zyxwvutsrqpo
zyxwvutsrq
zyxwvutsrqpon
zyxwv
Botanical ~7ournalofthe Linnean Sociely (1992), 108: 55-81. With 5 figures
Elucidative studies on the generic concept of
Senecio (Asteraceae)
zy
P. LESZEK D. VINCENT*
Department o f Botany, University o f the Witwatersrand, PIBag 3, Wits, 2050, Republic
o f South Africa
AND
FIONA M. GETLIFFE, F.L.S.
zyxwvutsrqp
zyxwvutsr
zyxwvutsr
zyxw
Herbarium Pac$cum, Department o f Botany, Bernice P. Bishop Museum, P.O. Box
19000-R, Honolulu, Hawaii 96819, U.S.A.
Receiued j’iily 1987, reuised and accepted,for publication Ocloher 1990
VINCENT, 1’. L. D. & GEILIFFE, F. M., 1992. Elucidative studies on the generic concept
of Senecio (Asteraceae).This paper presents the results of studies of the generic concept of Senecio
Aenm slricto. ‘l’he sample of taxa studied consisted of 93 Natal senecios (including seven varieties and
two forms), fivr Cape heterochromous senecios and nine non-southern African senecios, including
the type of thc gcnus, S. uulgaris L. Also included in thc study were six species from taxonomically
closely related genera in the tribe Senecioneae and one Senecio of uncertain taxonomic position. T h e
phenotype of these taxa was investigated with respect to a large number (122) of morphological and
micromorphological characters. Six characters were selcrtcd as being taxonomically important with
respect to elucidating the generic concept of Senecio sensu Jlrictu. T h e generic concept of Senecio has
hcen provisionally re-circumscribed and the generic status of rach of the senrcios and non-senrcios
studied has been tested according to this concept of Senecio sensu slricto sensu Vincent. In the light of
this concept of Senecio, the following species are recommended for exclusion from Senecio sensu strict0
sensn Vincent: S. ciaampelinus, S. transuaalensis, S. syringiflius and S.hockii. The following species are
considered to be peripheral to Senecio sensu slrzcto sensu Vincent: S. tanacetopsis, S. semzniueus,
S. medley-woodii, S. lamoides, S.helrninthioides, S. barbertonicus, S. breuilorus, S. uiminalts, S. radicans and
S.fu1gens. Before any taxonomir changes are made to the current composition of Senecio, the concept
of Senecio sensu striclo sensu Vincmt, is being tested on a worldwide sample of thr genus.
ADDITIONAL KEY WORDS:-Compositae
-
numerical taxonomy.
CONTENTS
Introduction .
.
.
Material and methods .
Results and observations
Discussion.
.
.
.
Conclusion
.
.
.
Acknowledgements
.
Rcferences
. . .
Appendix .
. . .
.
.
.
.
.
.
.
. 56
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
. 57
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
0024-4074/92/010055
+ 27 $03.00/0
55
.
.
.
.
.
.
.
.
.
. 76
74
74
zyxwv
.
.
*Author for correspondence.
63
73
.
.
.
.
75
0 1992 The Linnean Society of London
56
zyxwvutsrqp
zyxwvuts
zy
P. L. D. VINCENT AND F. M. GETLIFFE
INTRODUCTION
The genus Senecio was formally circumscribed by Linnaeus in Genera Plantarum
(1754), although a number of species with this name appeared in Species
Plantarum (Linnaeus, 1753) including the type Senecio vulgaris L. (Farr, Leussink
& Stafleu, 1979). Since then, the genus has been described from many parts of
the world and the generic concept has been subject to varied interpretations.
Thus, Senecio now includes species which exhibit a range of character variation
that overlaps, and even exceeds, the combined ranges of species currently
referred to several other genera (Jeffrey et al., 1977).
This expended concept of Senecio results in an unwieldy and patently
paraphyletic genus and, in the spirit of the ‘New Synantherology’ (King &
Robinson, 1970), several studies in the Senecioneae (Robinson & Brettell,
1973a-d, 1974; Robinson, 1974, 1975; Jeffrey et al., 1977; Nordenstam, 1978;
Jeffrey, 1979; Wetter, 1983) have attempted to improve generic concepts in
many groups within the Asteraceae. I n particular, Jeffrey (1979) advocated the
application of “uniform criteria” to studies of the Senecio complex, throughout its
geographic range. O n this basis, it was suggested, systematic conclusions on
generic and sectional limits might be more valid. After preliminary surveys,
Jeffrey et al. (1977) concluded that:
1. The species currently placed in Senecio can be subdivided into 16 groups,
only one of which (group IX) should be called Senecio, since it contains the type
species.
2. The remaining groups might be distinct genera, possibly new or possibly
congeneric with other widely recognized genera in the family.
Jeffrey (1979), re-evaluated group IX in the light of evidence from studies of
the pappus (Drury, 1967), chemical characters (Robins, 1977a, b; Bohlmann et
al., 1979) and succulent species (Jeffrey, 1979). As a result, group I X was further
subdivided into two sections: “eusenecionoids” which includes the type species,
and “gynuroids”.
It is estimated that Senecio consists of “perhaps 1000 species” (Jeffrey & Chen,
1984), which is considerably fewer than earlier estimates of c. 1500 species
(Nordenstam, 1978) or the c. 3000 species of the earlier estimates by Jeffrey
Ueffrey et al., 1977). Thus the application of the “uniform criteria” appears to
result in a narrower concept of the genus which may be further refined, as
attention is drawn to the fact that the “generic limits in many areas are
uncertain” (Jeffrey et al., 1977). I n contrast, Barkley (1985), argued for the
retention of the broad concept of Senecio, but this has not yet influenced the
direction of research on this genus.
Unfortunately, Jeffrey & Chen (1984) do not comment explicitly on their
concept of Senecio sensu strict0 but Nordenstam (1978) did mention characteristics
that he perceived as being characteristic of the genus in its narrower sense. These
are ‘senecioid’ filament collars, ‘radial’ endothecial tissue, discrete stigmatic
areas in a ‘banded’ configuration, truncate style branches and a haploid
chromosome number of 10.
The present study was undertaken with all but the last of these features in
mind, and with the intention of contributing to what must finally be a
zyxwvu
zy
zyxwv
GENERIC CONCEPT OF SENECIO
57
worldwide revision of the generic concept of Senecio, based on the concept of
“uniform criteria” as proposed by Jeffrey et al. (1977).
MATERIAL AND METHODS
The number and selection o f taxa
Eighty-four species of the c. 130 species of the genus Senecio in Natal, together
with five species of Senecio from the Cape, nine Senecio species from outside South
Africa and six species from other genera in the tribe Senecioneae (Table 1) were
investigated. Most of the material came from NU, the remainder from BOL, K ,
N H and PRE.
The Natal senecios studied included a11 the purple senecios described by
Hilliard (1977) and a large number of yellow senecios. In the selection of the
latter, both putatively closely allied and distantly allied species (on the basis of
similar morphology) were selected. At the start of this study some of these
senecios had an uncertain taxonomic position: S. cissampelinus (DC.) Sch. Bip.,
S.fulgens
D. Hooker) Nicholson, S. transvaalensis Bolus & S. viminalis
Bremekamp, incertae sedis,jde Hilliard, 1977, 1978 and S. syringiflius 0. Hoffm.
incertae sedis, fide Jeffrey personal communication, 1984, Jeffrey, 1986 and
S. hookii De Wild. & Muschl., incertae sedis, jide Jeffrey personal communication,
1984; Emilia hockii (De Wild & Muschl.) C. Jeffrey,jide Jeffrey, 1986.
The species descriptions in Hilliard ( 1977) were used to select representative
specimens of the Natal species. Harvey’s descriptions in Flora Capensis (Harvey,
1865) were used in the selection of the Cape senecios. While we decided which
non-southern African senecios and related genera to include in this study, the
selection of the representatives of these taxa was done on our behalf by M r
C. Jeffrey of The Royal Botanic Gardens, Kew (C. Jeffrey, personal
communication, 1984). A complete list of all the taxa investigated is given in
Table 1.
zyxwvut
zyx
u.
Morphology and micromorphology
A minimum of three representatives of each taxon were studied in detail. I n
the selection of the representatives an attempt was made to select specimens from
widely different localities or habitats, to enable the effect of any environmentally
induced variation to be encountered.
Preparation of vegetative and joral tissue
Samples of a t least four leaves, four bracts, four calyculus bracts and six to
eight florets of each specimen, were cleared in c. 85% lactic acid at c. 120°C. The
clearing process was carefully monitored to achieve adequate clearing, without
excessive softening of the tissue.
A portion of cleared leaf tisssue (c. 15 x 15 mm) of each specimen sampled was
mounted in lactic acid ( 85%).
The florets were dissected such that the corollas, the anthers and attached
filaments and the cypselae, usually with the style still attached, were all
displayed separated from each other.
Light microscope observations were made using Reichert and Olympus
compound microscopes (the latter with phase contrast objectives) and a Zeiss
dissecting microscope.
-
58
zyxwvutsrq
zyxwvu
zyxwv
P. L. D. VINCENT AND F. M. GE'ILIFFE
TABLE
1. Complete list of t h e species a n d varieties of Seneciu a n d related genera included i n this
study (SUM = 101).
zyxwvutsrqp
Natal senecios: The numbers in the left-hand margin correspond to the species numbers for Senecio in Hilliard
(1977) and are for identification purposes only.
1. S. madagascariensis Poiret
2. S. skirrhodon DC.
3. S. inaequidens DC.
4. S. harueianus MacOwan
5. S. polyanthemoider Schultz Bipontinus
6. S.pterophorus DC.
7 . S.juniperinus L. fil.
9. S. serratuloides DC.
10. S. microglossus DC.
14. S.pandur2frmis Hilliard
15. S. achilleifolius DC.
16. S. tanacetopsis Hilliard
17. S.seminiueus Wood & Evans
18. S. haygarthii Hilliard
20. S. medley-woodii Hutchinson
22. S.purpureus L.
23. S.gerrardii Harvey
24. S. erubescens Aiton
(24A) var. erubescens
(24B) var. crepidfolius DC.
(24C) var. incisus DC.
(24D) var. dzchotumus DC.
25. S. sandersonii Harvey
26. S.glanduloso-lanosus Thellung
27. S. uariabilis Schultz Bipontinus
28. S. umgeniensis Thellung
29. S.glanduloso-pilosus Volkens & Muschler
30. S. subcuriaceus Schlechter
3 1. S. cathcartensis 0. Hoffmann
32. S. ngoyanus Hilliard
33. S.polyodon DC.
(33A) var. polyodon
(33B) var. subglaber (0.Kuntze) Hilliard & Burtt
34. S. aff. speciosus
35. S. speciosus Willdenow
36. S. macrocephalus DC.
37. S. barbatus DC.
38. S.poseideonis Hilliard & Burtt
(38A) radiate forms
(38B) discoid forms
39. S. aff. poseideonis
40. S. arabidifolius 0. Hoffmann
4 1 . S. rhyncholaenus DC.
42. S. subrubriflorus 0. Hoffmann
43. S. hastatus L.
46. S. hirsutilobus Hilliard
47. S. ingeliensis Hilliard
48. S. natalicola Hilliard
50. S. hieracioides DC.
5 1. S. consanguineus DC.
53. S.chysocoma Meerburgh
55. S.asperulus DC.
56. S. hypochoerideus DC.
57. S. mooreanus Hutchinson
63. S.paludajinis Hilliard
64. S. ajinis DC.
65. S.lydenburgensis Hutchinson & Burtt Davy
67. S. anomalochrous Hilliard
72. S. breuidentatus M. D. Henderson
zyxwvu
zy
zyxwv
zyxwvutsr
zyxwvuts
zyxwvuts
zyxwvutsrq
zyxwvuts
GENERIC C O N C E P I OF SENECIO
59
TABLE
1. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHG
continued
~
73.
74.
75.
76.
77.
8I .
82.
83.
85.
86.
88.
89.
90.
97.
98.
99.
10 1.
102.
103.
104.
107.
108.
109.
110.
1 1 I.
116.
1 1 7.
118.
119.
120.
12 I .
122.
124.
S. praeteritus Killick
S. coronatus (Thunberg) Harvey
S. macrospermus DC.
S. dregeanus DC.
S. discodregeanus Hilliard & Burtt
S. caudatus DC.
S.mauricei Hilliard & Burtt
S.saniensis Hilliard & Burtt
S. heliopsis Hilliard & Burtt
S. albanensis DC. var. doroniciflorus (DC.) Harvey
S. inornatus DC.
S. oxjri~/olius DC. (referred to Senecio sect. Peltati, incertae sedisjde Jeffrey, 1986: 934)
S. rhomboideus Harvey
S.bupleuroides DC.
S. urophjllus Conrdth
S. scitus Hutchinson & Burtt Davy
S. glaberrimus DC.
S.IatifoliuJ DC.
S. retrorsu~DC.
S.othonniflorus DC.
S. brachypodus DC.
S.pleistocephalus Spencer Moore
S. mikanioides [Otto ex] Harvey
( = Delairea odorata Lem.,fide Jeffrey, 1986: 933)
S. deltoideus Lessing
S. tamoides DC. (incertae sedisjde Jeffrey, 1986: 934)
S.helminthioides (Schultz Bipontinus) Hilliard
S. barbertonicus Klatt
S. breuilorus Hilliard
S. uiminalis Bremekamp (referable to Kleinia subg. Kleinia, cf. Jeffrey, 1986: 924)
S. radicans (L,fil.) Schultz Bipontinus (referred to Senecio sect. Rowleyana,fide Jeffrey, 1986: 935)
S. cissampelinus (DC.j Schultz Bipontinus
( = Mikaniopsis cissampelina (DC.) C. Jeffreyjde Jeffrey, 1986: 879)
S.,/ulgens (Hooker fil.) Nicholson
( = Kleiniafulgens (Hook. f.) Nich., subg. Notoniafide Jeffrey, 1986: 935)
S. transuaalensis Bolus
( = Emilia transuaalensis (Bolus) C. Jeffrey,jde Jeffrey, 1986: 919)
The numbers in the left hand margin of all the following species are for identification purposes only when
referred to in the text.
Cape
200.
20 1.
202.
203.
204.
heterochromous non-yellow (purple) senecios
S. grandijorus Berg.
S. c a ~ ~ ~ e DC.
~~ius
S. arenarins Thunb.
S. multibracteatus Harv.
5. elegans L.
Non-Southern African senecios
210. S. vulgaris L. (U.S.A., U.K., Europe & E. Asia)
212. S. uiscosus L. (U.S.A. & U.K.)
213. S. gallicus Chaix (Spain & Portugal)
214. S. califrnicus DC. (U.S.A.)
215 S. vernalis Waldst. & Kit. (Europe)
216. S. sylvaticus L. (U.S.A. & U.K.)
217. S. desfontainei Druce (Middle East & E. Asia)
2 19. S. erechtitoides Bak. (Madagascar)
220. S. lautus Willd. (Australia)
Senecios of uncertain taxonomic position and other genera in the tribe Senecioneae
223. S. syringifolius 0. Hoffm.
( = S. syringifolius 0. Hoffm. incerfue sedis,fide Jeffrey, 1986: 885)
60
zyxwvutsrqpon
zyxwvutsr
zyxwvuts
zyxwvu
P. L. D. VINCENT AND F. M. CETLIFFE
TABLE
1. --continued
224. S. hockii De Wild. & Muschl.
( = Emifia hockii (De Wild & Muschl.) C. Jeffrey,fide Jeffrey, 1986: 912)
225. Crassocephahm c e n u u m (L.fi1.) Moench
226. Emiliajammea Cass.
227. Gynura divaricala (L.) DC.
228. Notonia grandiflora DC.
( = Kfeinia grandgora (DC.) N. Rani,fide Jeffrey, 1986: 935)
229. Kleinia neriifolia Haw.
232. Cineraria geifalia L.
TABLE2. Complete list of all characters investigated in this
study of Senecio and related genera
Character
Vegetative habit
Rootstock
Plant height (max.)
Leaf distribution
Leaf length (max.): Radical
Cauline
Leaf width (max.): Radical
Cauline
Leaf shape:
Radical
Cauline
Leaf apex:
Radical
Cauline
Leaf margin:
Radical
Cauline
Leaf base:
Radical
Cauline
Leaf venation:
Radical
Cauline
Leaf texture:
Radical
Ca u I i n e
Leaf attachment:
Radical
Cauline
Leaf pubescence:
Abaxial-radical
Abaxial-cauline
Adaxial-radical
Adaxial-cauline
Leaf trichomes:
Base
Apex
Appendages
Glandular
Flowering stem:
Appearance
Pubescence
Glandular
Peduncles:
Appearance
Leaf stomata
Inflorescence bracts
Capitula:
Length
Diameter
Radiate/discoid
Homo/heterogamoIUS
Arrangement
Character
no.
008
009
010
01 1
012
013
014
015
016
017
018
019
020
02 1
022
023
024
025
026
027
028
029
030
03 1
032
033
034
035
036
037
038
039
040
04 I
042
043
044
045
046
047
048
zyxwvu
zyxwv
zyxw
61
GENERIC CONCEPT OF SENECIO
TABLE
2.-continued
Character
Involucre shape
Involucral bracts:
Number
Length
Shape
Pubescence
Trichomes
Apex vestiture
Length rel. to disc florets
Glandular
Colour
Calyculus bracts:
Number
Length
Position
Pubescence
Trichomes
Apex vestiture
Glandular
Colour
Shape
Flower colour:
Ray florets
Disc florets
Floret number
Ray florets:
Corolla length
Corolla shape
Limb venation
Corolla trichomes
Resinous
Limb apex
Limb shape
Limb epid. cell shape
Limb orientation
Floret number
Disc florets:
Corolla length
Corolla shape
Corolla venation
Corolla trichomes
Resinous
Lobe apex
Gynoecium (ray): Style-arm length
Style-arm apices
Stigmatic surface
Style base
Nectary
Ovary vestiture
Ovary wall crystals
Style-arm length
(Disc):
Style-arm apices
Stigmatic surface
Style base
Nectary
Ovary vestiture
Ovary wall crystals
Distal epidermal cell shape
Disc corolla:
Proximal epidermal cell shape
Androecium (disc): Anther apex
Anther base appearance
Anther length
Endothecial tissue: Cell shape
Cell wall configuration
Character
no.
049
050
05 1
052
053
054
055
056
05 7
058
059
060
06 1
062
063
064
065
066
06 7
068
069
070
07 1
072
073
074
075
076
077
078
079
080
08 1
082
083
084
085
086
087
088
089
090
09 1
092
093
094
095
096
097
098
099
100
zyxw
101
102
103
104
105
106
107
zyxwv
zyxwvuts
P. L. D. VINCE NI' AND F. M. GETLIFFE
62
TABLE
2.
continued
Charactri
Length
Basal width
Shape
Cypsela (ray):
Lrngth
Shape
Ribbing
Vestiturr
Pericarp cells (shapr)
(Disc):
Length
Shapr
Ribbing
Vestiture
Pericarp crlls (shape)
Uniforrn/dimorphic
Pappus:
Pappus setar (ray): Length
Appearance
Distal rell number
Apex
Length
(Disc):
Appearance
Distal ccll number
Apex
Filament collars:
ti
Character
no.
108
109
110
Ill
112
1 I3
I I4
I I5
116
117
I18
119
I20
121
I22
I23
I24
I25
I26
127
128
129
zyxwvutsrq
zyx
Characters 001 to 007 inclusive were inadmissible characters (Sneath
Sokal, 1973: 103).
Recording o f characters
One hundred and twenty-two characters were recorded for each taxon
investigated. A list of these characters is given in Table 2. Since the details of the
character states for each of these characters are, in most instances, quite lengthy,
these have not been provided, but can be obtained from PLDV.
The coding of the qualitative character states was achieved using
NOMINAL-LEVEL
MEASUREMENT
(non-parametric).
In
this
measurement procedure no assumption is made with respect to ordering or
distances between character states. Thus, the codes for the character states of the
qualitative characters are arbitrary values separating the character states for
each respective character.
The maximum value for the quantitative characters was recorded rather than
the mean because it indicates the maximum genetic potential of each character.
Some of these data were obtained from Hilliard (1977).
Number of measurementslobservations o f each character
At least five measurementslobservations were made of each character from
each of between three and five representatives of each taxon.
Master data matrix
Using these data, a master data matrix ( M D M ) (Appendix B) was developed
for the entire sample set. Where character plasticity in the qualitative characters
was encountered amongst the representatives of a taxon, either the average state
was chosen or a new state was described which encompassed the variation
zyxwvu
zyxwv
GENERIC CONCEPT OF SENECIO
63
encountered. This ‘average state’ or ‘new state’ code was entered into the MDM.
Variation in the continuous quantitative characters was accommodated in the
MDM after first having been subdivided into size classes. This subdivision into
size classes was achieved by subjecting the original data to frequency analyses,
with the aid of the programme; Statistical Package for the Social Sciences
(SPSS) by Nie et al. (1975). These analyses enabled peaks and troughs to be
noted in the distribution of the data for each quantitative character. Care was
taken to avoid the danger of dichotomizing the peaks into two or more size
classes when one size class was judged to be a more accurate biological
representation of the data producing the peak. Naturally the data of meristic
quantitative characters was entered directly into the MDM.
zyxwvuts
Selection of the taxonomica1l;v important charactersfor deJining the generic Limits of Senecio
The data in the M D M were subjected to frequency analyses using the SPSS
programme (1.c.). These analyses elucidated the distribution of the data for each
character for all the taxa investigated. The typology of specimens of S. vulgaris
was kept in mind when scrutinizing the character distribution patterns in search
of correlations between characters. The search for the aforementioned
correlations was assisted by ordination analyses using the SPSS programme
(I.c.). T h e correlations with the typology of S. vulgaris led to the recognition of
the putatively important characters for the circumscription of the generic
concept of Senecio. T h e data in the M D M , for all the taxa studied, can be
obtained from PLDV upon request. T h e data of the six characters of generic
importance (generic concept data) are provided in Appendix B.
zyxwvuts
Analysis of generic concept data
The techniques used in the analysis of the generic concept were those
contained in the package: numerical taxonomy system of multivariate statistical
programmes (NTSYS-pc, ver. 1.50) (Rohlf, 1988). The data were standardized
(Sneath & Sokal, 1973) using the STAND (default options) subroutine, resulting
in a mean of zero and a standard deviation of unity for each character. T h e data
were standardized, since, in most instances, nominal-level (non-parametric)
coding was used for the six characters. Standardization of the data would enable
each character to contribute toward the overall resemblance inversely in
proportion to its variability amongst the entire set of OTUs. Correlation and
distance matrices were computed using S I M I N T (for interval data). Cluster
analysis was performed using the SAHN subroutine on a distance matrix.
Ordination of the data, using non-metric multidimensional scaling (MDS), was
used to evaluate the data. The results of a principal components analysis (PCA)
served as the initial matrix for the MDS. A minimum-length spanning tree
(MST), computed using the distance matrix, was superimposed on the final
ordination. The M S T helps to detect local distortions-pairs
of points which
look close together in a plot but actually are far apart if other dimensions are
taken into account.
RESULTS AND OBSERVAL‘IONS
We confirm that three of the morphological characters used by Nordenstam
(1978), to define Senecio, hold true for the sample studied. In addition, the
64
zyxwvuts
zyxwvutsrqpon
zyxwvu
zy
P. L. D. VINCENT AND F. M. GETLIFFE
following characters are also diagnostic at the generic level: the shape of the
anther apex, the length of the filament collars and the shape of the cypselae of
the disc florets.
According to this study, Senecio sensu stricto sensu Vincent is defined on the basis
of the moderate to considerable correlation of one or more character states of the
following six characters:
1. Style arm apices truncate with a subterminal semicircle of spreading
papillae (Character 095, states 9, 11, 13, 14).
2. Anther apices short (Character 103, state 1).
3. Endothecial tissue-inner anticlinal (Character 107, state 6).
4. Filament collars 0.3-0.9 mm long (Character 108, states, 1, 2, 3).
5. Filament collars baluster-form (Character 110, states 1-6, 9).
6. Cypsela (disc) shape-fusiform and cylindrical (Character 117, states 1, 2,
4).
Similarly, the character states which are used to identify senecios which are
peripheral to Senecio sensu stricto sensu Vincent (Table 3), also show moderate to
considerable positive correlation with each other. I t is this correlation amongst
these characters (determined via frequency and ordination analyses of the
MDM) that was used to assign them generic importance. Each of these
characters comprise a number of character states (see Appendix A ) .
These six characters are all 'good' taxonomic characters (Davis & Heywood,
1963) for most of the senecios investigated (Table 1) because (a) they are not
subject to wide variation; (b) they do not appear to be susceptible to
environmental modification; (c) they appear to have a low intrinsic genetic
variability.
A summary of which character states circumscribe Senecio sensu stricto and those
which are used to identify senecios which are peripheral to Senecio sensu stricto is
provided in Table 3.
zyxwvutsrq
Descr$tion of the six characters of generic importance
Gynoecium (disc) : style-arm apices (char 095)
Twelve distinct character states were recorded for this character. The stylearms of the majority of the senecios investigated (74 species and six varieties) are
truncate with a sub-apical semicircle (incomplete ring) of spreading papillae
(state 9). There is a small sub-apical region on the inner (upper) surface of each
style-arm which is devoid of papillae, but appears to be physiologically
receptive. This 'nude' sub-apical region corresponds to the width of the stigmatic
surface. Some representatives of some species have convex to conical apices to the
style-branches as opposed to truncate apices. The apices in all the specimens
examined are distinctly verrucose, this pattern ending abruptly at the semicircle
of spreading papillae.
A number of (13 species) of the remaining senecios have style-arms which are
very similar to those described above but they differ in that the outer surface,
proximal to the semicircle of spreading papillae, is also verrucose, the verrucae
facing distally at c. 45" to the outer surface (state 1 1 ) . The appearance of these
style-arms is distinct from those of character state 9.
The majority of the remaining character states occur only once and are
therefore irrelevant in defining groups.
zyxwvu
zyxwvu
zyxwv
zyxwvuts
zyxwvu
GENERIC CONCEPT O F SENECIO
65
TABLE
3. An analysis of t h e character states of t h e six taxonomically significant characters of
generic importance ( C h a r s 095, 103, 107, 108, 110, 117-see Appendix A ) , with respect to t h e
concept of Senecio sensu stricto sensu Vincent
Status
Character and state(s)
zyxwvutsrqpo
Senecio
sensu stricto
Gynoecium (disc): style-arm apices
(Char 095)
Character states: 9, 1 I , 13 & 14
Yes
No
Character states: 10, 15, 16, 17, 19
Character states 20, 21 and 22 are outside the concept of Senecio sensu Vincent
Androecium (disc): anther apex
(Char 103j
Character state: 1
Yes
No
Character state: 3
Character states 4 and 5 are outside the concept of Senecio sensu Vincent.
Endothecial tissue: cell wall configuration
(Char 107)
Character state: 6
Yes
Character state 10 and 11 are outside the concept of Senecio sensu Vincent.
Senecio
(peripheral)
No
Yes
No
Yes
zyxwvutsr
Filament collars: length
(Char 108)
Character states: 1, 2, 3
Character states: 4, 5
Yes
No
Filament collars: shape
(Char 110)
Yes
Character states: 1 , 2, 3, 4, 5, 6
Character states 7, 8 and 9 are outside the concept of Senecio sensu Vincent.
Cypsela (disc): shape
(Char 117)
Character states: 1 , 2 & 4
Character state: 5
Character state 6 is outside the concept of Senecio sensu Vincent.
Yes
No
Yes
No
Yes
Yes
Yes
Yes
zyxw
Androecium (disc) : anther apex (char 103)
This character refers to the shape of the apical appendage (the sterile
connective extension) of each anther. Four distinct character states were
recorded for this character, three of these (states 1, 3, 4) being found amongst
the senecios and some of the related genera studied. T h e fourth character state
(state 5) is restricted to Gynura diuaricata.
The majority (85 species and six varieties) of the senecios produce relatively
short anther apices (state 1). A further ten senecios produce anther apices that
have the same length: width ratio as those of character state 1, but the anther
apices are approximately twice the size.
Endothecial tissue: cell wall conjguration (char 107)
As has been discussed in a previous paper (Vincent & Getliffe, 1988), the
endothecium of all but two of the taxa studied, exhibited what we have termed
“inner anticlinal” endothecial thickening (state 6 ) . This variant of “radial
66
zyxwvutsr
zyxwvutsrqp
P. L. D. VINCENT AND F. M. GE‘TLIFFE
zyxwvutsrqp
zyxwvutsrqp
Figure 1. Inner anticlinal configuration of endothecium; view of adaxial surface of anther;
longitudinal axis of anther vertical; connective down right hand side. Sample from Senecio speciosus
Willdenow (Hiiliard 1061 N U ) .
zyxwvu
thickening” (Dormer, 1962), has the ribs of thickening restricted to the inner
anticlinal wall of each endothecial cell (i.e. the anticlinal wall nearest the
connective (Fig. 1). This configuration appears to have been illustrated in
Wetter’s account of the endothecium (Wetter, 1983), though he did not
distinguish it from “radial” thickening. Likewise, we believe that Nordenstam
(1978) also illustrated this variant of “radial thickening” without comment.
The exceptions were S. cissumpelinus, in which polarized thickening was found,
and Gynura diuaricata which was conspicuously different, having cells with a
distinct imperforate base plate with rib-like extensions from the edge of the plate.
This configuration appears very similar to that noted in Valeriana capensis Thunb.
(Valerianaceae) (Noel, 1983).
Filament collars: length (char 108)
The filament collar (Koyama, 1967; Drury, 1973, 1975; Nordenstam, 1978)
alternatively called the anther collar (Robinson & Brettell, 1973d; Jeffrey et al.,
1977) is a “downward extension of connective tissue which forms an incomplete
monostromatic collar of thick-walled cells” (Drury, 1975).
zyxwvu
zyxwv
zy
GENERIC CONCEPT OF SENECIO
Figure 2. A ‘cacalioid’ (cylindrical) filament collar. Sample from Senecio medley-woodiz Hutchinson
(Hilliard s.n. NU).
67
zyx
The majority of the senecios investigated have filament collars in the size
range 0.3-0.6 mm, the second most frequent size range being 0.7-0.9 mm.
Nordenstam (1978) has drawn attention to the occurrence of “very
elongated” filament collars in Notonia DC., Notoniopsis B. Nord. and Kleinia Mill.
These “very elongated” filament collars are quite possibly in the size range
1.O-1.4 mm, for the specimens of Kleinia grandzjora and Kleinia nerizfolia included
in the present study both have filament collars in this size range (state 4).
However, both S. tamoides and S. barbertonicus have filament collars which exceed
considerably the lengths of the filament collars of those genera examined by
Nordenstam (1978).
Filament collars: shape (char 110)
The fine structure of the filament collar has only recently attracted the
attention of synantherologists. Koyama (1966, 1967), drawing on the earlier
observations of Kitamura (1937), commented on differences in the filament
collar shape found in senecioid genera and within the sections of Senecio itself.
Drury (1973, 1975), described two types of filament collar: ‘cylindrical’, when
the diameter of the filament collar and the filament proper are the same and the
68
zyxwvutsrqpo
P. L. D. VINCENT AND F. M. GETLIFFE
zyxwv
zyxwv
zyxwvu
zyx
Figure 3. A ‘senecioid’ (balusterform) filament collar. Sample from Senecio speciosus (Hilliard 1061
NU).
zyxwvut
cells of the collar are all the same size (Fig. 2); and ‘balusterform’, when “the
basal cells of the filament collar are enlarged so that the diameter of the collar a t
the proximal end is greater than the filament proper” (Drury, 1975) (Fig. 3 ) .
Nordenstam ( 1978) noted that “cylindrical” filament collars generally
characterize cacalioid genera, but that they were also found in several genera
traditionally associated with Senecio (e.g. Brachyglottis J.R. & G. Forst.,
Nemosenecio (Kitam.) B. Nord., Sinosenecio B. Nord., Tephroseris Reichenb. &
Urostemon B. Nord.) Nordenstam (1978) also drew attention to the term
‘cylindrical’ pointing out that ‘cylindrical’ filament collars are, “strictly
speaking, not cylindrical since they are neither hollow nor circular in transect.
Instead they tend to be somewhat flattened, often with adaxially involute
margins, and more or less reniform in transect”. Since there is a strong
correlation between the ‘cylindrical’ filament collar and several other characters
typical of cacalioid genera, this shape of filament collar is called ‘cacalioid’
(Nordenstam, 1978) and is the term adopted in the present study.
Basally dilated filament collars “are found in Senecio sensu strict0 and in many
more or less senecioid genera, such as Crassocephalum Moench, Culcitium Humb. &
Bonpl., Dendrosenecio (Human ex Hedb.) B. Nord., Dorobaea Cass., Emilia Cass.,
zyxwvu
zy
zyxwvuts
GENERIC CONCEPT OF SENECIO
69
Euryops Cass., Jacmaia B. Nord., Odontocline B. Nord., Othonna B. Nord., Othonna
L., Stilpnogyne DC. and Xyridopsis B. Nord.” (Nordenstam, 1978). Since the
variation in the degree of dilation of these filament collars amongst members of
the Senecioneae, falls under the descriptive term ‘balusterform’, the nondescriptive term ‘senecioid’ is used by Nordenstam (1978) for all filament collars
which are basally distended. ‘Senecioid’ and ‘cacalioid’ filament collars are
produced by the taxa investigated in this study (Table 1). All the ‘senecioid’
filament collars have a distinctly balusterform shape. However, within this
balusterform shape, both narrow and conspicuously broad shapes are found
(states 1-6, 9).
Two distinct narrow balusterform filament collars occur (states 2, 9). I n the
first form (state 2), the cells of the monostromatic collar are virtually
isodiametric and increase in size three to five times towards the proximal end of
the filament collar. In the second form (state 9), the cells of the monostromatic
collar are noticeably elongate and somewhat irregularly shaped, increasing to
approximately twice the size in the proximal end of the filament collar.
The conspicuously broad balusterform filament collars were distinctly
distended into a basal bulge (state 3 ) .
The two cacalioid forms of filament collar recorded (states 7, 8 ) , differ only in
that in the one form (state 7), the width of the filament collar and the filament
proper are the same, while in the second form (state 8), the width of the filament
collar is consistently slightly greater than that of the filament proper.
While most of the taxa investigated possessed one of the several forms of
filament collar in a consistent manner, a few senecios exhibited intraspecific
variation in that they produced senecioid filament collars varying between
narrowly balusterform and noticeably broadly balusterform. This intraspecific
variation is recorded in character states 4-6.
Nordenstam (1978) observed that the cacalioid form of filament collar was
nearly always associated with ‘polarized’ endotheciai tissue and the senecioid
form with ‘radial’ endothecial tissue. A notable exception was the genus
Graphistylis B. Nord. (Nordenstam, 1978).
This association between the senecioid form of filament collar and inner
anticlinal endothecial tissue was also noted amongst the majority of the taxa
studied. However, we recorded nine more exceptions to the aforementioned
association. Senecio cissampelinus possessed senecioid filament collars associated
with an intermediate form of endothecial tissue which is predominantly
‘polarized’ with a limited amount of inner anticlinal endothecial tissue. T h e
remaining eight taxa had cacalioid filament collars associated with inner
anticlinal endothecial tissue. They are: S. medley-woodii, S. tamoides, S. viminalis,
S.fulgens, S. sylvaticus, Gynura divaricata, Kleinia grandgora and Kleinia nerizflia.
Cypsela (disc): shape (char 117)
The predominant shape of the cypsela of the disc florets amongst all the
senecios investigated is ‘fusiform and cylindrical’ (state 2). Four character states
represent the variation in this character. Infraspecific variation was noted with
respect to character states 1 and 2. Where either of these two character states is
used, the choice represents the more frequent occurrence of the two shapes
(states) for the particular taxon concerned.
zy
70
zyxwvuts
zyxwv
zyxwvutsrqpon
P. L. D. VINCENT AND F. M. GETLIFFE
Groupings of the species
The application of the aforementioned generic concept of Senecio to the
senecios investigated, results in the formation of 11 groups and several
ungrouped species of Senecio (Appendix C). Outside this concept are ten species
we regard as peripheral to the genus sensu strict0 (Appendix C ) and the
remaining four species (S. cissampelinus, S. transvaalensis, S. syringzfolius and
S. hockii) proved unequivocally to be excluded from Senecio.
Group I consists of several morpho-biological series namely: A, leafy-stemmed
herbs or shrubs with fpinnately divided, not strongly glandular leaves and
yellow or heterochromous capitula (species 1, 3, 4, 7, 200-204, 210, 212, 213,
2 15, 2 17, 2 19, 220); B, subcuspose herbs with pinnately lobed strongly glandular
leaves and yellow or more often whitish, dull yellow, mauve or purple or blue
homochromous capitula (species var./form 22, 24A, 27, 29, 30, 32, 33B, 34, 35,
38B, 48, 50, 51); C, subscapose herbs with unlobed crenate-serrate fibrous, not
markedly glandular leaves and yellow or purple homochromous capitula (species
63, 64, 65, 67, 73, 77, 81, 88); D, peltate-leaved non-glandular subsucculent
herbs (species 89); E, rhomboid-leaved, non-glandular subsucculent herbs
(species 90). However, these series are often repeated in other groups which may
be analysed as follows:
Group 11: series A (species 15); series F, succulent scandent shrubs
(species 107); series G, subsucculent scandent yellow discoid herbs with cordate
leaves (species 109). [In another phenetic study (Drury, 1967), species 109
clustered with species 11 1 and a species close to 1 16. Species 1 1 1 and 116 are
here grouped among the peripheral senecios.]
Group 111: series H, leafy-stemmed, non-glandular, entire-leaved, yellowheaded homochromous herbs (species 97-99, 102, 103).
Group IV: series B (species 25 & 28), series C (species 85).
Group V: series B (species var. 31, 33A, 42); series C (species 76); series H
(species 104).
Group VI: series A (species 9, 53, 214); series B (species 37, 39); series C
(species 14, 74, 82); series H (species 101).
Group VII: series A (species 2 ) ; series B (species var./form 23, 24B, 24C, 24D,
36, 38A, 40, 56, 57).
Group VIII: series B (species 41, 43, 46).
Group IX: series B (species 47); series H (species 75).
Group X: series B (species 26, 83); series C (species 86).
Group XI: series B (species 55); series C (species 72).
zyxwvu
Ungrouped species
Series A (species 5, 6, 10, 18, 216) [species 216 is known to be capable of
hybridization with species 210 of Group I (Hegi, 1954)l; series F (species 108);
series I scandent non-succulent herbs with cordate leaves (species 110).
Peripheral species
Series A (species 16, 17, 1 18); series J, perennial, succulent, yellow-flowered,
tomentose herbs (species 20); series K, scandent subsucculent cordate-leaved,
yellow, radiate herbs (species 11 1, 116); series F or near it (species 11 7); whiteflowered succulent shrubs referable to Kleinia subgenus Kleinia (species 1 19);
white-flowered succulent herbs currently referred to Senecio section Rowleyana
zyxwvut
zyxwvu
zyxwv
zyxwvuts
GENERIC CONCEPT OF SENECIO
71
Ueffrey, 1986) (species 120); and orange-flowered succulent herbs currently
referred to Kleinia subgenus Notonia (Jeffrey, 1986) (species 122).
The peripheral species share some of the character states associated with
Senecio sensu stricto sensu Vincent suggesting a dendritic or paraphyletic
development of the genus in its broadest sense.
Species excluded f r o m Senecio
Senecio transvaalensis (species 124) shows considerable affinity with the tropical
African genus Emilia Cassini, for it has lilac florets, style-branch apices with a
median fascicle of fused papillae and distinctly 5-angled cypselae. These
characteristics suggest that it is best placed in Emilia, as suggested by Hilliard
(1977, 1978). Since completion of our research, the new combination Emilia
transuaalensis (Bolus) C . Jeffrey (basionym: S. transuaalensis) has been published
(Jeffrey, 1986).
Senecio cissumpelinus (species 121) shows considerable affinity with the tropical
African genus Mikaniopsis Milne-Redhead, its prehensile petioles with their
thickened bases being a notable feature of the vegetative habit of this genus.
However, the genus Mikaniopsis and the Asiatic genus Cissampelo~sis(DC.) Miq.,
have been under study by Jeffrey (C. Jeffrey, personal communication, 1984)
with the view to elucidating the relationship between these two genera. Since the
completion of our research, the new combination, Mikaniopsis cissampelina (DC.)
C. Jeffrey (syn.: S. cissampelinus (DC.) Sch. Bip.) has been published (Jeffrey,
1986).
O n the basis of our analysis, S.syringiflius (species 223) seems to be allied with
Crassocephalum Moench, while Jeffrey ( 1986) regards the position of S. syringiflius
as incertae sedis.
Senecio hockii (species 224) appears to be allied with Emilia Cass., as suggested
by Jeffrey et al. (1977). Subsequently, Jeffrey (1986) published the new
combination Emilia hockii (De Wild. & Muschl.) C. Jeffrey (basionym: 5’. hockii
De Wild. & Muschl.).
zyxwv
zyxwv
zy
Numerical analysis
The concept of numerical taxonomy has progressed from simple comparisons
of similarities and differences to more sophisticated multivariate techniques (for
a review, see Sneath & Sokal, 1973). Numerical techniques can be useful in
understanding the distribution and amount of variation of morphological
characteristics (McNeil, 1979). Methods such as clustering and ordination can
assist in developing taxonomic concepts, but they do not themselves define the
taxonomic boundaries. Each analysis must be carefully interpreted, and can only
serve as a guideline in taxonomic delimitations.
A PCA on the non-metric MDS, overlayed with a minimum spanning tree, of
these six characters (Figs 4, 5), helps to illustrate how these characters
discriminate Senecio sensu stricto from: the peripheral senecios, the senecios which
should be excluded from Senecio and the related genera.
The three factors of the initial PCA (prior to the non-metric MDS), accounted
for 80.6% of the trace (factor 1 : 45.2%, factor 2 : 19.5y0, factor 3 : 15.90/). This
indicates that the three components (Figs 4, 5) are a reasonably accurate
representation of the six characters analysed. Component I has correlations
72
zyxwvutsrqpon
zyxwvuts
zyxwvuts
P. L. D. VINCENT AND F. M. GETLIFFE
zyxwvutsrqponm
zyxwvutsrq
zyxwvutsrqp
zyxwvutsrqpon
-1.0
1.5
0.0
2 .o
1.0
/
121
A
=
8
=
22
16 I 7
I 3 4 7 2 2 2 4 A 27 29 30 32 338 34 35 388
48 50 5 1 63 6 4 65 67 73 17 81 8 8 8'3 90 20(
201 202 203 204 210 2 1 2 213 215 2 1 7 219
220
C
= 9 1 4 1 5 37 39 53 7 4 82 101 107 109 2 1 4
D = 10 91 98 99 102 103
0.7
E
F
G
=
H
=
I
=
=
=
25 28 85
3 1 339 4 2 16 104
2 2 3 2LB 24C 240 36 3 8 1 4 0 56 5 7
4 1 4 3 46
55 72
...
I-
H
!
m
BCO:':
6:?
GH
- 0.1
I J
K
rstr.
[:\
N O N - M o TAXA
22s
PERIPHERAL
\
SENECIOS
L
I
-0.8
-1.6
COMPONENT I
Figures 4,5. The first three principal component axes (rotated) from analysing all the taxa studied
(Table I ) , with respect to the six characters ofgeneric importance (Table 3). Fig. 4. Component I
vs Component 11. Fig. 5. Component I vs Component 111. A minimum-length spanning tree
(MST) is superimposed over the plot in Fig. 5 for non-coincident taxa. The taxa are identified by
their relevant number (Table 1). The boundaries between Senecio sensu stricto sensu Vincent, the
peripheral senecios and non-Senecio taxa, according to the concept of Sen& sensu stricto sensu Vincent,
are demarcated on these two figures.
greater than 0.5 for: STYLE-ARM APICES (CHAR 095) (0.86), ANTHER
APEX (CHAR 103) (0.91), FILAMENT COLLARS: LENGTH (CHAR 108)
(0.64) and FILAMENT COLLARS: SHAPE (CHAR 110) (0.74).
Component I1 has correlations greater than 0.5 for: ENDOTHECIAL
TISSUE: CELL WALL CONFIGURATION (CHAR 107) (0.58),
FILAMENT COLLARS: LENGTH (CHAR 108) (0.53) and CYPSELA
(DISC): SHAPE (CHAR 117) (0.67). Component I11 has correlations greater
than 0.5 for: ENDOTHECIAL TISSUE: CELL WALL CONFIGURATION
zyxwvu
zyxwvu
zyxwv
zyxwvut
GENERIC CONCEPT OF SENECIO
-1.0
0.0
1.0
73
2.0
1.5
0.7
zyxwvuts
zyxwvutsrqpon
-
I
I-
H
31
a
0
6
v
-0.1
- 0.8
\
A
8
C
D
=
I
=
F
=
=
=
=
G =
H =
I
=
J =
K =
L =
zyxwvutsrqpon
16 1 7
1 1 8 120
1 5 55 72 1 0 9
1 3 4 7 22 24A 27 29 30 3 2 338 34 35 388
4 8 50 5 1 6 3 6 4 65 6 7 7 3 7 7 81 88 8 9 90
200 2 0 1 202 203 204 210 2 1 2 213 215 217
2 1 9 220
9 1 4 3 7 39 5 3 14 82 I01 214
2 6 8 3 86 97 9 8 99 102 103
4 7 75
2 5 28 85
2 2 8 229
2 23 248 24C 240 36 38A 40 56 57
41 4 3 46 108
3 1 331 42 76 104
- 1.6
(OM
IENT
1
Figure 5
(CHAR 107) (0.61) and CYPSELA (DISC): SHAPE (CHAR 117) (0.72).
Since the components of a PCA are arranged in numerical order of importance,
the composition of the three components of this PCA indicates, to a degree, the
relative importance of each of the six characters with respect to each other. It is
noted from Figs 4 and 5 that the boundary between Senecio sensu stricto (sensu
Vincent) and the non-Senecio taxa is relatively distinct. While the affinity of some
of the peripheral senecios is indistinct and the affinities of the peripheral senecios
as a whole appears dendritic.
DISCUSSION
While Barkley (1983, 1985) argues for the retention of Senecio as a broad
concept, the implications of this in the African context are biologically
74
zyxwvutsrqpo
zyx
P. L. D. VINCENT AND F. M. GETLIFFE
untenable. As Jeffrey (1986) points out, the only means of retaining Senecio as a
broad concept for the African species in an other than paraphyletic manner
would be to include all of the African Senecioneae within Senecio. This would
render meaningless the biological and other information associated with such
names as Crassocephalum, Gynura and a number of other members of the African
Senecioneae. Consequently segregation seems to be the correct course to follow,
primarily on biological grounds. An implication of this choice is that Senecio will
probably remain a paraphyletic “core genus” as criticized by Funk (1983).
Nevertheless, while Funk (1983) contends that the segregates must be defined by
synapomorphies, an evolutionary situation can be imagined in which a
monophyletic core group of species exists which is defined only by those
synapomorphies which are common to the whole subtribe Senecioniinae Ueffrey,
1986). While in a cladistic analysis such a core group would appear
paraphyletic, Jeffrey (1986) argues that it would be falsely so even though there
would be no way of verifying that falsity.
CONCLUSION
The generic concept of Senecio sensu Vincent delimits a natural assemblage.
While this concept also supports many of the suggestions of Jeffrey et al. (1977)
and Jeffrey (1979)) it also indicates that the concept of Senecio held by these
authors could still be too broad. T h e concept also supports some of the
suggestions by Nordenstam (1978) on characters he has perceived as being
characteristic of Senecio sensu stricto and elaborates upon them.
It is obvious that the concept needs to be tested on further representatives of
Senecio from many more areas of its cosmopolitan distribution, before any of the
consequences resulting from the application of this concept are put forward in
new taxonomic delimitations. Nevertheless, it is encouraging to see that the
application of this concept appears to be meeting the need of circumscribing the
assemblage of Senecio in Natal into a more homogenous group. Further phenetic
studies are currently underway on the worldwide Senecio flora and
representatives of related genera in the Senecioneae, to test the concept of Senecio
sensu stricto here presented.
Perhaps these further studies will indicate that Senecio sensu stricto sensu Vincent
is a paraphyletic “core genus” as criticized by Funk (1983) and as suggested by
Jeffrey (1986). However, it must be remembered that a taxonomic system in
practice will always remain a hypothetical reflection of phylogeny.
ACKNOWLEDGEMENTS
The authors acknowledge financial support from the CSIR and from the
University of Natal, while the senior author was a postgraduate student of the
latter. Funding from the University of the Witwatersrand is also gratefully
acknowledged. The use of the research facilities of the Department of Botany,
University of Natal, Pietermaritzburg and the University of the Witwatersrand
is also much appreciated. The r61e played by Prof. 0. M. Hilliard in introducing
the senior author to the very exciting field of synantherology, particularly studies
in Senecioneae, is very gratefully acknowledged.
zyxwvu
zyxw
zyxwvutsr
zyxw
zyxwvutsrqp
zyxwvut
zyxwv
zyxwvutsrq
GENERIC CONCEPT O F SENECIO
75
REFERENCES
BARKLEY, T. M., 1983. Generic concepts in the Senecioneae. American Journal ofBotany, 70f.5): 99.
BARKLEY, T. M., 1985. Generic boundaries in the Senecioneae. Taxon, 34: 17-21.
BOHLMANN, F.,ZDERO, C., BERGER, D., SUWITA, A., MAHANTA, P. &JEFFREY, C., 1979. Neue
Furanoeremophilane und Weitere Inhaltsstoffe aus Sudafrikanischen Senecio-arten.Phytochemistry, 18: 79-93.
DAVIS, P. H. & HEYWOOD, V. H., 1963. Principles of Angiosoperm ?axonomy, New York: D. Van Nostrand
Company, Inc.
DORMER, K. J., 1962. 'The fibrous layer in the anthers of Compositae. New Pbtologist, 61: 150-153.
DRURY, D. G., 1967. A taxonomic study ofCompositae with special reference to Senecio. Unpublished Ph.D Thesis of
the University of Southampton.
DRURY, D. G., 1973. Annotated key to the New Zealand shrubby Senecioneae-Compositae and their wild
garden hybrids. N e m zealand Journal of Botany, 11: 731-784.
DRURY, D. G., 1975. A comparison of Senecio kirkii (New Zealand) and Senecio insularis (Lord Howe Island)
with senecios endemic to the island of St. Helena. N e w zealand Journal of Botany, 13: 769-780.
FARR, E. R., LEUSSINK, J. A. & STAFLEU, F. A. (Eds), 1979. Index Nominum Genericorum (Plantarum), 3.
The Hague: Dr W. Junk b.v.
FUNK, V. A,, 1983. Cladistics and generic concepts in the Cornpositae. American Journal ofBotany, 70(5): 99.
HARVEY, W. H., 1865. Senecio. In W. H. Harvey & 0. W. Sonder, Flora Capmsis, 3: 346408.
HEGI, G., 1954. Illustrierte Flora von Mitteleuropa, 6: 795.
HILLIARD, 0. M., 1977. Compositae in Natal. Pietermaritzburg: University of Natal Press.
HILLIARD, 0. M., 1978. The geographical distribution of Compositae native to Natal. Notes. Royal Botanic
Garden (Edinburgh), 36: 4077425.
JEFFREY, C., 1979. Generic and sectional limits in Senecio (Compositae): 11. Evaluation of some recent
studies. Kew Bulletin, 34: 49-58.
JEFFREY, C., 1986. The Senecioneae in East Tropical Africa. Notes on Compositae: IV. Kew Bulletin, 41:
873-943.
JEFFREY, C. & CHEN, Y. L., 1984. Taxonomic studies on the tribe Senecioneae (Compositae) of Eastern
Asia. Kew Bulletin, 39: 205-454.
JEFFREY, C., HALLIDAY, P., WILMOT-DEAR, M. &JONES, S. W., 1977. Generic and sectional limits
in Senecio (Compositae): I. Progress Report. Kew Bulletin, 32: 47-67.
KING, R. M . & ROBINSON, H., 1970. The new synantherology. Taxon, 19: 6-1 1.
KITAMURA, S., 1937. Les Senecio du Japon. Acta Phytotaxonomica et Geobotanica, 6: 265-275.
KOYAMA, H., 1966. O n the section Nemosenecio of the genus Senecio. Acta Phytotaxonomica et Geobotanica, 22:
15-20.
KOYAMA, H., 1967. Taxonomic studies on the tribe Senecioneae of Eastern Asia. I. General part. Memoirs.
College of Science. Kyoto University, Series B, 33: 181-209.
LINNAEUS, C., 1753. Species Plantarum. Stockholm.
LINNAEUS, C., 1754. Genera Plantarum, ed. 5. Stockholm.
McNEILL, J., 1979. Purposeful phenetics. Systematic ,zbology, 28: 465-482.
NIE, N. H., HADLAI-HULL, C., JENKINS, J. G., BREMMER, K. S. & BENT, D. H., 1975. Statistical
Package f o r the Social Sciences, 2nd edn. New York: McGraw Hill.
NOEL, A. R . A,, 1983. The endothecium-a neglected criterion in taxonomy and phylogeny? Bothalia, f 4 :
833-838.
NORDENSTAM, B., 1978. Taxonomic studies in the tribe Senecioneae (Compositae). Opera Botanica, 44:
1-84.
ROBINS, D. J., 1977a. Alkaloids of the Compositae. Compositae Nemsletter, 5: 1-1 1.
ROBINS, D. J., 1977b. Senecioneae-chemical review. In V. H. Heywood., J. B. Harborne & B. L. Turner
(Eds), The Biology and Chemistv of the Compositae, 2: 831-850. London: Academic Press.
ROBINSON, H., 1974. Studies in the Senecioneae (Asteraceae). VI. The genus Arnoglossum. Phytologia, 28:
294295.
ROBINSON, H., 1975. Studies in the Senecioneae (Asteraceae). V I I . Additions to the genus Roldana.
Phytologia, 32: 33 1-332.
ROBINSON, H. & BRETTELL, R. D., 1973a. Studies in the Senecioneae (Asteraceae). I. A new genus,
Pittocaulon. Phytologia, 26: 45 1 4 5 3 .
ROBINSON, H. & BRETTELL, R . D., 1973b. Studies in the Senecioneae (Asteraceae). 11. A new genus,
Nelsonianthus. Phytologia, 27: 53-54.
ROBINSON, H. & BRETTELL, R. D., 1973c. Studies in the Senecioneae (Asteraceae). 111. The genus
Psacalium. Phytologia, 27: 254-264.
ROBINSON, H. & BRETTELL, R . D., 1973d. Studies in the Senecioneae (Asteraceae). IV. The genera
Mesadenia, Syneilesis, Miracacalia, Koyamacacalia and Sinacalia. Pbtologia, 27: 265-276.
ROBINSON, H. & BRETTELL, R., 1974. Studies in the Senecioneae (Asteraceae). V. The genera
Psacaliopsis, Barkltyanthus, 7elanthophora and Roldana. Phytologia, 27: 402-439.
ROHLF, F. J., 1988. NTSYS-pc: Numerical 7axonomy and Multivariate Analysis System. New York: Applied
Biostatistics Inc.
zyx
76
zyxwvutsrqp
zyxwvutsrq
zyxwvuts
zy
P. L. D. VINCENT AND F. M. GETLIFFE
SNEATH, P, H. A. & SOKAL, R. R., 1973. Numerical Taxonomy. San Francisco: W. H. Freeman & Co.
VINCENT, P. L. D. & GETLIFFE, F. M., 1988. The endothecium in Senecio (Asteraceae).Botanical Journal of
the Linnean Society, 97: 63-7 1 .
WETTER, M. A,, 1983. Micromorphological characters and generic delimitation of some new world
Senecioneae (Asteraceae). Briltonia, 35(1): 1-22.
APPENDIX A. CHARACTERS AND CHARACTER STATES
Characters and character states of the six characters of generic importance.
The numbers in the left-hand margin or adjacent to illustrations correspond to
the character state codes. The character state codes for each taxon studied
(Table l ) , for the six characters of generic importance, are provided in
Appendix B.
The numerical sequence of the character state codes, within each character, is
incidental (unweighted).
Gynoecium (disc): style-arm apices (Char 095)
1
zyxwvu
zyxwvuts
GENERIC CONCEPT OF SENECIO
Androecium (disc): anther apex (Char 103)
77
zyxwvutsrqpo
Endothecial tissue: cell wall configuration (Char 107)
6. Inner anticlinal configuration-marked peg-like pattern, sometimes with a median ‘T’ shaped process on
some of the inner anticlinal ribs (see Vincent & Getliffe, 1988).
10. Polarized endothecial thickening and a little inner anticlinal thickening (ribs) o n some cells.
11. Imperforate base plate with a number of rib-like extensions from both anticlinal edges of the base plate.
Filament collars: length (Char 108)
1. 0.2 mm or less
2. 0.3-0.6 mm
3. 0.7-0.9 mm
4. 1.0-1.4 mm
5. 1.6-2.3 mm
Filament collaw: shape (Char 110)
zyxwvutsrqponm
zyxwv
PREOOM. TYPE 2
B U T A L S O TYPE 1
0
PREOOM. T Y P E 1
B U T ALSO T Y P E 2
PREOOM. T Y P E 1
B U T ALSO
TYPE
@
3
7n
zyxwvutsrqpon
zyxwvutsrqpo
P. L. D. VINCENT AND F. M . GETLIFFE
Cypsela (Disc): shape (Char 117)
I . Oblong and cylindrical.
2. Fusiform and cylindrical.
4. Narrowly turbinate.
5. Cylindric and 5-angled.
6. Obovate, laterally compressed, margins winged
APPENDIX B. MASTER DATA MATRIX ( M D M )
zyxw
zyxwvutsrqponmlk
zyxwvutsrq
zyxwvutsrq
Master data of the six characters of generic importance for all the taxa investigated.
The numerical values in this matrix correspond to the relevant character states of the
character concerned. See Appendix A for a description/illustration of the character states of the
six characters of generic importance.
095
103
Characters
107
ion
110
117
I
2
1
2
2
2
2
2
2
2
2
2
2
2
4
2
2
~~
1.
2.
3.
4.
5.
6.
7.
9.
10.
14.
15.
16.
17.
in.
20.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
S. madagascariensis
S. skirrhodon
S. inaequidens
S.harueianus
S.polyanthemoides
S. pterophorus
S. juniperinus
S. serratuloides
S. microglossus
S. pandurnf rmis
S. achilleijolius
S. tanacetopsis
S.seminiveus
S.haygarthii
S. medlg-woodii
S. purpureus
S. gerrardii
9
9
9
9
9
9
9
9
9
9
II
17
17
13
1
1
1
1
1
I
1
1
1
1
1
3
3
1
14
9
9
1
9
1
9
9
9
9
9
9
1
1
1
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
2
2
2
2
6
6
6
6
6
6
6
6
6
6
6
6
2
2
2
2
2
3
2
2
2
2
2
2
6
6
6
6
6
6
2
2
2
2
2
2
1
1
2
2
2
2
2
2
2
2
3
2
2
I
1
1
1
6
1
2
3
2
1
1
1
1
7
1
1
1
S. erubescens
(24A) var. erubescens
(24B) var. crepidijoliur
(24C) var. incisus
(24D) var. dichotomus
S. sandersonii
S. glanduloso-lanosus
S. uariabilis
S. umgeniensis
S.glanduloso-pilosus
S. subcoriaceus
S. cathcartensis
S. ngoyanus
S. po!yodon
(33A) var. polyodon
(33B) var. subglaber
S. sp. aff. S. speciosus
S.speciosus
S. macrocephalus
S. barbatus
S. poseideonis
(38A) radiate forms
(38B) discoid forms
S. aff. poseideonis
S. arabidijolius
S. rhyncholaenus
S. svbrubrQ7orus
S. hastatus
9
1
1
1
1
1
1
9
9
9
9
1
1
1
9
9
9
9
9
9
1
9
9
9
9
9
9
9
I
1
1
1
1
1
1
2
2
2
2
2
2
2
1
1
I
1
4
2
1
4
1
1
5
1
5
1
1
1
1
2
1
2
1
1
1
2
2
2
2
2
2
2
2
2
2
2
2
1
2
1
zyxwvutsr
1
1
1
1
I
1
1
2
1
2
5
2
2
2
1
1
2
1
zyxwvu
zyxwv
zyxwvutsrqp
zyxwvutsrq
zyxwvutsr
GENERIC CONCEPT OF SENECIO
79
APPENDIX B.-continued
Charactrrs
095
103
107
108
I10
2
2
2
1
117
~~
46.
47.
48.
50.
51.
53.
55.
56.
57.
63.
64.
65.
67.
72.
73.
74.
75.
76.
77.
81.
82.
83.
85.
86.
88.
89.
90.
97.
98.
99.
101.
102.
103.
104.
107.
108.
109.
110.
111.
116.
117.
118.
119.
120.
121.
122.
124.
200.
201.
202.
203.
204.
210.
212.
213.
214.
215.
216.
S.hirsutilobus
S. ingeliensis
S. natalicola
S.hieraciaides
S. consanguineus
S. chrysacoma
S. asperulus
S.hypochoerideus
S. mooreanus
S.paludajinis
S.ajinis
S. lydenburgensis
S.anomalochrous
S. breuidentatus
S.praelerihs
S. coronatus
S. macrospermus
S. dregeanus
S. discodregeanus
S.caudutus
S.mauricei
S. saniensis
S. heliopsis
S. albanensis
(var. doranic$ororus)
S. inornalus
S. oxyriijolius
S. rhomboideus
S. bupleuroides
S. urophyllus
S.scifus
S.glaberrimus
S. latlfolius
S. retrorsus
S.olhonniflorus
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
11
I1
11
9
11
1
1
1
I
I
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
I
1
1
1
1
1
1
1
1
1
1
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
10
6
6
6
6
6
6
6
6
6
6
6
6
6
3
2
2
2
2
3
2
2
2
2
2
2
1
1
2
I
1
1
1
3
2
1
1
1
1
1
2
2
3
2
5
2
2
2
2
3
2
1
I
2
2
4
3
2
2
2
2
2
2
2
2
2
2
2
2
2
2
5
2
1
1
1
2
2
2
2
2
2
5
1
4
2
2
2
i
1
2
2
2
2
2
1
1
2
2
2
2
2
2
2
1
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
1
2
2
2
2
2
2
2
2
2
1
zyxwvutsr
zyxwvutsrq
S. brachypadus
S. pleistocephalus
S. mikanioides
S. deltoideus
S. tamoides
S.helminlhioides
S. barbertonicus
S.brevilorus
S. uiminalis
S. radicans
S.cissampelinus
S. fulgens
S. transuaalensis
S. grandiJlorus
S. cakilefolius
S. arenarius
S. multibracteatus
S.elegans
S. uulgaris
S. uiscosus
S. gallicus
S. c a l i j h c u s
S. uernalis
S.syluaticus
11
9
I1
II
11
14
I1
11
I1
10
15
9
11
16
19
9
9
9
9
9
9
9
9
9
9
9
1
I
1
1
1
3
3
3
3
3
3
3
3
4
1
1
1
1
1
I
I
1
1
1
1
5
3
4
4
2
4
2
2
2
2
2
2
2
2
2
2
2
2
1
1
1
n
n
2
1
8
3
1
1
I
1
1
1
I
1
2
1
8
5
2
2
2
2
2
2
2
2
2
2
2
zyxwvutsrq
80
zyxwvutsrqpon
zyxwvut
zyxwvu
P. L. D. VINCENT AND F. M. GETLIFFE
zyx
zyxw
zyxwvutsr
APPENDIX B.-continued
~
095
21 7.
219.
220.
223.
224.
225.
226.
227.
228.
229.
232.
S.desfontainei
S.erechtitoides
S. lautus
103
9
9
9
14
19
20
14
21
22
22
9
S. hockii
S.syringifolius
Crassocephalum cernuum
Emilia Jammea
Gynura diuaricata
Notonia grandiJora
Kleinia neriijolia
Cineraria geijolia
1
1
1
3
3
4
4
5
3
3
1
Characters
107
108
6
6
6
6
6
6
6
11
6
6
6
2
2
2
3
2
3
2
2
4
4
2
110
117
1
1
1
2
2
2
2
2
2
2
2
1
1
6
2
1
9
2
8
8
8
1
APPENDIX C
Complete list of the species and varieties of Senecio included in this study (Table 1) arranged according to the
new concept of Serrecio wzsu slricto sensu Vincent, together with an indication of which species are excluded from
Senecio sensu stricto Jensu Vincent. T h e I 1 groups and ungrouped species of Senecio, together with the peripheral
and excluded senecios are distinguished. ’lhe listing of the character states after each group follows the order of
the characters in Table 3. Taxa are identified by their respective number used in Table 1.
Senecio sensu stricto senm Vincent
Group 1:
Taxa:
Character states: 9, 1, 6, 2, I , 2
I , 3, 4, 7, 22, 24A, 27, 29, 30, 32, 33B, 34, 35, 38B, 48, 50, 51, 63, 64, 65, 67, 73, 77, 81, 88,
89, 90, 200, 201, 202, 203, 204, 210, 212, 213, 215, 217, 219, 220.
Group 11:
Taxa:
Character states: 11, 1, 6, 2, 1, 2
15, 107, 109.
Group 111:
Taxa:
Character states: 11, 1, 6, 2, 2, 2
97, 98, 99, 102, 103.
Group IV:
Taxa:
Character states: 9, 1, 6, 2, 4, 2
25, 28, 85.
Group V:
Taxa:
Character states: 9, 1, 6, 2, 5, 2
31, 33A, 42, 76, 104.
Group V1:
Taxa:
Character states: 9, I , 6, 2, 2, 2
9, 14, 37, 39, 53, 74, 82, 101: 214
Group VII:
‘Taxa:
Character states: 9, 1, 6, 2, 1, 1
2, 23, 24(A-D), 36, 38A, 40, 56, 57
Group V I I I : Character states: 9, 1, 6, 2, 2, 1
Taxa:
41, 43, 46.
Group IX:
Taxa:
Character states: 9, 1, 6, 3, 2, I
47, 75.
Group X:
‘l’axa:
Character states: 9, 1, 6, 3, 2, 2
26, 83, 86.
Group XI:
I axa:
Character states: 9, I , 6, 3, 1, 2
55, 72.
r
.
zyxwvutsrq
zyxwvutsrq
zyxwvuts
zyxwvu
GENERIC C O N C E P T OF SENECIO
APPENDIX C.-continued
Ungrouped species of Senecio sensu stricto sensu Vincent
5. Character states: 9, 1, 6, 1, 1, 2
6. Character states: 9, 1, 6, 1, 6, 2
10. Character states: 9, 1, 6, 2, 3, 2
18. Character states: 13, 1, 6, 2, 1, 4
108. Character states: 11, 1, 6, 2, 1, I
110. Character states: 14, 1, 6, 2, 1, 2
216. Character states: 9, 1, 6, 2, 8, 2
Peripheral senecios
16. Character states:
17. Character states:
20. Character states:
1 11. Character states:
116. Character states:
117. Character states:
118. Character states:
119. Character states:
120. Character states:
122. Character states:
Species excluded from
121. Character states:
124. Character states:
223. Character states:
224. Character states:
3, 6, 2, 1, 2
3, 6, 2, 1, 2
1 , 6, 3, 7, 2
3, 6, 5, 8, 2
3, 6, 4, 2, 2
3, 6, 5, 2, 2
3, 6, 3, 2, 2
15, 3, 6, 4, 8, 2
9, 3, 6, 4, 2, 2
16, 3, 6, 4, 8, 1
17,
17,
14,
11,
11,
11,
10,
Senecio
11, 3,
19, 4,
14, 3,
19, 3,
10, 2, 1, 2 ( = Mikaniopsis cissampelina (DC.) C. Jeffrey)
6, 2, 3, 5 ( = Emilia tmmzrvaalensis (Bolus) C. Jeffrey)
6, 3, 2, 2 ( = S. syringifolius 0. Hoffm. incertae sedir)
6, 2, 1, 2 ( = Emilia hockii (De Wild. & Muschl.) C. Jeffrey)
81