characteristic silvery hairy
inflorescence.
Habitat: Restricted to relatively dry, usually sandy soils, or on well-drained clay soil
sites such as railroads and road embankments and dry banks of ditches and gullies.
Type: Type locality, Jamaica.
Representative specimens:
Jamaica: Isla de Pinos, near Nueva Gerona, 02.iii.1904, Curtis (BM); Hall's Delight,
above Hope, altitude 1000ft, 25.x.1912, Harris 11293 (BM); below New Castle,
altitude 3500ft, 03.xi.1912, Harris 11396 (E); Mavis Bank, altitude 3500ft,
26.xi.1912, Harris 11449 (E); Rams Horn Range, xi.1912, Hitchcock 267 (BM).
Distribution: Tropical regions of the Americas and the Caribbean Islands.
3. Bothriochloa decipiens* (Hackel) C. E. Hubbard, in Kew Bull. 1934: 444 (1934).Blake, In Univ. Queensl. Papers 2(3): 32 (1944).-Burbidge, Australian grasses: 146
(1968).-Wheeler, Jacobs & Norton, Grasses of New South Wales: 128 (1982).-Simon,
A key to Australian grasses (ed.2): 77 (1993).-Simon & Latz, A key to the grasses of
the Northern Territory, Australia: 20 (1994).
Synonyms: Andropogon pertusus (L.) Willd. var. decipiens Hack., in DC., Monogr.
Phan. 6: 483 (1889); Andropogon decipiens Domin, in Biblioth. Bot. 20: 266 (1915).
*
decipiens. deceiving
240
There are two varieties in this species, var. decipiens and var. cloncurrensis, which
key out as follows:
Cuims tall and stout, up to 170cm high, always glabrous and smooth; leaves always
glabrous, up to 35cm long and 6mm wide; racemes up to 15 in number.
var. cloncurrensis
Culms up to 150cm tall, slender; leaves 25cm long and 2-5mm wide, with a few hairs
from tubercles on the lower surface; racemes 3-5 in number.
var. decipiens
Tufted perennial, 30-170cm tall. Cuims erect or geniculate at base, slender and
channelled along one side, branched above; culm nodes glabrous; lowest internodes
with a few hairs often from tubercles. Leaf blades 6-35cm long, 2-6mm wide, slightly
narrowed towards the base, revolute, lower surface hairy from minute tubercles,
upper glabrous to scabrous near the apex particularly on the margins; leaf sheaths
glabrous, keeled, striate, most shorter than the internodes except for the lowermost
which are longer than the internodes; ligule an unfringed membrane, rounded,
truncate, about 1mm long.
Inflorescence consists of 3-5 subdigitate erect racemes; main inflorescence axis
slender, angular; racemes 10-12-jointed, pedunculate; peduncles alternate or the
lowermost opposite, bearded at the nodes; racemes silvery hairy, 4-7cm long; rachis
internodes with a translucent longitudinal groove, hairy, hairs increasing upwards to
5mm long. Sessile spikelet lanceolate, up to 5.5mm long (4.5-5.5mm long); callus
short, shortly bearded; lower glume with a deep circular, pit above the middle, pilose
on the back in the lower third, ciliate on the lateral keels towards the apex, flat; upper
glume scabrous near tip, compressed laterally, sharply carinate. Florets 2: lower floret
empty, lemma oblong, obtuse; upper floret bisexual, lemma 2.3mm long, stipiform
with narrowly hyaline margins, single nerved, awned; awn 13mm long; epaleate;
lodicules 2, fleshy, glabrous; anther 1, 1.5mm long; ovary narrowly obovate.
Pedicellate spikelet neuter, represented by a subulate lower glume, 3.5nm'i long,
scabrous; pedicel with a translucent longitudinal groove, 4mm long, hairy.
241
Chromosome number: 2n =40 (De Wet, Borgaonkar & Richardson, 1963).
Diagnostic characters and affinities: This species has a single anther placed
between the two feathery stigmas. Apparently both varieties have unpitted forms
(Blake, 1944). I was only able to see the unpitted forms of var. cloncurrensis,
indicated by asterix in the specimens. In this form the lower glume of the sessile
spikelet is concave on the back, otherwise it has similar features to the pitted form.
The species is closely related to B. pertusa. It differs in having more hairy
racemes and narrowly lanceolate sessile spikelets, and in the narrow linear pedicellate
spikelets being neuter and reduced to the lower glume.
Habitat: Abundant in Eucalyptus forests of Queensland.
Uses: Palatable to stock when desirable grasses are not available (Blake, 1944).
Type: Australia, Queensland, Darling Downs District, Oakey, 1930, Donges 32 (K!).
Representative specimens:
Australia:
var. decipiens:
New South Wales: Triangie District (no exact location), 2511966, Campbell NSW
89959 (K); Mt. Lindsay basins, 23.i.1932, Davies s.n. (K).
Queensland: Moreton District, Woolowin, Brisbane, altitude 76ft, 14.v.1930,
Hubbard 2550 (K); Maranoa District, between Amby and Eurella, altitude 125 Oft,
06.i.1931, Hubbard& Winders 6329 (K).
var. cloncurrensis:
Queensland: Port Curtis District, Charon's Ferry Station, near Styx, altitude 40-50ft,
01 .vi. 1942, Blake 14614 (K); Cook District, Lyndbrook, in railway enclosure,
abundant Lyndbrook to Einasleigh, 20.xi.1942, *Blake 14709 (K); North Kennedy
District, Regional Station, Ayr, 07.vi. 1950, Cowdry 3 (K); Horse Creek, near
Dimbulah, on Mareeba road, altitude 400m, 21.iv.1962, *McKee 9255 (K); North
242
Kennedy District, Mt. Woodhouse, about 40km south-west of Ayr, altitude 1200m,
20.vi.1949, Smith 04313 (K).
Distribution: Australia.
4. Bothriochloa macera (Steud.) S. T. Blake, in Proc. Roy. Soc. Queensl. 80: 64
(1969).-Burbidge, Australian grasses 3: 204 (1970).-Wheeler, Jacobs & Norton,
Grasses of New South Wales: 128 (1982).-Simon, A key to Australian grasses (ed.2):
77 (1993).
Synonyms: Andropogon macer Steud., Syn. P1. Glum. 1: 371 (1854); Bothriochloa
ambigua S. T. Blake, in Univ. Queensl. Papers, Dept. Biol. II No.3: 29 (1944).
Tufted perennial grass, 50-60cm tall. Culms geniculate or erect, glabrous, branched
from the nodes; cuim nodes glabrous; culm internodes (except the terminal flowering
one) channelled on one side. Leaf blades flat, up to 15cm long, 2.5-5.5mm wide,
scabrous on the upper surface, sparsely hairy beneath, the hairs arising from
tubercles; leaf sheaths keeled, striate, auriculate, mostly shorter than the internodes;
ligule a fringed membrane, 1.5mm long, with 2mm long cilia.
Inflorescence of 3-6 (mostly 4) subdigitate racemes; main inflorescence axis glabrous
at the nodes; racemes solitary or paired, usually with 1 below the others, glabrous, 12jointed; rachis internodes and pedicels with a translucent longitudinal groove, 4.5mm
long, margins with hairs increasing upwards to 4.5mm long. Sessile spikelet
lanceolate-linear, 7mm long (including callus), narrowed to both sides from the
middle; callus 0.75mm long, bearded; lower glume narrowly rounded at the tip,
concave on the back, 6.2mm long, loosely hairy on the lower third, spinulose alpng
the keels above the middle, sometimes with a deep circular pit above the middle;
upper glume sharply carinate and depressed on either side of the keel, smooth, 6mm
long. Florets 2: lower floret empty, lemma ovate-lanceolate, 4.5mm long; upper floret
bisexual, lemma stipiform, entire, 2.7mm long, awned from the tip; awn 22mm long;
243
palea absent; lodicules 2, fleshy, cuneate; no anthers seen; caryopsis oblanceolate,
3.25mm long. Pedicellate spikelet neuter; lower glume 6mm long, lanceolate acute,
scabrous.
Diagnostic characters and affinities: Apparently pitted and unpitted forms appear to
be nearly equally common. B. macera is closely related to B. decipiens, differing
chiefly in the slightly differently shaped, and on the whole, rather larger sessile
spikelet which is more gently narrowed towards the distinctly rounded tip, with 3
anthers and in the slightly more developed pedicellate spikelet.
Habitat: Common in disturbed grasslands. It is evidently a drought resistant species
which thrives on soils of low fertility, and which under some conditions produces a
fair amount of forage (Blake, 1944).
Type: Australia, New South Wales, Central Western Slopes, Merriwa, iii.1914, Cheel
(NSW, BRI).
Representative specimens:
Australia:
New South Wales: Glen Davis, altitude 300m, 24.iv.1977, Coveny & Hind 9271 (K);
Canberra, 02.vi.1962, McKee 9009 (K); lower eastern slopes of Black mountains,
Canberra, altitude 200ft, 23 .xii. 1960, Pullen 2504 (K); 11 miles south-west of Scone,
16.iii.1960, Story 7074 (K).
Victoria: lower Deddick river, about 1'/2 miles from Snowy river bridge, 20.i.1953,
Melville & Wakefield 3001 (K); 17 km west of Werribie, 02.v.1978, Muir 5723 (K).
Distribution: Australia: New South Wales; Queensland; Victoria.
244
5. Bothriochloa biloba* S. T. Blake, in Univ. Queens!. Papers, Dept. Biol. II No.3: 27
(1944).-Wheeler, Jacobs & Norton, Grasses of New South Wales: 128 (1982).-Simon,
A key to Australian grasses (ed.2): 76 (1993).
Tufted perennial. Culms erect or genicu!ate at base, up to 1 m tall, simple or sparingly
branched from the lower nodes; culm nodes glabrous; culm internodes channelled on
one side. Leaf blades narrowly linear, tapering to a fine point, flat, up to 20cm long,
3-5mm wide, smooth beneath, upper surface with scattered tubercle-based hairs; leaf
sheaths loose, glabrous, smooth, striate; ligule an unfringed membrane, 1 -2mm long,
truncate, surrounded by long ciliate hairs.
Inflorescence of 3-6 subdigitate racemes; main inflorescence axis glabrous or ciliate
on the upper part, hairy at the nodes; racemes 15-jointed; raceme peduncles similar to
the main inflorescence axis but slender; rachis internodes and pedicels hairy on the
sides, with a translucent longitudinal groove, somewhat rounded at the back; racemes
villous with white hairs. Sessile spikelet lanceolate, widest at or below the middle,
7mm long (6.4-8mm long); callus bearded, 1mm long, hairs up to 3.5mm long; lower
glume minutely truncate, pilose on the lower third or half, margins ciliate in the upper
part, slightly depressed longitudinally; upper glume mucronate due to the middle
nerve drawn into a short tip, sharply carinate, the keel scaberulous towards the apex.
Florets 2: lower floret empty, lemma broadly ovate, acute, 4mm long; upper floret
bisexual, lemma 3.7mm long, narrowly linear but more or less stipiform,
membranous below and thickened upwards, bibbed, lobes filiform, 0.75mm long,
awned from the sinus; awn scaberulous, 25mm long (including lemma); palea ovate,
more or less jagged and ciliate at the apex, 1.5mm long; lodicules 2, fleshy, cuneate;
anthers 3, 2mm long; ovary elliptic-oblong. Pedicellate spikelet neuter, inrolled-
*
biloba, named becuase of the bibbed upper lemma of the sessile spikelet, a feature uncommon in
the genus Bothriochloa (found also in B. erianthoides).
245
subulate, 5.3mm long; lower glume scabrous, minutely ciliolate on the lateral keels;
pedicel with a longitudinal translucent groove, villous with hairs increasing upwards,
5.2mm long.
Diagnostic characters and affinities: The body of the upper lemma of the sessile
spikelet is more or less stipiform but slightly wider than in other species of
Bothriochloa; the upper lemma has two fine lobes; and the palea is well developed. In
most other species of Bothriochloa, the upper lemma is entire and very narrow,
scarcely wider than the body of the awn.
Habitat: Common in cleared areas of Eucalyptus forests.
Type: Australia, New South Wales, Western Plains, Wirialda, 17.iv.1937, Glenfield
Vet. Res. Station (NSW, BRI).
Representative specimens:
Australia:
New South Wales: Hunter Valley, 16 miles south-east of Coolah, 25.ii.1960, Story
7056 (K); Rookwood (near Sydney), 31 .iii. 1935, Vickery 12242 (K).
Queensland: Darling Downs District, near Oakey, towards Toowomba, 05.iv.1958,
Blake 20340 (K); 13 miles south of Toowomba, De Wet Okla. 10942 (K); Darling
Downs District, Allora, 24.iii.1953, Melville, Blake & Everist 3423 (K).
Distribution: Australia.
6. Bothriochloa radicans* (Lehm.) A. Camus, in Ann. Soc. Linn. Lyon 1930 n.s.
LXXVI: 164 (1931).-Sturgeon, A revised list of grasses of Southern Rhodesia Part
IV: 9 (1954).-Chippindall, in Meredith (ed.), The grasses and pastures of South
Africa: 48 (1955).-Bogdan, A revised list of Kenya grasses: 56 (1958).-Napper,
*
radicans, r000ting, putting forth aerial roots.
246
Grasses of Tanganyika: 100 (1965).-Hood, A guide to the grasses of Zambia: 57
(1967).-Clayton, in Polhill (ed.), F!. Trop. East Aft., Gramineae (3): 721 (1982).Muller, Grasses of South West Africa/Namibia: 86 (1984).-Chaudhary, Grasses of
Saudi Arabia: 420 (1989).-Gibbs Russell et al, Grasses of Southern Africa: 63 (1991).
Synonyms: Andropogon radicans Lehm., in Ind. Sem. Hort. Hamb. (1828);
Andropogon ischaemum var. radicans Hackel, in DC., Monogr. Phan. 6: 476 (1889);
Amphilophis radicans Stapf, in Pram, F!. Trop. Aft. IX: 172 (1917).
A tufted, often stoloniferous perennial of up to 70cm tall; has a shrubby growth form
and or can be decumbent (including rooting at the nodes). Cuims herbaceous and
branching from the lower nodes to form cushions; culm nodes exposed and bearded,
the lower with small velvety hairs and the upper glabrous; culm internodes terete.
Leaf blades linear-lanceolate, 6-20cm long, 2-6mm wide, gradually tapering to a fine
point; sheaths glabrous except at the usually bearded mouth, those of the basal leaves
loose and open, others closely clasping the culms; leaf ligule a ring of white hairs.
The inflorescence is of clustered 5-16 hairy subdigitate racemes up to 70mm long; the
main inflorescence axis is slender and shorter than the lower racemes; racemes 10-20jointed, shortly pedunculate; rachis intemodes and pedicels linear and have a
longitudinal translucent groove in the middle, hairy; rachis joints densely long hairy.
Sessile spikelet lanceolate, 2.5-4mm long; lower glume slightly concave on the back
and hairy on the lower half, firmly membranous; upper glume with an acute apex,
with a single keel, minutely hispid along this keel towards the apex. Florets 2: lower
floret empty, lemma oblong-linear, 3mm long, ciliolate at the apex; upper floret
bisexual, lemma stipiform, glabrous, 1.5mm long, awned; awn twisted, geniculate,
25nmi long; epaleate; lodicules 2, fleshy and glabrous; callus short and blunt; anthers
3, 1.5mm long; ovary oblong. Pedicellate spikelet neuter; lower glume 3.5mm long,
oblong, with sparse long white hairs along the lateral keels and ciliolate towards the
apex on these keels; upper glume flimsy, ciliate on margins, lanceolate, 3mm long;
lower lemma oblong, obtuse, truncate, 2mm long; pedicels linear.
247
Chromosome number: 2n =40 (De Wet & Higgins, 1963).
Diagnostic characters and affinities: This species is similar to B. ischaemum, except
that B. ischaemum tends to have glabrous black nodes as a noticeable feature.
Habitat: Open deciduous bushland.
Uses: Native pasture species.
Type: Cultivated in Hamburg, seed from South Africa (whereabouts uncertain).
Representative specimens:
Botswana: Northern, Sigare Pan, 30 miles west of Nata river, altitude 896m,
26.iv. 1957, Drummond & Seagrief 5240 (K, SRGH); Nxai Pan National Park,
14.iii.1976, Smith 1654 (K).
Zimbabwe: Nuanetsi District, Nuanetsi Ranch, Bubi Section, 23.ii.1967, Cleghorn
1421 (K); Hlekwani, approximately 10 km south-west of Bulawayo, altitude 1350m,
19.xii.1990, Laegaard 1528 (K).
Distribution: Tropical Africa; introduced to tropical America.
7. Bothriochloa bladhii* (Retz.) S. T. Blake, in Proc. Roy. Soc. Queensl. 80: 62
(1969).-Clayton, In. Hepper (ed.), F!. West Trop. Afr. ed.2(3): 470 (1972).-Clayton, in
Polhill (ed.), Fl. Trop. East Mr., Gramineae (3): 719 (1982).-Lowe, The flora of
Nigeria, Grasses: 225 (1989).- Gibbs Russell et a!, Grasses of Southern Africa: 62
(1991).
Synonyms: Andropogon bladhii Retz., Obs. Bot. 2: 27 (1781); Andropogon
intermedius R. Br., Prdr.: 202 (1810). Hack., in DC., Monogr. Phan. 6: 485 (1889);
Andropogon punctatus Roxb., Hort. Beng.: 7 (1814) and F!. Ind. 1: 268 (1820);
Andropogon haenkei J. S. Presl, in C. B. Pres!, Reliq. Haenk. 1: 340 (1830); Rhap his
*
bladhii, named for Bladh, who collected the type in China.
248
stricta Nees, in Hook. J. Bot. Kew. Misc. 2: 99 (1850); Andropogon inundatus F.
Muell., Linnaea 25: 444 (1853); Andropogon leptanthus Steud., Syn. P1. Glum. 1:
391 (1854); Andropogon perfosus Nees & Meyen ex Steud., Syn P1. Glum. 1: 391
(1854); Bothriochloa anamitica 0. Kuntze, Rev. Gen. P1. 2: 762 (1891); Sorghum
intermedium (R. Br.) 0. Kuntze, Rev. Gen. P1. 2: 792 (1891), based on Andropogon
intermedius; Sorghum intermedium (R. Br.) 0. Kuntze var. haenkei (Presl) 0.
Kuntze, Rev. Gen. P1. 2: 792 (1891), based on Andropogon haenkei; Amphilophis
intermedia (R. Br.) Stapf, in Agric. News W. Indies 15: 179 (1916), based on
Andropogon intermedius; Amphilophis glabra (Roxb.) Stapf var. haenkei (PresL) E.
G. & A. Camus, in Lecomte, F!. Gen. Indo-Chine 7: 311 (1922), based on
Andropogon haenkei; Amphilophis haenkei (Presl) Haines, Bot. Bihar & Orissa: 1029
(1924), based on Andropogon haenkei; Bothriochloa intermedia (R. Br.) A. Camus,
Ann. Soc. Linn. Lyon n.s. 76: 164 (1931); S. T. Blake, in Univ. Queens!. Papers,
Dept. Biol. II No.3: 37 (1944), based on Andropogon intermedius; Bothriochloa
glabra (Roxb.) A. Camus, in Ann. Soc. Linn. Lyon n.s. 76: 164 (1931); Bor, Grasses
Burma Ceylon India & Pakistan: 107 (1960); Bothriochloa inundata (F. Muell.) J.M.
Black, Trans. & Proc. Roy. Soc. S. Aust. 60: 163 (1936), based on Andropogon
inundatus; Bothriochloa caucasica (Trin.) Hubbard, in Kew Bull. 1939: 101 (1939);
Bor, Grasses Burma Ceylon India Pakistan: 106 (1960); Bothriochloa glabra (Roxb.)
A. Camus subsp. haenkei (Presl) Henn, Blumea 3: 456 (1940), based on Andropogon
haenkei; Bothriochloa haenkei (Presl) Ohwi, Acta Phytotax. & Geobot. 11: 168
(1942), based on Andropogon haenkei; Dichanthium bladhii (Retz.) Clayton, in Kew
Bull. 32: 3 (1977).
A variable and tufted perennial. Culms rather straggling, up to im tall, culm nodes
glabrous, rarely bearded; culm internodes terete or channelled on one side. Leaf
blades to 30cm long and 10mm wide, tapering gradually to a slender tip, almost
glabrous; leaf sheaths bearded at the nodes; ligule a ring of white hairs.
249
Inflorescence a panicle, with numerous slender branches 2-5cm long, along a
common axis 3-10cm long; lower branches sometimes rebranched; rachis internodes
and pedicels with a translucent longitudinal groove, hairy, hairs up to 2.5mm long.
Spikelets usually dark purple or strongly flushed with purple; sessile spikelet 3-4mm
long, with a ring of short white hairs at the base; lower glume hairy below the middle
on the back, rarely glabrous, rigidly ciliate in the margins on the upper part, depressed
along the middle, with a deep or shallow pit in the upper part, the pit sometimes not
occurring in all spikelets, rarely absent altogether; upper glume lanceolate, margins
sparsely ciliate in the upper part or glabrous, depressed longitudinally on both sides
of the median keel, keel scabrid in the upper part. Florets 2: lower floret empty,
lemma oblong, obtuse, glabrous; upper floret bisexual, lemma stipiform; awn 1018mm long, geniculate, weakly twisted, glabrous; epaleate; lodicules 2, tiny, fleshy,
glabrous; anthers 3, 1.5 mm long; ovary oblong. Pedicellate spikelet usually smaller
than the sessile and reduced to 1 or 2 glumes, rarely well developed and with the
lower glume pitted.
Diagnostic characters and affinities: The only species of Bothriochloa with a
paniculate inflorescence. B. bladhii is often confused with Capillipedium species
because of the particulate nature of the inflorescence and the somewhat capillaceous
inflorescence branches. It has 10-25 jointed panicle branches compared to 1-10
jointed ones in Capillipedium species. Work on genetic barriers of B. bladhii (as B.
intermedia) has established that this species is capable of crossing with other species
of Bothriochloa and species of the related genera Dichanthium and Capillipedium
(De Wet & Harlan, 1966; Faruqi, 1969). This led De Wet & Harlan (1966) to unite
the three genera taxonomically using the biological species concept, and calling the
species a compilospecies.
Habitat: Streamsides, swamp margins and cracking clays.
Type: China, Bladh (LD, K!).
Representative specimens:
Bangladesh: Dacca District, Norsingdi, 23.ix.1975, Huq & Rahman H1624 (E).
250
Bhutan: Lingmenthang, Shongar Chu, altitude 950m, 02.vii.1979, Grierson & Long
2423 (E); Thimpu District, Ramtokto, Namselling, altitude 2300m, Parker 7060 (E).
Botswana: Tubu Island, east of Gumare, 15.i.1983, Astle 8228 (SRGH); Mahalapye,
Madibeng, altitude 3200ft, 11 .xii. 1959, De Beer 870 (BM); Tati river, south-east of
Francistown, altitude 980m, Long 12246 (E).
Malawi: Blantyre District, Chichiri Campus, University of Malawi, Limbe,
04.vii.1971, Banda 1160 (MAL); Kasungu Game Reserve, altitude 3300ft, 18.ii.1970,
Hall-Martin 563 (PRE); Lilongwe District, Diampwe river, 18.iv.1952, Jackson 764
(MAL); Ruo-Tuchila plain, Mulanje, altitude 2000ft, 22.v.1934, Lawrence 143
(MAL).
Mozambique: Chitengo, 13km to Acampamento, 02.v.1978, Diniz 164 (WAG); Sul
do Save, liha Mariana, 07.iv.1954, Myre & Cavalho 1776 (PRE).
Nepal: Between Kunjuri and Tambawa, 27°22'N 87°50'E, altitude 1500-1900m,
01.x.1989, Keke 1140 (E); near Maikot, altitude 8500ft, 08.x.1954, Stainton, Sykes &
Williams 4756 (E).
Zambia: Lusaka District, Mount Makulu, 30.i.1973, Delmotte 949 (SRGH);
Namwala District, Kafue river at Kialala, Kafue National Park, 08.xii.1963, Mitchell
24/29 (BM; SRGH); Ngomi area, Fort Jameson, iii. 1962, Verboom 588 (WAG).
Zimbabwe: Lomagundi, Saroi-Sinaia road, altitude 3 SOOft, ii. 1972, Davies 3180
(SRGH); Lomagundi District, Alaska, General Office, altitude 3985fl, 16.ii.1966,
Jacobsen 2827 (PRE); Bubi District, Gwampa vlei, altitude 3000ft, i.1956, Goldsmith
45/56 (PRE); Wankie District, Matetsi Safari Area LK 82096, altitude 3200ft,
14.iii.1978, Gonde 131 (PRE); Sipolilo, Nyamunyeche Estate, Gwenzi dam,
20.iii.1979, Nyariri 772 (PRE).
Distribution: Africa, India to Australia, and the Pacific; introduced to United States.
251
8. Bothriochloa pertusa* (L.) A. Camus, in Ann. Soc. Linn. Lyon 1930 n.s 76: 164
(193 1).-Bor, Flora of Assam 5: 368 (1940).-Bogdan, A revised list of Kenya grasses:
57 (1958).-Jackson & Wiehe, An annotated check-list of Nyasaland grasses: 31
(1958).-Bor, Grasses Burma Ceylon India & Pakistan: 109 (1960).
Synonyms: Holcus pertusus L., Mant. Alt.: 301 (1771); Andropogon pertusus (L.)
Willd., Sp. P1. 4: 922 (1806); Andropogon angustfolius Parl., F!. Paterm. 1: 269
(1845); Lepeocercis pertusa (L.) Hassk., P1. Jay. Rar.: 52 (1848); Elionurus pertusus
Nees ex Steud., Syn. P1. Glum. 1: 364 (1854); Amphilophis pertusa (L.) Nash ex
Stapf, in Agric. News W. Ind. 15: 179 (1916).
Perennial grass, shortly rhizomatous. Culms slender, up to 30cm tall, branched above,
suberect or geniculately ascending or often weak, trailing and rambling over the
ground, rooting and throwing up tufted or single shoots from the nodes; cuim nodes
exposed, hairy. Leaf blades mostly basal, lanceolate, up to 15cm long, 3-4mm wide,
flat, glabrous, pubescent or hirsute, often with tubercle-based hairs; leaf sheaths of
innovation shoots compressed, more or less keeled, the others terete, glabrous,
smooth, usually bearded at the mouth; ligule very short, truncate, an unfringed
membrane.
Inflorescence of digitate racemes, to 5cm long; main inflorescence axis shorter than
the racemes; racemes 6-10-jointed,' pedunculate; peduncles glabrous, 4mm long,
smooth; rachis internodes and pedicels with a longitudinal translucent groove, hairy,
hairs much longer at the tip, up to 3.5mm long. Sessile spikelet oblong, 4mm long
(including the callus); lower glume with a single very pronounced pit, hairy below the
middle, 3.5mm long, keels rigidly ciliolate upwards, flat on the back; upper glume
3.5mm long, lanceolate, acute, margins ciliate upwards. Florets 2: lower floret empty,
lemma oblong, obtuse, 2.8mm long; upper floret bisexual, lemma stipiform, 1.5mm
*
pertusus, having holes or slits, perforated, alluding to the glume pits.
252
long, awned from the tip; awn 18mm long; lodicules 2, fleshy, opaque, cuneate;
anthers not seen; ovary lanceolate. Pedicellate spikelet male, much like the sessile in
shape and size; glumes equal, 3.5mm long; lower glume sometimes shallowly pitted
(0-1 pit, rarely 2), glabrous on the back, ciliate on the lateral keels; upper glume
thinly membranous, narrowly lanceolate; lower lemma like that of the sessile spikelet,
a bit shorter; upper lemma linear-lanceolate; anthers 3, 1.7mm long.
Chromosome number: 2n = 40 (De Wet & Higgins, 1963; De Wet, Borgaonkar &
Richardson, 1963).
Diagnostic characters and affinities: Very close to B. inscuipta, except that B.
inscuipta has a glabrous lower glume of the sessile spikelet, and one to four shallow
pits on the lower glume of the pedicellate spikelet. B. pertusa is smaller with slightly
smaller inflorescence and shorter spikelets.
Habitat: Grassland on clay soils and open woodland.
Uses: Used as fodder and hay, and in erosion control.
Type: Type not seen. Type locality unknown.
Representative specimens:
Kenya: on the Machakos Talta Plains, 02.vii.1938, Pole Evans & Erens 1067 (B).
India: Dehra Dun, Saharanpur, 22.x.1888, Duthie 7686 (E).
Pakistan: Harara District, south-east of Haripur on road to Taxila, 09.iv.1958, Burt
B519 (E).
Somalia: 80km east of Hargesia, on road to Burao, 21.xii.1969, Lavranos 7392 (B).
Thailand: Ayuthia, 80km north of Bangkok, altitude lOOm, 12.iii.1958, S4renson,
Larsen & Hansen 2061 (E); Rachasima, 25km north of Korat, altitude 400m,
24.iii.1958, Sorenson, Larsen & Hansen 2477 (B).
Distribution: Eastwards from Arabia to Southeast Asia, Tropical Africa, and the
Caribbean Islands.
253
9. Bothriochloa kuntzeana* (Hack.) Henn, in Blumea 3: 456 (1940).-Bor, Grasses
Burma Ceylon India & Pakistan: 108 (1960).
Synonyms: Andropogon kuntzeanus Hack., in DC., Monogr. Phan. 6: 478 (1889);
Andropogon pertusa var. longfolia Hack., in DC., Monogr. Phan. 6: 482 (1889);
Hook.f., Flor. Brit. Ind. 7: 175 (1897); Eulalia clarkei Haines, Bot. Bihar & Orissa:
1017 (1924). Type: India, 28.ix.1873, Clarke 20557 (K!); Amphilophis kuntzeana
(Hack.) Haines, Bot. Bihar & Orissa: 1031 (1924); Bothriochloa longfolia (Hack.)
Bor, Grasses Burma Ceylon India & Pakistan: 108 (1960); Dichanthium kuntzeana
(Hack.) Jain & Deshpande, in Bull. Bot. Soc. Ind. 20: 134 (1979)
Perennial grass. Culms erect, solitary, up to 1.5m tall, moderately robust; culm nodes
bearded with a ring of silky hairs; culm internodes terete or slightly angular. Leaf
blades mostly cauline, linear, about 30cm long, scabrid on both surfaces; leaf sheaths
keeled, striate, sometimes with tubercle-based hairs (falling off early), tight, the lower
longer than the internodes, but upper shorter than the internodes; ligule a short
unfringed membrane.
Inflorescence of 9-15 subdigitate racemes, silky; main inflorescence axis angular,
glabrous; racemes 20 to many-jointed; raceme peduncles glabrous; racemes up to
8cm long; rachis internodes and pedicels with a longitudinal translucent groove, hairy
along the raised margins. Sessile spikelet 5.5mm long; lower glume lanceolate,
glabrous on the back, shortly ciliate on the keels towards the apex, with a single pit
above the middle, 5mm long; callus 0.5mm long, with long white hairs to 2mm long;
upper glume sharply keeled and scabrous along this keel towards the tip, glabrous,
mucronate, 5mm long, lanceolate, ciliate along the hyaline margins. Florets 2: lower
floret empty, lemma narrowly ovate, 4mm long; upper floret bisexual, lemma
stipiform, 2.5mm long, firm upwards, awned from the tip; awn 15mm long, bristle-
*
kuntzeana, named for 0. Kuntze, a famous grass botanist.
254
like, minutely hispidulous; epaleate; lodicules 2, fleshy, cuneate; anthers 3, 2.3mm
long; ovary ovate. Pedicellate spikelet male; lower glume with a single slit-like pit,
narrowly lanceolate, 5.5mm long, glabrous on the back and shortly ciliate on the
lateral keels towards the apex; upper glume almost flat on the back, 5.5mm long,
smooth; lower lemma oblanceolate, 4mm long; upper lemma absent; anthers 3,
2.5mm long; pedicel hairy, hairs up to 3.5mm long.
Chromosome number(s): 2n = 20, 40 (De Wet, Borgaonkar & Richardson, 1963).
Diagnostic characters and affinities: Characterised by long, silky racemes.
Habitat: Occurs in grass plains.
Type: India, 28.ix.1873, Clarke 20557 (K!).
Representative specimens:
India: Deccan, (no date), Ralph 496 (K); (no date and location given), Stocks sm. (E,
K); Deccan, v.1880, Woodrow s. n. (K).
Distribution: India (Madhya Pradesh, Bombay, Deccan Peninsula).
10. Bothriochloa insculpta* (Hochst.) A. Camus, in Ann. Soc. Linn. Lyon 1930 n.s.
LXXVI: 165 (193 1).-Sturgeon, A revised list of grasses of Southern Rhodesia Part
IV: 9 (1959).- Chippindall, in Meredith (ed.), The grasses and pastures of South
Africa: 483 (1955).-Bogdan, A revised list of Kenya grasses: 56 (1958).-Jackson &
Wiehe, An annotated check-list of Nyasaland grasses: 31 (1958).-Bor, Grasses of
Burma Ceylon India & Pakistan: 107 (1960).-Napper, Grasses of Tanganyika: 100
(1965).-Hood, A guide to the grasses of Zambia: 57 (1967).-Clayton, in Polhill (ed.),
F!. Trop. East Aft., Gramineae (3): 720 (1982).-Chaudhary, Grasses of Saudi Arabia:
419 (1989).-Gibbs Russell eta!, Grasses of Southern Africa: 63 (1991).
*
inscuiptus, engraved, cut into, with sunken markings, alluding to the glume pits.
255
Synonyms: Andropogon inscuiptus Hochst. ex A. Rich., Tent. Fl. Abyss. II: 458
(1851); Andropogon pertusus (L.) Wilid. var. capensis Hackel, in DC., Monogr. Phan
6: 482 (1889); Andropogon pertusus (L.) Wilid. var. inscuiptus subvar. trfoveolatus
Hackel, in DC., Monogr. Phan. 6: 482 (1889); Andropogon pertusus (L.) Wilid. var.
capensis Durand & Schinz, Consp. F!. Aft. V: 718 (1895); Andropogon pertusus (L.)
Wilid. var. inscuiptus Durand & Schinz, Consp. F!. Aft. V: 718 (1895); Andropogon
pertusus (L.) Wilid. var. capensis Stapf, in Dyer, Fl. Cap. VII: 345 (1897);
Amphilophis inscuipta (Hochst.) Stapf, in Pram, Fl. Trop. Aft. IX: 176 (1917).
Tufted perennial. Cuims up to 2m tall, either decumbent and rambling or developing
into stout woody stolons, branching and rooting at the prostrate nodes; culm nodes
bearded with a ring of white hairs; internodes channelled on one side. Leaf blades 430cm long, 2-8mm wide, glabrous on both surfaces, occasionally with sparse
tubercle-based hairs on the lower surface, margins rough; leaf sheaths glabrous,
striate, clasping the internodes tightly, bearded at the mouth; ligule an unfringed
membrane, acute.
Inflorescence of 3-20 subdigitate pedunculate racemes; main inflorescence axis up to
3cm long, bearded at the joints; racemes 10-20-jointed; raceme peduncles glabrous;
raceme length variable but up to 10cm long; lower racemes longer than the main
inflorescence axis; rachis internodes and pedicels with a longitudinal translucent
groove, hairy on the margins, hairs increasing upwards to 3mm long. Sessile spikelet
narrowly elliptic, 4.5mm long; lower glume with a single deep pit on the back just
above the middle, glabrous, shortly ciliate on the lateral keels towards the apex, 4mm
long; callus bearded, obtuse, 0.5mm long; upper glume sharply carinate, minutely
ciliate on this keel towards the apex, glabrous, narrowly lanceolate, 4.5mm long.
Florets 2: lower floret empty, lemma 3.5mm long, broadly ovate; upper floret
bisexual, lemma stipiform, tapers downwards, 2mm long, awned from the tip; awn
25mm long, glabrous; epaleate; lodicules 2, fleshy, cuneate; anthers 3, 1.8mm long;
ovary ovate. Pedicellate spikelet neuter, glabrous; lower glume with 0-4 slit-like pits
256
(shallow), glabrous, oblong, 5nm long, with an irregular apex, sharply ciliate along
the lateral keels; upper glume lanceolate, 4.5mm long, flimsy with ciliate margins;
lower lemma oblong, truncate, 3mm long; pedicel hairy, with a longitudinal
translucent groove.
Chromosome number(s): 2n = 50 (De Wet, Borgaonkar & Richardson, 1963); 60
(De Wet & Higgins, 1963).
Diagnostic characters and affinities: This species may be confused with B. pertusa.
It differs in having the lower glume of the pedicellate spikelet always with 1-4 pits
and that of the sessile spikelet always glabrous on the back.
Habitat: Overgrazed grasslands and weedy places.
Uses: Cultivated fodder; native pasture species.
Type: Ethiopia, Mt. Sholloda (Selleuda), Schimper 80 (K!).
Representative specimens:
Botswana: Northwest District, Tubu Island, east of Gumare, 15.i.1983, Astle 8228
(SRGH); South-east District, altitude 1000m, 04.iii.1987, Long & Rae 34 (E);
Ngamiland District, altitude 910m, 31 .iii. 1987, Long & Rae 510 (E); Northwest
District, Savuti-Selinda road, 26.ii.1983, Smith 4141 (SRGH).
Malawi: Ncheu District, South Ncheu, 20.iv. 1956, Jackson 1851 (MAL).
Mozambique: Gaza District, Chibuto, Maniquenique, Estacão Experimental do
C.I.C.A., 13.vi.1950, Lemos & Balsinhas 100 (PRE, WAG); S'uldo S'ave, Licilo,
10.vii.1947, Pedro & Pedrogao 1404 (SRGH).
Zambia: Munibwa District, Lubungu Pontoon, Kafue river, Kafue National Park,
07.iv.1963, Mataundi 19/80 (SRGH).
Zimbabwe: Gokwe District, 3 miles north of Gokwe, 30.v.1963, Bingham 674
(SRGH); Balulima-Mangwe District, Dombodema Mission Farm, Tjompani Store,
altitude 1300m, 29.iv.1972, Norrgrann 145 (WAG); Lower Sabi, altitude 1500ft,
27.i-2.ii. 1948, Rattray 1262 (WAG).
257
Distribution: Tropical Africa, Arabia and India; introduced to Australia.
8.2.13. Euclasta Franch., in Bull. Soc. Hist. Nat. Autun 8: 335 (1895). The type
species is Euclasta condylotricha (Hochst.) Stapf.
Synonym: Indochloa Bor, in Kew Bull. 9: 75 (1954).
Inflorescence terminal and axillary, of delicate solitary or subdigitate pedunculate
racemes; racemes consist of spikelet pairs, one sessile, the other pedicellate; 1-3 large
homogamous spikelet pairs at the base of the raceme; pedicellate spikelet larger than
the sessile; rachis intemodes and pedicels with a translucent longitudinal groove;
lower glume papery, more or less flat; lower floret empty with a hyaline lemma;
upper floret bisexual, lemma entire, stipiform; pedicellate spikelet male or neuter.
Species 3.
A key to the species:
Inflorescence of subdigitate pedunculate racemes; homogamous spikelet pairs male or
1. condylotricha
neuter.
Inflorescence a delicate solitary raceme (rarely two); homogamous spikelet pairs
neuter:
Raceme peduncle glabrous; sessile spikelet 3.5mm long.
Raceme peduncle pilose; sessile spikelet 5mm long.
2. clarkei
3. oligantha
1. Euclasta condylotricha (Hochst.) Stapf, in Pram, Fl. Trop. Aft. IX: 181 (1917).Jackson & Weihe, An annotated check-list of Nyasaland grasses: 41 (1958).-Napper,
Grasses of Tanganyika: 100 (1965).-Hood, A guide to the grasses of Zambia: 65
(1967).-Clayton, in Hepper (ed.), Flora of West Tropical Africa, ed.2 (3): 471
258
(1972).-Clayton, in Polhill (ed.), Flora of Tropical East Africa, Gramineae(3): 722
(1982).-Lowe, Flora of Nigeria, Grasses: 237 (1989).
Synonyms: Andropogon condylotrichus Hochst., in Steud., Syn. P1. Glum. 1: 377
(1854); Andropogonpiptatherus Hack., in Mart. et Eichl., Fl. Bras. 2: 3, 293 (1883);
Sorghum piptatherum (Hack.) 0. Ktze., Rev. Gen. P1. 2: 792 (1891); Euclasta
glumacea Franch., in Bull. Soc. Hist. Nat. Autun 8: 336 (1895); Euclasta graminea
Th. et Hel. Durand, Syll. Fl. Congol.: 649 (1909); Amphilophis piptatherus (Hack.)
Nash, in North Amer. Fl. 17: 127 (1912).
Annual. Culms up to 1.5m tall, ascending but geniculate at the base, always bearded
at the nodes. Plant shallowly rooted and rooting at the nodes (stilt rooted). Leaf
blades linear-lanceolate, somewhat flaccid, up to 25cm long, 2-10mm wide,
contracted at the base and drawn to a setaceous tapering point, glabrous on both
surfaces or with very short sharp hairs along the midrib and immediate adjacent
veins; midrib slender, whitish and prominent; leaf sheaths striate, glabrous, closely
clasping the internodes for about half their length, with long tubercle-based hairs at
the mouth, in the collar region; ligule ciliolate.
Inflorescence of subdigitate racemes; rachis internodes and pedicels with a
translucent longitudinal groove; lowest 1-3 spikelet pairs of each raceme
homogamous, male or neuter. Sessile spikelet bisexual; lower glume pale to whitish,
elliptic-oblong, minutely truncate, 4mm long, with S nerves ending below the hyaline
apex, hairy on the margins and lower half, 2-keeled, the keels scaberulous; upper
glume ovate, strongly carinate, opaque, glabrous, 4mm long. Florets 2: lower floret
empty, lemma ovate, carinate and ciliate along this keel and the margins; upper floret
bisexual, lemma stipiform; palea present, broadly ovate, hyaline: awn 37mm long,
minutely hairy; lodicules 2, fleshy, glabrous; anthers 3, 0.75mm long; ovary ovate,
2mm long. Pedicellate spikelet male; lower glume oblanceolate, many nerved (7-10nerved), hairy on the lower 1/3, 5.5mm long; upper glume 4-nerved, 4-4.5mm long.
259
Upper floret male, lemma slightly shorter than the glume; anthers 3, linear, 1.5mm
long; lodicules 2, fleshy; pedicel grooved, hairy on the sides.
Chromosome number: 2n = 20 (Davidse & Pohl, 1974; Olorode, 1975); 40 (Bor,
1960).
Diagnostic characters and affinities: The flaccid leaves, stilt roots and the
subdigitate inflorescence distinguishes this species from the other Euclasta species.
Habitat: Occurs mainly on roadsides and shallow rock escarpments.
Type: Ethiopia, Dscheladscheranne, Schimper 2011 (BM!).
Representative specimens:
Zaire: Mweue District, Kakimba, altitude 120m, 09.v.1957, Liken 2876 (BM).
Zambia: Fort Jameson, Jumbe area, 24.iii.1963, Verboom 941 (BM).
Zimbabwe: Gokwe District, 10 miles from Gokwe, near top of basalt escarpment in
Bonye Gorge, 25.iii.1963, Bingham 571 (BM); Gatooma District, Sanyati Reserve,
12.iii.1959, Cleghorn 451 (BM).
Distribution: Tropics of Africa.
2. Euclasta clarkei* (Hack.) T. A. Cope, in Kew Bull. 35(3): 704 (1980).
Synonyms: Andropogon clarkei Hack., in Ost. Bot Zeit. 41: 49 (1891); Dichanthium
clarkei (Hack.) Haines, Bot. Bihar & Orissa: 1040 (1924); Indochloa clarkei (Hack.)
Bor, in Kew Bull. 9: 76 (1954).
A straggling annual with very slender culms, often rooting from the nodes. Culms up
to 45cm tall, shiny, glabrous except at the bearded nodes, branched. Leaf blades
*
clarkei, named for C.B. Clark, who collected the type specimen in Parasnath, India.
260
linear-lanceolate, up to 8cm long and 7.5mm wide, hairless, scabrous, with serrated
margins; lower leaf sheaths lax, and slipping from the culms, upper leaf sheaths
clasping the cuims very tightly and somewhat inflated, often with inflorescence in the
axil, smooth and glabrous, markedly striate; ligule membranous, 1.5mm long, soon
splitting into rigid hairs.
Inflorescence a single raceme, greenish in colour; raceme peduncle glabrous; lowest
pair of spikelets homogamous, neuter, each pair consisting of a sessile and pedicellate
spikelet, upper pairs differ in sex, with the sessile bisexual and the pedicellate neuter;
raceme terminates in a triplet of two pedicellate spikelets and one sessile spikelet;
rachis internodes with a translucent longitudinal groove, covered with dense white
hairs. Sessile spikelet bisexual; lower glume 3.4mm long, flat or slightly depressed on
the back, 5-nerved, covered with small glistening white hairs, has inflexed margins
and slightly winged keels, elliptic-acute or truncate; upper glume 3.5mm long, 3nerved, hyaline, glabrous, keeled on the back and sides, ovate-truncate, ciliate on the
inflexed margins and on the median keel (on the back). Florets 2: lower floret empty,
lemma hyaline, 3-3.5mm long, ciliate at the tip; upper floret bisexual, stipiform; awn
geniculate, 2.5cm long, with a twisted column; epaleate; lodicules 2, fleshy, glabrous;
anthers 3, 1mm long; ovary oblong, 2mm long. Pedicellate spikelet neuter; pedicels
hairy, but glabrous on the sessile spikelet side, has a translucent longitudinal groove;
lower glume broadly elliptic-acute, many-nerved, 5mm long, glabrous; upper glume
4mm long, 3-nerved; lower lemma 4mm long, nerveless, ciliate at the tip.
Habitat: Common in shaded places.
Type: India, Parasnath, altitude 4200ft, 07.x.1883, Clarke 33700 (K!).
Representative specimens:
India: Maha Deo caves, altitude 4100ft, 21.x.1958, Bharadwaja 32 (K); Mt. Abu,
14.x.1916, Blatter 2597 (K); Maharastetra, Katepani forest, Ambavane, Poona
District, 26.x.1964, Reddi 99496 (K).
261
Oman: the Fort Nashib road, altitude lOOft, 03.x.1992, Collenette 8382 (K); Dhofar,
altitude 750m, 07.xi.1991, Cope 536 (E); Dhofar, 12km on the lower Dhalqut road,
02.x.1993, McLeish 2835 (K).
Distribution: India and Oman.
3. Euclasta oligantha (Hochst. ex Steud.) Setshogo comb. nov.
Synonyms: Andropogon oliganthus Hochst. ex Steud, Syn. P1. Glum 1: 368 (1854).
Heteropogon oliganthus (Hochst. ex Steud.) Blatter & McCann, in J. Bombay Nat.
Hist. Soc. 32: 623 (1928); Indochloa oligantha (Hochst. ex Steud.) Bor, in Kew Bull.
9: 79, t. 2 (1954); Dichanthium oliganthum (Hochst. ex Steud.) T.A. Cope, in Kew
Bull. 35(3): 703 (1980).
An annual of up to 20cm tall, erect. Culms slender; culm nodes hidden, with long
white hairs where exposed; culm internodes also hidden, glabrous where exposed.
Leaf blades linear-acuminate, up to 4.5mm wide and up to 5cm long with thickened
cartilaginous and serrated margins, and a thickened mid-nerve drawn into a median
keel of the culm sheaths, hairy with tubercle-based hairs (becoming hirsute); leaf
sheaths clasping the culms, hairy towards the margins, glabrous on the back; ligule an
unfringed membrane, continues into the flimsy inflexed leaf sheath margins.
Inflorescence consists of a single raceme, rarely two; inflorescence peduncle pilose;
lowest 2-3 spikelet pairs homogamous, neuter; rachis internodes and pedicels with a
translucent longitudinal groove, flattened, densely ciliate on the margins. Sessile
spikelet: lower glume 4.5mm long, rounded on the back, oblong-truncate, covered on
the sides and base with short white hairs, slightly two-keeled towards the apex; upper
glume boat-shaped, slightly longer than the lower (5mm long), keeled towards the tip,
glabrous and shiny, ciliate towards the tip on the slightly inturned margins. Florets 2:
lower floret empty, lemma hyaline, 2.5mm long; upper floret bisexual, stipiform,
2mm long; awn 3.8cm long, column minutely ciliate and bristle scabrid; palea a very
262
small oblong or obtuse hyaline scale, oblanceolate, 2mm long; lodicules 2, fleshy,
cuneate; anthers 3, 1mm long; ovary oblanceolate, 2mm long. Pedicellate spikelet
neuter; lower glume 5.5mm long, elliptic-truncate with inturned hyaline margins,
ciliate on the upper half and serrated on the outer edges of the intumed margins, many
nerved; upper glume 4.5mm long, oblanceolate, with an acute apex, ciliate on the
inturned margins.
Diagnostic characters and affinities: This species differs from the other two
Euclasta species by having the hairy inflorescence peduncle.
Habitat: Common in overgrazed grasslands or along paths.
Type: India, Nilgri Hills, Hohenacker 1288 (K! BM!).
Representative specimens:
India: Madras, Anaimudi ridge, Travancore High Range, altitude 7000ft, xii.1963,
Barnes 663 (K); Madras, Pukaa, altitude 6000ft, x.1886, Gamble 18308 (K); Madras
(Tamilnadu), Dindigul District, Levinge path, altitude 2300m, 18.xii.1989, Mathew &
Mathew 54067 (K); Kalbutty bababood, altitude 6000ft, x.1908, Meebold 10763 (K);
Kerala, Eravikulan National Park, altitude 2100m, 27.xi.1979, Rice 55 (K).
Distribution: Southern India.
8.2.14. Dichanthium Willemet, in Usteri, Ann. Bot. 18: 11 (1796).-De Wet &
Harlan, in Bol. Soc. Arg. Bot. 12: 206-227 (1968). The type species is Dichanthium
annulatum (Forssk.) Stapf.
Synonyms: Lepeocercis Trin., Fund. Agrost.: 203 (1822); Diplasanthum Desv.,
Opusc.: 66 (1831); Eremopogon Stapf, in Pram, F!. Trop. Afr. IX: 182 (1917).
Annuals or perennials. Inflorescence of single, digitate or subdigitate racemes;
racemes rarely solitary but more often in pairs-to-many, arranged on a short primary
axis; pedunculate or subsessile, sometimes with the lower branches divided; racemes
263
always terminate in a triad of 1 sessile and 2 pedicellate spikelets; spikelets highly
imbricate, paired, one sessile, the other pedicellate, subequal to equal in size and
shape, usually differing in sex, except the lower 1-6 pairs, which are with rare
exceptions homogamous male or neuter; rachis intemodes and pedicels solid. Sessile
spikelet callus obtuse; lower glume papery to cartilaginous, broadly convex to
slightly concave, sometimes pitted, acute to broadly obtuse; upper lemma hyaline
base of the awn, entire; awn glabrous. Pedicellate spikelet much like the sessile, male
or neuter, awnless, rarely bisexual and awned (e.g. D. annulatum var. fecundum).
Species ± 20.
A key to taxa sampled:
Inflorescence a terminal solitary raceme; lower glume of sessile spikelet pitted,
glabrous:
Lower glume of pedicellate spikelet with a conspicuous pit. 1. micranthum
Lower glume of pedicellate spikelet never pitted.
2.foveolatum
Inflorescence of 2-many digitate or subdigitate racemes; lower glume of sessile
spikelet not pitted:
Inflorescence of pedunculate subdigitate racemes:
Lowest 1-6 spikelet pairs homogamous; raceme peduncles and
flowering cuim glabrous:
Cuim nodes bearded; rachis internodes hairy; lower glume of
sessile spikelet concave, pubescent in the lower third, with
long tubercle-based cilia above the middle, along the margins
and forming a subapical fringe.
3. annulatum
Cuim nodes glabrous; rachis intemodes glabrous; lower glume
of sessile spikelet narrowly winged on the lateral keels, without
tubercle-based cilia.
264
4. aristatum
Lowest 1-3 spikelet pairs homogamous; raceme peduncles and
5. caricosum
flowering culm pubescent.
Inflorescence digitate with 1-15 sessile to subsessile white villous racemes.
6. sericeum
1. Dichanthium micranthum T. A. Cope, in Pub!. Cairo Univ. Herb. 78: 325 (1977).
A perennial tussock grass of up to 30cm tall, shortly rhizomatous. Cuims slender,
branched upwards; culm nodes exposed, mostly glabrous (some bearded); culm
internodes glabrous. Leaf blades linear-acute and mostly basal, forming the tussocks,
glabrous, 1.5-2.5mm wide; leaf sheaths with a few tubercle-based hairs, otherwise
glabrous, lower older sheaths lax; ligule a ciliate membrane.
Inflorescence a solitary raceme, 10-12-jointed. Sessile spikelets narrowly elliptic and
shorter (2.5mm long) than the pedicellate ones (3mm long); lower glume with a
single pit on the back, 2mm long, slightly concave to flat on the back, glabrous; upper
glume slightly longer than the lower (2.5 mm long), with a single median keel,
glabrous. Florets 2: lower floret empty, lemma hyaline; upper floret bisexual, lemma
stipiform; awn 20mm long, minutely pubescent, with a twisted base; anthers 3,
1.25mm long. Pedicellate spikelet longer than the sessile one (3mm long), obovate in
shape, male; lower glume with a conspicuously deep pit, with very fine short teeth
(pectinato-ciliatis) on the margins towards the apex; anthers 3, 2mm long.
Diagnostic characters and affinities: This species is very similar to D. foveolatum in
having solitary racemes but differs in the smaller spikelets and shorter callus hairs,
and the longer and broader pedicellate spikelets than the sessile ones. Also the lower
glume of the pedicellate spikelet is always pitted in D. micranthum but never pitted in
D. foveolatum.
Habitat: Occurs in coastal areas, and occasionally in rocky hill slopes.
265
Type: Oman, Plain behind Salalah, Dhufar, 26.ix. 1943, Vesey -FitzGerald 12316/1
(K!).
Representative specimens:
Oman: Raysat, 01.xi.1983, Lawton 2513 (E); Zeak, 05.vii.1982, Maconochie 3570
(E); Taqua, altitude 10m, 11 .x. 1984, McLeish 101(E); Dhofar, wadi 5km east of
Salalah, altitude 2m, 09.ix.1985, Miller 7546 (E); Dhofar, altitude lOm, 01.xi.1989,
Miller & Nyberg M.9020 (E).
Distribution: Arabia.
2. Dichanthiumfoveolatum* (Delile) Roberty, in Boissiera 9: 170 (1960).-Clayton &
Renvoize, in Polhill (ed.), Fl. Trop. East Afr., Gramineae (3): 723 (1982).-Cope, in
Nasir & Au, F!. Pak. 143: 279 (1982).-Chaudhary, Grasses of Saudi Arabia: 415
(1989).-Cope & Hosni, A key to Egyptian grasses: 54 (1989).
Synonyms: Andropogon foveolatus Delile, Memoires Botaniques, extracts de la
Descriptionde L'Egypte, Paris: 16 (1813); Eremopogonfoveolatus Stapf, in Pram, F!.
Trop. Afr. IX: 183 (1917).
A tufted perennial of about 50cm tall, with very short rhizomes. Cuims slender and
geniculate; cuim nodes exposed and bearded. Leaf blades linear, flat, glabrous except
on upper surface in the ligule area; leaf sheaths glabrous; ligule a ciliate membrane.
Inflorescence a terminal solitary raceme issuing from the tip of the cuims and
branches; rachis internodes and pedicels fihiform, hairy except on the spikelet side;
lowest 0-1 spikelet pair homogamous, male or neuter. Sessile spikelet lanceolate,
shinning with a reddish tip; callus hairy, hairs extend to about '/2 the length of the
spikelet; lower glume glabrous, always with a pit just above the middle, 4mm long;
*
foveolatus, minutely pitted, alluding to the pitted glumes.
266
upper glume carinate, 3-nerved. Florets 2: lower floret empty, lemma hyaline, 3mm
long, acute, emarginate; upper floret bisexual, lemma stipiform, 2mm long; awn
14mm long with a twisted brown column; lodicules 2, fleshy; anthers 3, 3mm long;
ovary oblong. Pedicellate spikelet about equal or slightly longer than the sessile;
glumes as for the sessile spikelet except that the lower glume is never pitted.
Chromosome number: 2n =40 (Mehra, Subramanyam & Swaminanthan, 1962).
Diagnostic characters and affinities: Similar to D. micranthum, except that the
lower glume of the pedicellate spikelet is never pitted in D. foveolatum.
Habitat: Sandy and rocky low lying areas subject to occasional flooding.
Type: Egypt, Delile (K!).
Representative specimens:
Bahrain: Jebel Dukha, 26.vii.1984, Alder 25 (E).
Oman: Wadi Dawkah, altitude 500m, 21.ix.1983, Gallagher 6898/A (E).
Pakistan: Baluchistan, 13.iv.1965, Lamond 443 (E); Sind, 05.v.1965, Lamond 872
(K).
Somalia: Ambouli Palm Grove, near Djibouti, 14.iv.1974, Lavranos 11481 (E).
Yemen: Hodeida, altitude 50m, 02.x.1978, Miller 302 (E).
Distribution: Arabia and East Africa.
3. Dichanthium annulatum* (Forssk.) Stapf, in Pram, F!. Trop. Afr. IX: 178 (19 17).Bor, Flora of Assam 5: 371 (1940).-Sturgeon, A revised list of grasses of Southern
Rhodesia Part IV: 10 (1954).-Bogdan, A revised list of Kenya grasses: 57 (1958).-
*
annulatus, marked with rings, surrounded by raised rings or bands, having a ring, probably
alluding to the subapical and marging rings of hairs on the lower glumes.
267
Bor, Grasses Burma Ceylon India & Pakistan: 133 (1960).-Hood, A guide to the
grasses of Zambia: 65 (1967).-Clayton, in Hepper (ed.), Fl. West Trop. Mr. (ed.2) 3:
471 (1972).-Clayton, in Polhill (ed.), F!. Trop. East Mr., Gramineae (3): 725 (1982).Chaudhary, Grasses of Saudi Arabia: 417 (1989).-Cope & Hosni, A key to Egyptian
grasses: 54 (1989).-Gibbs Russell et al, Grasses of Southern Africa: 105 (1991).
Synonyms: Andropogon annulatus Forssk., F!. Aegypt.- Arab.: 173 (1775);
Andropogon bladhii Retz., Obs. Bot. 2: 27 (1781); Andropogon scandens Roxb.,
Hort. Beng.: 7 (1814); Andropogon obtusus Nees, in Hook. et Am., Bot. Beech Voy.:
243 (1838); Lepiocercis annulata (Forssk.) Nees, F!. Mr. Aust.: 98 (1841);
Andropogon papillosus Hochst., in Rich., Tent. F!. Abyss. 2: 457 (1851);
Andropogon annulatus var. monostachya F. Muell. ex Benth., F!. Austr. 7: 531
(1878); Andropogon annulatus var. bladhii (Retz.) Hack., in DC., Monogr. Phan. 6:
572 (1889); Andropogon annulatus var. decalvatus Hack., in DC., Monogr. Phan. 6:
572 (1889); Andropogon grandispiculatus Domin, Biblioth.. Bot. 85: 269 (1915);
Dichanthium papillosum (Hochst.) Stapf, in Pram, F!. Trop. Afr. IX: 179 (1917);
Dichanthiumfecundum S. T. Blake, in Univ. Queens!. Papers, Dept. Biol. II No.3: 51
(1944); Dichanthium annulatum var. decalvatum (Hack.) Maire et Weiller, Fl. de
L'Afr. du Nord 1: 282 (1952).
This species has three varieties keyed out as follows:
Some pedicel!ate spikelets on each raceme bisexual and often shortly awned.
var. frcundum*
Pedicel!ate spikelets always male or neuter.
Spikelets pilose, with a distinct subapical fringe of long tubercle-based cilia.
var. papillosum
*
fecundus, fertile, alluding to the fertile pedicellate spikelets.
Spikelets less pilose, with tubercle-based cilia mainly along the margins of
var. annulalum
the lower glume.
A tufted, decumbent or erect perennial of up to 2m tall with distinct short rhizomes.
Culms woody and persistent, robust, geniculate ascending, simple or branched; culm
nodes exposed and with a ring of hairs; cuim internodes glabrous. Leaf blades linearlanceolate, 4-4.5mm wide and up to 30cm long, glabrous along most of their length
with long hairs in the ligule area; leaf sheaths glabrous, or sparsely hairy all over;
ligule an unfringed membrane, about 2mm long.
Inflorescence consists of subdigitate 2-to-many racemes arranged on a short primary
axis; rachis internodes and pedicels solid, usually hairy. Sessile spikelets imbricate,
oblanceolate, 2-6mm long, the lowest 1-6 usually homogamous, male or neuter, the
upper bisexual; lower glume slightly concave, pubescent below the middle, with long
tubercle-based cilia above the middle, along the margins and across the apex; upper
glume slightly shorter than the lower, glabrous except for the keel and apex which are
often shortly ciliate. Florets 2: lower floret empty, lemma ovate, 3.8mm long, hyaline;
upper floret bisexual, lemma stipiform, about 3mm long; awn 25mm long, both
column and bristle minutely hispidulous; lodicules 2, fleshy, cuneate, highly
vascularised; anthers 3, 1.5mm long; caryopsis elliptic-oblong. Pedicellate spikelets
equal or subequal in size and shape to the sessile ones, male or neuter or bisexual.
When bisexual, the pedicellate spikelet was found to be larger than the sessile one in
all its parts.
Chromosome number(s): 2n = 20 (De Wet & Harlan, 1966; De Wet, Borgaonkar &
Richardson, 1963); 40 (De Wet, Borgaonkar & Richardson, 1963).
Diagnostic characters and affinities: The three varieties are distinguished mainly on
the basis of the glume characteristics and the sexuality of the pedicellate spikelets. D.
*
papillosus, referring to the tubercles (pimples, papillae) from which the cilia arise.
269
annulatum var. fecundum is the only other member of the subtribe (in addition to
Sorghastrum fuscescens and S. pogonostachyum) which has bisexual pedicellate
spikelets.
Habitat: Dry open places subject to overgrazing or disturbance. Also in swampy
depressions.
Type: Egypt, Forsskál (C).
Representative specimens:
var. annulatum:
Africa:
Somalia: Bosaso, 29.xi.1971, Lavranos 9049 (E).
Middle East:
Iran: S. Lorestan-Sheshom, 32°40'N 48°15'E, altitude 300-350m, 25.iv.1963, Jacobs
6411 (E).
Iraq: Baghdad, 7km on Hillah road, 29.ix.1954, Wheeler Haines 36 (E).
Israel: Lower Galilee, between Tabigha and Rosh Pinna, altitude Om, 18.iv.1957,
Lorch & Grizi 714 (E).
Oman: Wadi Bani Khalid, Shangiyah, 22°34'N 59°06E, altitude 550m, 14.iv.1983,
Gallagher 6733/7 (E).
Saudi Arabia: Sandia city compound, Jiddah, altitude 20ft, 27.ix.1983, Collenette
4584 (E).
Turkey: Hatay, Dortyol to Iskenderun, altitude 1 OOm, 06.v. 1965, Goode & Jones 565
(B).
Yemen Arab Republic: Taiz Province, Bani Sheba, west of Turbah, altitude 1200m,
18.iii.1984, Miller & King 5261 (E); near Shemlaan near San's, altitude 2300m,
15.viii.1977, Wood 1806 (E).
South East Asia:
270
Pakistan: Harara, Abbttabad, Nasrav Valley in hills east of town, altitude 4200ft,
22.v.1958, Burt 664 (E); Baluchistan, Bela, Diwana to Hab., 0l.v.1965, Lamond 798
(E).
Nepal: Mayangdi Khola (BENI), altitude 8200ft, 22.iv.1954, Stainton, Sykes &
Williams 111(E).
var. papillosum:
Botswana: North-east District, altitude 1000m, 19.iv.1987, Long & Rae 802 (K);
North-west District, 28.i.1980, Smith 3498 (K).
Zambia: Livingstone Game Park, 14.ii.1961, Mitchell 5/78 (K).
Zimbabwe: Belingwe, altitude 3000ft, 01 .xi. 1973, Gosden 10 (K); Gokwe District,
Sengwa Research Station, 10.i.1975, Guy 2256 (K).
var. fecundum:
Australia:
Northern Territory: Q. T. Station, 1697'S 135°3'E, altitude 177m, 09.v.1947, Blake
17685 (K); Timber Creek, 19.v.1971, Byrnes 2208 (K); 1 mile north of Wollogorang
Station, 02.vi.1948, Perry 1158 (K); 2 miles south of McArthur River Station,
25.vii.1948, Perry 1738 (K); Brunette Downs, 13.i.1972, Redhead Mck B282 (K).
Queensland: Forest Home Station, Gilbert river, 01.1.1931, Brass 1817 (K); North
Kennedy District, Lansdown Pasture Research Station, 35 miles south of Townsville,
06.iv.1965, Lazarides 7176 (E); Burke District, "Sutherland", 45 miles north-west of
Maxwelton, 26.i. 1966, Pedley 1944 (K).
Distribution: This species is the widely distributed of Dichanthium species. Var.
annulatum extends from western North Africa to Indonesia. Introduced to central
Africa and Australia (De Wet & Harlan, 1966). Var. papillosum occurs in the tropics
and subtropics of Africa, and var. fecundum is Australian in distribution (widely
distributed in the more tropical regions of Western Australia, Northern Territory and
Queensland).
271
4. Dichanthium aristatum* (Poir.) C. E. Hubbard, in Kew Bull. 1939: 654 (1939).Sturgeon, A revised list of grasses of Southern Rhodesia Part IV: 10 (1954).Chippindall, in Meredith (ed.), The grasses and pastures of South Africa: 481 (1955).Bor, Grasses Burma Ceylon India & Pakistan: 134 (1960).-Clayton, in Polhill (ed.),
F!. Trop. East. Aft., Gramineae (3): 723 (1982).-Gibbs Russell et a!, Grasses of
Southern Africa: 105 (1991).
Synonyms: Dichanthium nodosum Willem., in Usteri, Ann. Bot. 18: 11 (1796);
Andropogon aristatus Poir, in Lamk., Encycl. Met. Bot. Suppi. 1: 585 (1810);
Andropogon mollicomus Kunth, Rev. Gram. 1: 365 (1830); Diplasanthum lanosum
Desv., Opusc.: 67 (1831); Lepiocercis mollicoma (Kunth.) Nees, Edinb. New.
Philom. Joum. 18: 185 (1835); Andropogon caricosus subsp. mollicomus Hack. var.
mollicomus (Kunth.) Hack., in DC., Monogr. Phan. 6: 569 (1889); Andropogon
nodosus (Wilem.) Nash, N. Amer. Fl. 17: 122 (1912); Dichanthium caricosum var.
mollicomus (Hack.) Haines, Bot. Bihar & Orissa 5: 1039 (1924).
A tufted perennial of up to 1.1 m tall, shortly stoloniferous. Cuims robust, erect,
sometimes decumbent; cuim nodes exposed and glabrous. Leaf blades linearlanceolate, up to 25cm long and 3mm wide, glabrous except at the base in the ligule
vicinity where there are long tubercle-based cilia; leaf sheaths glabrous; ligule a
scarious membrane.
Inflorescence of 1-5 racemes arranged subdigitately on a short primary axis; raceme
peduncles and cuim below the inflorescence pubescent; rachis internodes and pedicels
solid; rachis joints hairy; rachis internodes glabrous and pedicels hairy on one side
(glabrous on the side facing the sessile spikelet); lowest 1-6 spikelet pairs
homogamous, male or neuter. Sessile spikelet obovate, 2-5mm long; lower glume
5mm long, pilose below the middle, glabrous or shortly ciliate above the middle
*
aristatus, awned, abruptly terminated in a hard, straight, subulate point of various lengths.
272
along the margins and near apex, narrowly winged towards the apex; upper glume as
long as the lower, glabrous or ciliate along the margins and keel. Florets 2: lower
floret empty, lemma hyaline; upper floret bisexual, lemma stipiform; awn 22mm
long, minutely pubescent; callus short, blunt; lodicules 2, fleshy; anthers 3, 2mm
long; ovary ovate. Pedicellate spikelet male or neuter, subequal in size and shape to
the sessile one.
Chromosome number(s): 2n = 20, 40 (De Wet, Borgaonkar & Richardson, 1963);
60 (Fedorov, 1974).
Diagnostic characters and affinities: Characterised by robust, suberect plants
having the peduncles to the racemes villous, and the cuim below the inflorescence
strongly pilose. The species is related to and sometimes regarded as conspecific with
D. caricosum.
Habitat: Damp places in disturbed ground.
Uses: Fodder grass.
-
Type: Mauritius, Commerson (Fl).
Representative specimens:
South Africa: Transvaal: Pretoria, 13.i.1935, Chippindal 21(K); Pretoria, Prinshof
Experiment Station, iii.1943, Codd 2226 (K); Bon Accord Dam picnic area,
07.xii. 1980, Crook 2347 (K); Ubombi District, Pongolo Poort dam site, altitude 200250ft, Ward 3808 (E).
Zambia: Luangwa Valley, near the Lunda Plain, altitude 2500ft, 04.v.1966, Astle
4865 (K); Mpika District, Luangwa Game Reserve, 27.iv.1965, Mitchell 2689 (K).
Zimbabwe: Harare District, Harare, university campus on southern slope, 14.iii.1985,
Bennett s.n. (K).
Distribution: India, introduced to Australia, Africa and America.
273
5. Dichanthium caricosum (L.) A. Camus, Bull. Mus. Hist. Nat. Paris 27: 519
(192 1).-Bor, Flora of Assam 5: 370 (1940).-Bor, Grasses Burma Ceylon India &
Pakistan: 134 (1960).-Napper, Grasses of Tanganyika: 100 (1965).-Clayton, in Polhill
(ed.), Fl. Trop. East Afr., Gramineae (3): 725 (1982).
Synonyms: Andropogon caricosus L., Sp. P1. ed. 2: 1480 (1763); Andropogon
serratum Retz., Obs. Bot. 5: 21(1789); Andropogonfihiformis Pers., Syn. P1. Glum.
1: 103 (1854); Heteropogon concinuus Thw., Enum. P1. Zeyl.: 368 (1864);
Andropogon caricosus subsp. genuinis Hack., in DC., Monogr. Phan. 6: 568 (1889);
Apocepis pallida Hook.f.. Fl. Brit. Ind. 7: 143 (1896); Dichanthium caricosum (L.)
Stapf, in Ridley, Fl. Malay Penins. 5: 210 (1925); Dichanthium pallidum (Hook.f.)
Stapf ex Fischer, in Gamble, Fl. Pres. Madras 10: 1740, 1741 (1934); Dichanthium
theinlwinii Bor, in Kew Bull. 1949: 223 (1949).
Perennial. Culms usually robust, ascending from a creeping base, often forming tufts
at the rooted nodes. Leaf blades linear-lanceolate, up to 20cm long and 5mm wide,
glabrous or sparingly hairy; leaf sheaths hairy along the margins and at the node,
more rarely completely glabrous; ligule a ciliate membrane.
Inflorescence subdigitate; raceme peduncles and culm below inflorescence glabrous;
racemes solitary, paired or more often 3-10, arranged along a short primary axis, with
15-20 spikelet pairs. Sessile spikelet 3-4mm long, obovate, lowest 1-3 homogamous,
male or neuter, the upper ones bisexual; glumes subequal; lower glume ciliate below
the middle, glabrous above or with scattered longer hairs near the apex and along the
margins, rarely with scattered long hairs all over; upper glume often slightly longer
than the lower one, ciliate along the keel. Florets 2: lower floret empty, lemma
hyaline; upper floret bisexual, lemma stipiform, entire; awn 25mm long, glabrous; no
anthers seen; ovary elliptic, tiny. Pedicellate spikelet male or neuter, subequal in
shape and size to the sessile one.
Chromosome number(s): 2n = 20 (De Wet & Harlan, 1966); 40 (De Wet,
Borgaonkar & Richardson, 1963).
274
Diagnostic characters and affinities: Morphologically D. caricosum and D.
aristatum are easily distinguishable on the basis of the peduncle pubescence, which is
glabrous in D. caricosum. On the basis of other characters, however, many specimens
may equally well be included with either of the species. The plant is much more
robust than D. aristatum. The spikelets seem to be bigger and markedly obovate.
Habitat: Swampy places, open humid woodland, black cotton soils.
Uses: Native pasture species.
Type: India, Burmann (LINN).
Representative specimens:
Singapore: Nelson road, Tanjong Pagar, railway goods yard, 08.v.1965, Sinclair
10784 (BM; E; K;).
Thailand: Payap, Fang, wayside, altitude 350m, 28.ii.1958, Sorensen, Larsen &
Hansen 1785 (E).
Distribution: India, Burma, Sri Lanka, Malay Peninsula, China, New Guinea and
Fiji.
6. Dichanthium sericeum* (R. Br.) A. Camus, in Bull. Mus. Hist. Nat. Paris 27: 549
(192 1).-Burbidge, Australian grasses 1: 146 (1966).-Burbidge, Australian grasses:
146 (1968).-Lazarides, The grasses of Central Australia: 70 (1970).-Wheeler, Jacobs
& Norton, Grasses of New South Wales: 163 (1982).-Simon, A key to Australian
grasses (ed.2): 98 (1993).-Simon & Latz, A key to the grasses of the Northern
Territory, Australia: 26 (1994).
Synonyms: Andropogon sericeus R. Br., Prodr. Fl. Nov. Holl. 1: 201 (1810);
Andropogon affine R. Br., Prodr. Fl. Nov. Holl. 1: 201 (1810); Andropogon
*
sericeus, silky, with long straight close-pressed hairs.
275
chrysantherus F. Muell., Linnnaea 25: 443 (1852); Andropogon jubatus Balansa,
Bull. Soc. Bot. Fr. 19: 322 (1872); Andropogon acutiusculus Hack., in DC., Monogr.
Phan. 6: 575 (1889); Andropogon sericeus var. mollis F. M. Bailey, Queens!. Agric.
Journ. 30: 316 (1913); Dichanthium acutiusculum (Hack.) A. Camus, Bu!!. Mus. Hist.
Nat. 27: 549 (1921).
A perennial of about im tall, shortly rhizomatous, erect or slightly decumbent; cuim
nodes exposed and bearded. Leaf blades linear-lanceolate, up to 15cm long, glabrous
to pilose all over; leaf sheaths glabrous to pilose; ligule an unfringed membrane of up
to 1.5mm long.
Inflorescence of 1-15 sessile to subsessile white villous digitate racemes; lowest 1-6
spikelet pairs homogamous, male or neuter. Sessile spikelet elliptic-oblong, truncate
at the apex; lower glume obtuse, pilose or rarely glabrous on the back below the
middle, always with tubercle-based cilia (up to 6mm long) along the margins and
forming a transverse subapical fringe; upper glume slightly longer than the lower,
glabrous, ovate, carinate. Florets 2: lower floret empty, lemma hyaline; upper floret
bisexual, lemma 2mm long, stipiform; awn 30mm long; epaleate; anthers 2 or 3, 2mm
long; ovary obovate. Pedicellate spikelet obovate-oblong, 4mm long, male or neuter;
lower glume obovate, linear to narrowly ovate, with a transverse subapical fringe of
cilia; anthers 2.
Chromosome number: 2n = 20 (De Wet, Borgaonkar & Richardson, 1963; Gould,
1975).
Diagnostic characters and affinities: Racemes are densely covered with long, white
silky hairs, sessile or very shortly pedunculate; lower glume of the sessile spikelet
covered with long, silvery hairs with an arch of extremely long hairs at apex, many of
which arise from tubercles.
Habitat: Occurs mainly in open grassland on heavy blackish soil and somewhat and
places.
276
Uses: Native pasture species.
Type: Australia, Port Jackson, Brown 6178 (BM!)..
Representative specimens:
Australia: Leichhardt District, vi. 1956, Bissett 1 (B); Keppel Bay, 1802-1805, Brown
6178 (sheet 1) & Paterson river (sheet 2) (isolectotypes of Andropogon sericeus R.
Br.( labelled A on the sheet) and Andropogon affinis (R. Br.) A. Camus (labelled B
on the sheet); E); Upper Murray District, v.1913, Peck s.n. (E).
Kenya: Kabete Experimental Farm, 27.ix.1915, Dowson 203 (K).
Distribution: Widely distributed in Australia, and introduced to tropical East Africa.
277
Chapter 9. Distribution and ecogeography of the
Andropogoneae and the Sorghinae.
9.1. The Andropogoneae.
The Paniceae and Andropogoneae both appear to be 'natural' tribes, which probably
originated from a common panicoid stock in the warmer parts of the eastern
hemisphere, possibly in the East Africa-Madagascar region (Hartley, 1958). It is
thought that the Panicoideae then spread throughout the tropical and subtropical parts
of both hemispheres. In this spread the tribe Paniceae formed the advance guard,
reaching the western hemisphere at an early stage, and developing rapidly there under
favourable conditions. The Andropogoneae, possibly formed at .a later stage from the
panicoid stock, spread rather towards the Indo-Malaysian region, and reached the
American continent only comparatively recently (Hartley, 1958).
There is evidence that climatic factors, including winter temperature and
possibly rainfall, are of importance in relation to the distribution of the
Andropogoneae (Hartley, 1950). The effect of temperature is reflected in the
predominantly tropical and subtropical distribution of the tribe, while its abundance
in different parts of the tropical zone appears to be influenced by rainfall. The
Andropogoneae reach their highest relative development in monsoonal climates, with
a short season of heavy rainfall (Hartley, 1958). He lists the most conspicuous factors
in the distribution of the Andropogoneae as:
the concentration of the tribe in the tropical and subtropical parts of the world;
the region of maximum abundance in Indo-Malaysia;
and
the relatively lower concentration in the western hemisphere than in the eastern
hemisphere.
278
The general distribution of the Andropogoneae throughout the whole world is
shown in Figure 9.1 (adapted from Hartley, (1958)). The percentages indicate the
proportion of the species of the Andropogoneae in the total grass flora of the region.
The main centre of specific differentiation is in southern Asia. This region is also rich
in taxa which have been regarded as being the most primitive members of the tribe.
These include the subtribe Saccharinae, and especially the genus Miscanthus, which
not only has a generalised type of inflorescence from which the more specialised
forms within the tribe may have been derived, but which also shows relationships to
other tribes of grasses (Hartley, 1958).
While the percentage of the species of the Andropogoneae in the flora of
Africa does not in anyway approach the high levels reached in some parts of southern
Asia, this may be due to the absence in Africa of comparable levels of summer
rainfall and is not necessarily to be accepted as evidence that the tribe has spread to
Africa from a centre of origin in Asia. The presence in tropical Africa of the genus
Miscanthidium, closely related to, or perhaps congeneric with Miscanthus, indicates
that the primitive forms of the tribe reached Africa at a very early stage, if they did
not, indeed, originate there (Hartley, 1958). This has brought in the probability of a
somewhat parallel evolution and spread occurring in Africa and Asia, with the tribe
throughout retaining its tendency to develop maximum differentiation in regions of
high winter temperature and abundant summer moisture.
Distribution of the Andropogoneae in the western hemisphere is quite
different compared with the eastern hemisphere. Species of the tribe form a very
small percentage of the grass flora. Also the more primitive taxa of the tribe are
absent from the American continent. It is noteworthy that only one of the 87 genera
(Agenium) is recorded as endemic there (Hartley, 1958). This strongly suggests that
the tribe has not reached its full potential development in this region and that it has
spread to America from a centre or centres of origin in the old world.
279
20
,00• u01
00
60•
11
40
20
0
ir
10
10
20
100.
'40
160
60
20%
II
-
0.
'-•
--.--- -__
30%
--
110
100
20
.L
--
0%
10%—
or
.%:
20%
10
000 10)
1000
\R0
lTh
0
1
too,
go'
0
20
0
0
] - regions of maximum
Figure 9.1. World distribution of the Andropogoneae. [
abundance (>40%). The percentages indicate the proportion of the species of the
Andropogoneae in the total grass flora of the region (after Hartley, 1958).
0
Mo
20.10
No
9.2. The Sorghinae.
From a geographical point of view, the genus is the most important and illuminating
of all categories. The species is, generally speaking, too small a unit to be of much
use in the consideration of world-wide plant distribution problems. The family, on
the other hand, is more often than not too large a category for convenient handling.
The genus on the contrary, tends to possess the advantages of both the family and
species without their disadvantages. Genera are mostly of convenient size, both
taxonomically and geographically.
The genera of the subtribe are distributed mainly in the Paleotropical floristic
kingdom, with a few spreading into the Holoarctic, Neotropical and Australian
kingdoms (Table 9.1). The monotypic genera and those with few species have
restricted distributions and to some extent endemic. At specific level, the distribution
discussed is those of species sampled. All the distribution data (maps and tables) are
based on the information obtained from herbarium sheets, unless otherwise stated.
The floristic nomenclature is based on Takhtajan (1986) (Figure 9.2).
9.2.1. Cosmopolitan genera.
These are genera which naturally extend somewhat beyond the limits of the region in
which the bulk of their range and the majority of their species occur. In the
Sorghinae, these genera are represented by Bothriochloa Kuntze, Chrysopogon Trin.
and Sorghum Moench.
Bothriochloa: Other than Sorghum, this is the widely distributed of Sorghinae
genera. It is found throughout the tropics of the world (Figure 9.3). To the south, its
range reaches the sub-Antarctic, while to the north it goes as far as the North
American Atlantic region. The genus has its origins in the Old World but was
introduced into the Western Hemisphere (Celarier & Harlan, 1955). These are
grasses of open grassy places.
281
Table 9.1. Distribution of Sorghinae genera.
Abbreviations:
Ast. - Asthenochloa
Chr. - Chrysopogon
Bot. - Bothriochloa
Cle. - Cleistachne
Cap. - Capillipedium
Die - Dichanthium
Euc. - Euclasta
Hem. - Hemisorghum
Psd. - Pseudodichanthium
Pss. - Pseudosorghum
Sor. - Sorgha strum
Soru. - Sorghum
Vet. - Vetiveria
+- present
282
Floristic Kingdom
Subkingdom
Subkingdom
Boreal
Euro-Siberian
Ast.
Eastern Asian
I
Atlantic North
Canadian-
American
Appalachians
n
e
r
a
Cap.
Chr.
Cle.
Die.
Euc.
Hem.
PSd.
Pss.
+
+
Tethyan
+
+
+
+
Macaronesian
C
Meditarranean
t
Irano-Turanian
+
+
+
+
+
+
+
+
+
Madrean
C
+
African
Saharo-Sidian
a
Sudano-Angolan
+
+
+
Sahelo-Sudanian
+
I
Somalo-Ethiopian
+
+
+
e
South Tropical
+
+
+
African
o
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
West African
r
Rainforest
+
Namib-Karoo
+
+
+
Madagascan
+
Indomalesian
Indian
+
+
+
+
+
+
+
+
+
j
Indo-Chinese
C
Malesian
a
Papuan
Polynesian
+
+
+
+
+
+
+
+
+
+
+
+
+
Hawaiian
Neotropical
+
+
+
+
+
+
+
+
+
+
+
+
Fijian
+
Caribbean
+
Central Brazilian
+
+
Andean
Cape
+
+
+
+
+
+
+
.
North and East
Tropical North
Australian
and East
+
+
+
+
+
+
+
Australian
Central Australian Temprate and
South-Eastern
+
Australian
Antarctic
+
Kalaharian
It
Australian
+
Central
Grasslands
I
Vet.
Southern Atlantic
a
P
Sorti. Spa.
+
+
North American
P
Sor.
+
o
j
Bot.
European
o
r
e
Subregions
regions
H
G
I
Patagonian
+
+
+
+
Figure 9.2. Floristic regions of the World (after Takhtajan (1986)).
1. Circumboreal Region. 2. Eastern Asiatic Region. 3. North American Atlantic
Region. 4. Rocky Mountain Region. 5. Macaronesian Region. 6. Mediterranean
Region. 7. Saharo-Arabian Region. 8. Irano-Turanian Region. 9. Madrean Region.
10. Guineo-Congolan Region. 11. Uzambara-Zululand Region. 12. SudanoZambesian Region. 13. Karoo-Namib Region. 14.St. Helena and Ascension Region,
15. Madagascan Region. 16. Indian Region. 17. Indo-Chinese Region. 18. Malesian
Region. 19. Fijian Region. 20. Polynesian Region. 21. Hawaiian Region. 22.
Neocaledonian Region. 23. Caribbean Region. 24. Region of the Guyana Highlands.
25. Amazonian Region. 26. Brazilian Region. 27. Andean Region. 28. Cape Region.
29. North East Australian Region. 30. Southwest Australian Region. 31. Central
Australian or Eremaean Region. 32. Fernándezian Region. 33. Chile-Patagonian
Region. 34. Region of the South Suantarctic Islands. 35. Neozylandic Region.
284
ISO 40
1218
100'
or
60
30
20' 0' 20 40' 60° 10' 100
120'
40
00 180'
30' 180'
U)
co
('.4
25
26
32
33
28
3
ISO'
140,
00'
Figure 9.2. Floristic regions of the World (after Takhtajan (1986)).
40
so
-'<
- •
f'
o
too
too
L
a,•
/
10
10
/
/
/
0
20
0
I 1
:0-
r
10
10
\
100 .
\
120
110
I0
—\
tao
ISO
l00
.
/
—
---
--
----
20
1
100 oon 000
IIKIO
Figure 9.3. World distribution of the genus Bolhriochloa (---)
10".
Several species of Bothriochloa are of value in rangelands. B. ischaemum has
been utilised as a sown forage grass for many years on the southern Great Plains of
the USA (Hacker & Loch, 1995). The distribution of this species is widespread
throughout the subtropical and temperate regions of Eurasia. It is found occasionally
in the Atlas mountains of North Africa, throughout southern Europe, western Siberia,
and central Asia, Mongolia and North China. Its northern limit in Europe is Belgium
and Germany. Its southern boundaries extend from Syria and Lebanon, and then east
through Iraq, Iran, Afghanistan, the Himalayas of North India, and into South China.
In the tropics, only B. inscuipta (creeping bluegrass) and B. pertusa (Indian
bluegrass) have been developed as species for sown pastures. B. inscuipta is a
stoloniferous perennial species native to tropical Africa, Arabia and Southeast Asia.
It grows on a range of soils, including clays. This species was introduced in Australia
in the 1930's (Hacker & Loch, 1995). B. pertusa is also stoloniferous and closely
related to B. inscuipta though generally smaller in stature. B. pertusa is essentially an
Asian species. It also occurs in eastern Africa, but it is uncertain whether it is native
to that continent or introduced. It is also naturalised in tropical America and
Australia.
B. bladhii has a very extensive distribution in the Old World tropics. It is
widely distributed throughout Africa, is found in India, Pakistan and Sri Lanka. It is
also present in Burma, Malay Peninsula, China and Australia. It is also found in
abundance in most of the Pacific islands. B. radicans is a native of the tropics and
subtropics of eastern and southern Africa. It has also been introduced to tropical
America. B. macera, B. biloba and B. decipiens occur only in Australia. According to
Blake (1944), B. macera and B. biloba are particularly abundant in New South Wales
and some neighbouring parts of Victoria, but extends to Queensland and South
Australia. The species B. decipiens has two varieties. Both these varieties are
endemic to Australia. B. decipiens var. decipiens is very abundant in the Eucalyptus
forests of Queensland, particularly towards the south and east and extending into
northern New South Wales (Blake, 1944). B. decipiens var. cloncurrensis occurs
287
only in Queensland. B. kuntzeana is a species restricted to the Indian subcontinent. B.
saccharoides is found in tropical America and the Caribbean islands. It is not clear
whether this species is native there or has been introduced.
Chiysopogon: This genus is found mainly in the tropics of the Old World with a
large concentration of species in the Indian subcontinent (Figure 9.4). Only one
species, C. aciculatus, also occurs in temperate southern Europe, where it is
commonly used in lawns. This species is also found in the tropics of Asia, Africa,
Polynesia and Australia. Four species, C. sylvaticus, C. fallax, C. latfolius and C.
pallidus are all endemic to Australia. They occur in abundance in Queensland, the
Northern Territory and Western Australia. They are found in varied habitats, from
stony hillsides through savannah woodlands to roadsides and ledges. One species, C.
orientalis, occurs in the Indian subcontinent. The remaining species, C. aucheri, C.
gryllus, C. plumulosus, C. fulvus and C. serrulatus are distributed in the Arabian
Peninsula, tropical Africa and the Indian subcontinent. C. serrulatus is the only one
in the group whose range extends to southern Africa.
Sorghum: The genus Sorghum is largely indigenous to the Old World tropics and
subtropics (Figure 9.5). Of the five subgenera of the genus, only (Eu)Sorghum and
Parasorghum have representatives in the Western Hemisphere. The other subgenera
are restricted to the Old World tropics.
Parasorghum spreads from southern and eastern Africa through India and
south-east Asia to Australia, and also to western Guatemala and Mexico. In this
subgenus was sampled the species S. nitidum and S. versicolor. S. nitidum is found in
Australia (eastern Queensland), China, Japan, India and most of the Indonesian
islands. S. versicolor is characteristic of deciduous bushland or wooded grasslands of
eastern, central and southern Africa.
Heterosorghum is restricted to Papua New Guinea, the Philippines, and
Australia. Both Chaetosorghum and Stiposorghum are confined to Australia.
Stiposorghum is represented by species S. plumosum and S. stipoideum in the
sample. S. plumosum is distributed in Western Australia, the Northern Territory and
nfl-
n.
10
50
t
..
fl
It
SQ
5
r Or in,
50 '411 III
IOU
'0
rFOMMENW
IV
IM
F
F.
I
I— 10001d)
1000
IW S
I40
70
U0
-
Figure 9.4. World distribution of the genus Chrysopogon (-
)
0
~Iwo
5,
-
7
_-
-
--
a
/RI(
25
C
C)
co
5-4 1
60
10
00
10
60
0
10
0
11117 0
60
60
/
00
Figure 9.5. World distribution of the genus Sorghum (_
00:
10
_)•
60
fe.
00
60
100
100 000 000
Queensland chiefly north of latitude 21°, but extending further south in Queensland.
(Lazarides et al, 1991). A new record of the presence of this species in the Indonesian
island of Sumba was made in this study. S. stipoideum is distributed in Western
Australia and the Northern Territory.
(Eu)Sorghum is distributed through southern Africa to India, south-east Asia
and the Philippines, "but did not reach either Australia or America until carried there
by man in the past few centuries"(Doggett, 1970). All taxa of significance as crops or
sown pastures are in this subgenus. A detailed comparative morphological study of
grain sorghums, S. bicolor subsp. bicolor, led Harlan and De Wet (1972) to recognise
five basic races, and ten hybrid races that combine characteristics of any two or more
basic races. Race bicolor sorghums are grown across the range of Sorghum
cultivation in Africa and Asia. Race kqfIr sorghums are important staples across the
eastern and southern savannah from Tanzania to South Africa. Race caudatum is
widely grown in Chad, Sudan, north-eastern Nigeria and Uganda. Race durra is
widely grown along the fringes of the southern Sahara, across and West Africa, the
Near East and parts of India. Finally race guinea is grown in the high rainfall areas of
West Africa and the savannah of Malawi. S. bicolor subsp. drummondii occurs as a
weed in Africa wherever grain sorghums and their closest relatives are sympatric (De
Wet, 1978). S. bicolor subsp. arundinaceum extends across the African savannah. De
Wet (1978) reports that it was introduced to tropical Australia, parts of India and the
New World. Four races are recognised in this subspecies: race aethiopicum of the
and African Sahel; race virgatum of north-eastern Africa; race arundinaceum of the
African tropical forest; and race verticillflorum of the African savannah.
S. halepense (Johnson grass) is a perennial with stout, creeping rhizomes. It is
a native of the Mediterranean regions and common in all warm countries, including
the southern United States. It is often a troublesome weed but reckoned an excellent
forage grass. It is said to cross easily with S. bicolor in Argentina, and the hybrid
thus formed is called S. almum Parodi (Hacker & Loch, 1995).
291
9.2.2. Discontinuous genera.
Discontinuity is the occurrence of a species or other unit in two or more separated
regions. When one plant group is found distributed over two or more widely
separated regions, its discontinuity is significant, only if it can be assumed that its
range was formerly continuous and the subsequent disjunction has resulted from
natural causes. This therefore means that the group is of monophyletic origin and has
therefore not only had a common ancestry but also a single point of origin.
Capillipedium Stapf: Capillipedium occurs in the tropics of the Old World with a
high concentration of the species in the Indomalesian region (Figure 9.6). Of the 14
or so species of the genus, it is only C. parvflorum and C. spicigerum that occur
outside the region in Africa and Australia, with C. spicigerum endemic to Australia.
C. Jiliculme and C. hue gelii are endemic to India. C. assimile has a much wider
distribution than the two species mentioned above, for it covers most of Southeast
Asia. C. parvflorum is the widely distributed of Capillipedium species. In addition
to the Indian subcontinent, its range extends westwards to eastern and southern
Africa, and southwards to Australia.
Cleistachne Benth.: The genus consists of a single species, C. sorghoides. It is
restricted in distribution to eastern and southern tropical Africa and western India
(Figure 9.7).
Dichanthium Willemet: Dichanthium is widespread throughout the tropics and
subtropics of the Old World (Figure 9.8). Most of the species occur in Africa, Arabia
and the Indian subcontinent. D. annulatum is native through most of Africa, east
China and Southeast Asia and south to tropical eastern Australia. It has also been
introduced to the southern parts of the USA (Hacker & Loch, 1995). This species has
three varieties. These are var. annulatum, var. fecundum and var. papillosum. Var.
annulatum occurs in all the places where the species is known to exist. Var.
fecundum occurs only in Australia, while var. papillosum is native to the tropics and
subtropics of Africa. D. aristatum is indigenous to regions from tropical India
292
so,
/
L
T
/7
100'
'
'\
1
Jc
//ri
it
(Y)
C)
-
:T
:
—
—
3c 3c
/---
—
10,00
o
oW
Figure 9.6. World distribution of the genera
Pseudodichanthium (-.-.-) and Vetiveria (-).
I0
,o
o
Capillipedium (x x x),
60
0
00Wb0
Q
Ar
0
0h10ISO,
KO -
b0100
L_ -_----
-
\
y
V
NN
/
20 ,
-
-
-.
00
Ion
•
$Q•
.
.
.
o'6
'o
so
0
us
oo
00'
Figure 9.7. World distribution of the genera Cleistachne (-.-.-) and
Sorghastrum (-).
).
5o l.
1
100 000 000
(t-.00o6)
ISO'
ISO'
ISO'
ISO
I00
00
90
Q
00
70
.o
/
-
10
IF
40
OF
0
70
•
/
AT
80
IF
00
70
'0
10
10
110
ISO,
_._-___.-
r:---
:
/
-
- - -
-
--
--
70
—
I
tI
.'
,•_._.._,1
(
.
... ..
• ..
I
0
\
I
Se
.000
00
70
Figure 9.8. World distribution of the genera Dichanthiuni (Euclasta (-.-'-).
40
-) and
0
00
10
ISO
0
1000
eastwards to Indonesia, "... but has become naturalised in the Philippines, Fiji,
Australia, Africa and the Americas " (Hacker & Loch, 1995). D. caricosum is a
tropical species native to southern Asia from India to Malaysia and China. It also
occurs in the West Indies, Cuba and Fiji, where it probably has been introduced. D.
sericeum (Queensland bluegrass) is native to subtropical and tropical eastern
Australia, Papua New Guinea and the Philippines. In Queensland, the species often
dominates in open grasslands. The species D. micranthum and D. foveolatum occur
mainly in the Arabian Peninsula, with the latter species extending southwards to
tropical east Africa. These are species of low lying coastal areas.
Euclasta Franch.: This is one genus that shows a striking discontinuity. It is found
only in east tropical Africa and western India (Figure 9.8). Of the three species in the
genus, two, E. clarkei and E. oligantha, are found only in western India, and only E.
condylotricha, extends to Madagascar and tropical east Africa.
Pseudodichanthium Bor: This genus was once thought to be endemic to India
(Mehrota & Jam, 1980), but recent collections have shown that it also occurs in the
Saharo-Arabian region, in the country of Oman (Figure 9.6). The single species, P.
serrafalcoides, is found in monsoon forests. Oman tends to have affinities with India
in terms of climate in contrast to most of the Arabian Peninsula whose affinities are
closer to continental Africa (Cope, personal communication).
Sorghastrum Nash: Species of the genus are found in central and western Africa and
north, central, and south America, including the Caribbean (Figure 9.7). In the
Western Hemisphere, Sorghastrum appears to be restricted to the regions of the two
continents facing the Atlantic ocean. The range of distribution of this genus would be
continuous, in a sense, if the ocean barrier did not exist. In fact, two species, S.
incompletum and S. stipoides, have been reported from both hemispheres. The
species S. pogonostachyum, S. fuscescens, S. friesii and S. incompletum var.
bipennatum are all African in distribution. S. elliottii and S. nutans occur mainly in
the USA. In fact, S. nutans is said to be one of the important grasses of the prairies
'I.
and an indicator of fodder in good condition (Gould, 1975). S. minarun and S.
setosum are found in South America and extend northward to the Caribbean islands.
Vetiveria Bory: This is a small genus, probably with fewer that ten species, but is
widespread throughout the Old World tropics (Figure 9.6). One species, V
zizanioides, is a native of tropical Asia. It has been extensively cultivated for its
aromatic roots, and is now widely distributed throughout the tropics of world. It is
'Khas-Khas' or 'Khus-Khus', the source of 'Vetiver' oil. V. fulvibarbis and V.
nigritana are African species. V fulvibarbis is mainly West African in distribution,
while V. nigritana, occurs in both west, east and southern Africa. V. Jllipes and V
elongata are restricted to Australia, where they tend to favour habitats near
watercourses.
9.2.3. Endemic genera.
Endemism, in botanical terms, refers to the naturally growing taxa (species, genus, or
family) confined to a particular area (Mehrota & Jam, 1980). In this account of the
distribution of genera, endemic genera are those which are confined to only one of
the floristic regions of the world, or whose ranges are not much greater than the
average size of the continental ranges therein mentioned. Generic endemism is here
therefore interpreted broadly in terms of Takhtaj an's floristic regions (Figure 9.2).
Four genera in the Sorghinae are endemic to certain regions. These are Asthenochloa
BUse, Hemisorghum C. E. Hubbard, Pseudosorghum A. Camus and Spathia Ewart
(Figure 9.9).
Asthenochloa: This genus is found in the south-east Asian islands of Java and the
Philippines of the Malesian region. Even though it has been poorly collected, it
seems to be found only in this part of the world. It occurs in damp and shaded places,
rarely on banks of terraces, and of roads.
Hemisorghum: Hemisorghum is distributed in the Indian region, from Burma to
Indo-China. H venustum is apparently restricted to the island of Sri-Lanka. It is
297
occasionally met with in southern India. The species of this genus tend to prefer
shallow soils, mainly along river banks and disused roadsides.
Pseudosorghum: The genus occurs in tropical south-east Asia. Collections have
mainly been made from the Indian sub-continent, and most of the Indonesian islands.
The species are found in fairly grassy places.
Spathia: This is an Australian genus. So far, collections indicate that it occurs in the
Northern Territory and Queensland. It is common in grasslands.
298
,.
T.
n.
.ft•
Inn.
I,n•
1401
Ian.
100
50
1S0
'-----•--
i---.i4 --
-
.
-
SMEArpom
-7zISO'
1*0'
Ifl
Figure 9.9. World distribution of the genera AsthenochlOa ( A ),
-) and Spathia (- .-)
Hemisorghum (x x x), Pseudosorghum (
1— 1000
1,)
Chapter 10. References.
Adanson, M. (1763). Families de Plantes 2: 38. Paris.
Arber, A. (1934). The Gramineae: A study of cereal, bamboo, and grass. Cambridge
University Press. New York. USA.
Avdulov, N.P. (1931). Karyo-systematische Untersuchung de Familie Graminen.
Bull. Appl. Bot. P1. Bree. Suppl. 43: 428.
Backer, C.A. & B.C.B. van den Brink (1968). Poaceae (Gramineae). In Flora of
Java 3: 495-461. Wolters-Noordhoff, N.V.-Groningen.
Baum, D. (1992). Phylogenetic species concepts. Trends in Ecology and Evolution 7:
1-2.
Bentham, G. & J.D. Hooker (1883). Genera Plantarum. L. Reeve. London. UK.
Bews, J.W. (1929). The World's grasses. Longmans, Green & Co. London. UK.
Blake, S.T. (1944). Monographic studies in the Australian Andropogoneae, Part I.
University of Queensland Papers 2(3): 1-55.
Blake, S.T. (1969). Taxonomic and nomenclatural studies in the Gramineae. I. Proc.
R. Soc. Qd. 80(6): 55-84.
Bor, N.L. (1940). Flora ofAssam 5. Calcutta. India.
Bor, N.L. (1940). Three new genera of Indian grasses. Indian Forester 66: 267-272.
Bor, N.L. (1954). Notes on Asiatic grasses: XVI. Kew Bull. 1954: 73-81.
Bor, N.L. (1960). Grasses of Burma, Ceylon, India and Pakistan. Pergamon Press.
Oxford. 767p.
Borgaonkar, D.S. & J.M.J. De Wet (1964). Cytogenetical study of hybrids between
Dichanthium annulatum and D. fecundum. III. Oyton 21(l): 55-60.
300
Bowden, W.M. (1960). Chromosome numbers and taxonomic notes on Northern
grasses. Can. J Bot. 38: 341-557.
Bremer, K. (1978). Oreolysera and Anthistrixia, new and old South African genera
of Compositae. Bot. Notiser 131: 449-453.
Bremer, K & H. Wanntorp (1978). Phylogenetic systematics in botany. Taxon
27(4): 317-329.
Briggs, D. & S.M. Walters (1984). Plant variation and evolution. Second edition.
Cambridge University Press. Cambridge. UK. 412 pp.
Brown, W.V. (1950). A cytological study of some Texas Gramineae. Bull. Torrey
Bot. Club 77(2): 63-76.
Brown, W.V. (1975). Variations in anatomy, associations and origins of Kranz
tissue. Amer. I Bot. 62(4): 395-402.
Brown, W.V. (1977). The Kranz syndrome and its subtypes in grass systematics.
Mem. Torr. Bot. Club 23: 1-97.
Carnahan, H.L. & H.D. Hill (1961). Cytology and genetics of forage grasses. Bot.
Rev. 27(1): 1-162.
Celarier, R.P. (1958). Cytotaxonomy of the Andropogoneae III: Subtribe Sorgheae,
genus Sorghum. Cytologia 23: 395-418.
Celarier, R.P. (1959). Cytotaxonomy of the Andropogoneae IV: Subtribe Sorgheae.
Cytologia 24(3): 285-303.
Celarier, R.P., J.M.J. De Wet & D.S. Borgaonkar (1961). Intergeneric hybrids in
the Bothriochloininae I. Bothriochloa intermedia and Dichanthium annulatum.
Cytologia 26: 170-175.
Celarier, R.P. & J.R. Harlan (1955). Studies on the Old World bluestems. Okla.
Agric. Exp. Sta. Tech. Bull. T-58: 1-33.
301
Chippandall, L.K.A. (1955). Part 1: A guide to the identification of grasses in South
Africa. In Meredith, D (editor), The Grasses and Pastures of South Africa. Central
News Agency. Johannesburg. South Africa.
Church, G.L. (1929). Meiotic phenomena in certain Gramineae II. Paniceae and
Andropogoneae. Bot. Gaz. 88: 64-84.
Clayton, W.D. (1961). Proposal to conserve generic name Sorghum Moench
(Gramineae) versus Sorghum Adans. (Gramineae). Taxon 10(8): 242.
Clayton, W.D. (1969). A revision of the genus Hyparrhenia. Kew Bulletin
Additional Series II. HMSO. London.
Clayton, W.D. (1972). The awned genera of the Andropogoneae: Studies in the
Gramineae XXXI. Kew Bull. 27: 457-474.
Clayton, W.D. (1982). Gramineae (3), in Polhill, R.M. (editor), Flora of Tropical
East Africa: 718-739.
Clayton, W.D. (1983). The genus concept in practice. Kew Bull. 38: 149-153.
Clayton, W.D. & S.A. Renvoize (1986). Genera Graminum: Grasses of the World.
Kew Bull. Add. Ser. 13. HMSO. London. 389p.
Clayton, W.D. & S.A. Renvoize (1992). Classification and evolution of the grasses.
In Chapman, G.P (editor), Grass evolution and domestication. Cambridge University
Press: 3-37.
Clifford, H.T. & L. Watson (1977). Identifying grasses: Data, Methods and
Illustrations. 146 pages. St Lucia: University of Queensland Press. Australia
Cope, T.A. (1980). New combinations in Asiatic grasses. Kew Bull. 35(3): 701-704.
Davila, P.D. (1988). Systematic revision of the genus Sorghastrum (Poaceae:
Andropogoneae) PhD. Thesis. Iowa State University, Ames, Iowa. Arb, Mich.,
University Microfilms International.
302
Davila, P.D. & L.G. Clark (1990). Scanning Electron microscope survey of leaf
epidermis of Sorghastrum (Poaceae: Andropogoneae). Amer. Journ. Bot. 77(1): 499511.
Davidse, G. & R.W. Pohl (1972). Chromosome numbers and notes on some Central
American grasses. Can. I Bot. 50: 273-283.
Davidse, G. & R.W. Pohl (1974). Chromosome numbers, meiotic behaviour and
notes on tropical American grasses (Gramineae). Can. I Bot. 52(2): 317-328.
Davis, P.H. & V.H. Heywood (1963). Principles of Angiosperm taxonomy. London.
UK.
Dc Queiroz, K. & M.J. Donoghue (1988). Phylogenetic Systematics and the species
problem. Cladistics 4: 317-338.
Dc Queiroz, K. & M.J. Donoghue (1990). Phylogenetic systematics and species
revisited. Cladistics 6: 83-90.
Dc Wet, J.M.J. (1978). Systematics and evolution of Sorghum sect. Sorghum. Amer.
I Rot. 65(4): 477-484.
Dc Wet, J.M.J. & L.J. Anderson (1956). Chromosome numbers in Transvaal
grasses. Cytologia 21: 1-10.
De Wet, J.M.J., D.S. Borgaonkar & H.R. Chheda (1961). Intergeneric hybrids in
the Bothriochloininae II. Bothriochloa and Capillipedium. Cytologia 26: 268-273.
Dc Wet, J.M.J., D.S. Borgaonkar & W.L. Richardson (1963). Chromosome
number and mode of reproduction in the Bothriochloininae. Caryologia 16(1): 47-55.
Dc Wet, J.M.J & J.R. Harlan (1962). The compilospecies concept. Evolution 17:
494-501.
Dc Wet, J.M.J & J.R. Harlan (1966). Morphology of the compilospecies
Bothriochloa intermedia. Amer. I Rot. 53(1): 94-98.
Dc Wet, J.M.J & J.R. Harlan (1967). Essential oils as taxonomic aids in the
classification of Dichanthium parvflorum. Amer. I Bot. 54(3): 384-387.
303
Dc Wet, J.M.J & J.R. Harlan (1970). Bothriochloa intermedia - A taxonomic
dilemma. Taxon 19: 339-340.
De Wet, J.M.J, J.R. Harlan & E.G. Price (1970). Origin of variability in the
Spontanea complex of Sorghum bicolor L. Amer. I Bot. 57(6): 704-707.
Dc Wet, J.M.J. & M.L. Higgins (1963). Species relationships within the
Bothriochloapertusa complex. Ohyton 20(2): 205-211.
De Wet, J.M.J. & J.P. Huckabay (1967). The origin of Sorghum bicolor. II.
Distribution and domestication. Evolution 21: 787-802.
Doggett, H. (1970). Sorghum. Longmans Green & Co. Ltd. i-xvi, 403p.
Donoghue, M.J. (1985). A critique of the biological species concept and
recommendations for a phylogenetic alternative. Bryolo gist 88: 172-181.
Duval-Jouve, M.J. (1875). Histotaxie des fuilles des Graminees. Ann. Sci. Nat. Bot.
Ser. 6(1): 294-371.
Ellis, R.P. (1976). A procedure for standardising comparative leaf anatomy in the
Poaceae. I. The leaf-blade as viewed in transverse section. Bothalia 12(1): 65-109.
Ellis, R.P. (1979). A procedure for standardising comparative leaf anatomy in the
Poaceae. II. The epidermis as seen in surface view. Bothalia 12(4): 641-671.
Ellis, R.P. (1986). A review of comparative leaf-blade anatomy in the systematics of
the Poaceae: The past twenty-five years. In Soderstrom, T.R., K.W. Hilu, C.S.
Campbell & M.E. Backsworth (editors), Grass systematics and Evolution.
Smithsonian Institute Press: 3-10.
Esau, K. (1965). Plant anatomy. 2nd edition. John Wiley & Sons, Inc. New York
Estes, J.R. & R.J. Tyrl (1986). Concepts of taxa in the Poaceae. In Soderstrom, T.R.,
K.W. Hilu, C.S. Campbell & M.E. Backsworth (editors), Grass systematics and
evolution. Smithsonian Institution Press. Washington, DC.: 325-333.
Ewart, A.J. & O.B. Davies (1917). The Flora of Northern Territory: 26. MacCarron,
Bird & Co., Melbourne. Australia. 387p.
304
Faruqi, S.A., H.B. Quraish & N. Halai (1979). Chromosome numbers and
morphological characteristics of some Andropogoneae of Pakistan. Cytologia 44:
585-605.
Fedorov, A. (1974). Chromosome numbers offiowering plants. Otto Koeltz Science
Publishers. Koenigstein. Germany.
Fournier, E. (1881). Mexicanas Plantes 2: 56-57.
Funk, V.A. (1985a). Cladistics and generic concepts in the Compositae. Taxon 34:
72-80.
Funk, V.A (1995). Cladistic methods. In Wagner, W.L. & V.A. Funk, Hawaiian
Biogeography: Evolution on a hotspot Archipelago. Smithsonian Institution,
Washington DC.
Garber, E.D. (1950). Cytotaxonomic studies in the genus Sorghum. Univ. Calif Pub.
Bot. 23: 283-362.
Garber, E.D. (1954). Cytotaxonomic studies in the genus Sorghum III. The
polyploid species of the subgenera Parasorghum and Stiposorghum. Bot. Gaz. 115:
336.
Gould, F.W. (1959). The glume pit of Andropogon barbinodis Lag. Brittonia 11:
182-187.
Gould, F.W. (1968). Grass Systematics. McGraw-Hill Inc., New York.
Gould, F.W. (1975). The grasses of Texas. Texas A & M University Press, USA.
Gould, F.W. & R.B. Shaw (1983). Grass systematics. 2nd edition. A & M
University Press, College Station, Texas, USA.
Gould, F.W. & T.R. Soderstrom (1974). Chromosome numbers of some Ceylon
grasses. Can. J. Bot. 52: 1075-1090.
Grant, W.F. (1960). The categories of classical and experimental taxonomy and the
species concept Rev. Canad. Biol. 19: 241-262.
305
Gupta, S.C., J.R. Harlan, J.M.J. Dc Wet & C.O. Grassi (1976). Cytology of
backcross of four individuals derived from Saccharum-Sorghum hybrid. Caryologia
29: 351-359.
Hacker, J.B. & D.S. Loch (1995). Tropical and subtropical grasses: various genera
(Andropogoneae). In Smartt, J & N.W. Simmonds (editors.), Evolution of crop
plants, 2nd edition. Longman Group. UK.
Harlan, J.R. (1963). Two kinds of gene centres in Bothriochloininae. Amer. Nat. 97:
Harlan, J.R. & J.M.J. De Wet (1972). A simplified classification of cultivated
sorghum. Crop Sc. 12: 172-176.
Hartley, W. (1950). The global distribution of the tribes of the family Gramineae in
relation to historical and environmental factors. Austr. Journ. Agri. Res. 1: 355-373.
Hartley, W. (1958). Studies on the Origin, Evolution, and Distribution of the
Gramineae. I. The Tribe Andropogoneae; II. The Tribe Paniceae. Australian Journal
ofBotany6: 116-128 and 343-358.
Hattersley, P.W. & L. Watson (1992). Diversification of photosynthesis. In
Chapman, G.P. (editor), Grass evolution and domestication. Cambridge University
Press. Cambridge. UK.
Hennig, W. (1966). Phylogenetic Systematics. University of Illinois Press. Urbana.
263 pp.
Heslop-Harrison, J. (1959). Photoperiod and fertility in Rottboellia exaltata L.
Annals of Botany 23: 345-349.
Heslop-Harrison, J. (1961). The function of the glume pit and the control of
cleistogamy in Bothriochloa decipiens (Hack.) C. E. Hubbard. Phytomorphology 11:
378-383.
OR
Heslop-Harrison, Y. & J.S. Heslop-Harrison (1996). Lodicule function and
extension in the grasses: Potassium ion movement and tissue specialisation. Annals of
Botany 77: 573-582.
Heywood, V.H. (1958). The presentation of taxonomic information: a short guide for
contributors to Flora Europea. Leicester University Press. UK.
Heywood, V.H. (1993). Flowering plants of the world. Andromeda Oxford Ltd.
Abingdon. UK 136p.
Hitchcock, A.S. (1935). Manual of the grasses of the United States. First Edition.
United States Government Printing Office. Washington. USA.
Hitchcock, A.S. (1936). Manual of the grasses of the West Indies. United States
Government Printing Office. Washington. USA.
Hitchcock, A.S. (1951). Manual of the grasses of the United States. Second Edition.
Revised by A. Chase. United States Government Printing Office. Washington. USA.
Holmgren, P.K., W. Keuken & E.K. Schofield (1981). Index Herbariorum I.
Edition 7. Bohn, Scheltema & Holkema Publications. Utrecht. Netherlands.
Hubbard, C.E. (1934). Gramineae. In Hutchinson, J., The families of flowering
plants Ii Monocotyledons. Macmillan & Co. Ltd. London. UK.
Hubbard, C.E. (1984). Grasses: A guide to their structure, identification, uses and
distribution in the British Isles. 3rd edition. Harmonds Worth. Penguin Books. UK.
Ivanyukovich, L.K. (1986). Anatomical characteristic features of caryopsis in
different species of Sorghum. Sborn. Nauch. Jr. Prikl. Bot. Genet. Selek. 103: 105112.
Johansen. D.A. (1940). Plant Microtechnique. McGraw-Hill Book Company. New
York. i-xi, 523p.
Keng, Y.L. (1939). The gross morphology of the Andropogoneae. Sinensia 10: 274343.
307
Kellogg, E.A & C.S. Campbell (1987). Phylogenetic analysis of the Gramineae. In
Soderstrom, T.R., K.W. Hilu, C.S. Campbell & M.E. Backsworth (editors), Grass
sysytematics and evolution. Washington DC: Smithsonian Institution Press. USA.
Kubitzki, K. (editor) (1977). Flowering plants: Evolution and classification of
higher categories. Springer-Verlag Wien/New York. 416 pp.
Larsen, K. (1963). Cytological studies in vascular plants of Thailand. Dansk
BotaniskArkiv 20(3): 211-275.
Lazarides, M., J.B. Hacker & M.H. Andrew (1991). Taxonomy, Cytology and
Ecology of Indigenous Australian sorghums (Sorghum Moench: Andropogoneae:
Poaceae). Aust. Syst. Bot. 4: 591-635.
Legendre, P. (1971). Circumscribing the concept of genus. Taxon 20: 137-139.
Legendre, P. & P. Vaillancourt (1969). A mathematical model for the entities
species and genus. Taxon 18: 234-252.
Linder, H.P. & I.K. Ferguson (1985). On the pollen morphology and phylogeny of
the Restionales and Poales. Grana 24: 65-76.
Linnaeus, C. (1753). Species Plantarum. Ed. 1. Holmiae: p1047.
Maddison, W.P. & D.R. Maddison (1992). MacGlade, Version 3.0. Analysis of
phylogeny and character evolution. Sinauer, Sunderland, Massachusetts. USA.
Mainra, A.K. & H.S. Naimawatee (1983). Species relationships in Sorghum by
protein and esterase electrophoresis. Perspect. Cytol. Genet. 3: 113-115.
Mayr, E. (1942). Systematics and the origin of species. Columbia University Press.
New York.
Mehra, K.L. (1955). Chromosome numbers in the tribe Andropogoneae. md i pi.
Genet. Breed. 15: 144.
Mehra, K.L. (1955). Chromosome races in Chrysopogon montanus. Current Sc. 24:
95-96.
-
Mehra, K.L., K.N. Subramanyam & A.S. Swaminanthan (1962). Cytogenetical
studies in some members of Andropogoneae. I Ind. Rot. Soc. 41: 491-503.
Mehra, P.N. & S.L. Gill (1968). In IOPB chromosome reports XVI. Taxon 17: 199204.
Mehrota, A. & S.K. Jam
(1980). Endemism in Indian grasses- Tribe
Andropogoneae. Bull. Bot. Surv. India 22 (1-4): 51-58.
Metcalfe, C.R. (1960). Anatomy of the monocotyledons: I. Gramineae. Clarendon
Press. Oxford. xi, 731p.
Minelli, A. (1993). Biological systematics: The state of the art. Chapman & Hall.
London. UK.
Moench, C. (1794). Methodus Plantes horti botanici et agri Marbugensis. Marbugi
Cattorum, Germany. p207.
Olorde, 0. (1975). Additional chromosome counts in Nigerian grasses. Brittonia 27:
63-68.
Pankhurst, R.J. (1991). Practical taxonomic computing. Cambridge University
Press. Cambridge. UK. 202pp.
Pereira, F.A.I. (1982). Estudio taxonomico e reproductivo des especies do genero
Sorghastrum Nash (Gramineae: Andropogoneae) ocorrentes no Rio Grande do Sul,
Brazil. MSc. Thesis, Universidade Federal do Rio Grande do Sul.
Pereira, F.A.I. (1986). Chromosome number reports. Taxon 35(1): 195-196.
Persoon, C.H. (1805). Synopsis Plantarum, seu Enchiridium Botanicum. Paris.: 73201.
Pilger, R (1940). Gramineae III (Unterfam. Panicoideae). In Die Narturlichen
Pflanzenfamilien, by Engler, A. & K. Pranti. Vol. 14e, p208. (Wilhelm Engleman:
Leipzig).
Pohl, R.W. & G. Davidse (1971). Chromosome numbers of Costa Rican grasses.
Brittonia 23: 293-324.
309
Potztal, E. (1956). Nachtrag zu Gramineae III. In Die Nartulichen Pflanzenfamilien,
by A. Engler & K. Prantl. Vol. 14d, p255. (Duncker und Humbolt: Berlin)
Purvis, M.J., D.C. Collier & D. Walls (1964). Laboratory Techniques in Botany.
Butterworths. London. i-vu, 371p.
Quicke, D.L.J. (1993). Principles and Techniques of Contemporary Taxonomy.
Blackie Academic & Professional. Glasgow. UK.
Rafinesque, C.S. (1830). Plantae Asiatciae rariores. Bull. Bot. Geneve 1: 221.
Rahman, M. (1988). Taxonomic studies in the genus Panicum L. PhD. Thesis.
University of Edinburgh. Edinburgh. UK.
Raman, V.S., P. Chandrasekharan & D. Krishnaswami (1959). Note on some
chromosome numbers in Gramineae. Curr. Sc. 28: 127-128.
Renvoize, S.A. (1981). The subfamily Arundinoideae and its position in relation to a
general classification of the Gramineae. Kew Bulletin 36: 85-102.
Ridley, H.N. (1930). Dispersal ofplants throughout the world. L. Reeve & Co. Ltd.
Ashford. Kent. UK. 744pp.
Sangster, A.G. (1983). Anatomical features and silica depositional patterns in the
rhizomes of the grasses Sorghastrum nutans and Phragmites australis. Can. J Bot.
61(3): 752-761.
Saura, F. (1944). Cariologia de Gramineas en Argentina. Revista Fac. Agron.
Veterin. II: 51-63.
Simon, B.K. (1989). Studies in Australian Grasses: 4. Taxonomic and Nomenclatural
studies in Australian Andropogoneae. Austrobaileya 3(1): 79-99.
Simon, B.K. (1990). A key to Australian grasses. Queensland Government. Brisbane.
Australia.
Simpson, G.G. (195 1).The species concept. Evolution 5: 285-298.
310
Simpson, G.G. (1961). Principles of animal taxonomy. Columbia University Press.
New York.
Sivarajan, V.V. (1991). Introduction to the principles of Plant taxonomy. 2nd
edition. Cambridge University Press. Cambridge.
Sneath, P.H.A. & R.R. Sokal (1973). Numerical taxonomy. W. H. Freeman. San
Francisco.
Snowden, J.D. (1935). A classification of the cultivated Sorghum. Kew Bull. XXI:
221-254.
Snowden, J.D. (1936). The cultivated races of Sorghum. Allard & Son. London.
Snowden, J.D. (1955). The wild fodder Sorghums of the section Eusorghum. Journ.
Linn. Soc. London 55: 191-260.
Sokal, R.R. & P.H.A. Sneath (1963). Principles on Numerical Taxonomy. W. H.
Freeman. San Francisco.
Sporne, K.R. (1948). Correlation and classification in dicotyledons. Proc. Linn. Soc.
Lond. 160: 40-47.
Sporne, K.R. (1959). A new approach to the approach of the primitive flower. New
Phytol. 48: 259-276.
Stace, C.A. (1989). Plant taxonomy and biosystematics. Second edition. Edward
Arnold. London. 264 pp.
Stapf, 0. (1917). Gramineae, In Pram, D., Flora of Tropical Africa IJ( London.
Stebbins, G.L. (1950). Variation and evolution in plants. Columbia University Press.
New York. 643 pp.
Stebbins, G.L. (1951). Natural selection and the differentiation of angiosperm
families. Evolution 5: 299-324.
Stebbins, G.L. (1956). Cytogenetics and Evolution of the grass family. Amer. J Bot.
43(10): 890-905.
311
Stebbins, G.L. (1956). Taxonomy and the evolution of genera, with special reference
to the family Gramineae. Evolution 10: 235-245.
Steudel, E.G. (1855). Synopsis Plantarum Glumacearum I. Metzler, Stuttgart.
Stevens, P.F. (1985). The genus concept in practice. But for what practice? Kew Bull.
40: 457-465.
Swofford, D.L. (1993). PA UP. Phylogenetic Analysis Using Parsimony, Version 3.1.
Computer program distributed by the Illinois National History Survey, Champaign,
Illinois, USA.
Takhtajan, A. (1986). Floristic regions of the world. University of California Press.
Berkeley. USA.
Tateoka, T. (1965). Chromosome numbers of some East African grasses. Amer. J.
Bot. 52(8): 864-862.
Tsvelev, N.N. (1976). Grasses of the Soviet Union. Part II. Nauka Publishers.
Leningrad. Russia.
Valentine, D.H.
& A.
Love (1958). Taxonomic and biosystematic categories.
Brittonia 10: 153-166.
Vavilov. N.I. (1935). The origin, variation, immunity and breeding of cultivated
plants, trans. K.S. Chester, Chronica Botanica, 13. Ronald Press Co., New York.
Vinall, H.N., J.C. Stephens & J.H. Martin (1936). Identification, history and
distribution of common Sorghum varieties. USDA Tech. Bull. 506. Washington D.C.
Watson, L., H.T. Clifford & M.J. Dallwitz (1985). The classification of the
Poaceae: Subfamilies and Supertribes. Australian Journal of Botany 33: 433-484.
Watson, L. & M.J. Dallwitz (1980). Australian grass genera: Anatomy, Morphology
and Keys. 209 pages. Canberra: Research School of Biological Sciences, Australian
National University.
Watson, L. & M.J. Dallwitz (1992). The grass genera of the World. Wallingford:
CAB International. 1035p.
312
Wheeler, D.J.B., S.W.L. Jacobs & B.E. Norton (1982). Grasses of New South
Wales. University of New England. Armidale. Australia.
Wheeler, Q.D. (1990b). Ontogeny and character phylogeny. Cladistics 6: 225-268.
Wheeler, Q.D. & K.C. Nixon (1990). Another way of looking at the species
problem: A reply to De Queiroz and Donoghue. Cladistics 6: 77-81.
Wiley, E.O. (1981). Phylogenetics: The theory and practice of phylogenetic
systematics. John Wiley & Sons. New York.
313
Chapter 11. Appendices.
Appendix 11.1. A manuscript sent to the Flora Zambesiaca
Editorial Committee at Kew for publication.
Sorghinae of Flora Zambesiaca.
M.P. Setshogo*
Sorghinae Bluff, Nees & Schauer, Comp. Fl. Germ. ed. 2, 1: 46(1836).
Bothriochloinae Keng, in Sinensia 10: 282 (1939).
Inflorescence terminal or rarely axillary, of single or digitate or paniculate racemes, the latter often in
whorls; racemes with fragile rachis and slender internodes, sometimes reduced to triads or single
spikelets, occasionally with homogamous pairs. Spikelets paired, dissimilar, seldom plumose. Sessile
spikelet bisexual, usually dorsally compressed, the callus usually obtuse with cupuliform or truncate
articulation but the latter sometimes oblique (always so when callus pungent); lower glume usually
firm, more or less convex on the back and abruptly rounded on the flanks (except Hem isorghum);
lower floret reduced to a barren lemma; upper lemma linear to oblong, entire or bibbed, sometimes
stipiform, in which case it is reduced to a hyaline base of the awn, usually with a glabrous awn.
Pedicellate spikelet male, barren, or (rarely) bisexual, sometimes much reduced, rarely with a small
callus.
Genera 14. Only 9 of these genera occur in FZ.
A key to FZ Sorghinae genera:
Spikelets in racemes which are solitary at ends of branches; the racemes collected into whorled
panicles; rachis internodes and pedicels not furrowed.
Spikelets solitary at the ends of branches.
Pedicels entirely absent.
1. Cleistachne
Pedicels present; pedicellate spikelets absent.
2. Sorghastrum
Spikelets in pairs or threes; pedicellate spikelets often more or less reduced.
*
Instutute of Cell and Molecular Biology, Department of Botany, University of Edinburgh,
Daniel Rutherford Building, Kings Buildings, Mayfield Road, Edinburgh. EH9 3JH.
314
Spikelets dorsally compressed, in panicles of many pairs or in threes of which one is
hermaphrodite; racemes very much less than 30 cm. long, readily disarticulating;
rachis nodes and pedicels ciliate; lower glume of sessile spikelet rounded on the
3. Sorghum
sides, 2-keeled at tip; lodicules hairy.
Spikelets laterally compressed.
Spikelets in groups each consisting of a sessile and two more or less
reduced pedicellate spikelets or a raceme of several pairs; lower glume
4. Chrysopogon
smooth, not tuberculate.
Spikelets with many pairs of spikelets arranged in whorled racemes; lower
5. Vetiveria
glume tuberculate.
Spikelets in racemes which are digitate or in pairs or solitary and terminal; the racemes not collected
into whorled panicles; sometimes the rachis internodes and pedicels with a translucent median furrow.
Spikelets arranged in false racemes terminating the branches and branchlets of a true panicle;
6. Capillipedium
rachis intemodes and pedicels with a median translucent furrow.
Spikelets arranged in false racemes which are either solitary and terminal or subsessile,
arranged along a central axis but not at the tips of capillary branches.
Racemes several to many, if solitary then the sessile spikelets imbricate; rachis
intemodes and pedicels with a translucent longitudinal furrow.
All pairs of spikelets heterogamous, rarely homogamous; racemes
7. Bothriochloa
ascending; glumes sometimes pitted.
The lowest 1-3 pairs of spikelets homogamous; racemes nodding; glumes
8. Euclasta
never pitted.
Rachis internodes and pedicels without a translucent furrow; lowest 1-6 pairs of
spikelets homogamous; all spikelets, both sessile and pedicellate, in the racemes
9. Dichanthium
more or less alike; glumes herbaceous, greenish.
1. Cleistachne Benth., in Hook. Ic. P1. 14: t 1379 (1882). Type species: Cleistachne sorghoides Benth.
Coarse annual. Inflorescence a panicle, its primary branches bearing racemes at regular intervals along
their length, each raceme reduced to a single sessile spikelet (whose apparent pedicel is homologous to
the raceme peduncle); spikelets solitary, sessile; female fertile florets one; upper lemma bibbed,
stipiform; lodicules 2, ciliate; pedicellate spikelet and pedicel completely suppressed.
Species 1.
1. Cleistachne sorghoides Benth., in Hook. Ic. P1. 14: t 1379 (1882).-Hack., in DC., Monogr. Phan. 6:
653 (1 889).-Stapf, in Prain, Flora of Trop. Africa IX: 154 (1917).-Sturgeon, A revised list of grasses
of Southern Rhodesia Part IV: 8 (1954).-Chipindall, in Meredith, The grasses and pastures of South
Africa: 468 (1955).-Bogdan, A revised list of Kenya grasses: 56 (1958).-Jackson & Wiehe, An
annotated check-list of Nyasaland grasses: 33 (1958).-Bor, Grasses of Burma Ceylon India &
Pakistan: 119 (1960).-Napper, Grasses of Tanganyika: 98 (1965).-Hood, A guide to the grasses of
Zambia: 58 (1967).-Clayton, in Polhill (ed.), Flora of Tropical East Africa, Gramineae (3): 734
(1982).-Gibbs Russell et al, Grasses of Southern Africa: 87 (1991). Type: Mozambique, Shupanga,
Kirk (K).
315
Tall erect stilt rooted annual up to 3m. tall. Culms slender, cuim nodes and intemodes glabrous. Leaf,
blades linear-lanceolate, variable in width but often expanded to greater than 2 cm. wide, hairy on both
surfaces, but less so on the abaxial surface, with prominent midribs; leaf sheaths hairy or glabrous,
when hairy, the hairs just forward of the horizontal; leaf ligule an unfringed membrane.
Inflorescence a very compact panicle; racemes markedly reduced to single sessile spikelets (the
raceme peduncle equivalent to the spikelet pedicel); raceme peduncles and rachis internodes covered
with hair; spikelet colour variable but ranges from straw coloured to chestnut brown. Spikelets oblong;
lower and upper glumes the same size, slightly coriaceous; lower glume minutely truncate, finely 7nerved, covered with uniform short and dense hairs; upper glume glabrous below, hispid towards the
apex, clasped lengthwise by the lower; callus obscure, shortly bearded. Florets 2: lower floret empty,
lemma hyaline, oblong, ciliate, 2-nerved; upper floret bisexual, lemma stipiform, finely 3-nerved,
awned; awn about 5-8 times the length of the lemma, geniculate, with a twisted column hairy along
the spiral and a serrated bristle; palea shorter than the lodicules; lodicules 2, ciliate; anthers not seen;
caryopsis obovate oblong, truncate, brown. Pedicellate spikelet and pedicel absent.
Zambia. (no locality given) dambo, 06.vi.1931, Fraser 14 (K). Zimbabwe. N: Gokwe District,
19.iii.1984, Mahiangu 981 (K); Henderson Research Station, l0.iii.1964, West 4743 (BM); Banket
District, 15.iv.1947, Wild 1891 (K). C: Harare (Salisbury), Teviotdale, 20.iii.1965, Bingham 1429
(BM). E: Chimanimani (Melsetter) District, Muchira river, Tarka Forest Reserve, altitude 1100 m.,
iii.1971, Goldsmith 12/71 (BM). Malawi. C: Dedza District, Mua-Livulezi Forest, 09.iv.1964, Adlard
607 (MAL); Lilongwe District, Agricultural Research Station, 07.iv.1951, Jackson 457 (MAL). 5:
Chiradzulu, altitude 1250 in., 1970, Brummitt & Banda 9831 (K); Zomba District, lower slopes of
Zomba Plateau, above Salisbury bridge, Zomba, altitude 1140 m., 26.iv.1980, Brummitt, Masiye &
Tawakali 15560 (MAL). Mozambique. N: Niassa District, altitude 1000 m., 23. iii. 1964, Torre &
Paiva 11359 (K).
Tropical Africa and India. Grassland or savannah woodland.
2. Sorghastrum Nash, in Britton, Fl. North. States: 71(1901). Type species: Sorghastrum nutans (L.)
Nash.
Poranthera Raf., in Bull. Bot. Geneve 1: 221 (1830).
Divogon Steud., Nom. Bot. ed. 2, 1: 518 (1840).
Annuals or perennials. Inflorescence a panicle, the subdivided primary branches bearing short
racemes, these sometimes reduced to triads. Sessile spikelet callus obtuse or pungent (with involucral
hairs in S. pogonostachyum); lower glume coriaceous, convex, keeled at the tip; upper glume broadly
convex; upper lemma mostly entire, sometimes bibbed, awned; lodicules glabrous. Pedicellate
spikelet usually reduced to a barren pedicel, but present and bisexual in S. fuscescens and S.
pogonostachyum.
Species ± 16.
Pedicellate spikelet bisexual, some pedicels barren.
Cuim nodes glabrous; racemes up to 7-jointed; spikelet surrounded by an involucre of silvery
hairs to 3/4 the length of the sessile spikelet; awn 4 mm. long; upper lemma entire.
1. pogonostachyum
Culm nodes hairy; racemes 3-15-jointed; callus with a short beard of hairs only; awn 25 mm.
2. fuscescens
long; upper lemma bibbed.
316
Pedicellate spikelet absent, all pedicels barren.
Racemes single-jointed; sessile spikelet 4-5.5 mm. long; upper lemma entire; awn 25-40 mm.
3. incompletum var. bipennatum
long.
Racemes 2-many-jointed.
Leaf sheaths not produced into auricles; spikelets 5-7(-8) mm. long; awn 3-8(-l0)
4. friesli
mm. long, straight or bent and twisted.
Leaf sheaths with auricles; spikelets 4-7(-8) mm. long; awns 8-16 mm. long, never
5. stipoides
straight, always bent and twisted.
Sorghastrum pogonostachyum (Stapf) Chippind. & Pole Evans, in Bot. Surv. S. Aft., Mem.: 247
(1948). Type: Angola, country of the Ganguellas and Ambuellas, Gossweiler 2225 (K).
Sorghum pogonostachyum Stapf, in Pram, Flor. Trop. Afr. IX: 144 (1917).
Mischanthidium gracilius Napper, in Kirkia 3: 120 (1963).
Perennial grass with a creeping rhizome, sending up a small tuft of culms from the base. Culms about
I m. tall; cuim nodes glabrous. Leaf blades narrow, glabrous, smooth except at the slightly rough tips;
leaf sheaths firm, tight, glabrous, smooth, striate; ligule short, stout, pubescent.
Inflorescence a narrow panicle up to 15 cm. long; racemes loose, containing spikelet pairs, i.e., sessile
and pedicellate spikelets. Sessile spikelet lanceolate; lower glume chartaceous, sparsely pilose on the
back, 5.5 mm. long; upper glume glabrous; callus short, round, with hairs 4 mm. long. Florets 2: lower
floret empty, lemma oblong, ciliate on the margins; upper floret bisexual, lemma entire, 3-nerved,
awned from the tip; awn bristle-like, 4 mm. long; epaleate; lodicules 2, minute, fleshy, glabrous;
anthers 3, 3 mm. long; ovary oblong. Pedicellate spikelet lanceolate, also with two florets of which the
upper is bisexual; anthers 3, 3 mm. long, pedicel pilose, some barren while others bear the bisexual
pedicellate spikelets.
Zambia. C: Central Province, Chakwenga Headwaters, 100-129 km. east of Lusaka, 27.x.1963,
Robinson 5770 (PRE); Mkushi District, Mkushi river dambo, 27.x.1967, Simon & Williamson 1222
(PRE).
East and central Africa. Wet places, especially along water courses.
Note: Many, if not most, of the pedicels bear perfect spikelets.
Sorghastrum fuscescens (Pilger) Clayton, in Kew Bull. 30(3): 509 (1975). Type: Tanzania,
Njombe District, Msima, Schlieben 1010 (K).
Miscanthidiumfuscescens Pilger, in Notizdbl. Bot. Bart. Berlin II: 806 (1933).
Tufted perennial of up to 2 m. tall. Culm nodes shortly hairy. Leaf blades linear, 10-60 cm. long, 2-7
mm. wide, densely and coarsely pubescent; leaf sheaths also pubescent; ligule a short unfringed
membrane.
Inflorescence a panicle of about 20-30 cm. long; the internodes of the main inflorescence axis pilose in
the upper third, the rest glabrous, the joints pilose; racemes loose, 3-10 cm. long, with 3-15 spikelet
pairs; raceme peduncles pilose on the upper third of their length; raceme intemodes and pedicels
pilose along their whole lengths; both pedicellate and sessile spikelets alike with a developed brown
awn. Sessile spikelet 6 mm. long, lanceolate; lower glume coriaceous, pilose on the back, lanceolate,
317
5.5 mm. long, brown; upper glume keeled in the middle, pilose on the keel only, the rest of the body
glabrous, slightly longer than the lower, 6 mm. long, ciliate on the margins. Florets 2: lower floret
empty, lemma ovate, ciliate on the margins; upper floret bisexual, lemma bibbed, ciliate on the
margins, awned; awn 25 mm. long, bigeniculate, glabrous; lodicules 2, opaque, fleshy; epaleate; callus
1 mm. long, rounded with short hair; anthers 3, 4 mm. long; ovary oblong. Pedicellate spikelet similar
to the sessile, also with a bisexual upper floret; glumes coriaceous; anthers 2 or 3, 4 mm. long;
pedicels pilose, some barren while others bear the spikelet.
Zambia. N: Mporokoso District, near Muzombwe, western side of Mweru-wa-Ntipa, altitude 3500 ft.,
15 iv. 1961, Phipps & Vesey-Fitzgerald 3218 (BM); Abercorn District, Kawimbe, altitude 5600 ft.,
26.iii.1959, Webster A239 (K). Malawi. C: Dedza, Linthipe river, 10.iii.1951, Jackson 426 (K).
East and central Africa. Grassy hillsides and lake margins.
Note: Many, if not most, of the pedicels bear perfect spikelets. This species together with S.
pogonostachyum differ with other members of Sorghastrum in this respect, but otherwise conform in
most other characters.
3. Sorghastrum incompletum (Presl.) Nash var. bipennatum (Hackel) P. D. Davila, Systematic
revision of the genus Sorghastrum (1988). Type: Sudan, Seriba Ghattas, Schweinfurth 2486 (K).
Andropogon bipennatus Hackel, Androp. Nov. Fl. 68 (8): 142 (1883).
Sorghum bipennatum (Hackel) Kuntze, Re. Gen. P1. 2: 791 (1891), based on Andropogon
bipennatus Hackel.
Sorghum bipennatum (Hackel) Stapf, Fl. Top. Aft. IX: 144 (1917), based on Andropogon
bipennatus Hackel.
Sorghastrum bipennatum (Hackel) Pilger, in Notizdbl. Bot. Gail. Berlin XIV: 96 (1938).
A weak annual up to 1.5 m. tall (rarely more than 2 m.), often decumbent and rooting at the nodes
below and with short stilt roots. Culms unbranched; cuim nodes with short shiny hairs. Leaf blades up
to 30 cm. long, linear-lanceolate, 3-10 mm. wide; leaf sheaths striate; ligule a tough membrane.
Inflorescence a loose, narrowly lanceolate panicle; raceme peduncles capillary, glabrous; racemes
reduced to a single sessile spikelet, always accompanied by 2 barren pedicels; spikelets golden in
colour, narrowly ovate, 4-5.5 mm. long, with a rounded base; lower glume coriaceous, dark brown,
ovate, pilose with white hairs on the back, 5.5 mm. long; upper glume smooth on the back, also dark
brown, boat-shaped. Florets 2: lower floret empty, lemma hyaline, ciliate along the margins; upper
floret bisexual, lemma entire, awned; awn 40 mm. long, smooth, scabrid along the spiral, bigeniculate;
palea single-nerved, small; lodicules 2, fleshy; anthers not seen; ovary ovate. Pedicellate spikelet
absent; pedicels 2, 4 mm. long, barren, pilose with white hairs.
Zambia. C: Luangwa valley, Mftiwe Camp, altitude 2000 ft., 01.iv.1966, Astle 4765 (K). B: Chipata
District (no exact location given), 17.iii.1970, Abel 182 (SRGH). Zimbabwe. N: Gokwe District (no
exact location given), 15 .iii. 1984, Mahiangu 975 (SRGH, K). Malawi. C: Dedza District, MuaLivulezi forest, 09.iv.1964, Adlard 607 (SRGH); Lilongwe District, Kampini, 17.iv.1950, Jackson 761
(MAL). 5: Zomba, Naisi road, altitude 3500 ft., 14.v.1949, Wiehe N/106 (K). Mozambique. N:
Malema District, 09.iv.1962, Lemos & Marrime 330 (PRE, K); Cabo Delgado Province: Montepuez
area, 21.iv.1984, Timberlake 3408 (SRGH); Montepuez, 09.iv.1964, Torre & Paiva 11.833 (K).
African variety of Sorghastrum incompletum: tropical areas of Africa; Sudano-Zambesian and GuineoCongo regions. Wooded grasslands.
1E
Sorghastrum friesii (Pug.) Pug., Notizdbl. Bot. Gart. Berlin XIV: 96 (1938). Type locality,
Zimbabwe.
Andropogonfriesii Pilger, in R. E. Fries, Erg. Schwed. Rhod.-Kongo Exped. 1911-1912 I:
195 (1916).
Sorghum micratherum Stapf, in Pram, Flor. Trop. Afr. 9: 142 (1917), based on Andropogon
nutans var. angolense Rendle, Cat. Afr. P1. Welw. 2(1): 152 (1899).
Sorghum friesii (Pilger) Hubbard, in Kew Bull. Misc. Inf.: 109 (1934), based on Andropogon
friesii Pilger.
Tufted, shortly rhizomatous perennial. Culms up to 120 cm. tall, often rooting from the lower nodes;
cuim nodes many (crowded) in the lower part, with inconspicuous velvety hairs; culm internodes
terete, hollow. Leaf blades glabrous, to 20 cm. long and 2-6 mm. wide, linear to linear-lanceolate,
usually reflexed, leaf base expanded; leaf sheaths non-auriculate, glabrous; ligule an unfringed
membrane, 0.5-1.5 mm. long.
Inflorescence a loose open panicle; primary branches divided; raceme peduncles glabrous, sometimes
hairy towards the apex; racemes (2-)4-7-jointed; rachis internodes and pedicels hairy. Sessile spikelet
5-7(-8) mm. long, lanceolate; callus obtuse, bearded; glumes papery; lower glume hirsute on the back,
slightly depressed longitudinally; upper glume glabrous, boat-shaped. Florets 2: lower floret
represented by a hyaline lemma, lemma ovate, 5 mm. long, finely 2-nerved, ciliate on the upper
margins, apex emarginate; upper floret bisexual, lemma bibbed, 4 mm. long, lobes glabrous, awned;
awn 3-8(-10) mm. long, straight, glabrous; epaleate; lodicules 2, tiny, fleshy, glabrous; anthers 3, 4-5
mm. long; ovary oblong; pedicel 5 mm. long, shorter than the sessile spikelet, hairy, curved towards
the base.
Botswana. N: Ngamiland, Boro flood plain, altitude 940 m., Biggs M517 (SRGFI); 11km south of
Gumare on road to Nokaneng, 18.iii.1976, Ellis 2680 (K); Boro river at Buffalo fence north of Maun,
altitude 920 m., 28.ii.1987, Long & Rae 464 (E); Lediba at Matsaudi aga Mmatalelo Island,
30.iii.1973, Smith 503 (K); Floodplain near Kwara Bochai river, 02.vi.1973, Smith 562 (E); Xudum
Drainage, Okavango, altitude 3300 ft., 15 .iii. 1961, Vesey-Fitzgerald 3241 (BM); Okavango river, 19
km. north of Shakawe on Botswana (Bechuanaland) border, 1 6.iii. 1965, Wild & Drummond 7091
(BM); Kwando, James Camp, 18°22'S 23°32'E, 11.iv.1975, Williamson 16 (K). Zambia. N: Abercom
District, Saisi river (near Jericho), 27.ii. 1958, Vesey-Fitzgerald 1547 (K). C: Mushwishi Agricultural
Station, Broken Hill, altitude 3800 ft., 08.xii.1953, Hinds 175 (K); roadside at Munshiwemba,
17.i. 1942, Stohr 744 (PRE). Zimbabwe. N: Gokwe District, about V2 mile north of Gokwe, 11 .ii. 1963,
Bingham 484 (SRGH); Sipolilo, Nyamunyeche Estate, Karoi vlei, 07.ii.1979, Nyariri 673 (PRE). W:
Shangani, Gwampa vlei at Tunke pan, Gwampa Forest Reserve, altitude 3000 ft., ii. 1956, Goldsmith
31/56 (K); Matobo District, Mtsheleli Dam, altitude 4400 ft., 20.ii.1965, Simon 160 (BM). C: Gwelo
District, G.T.C. sand, altitude 4600 ft., 10.iii.1967, Biegel 1989 (BM). 5: Victoria District, 30.iii.1973,
Chiparawasha 665 (K).
A southern African species. Typically favours wet areas such as swamps, road drains and stream
banks. Often on poor sandy soils.
Sorghastrum stipoides (Kunth) Nash, in N. Amer. Fl. 17: 129 (1912).-Clayton & Renvoize, in
Polhill, Fl. Trop. E. Afr. Gramineae Part 3: 732, fig. 169 (1982). Type: Colombia, Popaya to
Almaguer, Humboldt & Bonpland (P, K).
Andropogon stipoides Kunth, in H.B.K., Nov. Gen. Sp. 1: 189 (1816).
Andropogon trichopus Stapf, in Kew Bull. 1897: 287 (1897).
319
Sorghum trichopus (Stapf) Stapf, in F!. Trop. Afr. 9: 141 (1917).
Sorghum rigidfo1ium Stapf, in Fl. Trop. Afr. 9: 143 (1917).
Sorghastrum trichopus (Stapf) Pug., in Engl. & Pranti, Pflanzenfam., edn 2, l4e: 142 (1940);
Clayton, in Hepper, F!. West Trop. Aft., edn2, 3: 468 (1972).
Sorghastrum rigidfolium (Stapf) Chippindall & Pole Evans, in Mem. Bot. Surv. S. Aft. 22:
247 (1948); Chippindall, in Meredith, Grasses and pastures of South Afiica: 468 (1955).
A robust tufted perennial up to 1.5 m. tall with a hard creeping rhizome. Cuims erect, usually
unbranched, the internodes glabrous, nodes with a ring of inconspicuous adpressed white hairs. Leaf
blades usually glabrous, 15-45 cm. long and 3-6 mm. wide, usually inrolled, rigid and conspicuously
narrowed towards the base; leaf sheaths auriculate; ligule an unfringed membrane, 1.5-4 mm. long.
Inflorescence a linear-lanceolate panicle, about 20. cm. long, primary branches subverticillate, with
capillary peduncles; racemes 1-4-jointed; rachis internodes violet-hirsute. Spikelets all alike,
lanceolate, 4-7(-8) mm. long, the laterals each accompanied by a single densely violet-hirsute pedicel,
the terminal on each raceme by two such pedicels; callus rounded, hirsute; lower glume coriaceous,
tawny, green-nerved, pale violet-hirsute, broadly convex across the back, emarginate; upper glume
about the same length as the lower glume, paler, glabrous proximally, shortly hirsute distally,
emarginate. Florets 2: lower floret reduced a hyaline lemma; upper floret bisexual, with a linear
bibbed lemma, ciliate on the margins; awn geniculate, flattened distally, 8-16 mm. long, glabrous;
palea hyaline, minute; bodicules 2, fleshy, glabrous; anthers 3; caryopsis oblong.
Zambia. N: Mbala District, Uninji pans, altitude 5000 ft., 16.v.1968, Sanane sn. (K); wet pasture near
Lunzuo river bridge, 20 miles from Abercorn, altitude 5000 ft., 5.iv.1959, Webster A283 (K). Malawi.
N: Karonga, Mwemtete Dambo, 26.vi. 1951, Jackson 552 (K). C: Nkhotakota District, Katimbira
village, T.A. Kanyenda, 13.v.1986, Patel & Kwatha 3194 (MAL). S: Zomba, near Tongani,
20.iv.1950, BWN/492 (MAL); Mposa Dambo, 13.v.1952, Jackson 814 (K). Mozambique. MS: Beira
District, Cheringoma Coastal Area, Zuni Drainage, 5 km. west of Nyamaruza Camp, v.1973, Tinley
2842 (K). M: Sul do Save, Maputo, 20.ii.1952, Mrye & Carvaiho 1148 (K); Mahotas, Marracüene,
xii.1945, Pimenta 52565 (PRE, SRGH).
Tropical America and Tropical Africa. Low lying areas subject to flooding.
3. Sorghum Moench, Meth.: 207 (1794) nom conserv; Hubbard, in Hook. Ic. P1. 34: t 3364 (1938);
Garber, in Univ. California Pub!. Bot. 23: 283-361 (1950); Snowden, in J. Linn. Soc. Bot. 55: 191-260
(1955); Celarier, in Cytologia 23: 395-418 (1959); Ivanyukovich & Doronina, in Trudy Prikl. Bot.
Genet. Selek. 69: 18-27 (1980). Type species: Sorghum bicolor (L.) Moench.
Blumenchia Koel., Decs. Gram.: 28 (1802).
Sarga Ewart & White, in Proc. Roy. Soc. Victoria 23: 296 (1911).
Annuals or perennials, tufted or sometimes rhizomatous, mostly robust. Inflorescence a large panicle,
its primary branches simple or subdivided, bearing short dense racemes with hairy internodes. Sessile
spikelet dorsally compressed; callus obtuse or pungent; lower glume cartilaginous, convex, rounded on
the flanks but becoming 2-keeled near the tip, usually hairy; upper lemma awned or awnless; lodicules
ciliate. Pedicellate spikelet male or neuter, or reduced to a lower glume.
Species ± 20.
Nodes glabrous or pubescent.
320
Racemes tough or tardily disarticulating; gram large, commonly exposed by the gaping
1. bicolor
glume; annuals; cultivated.
Racemes fragile; grain enclosed by the glumes; perennials or annuals; weeds or wild.
Perennial with abundant creeping rhizomes; sessile spikelet epaleate; lodicules
2. halepense
ciliate; weed.
Annual; palea 2 mm. long; lodicules glabrous; wild.
3. arundinaceum
Nodes bearded.
Racemes 3-7-jointed; spikelets variously coloured; sessile spikelet 5-7 mm. long; callus short
4. versicolor
and blunt; awn 25-40 mm. long, glabrous.
Sorghum is the only genus of the Sorgh/nae for which an extensive infra-generic classification exists.
The most comprehensive subgeneric treatment of the genus was initiated by Garber (in Univ. Cal. Pub.
Bot. 23: 283-362 (1950)) who recognised six subgenera. His subgenus Sorghastrum has now been
accorded a generic status, and therefore only five subgenera now remain in the genus and these are:
(Eu)Sorghum (Stapf) Garber - Old World, e.g. S. bicolor, S. halepense (introduced to the
Americas), etc.;
Chaetosorghum Garber - Australian; monotypic, e.g. S. macrospermum;
Heterosorghum Garber - Australian; monotypic, e.g. S. lax(florum;
Parasorghum (Snowden) Garber - Old World; e.g. S. nitidum, S. versicolor, etc.;
Stiposorghum Garber - Australian; S. intrans, S. plumosum, etc.
The genus is represented by species of subgenera (Eu)Sorghum and Parasorghum in FZ. As is often
the case with genera which include domesticated taxa, the subgenus (Eu)Sorghum contains a lot of
nomenclatural inconsistencies, which has resulted in the existence of an extensive synonymy. The
subgenus consists of cultivated grain sorghums, a complex of closely related annual taxa from Africa,
and a complex of perennial taxa from Southern Europe and Asia. The subgenus is usually divided into
two groups, the Halepensia complex, and the Arundinacea complex (De Wet, in Amer. J. Bot. 65(4):
477-484 (1978)). The subgenus is now considered to consist of three species: two rhizomatous taxa, S.
halepense, and S. propinquin (both in the Halepensia), and the large and complex S. bicolor to include
all annual wild, weedy and cultivated taxa (in the Arundinacea). Snowden (The cultivated races of
sorghum (1936)) recognised 24 'species' which are all now considered variants of S. bicolor.
The grain sorghums are artifacts of human selection and cultivation (De Wet & Huckabay, in
Evolution 21: 787-802 (1967); De Wet, Harlan & Price, in Amer. J. Bot. 57(6): 704 (1970)), and the
different cultivated kinds deserve at most racial status (Harlan & De Wet, in Crop Science 12: 172-176
(1972)). The species S. bicolor has three subspecies:
S. bicolor subsp. bicolor (grain sorghums). Based on S. bicolor (L.) Moench. A detailed
comparative morphological study of grain sorghums led Harlan & De Wet (1972) to recognise five
basic races, and ten hybrid races that combine characteristics of any two or more basic races. The five
basic races are Bicolor, Guinea, Caudatum, Kaffir, and Durra.
S. bicolor subsp. arundinaceum. Based on S. arundinaceum (Desv.) Stapf. This subspecies
incorporates all wild sorghums associated with the grain sorghums. It therefore represents the 'wild'
component of S. bicolor. De Wet (in Amer. J. Bot. 65(4): 477-484 (1978)) recognises four races in this
subspecies: Aethiopicum, Virgatum, Arundinaceum and Verticilliflorum.
S. bicolor subsp. drummond/i. Based on S. drummond// (Steud.) Millsp. et Chase. This subspecies
incorporates derivatives from hybridisation between subspecies bicolor and arundinaceum. It occurs
as a weed in Africa wherever cultivated grain sorghums and their closest wild relatives are sympatric
(De Wet, in Amer. J. Bot. 65(4): 477-484 (1978)). This species is extremely variable, a variability
321
undoubtedly enhanced by human selection of grain races, which then introgress with the wild species
(Clayton, in Polhill (ed.), Fl. Trop. East Afr., Gramineae (3): 726 (1982)).
The races of Sorghum bicolor grade morphologically and ecologically so completely into one another
that they do not deserve formal taxonomic status.
In the present treatment of the genus, only S. bicolor subsp. arundinaceum (as S. arundinaceum) has
been investigated. There is no doubt that S. bicolor subsp. bicolor occurs in FZ, but no suitable
material was seen to enable a thorough investigation of this subspecies.
2. Sorghum halepense (L.) Pers., Syn. P1. 1: 101 (1805).-Bor, Flora of Assam 5: 352 (1940).Sturgeon, A revised list of the grasses of Southern Rhodesia Part IV: 6 (1954).-Chippindall, in
Meredith (ed.), The grasses and pastures of South Africa: 460 (1955).-Clayton, in Hepper (ed.), Fl.
West Trop. Afr. ed.2(3): 467 (1972).-Gibbs Russell et al, Grasses of Southern Africa: 320 (1990).Cope & Hosni, A key to Egyptian grasses: 54 (1991). Type locality, Syria.
Holcus halepensis L., Sp. P1. ed. 1: 1047 (1753).
Andropogon arundinaceus Scop., F!. Cam. ed. 2(u): 274 (1772).
Milium halepense (L.) Cay., Descr. P1.: 306 (1802).
Blumenbachia halepensis (L.) Koel., Descr. Gram.: 29 (1802).
Andropogon halepensis (L.) Brot., Fl. Lusit. I: 89 (1804).
Andropogon Sorghum (L.) Brot. subsp. halepensis (L.) Hack. var. halepensis (L.) Hack.
subvar. genuinus Hack., in DC., Monogr. Phan. 6: 502 (1889).
Andropogon halepensis (L.) Brot. var. genuinus (Hack.) Stapf, in Hook., F!. Brit. Ind. 7: 183
(1896).
Andropogon halepensis (L.) Brot. lypicus Asch. et Graebn., Syn. Mitteleur. Fl. II Abt. 1: 47
(1898).
Sorghum miliaceum (Roxb.) Snowden, in J. Linn. Soc. (Bot.) 55: 205 (1955).
Sorghum miliaceum var. parvispiculum Snowden, in J. Linn. Soc. (Bot.) 55: 209 (1955).
Perennial with abundant creeping rhizomes. Culms slender to robust, mostly simple but sometimes
branched, 0.5-3.5 m. tall, up to 2 cm. wide; cuim nodes exposed, glabrous. Leaf blades linear,
narrowed to a point at the apex, 20-90 cm. long, 0.5-4 cm. wide; leaf sheaths glabrous; ligule an
unfringed membrane.
Inflorescence a panicle; lower primary branches naked for about 2-5 cm., glabrous; racemes 1 to 5jointed; rachis internodes and pedicels slender, densely ciliate with whitish hairs, hairs up to 1 mm.
long. Sessile spikelet elliptic, 4-6.5 mm. long, almost glabrous to densely hairy, awned or awnless;
glumes coriaceous; lower glume 2-keeled, keel wings ending in minute teeth, forming with the pointed
apex a 3-toothed tip, hairy on the back, dorsally compressed, broadly lanceolate; upper glume sparsely
hairy on the back, boat-shaped; callus short and blunt, shortly bearded. Florets 2: lower floret empty,
lemma lanceolate, ciliate; upper floret bisexual, lemma bibbed, ciliate, awned or awnless; awn (when
present) 12 mm. long, glabrous, untwisted; epaleate; lodicules 2, fleshy, ciliate on the apical corners;
anthers not seen; ovary ovate, small. Pedicellate spikelet male, narrower than the sessile, 5.5 mm.
long; glumes papery; lower glume lanceolate, ciliate on the lateral keels, glabrous, 5.5 mm. long;
upper glume sharply carinate, glabrous, 5.5 mm. long; lower lemma broadly lanceolate, ciliate on the
margins, with a truncate apex, 4 mm. long; upper lemma ovate, ciliate on margins, 3 mm. long;
lodicules 2, fleshy, glabrous; anthers 3, 2.5 mm. long; pedicels hairy on the margins.
322
Zimbabwe. N: Lomagundi District, altitude 3500 ft., 8.vii.1921, NO. 3149 (K). W: Insiza, altitude
4500 ft., iii.1920, Capstick 2328 (K); Wankie District, Victoria Falls National Park, pathside in
Hyparrhenia zone near Chimunzi, altitude 2800 ft., 21.iv.1970, Simon & Hill 2129 (K). C: Harare
(Salisbury) District, Harare (Salisbury) Experiment Station, 5 .vii. 1921, Mundy SOb (K); cultivated at
Department of Agriculture Experimental Station, altitude 4800 ft., 28.xi.1931, Trapuds s.n. (K). E:
Chipinga District, East Sabi, upper Rupembe, altitude 1300 ft., 22.1.1957, Phipps 86 (K). Malawi. S:
Port Herald District, between Muone & Shire river, altitude 260 ft., 20.iii.1960, Phipps 2586 (K).
Mozambique. T: Boruma (Zambezi), v.1891, Menyharth 1046 (K).
Tropics of Old and New Worlds. In moist areas on river banks, in clay soils and wet sandy soils.
Cultivated fodder.
Note: The distinguishing features of this species are the abundant slender wide-spreading rhizomes,
narrow leaf blades, and the somewhat small contracted panicles. A frequent variation of the spikelets
is the presence or absence of an awn to the upper lemma. The majority of the plants have awned
spikelets, but sometimes in specimens of the same gathering and number, the spikelets of one panicle
are awned and those of the other awnless. It is also not uncommon to find most of the spikelets in a
panicle awnless but accompanied by a few that are distinctly awned. It is on this basis that some
authors have recognised two forms or varieties, S. halepense subsp. halepense (the typical awned
form), and S. halepense subsp. muticus (the awnless form).
3. Sorghum arundinaceum (Desv.) Stapf, in Pram, Fl. Trop. Afr. LX: 114 (1917).-Sturgeon, A
revised list of grasses of Southern Rhodesia Part IV: 6 (1954).-Clayton, in Hepper (ed.), Fl. West
Trop. Afr. ed.2(3): 467 (1972).-Clayton, in Polhill (ed.), Fl. Trop. East Afr., Gramineae (3): 727
(1982).-Lowe, Flora of Nigeria, Grasses: 277 (1989)-Cope & Hosni, A key to Egyptian grasses: 54
(1991). Type: Ghana, Isert (B).
Andropogon arundinaceus Willd, Sp. P1. 4: 906 (1805).
Rhaphis arundinacea Desv., Opusc.: 69(1831).
Andropogon sorghum var. effusus Hack., in DC., Monogr. Phan. 6: 503 (1889).
Andropogon sorghum effusus Piper, in Proc. Biol. Soc. Wash. 28: 35 (1915).
Sorghum verticillflorum (Steud.) Stapf, in Pram, Flor. Trop. Afr. IX: 114 (1917).
Sorghum stapfii (Hook.f.) Fischer, in Gamble, F.P.I. 10: 1735 (1934).
Sorghum pugionfolium Snowden, in J. Linn. Soc. (Bot.) 5: 240 (1955).
Short-lived perennial or an annual. Culms 0.3-4 m. tall, often robust, branched; culm nodes mostly
glabrous, sometimes pubescent. Leaf blades variable, often large, 5-75 cm. long, 5-7 mm. wide, flat,
broadly lanceolate, glabrous on both surfaces, with a prominent whitish midrib, tapering to a fine
point; leaf sheaths glabrous, not keeled; leaf ligule a fringed membrane, edged with a fringe of fine
hairs and hairy on the back.
Inflorescence a broadly spreading panicle, 10-60 cm. long; main inflorescence axis angular, glabrous;
primary branches divided, pubescent at the nodes; raceme peduncles glabrous; racemes 2-7-jointed;
rachis internodes and pedicels hairy. Sessile spikelet (4-)7(-9) mm. long, lanceolate to narrowly ovate;
glumes coriaceous; lower glume white pubescent, sometimes tomentose or fulvously pubescent,
slightly depressed longitudinally on the back; upper glume glabrescent or with sparse hairs on the
back. Florets 2: lower floret empty, lemma lanceolate, 5.5 mm. long, ciliate on the margins; upper
floret bisexual, lemma deeply lobed, ciliate on the lobes and margins, 3 mm. long; awn 20 mm. long,
glabrous; palea present, 2 mm. long, ciliate on margins and apex; lodicules 2, fleshy, glabrous; no
anthers seen; caryopsis obovate, 3 mm. long. Pedicellate spikelet neuter, linear to lanceolate, 6.5 mm.
323
long; glumes papery; lower glume glabrous; upper slightly shorter than the lower, glabrous; lower
lemma glabrous, with a truncate apex; pedicel hairy.
Botswana. N: Tati river, south-east of Francistown, altitude 980 m., 04.iii.1985, Long 12242 (B);
Chadum valley, South-west African border, 16 km. west of Knau Knau, 14.iii. 1965, Wild &
Drummond 7024 (BM, SRGH). Zambia. N: Mpika District, Mfuwe, 26.iv.1965, Mitchell 2683 (BM,
SRGH); 04.i.1963, Rensburg 1161 (WAG). Zimbabwe. N: Urungwe District, Mana pools floodplain,
31 .iii. 1981, Dunham 73 (SRGH). C: Harare (Salisbury): Crowborough Sewage farm, Mufakose
township, 06.ii.1974, Lenton s.n. (WAG); Harare (Salisbury) Research Station, cultivated, 22.ii.1974,
Simon 2386 (PRE). Malawi. C: Nkhota Kota, altitude 490 m., I 7.vi. 1970, Brummitt 11519 (MAL);
Dedza District, Masasa controlled area, 22.v.1989, Chikuni, Pate! & Nachamba 79 (MAL); Salima
District, Lifidzi Breeding Centre, 21.v.1985, Pate! & Nachamba 2192 (MAL); Salima District,
Chipoka, 25.v.1972, Salubeni 1811 (MAL). 5: Tanga District, Palombe river, 06.v.1952, Jackson 802
(MAL); Njanje District, Mwanalundu Hills, ll.i.1990, Salubeni & Nachamba 5615 (MAL).
Mozambique. Z: Namagoa, Mocuba, (no date), Faulkner 37 (PRE). T: Baroma Province, Msusa,
Zambesi valley, 25.vii.1950, Chase 2805 (BM); Tete, Cabora Bassa, altitude 215-220 m., 09.v.1972,
Pereira & Correia 2469 (WAG). GI: Gaza Province, Chibute, Miniquenique, Estacao Experimental do
C.I.C.A., 13.vi.1960, Lemos & Balsinhas 98 (PRE). M: Maputo Province, Umbeluzi State farm, near
Boane, 19.ii.1985, Timberlake 3386 (SRGH).
Throughout Africa, extending eastwards to Australia. Swampy soils, streamsides, disturbed places and
old farmland.
Note: This species is assumed to be the wild progenitor of the grain sorghum, S. bicolor. The
variability in the species can be attributed to human selection of grain races, and introgression with the
wild species.
4. Sorghum versicolor Anderss., in Peters, Reise Mossamb.: 563 (1863).-Stapf, Fl. Trop. Afr. IX: 138
(1917).-Sturgeon, A revised list of the grasses of Southern Rhodesia Part IV: 6 (1954).-Chippindall, in
Meredith (ed.), The grasses and pastures of South Africa: 459 (1955).-Bogdan, A revised list of Kenya
grasses: 55 (1958)-Jackson & Wiehe, An annotated check-list of Nyasaland grasses: 60 (1958),Napper, Grasses of Tanganyika: 96 (1965).-Hood, A guide to the grasses of Zambia: 58 (1967).Clayton, in Polhill (ed.), Fl. Trop. East Afr., Gramineae (3): 729 (1982).-Gibbs Russell et al, Grasses
of Southern Africa: 302 (1990). Type: Mozambique, Boror, Peters (whereabouts uncertain).
Andropogon serratus var. versicolor Hack., in DC., Monogr. Phan. 6: 522 (1889).
Sorghum purpureo-sericeum (A. Rich.) Aschers. & Shweinf. var. trinervatum Chiov., Fl.
Somal 2: 439 (1932).
An annual of up to 2.5 m. tall. Cuims slender, with a very conspicuous ring of long, silky white hairs
spreading round the cuim from each node. Leaf blades hairy on both surfaces, occasionally glabrous
on the upper surface, 10-30 cm. long, up to 15 mm. wide; leaf sheaths glabrous, bearded with long
cilia at the mouth; ligule an unfringed membrane, to 3 mm. long.
Inflorescence an open panicle to 25 cm. long; branches simple; main inflorescence axis glabrous,
smooth, bearded at the nodes; raceme peduncles glabrous, smooth, with prominent pulvini; racemes 37-jointed; rachis internodes and pedicels long hairy. Sessile spikelet elliptic-oblong, (5-)5.5(-7) mm.
long; lower glume coriaceous, glossy, reddish brown to black, pilose on the back, bearded from the
callus with pallid to reddish hairs; upper glume also coriaceous and same colour as lower, tapers to a
fine apex, pilose above the middle, glabrous and shiny below. Florets 2: lower floret empty, lemma 4
mm. long, ciliate on the margins; upper floret bisexual, lemma 3 mm. long, deeply lobed, ciliate on the
lobes and margins, awned; epaleate; awn (25-)40 mm. long, twisted, scabrid along the spiral; lodicules
2, fleshy, densely ciliate from the apical corners; anthers 3, 2.5 mm. long; ovary oblong. Pedicellate
324
spikelet neuter, smaller than sessile [(3-)4(-5) mm. long], lanceolate, greenish; lower glume
membranous, hirsute on the back; upper glume also membranous, glabrous to scabrid, enclosed by the
lower; pedicel hairy.
Botswana. N: 6 km. south-east of Tsau, 18.iii.1965, Wild & Drummond 7132 (BM, SRGH). Zambia.
C: Mumbwa District, Sala Reserve, Cheta river, 26.iii.1963, Vesey-Fitzgerald 4019 (SRGH). S:
Kalomo District, Siantambo, 07.ii.1963, Mitchell 17/63 (BM). Zimbabwe. W: Bubi District, Gwampa
Forest Reserve, ii. 1956, Goldsmith 6/56 (PRE). C: Gwelo District, Base, Gwelo kop., altitude 4700 ft.,
05.iii.1967, Biegel 1959 (BM). E: Chimanimani (Melsetter) District: Mutambara Tribal Trust Land on
main road near Lisnacloon farm, altitude 1000 m., 26.iii.1969, Crook 850 (PRE); Sabi Valley,
Nyanyadzi, 26.iii.1956, Whellan 1009 (WAG). S: Victoria District: Fort Victoria, Victoria Reserve
Zone, 19.iii.1956, Cleghorn 185 (BM); Fort Victoria area, 02.i.1949, Robinson 336 (SRGH). Malawi.
N: Northern Province, Nzimba District, 25 miles south of Rumphi on Ml, altitude 4000 ft., 04.v.1974,
Pawek 8571 (PRE, WAG); Zomba District, near Macheleni hill, Tembenu village, 05.iii.1978, Seyani
& Patel 800 (MAL). Mozambique. N: Malema, Mutuáli, Estacão Experimental do Instituto do
Algodo, 05.iv.1962, Lemos & Marrime 318 (PRE). T: Tete, Cabora Bassa, altitude 230-330 m.,
02.v.1972, Pereira & Correia 2328 (WAG). M: Maputo (Laurenco Marques), Macaëne-Magude,
22.i.1948, Torre 7179 (BM).
Eastern, central and southern Africa. In deciduous bushland or wooded grassland, commonly on
waterlogged soils or black clays.
Note: This species and S. purpureo-sericeum (A. Rich.) Aschers. & Schweinf, are so very similar that
they may just represent a variation in one species. Differences between them mainly comes from their
spikelet length (shorter in S. versicolor), their apparent inability to hybridise (Garber, 1950), and their
differing geographical distributions.
4. Chrysopogon Trin., Fund. Agrost.: 187 (1820); nom. conserv. Type species: Chrysopogon gryllus
(L.)Trin.
Rhaphis Lour., Fl. Cochin.: 552 (1790).
Pollinia Spren., P1. Pugill. 2: 10 (1815).
Centrophorum Inn., Fund. Agrost.: 106 (1822).
Trianthium Desv., Opusc.: 69(1831).
Chalcelytrum Lunell, in Amer. Midl. Nat. 4: 212 (1915).
Tufted perennials; ligule a short membrane or a line of hairs. Inflorescence a panicle, its primary
branches whorled, simple, each bearing a raceme reduced to a triplet of spikelets (rarely racemes with
2 sessile spikelets, e.g. C. sylvaticus); spikelets in triplets, or in triplets and pairs; the triplet consists of
a sessile spikelet and two pedicellate spikelets; pedicels linear filiform, never longitudinally grooved.
Sessile spikelet laterally compressed, its callus elongated, acute to pungent; glumes subequal,
unequally awned or aristate or awnless; lower glume cartilaginous to coriaceous, often spinulose on
the keels and muricate on the back; upper lemma entire or bibbed, awned; awn glabrous to pubescent
and usually prominent. Pedicellate spikelet male or neuter; glumes awned or aristate or awnless.
Species 26.
1. Chrysopogon serrulatus Trin., in Mem. Acad. Sci. Petersb. Ser. 6(2): 318 (1832).-Bor, Grasses
Burma Ceylon India & Pakistan: 118 (1960).-Clayton & Renvoize, in Polhill (ed.), Flor. Trop. East
325
Africa: 736 (1`982).-Gibbs Russell et at, Grasses of Southern Africa: 85 (1991). Type: Nepal, Wallich
(LE).
Andropogon trinii Steud., Syn. P1. Glum. 1: 395 (1854).
Andropogon ciliolatus Steud., Syn. P1. Glum. 1: 396 (1854).
Andropogon caerulens Steud., Syn. P1. Glum. 1: 395 (1854).
Chrysopogon ciliolatus (Steud.) Boiss., F!. Orient. 5: 458 (1884).
Andropogon trinii Steud. var. increscens Hack., in DC., Monogr. Phan. 6: 558 (1889).
Andropogon monticola Schutt. var. trinii Hook.f., Fl. Brit. Ind. 7: 193 (1896).
Chrysopogon montanus Trin. var. serrulatus (Trin.) Stapf, in Pram, Fl. Trop. Afr. IX: 160
(1917).
Chrysopogon montanus Trin var. tremulus (Hack.) Stapf, in Pram, Flora of Tropical Africa
IX: 160 (1917).
Chrysopogonfulvus (Spreng.) Chiov. var. tremulus (Hack.) Chiov., Rac. Bot. Miss. Consol.
Kenya: 325 (1935).
Chrysopogonfulvus (Spreng.) Chiov. var. serrulatus (Trin.) R. B. Stewart, in Brittonia 5: 446
(1945).
Perennial. Cuims up to 150 cm. tall, sometimes robust. Leaf blades glabrous except in the ligule area
where there are sparse tubercle-based hairs on the margins on both surfaces, linear-lanceolate, 2-10
mm. wide; leaf sheath glabrous, rarely hairy at the top; leaf ligule a ciliolate rim.
Inflorescence particulate, up to 15 cm. long; consists of racemes with peduncles fulvously bearded at
the tips. Sessile spikelet narrowly oblong, 6 mm. long; lower glume awnless, slightly hispid at the tip,
6 mm. long; upper glume with few white hairs on the back, with a glabrous awn about 12 mm. long,
obscurely hispidulous. Florets 2: lower floret empty, lemma hyaline, single-nerved with sparsely
ciliolate margins; upper floret bisexual, lemma hyaline, bibbed, awn 2.5 cm. long, slightly
puberulous, ovary about 1 mm. long, completely enclosed by 2 fleshy lodicules, anthers 3, 3 mm.
long. Pedicellate spikelet male or neuter, glumes awned; lower glume 6 mm. long with white short
hairs on the back, awn 5 mm. long, hispidulous; upper glume glabrous, 6.5 mm. long, awn 2.5 mm.
long, obscurely hispidulous, lemma hyaline, sparsely ciliolate on the margins; anthers 3, 3 mm. long;
pedicel 2.5 mm. long, fWvously villous unilaterally.
Botswana. SE: Gaborone, Aedume Park, altitude 1050 m., 09.xi.1977, Hansen 3276 (K). Zimbabwe.
N: Urungwe District, altitude 1700 ft., 16.x.1957, Phipps 797 (K). W: Nyamandhlovu District (no
exact location), 21.ix.1956, Plowes 1880 (K). 5: Nuanetsi District, Sengwe Tribal Trust Land,
15.xi.1973, Cleghorn 2910 (K); Gwanda District, Tuli Pasture Research Sub-Station, altitude 2500 ft.,
5.i.1955, Oates 1742 (E).
Tropical Africa; Northwest India. Common in shallow sandy soils and rocky landscape.
Note: This species is very similar to C. plumulosus except that it has a longer callus (up to 1.5 mm.
long), longer upper glume of sessile spikelet (4-7 mm. long) and the glabrous or obscurely hispidulous
(not plumose) awns of the glumes of both the sessile and pedicellate spikelets. Also the upper lemma
apex is bibbed in C. serrulatus and entire in C. plumulosus.
5. Vetiveria Bory, in Lem.-Lisanc., Bull. Sci. Soc. Philom: 42 (1822). Type species: Vetiveria
zizanioides (L.) Nash.
326
Lernomandia Steud., in Flora 33: 229 (1850).
Mandelorna Steud., Syn. P1. Glum. 1: 359 (1854).
Tufted perennials with tough rhizomes; lower leaf sheaths much compressed, flabellate-imbricate;
conduplicate leaf blades that do not bend away from the culm at the collar region but just a lateral
indentation marks the end of the sheath and the beginning of the blade; leaf sheaths keeled.
Inflorescence a panicle, its primary branches whorled, simple or each bearing a raceme; racemes (2-)3
to many-jointed, pedunculate; spikelets paired, compressed laterally, differing in sex, one sessile, the
other pedicellate; rachis internodes and pedicels slender and gradually thickened upwards. Sessile
spikelet laterally compressed, its callus obtuse to pungent, often large and conical; lower glume papery
to coriaceous, muticus, spinulose on the keels and muricate or mammilate on the back; upper glume
aristate; upper lemma entire or bibbed, awned or awnless. Pedicellate spikelet male or neuter.
Species 10.
Sessile spikelet mucronate or awnless.
1. zizanioides
Sessile spikelet awned; awn 5 mm. long.
2. nigritana
1. Vetiveria zizanioides (L.) Nash, in Small, Fl. Southeast U. S.: 67 (1903).-Stapf, in Kew Bull. 1906:
346-349, 362 & in Pram, Flor. Trop. Africa IX: 157 (1917).-Sturgeon, A revised list of the grasses of
Southern Rhodesia Part IV: 8 (1954).-Chippindall, in Meredith (ed.), The grasses and pastures of
South Africa: 470 (1955).-Jackson & Wiehe, An annotated check-list of Nyasaland grasses: 70
(1958).-Bor, Grasses Burma Ceylon India & Pakistan: 258 (1960).-Napper, Grasses of Tanganyika: 99
(1965).-Clayton, in Hepper (ed.), Flor. West Trop. Africa ed.2 (3), Gramineae: 470 (1972)-Lowe, The
flora of Nigeria, Grasses: 291 (1989). Type: India, Koenig (LINN).
Phalaris zizanioides L., Mant. P1. 2: 183 (1771).
Andropogon muricatus Retz., Obs. Bot. 3: 43 (1783).
Agrostis verticillata Lamk., Ecycl. Meth. Bot. 1: 59(1783).
Anatherum muricatum (Retz.) P. Beauv., Ess. Agrost.: 150 (1812).
Vetiveria odoratissima Lem.-Lisanc., in Bull. Soc. Philom. (Paris): 43 (1822).
Vetiveria odorata Vivey, J. Pharm. 13: 501 (1827).
Andropogonfestucoides J. S. Presl ex C. B. Presl, Rel. Haenk. 1: 340 (1830).
Vetiveria muricata (Retz.) Griseb., Fl. Brit. West Ind.: 560 (1864).
Vetiveria arundinacea Griseb., Fl. Brit. West. Ind.: 559 (1864).
Sorghum zizanioides (L.) 0. Ktze., Rev. Gen. P1. 2: 791 (1891).
Andropogon zizanioides (L.) Urban, Symb. Antill. 4: 79 (1903).
Holcus zizanioides (L.) 0. Ktze. ex Stuck., in Ann. Mus. Nac. Buenos Aires 11: 48 (1904).
Anatherum zizanioides (L.) Hitchc. et Chase, in U.S. Nati. Herb. Contrib. 18: 285 (1917).
Perennial grass of up to or greater than 2 m. tall; culms erect, usually sheathed all along. Leaf blades
linear, acute, up to 90 cm. long, 4-20 mm. wide, erect, rigid, firm or somewhat spongy, striate; leaf
sheaths compressed, particularly the lower ones, sharply keeled, fan shaped, imbricate, very smooth;
ligule a scarious rim.
327
Inflorescence a long oblong panicle of up to and over 30 cm. long, contracted; the main inflorescence
axis stout, glabrous, smooth; whorls 6-10 with up to 20 pedunculate racemes; peduncles fihiform,
slightly scabrous; racemes very slender; rachis internodes about the same length as spikelets or slightly
longer, glabrous; pedicels shorter than the sessile spikelets, glabrous. Sessile spikelet linear-lanceolate,
variable in colour (yellowish, olive, violet brown, purplish); callus obtuse, 1 mm. long or less,
glabrous; lower glume muricate on the back, spinulose on the keels; upper glume spinulose on the
keel. Florets 2: lower floret empty, lemma as long as the glumes, acute, retrorsely ciliate; upper floret
bisexual, lemma up to 3 mm. long, narrow, oblong-lanceolate, mucronate, glabrous; lodicules 2,
fleshy, glabrous; anthers 3, 2-3 mm. long; ovary ovate, 3.5 mm. long. Pedicellate spikelet sparingly
aculeolate or almost smooth, male; lower lemma same as for sessile spikelet; upper lemma acute,
entire; anthers 3, 2.5 mm. long; pedicel glabrous.
Zimbabwe. N: Mazoe District, Makalanga, cultivated, 22.v. 1950, Pollitt s.n. (K). E: Chipinga District,
Newcastle Block, La Lucie Farm, introduced, cultivated, 19.iii. 1981, Nicoll s.n. (K).
Southeast Asia to Tropical Africa; introduced to most parts of the world. Mostly in cultivated fields
and basically in places associated with the function it is intended for, e.g., it might be grown as edging
to roads. Principally used for thatching; not grazed except when young and tender. Its root is the
scented khas khas which is used for tattis or grass screens which are kept wetted to cool rooms.
Commercial oils are extracted from its roots and used in perfumery.
2. Vetiveria nigritana Stapf, in Pram, F!. Trop. Afr. IX: 157 (1917).-Sturgeon, A revised list of the
grasses of Southern Rhodesia Part IV: 8 (1954).-Chippindall, in Meredith (ed.), The grasses and
pastures of South Africa: 469 (1955).-Napper, Grasses of Tanganyika: 99 (1965).-Hood, A guide to
the grasses of Zambia: 58 (1967).-Clayton, in Hepper (ed.), Flor. West Trop. Africa ed.2 (3),
Gramineae: 470 (1972).-Clayton, in Polhill (ed.), Flor. Trop. East Africa Gramineae (3): 739 (1982).Lowe, The flora of Nigeria, Grasses: 291 (1989)-Gibbs Russell et al, Grasses of Southern Africa: 353
(1991). Type: Nigeria, Nun R., Vogel (K).
Andropogon nigritanus Benth., in Hook., Niger Fl.: 573 (1849).
Andropogon squarrosus var. nigritanus Hack., In DC., Monogr. Phan. 6: 544 (1889).
Mandelorna insignis Steud., Syn. Pt. Glum. 1: 359 (1854).
Vetiveria zizanioides (L.) Nash var. nigritana (Benth.) A. Camus, in Bull. Mus. Hist. Nat.
Paris 25: 674 (1919).
Tufted perennial. Culms 1.5-3 in. tall, unbranched: upper culm nodes exposed, glabrous, lower ones
hidden by leaf sheaths. Leaf blades narrow, up to 90 cm. long and 7 mm. wide, cauline; leaf sheaths
sharply keeled; ligule a scarious rim.
Inflorescence a terminal panicle of whorled pedunculate racemes; racemes many-jointed; the main
inflorescence axis and raceme peduncles minutely ciliolate towards the apex (upward facing cilia);
whorls 8-10 and with up to 15 racemes; rachis internodes and pedicels gradually thickened upwards.
Sessile spikelet: lower glume compressed laterally, coriaceous, spinulose on the back, narrowly linearlanceolate, 7 mm. long; upper glume sharply keeled in the middle, spinulose along the keel, coriaceous
to chartaceous, but with an inflexed margin, drawn to a short aristate tip, 7 mm. long. Florets 2: lower
floret represented by a hyaline lemma, lemma ovate; upper floret bisexual, lemma hyaline, with a
bibbed apex; awn 5 mm. long, slightly exserted from the glumes or enclosed, bristle-like, glabrous on
the lower half (column) and minutely toothed on the bristle; lodicules 2, fleshy, glabrous; anthers not
seen; ovary oblong-lanceolate. Pedicellate spikelet neuter, shorter than the sessile one; both glumes
less coriaceous, and less spinulose; lower glume 5 mm. long, sparingly aculeolate along the keel
328
towards the apex or almost smooth; upper glume smooth with flexible ciliate margins; lower lemma
hyaline with ciliate margins; pedicel solid, glabrous, thickened upwards.
Botswana. N: Okavango river, Old Mohembo, 16 km. north of Shakawe on Botswana border,
19.iii.1965, Wild & Drummond 7074 (BM). Zambia. B: Mankoya District, near resthouse, Mankoya,
20.xi.1959, Drummond & Cookson 6644 (E, BM). C: Kafue District (no exact location), 09.ii.1963,
Van Rensberg 1359 (BM). Zimbabwe. N: Gokwe District, Copper Queen N.R.A, near Morowa river,
21.xii.1963, Bingham 901b (BM). Malawi. S: Port Herald, Tengani, 28.xi.1950, Jackson 310 (MAL);
Khonjeni Estate, Thyolo, altitude 2400 ft., 16.xii.1974, Schwarz 14 (M-AL). Mozambique. 1: Boroma
Province, Ulere station, Zambezi river, lO.vii.1956, Chase 2671 (BM); Chitengo, 13 km. to
Acampamento, 02.v.1978, Diniz 163 (WAG).
Tropical Africa. Occurs mainly on river banks, occasionally on roadsides and forested areas.
6. Capillipedium Stapf, in Pram, Fl. Trop. Aft. IX: 169 (1917). Type species: Capillipedium
parvfiorum (R. Br.) Stapff,
Filipedium Raiz. & Jam, in J. Bombay Nat. Hist. Soc. 49: 682 (1951).
Annuals or perennials; mostly rambling. Inflorescence a delicate, loose panicle, bearing short 1-5(-8)jointed racemes at the ends of capillary primary and secondary branches; rachis intemodes and
pedicels finely fihiform, with a translucent longitudinal groove. Sessile spikelet callus obtuse; lower
glume cartilaginous, broadly convex to slightly concave, acute or obtuse; upper lemma hyaline base of
the awn, entire; awn glabrous. Pedicellate spikelet male or neuter.
Species ± 14.
1. Capillipedium parviflorum (R. Br.) Stapf, in Pram, F!. Trop. Aft. IX: 169 (1917).-Bor, Flora of
Assam 5: 362 (1940)-Sturgeon, A revised list of grasses of Southern Rhodesia Part IV: 8 (1954).Bogdan, A revised list of Kenya grasses: 56 (1958).-Jackson & Wiehe, An annotated check-list of
Nyasaland grasses: 32 (1958).-Bor, Grasses Burma Ceylon India & Pakistan: 112 (1960).-Napper,
Grasses of Tanganyika: 99 (1965)-Clayton, in Polhill (ed.), Fl. Trop. East Afr., Gramineae (3): 718
(1982). Type: Australia, North Coast, Brown 6188 (BM!, K!).
Holcusparvflorus R. Br., Prodr.: 199 (1810).
Sorghum parv ?Ilorum Beauv., Agros: 132 (1812).
Anatherumparvflorum Spreng., Syst. 1: 290 (1825).
Holcus caerulescens Gaud., in Freycin. Voy. Bot.: 411 (1829).
Andropogon micranthus Kunth, Rev. Gram. I: 165 (1829); Steud., Syn. P1. Glum. 1: 396
(1854); Hack., in DC., Monogr. Phan. 6: 488 (1889); Hook.f., Fl. Brit. Ind. VII: 178 (1896).
Andropogon alternans J. S. Presi., in C. B. Presi., Reliq. Haenk. 1: 342 (1830).
Rhaphis caerulescens Desv., Opusc.: 69(1831).
Chrysopogon violascens Trin., in Mem. Ac. Petersb. 6 me ser. II: 319 (1832).
Andropogon quartianianus A. Rich., Tent. Fl. Abyss. II: 469 (1851).
Andropogon violascens Nees ex Steud., Syn. P1. Glum. 1: 396 (1854).
Andropogonparvispicus Steud., Syn. P1. Glum. 1: 397 (1854).
Andropogon capillflorus Steud., Syn. P1. Glum. 1: 397 (1854).
329
Rhaphis villosula Nees ex Steud., Syn. P1. Glum. 1: 397 (1854).
Rhaphis microstachya Nees ex Steud., Syn. P1. Glum. 1: 397 (1854).
Andropogon serratus Miq., Ann. Mus. Lugd.-Bat. II: 290 (1866).
Chrysopogonparv(florus Benth., Fl. Austr. VII: 537 (1878).
Chrysopogonparvispicus (Steud.) Wats., in Atkins., Gaz. N. W. Ind.: 392 (1882).
Tufted perennial, mostly 75-100 cm. tall. Culms with bearded nodes. Leaf blades glaucous, the blades
expanded, erect and straight, linear from an often narrowed and slightly contracted base, 30 cm. long,
2.5 mm. wide, fmely pointed, margins scabrid; leaf sheaths tight or at length slipping from the cuims,
the lowermost ones longer, the others usually shorter than the internodes, more or less with tuberclebased hairs and frequently villous on the collar; ligule a very short, truncate, ciliolate membrane.
Inflorescence a profusely divided delicate panicle; main inflorescence axis terete; primary branches
alternate, opposite or semiverticillate, divided from low down and up to the third degree; raceme
peduncles filiform, glabrous. Racemes 1-jointed, reduced to 1 sessile and two pedicellate spikelets,
rarely 2-jointed with 2 sessile and 3 pedicellate spikelets; rachis intemodes very similar to the raceme
peduncles and glabrous. Sessile spikelet narrowly oblong, 2.8-5 mm. long; lower glume with short
hairs on the back, shallowly concave in the middle, ciliate upwards on the keels, truncate, with 2
intracarinal nerves; upper glume single-keeled, scabrid on the keel, the median nerve drawn into a
mucro. Florets 2: lower floret empty, lemma 2-2.5 mm. long, ovate; upper floret bisexual, lemma
stipiform, linear, flattened, entire, single-nerved; awn 20 mm. long (including the upper lemma),
minutely hispidulous; lodicules 2, fleshy, glabrous; anthers 3, 1.5 mm. long; caryopsis oblong, 2.1
mm. long. Pedicellate spikelet lanceolate, male or neuter, 2.5-3.5 mm. long; lower glume shortly hairy
on the keels, truncate, with 3 intracarinal nerves, acute; upper glume glabrous, lanceolate, with 2
intracarinal nerves; lower lemma oblong, truncate; anthers 3, 1.5 mm. long; pedicels with a median
translucent longitudinal groove, subequal, hairy on the raised margins.
Zambia. E: Fort Jameson, 25.v.1963, * Verboom 761 (SRGH, BM). Zimbabwe. C: Gwelo District, 6
miles south from Gwelo, altitude 4600 ft., 18.iii.1967, Biegel 2000 (BM); C: Harare (Salisbury)
District, Ballantyne Park, altitude 4950 ft., 26.iii.1965, Crook 730 (BM); Harare (Salisbury),
university grounds near main entrance gate, altitude 1500 m., 01.i.1968, Crook P55 (PRE);
Marlborough, ii.1954, Kerr 6 (PRE); wet ground bordering Gwebi river, Mazoe road, 22.ii.1974,
Simon 2385 (SRGH, PRE). Malawi. C: Dowa District, Uperere Mission, Chankalamu Dambo,
02.xi. 1950, Jackson 249 (MAL); Agricultural Research Station near Lilongwe, 30.ix. 1952, Kantikana
4 (BM); Chipata mountain, altitude 6000 ft., 04.v. 1963, * Verboom 978 (SRGH).
* - racemes 2-jointed (i.e.. with 2 sessile and 3 pedicellate spikelets).
Widely distributed throughout the tropics of the Old World, China and Japan. Grassland.
7. Bothriochloa Kuntze, Rev. Gen. P1. 2: 762 (1891). Type species: Bothriochloa bladhii (Retz.) S. T.
Blake.
Andropogon subgen. Gymnoandropogon Nees, Fl. Afr. Austr.: 103 (1841).
Gymnandropogon (Nees) Duthie, in Atkinson, Gaz. N.W. Prov. & Oude 10: 638 (1882).
Amphilophis Nash, in Britton, Man. Fl. North. States: 71(1901).
Perennials. Inflorescence of few to many-jointed pedunculate digitate or subdigitate racemes, the
branches undivided or the lower ones sparsely branched; racemes without homogamous spikelet pairs,
330
always terminate in a triad; spikelets paired, one sessile, the other pedicellate, similar in shape or the
pedicellate reduced and smaller, always differing in sex; rachis internodes and pedicels linear-filiform,
with a translucent longitudinal groove. Sessile spikelet callus obtuse; lower glume mostly
cartilaginous, broadly convex to slightly concave, sometimes with 1-3 circular pits on the back, acute;
upper lemma hyaline base of the awn, entire (bibbed in B. biloba and B. erianthoides). Pedicellate
spikelet much like the sessile or smaller, male or neuter.
Species ± 35.
1. radicans
Lower glume of the sessile spikelet not pitted.
Lower glume of the sessile spikelet pitted.
Inflorescence paniculate; lower glume of pedicellate spikelet not pitted
2. bladhil
Inflorescence subdigitate; lower glume of pedicellate spikelet with 0-4 slit-like pits.
3. insculpta
1. Bothriochloa radicans (Lehm.) A. Camus, in Aim. Soc. Linn. Lyon 1930 n.s. LXXVI: 164
(1931).-Sturgeon, A revised list of grasses of Southern Rhodesia Part IV: 9 (1954).-Chippindall, in
Meredith (ed.), The grasses and pastures of South Africa: 48 (1955).-Bogdan, A revised list of Kenya
grasses: 56 (1958).-Napper, Grasses of Tanganyika: 100 (1965).-Hood, A guide to the grasses of
Zambia: 57 (1967).-Clayton, in Polhill (ed.), Fl. Trop. East Afr., Gramineae (3): 721 (1982).-Müller,
Grasses of South West Africa/Namibia: 86 (1984).-Chaudhary, Grasses of Saudi Arabia: 420 (1989).Gibbs Russell et al, Grasses of Southern Africa: 63 (1991). Type: cultivated in Hamburg, seed from
South Africa (whereabouts uncertain).
Andropogon radicans Lehm., in Ind. Sem. Hort. Hamb. (1828).
Andropogon ischaemum var. radicans Hackel, in DC., Monogr. Phan. 6: 476 (1889).
Amphilophis radicans Stapf, in Pram, Fl. Trop. Afr. IX: 172 (1917).
A tufted, often stoloniferous perennial of up to 70 cm. tall; has a shrubby growth form and or can be
decumbent (including rooting at the nodes). Culms herbaceous and branching from the lower nodes to
form cushions; culm nodes exposed and bearded, the lower ones with small velvety hairs and the
upper glabrous; cuim internodes terete. Leaf blades linear-lanceolate, 6-20 cm. long, 2-6 mm. wide,
gradually tapering to a fine point; sheaths glabrous except at the usually bearded mouth, those of the
basal leaves loose and open, others closely clasping the culms; leaf ligule a ring of white hairs.
The inflorescence is of clustered 5-16 hairy subdigitate racemes up to 70 mm. long; the main
inflorescence axis is slender and shorter than the lower racemes; racemes 10-20-jointed, shortly
pedunculate; rachis internodes and pedicels linear, with a longitudinal translucent groove in the
middle; rachis joints densely long hairy. Sessile spikelet lanceolate, 2.5-4 mm. long; lower glume
slightly concave on the back and hairy on the lower half, firmly membranous; upper glume with an
acute apex, with a single keel, minutely hispid along this keel towards the apex. Florets 2: lower floret
empty, lemma oblong-linear, 3 mm. long, ciliolate at the apex; upper floret bisexual, lemma stipiform,
glabrous, 1.5 mm. long, awned; awn twisted, geniculate, 25 mm. long; epaleate; lodicules 2, fleshy
and glabrous; callus short and blunt; anthers 3, 1.5 mm. long; ovary oblong. Pedicellate spikelet
neuter; lower glume 3.5 mm. long, oblong, with sparse long white hairs along the lateral keels and
ciliolate towards the apex on these keels; upper glume flimsy, ciliate on margins, lanceolate, 3 mm.
long; lower lemma oblong, obtuse, truncate, 2 mm. long; pedicels linear.
Botswana. N: Sigara Pan, 30 miles west of Nata river, altitude 896 m., 26.iv.1957, Drummond &
Seagrief 5240 (K, SRGH); Nxai Pan National Park, 14.iii.1976, Smith 1654 (K). Zimbabwe. W:
331
Hlekwani, approximately 10 km. south-west of Bulawayo, altitude 1350 m., 19.xii.1990, Laegaard
1528 (K). S: Nuanetsi District, Nuanetsi Ranch, Bubi Section, 23.ii.1967, Cleghorn 1421 (K).
Tropical Africa; introduced to tropical America. Open deciduous bushland.
Native pasture species.
2. Bothriochloa bladhii (Retz.) S. T. Blake, in Proc. Roy. Soc. Queensl. LXXX: 62 (1969).-Clayton,
in Hepper (ed.), Fl. West Trop. Afr. ed.2(3): 470 (1972)-Clayton, in Polhill (ed.), Fl. Trop. East Aft.,
Gramineae (3): 719 (1982).-Lowe, The flora of Nigeria, Grasses: 225 (1989).- Gibbs Russell et a!,
Grasses of Southern Africa: 62(1991). Type: China, Bladh (LD, K).
Andropogon bladhii Retz., Obs. Bot. 2: 27 (1781).
Andropogon intermedius R. Br., Prdr.: 202 (1810). Hack., in DC., Monogr. Phan. 6: 485
(1889).
Andropogon punctatus Roxb., Hort. Beng.: 7 (1814) and Fl. Ind. 1: 268 (1820).
Andropogon haenkei J. S. Presl, in C. B. Presi, Reliq. Haenk. 1: 340 (1830).
Rhaphis stricta Nees, in Hook. J. Bot. Kew. Misc. 2: 99(1850).
Andropogon inundatus F. Muell., Linnaea 25: 444 (1853).
Andropogon leptanthus Steud., Syn. P1. Glum. 1: 391 (1854).
Andropogonperfosus Nees & Meyen ex Steud., Syn P1. Glum. 1: 391 (1854).
Bothriochloa anamitica 0. Kuntze, Rev. Gen. P1. 2: 762 (1891).
Sorghum intermedium (R. Br.) 0. Kuntze, Rev. Gen. P1. 2: 792 (1891), based on Andropogon
intermedius.
Sorghum intermedium (R. Br.) 0. Kuntze var. haenkei (Presl) 0. Kuntze, Rev. Gen. P1. 2: 792
(1891), based on Andropogon haenkei.
Amphilophis intermedia (R. Br.) Stapf, in Agric. News W. Indies 15: 179 (1916), based on
Andropogon intermedius.
Amphilophis glabra (Roxb.) Stapf var. haenkei (PresL) E. G. & A. Camus, in Lecomte, Fl..
Gen. Indo-Chine 7: 311(1922), based on Andropogon haenkei.
Amphilophis haenkei (Presl) Haines, Bot. Bihar & Orissa: 1029 (1924), based on Andropogon
haenkei.
Bothriochloa intermedia (R. Br.) A. Camus, Ann. Soc. Linn. Lyon n.s. 76: 164 (1931); S. T.
Blake, in Univ. Queensl. Papers, Dept. Biol. II No.3: 37 (1944), based on Andropogon
intermedius.
Bothriochloa glabra (Roxb.) A. Camus, in Ann. Soc. Linn. Lyon n.s. 76: 164 (1931); Bor,
Grasses Burma Ceylon India & Pakistan: 107 (1960).
Bothriochloa inundata (F. Muell.) J.M. Black, Trans. & Proc. Roy. Soc. S. Aust. 60: 163
(1936), based on Andropogon inundatus.
Bothriochloa caucasica (Trin.) Hubbard, in Kew Bull. 1939: 101 (1939); Bor, Grasses Burma
Ceylon India Pakistan: 106 (1960)
Bothriochloa glabra (Roxb.) A. Camus subsp. haenkei (Presl) Henr., Blumea 3: 456 (1940),
based on Andropogon haenkei.
Bothriochloa haenkei (Presl) Ohwi, Acta Phytotax. & Geobot. 11: 168 (1942), based on
Andropogon haenkei.
332
Dichanthium bladhii (Retz.) Clayton, in Kew Bull. 32: 3 (1977).
A variable and tufted perennial. Cuims rather straggling, up to 1 in. tall, cuim nodes glabrous, rarely
bearded; culm internodes terete or channelled on one side. Leaf blades to 30 cm. long and 10 mm.
wide, tapering gradually to a slender tip, almost glabrous; leaf sheaths bearded at the nodes; ligule a
ring of white hairs.
Inflorescence a panicle, with numerous slender branches 2-5 cm. long, along a common axis 3-10 cm.
long; lower branches sometimes rebranched; rachis internodes and pedicels with a translucent
longitudinal groove, hairy, hairs up to 2.5 mm. long. Spikelets usually dark purple or strongly flushed
with purple; sessile spikelet 3-4 mm. long, with a ring of short white hairs at the base; lower glume
hairy below the middle on the back, rarely glabrous, rigidly ciliate on the margins in the upper part,
depressed along the middle, with a deep or shallow pit in the upper part, the pit sometimes not
occurring in all spikelets, rarely absent altogether; upper glume lanceolate, margins sparsely ciliate in
the upper part or glabrous, depressed longitudinally on both sides of the median keel, keel scabrid in
the upper part. Florets 2: lower floret empty, lemma oblong, obtuse, glabrous; upper floret bisexual,
lemma stipiform; awn 10-18 mm. long, geniculate, weakly twisted, glabrous; epaleate; ovary oblong;
lodicules 2, tiny, fleshy, glabrous; anthers 3, 1.5 mm. long; ovary oblong. Pedicellate spikelet usually
smaller than the sessile and reduced to 1 or 2 glumes, rarely well developed and with the lower glume
pitted.
Botswana. N: Tubu Island, east of Gumare, 15.i.1983, Astle 8228 (SRGH)); Tati river, south-east of
Francistown, altitude 980 in., Long 12246 (E). SE: Mahalapye, Madibeng, altitude 3200 ft.,
11.xii.1959, De Beer 870 (BM). Zambia. C: Lusaka District, Mount Makulu, 30.i.1973, Delmotte 949
(SRGH). E: Ngomi area, Fort Jameson, iii.1962, Verboom 588 (WAG). 5: Namwala District, Kafue
river at Kialala, Kafue National Park, 08.xii.1963, Mitchell 24/29 (BM; SRGH). Zimbabwe. N:
Lomagundi District, Saroi-Sinaia road, altitude 3500 ft., ii.1972, Davies 3180 (SRGH); Alaska,
General Office, altitude 3985 ft., 16.ii.1966, Jacobsen 2827 (PRE); Sipolilo, Nyamunyeche Estate,
Gwenzi dam, 20.iii.1979, Nyariri 772 (PRE). W: Bubi District, Gwampa vlei, altitude 3000 ft., i.1956,
Goldsmith 45/56 (PRE); Wankie District, Matetsi Safari Area LK 82096, altitude 3200 ft., 14.iii.1978,
Gonde 131 (PRE). Malawi. C: Kasungu Game Reserve, altitude 3300 ft., 18.ii.1970, Hall-Martin 563
(PRE); Lilongwe District, Diampwe river, 18.iv.1952, Jackson 764 (MAL). S: Blantyre District,
Chichiri Campus, University of Malawi, Limbe, 04.vii. 1971, Banda 1160 (MAL); Ruo-Tuchila plain,
Mulanje, altitude 2000 ft., 22.v.1934, Lawrence 143 (MAL). Mozambique. 1: Chitengo, 13 km. to
Acampamento, 02.v.1978, Diniz 164 (WAG). M: Sul do Save, liha Mariana, 07.iv.1954, Myre &
Cavaiho 1776 (PRE).
Africa, India to Australia, and the Pacific; introduced to United States. Streamsides, swamp margins
and cracking clays.
Note: B. bladhii is often confused with Capillipedium species because of the paniculate nature of the
inflorescence and the somewhat capillaceous inflorescence branches. It has 10-25 jointed panicle
branches compared to 1-10 jointed ones in Capillipedium species. Work on genetic barriers of B.
bladhii (as B. intermedia) has established that this species is capable of crossing with other species of
Bothriochloa and species of the related genera Dichanthium and Capillipedium (De Wet & Harlan,
1966; Faruqi, 1969). This led De Wet & Harlan (1966) to unite the three genera taxonomically using
the biological species concept, and calling the species a compilospecies.
3. Bothriochloa insculpta (Hochst.) A. Camus, in Ann. Soc. Linn. Lyon 1930 n.s. LXXVI: 165
(1931).-Sturgeon, A revised list of grasses of Southern Rhodesia Part IV: 9 (1959).- Chippindall, in
Meredith (ed.), The grasses and pastures of South Africa: 483 (1955).-Bogdan, A revised list of Kenya
grasses: 56 (1958).-Jackson & Wiehe, An annotated check-list of Nyasaland grasses: 31 (1958).-Bor,
Grasses of Burma Ceylon India & Pakistan: 107 (1960).-Napper, Grasses of Tanganyika: 100 (1965).-
333
Hood, A guide to the grasses of Zambia: 57 (1967).-Clayton, in Polhill (ed.), F!. Trop. East Afr.,
Gramineae (3): 720 (1982).-Chaudhary, Grasses of Saudi Arabia: 419 (1989).-Gibbs Russell et a!,
Grasses of Southern Africa: 63 (1991). Type: Ethiopia, Mt. Sholoda (Selleuda), Schimper 80 (K).
Andropogon inscuiptus Hochst. ex A. Rich., Tent. Fl. Abyss. II: 458 (1851).
Andropogon pertusus (L.) Wilid. var. capensis Hackel, in DC., Monogr. Phan 6: 482 (1889).
Andropogon pertusus (L.) Wilid. var. inscuiptus subvar. trfoveolatus Hackel, in DC.,
Monogr. Phan. 6: 482 (1889).
Andropogon pertusus (L.) Wilid. var. capensis Durand & Schinz, Consp. Fl. Afr. V: 718
(1895).
Andropogon pertusus (L.) Wilid. var. insculpths Durand & Schinz, Consp. F!. Afr. V: 718
(1895).
Andropogonpertusus (L.) WilId. var. capensis Stapf, in Dyer, Fl. Cap. VII: 345 (1897).
Amphilophis inscuipta (Hochst.) Stapf, in Pram, F!. Trop. Afr. IX: 176 (1917).
Tufted perennial. Cuims up to 2 m. tall, either decumbent and rambling or developing into stout
woody stolons, branching and rooting at the prostrate nodes; cuim nodes bearded with a ring of white
hairs; internodes channelled on one side. Leaf blades 4-30 cm. long, 2-8 mm. wide, glabrous on both
surfaces, occasionally with sparse tubercle-based hairs on the lower surface, margins rough; leaf
sheaths glabrous, striate, clasping the internodes tightly, bearded at the mouth; ligule an unfringed
membrane, acute.
Inflorescence of 3-20 subdigitate pedunculate racemes; main inflorescence axis up to 3 cm. long,
bearded at the joints; racemes 10-20-jointed; raceme peduncles glabrous; raceme length variable but
up to 10 cm. long; lower racemes longer than the main inflorescence axis; rachis internodes and
pedicels with a longitudinal translucent groove, hairy on the margins, hairs increasing upwards to 3
mm. long. Sessile spikelet narrowly elliptic, 4.5 mm. long; lower glume with a single deep pit on the
back just above the middle, glabrous, shortly ciliate on the lateral keels towards the apex, 4 mm. long;
callus bearded, obtuse, 0.5 mm. long; upper glume sharply carinate, minutely ciliate on this keel
towards the apex, glabrous, narrowly lanceolate, 4.5 mm. long. Florets 2: lower floret empty, lemma
3.5 mm. long, broadly ovate; upper floret bisexual, lemma stipifonn, tapers downwards, 2 mm. long,
awned from the tip; awn 25 mm. long, glabrous; epaleate; lodicules 2, fleshy, cuneate; anthers 3, 1.8
mm. long; ovary ovate. Pedicellate spikelet neuter, glabrous; lower glume with 0-4 slit-like pits
(shallow), glabrous, oblong, 5 mm. long, with an irregular apex, sharply ciliate along the lateral keels;
upper glume lanceolate, 4.5 mm. long, flimsy with ciliate margins; lower lemma oblong, truncate, 3
mm. long; pedicel hairy, with a longitudinal translucent groove.
Botswana. N: Tubu Island, east of Gumare, 15.i.1983, Astle 8228 (SRGH); Ngamiland District,
altitude 910 m., 31.iii.1987, Long & Rae 510 (E); Savuti-Selinda road, 26.ii.1983, Smith 4141
(SRGH). SE: South-east District, altitude 1000 m., 04.iii.1987, Long & Rae 34 (B). Zambia. C:
Munibwa District, Lubungu Pontoon, Kafue river, Kafue National Park, 07.iv. 1963, Mataundi 19/80
(SRGH). Zimbabwe. N: Gokwe District, 3 miles north of Gokwe, 30.v.1963, Bingham 674 (SRGH).
W: Balulima-Mangwe District, Dombodema Mission Farm, Tjompani Store, altitude 1300 m.,
29.iv.1972, Norrgrann 145 (WAG). E: Lower Sabi, altitude 1500 ft., 27.i - 2.ii.1948, Rattray 1262
(WAG). Malawi. C: Ncheu District, South Ncheu (Ntcheu), 20.iv.1956, Jackson 1851 (MAL).
Mozambique. GI: Gaza District, Chibuto, Maniquenique, Estaço Experimental do C.I.C.A.,
13.vi.1950, Lemos & Balsinhas 100 (PRE, WAG); S'uldo S'ave, Licilo, 10.vii.1947, Pedro &
Pedrogao 1404 (SRGH).
Tropical Africa, Arabia and India; introduced to Australia. Overgrazed grasslands and weedy places.
334
Cultivated fodder; native pasture species.
Note: This species may be confused with the east African B. pertusa. It differs in having the lower
glume of the pedicellate spikelet always with 1-4 pits and that of the sessile spikelet always glabrous
on the back.
8. Euclasta Franch., in Bull. Soc. Hist. Nat. Autun 8: 335 (1895). Type species: Euclasta
condylotricha (Hochst.) Stapf.
Indochloa Bor, in Kew Bull. 9: 75 (1954).
Inflorescence terminal and axillary, of delicate solitary or subdigitate pedunculate racemes; racemes
consist of spikelet pairs, one sessile, the other pedicellate; 1-3 large homogamous spikelet pairs at the
base of the raceme; pedicellate spikelet larger than the sessile; rachis internodes and pedicels with a
translucent longitudinal groove; lower glume papery, more or less flat; lower floret empty with a
hyaline lemma; upper floret bisexual, lemma entire, stipiform; pedicellate spikelet male or neuter.
Species 3.
1. Euclasta condylotricha (Hochst.) Stapf, in Pram, F!. Trop. Aft. IX: 181 (1917).- Jackson & Weihe,
An annotated check-list of Nyasaland grasses: 41 (1958).-Napper, Grasses of Tanganyika: 100
(1965).-Hood, A guide to the grasses of Zambia: 65 (1967).-Clayton, in Hepper (ed.), Flora of West
Tropical Africa, ed.2 (3): 471 (1972).-Clayton, in Polhill (ed.), Flora of Tropical East Africa,
Gramineae(3): 722 (1982).-Lowe, Flora of Nigeria, Grasses: 237 (1989). Type: Ethiopia,
Dscheladscheranne, Schimper 2011 (BM!).
Andropogon condylotrichus Hochst., in Steud., Syn. P1. Glum. 1: 377(1854).
Andropogonpiptatherus Hack., in Mart. et Eichl., Fl. Bras. 2: 3, 293 (1883).
Sorghum piptatherum (Hack.) 0. Ktze., Rev. Gen. P1. 2: 792 (1891).
Euclasta glumacea Franch., in Bull. Soc. list. Nat. Autun 8: 336 (1895).
Euclasta graminea Th. et Hel. Durand, Syll. Fl. Congo!.: 649 (1909).
Amphilophispiptatherus (Hack.) Nash, in North Amer. Fl. 17: 127 (1912).
Annual. Cuims up to 1.5 m. tall, ascending but geniculate at the base, always bearded at the nodes.
Plant shallowly rooted and rooting at the nodes (stilt rooted). Leaf blades linear-lanceolate, somewhat
flaccid, up to 25 cm. long, 2-10 mm. wide, contracted at the base and drawn to a setaceous tapering
point, glabrous on both surfaces or with very short sharp hairs along the midrib and immediate
adjacent veins; midrib slender, whitish and prominent; leaf sheaths striate, glabrous, closely clasping
the internodes for about half their length, with long tubercle-based hairs at the mouth, in the collar
region; ligule ciliolate.
Inflorescence of subdigitate racemes; rachis internodes and pedicels with a translucent longitudinal
groove; lowest 1-3 spikelet pairs of each raceme homogamous, male or neuter. Sessile spikelet
bisexual; lower glume pale to whitish, elliptic-oblong, minutely truncate, 4 mm. long, with 5 nerves
ending below the hyaline apex, hairy on the margins and lower half, 2-keeled, the keels scaberulous;
upper glume ovate, strongly carinate, opaque, glabrous, 4 mm. long. Florets 2: lower floret empty,
lemma ovate, carinate and ciliate along this keel and the margins, glabrous; upper floret bisexual,
lemma hyaline, stipiform; palea present, broadly ovate, hyaline: awn 37 mm. long, minutely hairy;
lodicules 2, fleshy, glabrous; anthers 3, 0.75 mm. long; ovary ovate, 2 mm. long. Pedicellate spikelet
male; lower glume oblanceolate, many nerved (7-10 nerved), hairy on the lower 1/3, 5.5 mm. long;
335
upper glume 4-nerved, 4-4.5 mm. long. Upper floret male, lemma slightly shorter than the glume;
anthers 3, linear, 1.5 mm. long; lodicules 2, fleshy; pedicel grooved, hairy on the sides.
Zambia. E: Fort Jameson, Jumbe area, 24.iii.1963, Verboom 941 (BM). Zimbabwe. C: Gatooma
District, Sanyati Reserve, 12.iii.1959, Cleghorn 451 (BM). S: Gokwe District, 10 miles from Gokwe,
near top of basalt escarpment in Bonye Gorge, 25.iii.1963, Bingham 571 (BM).
Tropics of Africa. Occurs mainly on roadsides and shallow rock escarpments.
9. Dichanthium Willemet, in Usteri, Ann. Bot. 18: 11 (1796).-De Wet & Harlan, in Bol. Soc. Arg.
Bot. 12: 206-227 (1968). Type species: Dichanthium annulatum (Forssk.) Stapf.
Lepeocercis Trin., Fund. Agrost.: 203 (1822).
Diplasanthum Desv., Opusc.: 66 (1831).
Eremopogon Stapf, in Pram, Fl. Trop. Aft. IX: 182 (1917).
Annuals or perennials. Inflorescence of single, digitate or subdigitate racemes; racemes rarely solitary
but more often in pairs-to-many, arranged on a short primary axis; pedunculate or subsessile,
sometimes with the lower branches divided; racemes always terminate in a triad of 1 sessile and 2
pedicellate spikelets; spikelets highly imbricate, paired, one sessile, the other pedicellate, subequal to
equal in size and shape, usually differing in sex, except the lower 1-6 pairs, which are with rare
exceptions. homogamous male or neuter; rachis intemodes and pedicels solid. Sessile spikelet callus
obtuse; lower glume papery to cartilaginous, broadly convex to slightly concave, sometimes pitted,
acute to broadly obtuse; upper lemma hyaline base of the awn, entire; awn glabrous. Pedicellate
spikelet much like the sessile, male or neuter, awnless, rarely bisexual and awned (e.g. D. annulatum
var. fecundum).
Species ± 20.
Inflorescence of 2-many digitate or subdigitate racemes:
Cuim nodes bearded; rachis intemodes hairy; lower glume of sessile spikelet concave,
pubescent in the lower third, with long tubercle-based cilia above the middle, along the
1. annulatum
margins and forming a subapical fringe.
CuIm nodes glabrous; rachis intemodes glabrous; lower glume of sessile spikelet narrowly
.
2. aristatum
winged on the lateral keels, without tubercle-based cilia.
Inflorescence digitate with 1-15 sessile to subsessile white villous racemes.
3. sericeum
1. Dichanthium annulatum (Forssk.) Stapf, in Prain, Fl. Trop. Afr. IX: 178 (1917).-Bor, Flora of
Assam 5: 371 (1940).-Sturgeon, A revised list of grasses of Southern Rhodesia Part IV: 10 (1954).Bogdan, A revised list of Kenya grasses: 57 (1958).-Bor, Grasses Burma Ceylon India & Pakistan:
133 (1960).-Hood, A guide to the grasses of Zambia: 65 (1967).-Clayton, in Hepper (ed.), Fl. West
Trop. Afr. (ed.2) 3: 471 (1972).-Clayton, in Polhill (ed.), Fl. Trop. East Mr., Gramineae (3): 725
(1982).-Chaudhary, Grasses of Saudi Arabia: 417 (1989)-Cope & Hosni, A key to Egyptian grasses:
54 (1989).-Gibbs Russell et al, Grasses of Southern Africa: 105 (1991). Type: Egypt, Forsskâl (C).
Andropogon annulatus Forssk., Fl. Aegypt.- Arab.: 173 (1775).
Andropogon bladhii Retz., Obs. Bot. 2: 27 (1781).
Andropogon scandens Roxb., Hort. Beng.: 7 (1814).
Andropogon obtusus Nees, in Hook. et Am., Bot. Beech Voy.: 243 (1838).
336
Lepiocercis annulata (Forssk.) Nees, Fl. Afr. Aust.: 98 (1841).
Andropogon papillosus Hochst., in Rich., Tent. Fl. Abyss. 2: 457 (1851).
Andropogon annulatus var. monostachya F. Well. ex Benth., Fl. Austr. 7: 531 (1878).
Andropogon annulatus var. bladhii (Retz.) Hack., in DC., Monogr. Phan. 6: 572 (1889).
Andropogon annulatus var. decalvatus Hack., in DC., Monogr. Phan. 6: 572 (1889).
Andropogon grandispiculatus Domin, Biblioth.. Bot. 85: 269 (1915). Type: Australia,
Queensland, Flinders river, ii. 1910, Domin s. n. (K!).
Dichanthium papillosum (Hochst.) Stapf, in Pram, Fl. Trop. Afr. IX: 179 (1917).
Dichanthiumfecundum S. T. Blake, in Univ. Queens!. Papers, Dept. Biol. II No.3: 51(1944).
Type: Australia, Queensland, North Kennedy District, Ayr, altitude 20ft, Blake 8337 (K!).
Dichanthium annulatum var. decalvatum (Hack.) Maire et Weiller, Fl. de LAfr. du Nord 1:
282 (1952).
This species has three varieties keyed out as follows:
Some pedicellate spikelets on each raceme bisexual and often shortly awned.
var. fecundum
Pedicellate spikelets always male or neuter.
Spikelets pilose, with a distinct subapical fringe of long tubercle-based cilia.
var. papillosum
Spikelets less pilose, with the tubercle-based cilia mainly along the margins of the lower
var. annulatum
glume.
Var. papillosum is the only one that occurs in FZ.
A tufted, decumbent or erect perennial of up to 2 m. tall with distinct short rhizomes. Cuims woody
and persistent, robust, geniculate ascending, simple or branched; cuim nodes exposed and with a ring
of hairs; culm intemodes glabrous. Leaf blades linear-lanceolate, 4-4.5 mm. wide and up to 30 cm.
long, glabrous along most of their length with long hairs in the ligule area; leaf sheaths glabrous, or
sparsely hairy all over; ligule an unfringed membrane, about 2 mm. long.
Inflorescence consists of subdigitate 2-to-many racemes arranged on a short primary axis; rachis
internodes and pedicels solid, usually hairy. Sessile spikelets imbricate, oblanceolate, 2-6 mm. long,
the lowest 1-6 usually homogamous, male or neuter, the upper bisexual; lower glume slightly concave,
pubescent below the middle, with long tubercle-based cilia forming a subapical arch; upper glume
slightly shorter than the lower, glabrous except for the keel and apex which are often shortly ciliate.
Florets 2: lower floret empty, lemma ovate, 3.8 mm. long, hyaline; upper floret bisexual, lemma
stipiform, about 3 mm. long; awn 25 mm. long, both column and bristle minutely hispidulous;
lodicules 2, fleshy, cuneate, highly vascularised; anthers 3, 1.5 mm. long; caryopsis elliptic-oblong.
Pedicellate spikelets equal or subequal in size and shape to the sessile ones, male or neuter.
var. papillosum:
Botswana. N: North-east District, altitude 1000 m., 19.iv.1987, Long & Rae 802 (K); North-west
District, 28.i.1980, Smith 3498 (K). Zambia. S: Livingstone Game Park, 14.ii.1961, Mitchell 5/78 (K).
Zimbabwe. N: Gokwe District, Sengwa Research Station, 101.1975, Guy 2256 (K). 5: Belingwe,
altitude 3000 ft., 01.xi.1973, Gosden 10 (K).
337
Occurs in the tropics and subtropics of Africa. Dry open places subject to overgrazing or disturbance.
Also in swampy depressions.
Dichanthium aristatum (Poir.) C. E: Hubbard, in Kew Bull. 1939: 654 (1939).-Sturgeon, A
revised list of grasses of Southern Rhodesia Part IV: 10 (1954).-Chippindall, in Meredith (ed.), The
grasses and pastures of South Africa: 481 (1955).-Bor, Grasses Burma Ceylon India & Pakistan: 134
(1960).-Clayton, in Polhill (ed.), F!. Trop. East. Aft., Gramineae (3): 723 (1982)-Gibbs Russell et at,
Grasses of Southern Africa: 105 (1991). Type: Mauritius, Commerson (whereabouts uncertain).
Dichanthium nodosum Willem., in Usteri, Ann. Bot. 18: 11(1796).
Andropogon aristatus Poir, In Lamk., Encycl. Met. Bot. Suppl. 1: 585 (1810).
Andropogon mollicomus Kunth, Rev. Gram. 1: 365 (1830).
Diplasanthum lanosum Desv., Opusc.: 67 (1831).
Lepiocercis mollicoma (Kunth.) Nees, Edinb. New. Philom. Journ. 18: 185 (1835).
Andropogon caricosus subsp. mollicomus Hack. var. mollicomus (Kunth.) Hack,, in DC.,
Monogr. Phan. 6: 569 (1889).
Andropogon nodosus (Wilem.) Nash, N. Amer. Fl. 17: 122 (1912).
Dichanthium caricosum var. mollicomus (Hack.) Haines, Bot. Bihar & Orissa 5: 1039 (1924).
A tufted perennial of up to 1.1 m. tall, shortly stoloniferous. Culms robust, erect, sometimes
decumbent; culm nodes exposed and glabrous. Leaf blades linear-lanceolate, up to 25 cm. long,
glabrous except at the base in the ligule area where there are long tubercle-based cilia; leaf sheaths
glabrous; ligule a scarious membrane.
Inflorescence of 1-5 racemes arranged subdigitately on a short primary axis; raceme peduncles and
culm below the inflorescence pubescent; rachis internodes and pedicels solid; rachis joints hairy;
rachis intemodes glabrous and pedicels hairy on one side (glabrous on the side facing the sessile
spikelet); lowest 1-6 spikelet pairs homogamous, male or neuter. Sessile spikelet obovate, 2-5 mm.
long; lower glume 5 mm. long, pilose below the middle, glabrous or shortly ciliate above the middle
along the margins and near apex, narrowly winged towards the apex; upper glume as long as the
lower, glabrous or ciliate along the margins and keel; callus short, blunt. Florets 2: lower floret empty,
lemma hyaline; upper floret bisexual, lemma hyaline, stipiform; awn 22 mm. long, minutely
pubescent; lodicules 2, fleshy, glabrous; anthers 3, 2 mm. long; ovary ovate. Pedicellate spikelet male
or neuter, subequal in size and shape to the sessile one.
Zambia. N: Mpika District, Luangwa Game Reserve, 27.iv.1965, Mitchell 2689 (K). C: Luangwa
Valley, near the Lunda Plain, altitude 2500 ft., 04.v.1966, Astle 4865 (K). Zimbabwe. C: Harare
District, Harare, university campus on southern slope, 14.iii. 1985, Bennett s.n. (K).
India, introduced to Australia, Africa and America. Damp places in disturbed grounds.
Fodder grass.
Note: Characterised by robust, suberect plants having the peduncles to the racemes villous, and the
culm below the inflorescence strongly pilose. The species is related to and sometimes regarded as
conspecific with D. caricosum.
Dichanthium sericeum (R. Br.) A. Camus, in Bull. Mus. Hist. Nat. Paris XXVII: 549 (1921).Burbidge, Australian grasses 1: 146 (1966).-Burbidge, Australian grasses: 146 (1968).-Lazarides, The
338
grasses of Central Australia: 70 (1970).-Wheeler, Jacobs & Norton, Grasses of New South Wales: 163
(1982).-Simon, A key to Australian grasses (ed.2): 98 (1993).-Simon & Latz, A key to the grasses of
the Northern Territory, Australia: 26 (1994). Type locality, Australia.
Andropogon sericeus R. Br., Prodr. Fl. Nov. Holl. 1: 201 (1810).
Andropogon affine R. Br., Prodr. Fl. Nov. Ho!!. 1: 201 (1810).
Andropogon chrysantherus F. Muell., Linnnaea 25: 443 (1852).
Andropogonjubatus Balansa, Bull. Soc. Bot. Fr. 19: 322 (1872).
Andropogon acutiusculus Hack., in DC., Monogr. Phan. 6: 575 (1889).
Andropogon sericeus var. mollis F. M. Bailey, Queensl. Agric. Journ. 30: 316 (1913).
Dichanthium acutiusculum (Hack.) A. Camus, Bull. Mus. Hist. Nat. 27: 549 (1921).
A perennial of about 1 m. tall, shortly rhizomatous, erect or slightly decumbent; culm nodes exposed
and bearded. Leaf blades linear-lanceolate, up to 15 cm. long, glabrous to pilose all over; leaf sheaths
glabrous to pilose; ligule an unfringed membrane of up to 1.5 mm. long.
Inflorescence of 1-15 sessile to subsessile white villous digitate racemes; lowest 1-6 spikelet pairs
homogamous, male or neuter. Sessile spikelet elliptic-oblong, truncate at the apex; lower glume
obtuse, pilose or rarely glabrous on the back below the middle, always with tubercle-based cilia (up to
6 mm. long) along the margins and forming a transverse subapical fringe; upper glume slightly longer
than the lower, glabrous, ovate, carinate. Florets 2: lower floret empty, lemma hyaline; upper floret
bisexual, lemma 2 mm. long, stipiform; awn 30 mm. long; epaleate; lodicules 2, fleshy, glabrous;
anthers 2 or 3, 2 mm. long; ovary obovate. Pedicellate spikelet obovate-oblong, 4 mm. long, male or
neuter; lower glume obovate, linear to narrowly ovate, with a transverse subapical fringe of cilia;
anthers 2, 2 mm. long; pedicel hairy.
Zimbabwe. C: Harare (Salisbury) District, cultivated, altitude 4800 ft., xii.1919, Eyles 1971 (K).
Widely distributed in Australia, and introduced to tropical East Africa. Cultivated
Note: Racemes are densely covered with long, white silky hairs, sessile or very shortly pedunculate;
lower glume of the sessile spikelet covered with long, silvery hairs with an arch of extremely long
hairs at apex, many of which arise from tubercles.
339
Appendix 11.2. A mauscript sent to Kew Bulletin for
publication on the new combination recognised in the genus
Euclasta Franch.
A NEW COMBINATION OF EUCLASTA (POACEAE) FROM INDIA.
M.P. SETSHOGO*
Summary. A new combination in the genus Euclasta Franch., Euclasta oliganthus (Hochst. ex Steud.)
M.P. Setshogo, is made.
Euclasta oliganthus (Hochst. ex Steud.) M.P Setshogo comb. nov. Type: India, Nilgri Hills,
Hohenacker 1288 (K! BM!).
Synonyms: Andropogon oliganthus Hochst. ex Steud, Syn. P1. Glum 1: 368 (1854). Heteropogon
oliganthus (Hochst. ex Steud.) Blatter & McCann, in J. Bombay Nat. Hist. Soc. 32: 623 (1928);
Indochloa oligantha (Hochst. ex Steud.) Bor, in Kew Bull. 9: 79, t. 2 (1954); Dichanthium oliganthum
(Hochst. ex Steud.) T.A. Cope, in Kew Bull. 35(3): 703 (1980).
Distribution: Southern India.
The genus, Euclasta, was described by Bor (1954) as a new genus Indochloa Bor. This description
was based on Dichanthium clarkei (Hack.) Haines. This genus included the species, D. clarkei and
Heteropogon oliganthus (Hochst. ex Steud.) Blatt. et McCann. Bor thought that these species had been
misplaced in the genera Dichanthium and Heteropogon, for the similarities between the two far
outweighed the differences. The name Indochloa Bor, however, was rejected because the genus had
already been described as Euclasta by Franchet in 1895, who based his description on E. glumacea
(=E. condylotricha). In conforming with this new generic circumscription, Cope (1980) transferred the
*
Institute of Cell and Molecular Biology, Department of Botany, University of Edinburgh,
Daniel Rutherford Building, Kings Buildings, Mayfield Road, Edinburgh. EH9 3JH.
340
species Indochloa clarkei (Hack.) Bor to Euclasta. He then transferred Indochloa oligantha (Hochst.
ex Steud.) Bor to the genus Dichanthium.
As a result of a thorough revision of the genus Euclasta and other related Sorghinae genera
(including Dichanthium), I propose a new combination, Euclasta oliganthus (Hochst. ex Steud.) M.P.
Setshogo. This species has many features in common with E. clarkei, some of which are: (1) both are
straggling annuals; (2) the structure of the raceme is the same; (3) the lower pairs of neuter spikelets
are remarkably similar both in shape and in nervation; (4) the sessile bisexual spikelets are very
similar in shape, texture and nervation; and (5) the callus is obtuse and bearded. It differs from
Dichanthium species in the following: (1) the pedicellate spikelets are spreading rather than
overlapping; (2) rachis intemodes and pedicels have a translucent longitudinal groove; and (3) lower
glume of the sessile spikelet is characteristically oblong-truncate.
It is quite clear that Dichanthium is not the genus in which this species can be placed. It
clearly belongs in the genus Euclasta.
Key to the species of Euclasta:
Inflorescence of subdigitate pedunculate racemes; homogamous spikelet pairs male or neuter.
1. condylotricha
Inflorescence a delicate solitary raceme (rarely two); homogamous spikelet pairs neuter:
2. clarkei
Raceme peduncle glabrous; sessile spikelet 3.5 mm long.
Raceme peduncle pilose; sessile spikelet 5 mm long.
3. oliganthus
Acknowledgements.
I wish to thank Dr. P. M. Smith for scrutinising the manuscript. I also wish to thank Keepers of the
herbaria BM and K for the loans of material used in the study.
References.
Bor, N. L. (1954). Notes on Asiatic Grasses: XVI. Indochloa Bor, a new genus of Indian grasses. Kew
Bull. 9: 73-81.
Cope, T. A. (1980). New combinations in Asiatic Grasses. Kew Bull. 35(3): 701-704.
341
Appendix 11.3. Delta file for phenetic analysis.
*HEADING Species of Sorghinae
*NIJI ,ffiER OF CHARACTERS
OPTIONS
*PRINT WIDTH
* DATA BUFFER SIZE
*MM(II{IJM NUMBER OF STATES
*MAXIMUM NUMBER OF ITEMS
* CHARACTER TYPES
*NUMBERS OF STATES
*CHARACTER WEIGHTS
*KEY STATES
* CHARACTER DESCRIPTIONS
#1. Longevity!
annual!
perennial!
#2. Habit!
caespitose<tufted>/
spreading <decumbent>
#3. CuIm orientation!
erect!
geniculate/
#4. Perennating organs!
rhizomes!
stolons!
other!
342
#5. CuIm branching!
branched!
sparingly branched!
unbranched!
#6. Central culm nodes!
exposed!
hidden <by leaf sheath>!
#7. CuIm node pubescence!
hairy!
glabrous!
48. CuIm node hair type!
short hairs!
pilose!
#9. CuIm internode!
terete/
channelled on one side!
#10. Leaf sheath!
auriculate!
non auriculate!
#11. Leaves!
cauline!
basal!
#12. Leaf blade shape!
linear!
linear lanceolate/
lanceolate/
acicular <narrow and tapering>!
343
#13. Leaf blade folding!
flat!
folded <conduplicate>!
rolled!
#14. Leaf blade width <range (mm)>!
1. 1-5.9/
2.6-10.9/
3. 11-20/
4>20/
#15. Ligule!
I. absent!
2. unfringed membrane!
3. fringed membrane!
#16. Fringed membrane type!
ciliate!
lacerate <shortly toothed>!
#17. Inflorescence type!
paniculate!
racemose/
918. Spathe!
present!
absent!
#19. Paniculate inflorescence!
with simple branches!
with divided branches!
#20. Panicle internode!
1. long <panicle open or loose>!
344
2. short <panicle contracted>!
#21. Main inflorescence axis!
slender!
substantial<stout>!
#22. Racemose inflorescence!
digitate!
subdigitate!
single raceme!
#23. Racemes!
solitary!
paired!
fasciculate <clustered in groups of three or more>!
#24. Racemes <manner of disarticulation>!
falling entire!
disarticulating at the joint(s)!
#25. Number of spikelet pairs (nodes) on ultimate rachis axis!
1. 1!
2.2-3/
3.4-5/
4.6-10/
5.> 10!
#26. Raceme nodes!
densely long-hairy!
glabrous!
#27. Rachis internodes!
with a translucent longitudinal groove!
without a translucent longitudinal groove!
345
#28. Rachis internodes pubescence!
hairy!
glabrous!
scabrous!
#29. Spikelets/
solitary!
consistently paired!
consistently in triplets!
sometimes in triplets and paired!
#30. Spikelet type!
sessile only/
in pedicellate/sessile combinations!
in a triad <1 sessile and 2 pedicellate>/
sessile and accompanied by (a) barren pedicel(s)!
#31. Spikelet disposition!
imbricate!
distant <not overlapping>!
932. Sessile spikelet <in the heterogamous pairs>!
bisexual!
female-only!
male-only!
neuter!
#33. Pedicellate spikelet <in the heterogamous pairs>!
absent!
bisexual!
female-only!
male-only!
346
5. neuter!
#34. Lowest sessile spikelet <in the raceme>/
homogamous<male or neuter>!
heterogamous<bisexual>!
#35. Pedicel <development>!
well developed!
rudimentary!
absent!
#36. Pedicel pubescence!
hairy!
glabrous!
#37. Sessile spikelet!
compressed laterally!
not noticeably compressed!
compressed dorsiventrally!
#38. Callus!
absent!
long!
short!
#39. Callus point shape!
pungent!
blunt!
#40. Callus pubescence!
bearded!
not bearded!
#41. Lower glume<of sessile spikelet>!
1. two keeled!
347
2. not two keeled!
#42. Shape of the back of the lower glume/
convex!
concave!
flattened!
#43. Lower glume awns/
present!
absent!
#44. Lower glume wings!
present!
absent!
#45. Lower glume pitting/
with a conspicuous pit!
without a conspicuous pit!
#46. Lower glume!
hairy!
glabrous!
#47. Lower glume indumentum!
hairy all over/
hairy below the middle!
with a subapical fringe of hair!
with submarginal hairs!
#48. Lower glume back!
with aculeolate scales!
without aculeolate scales!
#49. Lower glume keels!
1. ciliate!
348
2. not ciliate!
#50. Upper glume/
awned!
awnless/
mucronate!
#51. Upper lemma <of upper floret>/
awned!
awnless!
mucronate!
#52. Upper lemma margins!
ciliate!
glabrous!
#53. Upper lemma width relative to awn base width!
wider than the awn base!
narrower than the awn base!
#54. Upper lemma apex!
entire!
bibbed!
#55. Upper lemma awn!
present!
absent!
#56. Upper lemma awn orientation!
geniculate!
not geniculate!
#57. Awn twisted!
yes!
no/
349
#58. Upper lemma awn pubescence!
hairless<glabrous or scabrous>!
hairy!
#59. Awn hairs/
long plumose!
not long plumose!
#60. Upper lemma awn length in relation to the upper lemma length!
much shorten
about as long!
much longer!
#61. Falea <of upper floret>I
absent!
relatively long!
conspicuous but relatively short!
#62. Palea margin and apex!
entire!
ciliate!
#63. Lodicules!
absent!
fleshy!
membranous!
#64. Lodicules margins and apex!
ciliate!
glabrous!
*DEPENDENT CHARACTERS
350
1,1:4 6,2:7-8 7,2:8 15,1:16 15,2:16 17,1:22 17,2:19-2125,1:27-28,34 35,3:36 41,2:49 46,2:47 51,2:53
55,2:56-60 58,1:59 61,1:62 63,1:64
*ITEM DESCRIPTIONS
Asthenochloa ten era!
1,2 2,2 3,1 4,3 5,2 6,1 7,2 9,1 10,2 11,1 12,3 13,1 14,2 15,3 16,1 17,1 18,2 19,2 20,2 21,1 23,1 24,2
25,1 26,1 29,2 30,4 31,2 32,1 33,1 35,2 36,1 37,3 38,3 39,2 40,1 41,1 42,3 43,2 44,2 45,2 46,1 47,1
48,2 49,2 50,3 51,1 52,1 53,1 54,2 55,1 56,1 57,1 58,1 60,3 61,1 63,1
Cleislachne sorghoides/
1,1 2,2 3,1 5,3 6,1 7,2 9,1 10,2 11,1 12,2 13,1 14,U 15,2 17,1 18,2 19,2 20,2 21,1 23,1 24,1 25,1 26,1
29,1 30,1 31,2 32,1 33,1 34,2 35,3 37,3 38,3 39,2 40,1 41,2 42,1 43,2 44,2 45,2 46,1 47,1 48,2 50,2
51,1 52,1 53,1 54,2 55,1 56,1 57,1 58,2 59,2 60,3 61,3 62,2 63,2 64,1
Euclasta clarkeil
1,1 2,2 3,2 5,1 6,1 7,1 8,1 9,1 10,2 11,1 12,2 13,1 14,2 15,3 16,1 17,2 18,2 22,3 23,1 24,2 25,4 26,1
27,1 28,1 29,2 30,2 31,2 32,1 33,5 34,1 35,1 36,1 37,3 38,3 39,2 40,141,142,2/3 43,2 44,1 45,2 46,1
47,1 48,2 49,2 50,2 51,1 52,2 53,2 54,1 55,1 56,1 57,1 58,1 60,3 61,1 63,2 64,2
Euclasta condylotrichal
1,1 2,2 3,1/2 5,1 6,1 7,1 8,U 9,1 10,2 11,1 12,2 13,1 14,1/2 15,3 16,2 17,2 18,2 22,2 23,1 24,2 25,4
26,1 27,1 28,1 29,2 30,2 31,2 32,1 33,4 35,1 36,1 37,3 38,3 39,2 40,1 41,1 42,3 43,2 44,2 45,2 46,1
47,2/4 48,2 49,2 50,2 51,1 52,2 53,2 54,1 55,1 56,1 57,1 58,2 59,2 60,3 61,2 62,1 63,2 64,2
Euclasta oliganthal
1,1 2,2 3,1/2 5,1/2 6,1/2 7,1 8,2 9,1 10,2 11,1 12,1 13,1 14,1 15,2 17,2 18,2 22,3 23,1 24,2 25,4 26,1
27,1 28,1 29,2 30,2 31,2 32,1 33,5 34,1 35,1 36,1 37,3 38,3 39,2 40,1 41,1 42,1/3 43,2 44,2 45,2 46,1
47,2/4 48,2 49,1 50,2 51,1 52,2 53,2 54,1 55,1 56,1 57,1 58,1 60,3 61,3 62,1 63,2 64,2
Hemisorghum mekongense/
1,2 2,1 3,1 4,3 5,2/3 6,1 7,2 9,1 10,2 11,1 12,2 13,1 14,4 15,3 16,2 17,1 18,2 19,1 20,1 21,2 23,3 24,2
25,5 26,2 27,2 28,2 29,2 30,2 31,2 32,1 33,5 34,2 35,1 36,2 37,3 38,3 39,2 40,2 41,1 42,1/3 43,2 44,2
45,2 46,2 48,2 49,2 50,2 51,2 52,154,1/2 55,2 61,2 62,1 63,2 64,2
Hemisorghum venustum/
351
1,2 2,1 3,1 4,3 17,1 18,2 19,2 20,1 21,1 23,U 24,2 25,5 26,U 27,2 28,2 29,2 30,2 31,2 32,1 33,5 34,2
35,1 36,2 37,3 38,3 39,2 40,2 41,1 42,3 43,2 44,2 45,2 46,2 48,2 49,2 50,2 51,1 52,2 53,1 54,2 55,1
56,2 57,2 58,1 60,3 61,1 63,2 64,2
Pseudodichanthium serrafalcoidesi
1,1 2,2 3,2 5,1 6,1 7,2 9,1 10,2 11,1 12,1 13,1 14,1/2 15,2 17,2 18,2 22,1 23,1 24,2 25,U 26,1/2 27,2
28,1/2 29,2 30,2 31,1 32,1 33,5 34,1 35,1 36,1/2 37,3 38,3 39,2 40,2 41,1 42,1 43,2 44,1 45,2 46,2
48,2 49,2 50,2 51,1 52,2 53,2 54,1 55,1 56,1 57,1 58,2 59,2 60,3 61,1 63,2 64,2
Pseudosorghumfasciculare/
1,1 2,2 3,2 5,1 6,1 7,2 9,1 10,2 11,1 12,2 13,1 14,1/2 15,2 17,1 18,2 19,1/2 20,2 21,1 24,2 25,5 26,1
27,2 28,1 29,2 30,2 31,2 32,1 33,4/5 34,1 35,1 36,1 37,3 38,3 39,2 40,1 41,1 42,U 43,2 44,2 45,2 46,2
48,2 49,2 50,2 51,1 52,1 53,2 54,2 55,1 57,1 58,1 60,3 61,2 62,1 63,2 64,2
Spathia neurosa/
1,1 2,1 3,1 5,2/3 6,1 7,1 8,1 9,2 10,2 11,1 12,1 13,1 14,1 15,3 16,2 17,2 18,1 22,1 23,3 24,2 25,2/3
26,1 27,2 28,1 29,2 30,2 31,1 32,1 33,5 34,1 35,1 36,1 37,3 38,3 39,2 40,1 41,1 42,1 43,2 44,2 45,2
46,147,1/3 48,2 49,2 50,2 51,1 52,2 53,2 54,1 55,1 56,1 57,1 58,1 60,3 61,3 62,1 63,2 64,2
Chrysopogon aciculatusl
1,2 2,1 3,2 4,1 5,1 6,1 7,2 9,1 10,2 11,2 12,4 13,1 14,1 15,3 16,2 17,1 18,2 19,1 20,1/2 21,1 23,3 24,2
25,1 26,1 29,3 30,3 31,2 32,1 33,4 35,1 36,2 37,1.38,2 39,1 40,1 41,1 42,1 43,2 44,2 45,2 46,2 48,2
49,1 50,1 51,1 52,2 53,1 54,1 55,1 56,2 57,2 58,1 60,3 61,2 62,1 63,2 64,2
Chrysopogon aucheri/
1,2 2,1 3,1 4,3 5,3 6,1 7,1/2 8,1 9,1 10,2 11,2 12,2 13,1 14,1 15,3 16,2 17,1 18,2 19,1 20,1 21,1 23,3
24,2 25,1 26,1 29,3 30,3 31,2 32,1 33,4 35,1 36,1 37,1 38,3 39,2 40,1 41,1 42,1 43,2 44,2 45,2 46,2
48,2 49,2 50,1 51,1 52,1 53,1 54,2 55,1 56,2 57,1 58,1 60,3 61,1 63,2 64,2
Chrysopogonfallax/
1,2 2,1 3,1 4,3 5,3 6,1 7,2 9,2 10,2 11,1 12,1 13,1 14,1 15,3 16,2 17,1 18,2 19,1 20,2 21,1 23,3 24,2
25,1 26,1 29,3 30,3 31,2 32,1 33,4 35,1 36,2 37,1 38,2 39,2 40,1 41,1 42,1 43,2 44,2 45,2 46,2 48,1
49,1 50,1 51,1 52,2 53,1 54,2 55,1 56,1 57,1 58,1 60,3 61,2 62,2 63,2 64,2
Chrysopogonfulvus/
1,2 2,1 3,1 4,3 5,1 6,1 7,2 9,1 10,2 11,1 12,1 13,1 14,U 15,3 16,1 17,1 18,2 19,1 20,1 21,1 23,3 24,2
25,1 26,1 29,3 30,3 31,2 32,133,4/5 35,1 36,1 37,1 38,3 39,2 40,1 41,2 42,1 43,2 44,2 45,2 46,2 48,2
50,1 51,1 52,2 53,1 54,2 55,1 56,1 57,1 58,U 59,U 60,3 61,1 63,2 64,2
352
Chrysopogon gryllusi
1,2 2,1 3,1 4,3 5,3 6,1 7,2 9,1 10,2 11,2 12,1/2 13,1 14,2 15,3 16,2 17,1 18,2 19,1 20,1 21,1 23,3 24,2
25,1/2 26,1 27,2 28,2 29,3/4 30,2 31,2 32,133,4/5 34,2 35,136,1/2 37,1 38,3 39,2 40,1 41,1 42,1 43,2
44,2 45,2 46,2 48,2 49,1 50,1 51,1 52,1/2 53,1 54,2 55,1 56,1 57,1 58,1 60,3 61,2 62,1 63,2 64,2
Chrysopogon latfolius/
1,2 2,1 3,1 4,3 5,1 6,1 7,2 9,2 10,2 11,1 12,1 13,1/2 14,2/3/4 15,3 16,1 17,1 18,2 19,1 20,2 21,2 23,3
24,2 25,1 26,1 29,3 30,3 31,2 32,1 33,4 35,1 36,2 37,1 38,2 39,1 40,1 41,1 42,1 43,2 44,2 45,2 46,2
48,1 49,1 50,1 51,1 52,2 53,1 54,1 55,1 56,1 57,1 58,2 59,2 60,3 61,2 62,1 63,2 64,2
Chrysopogon orientalist
1,2 2,1 3,1 4,3 5,1 6,1 7,2 9,1 10,2 11,2 12,1 13,1/3 14,1 15,3 16,2 17,1 18,2 19,1 20,1 21,1 23,3 24,2
25,1 26,1 29,3 30,3 31,2 32,1 33,4 35,1 36,2 37,1 38,3 39,2 40,1 41,1 42,3 43,1 44,2 45,2 46,2 48,2
49,2 50,1 51,1 52,2 53,1 54,1 55,1 56,2 57,2 58,1 60,3 61,1 63,2 64,2
Chrysopogon paiidus/
1,2 2,1 3,1 4,2 5,2 6,1 7,2 9,2 10,2 11,1 12,U 13,2 14,2 15,3 16,1 17,1 18,2 19,1 20,1 21,1 23,3 24,2
25,1 26,1 29,3 30,3 31,2 32,1 33,4 35,1 36,2 37,1 38,2 39,1 40,1 41,1 42,1 43,1/2 44,2 45,2 46,2 48,1
49,1 50,1 51,1 52,1 53,1 54,2 55,1 56,1 57,1 58,1 60,3 61,2 62,2 63,2 64,2
Chrysopogon plumulosus/
1,2 2,1 3,1 4,3 5,3 6,1 7,2 9,1 10,2 11,1 12,2 13,1 14,1 15,3 16,2 17,1 18,2 19,1 20,1 21,1 23,3 24,2
25,1 26,1 29,3 30,3 31,2 32,1 33,4/5 35,1 36,1 37,1 38,3 39,1 40,1 41,2 42,1 43,2 44,2 45,2 46,2 48,2
50,1 51,1 52,2 53,1 54,1 55,1 56,1 57,1 58,1 60,3 61,1 63,2 64,2
Chrysopogon serrulatusi
1,2 2,1 3,1 4,3 5,3 6,1 7,2 9,1 10,2 11,1 12,U 13,1 14,1/2 15,3 16,2 17,1 18,2 19,1 20,1 21,1 23,3 24,2
25,1 26,1 29,3 30,3 31,2 32,1 33,4/5 35,1 36,1 37,1 38,3 39,1 40,1 41,2 42,1 43,2 44,2 45,2 46,2 48,2
50,1 51,1 52,2 53,1 54,2 55,1 56,1 57,1 58,1/2 59,2 60,3 61,1 63,2 64,2
Chrysopogon sylvaticus/
1,2 2,1 3,1 4,3 5,1 6,1 7,2 9,1 10,2 11,2 12,1 13,2 14,2 15,3 16,2 17,1 18,2 19,1 20,1 21,1 23,3 24,2
25,1/2 26,1 27,2 28,2 29,3/4 30,2/3 31,2 32,1 33,5 34,2 35,1 36,2 37,1 38,3 39,1 40,1 41,1 42,1 43,2
44,2 45,2 46,2 48,2 49,1 50,1 51,1 52,2 53,1 54,2 55,1 56,1 57,1 58,1 60,3 61,2 62,1 63,2 64,2
Vetiveria elongatat
353
1,2 2,1 3,1 4,3 5,3 6,2 9,1 10,2 11,1 12,1 13,2 14,2 15,3 16,2 17,1 18,2 19,1 20,2 21,1 23,3 24,2
25,1/2/3/4 26,2 27,2 28,2 29,2 30,2 31,2 32,1 33,4 34,2 35,1 36,2 37,1 38,2 39,1 40,1 41,1 42,1 43,2
44,2 45,2 46,2 48,2 49,1 50,1/3 51,1 52,2 53,1 54,2 55,1 56,2 57,2 58,1 60,3 61,3 62,1 63,2 64,2
Veliveriaflilpesl
1,2 2,1 3,1 4,3 5,2/3 6,1 7,2 9,1 10,2 11,1 12,3 13,2 14,1/2 15,3 16,2 17,1 18,2 19,1 20,1 21,1 23,3
24,2 25,2/3/4 26,1 27,2 28,2 29,2 30,2 3,1,2 32,1 33,5 34,2 35,1 36,2 37,1 38,2 39,1 40,1 41,1 42,1
43,2 44,2 45,2 46,2 48,1 49,1 50,3 51,1 52,2 53,154,1/2 55,1 56,2 57,1 58,1 60,3 61,1 63,2 64,2
Veliveriafulvibarbisl
1,2 2,1 3,1 4,3 5,2 6,1 7,2 9,1 10,2 11,1 12,1 13,2 14,U 15,3 16,2 17,1 18,2 19,1 20,1 21,1 23,3 24,2
25,4 26,1 27,2 28,2 29,2 30,2 31,2 32,1 33,4 34,2 35,1 36,2 37,1 38,3 39,2 40,1 41,1 42,1 43,2 44,2
45,2 46,2 48,2 49,1 50,3 51,1 52,1 53,1 54,2 55,1 56,2 57,2 58,1 60,3 61,2 62,1 63,2 64,2
Vetiveria nigritanal
1,2 2,1 3,1 4,3 5,3 6,1 7,2 9,1 10,2 11,1 12,1 13,2 14,2 15,3 16,2 17,1 18,2 19,1 20,1 21,1 23,3 24,2
25,5 26,1 27,2 28,2 29,2 30,2 31,2 32,1 33,5 34,2 35,1 36,2 37,1 38,3 39,2 40,1 41,2 42,1 43,2 44,2
45,2 46,2 48,1 50,3 51,1 52,2 53,1 54,2 55,1 56,2 57,2 58,1 60,2 61,1 63,2 64,2
Vetiveria zizanioides/
1,2 2,1 3,1 4,3 5,3 6,2 9,1 10,2 11,1 12,1 13,2 14,1/2/3 15,3 16,2 17,1 18,2 19,1 20,2 21,2 23,3 24,2
25,5 26,2 27,2 28,2 29,2 30,2 31,2 32,1 33,4 34,2 35,1 36,2 37,1 38,3 39,2 40,2 41,1 42,1 43,2 44,2
45,2 46,2 48,1 49,1 50,2 51,2/3 52,2 53,1 54,1 55,2 61,1 63,2 64,2
427. Bothriochloa bioba/
1,2 2,1 3,1/2 4,3 5,2/3 6,1 7,2 9,2 10,2 11,1 12,1 13,1 14,1 15,2 17,2 18,2 22,2 23,1 24,2 25,U 26,1
27,1 28,1 29,2 30,2 31,2 32,1 33,5 34,2 35,1 36,1 37,3 38,3 39,2 40,1 41,1 42,2 43,2 44,2 45,2 46,1
47,2 48,2 49,1 50,3 51,1 52,2 53,1 54,2 55,1 56,1 57,1 58,1 60,3 61,3 62,2 63,2 64,2
#28. Bothriochloa decipiens/
1,2 2,2 3,1/2 4,3 5,1 6,1 7,2 9,2 10,2 11,1 12,2 13,1/3 14,1 15,2 17,2 18,2 22,2 23,1/2 24,2 25,U 26,1
27,1 28,1 29,2 30,2 31,2 32,1 33,5 34,2 35,1 36,1 37,3 38,3 39,2 40,1 41,1 42,3 43,2 44,2 45,1 46,1
47,2 48,2 49,1 50,2 51,1 52,2 53,2 54,1 55,1 56,1 57,1 58,1 60,3 61,1 63,2 64,2
429. Bothriochloa insculpta/
1,2 2,1 3,2 4,2 5,1 6,1 7,1 8,2 9,2 10,2 11,1 12,3 13,1 14,1/2 15,2 17,2 18,2 22,2 23,1/2/3 24,2 25,U
26,1 27,1 28,1 29,2 30,2 31,2 32,1 33,5 34,2 35,1 36,1 37,3 38,3 39,2 40,1 41,1 42,3 43,2 44,2 45,1
46,2 48,2 49,1 50,2 51,1 52,2 53,2 54,1 55,1 56,1 57,1 58,1 60,3 61,1 63,2 64,2
354
Bothriochloa ischaemum/
1,2 2,1 3,2 4,1 5,1/3 6,1 7,1/2 8,1 9,1 10,2 11,2 12,1 13,1 14,1 15,3 16,1 17,2 18,2 22,1/2 23,1/3 24,2
25,U 26,2 27,1 28,1 29,2 30,2 31,2 32,1 33,5 34,2 35,1 36,1 37,3 38,3 39,2 40,1 41,1 42,1 43,2 44,2
45,2 46,1 47,2 48,2 49,2 50,2 51,1 52,2 53,2 54,1 55,1 56,1 57,1 58,1 60,3 61,1 63,2 64,2
Bothriochloa kuntzeana/
1,2 2,2 3,1 4,3 5,3 6,1 7,1 8,2 9,1 10,2 11,1 12,1 13,1 14,U 15,2 17,2 18,2 22,2 23,1/3 24,2 25,U 26,1
27,1 28,1 29,2 30,2 31,2 32,1 33,4 34,2 35,1 36,1 37,3 38,3 39,2 40,1 41,1 42,3 43,2 44,2 45,1 46,2
48,2 49,1 50,3 51,1 52,2 53,2 54,1 55,1 56,2 57,2 58,1 60,3 61,1 63,2 64,2
Bothriochloa macera/
1,2 2,1 3,1/2 4,3 5,1 6,1 7,2 9,2 10,2 11,1 12,1 13,1 14,1 15,3 16,1 17,2 18,2 22,2 23,1/2 24,2 25,U
26,1 27,1 28,1 29,2 30,2 31,2 32,1 33,5 34,2 35,1 36,1 37,3 38,3 39,2 40,1 41,1 42,U 43,2 44,2 45,1/2
46,1 47,2 48,2 49,1 50,2 51,1 52,2 53,2 54,1 55,1 56,1 57,1 58,1 60,3 61,1 63,2 64,2
Bothriochloa pertusa/
1,2 2,2 3,1/2 4,1 5,1 6,1 7,1 8,2 9,1 10,2 11,2 12,3 13,1 14,1 15,2 17,2 18,2 22,1 23,3 24,2 25,U 26,1
27,1 28,1 29,2 30,2 31,2 32,1 33,4 34,2 35,1 36,1 37,3 38,3 39,2 40,1 41,1 42,3 43,2 44,2 45,1 46,1
47,2 48,2 49,1 50,2 51,1 52,2 53,2 54,1 55,1 56,1 57,1 58,2 59,2 60,3 61,1 63,2 64,2
434. Bothriochloa radicansl
1,2 2,1 3,2 4,2 5,1 6,1 7,1 8,1 9,1 10,2 11,112,2 13,1 14,1/2 15,3 16,1 17,2 18,2 22,1/2 23,3 24,2 25,U
26,1 27,1 28,U 29,2 30,2 31,2 32,1 33,5 34,2 35,1 36,U 37,3 38,3 39,2 40,1 41,1 42,2 43,2 44,2 45,2
46,1 47,2 48,2 49,2 50,2 51,1 52,2 53,2 54,1 55,1 56,1 57,1 58,1 60,3 61,1 63,2 64,2
Bothriochloa saccharoides/
1,2 2,1 3,1/2 4,3 5,U 6,U 7,U 8,U 9,U 10,2 11,1 12,1 13,1/2 14,1/2 15,2 17,2 18,2 22,2 23,1/2/3 24,2
25,U 26,1 27,1 28,1 29,2 30,2 31,2 32,1 33,5 34,2 35,1 36,1 37,3 38,3 39,2 40,1 41,1 42,1 43,2 44,2
45,2 46,2 48,2 49,1 50,2 51,1 52,2 53,2 54,1 55,1 56,1 57,2 58,2 59,2 60,3 61,1 63,2 64,2
Capiilpedium assimile/
1,22,23,24,3 5,1 6,1 7,1/2 8,1 9,1 10,2 11,1 12,2 13,1 14,1/2 15,3 16,2 17,1 18,2 19,220,121,1 23,U
24,2 25,2 26,1 27,1 28,1 29,2 30,2 31,2 32,1 33,5 34,2 35,1 36,1 37,3 38,3 39,2 40,1 41,1 42,2 43,2
44,2 45,2 46,2 48,2 49,1 50,2 51,1 52,2 53,2 54,1 55,1 56,2 57,2 58,1 60,3 61,1 63,2 64,2
Capillipediumfihiculme/
355
1,1 2,23,25,1 6,1 7,1 8,1 9,1 10,2 11,1 12,3 13,1 14,2/3 15,2 17,1 18,2 19,2 20,121,1 23,U24,225,U
26,1 27,1 28,1 29,230,2 31,2 32,1 33,4 34,2 35,1 36,1 37,3 38,3 39,2 40,1 41,1 42,1 43,2 44,2 45,2
46,2 48,2 49,1 50,2 51,1 52,2 53,2 54,1 55,1 56,1 57,1 58,1 60,3 61,1 63,2 64,2
Capillipedium huegelii/
1,2 2,2 3,1 4,3 5,1 6,1 7,1 8,1 9,1 10,2 11,1 12,1 13,1 14,U 15,3 16,2 17,1 18,2 19,2 20,2 21,1 23,U
24,2 25,U 26,1 27,1 28,1 29,2 30,2 31,2 32,1 33,5 34,2 35,1 36,1 37,3 38,3 39,2 40,1 41,1 42,3 43,2
44,2 45,2 46,2 48,2 49,1 50,2 51,1 52,2 53,2 54,1 55,1 56,2 57,2 58,1 60,3 61,1 63,2 64,2
Capihipedium parvflorum/
1,2 2,1 3,1 4,3 5,3 6,1 7,1 8,1 9,1 10,2 11,1 12,1 13,1 14,1 15,3 16,2 17,1 18,2 19,2 20,121,123,1/2/3
24,1 25,1 26,1 29,3 30,3 31,2 32,1 33,4/5 35,1 36,1 37,3 38,3 39,2 40,1 41,1 42,2 43,2 44,2 45,2 46,1
47,1 48,2 49,1 50,3 51,1 52,2 53,2 54,1 55,1 56,1 57,1 58,1 60,3 61,1 63,2 64,2
Capihhipedium spicigerum/
1,2 2,1 3,1 4,3 5,3 6,1 7,1/2 8,1 9,1 10,2 11,1 12,1 13,1 14,1/2 15,3 16,2 17,1 18,2 19,2 20,1 21,1
23,1/2/3 24,2 25,2/3/4 26,1 27,1 28,1 29,2 30,2 31,2 32,1 33,4 34,2 35,1 36,1 37,3 38,3 39,2 40,1 41,1
42,2 43,2 44,2 45,2 46,1 47,1 48,2 49,1 50,2 51,1 52,2 53,2 54,1 55,1 56,1 57,1 58,1 60,3 61,1 63,2
64,2
Dichanthium annulatumi
1,2 2,1/2 3,1 4,1 5,1/3 6,1 7,1 8,2 9,1 10,2 11,1 12,2 13,1 14,1 15,2 17,2 18,2 22,2 23,1/2 24,2 25,5
26,1 27,2 28,1 29,2 30,2 31,1 32,1 33,2/4/5 34,1 35,1 36,1 37,3 38,3 39,2 40,1 41,1 42,2 43,2 44,2
45,2 46,147,2/3/4 48,2 49,2 50,2 51,1 52,2 53,2 54,1 55,1 56,1 57,1 58,1 60,3 61,1 63,2 64,2
Dichanthium aristatum/
1,2 2,1 3,1 4,2 5,3 6,1 7,2 9,1 10,2 11,1 12,2 13,1 14,U 15,3 16,2 17,2 18,2 22,2/3 23,1/3 24,2 25,5
26,1 27,2 28,2 29,2 30,2 31,1 32,1 33,4/5 34,1 35,1 36,1 37,3 38,3 39,2 40,1 41,1 42,3 43,2 44,1 45,2
46,1 47,2 48,2 49,1 50,2 51,1 52,2 53,2 54,1 55,1 56,1 57,1 58,2 59,2 60,3 61,1 63,2 64,2
Dichanthium caricosum/
1,2 2,2 3,2 4,3 5,3 6,1 7,2 9,1 10,2 11,1 12,2 13,1 14,U 15,3 16,1 17,2 18,2 22,2 23,1/2/3 24,2 25,U
26,1 27,2 28,1 29,2 30,2 31,1 32,1 33,4/5 34,1 35,1 36,1 37,3 38,3 39,2 40,1 41,1 42,1 43,2 44,2 45,2
46,147,2/3/4 48,2 49,2 50,2 51,1 52,2 53,2 54,1 55,1 56,1 57,1 58,1 60,3 61,1 63,2 64,2
Dichanthiumfoveolatuml
356
1,2 2,1 3,2 4,1 5,3 6,1 7,1 8,1 9,1 10,2 11,1 12,1 13,1 14,U 15,3 16,1 17,2 18,2 22,3 23,1 24,2 25,5
26,1 27,2 28,1 29,2 30,2 31,1 32,1 33,5 34,1 35,1 36,1 37,3 38,3 39,2 40,1 41,1 42,2 43,2 44,2 45,1
46,2 48,2 49,2 50,2 51,1 52,2 53,2 54,1 55,1 56,1 57,1 58,1 60,3 61,1 63,2 64,2
#45. Dichanthium micranthum/
1,2 2,1 3,1 4,1 5,1 6,1 7,1/2 8,1 9,1 10,2 11,2 12,1 13,1 14,1 15,U 16,U 17,2 18,2 22,3 23,1 24,2 25,U
26,1 27,2 28,1 29,2 30,2 31,1 32,1 33,4 34,1 35,1 36,1 37,3 38,3 39,2 40,1 41,1 42,3/4 43,2 44,2 45,1
46,2 48,2 49,2 50,2 51,1 52,2 53,2 54,1 55,1 56,1 57,1 58,1 60,3 61,1 63,2 64,2
446. Dichanthium sericeum/
1,2 2,2 3,1/2 4,1 5,3 6,1 7,1 8,2 9,1 10,2 11,1 12,2 13,1 14,U 15,2 17,2 18,2 22,1 23,3 24,2 25,U 26,1
27,2 28,1 29,2 30,2 31,1 32,1 33,4/5 34,1 35,1 36,1 37,3 38,3 39,2 40,1 41,1 42,3 43,2 44,2 45,2 46,1
47,2/3/4 48,2 49,2 50,2 51,1 52,2 53,2 54,1 55,1 56,1 57,1 58,2 59,2 60,3 61,1 63,2 64,2
Sorghastrum ellio#ll/
1,2 2,1/2 3,1 4,3 5,3 6,2 9,1 10,1 11,1 12,U 13,1 14,1/2 15,2 17,1 18,2 19,1 20,1 21,1 23,3 24,1 25,1
26,1 29,3 30,4 31,2 32,1 33,1 35,1 36,1 37,3 38,3 39,2 40,1 41,1 42,1 43,2 44,2 45,2 46,1 47,1 48,2
49,2 50,3 51,1 52,2 53,1 54,1 55,1 56,1 57,1 58,1 60,3 61,1 63,2 64,2
Sorghastrumfuscescens/
1,2 2,1 3,1 4,3 5,3 6,1 7,1 8,1 9,1 10,1 11,1 12,1 13,1 14,1/2 15,2 17,1 18,2 19,1 20,1 21,1 23,3 24,2
25,2/3/4/5 26,1 27,2 28,1 29,2 30,2/4 31,2 32,1 33,1/2 34,2 35,1 36,1 37,3 38,3 39,2 40,1 41,1 42,1
43,2 44,2 45,2 46,1 47,1 48,2 49,2 50,2 51,1 52,1 53,1 54,2 55,1 56,1 57,1 58,1 60,3 61,1 63,2 64,2
Sorghastrum incompletum var. bipennatumi
1,2 2,1 3,1 4,3 5,3 6,1 7,1 8,1 9,1 10,1 11,1 12,2 13,1 14,1/2 15,2 17,1 18,2 19,1 20,1 21,1 23,3 24,1
25,1 26,2 29,3 30,4 31,2 32,1 33,1 35,1 36,1 37,3 38,3 39,2 40,1 41,1 42,1 43,2 44,2 45,2 46,1 47,1
48,2 49,2 50,2 51,1 52,2 53,1 54,1 55,1 56,1 57,1 58,1 60,3 61,3 62,1 63,2 64,2
Sorghastrum minarum/
1,2 2,1/2 3,1 4,3 5,3 6,1 7,1 8,1 9,1 10,1 11,1 12,2 13,1 14,2 15,2 17,1 18,2 19,2 20,2 21,1 23,3 24,2
25,1 26,1 29,3 30,4 31,2 32,1 33,1 35,1 36,1 37,3 38,2 39,1 40,1 41,1 42,1 43,2 44,2 45,2 46,147,2/4
48,2 49,2 50,2 51,1 52,2 53,1 54,2 55,1 56,1 57,1 58,2 59,1 60,3 61,1 63,2 64,2
Sorghastrum nulans/
1,2 2,2 3,1 4,1 5,2 6,1 7,1 8,1 9,1 10,1 11,1 12,1 13,1 14,1/2 15,2 17,1 18,2 19,2 20,2 21,1 23,3 24,2
25,U 26,1 29,2 30,4 31,2 32,1 34,2 33,1 35,1 36,1 37,3 38,3 39,2 40,141,142,2/3 43,2 44,2 45,2 46,1
47,1 48,2 49,2 50,2 51,1 52,1 53,1 54,1 55,1 56,1 57,1 58,1 60,3 61,1 63,2 64,2
357
Sorghaslrum pogonostachyum/
1,2 2,1 3,1 4,1 5,3 6,1 7,2 9,1 10,1 11,1 12,U 13,1 14,U 15,2 17,1 18,2 19,1 20,1 21,1 23,3 24,2
25,2/3/4 26,1 27,2 28,1 29,2 30,2 31,2 32,1 33,1/2 34,2 35,1 36,1 37,3 38,3 39,2 40,1 41,1 42,1 43,2
44,2 45,2 46,1 47,1 48,2 49,2 50,2 51,1 52,2 53,1 54,1 55,1 56,2 57,2 58,1 60,2 61,1 63,2 64,2
Sorghaslrum setosum/
1,2 2,1 3,1 4,1 5,3 6,1 7,2 9,1 10,1 11,1 12,2 13,1 14,2 15,2 17,1 18,2 19,1 20,1 21,1 23,3 24,2 25,U
26,1 27,2 28,1 29,2 30,4 31,2 32,1 33,1 34,2 35,1 36,1 37,3 38,3 39,2 40,1 41,1 42,1 43,2 44,2 45,2
46,1 47,1 48,2 49,2 50,2 51,1/2 52,2 53,1 54,1 55,2 61,1 63,2 64,2
Sorghum arundinoceum/
1,1/2 2,2 3,1 4,3 5,1 6,1 7,1/2 8,1 9,1 10,2 11,1 12,3 13,1 14,1/2 15,3 16,1 17,1 18,2 19,2 20,1 21,2
23,1 24,2 25,2/3/4 26,1 27,2 28,1 29,2 30,2 31,2 32,1 33,5 34,2 35,1 36,1 37,3 38,3 39,2 40,1 41,1
42,2 43,2 44,2 45,2 46,1 47,1 48,2 49,2 50,2 51,1 52,1 53,1 54,2 55,1 56,1 57,1 58,1 60,3 61,2 62,2
63,2 64,2
Sorghum hulepensel
1,2 2,2 3,1 4,1 5,2/3 6,1 7,2 9,1 10,2 11,1 12,1 13,1 14,1/2/3/4 15,2 17,1 18,2 19,2 20,1 21,2 23,1 24,2
25,1/2/3 26,U 27,2 28,1 29,2 30,2 31,2 32,1 33,4 34,2 35,1 36,1 37,3 38,3 39,2 40,1 41,1 42,3 43,2
44,2 45,2 46,1 47,1 48,2 49,1 50,2 51,1/2 52,1 53,1 54,2 55,1/2 56,2 57,2 58,1 60,3 61,1 63,2 64,1
Sorghum macrospermuml
1,1 2,2 3,1 5,1 6,1 7,2 9,1 10,2 11,1 12,4 13,1 14,4 15,3 16,1 17,1 18,2 19,1/2 20,1 21,2 23,1 24,2
25,4/5 26,1 27,2 28,1 29,2 30,2 31,2 32,1 33,5 34,2 35,1 36,1 37,3 38,3 39,2 40,1 41,1 42,1 43,2 44,2
45,2 46,2 48,2 49,2 50,2 51,1 52,1 53,1 54,2 55,1 56,1 57,1 58,1 60,3 61,1 63,2 64,1/2
Sorghum nitidum/
1,2 2,2 3,1 4,3 5,1 6,1 7,1 8,1 9,1 10,2 11,1 12,2 13,1 14,3 15,3 16,2 17,1 18,2 19,1 20,1 21,2 23,3
24,2 25,2/3 26,U 27,2 28,1 29,2 30,2 31,2 32,1 33,4/5 34,2 35,1 36,1 37,3 38,3 39,2 40,1 41,1 42,3
43,2 44,2 45,2 46,1 47,1 48,2 49,2 50,2 51,1/2 52,1 53,1 54,2 55,1/2 56,1 57,2 58,1 60,3 61,1 63,2
64,1
Sorghum versicolor/
1,1 2,2 3,1 5,3 6,1 7,1 8,2 9,1 10,2 11,1 12,U 13,1 14,3 15,2 17,1 18,2 19,1 20,1 21,2 23,3 24,2
25,2/3/4 26,1 27,2 28,1 29,2 30,2 31,2 32,1 33,5 34,2 35,1 36,1 37,3 38,3 39,2 40,1 41,1 42,3 43,2
44,2 45,2 46,1 47,1 48,2 49,2 50,2 51,1 52,1 53,1 54,2 55,1 56,1 57,1 58,1 60,3 61,1 63,2 64,1
Sorghum plumosumi
358
1,2 2,1 3,1 4,3 5,3 6,1 7,1 8,1 9,1 10,2 11,112,2 13,1 14,1/2/3 15,3 16,1 17,1 18,2 19,1 20,2 21,2 23,3
24,2 25,1/2/3/4 26,1 27,2 28,1 29,2 30,2 31,2 32,1 33,5 34,2 35,1 36,1 37,3 38,2 39,1 40,1 41,1 42,3
43,2 44,2 45,2 46,1 47,1 48,2 49,1 50,2 51,1 52,1 53,1 54,2 55,1 56,1 57,1 58,2 59,1 60,3 61,3 62,2
63,2 64,1
#60. Sorghum stipoideumi
1,1 2,2 3,1 5,1 6,1 7,1/2 8,1/2 9,1 10,2 11,1 12,U 13,1 14,3 15,2 17,1 18,2 19,1 20,1 21,2 23,3 24,2
25,1/2 26,1 27,2 28,1 29,2 30,2 31,2 32,1 33,4 34,2 35,1 36,1 37,3 38,2 39,1 40,1 41,1 42,3 43,2 44,2
45,2 46,1 47,1 48,2 49,2 50,2 51,1 52,1 53,1 54,2 55,1 56,1 57,1 58,2 59,1 60,3 61,1 63,2 64,1
359
Appendix 11.4
Table 11.1: A table showing character coding for clad istic
analysis.
Characters
Character states
1. Longevity
0. prennial
1. annual
2. Perennating organs
0. absent
1 .present
3. CuIm internode
0. terete
1. channelled on one side
4. Leaf sheath
0. auriculate
1. non auriculate
5. Leaf blade width
0. broad
1. narrow
6. Ligule
0. unfringed membrane
1. fringed membrane
7. Inflorescence type
0. paniculate
racemose
spatheate
8. Particulate inflorescence
0. with simple branches
1. with divided branches
9. Racemose inflorescence
0. subdigitate racemes
digitate racemes
a single raceme
10. Racemes (disarticulation)
0. breaking at the joints
1. falling entire
11. Rachis internodes
0. without a translucent groove
1. with a translucent groove
12. Spikelets
0. consistently in triplets
1. sometimes in triplets and pairs
360
consistently paired
solitary
13. Spikelet type
0. in a triplet of 1 seesile and 2
pedicellate
in a pedicellate/sessile pair
sessile and accompanied by barren
pedicels
sessile only
14. Pedicellate spikelet
0. present
1. absent
15. Lowest sessile spikelet
0. heterogamous
1. homogamous
16. Pedicel development
0. well developed
rudimentary
absent
17. Callus point shape
0. blunt
1. pungent
18. Lower glume (of sessile spikelet)
0. two keeled
1. not two keeled
19. Lower glume awn
0. absent
1. present
20. Lower glume wings
0. absent
1. present
21. Lower glume pitting
0. without a pit
1. with a pit
22. Upper glume
0. awnless
1. awned
23. Palea
0. present
1. absent
24. Lodicules
0. present
1. absent
25. Lodicule margins and apex
0. ciliate
1. glabrous
26. Upper lemma
0. awnless
1. awned
361