~ ,) .I Classification and Evolutionary history in Cyperaceae. To w n SEDGES ap INFLORESCENCES e """'- SPIKELETs rs ity of C SEDGE FLOWER. Robert Skelton U ni ve I I I I I I I I II I I :I I I I .·I .I I Botany Honours Systematics 2007 Supervisor: Dr. A.M. Muasya PERIGYNWM n of C ap e To w The copyright of this thesis vests in the author. No quotation from it or information derived from it is to be published without full acknowledgement of the source. The thesis is to be used for private study or noncommercial research purposes only. U ni ve rs ity Published by the University of Cape Town (UCT) in terms of the non-exclusive license granted to UCT by the author. -.1 I I I I I I I I I I I I I I I I I I I I - BOLUS LIBRARY C24 0008 5101 Abstract: (An analysis of I07 taxa reps~' ~cL, 'l rr -. ~ ~ , IIIIIIIIIIIIIIIIIIIIL. · JP 1;,~ trnL-F (intron and intergenic spacer) and rps16 intron is presented.@ecies f were sequenced for the first time in the rps16 intron region of the plastid genome in an attempt to further resolve the phylogenetic relationships within Cyperaceae. Cyperaceae are monophyletic and resolved into two subfamilies, the Mapanioideae and Cyperoideae. The overall topology is similar to those observed in other studies, with a few exceptions. Our study suggests that Cladium should be treated as a separate tribe, the Cladieae. Sclerieae is observed as sister to Schoeneae. Fuirena is observed here_a.s sister to either ofFuirena~d are J:/ '1}-l- _ 85 genera from !4 tribes, sequenced for tl)""'tid Eleocharideae or Abildgaardieae whereas the rest ." ~. closer to Cypereae. Our robust phylogeny allowed us to conduct phylogenetic optimizations of habitat preferences and adaptive character states, which enabled us to investigate the evolutionary history of the family. Cyperaceae most probably evolved in permanently moist habitats in tropical regions. The Cyperaceae have since radiated into a number of other habitats, notably open grassland, woodland and savannah. Diversification into these habitats was likely to have been influenced or facilitated by morphological features such as a change in the number of functional stamens or a switch to an annual life history. 1 ~ .... ~. I I I I I I I I I I I I I I I I I I I I I Introduction to Cyperaceae: The Cyperaceae is the third largest monocotyledonous family, after Orchidaceae and Poaceae. It contains 109 genera and approximately 5500 species (Muasya et al. in press). The Cyperaceae have an almost cosmopolitan distribution only being absent from Antarctica (Goetghebeur 1998). The species occur in a variety ofhabitats, although they are mostly dominant in many wetland ecosystems (Muasya et al. in press and Simpson et al. 2003). About 35% of the genera are monotypic, 26% have two to five species while 6% of the genera have over 200 species (Muasya et al. in press). Carex (1757 species) and Cyperus (686 species) are the largest genera (Muasya et al. in press). / Classification of the Cyperaceae ~:? r!l~.f Q.•' · "~.:o t~L · Historically family-level phylogenetic studies of the Cyperaceae have largely made use of morphological data (Muasya eta!.~ press). The determination of relationships within Cyperaceae on morphological grounds is difficult, largely due to the presence of highly reduced flowers and condensed inflorescences (Muasya et al. in press). In particular subfamily and tribal delimitation has been particularly inconsistent (Muasya et al. in press). For example Goetghebeur (1998) recognized four subfamilies: the Sclerioideae, Mapanioideae, Caricoideae and Cyperoideae. Bruhl (1995) on the other hand only recognized two subfamilies: the Caricoideae and the Cyperoideae. The two studies also differed in their tribal circumscription, as Bruhl (1995) recognized 12 tribes, while Goetghebeur (1998) recognized 14 tribes. Over the last 10 years a number of molecular studies have improved our knowledge of !(,.:;. the phylogeny of the family (Muasya et al. in press). DNA sequencing is assisting tremendously in building a b~ understanding of intra-familial relationships of the Cyperaceae (Simpson et al. 2003). The first family-wide DNA based phylogenetic tree was conducted by Muasya et al. (1998) using plastid DNA from the rbcL gene region. Subsequent broad suprageneric studies have so far sampled all subfamilies and tribes, although sampling effort is not evenly distributed among all tribes (Muasya et al. in press). Family-level studies have been based mainly on rbcL sequence data, which is a relatively slowly evolving plastid marker (Muasya et al. in press). Other plastid and 2 I I I I I I I I I I I I I I I I I I I I I nuclear regions have been used in tribal or subfamiliallevel studies (Muasya et al. in press). Studies of the subfamily Mapanioideae (e.g. Simpson et al. 2003) and a number of other tribes or genera have been based mainly upon the plastid regions rps 16 intron, trnL intron and trnl-F intergenic spacer (e.g. Verboom 2006). Slowly evolving plastid regions, such as rbcL, provide less resolution at the tribal or generic level than more rapidly evolving markers, such as rps16 intron, although they are more easily aligned across a family (Muasya et al. in press). The choice of a slowly evolving marker in the majority of studies of Cyperaceae may account for poor resolution of the relationship between a number of clades, such as Fuireneae and Schoeneae (Muasya et al. in press). Further, few studies have combined multiple molecular marker data sets or molecular and morphological data sets (e.g. Muasya et al. 2000a). One potential solution to poor resolution is to combine data from the same DNA regions (e.g. rbcL, trnl-F and rps16) for similar taxa (Muasya et al. in press). Muasya et al. (in press) also suggested that greater sampling effort is required to elucidate the phylogenetic relationships within ') certain trdbes, such as Cryptangineae, Bisboeckelerieae, Fuireneae, Schoeneae and Sclerieae (Muasya et al. in press). In Fuireneae there is a need for greater resolution of the relationship between the Schoenoplectus and Actinoschoenus and the Schoenoplectiella groups, which have been sh()wft~"lo (Muasya et al. in press). The relationship ben*~ G:YJ!,~schoenu Gay~!is to DidynzL~!:rum and reqlifres further investigation (Mua~yet ' I and ;heRyl(c~spo a polytomy with Cypereae ,, four clai:les (Cladium, al. in press). N\l~ is unresolved dV:J..JJ· 'rft' \rjv)~, [\v '"· ( -\'it''\ The evolutionary history of Cyperaceae A robust phylogeny allows one to gain an understanding of the evolutionary history of morphological characters and ecological states of a taxon (e.g. Linder and Rudall 2005). Linder and Rudall (2005) conducted phylogenetic optimizations of habitat preferences and adaptive character states, which enabled them to investigate the factors that have influenced the evolution of the Poales. They included a number of morphological characters, such as the growth form, type of rooting system, the thickness of trichomes, photosynthetic pathway and the presence or absence of annual or perennial life history 3 '""¢. I I I I I I I I I I I I I I I I I I I I I (Linder and Rudall2005). Their study included a single terminal for the Cyperaceae and thus a more extensive sampling effort is required to further evaluate the evolutionary history of the Cyperaceae. There are a number of interesting and informative characters that one could include in an evaluation of the evolutionary history of the Cyperaceae (Goetghebeur 1998). The Cyperaceous flower is generally a bisexual trimerous structure with 3 + 3 perianth parts, 3 (+ 3) stamens and a 3-carpellate pistil with a single, basal, anatropous ovule, although numerous variations occur (Goetghebeur 1998). The number of stamens varies from 0 (in female flowers) to 1-2-3-6 or more (Goetghebeur 1998). Other traits relevant to the Cyperaceae include the structure of inflorescence, the sexuality of the flower- is it unisexual (dioecious) or bisexual (monoecious)- and the presence or absence of a perianth (Simpson et al. 2003). Goetghebeur (1998) reported that the perianth structure is extremely variable in Cyperaceae. Bruhl and Wilson (in press) provide an up to date review of the distribution ofC3 and C4 photosynthetic pathways in Cyperaceae, the latter of which has bee{fuggested to have evolved"''llultiple times independently. Among the studied I 07 species w~th€3 gen~! and~er(Buli Jre consistently and W~1 c{!!Jere consistently c, and a rekcln~id <\~_ m press). "' \\) v Simpson eta/. (2003) investigated pollen development in the Mapanioideae and a few related taxa to infer pollen type. They reported finding at least two-types of pollen, the Mapanioid-type and "pseudomonad" pollen (Simpson eta/. 2003). Simpson eta/. (2003) also investigated the distribution of each pollen type within the family and found that the Mapanioid type is restricted to a few taxa in the tribe Hypolytreae, while the rest of the Cyperaceae had the pseudomonad type. The evolutionary significance of each type is not yet fully understood but there are some indications that pseudomonad pollen might be an adaptation to wind-pollination, as the shape is fairly streamlined. Simpson eta/. (2003) reported that the Mapanioid pollen is stickier than pseudomonad pollen, which together with floral morphology and habit indicate that Mapanioid pollen may be an adaptation to animal dispersal. While most of the Cyperaceae is adapted to anemophily, a few species are known to be visited by pollen-gathering insects or have traits indicating potential 4 I I I I I I I I I I I I I I I I I I I I I pollination by insects (Simpson et a!. 2003 and Goetghebeur 1998). It has even been suggested that entomophily may be linked to the forest habitat, as wind velocity is much reduced and alternative pollen vectors may be required (Goetghebeur 1998). In an investigation of the evolutionary history of the Poales, Linder and Rudall (2005) included habitat ecology as one of the characters of interest. They reported that it is likely that ancestral Poales grew in marshy or wet habitats, although the extant groups of Poales still found in this habitat are mostly species poor (Linder and Rudall2005). Species of a few genera, such as Prionium can still be found to dominate wetland habitats and most of the sedge families include at least a few species that prefer wetland habitats (Linder and Rudall 2005). Despite this it is uncertain whether these species represent reversals or the ancestral state within Cyperaceae (Linder and Rudall 2005). The majority of Cyperaceae species are known to have a perennial life history, while the annual species are restricted to a limited number of genera (Goetghebeur 1998). Despite this no study has investigated the likely state of the ancestral sedge. Aims and objectives: The aim of this study is to reconstruct a robust phylogeny of the family Cyperaceae, using three of the most commonly used plastid gene regions (rbcL, trnl-F and rps16 intron) and morphological characters. Subsequently we will use this phylogeny to review the current suprageneric classification of the family. In particular we will attempt to provide greater resolution to the relationship between a number of poorly resolved clades, such as Cryptangineae, Bisboeckelerieae, Fuireneae, Schoeneae and Sclerieae by providing sequences of previously unsequenced ~ewly will be added to sequence data lodged in GenBank the same plastid gene regions. sequenced taxa i?om~pxevus . \ ~ studies that have used ~- A second objective is use the robust phylogeny to reconstruct the evolutionary pattern of key morphological and ecological characters. We aim to conduct character optimization analyses which will allow us to infer the likelihood of each state at specific nodes. This 5 I I I I ·I I I I I I I I I I I I I I I I I will allow us to evaluate what the ancestral state was likely to be and to trace the evolution of particular traits within and between clades. Summary of research questions: 1) Are we able to provide greater resolution to poorly resolved nodes? 2) If we are able to provide greater resolution, is the current classification of the Cyperaceae in need of revision? 3) Can we trace the evolutionary history of specific characters within Cyperaceae? 4) Are we able to infer the likely state of the ancestral sedge using our phylogeny? 6 I I I I I I I I I I I I I I I I I, I I I I Methods: A total of 107 taxa were included in this study. They consist oftwo outgroups and 105 Cyperaceae species. The outgroup taxa were a species from the Juncaceae, Juncus - -1. effusus L. and a species from the Thiuniaceae, Prionium serratum Drege (Muasya et al. in press). The 105 species ofCyperaceae were taken from 85 genera from the 14 tribes and four subfamilies recognised by Goetghebeur (1998). The complete table of taxa, together with the type of DNA material used in the analysis, can be seen in appendix 1. DNA extraction and PCR: Total DNA was extracted from 0.02-0.08g silica dried samples or from 0.05-0.lg fresh ·vegetative (leaves or culm) material collected in the field or from herbarium specimens. DNA was also obtained from stock held at Kew Gardens. DNA extraction, amplification and sequencing were performed according to published procedures (e.g.: Muasya et al. in press). Three p~astid The rbcL i~tr6'fuegon ~.- regions were amplified using polymerase chain reaction (PCR). of the plastid genome was amplified using the rbcL74F, rbcL938R and the rbcL1368R primers designed by Muasya et al. (in press). The trnL intron region of the plastid genome was amplified using trn-c, trn- f, trn- d, trn-e and trnL-f designed by Taberlet et a!. ( 1991 ). The rps 16 intron region of the plastid genome was amplified using the primers rpsF and rpsR2 designed by Oxelman et al. (1997). The PCR reactions were performed in 30pJ volumes consisting of 18.6pJ of sterile water, 3pJ of lOx DNA polymerase buffer, lpJ each of the forward and reverse primers (at 10/lM), 1,2/ll dNTP (lOmM), 0,2/ll ofTaq DNA polymerase and 2/ll of template DNA. The amplification was carried out on an applied Biosystems GeneAmp 2700 thermal cycler (Applied Biosystems, Foster City. CA, USA). The programme used had an initial denaturation phase of 2 minutes at 94 oc, followed by 30-35 cycles of 60 seconds at 94 OC, 60 seconds at 52°C and 2 minutes at 72°C, followed by a final extension phase of7 minutes at 72°C. The PCR products were checked on a 1% agarose gel. Successful products were subsequently sent to the Macrogen (http://www.macrogen.com) laboratories in Kumchum-ku in Seoul, Korea for cycle sequencing using the same 7 I I I I I I I I I I I I I I I I I I I I I primers as were used in the PCR. Sequences were cleaned, assembled, edited and aligned both manually and using ClustalW in the BioEdit programme {Thompson et al. 1994). 25 rps16 sequences from taxa representing 25 genera were added to sequences downloaded from GenBank (www.genbank.com). oftfL fl,J~" Morphological and ecological data: Morphological and ecological<Clata were obtained by reading through published literature @etghebeur 1998j(.ble I provides a list ofthe characters used in the analysis. Tfiese characters were scored at the genus-level, such that even though the tips of the branches were represented by species, the character states indicated the state present within each genus. This was done for two main reasons. Firstly many Cyperaceae genera are monotypic and secondly the available literature on specific character states was not extensive enough to allow for a species-level evolutionary character reconstruction. Analysis: The plastid sequence and the morphological and ecological data matrices were analysed in three separate ways. Firstly the individual plastid region matrices were analysed separately using parsimony reconstruction, secondly a parsimony analysis was conducted on the combined morphological and ecological and plastid sequence matrices and thirdly a Bayesian analysis was conducted on the combined morphological and plastid seq~nc matrices . ..,..., I: )\~"' "'~1JY) qvV') ~ 1 · Parsimony analysis: The parsimony analysis of the individual matrices was done to ensure that the separate data sets produced topologically similar results so that the data sets could be combined and produce a result with little conflict (Wiens 1998). Heuristic analyses were carried out using PAUP* (Swofford 2002). Searches were conducted under Fitch (1971) parsimony, TBR (tree-bissection-reconnection) branch swapping, and random taxon addition. The analysis was run for 1000 replicates, holding no more than 10 trees per replicate. The parsimony analysis of the combined data sets used a full heuristic search 8 I I I I I I I I I I I I I I I I I I I I I with 1 000 random addition replicates, holding a single tree at each step with SPR branch swapping and saving no more than 10 trees per replicate. A single strict consensus tree was kept and compared to the Bayesian output, to ensure that a similar topology was observed using both methods. Bayesian analysis: Final phylogenetic relationships were inferred using a Bayesian approach. The trees - 7. were rooted on Juncus effusus and Prionium serrata. Mr Bayes version 3.12 (Huelsenbeck and Ronquist 2003) was used to per;:or:nJ the~Bay'" combined data set. The most complex model- tliz_GTR+I+G m~_L,. was used in the analysis. This was chosen as it has been found that ilie-:lccur;cy of a Bayesian model suffers more in response to under parameterisation than to over parameterisation (Huelsenbeck and Rannala 2004). In the analysis parameters were estimated separately for each of the gene regions using uniform prior probabilities. Each run consisted of four Markov chains, each chain had random starting seeds. One chain was cold while the other three were heated. The temperature parameter was set ~:L!o il1P~m . 6 The analysis was run for 10 generations and was sampled at every 1OOth generation, thus producing 10 000 sampled trees per run. To ascertain whether stationarity had occurred, a plot of the -log likelihoods against generation time was investigated and this was also used to determine the "burn-in" time. The average standard deviation of split frequencies between runs was also used to determine stationarity, and as they had dropped to a value less than 0.03 this indicated that the tree samples had become similar enough after 106 generations to be regarded as stationary.- ~ ttl' ~ &z l-d{~ · Analysis of morphological and ecological data: The final Bayesian tree was loaded into Mesquite Version 2.0 (Maddison and Maddison 2007). Ancestral state reconstructions were done by imposing character states onto the tree. Marginal probability reconstructions were calculated with model MKI (est.) reporting likelihoods as proportional likelihoods (PL). 9 -~ Table 1: Showing the moiphological and ecological characters used to infer the evolutionary history of the Cyperaceae (Goetghebeur 1998) State 1 State 2 Tropical Absent Present Temperate Absent Present Forest Absent Present Grassland, Woodland or Savannah Absent Present Inselberg Absent Present Mediterranean Absent Present Permanently moist Absent Present Anemophily Zoophily Flowers Unisexual Bisexual Photosynthetic pathway c3 c4 Life-form Annual Perennial Root structure Rhizome • Other (e.g. culm) Number of functional stamens 1 Pollen type Tetrad State 3 State 4 2 3 4 or more Pseudomonad Mapanioid Distribution: Habitat: (Stream/Marsh/Bog/Swamp) Ecological character: Pollination mode Anatomical and Morphological. Character: ? ~ 10 I I I I I I I I I I I I I I I I I I I I I Results: /" / Out of 3263 characters, 661 wer~ most parsimonious trees each oft~e:dngh ~ ,/~ '"''"\\ uninformative ~yd 1056 we~ .-~"',_ '>' ~ informative")'460 equally 5212 were kept. The consistency index (CI) was 0.50, and the retention index was 0.703. The phylogenies produced by the Bayesian and the Parsimony analyses were similar in topology (Fig. 1 and Fig. 2). Phylogenetic relationships among genera and tribes: The Cyperaceae are resolved as monophyletic (Fig. 1 and Fig. 2). The Mapanioideae are shown to be sister to the rest of the Cyperaceae, with strong support (PP = 1.00; Fig. 2). Goetghebeur's (1998) Caricoideae (Cariceae in Fig. 2) and Sclerioideae (Sclerieae, Cryptangieae, Trilepideae and Bisboeckelerieae in Fig. 2) are embedded in Cyperoideae (Fig. 2). Within the Mapanioideae there is strong support for Hypolytreae and Chrysitricheae as sister groups (PP = 1.00; Fig. 2). Capitularina (PP = 0.89) and Exocarya (PP = 0.94) are included in the Chrysitricheae (Fig. 2). There is strong support (PP = 1.00) for the monophyly of the Trilepideae with Afrotrilepis sister to Coleochloa (PP = 0.99; Fig. 2). There is strong support (PP = 1.00) for Cladium as sister to the rest of the Cyperaceae (Fig. 2). The next branching includes Sclerieae and Bisboeckelerieae as sister to the rest of the Cyperaceae (PP = 0.61; Fig. 2). There is weak support (PP = 0.68) for Schoeneae as sister to Cryptangieae (Fig. 2). There is strong support (PP = 1.00) for Rhynchosporeae as sister to the rest of the Cyperaceae (Fig. 2). (PP = 1.00) for a clade including Khaosokia to Carex being sister There is strong sup~rt - to the remaining Cyps,(Fig. 2). Khaosokia branches out first and closest to Dulicheae (PP = 1.00) and this group is sister to a grade including Scirpeae and Cariceae (PP = 0.99; Fig. 2). Cariceae is shown to be embedded within Scirpeae (PP = 0.52; Fig. 2). The position of Oreobolopsis and Trichophorum makes Scirpeae paraphyletic with Cariceae (PP = 0.76; Fig. 2). Cariceae are found to be monophyletic (PP = 1.00; Fig. 2). 11 I I I I I I I I I I I I I I I I I I I I I Strict () Fig. 1: Maximum parsimony strict consensus tree of Cyperaceae, showing the outgroup. 12 I I I I I I I I I I I I I I I I I I I I I Fig. 2: Bayesian phylogeny of Cyperaceae based on a combined data set of three plastid gene regions (rbcL, trnl-F and rps16) showing the outgroup and Cyperaceae subfamilies and tribes. 2a) Shows the Hypolytreae and Chrysitricheae of the Mapanioideae and the Trilepideae, Cladieae (*), Bisboeckelereae (1). Sclerieae, Cryptangieae, and a section of Schoeneae of the Cyperoideae. 2b) Shows the rest of the Schoeneae, Rhvnchosporeae, Dulicheae (2), and a section of Scirpeae. The presence of Khaosokia is indicated by#. 2c) Shows the rest of Scirpeae, Cariceae, Eleocharideae, Abildgaardieae, and a section of Fuireneae (Fuirena is indicated by +). 2d) Shows the rest of Fuireneae and Cypereae. Maximum likelihood (PP) values are shown at each node. (Figures 3-15 can be found before the Appendix.) Y~ 0~ ~r" ~ ~· 13 I I I I I I I I I I I I I I I I I I I I I ?---------------- Juncus effusus Prioni um serratum Diplasia karatifo lia ,.. .. p.98 • Hypolytrum nemorum r'0 ·78 Mapania cuspidata 1.00 11o 0 - r'J. 89 Capitu Iarin a invo lucrata .0 .94 Lepiro nia articul.ata ,1 .oo Chorizandra cymbaria (r '. ~ \ 10 ' ~ °Chrysitrix capensis Microdracoid es squamosus ~ .oo Trilep is lhotzkian a Afrotrilepis ~ (I0.99 Coleochloa.abys,sinica I ~ Scirpoden dron b ogn eri Exocarya sclerioides .60 01adium ~ariscu'Y Becquerella·cymosa 00 · Diplacrum africanum ~1 • * 1·00 ScIena . d.Jstans r 1 ·00 I. 1 .00 Scleria fo lias a. Scleria terrestris Didymiandrum stellatum 1 .oo Exochogyne amazonica 0.99 Lag en ocarpus al bani ger . 1.00 I F11 Gymnoschoenus sphaerocephalus 00 · Caustis dioica Evandra aristata Fig. 2a. 14 I I I I I I I I I I I I I I I I I I I I I 1.UU 1 00 0 .96 · · 1.00 .61 E. van dra an.stat a Gahn ia bani ens is Mesomelaena pseudostygia Ptilothrix deusta Schoenus ni gricans 1 · .oo . hoenus qua dran gu Ians . .oo Ep1sc 1 Tetraria compar Tricostularia pauciflora Oreobolus kue kenthal ii 1.oo Capeob olus brevicau lis r-----.,.94 0.95 Cyathocoma sp Lepidosperma aff. filiforme 0 ~ 5 Neesenbeckia punctoria .ocr Tetreria capillaris - 1.00 Lepid osperma tortu asum 0.68 Baumea rubiginosa i 1.00 .-----......_j Machaerina sp Pleurostachys sp .oo Rhynchospora nervosa 1 00 · 1 1 00 ,....----·_ 0 92 ' Rhynchospora alb a Rhynchospora browni i Khaosokia caricoides .=t*=: Blysmus compressus _:---1 1 ·00 Dulichium arundinaceum Oreobolopsis clementis 1.oo Trichophorum alpinum 0 99 · rf 1.00-Eriophorum vaginatum . . . . Fig. 2b 15 I I I I I I I I I I I I I I I I I I I I I 1.00 LIIUjJIIUI Ull.l 0.76 - 1.00 . 1:/) .a· 1 00 Amphiscirpus nevad ensis · 85 Zamei oscirpus sp · 92 0 Phylloscirpus acaulis i · 1.00 0.52 1.0 vay IIIOLUIII Scirpus polystachyus · 1.00 ... 1 ;00 0 0 I» 0 Phyll oscirpus deserticola Kobresia simpliciuscula Schoenoxiph ium sparteum (') I» Uncin ia nemoralis 87 ·· Carex monostachya :::!. 0 0 I» (1) (') .97 Carex conferta o. ~ (1) .... 0 Carex sylvatica 78 ~ I» Carex echi nochlo e ·. 0 ·94 Carex atrofusca · 0.62 Carex h astian a Fuirena welwitschii : . Eieocharis atropurp urea 1 00 Eleocharis margi nulata (1) • I '1 .00Bulbostylis atrosan guinea + ~ (1) 0 0 ::r' I» :::!. §~ Nemum spadiceum Actin oschoenus Arthrostylis a phylla 0 · ~rachystli 8 ~ strand brokensis 0: (JC! Abildgaardia ovata .84 Crosslandia setifolia 0.88 I I» 2l n;· 0.. I» (1) 1.o0Fimbristylis complanata Fimbristylis d ichotoma Bolboschoen us maritimus • Fig. 2c 16 I I I I I I I I I I I I I I I I I I I I I Actin oscirpus grossus lacustris 1 00 · Schoenoplectus 0.99 Schoenop lectiella sen ega len sis Scirpoides holoschoenus subsp. thunbergii 1.00 Hellmuthi a membranacea .oo Isol epis cernua •.._---lll.4 .oo 0.79 1.00 0 . h .. esmosc oenus 1 00 · . ,. sp1ra 1s Ficinia gracilis Kyll ingi ell a microcephal a .99 ~1.00 0 xycaryum cu bense Courtoisin a assimilis Cyperus dichro ostachyus Cyperus cuspidatus .oo Cype_rus cyp~oides 1 Rem1rea manttma 0.99 0.89 Alinula paradoxa Alinula lipocarparoides Queenslandiella hyalina Ascolepis capensis 1.00 .oo Lipocarp ha n ana Lipocarp ha h emisphaerica Kyllinga appendiculata Sphaerocyperus eri naceus .oo Pycreus flavescens Cyperus compressus 0.99 Cyperus congestus 1.00 Cyperus en dlichi i Fig. 2d 17 I I I I I I I I I I I I I I I I I I I I I 7 F) (]_i 6 There is strong Fuireneae as ~s/ter (Rl = 1.0~) for the Abildgaardieae clade including / ' Eleoch~tia to the rest of the Cyps (Fig. 2). Abildgaardieae and ~ shown to be monophyletic with strong support (PP=. = 1.00; Fig. 2):, 1ctinoschoenus is resolved into the Abildgar~(P l.O)an~ as .-.> si~'r I / and El~ocharide are to each other (PP = 0.75; Fig. 2). Our results show that Fuirena cannot be resolved as si.ster to either Abildgaardieae or Eleocharideae, as these three clades form a trichotomy (Fig. 2). The rest of the Fuireneae, including Bolboschoenus, Actonoscirpus, Schoenoplectus and Schoenoplectiella are shown to be sister to Cypereae, although these genera display a grade topology (Fig. 2). There is strong support (PP = 1.00) for Bolboschoenus to be resolved as a separate clade, sister to the remainder of the Fuireneae (Fig. 2). There is also strong support (PP = 1.00) for Actinoscirpus and Schoenoplectus to be resolved as sister to Schoenoplectiella (Fig. 2). Cypereae, including Schoenoplectiella, form a monophyletic clade (PP = 0.79; Fig. 2). Within Cypereae there is shown to be strong support (PP = 1.00) for two main clades: a Scirpoides- Ficinia group and a Kyllingiella - Cyperus endlichii group (Fig. 2). Distribution and Habitat: The majority (85%) of the genera included in this study were found to have a tropical distribution (Fig. 3). Further it is likely (PL = 0.99) that the ancestral sedge had a tropical distribution (Fig. 3). A number of genera are shown to occupy habitats in temperate regions (Fig. 4). It is shown that the ancestral sedge most likely occurred in a permanently moist habitat, such as a swamp, stream, bog or a marsh (PL = 0.54; Fig. 5). The Trilepideae and Abildgaardieae are shown to be the only two tribes that are predominantly not found in permanently moist habitats (PL = 0.60 and 0.77 respectively; Fig. 5). The majority (56%) of the genera were found to occur in grasslands, woodlands and savannahs (Fig. 6). While it is unlikely (PL = 0.09) that the ancestral sedge occurred in this type of habitat, there is a fairly high likelihood (PP = 0.53) that the ancestor to the Scirpeae, Cariceae, Dulicheae, Cypereae, Fuireneae, Eleocharideae and the Abildgaardieae; and the ancestors of Sclerieae and Bisboeckelerieae (PL = 0.63) occurred in grassland, woodland 18 I I I I I I I I I I I I I I I I I I I I I or savannah (Fig. 6). The Cariceae (PL = 0.87) and Mapanioideae (PL = 0.97) are shown to have a number of genera found within a forest habitat (Fig. 7). The ancestor of the Mapanioideae was likely (PL = 0.97) to have occurred within a forest environment (Fig. 7). The Trilepideae are shown to occur on Inselbergs (PL = 0.99; Fig. 8). Pollination biology: Our results show that the ancestral sedges were likely to be anemophilous (PL = 0.99; Fig. 10). Zoophily- or traits indicating pollination by insects and other invertebrates - is shown to have evolved at least 9 times independently within Cyperaceae (Fig. 10). Two genera in the Mapanioideae - Hypolytrum and Mapania - are shown to have evolved zoophilous traits independently (PL = 0.56; Fig. 10). Life history: The ancestral life-history state of the Cyperaceae is shown to have a high likelihood (PL = 0.97) of having been perennial (Fig. 11). Annualness is shown to have arisen multiple times independently within the Cyperaceae, in Sclerieae, Schoeneae, Rhynchosporeae, Abildgaardieae, Fuireneae and multiple times within Cypereae and Eleochirea e (Fig. 11 ). The majority of Cyperaceae are perennial (Fig. 11 ). ---- Photosynthetic pathway: The photosynthetic pathway in ancestral sedges was likely to be C 3 (PL = 0.99; Fig. 12). c4 photosynthetic pathway is shown to have evolved at least four times independently, in four separate tribes: Rhynchosporeae, Eleocharideae, Abildgaardieae and Cypereae (Fig. 12). Anatomieal an~omhlgic data: There is strong likelihood (PL = 0.98) that the ancestral state within the Cyperaceae was to have bisexual flowers (Fig. 13 ). Presence of a unisexual flower is shown to have evolved multiple times independently within Cyperaceae (Fig. 13). There is strong support (PL > 0.97) showing that unisexual flowers have evolved independently from an ancestor with bisexual flowers in Trilepideae, Sclerieae, Cryptangieae, Cariceae, 19 I I I I I I I I I I I I I I I I I I I I I Cypereae and Eleocharideae (Fig. 13). Our results show that unisexual flowers appear to have evolved at least 10 times independently within the Cyperaceae (Fig. 13). The ancestral sedge is shown to have three functional stamens (PL = 0.86; Fig. 14). The majority of species are shown to have three functional stamens (Fig. 14). The Mapanioideae are shown to have lost all but one of their functional stamens, with the exception of Dip/asia which has more than four (Fig. 14). Our results show that the two outgroups Juncus and Prionium have a tetrad pollen type (PL = 0.85; Fig. 15). It is shown to be highly likely (PL = 0.99) that the ancestral pollen type within the Cyperaceae was pseudomonad (Fig. 15). There is strong support (PL = 0.99) showing that the Mapanioid pollen type has arisen within the Hypolytreae (Fig. 15). Discussion: Classification: This study supports the monophyly of the Cyperaceae reported by a number of studies (e.g. Muasya et al. in press). Goetghebeur's (1998) classification of the Mapanioideae according to their peculiar floral structure, where the lateral bisexual flowers are provided with a pair of usually larger, keeled, laminar hypogynous scales, is supported by our analysis. This study and other analyses of DNA data support the recognition of Mapanioideae as sister to all other Cyperaceae (Fig. 2; Muasya et al. in press). Our results provide evidence that the Sclerioideae and Caricoideae are paraphyletic, and consequently the recognition of Caricoideae and Sclerioideae as subfamilies separate from Cyperoideae is not supported by our findings. Goetghebeur (1998) recognized the Cyperoideae as having at least one (sometimes all) bisexual flower.lA,~ per spikelet and as lacking a mapanioid lateral pair of keeled hypogynous scales. ~itlea were recognized as having strictly unisexual flowers enclosed by a utricle (Goetghebeur 1998). Our data support the proposal for a revised classification of Cyperaceae into two subfamilies, Mapanioideae and Cyperoideae sensu Muasya et al. (in press). 20 I I I I I I I I I I I I I I I I I I I I I This study broadly supports the tribal circumscriptions of Goetghebeur (1998) with a few modifications. Within the Mapanioideae Goetghebeur (1998) recognized two tribes, Hypolytreae and Chrysitricheae. Hypolytreae are characterized as large-leaved species, usually with many spikelets per inflorescence, and very poorly differentiated embryos. Chrysitricheae are characterised as having a much reduced vegetative apparatus and inflorescence, and a highly differentiated embryo. These two tribes are well supported by this study (Fig. 2) although the position of Capitularina and Exocarya differs from Goetghebeur (1998). Our study provides evidence that both genera should be included in the Chrysitricheae, a finding that is supported by Simpson et al. (2003) in a combined pollen and DNA data study. The position of the Trilepideae within the Cyperaceae is not well-resolved by this study (PP = 0.60; Fig. 2). It thus remains uncertain whether this tribe could be sister to the rest of the Cyperioideae sensu Muasya et al. (in press) or if Mapanioideae is the true sister clade to the Cyperioideae sensu Muasya et al. (in press). Goetghebeur (1998) characterized the Trilepideae as having a panicle composed of many dense spikes of many tiny spikelets with few distichous glumes. Goetghebeur (1998) reported that the perianth is similar to that of the Cryptangieae, in that it is usually formed by 3 fimbriate scales opposite the flat sides of the achene. This suggests that the · Trilepideae could be more closely related to the rest of the Cyperioideae sensu Muasya et al. (in press). A greater sampling effort is required to be able to fully resolve this issue. Previously Cladium was placed within Schoeneae due to a restricted number of bisexual flowers per spikelet and a fairly well-developed perianth (Goetghebeur 1998 and Bruhl 1995). This study provides evidence which suggests that Cladium should be treated as a separate tribe, Cladieae, which had been previously recognized (Bruhl 1995). Goetghebeur (1998) recognized the Sclerieae as having bisexual or unisexual spikelets and an achene surrounded at the base by a hypogynium and a cupula. This circumscription is supported by our findings, which show a well-resolved and supported Sclerieae clade (Fig. 2). Goetghebeur (1998) recognized Bisboeckelereae as having (sometimes connate) empty glumes surrounding the apparently terminal female flower and male spikelets with the unusual structure of having glumes with a single stamen. Our 21 I I I I I I I I I I I I I I I I I I I I I study supports the recognition of Bisboeckelereae (PP = 1.00; Fig. 2). Bisboeckelereae and Sclerieae are resolved to be sister (PP = 1.00), an observation reported from previous studies (e.g. Muasya et al. in press). In this study this clade is sister to Schoeneae, unlike in Muasya et al. (in press) where this clade was embedded in Schoeneae. The Cryptangieae were circumscribed by Goetghebeur (1998) as having unisexual spikelets, spirally arranged glumes and a perianth usually formed by 3-fimbriate scales opposite the flat sides of the achene. The monophyly of the Cyptangieae is found to be supported by our study (PP = 1.00; Fig. 2). Cryptangieae are poorly supported (PP = 0.68) as sister to Schoeneae (Fig. 2). Goetghebeur (1998) circumscribed the Schoeneae as having a restricted number of bisexual flowers per spikelet often provided with a well-developed perianth and flowers that are included by the wings of adjacent glumes. However, Schoeneae sensu Goetghebeur (1998) is found to be heterogenous in this study (Fig. 2). This is supported by Muasya et al. (in press) which resolved four clades within Schoeneae. Thus our study supports the call for a need of further revision and possible division of the tribe made by Goetghebeur (1998) and Muasya et al. (in press). Our study also provides support for the recognition of Gymnoschoenus as a tribe separate from the rest of Schoeneae and for the recognition ofRhyncho sporeae, proposed by Muasya et al. (in press). There is strong support (PP = 1.00) for a clade comprising Khaosokia and members of Dulicheae, Scirpeae and Cariceae. This relationship has been reported by other studies (e.g. Muasya et al. 1998 and Muasya et al. 2000). Goetghebeur's (1998) classification of the Dulicheae as having a fertile spikelet prophyll bearing a bisexual flower appears to be well supported by our study (Fig. 2). Our results also suggest that Khaosokia could be included in a tribe separate from Cariceae and as sister to Dulicheae, a finding supported by Simpson et al. (2005). There is some indication that Khaosokia may be closely related to Sumatroscirpus, a sedge found in Sumatra. Goetghebeur (1998) recognised Scirpeae as having fertile spikelets with spirally arranged glumes and flowers with hypogynous scales. This circumscription is not upheld by this and other studies conducted with DNA 22 I I I I I I I I I I I I I I I I I I I I I data (e.g. Muasya eta/. in press). Cariceae is found to be embedded within Scirpeae, as Oreobolopsis and Trichophorum are basal to both clades (PP = 0.76; Fig. 2). We propose that these two genera be included in a new tribe, which would resolve both Cariceae and Scirpeae as monophyletic. The monophyly of the Cariceae is supported by Goetghebeur (1998) which reported that the clade is easily recognized by the female flower being enclosed within a utricle. 7 ' The relatively close relationship between Fuireneae and Scirpeae shown by our analysis is supported by Goetghebeur (1998) which reported that both tribes have a similar floral morphology. Fuireneae are shown to have a grade topology, meaning that a number of taxa could be resolved into a number of separate tribes (Fig. 2): This could be as a result of inadequate sampling, although a likelier explanation, supported by Muasya et a/. (in press) is that this section should be split into a number of separate tribes. Muasya eta/. (in press) showed that Fuireneae could be split into several clades. Fuirena is observed here to be sister to either Eleocharideae or Abildgaardieae, whereas the rest of Fuireneae form a grade and are closer to Cypereae (PP = 1.00). Previous studies (e.g. Muasya eta/. in press) have shown Fuirena to resolve apart from other Fuireneae raising questions as to the monophyly of the tribe. It is evident that Fuireneae needs further evaluation. Goetghebeur (1998) recognized the Eleocharideae as having reduced vegetative apparatus, a fixed unispiculate inflorescence, a unique embryo type and a helophilous life-form. This tribe is well supported by our analysis as being monophyletic and sister to the Abildgaardieae (Fig. 2). The close relationship between Eleocharideae, Abildgaardieae and Fuireneae is supported by Goetghebeur (1998). Goetghebeur (1998) stated that the Eleocharideae shares a number of characters in common with Fuireneae, including a similar vegetative morphology, an embryo with broadened cotyledon and a bristle-like perianth. Eleocharideae were also reported to share the Kranz syndrome (having a green sheath around the vascular bundles), distichous glumes, a differentiated and thickened style base and moniliform stigmatic hairs (Goetghebeur 1998). Our 9. analysis supports the proposal to include Trachystylis and Arthrostylis and Actinoschoenus in Abildgaardieae made by Muasya e;;;; ~n press). Similar results were 23 I I I I I I I I I I I I I I I I I I I I I obtained by Ghamkar eta/. (in press) based on plastid and nuclear ribosomal (ITS) data. The monophyly of the Cypereae is well-supported, in agreement with Goetghebe ur (1998) which characterized Cypereae by the presence of the Cyperus-type embryo. Hellmuthia, a genus previously included in the Chrysitricheae, is resolved within the Cypereae, a finding supported by Muasya et al. (in press). Distribution and habitat: Our analysis provides evidence to support the hypothesis that the ancestral sedge grew in marshy or wet habitats in tropical regions (Fig. 5 and Fig. 3). A few of the tribes, such as the Chrysitricheae, are shown to occur predominantly in open swamps (Fig. 5). The Cyperaceae appear to have radiated into drier open habitats; such as grassland, woodland, savannahs and Mediterranean shrubland, while a few clacles appear to have moved into closed forest habitats (Fig. 6, Fig. 9 and Fig. 7). Life history: The evolution of an annual life history is a remarkable adaptation to seasonal climates. Although annuals are relatively rare in monocots, Linder and Rudall (2005) reported that the strategy has evolved independently in seven families of the Poales, including Cyperaceae. Despite this Linder and Rudall (2005) found that annualness has not become the dominant strategy within any of these seven families. This is supported by our analysis which found that only a few genera contained annual species, while the majority of the genera were perennial (Fig. 11). Hu et al. (2003) found that in Poaceae the shift between the annual and perennial life histories appears to be controlled for by only two genes, which suggests that it may be a relatively simple evolutionary transition. The fairly high number of shifts within the Cyperaceae found in our analysis supports this conclusion (Fig. 11 ). Anatomical and morphological data: The ancestral photosynthetic pathway in the Cyperaceae is C3, although the C4 photosynthetic pathway system appears to have evolved multiple times (Fig. 12). There is some controversy surrounding the ecological advantages of the C4 photosynthetic 24 I I I I I I I I I I I I I I I I I I I I I pathway. Stock et al. (2004) investigated the contribution of climatic factors and phylogenetic relationships affecting the geographical distribution of c3 and c4 genera of the Cyperaceae in South Africa. Their analysis showed that there is no simple relationship between Southern African C 3/C 4 species distributions and environmental factors. Similar results have also been reported from other regions of the world (Teeri et al. 1980; Ueno and Takeda 1992). Linder and Rudall (2005) reported that the parallel evolution of C4 in grasses and sedges is probably an adaptation to changes in atmospheric C02 concentrations as opposed to other environmental factors such as seasonal drought. However, with a lack of precise estimate of dating, this hypothesis is difficult to test and further investigation is needed to resolve why this photosynthetic pathway has evolved multiple times within Cyperaceae. Linder and Rudall (2005) reported that unisexual flowers are often associated with wind pollination and that the shift from bisexual to unisexual flowers has occurred numerous times during the evolution ofPoales. Our analysis shows that this pattern is present within the Cyperaceae, as the ancestral sedge was likely to have had bisexual flowers and unisexual flowers are shown to have arisen multiple times independently (Fig. 13). Faegri and Vander Pijl (1979) reported that increased pollen to ovule ratio is normally regarded as characteristic of wind-pollinated species and that this often correlates to an increase in stamen number. However Linder and Rudall (2005) showed that within Poales there have been numerous reductions in stamen number and that stamen number is remarkably labile in grasses. This led them to conclude that the number of functional stamens is an unlikely adaptive modification in wind-pollinated groups in which an increase in stamen number is expected, maximizing pollen output (Linder and Rudall 2005). Our data show that the number of functional stamens is also labile in the sedges, and that it is thus unlikely to be an adaptation to anemophily (Fig. 14). Simpson et al. (2003) reported that members of the Hypolytreae possess "sticky" pollen in monads, which they suggested might be evidence for zoophily. Our analysis supports this observation as the Hypolytreae are found to have evolved zoophily and are shown to 25 I I I I I I I I I I I I I I I I I I I I I have Mapanioid pollen (Fig. 10 and Fig. 15). There also appears to be a correlation between ocurrance in a forest habitat and zoophily (Fig. 7 and Fig. 10). Both states were found within a number of the same genera in the Hypolytreae and Cariceae (Fig. 7 and Fig. 10). Goetghebeur (1998) reported that entomophily is probably linked to the forest habitat as wind velocity is reduced compared to an open habitat. This relationship does however require further investigation, at the specific level. Conclusion: There was a call from Muasya et al. (in press) for a need to study the same DNA regions -such as rbcL, trnL-F and rps16- for similar taxa such that different data sets could be combined. This study has attempted to achieve this. Cyperaceae are monophyletic and resolved into two subfamilies, the Mapanioideae and Cyperoideae. The overall topology of our analysis is similar to those observed in other studies, with a few exceptions. Our study suggests that Cladium should be treated as a separate tribe, the Cladieae. Sclerieae is observed as sister to as opposed to being embedded in Schoeneae. Fuirena is observed here as sister to either Eleocharideae or Abildgaardieae whereas the rest of Fuireneae form a grade and are closer to Cypereae. 24 species were sequenced for the first time in the rps 16 intron region of the plastid genome in an attempt to further resolve the phylogenetic relationships within Cyperaceae. In addition we were able to reconstruct the evolutionary history of the family using a number of morphological and ecological characters. The ancestral sedge was shown to have bisexual flowers, 3 functional stamens, pseudomonad pollen and the c3 photosynthetic pathway. Further, the ancestral sedge is likely to have been an anemophilous perennial plant which grew in marshy or wet tropical habitats. Future prospects: A complete investigation of the evolutionary history of the Cyperaceae would require dating the phylogeny. There are currently a number of both molecular and 26 I I I I I I I I I I I I I I I I I I I I I paleontological methods with which one could investigate the temporal framework of lineages or to estimate ages of clades (Magallon 2004). Bremer (2000) and Janssen and Bremer (2004) attempted to date the major monocot groups based both on fossil data and rbcL sequence data. It would be informative to be able to date the phylogeny of the Cyperaceae, as this would allow one to infer when a particular trait arose and thus to further investigate the evolutionary history of the family. Acknowledgemen ts: Firstly I would like to thank Muthama Muasya for being a considerate, engaging and patient supervisor. His assistance, guidance and sacrifice of time was much appreciated throughout the project. Secondly to Timothy Moore and Tony Verboom for assistance with the data analysis and for stimulating discussion around the topic. References: Bremer, K. (2000) Early Cretaceous lineages of monocot flowering plants. Proceedings -~ . ,_ of the National Academy of Sciences, USA. 97 pp: 4707-4711. <....... . .. . . . _.~ Bruhl, J.J. (1995) Sedge genera of Cyperaceae. ~s thew~: u~"'r:_,(4\ L . relationships and a new classification of the syv_:125-305. Bruhl, J.J. and Wilson, K.L. (2006) Toward~-mp.ehnsiv survey ofC 3 and C4 photosynthetic pathways in C y p e r a c e a e Fitch, W.M. (1971) Toward defining the course of evolution: minimum change for a specific tree topology. Syst. Zoo/. 20. pp: 406. Faegri, K. and Vander Pijl, L. (1979) The principles of pollination biology. Oxford: Pergamon. 27 I I I I I I I I I I I I I I I I I I I I I Ghamkar, ~archnt, :;,: II /i A.D.; Wilson, K.L. and Bruhl, J.J. (2006) Phylogeny of Abildgaardi,tae (Cyperaceae) inferred from ITS and trnl-F data. Aliso 23. pp: 00-00. Goetghebeur, P. (1998) Cyperaceae. In: Kubitzki, K. The families and genera of vascular plants 4:164. Springer, Berlin. Huelsenbeck, J.P. and Ronquist, F. (2003) MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics 1~tp: 1572-1574. \J Huelsenbeck, J.P. and Rannala, B. (2004) Frequentist properties of Bayesian posterior probabilities of phylogenetic trees under simple and complex substitution models. Systematic Biology. 53. pp: 904-913. Janssen, T. and Bremer, K. (2004) The age of major monocot groups inferred from 800+ rbcL sequences. Botanical Journal of the Linnean Society. 14~385-9. Linder, H.P. and Rudall, P.J. (2005) Evolutionary history ofPoales. Annu. Rev. Evol. Syst. 36.€£)107-12 4. Magallon, S. (2004) Dating lineages: molecular and paleontological approaches to the temporal framework of clades. Int. J. Plant Sci. 165. pp: 7-21. Maddison, W.P. and Maddison, D.R. (2007) Mesquite: a modular system for evolutionary analysis. Version 2.0. Muasya, A.M.; Bruhl, J.J.; Simpson, D.A.; Culham, A. and Chase, M.W. (2000a) Suprageneric phylogeny of Cyperaceae: a combined analysis. In Wilson, K.L. and Morrison, D.A. (Eds.) Monocots: Systematics and Evolution. Pp: 593-601. CSIRO Publishing, Melbourne, Victoria, Australia. 28 I I I I I I I I I I I I I I I I I I I I I Muasya, A.M.; Simpson, D.A.; Verboom, G.A.; Goetghebeur, P.; Naczi, R.F.C.; Chase, M.W. and Smets, E. (in press) Phylogeny ofCyperaceae based on DNA sequence data: current progress and future prospects. Oxelman, B.; Liden, M. and Berglund, D. (1997) Chloroplast rpsl6 intron phylogeny of the tribe Sileneae (Cayophyllaceae). Plant Systematics and Evolution. 206. pp: 393-410. Simpson, D.A.; Furness, C.A.; Hodkinson, T.R.; Muasya, A.M. and Chase, M.W. (2003) Phylogenetic relationships in Cyperaceae subfamily Mapanioideae inferred from pollen and plastid DNA sequence data. A mer. J. Bot. 90. pp: 1071-1086. Simpson, D.A.; Muasya, A.M.; Chayamarit, K.; Parnell, J.A.N.; Suddee, S.; Wilde, B.D.E.; Jones, M.B.; Bruhl, J.J. and Pooma, R. (2005) Khaosokia caricoides, a new genus and species ofCyperaceae from Thailand. Bot. J. Linn. Soc. 149. pp: 357-364. Stock, W.D.; Chuba, D.K. and Verboom, G.A.(2004) Distribution of South African C3 and C4 species of Cyperaceae in relation to climate and phylogeny. Austral. Ecology. 29. pp: 313-319. Swofford, D.L. (2002) PAUP*: Phylogenetic analysis using parsimony (*and other methods). Version 4. Sinaur associates, Sunderland, Massachusetts. Taberlet, P.; Gielly, L.; Pautou, G.; Bouvet, J. (1991) Universal primers for amplification of three non-coding regions of chloroplast DNA. Plant Molecular Biology. 17. pp: 11051109. Teeri, J.A.; Stowe, L.G.; Livingstone, D.A. (1980) The distribution ofC4 species of Cyperaceae in North America in relation to climate. Oecologia 47. pp: 307-310. 29 I I I I I I I I I I I I Thompson, J.D. Higgins, D.G. and Gibson, T.J. (1994) Clustal W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position specific gap penalties and weight matrix choice. Nucleic Acids Research Ueno, 0. and Takeda, T. (1992) Photosynthetic pathways, ecological characteristics and the geographical distribution of the Cyperaceae in Japan. Oecologia. 89. pp: 195-203. Verboom, G.A. (2006) A phylogeny of the schoenoid sedges (Cyperaceae: Schoeneae) based on plastid DNA sequences, with special reference to the genera found in Africa. Mol. Phylogenet. Evol. 38. pp: 79-89. Wiens, J.J. (1998) Combining data sets with different phylogenetic histories. Systematic Biology. 47. pp: 568-581. I. I I I I I I I I 30 I I I I I I I I I I I I I I I I I I I I I '"rj 3 C/l $:ll =:-. 0 ::s 0 ~ (!) ::s (!) ...., $:ll ~ ;::,: ::s- $:ll .-. ...., 0 (") 2. Q. C/l .-. ::::. 0" c .-. a· ::s 0" ~ (") ;:>;"" (") ...., (") (b C/l 0 ::+ a· ::s 2. ;:>;"" (!) ::s0 0 Q. < 2. c(!) C/l 0' ...., 3 0 ::+ $:ll ::s ..... ::s 0 Q. (!) :n I I I I I I I I I I I I I I I I I I I I I >Tj ....... +:>. 0(!> ::l (!> ,...., ~ 0' s:: ::l 0.. s· ....... (!> a 2: ~ ("") ;:-;pt ipt~;,u .. ("") ;:::;· ("") ~ en 0 ::+ c;· ::l e:.. "' ~ :r 0 0 0.. < e:.. s:: (!> en 0' ,...., §" 0 ::+ ~ ::l ....... ::l 0 0.. (!> :n N:io.~t! · · ....... I I I I I I I I I I I I I I I I I I I I I >Tj =t. 3 Cll ~. 0 :J 0 ~ 3 0 Cll ....... ::r cr PJ ;:: :;: Cll PJ ::r ...... Cll 0 ~ 0 ~ Cll :J s:: (') ::r C' ,..., PJ Cll 3 Cll 0 ,..., Cll ....... ,..., (1) PJ 3 cr ~.,I PJ . (') ;;>;"" (') ,..., "'.... (') (1) Cll N:i~•n!lu.U Fig. 6: Optimisation of genera occurring in open habitats, such as grassland, woodland or savannah (black circles). Proportional likelihood values are shown for important nodes. I I I I I I I I I I I I I I I I I I I I I "Tj :-. 3 Vl $:ll ....... c;· = S' ...., (1) = (1) ...., $:ll 0 (".) (".) s:: ...., ::::!. = :;· (".) 0 Vl (1) 0.. S' ...., (1) Vl ....... ::r $:ll cr ;:;: ~ cr ~ (".) ~ (".) ...., (".) ~ Vl 0 :::4. c;· = e. < e. s:: (1) Vl S' ...., 3 0 :::4. $:ll = = ....... 0 0.. (1) :" ~ .cr - .. I I I I I I I I I I I I I I I I I I I I I :::1 ~. 3 VJ ~ ..-+ a· ::J 8"' ""1 ~ ::J ~ ""1 ~ 0 () () s:: ""1 ::::!. ::J 0 ::J ......... ::J VJ ~ 0'" ~ ""1 VJ 0'" ~ "' () m ~ () ,.IP*"' A\ ......IW . . .I1i..S. ""1 () ~ VJ "'!'I!IJ'IGI\Im • p rlllll 0 :::4. a· ::J ~ "' ~ s.: 0 0 0.. < ~ s:: ~ VJ '"'aon.rtlu-t:J"...:• h ~i\I.T' lci~ 8"' ""1 3 ...,...,,.«CCPI- 0 :::4. ~ ::J ..-+ ::J 0 0.. ~ ~ ~ .a .... ft bl l!; l I I I I I I I I I I I I I I I I I I I I I '"rj :::=. 3 Ul ~ c;· ::l 0,....., (t> ::l (t> ~ ..... 0 (") (") = ..... ::!. ::l ~ (t> 0.. ;:::;: (t> 3 0 :::4. ~ ..... ..... ~ ::l (t> ~ ::l ::l :r ..... ::l ~ Ci 0 ;:::;: 0.. (t> ~ ~ Ci ~ (") 7\" (") ..... (") ~ Ul 0 :::4. c;· ::s !::... ~ (t> :r 0 0 0.. < !::... = (t> Ul b' ..... Mil \JVlr VI&" ~;J. . to* ~ · ...... I I I I I I I I I I I I I I I I I I I I I N 0 0 ..., ~ en ~ 3 ::s en ~ (') ....... ;s· ~ 0 ,......, ::s 0 ,......, ::s en ~ (') ....... en ~ "' ..., 0::s- 0 0 0 ....... 0.. ::s- < ..., a ~ ::s c~ < ~ :::4. ~ cr ..., ~ ....... en 0' ..., 3 ~ en 0 :::4. ~ ::s ::s- ....... ;::+." ::s ~ ~ (') 0 0.. ..., :n ~ ~ en to ::s ~ (') (') ~ ...,~- ~ 0.. ....... ~ "' (') ~ ~ ::s en ~ ::s 3 0 0.. ()" -e ::s- ~ ~ - - ~ - .... I I I I I I I I I I I I I I I I I I I I I "T:: OJ ::-. 3 ~ V> ;:>\" $:IJ ....... ..., o· ::s (") 0 (D ....., V> (") ::s ~ 0... (=i" ::s- ~ (1) (1) ..., $:IJ (") (1) $:IJ (1) ....... $:IJ ::s 0... Cti' ..., (1) 8"' ....... ..., p) (1) 3 0... I ~ ::s- ;:::+." (1) ....... $:IJ ;:< $:IJ V> ::s0 ~-::s ..., (1) (1) ;:I < ::s E. §.: c(1) V> Cti' 8"' ..., 8"' ..., 3 3 0 :::+ $:IJ ::s ....... ::s 0 0... (1) :n ..U,.". IO!"' ""'·A I I I I I I I I I I I I I I I I I I I I I :n =-· 3 Ul 1:1) =-· 0 ::s 0 ..-+, n I'--' 0" 1:1) (') ;;-;- (') ;;-;- ;:: :;: Cl> Cl> (') ..... (') Cl> Ul Cl> ..... 1:1) (') Cl> 1:1) Cl> 1:1) ::s 0.. ..... Cl> C) ...... Cl> 0.. ...... 1:1) .. X 1:1) Ul ::s- 0 ~. ::s .lt>totr- • •l'f'• I I I I I I I I I I I I I I I I I I I I I 'Tj ....... w u Ul ....... :::!. cr c....... c;· ::I 0......., c ::I Ul (1) ;x: c e. cr ~ (') ;:>';" ~ ...... . (') (p Ul ~ ::I 0.. cr Ul (1) ;x: §' ce. (1) ...... ~ (') (1) ~ (1) Ul :; 0 §. ::I 0 ::::1- c;· ::I e. ;:>';" (1) :; 0 0 0.. < e. c(1) Ul 8" ...... ~ re w cw lA Fig. 14: Optimisation of the number of functional stamens, showing proportional likelihood values for important nodes. Green circles indicate presence of 3 functional stamens, Blue indicates presence of 2; Black indicates presence of 4 or more; white circles indicate presence of a single functional stamen. -- - - -- - - - - - - - -' - - - - - - Appendix 1: List oftaxa used in this analysis, showing the voucher number and the Genbank accession numbers for each of the three plastid DNA regions. Taxon GenBank accession numbers voucher rps16 intron rbcL This study (Skelton, Y12985 OR trnL-F intron/spacer I. Cyperoideae Suess. Abildgaardieae Lye Abildgaardia ovata (Burm.F.) Kral Kenya: Muasya et al. 684 (EA, K) AJ295754 R.) This study (Skelton, EF178537 Actinoschoenus repens Raynal Zambia: Robinson 3643 (K) Arthrostylis aphylla R.Br. Australia: Wilson 8249 (NSW) AY725939 Bulbostylis atrosanguinea Kenya: Muasya 1037 (EA, K) Y12992 Australia: Wilson 10147 (K) EF178538 Kenya: Muasya 1029 (EA, K) Y13009 Kenya: Muasya 1006 (EA, K) Y13008 R.) (Boeck.) C.B.Clarke Crosslandia setifolia EF178592 W.Fitzg. Fimbristylis complanata (Retz.) Link Fimbristylis dichotoma (L.) Vahl AJ295755 44 -~ - -: -~ -~ ... -- - - - - - - - - - - - - Nemum spadiceum (Lam.) Desv. Ex Ham. Trachystylis stradbrokensis (Domin) Kukenth. Y12945 WEST AFRICA: Baldwin 9766 (K) Australia: Wilson EF178539 8175 EF178591 (K) Bisboeckelereae Pax ex L.T. Eiten Becquerelia cymosa Brongn. Brazil: Thomas et al. 10284 (NY) Diplacrum africanum Tanzania: Vollensen 3967 (K) This study (Skelton, Y12948 R.) C.B.Clarke AY725942 Cariceae Kunth ex Dumort. AM085614 Carex atrofusca Carex conferta A.Rich. Kenya: Muasya 1055 (K) Carex echinochloe Kunze Kenya: Muasya 1051 (K) This, study (Skelton, R.) This study (Skelton, Y12999 Y12997 AF191818 R.) (1993) Carex hostiana DC. Chase et al. Carex monostachya A.Rich. Kenya: Muasya 1052 (K) Carex sylvatica Huds. Simpson et al. (2003) L12672 Y12998 AY344175 45 --1 ___ _., - - - - - - - -- - -~li Kobresia simpliciuscula (Wahlenb.) Mackenzie Plunkett et al {1995); Yen et al. {2000) U49232 Uncinia nemoralis K.L.Wilson Australia: Wilson et al. 9533 {K) AY725956 Kenya: Muasya 2566 {EA) EF178543 Thailand: Simpson et al. 1886 {K) AY725948 AF164948 Schoenoxiphium sparteum {Wahlenb.) C.B.Clarke Uncertain tribe aff. Caricieae Khaosokia caricoides D.A.Simpson, Chayam. & J.Parn. EF178535 Cryptangieae Benth. Didymiandrum stellatum {Boeck.) Gilly Exochogyne amazonica C.B.Clarke Lagenocarpus alboniger Venezuela: Liesner 23562 {GENT) Brazil: Aparecida da Silva 1986 {GENT) Brazil: Thomas 11111 {NY) EF178544 This study {Skelton, R.) EF178545 AY725949 {A.St.Hil.) C.B.Clarke Cypereae Dumort. This study {Skelton, Alinula lipocarparoides R.) Alinula paradoxa Goetgh. & Vorster Tanzania: Faden et al. 96/29 {K) AJ278290 AJ295756 46 - - - -1-1- -- - - -1 - - - - - - -'- - Kenya: Muasya 1009 (EA, K) AF449518 Y13003 Tanzania: Faden et al. 96/119 (K) AF449519 AY40590 Cyperus compressus L. Thailand: Muasya 1375 (K) AF449521 AF449506 Cyperus congestus Vahl Australia: Coveny et al. 17492 (K) AF449522 AF449507 Cyperus cuspidatus Kunth. Thailand: Muasya 1374 (K) AF449523 Ascolepis capensis (Kunth) Ridl. Courtoisina assimilis (Steud.) Maquet AJ295757 AY040595 AF449555/ AF449556/ AF449568 AF449508 AF449557/ AF449569 Cyperus cyperoides (L.) Thailand: Muasya 1277 (K) AF449524 AF449509 AF449558/ AF449570 Kuntze Kenya: Muasya 976 (EA, K) AF449525 Y12965 Cyperus endlichii Kuk. Kenya: Muasya 695 (K) AF449526 AF449510 Desmoschoenus spiralis New Zealand: Ford 44/94 Cyperus dichroostachyus A.Rich. Hook.f. Ficinia gracilis Schrad. /AF449571 AF449559/ 449572 AJ404701 (~) AJ295753 S. Africa: Muasya 2355 AF449537 EF178589 EF178595 Hellmuthia membranacea (Thunb.) R.W.Haines & Lye Isolepis cernua (Vahl) Roem. & Schult. var. cernua S. Africa: Weerderman et al. 269 (K); Muasya 1145 (K) Britain: Muasya 1058 (K) This study (Skelton, Y13000 Yl3014 AJ295815 R.) AJ295775 Kyllinga appendiculata K. Schum. Kenya: Muasya 1050 (EA, K) AF449542 Y13007 AJ295761 47 t' .l i J -~.J iiiiiiiJ IIIIJ -D ..:J ~ ~ __, - - ------· --:1 Kyllingiella microcephala (Steud.) R.W.Haines & Lye Zimbabwe: Muasya et al. 1118 (K) Lipocarpha hemisphaeric a Thailand: Muasya 1217 (K) ~ ~ AF449540 ~ ~ ~ ~ r-"1 rJ AY040592 AJ295807 AF449516 (Roth.) Goetgh. Lipocarpha nana (A.Rich.) --------- --. AF449565/ AF449577 Kenya: Muasya 972 (EA, K) AF449545 Y12990 zambia: Richards 13318 (K) This study (Skelton, R.) Y13006 Kenya: Muasya 1022 (EA, K) AF449547 Y13005 J.Raynal AJ295762 Oxycaryum cubense (Poepp. & Kunth) E. Palla Pycreus flavescens (L.) Rchb. Queenslandi ella hyalina AJ295763 Kenya: Mwachala 296 (EA) AY725953 (Vahl) Ballard Remirea maritima Aubl. Tanzania: Faden et al. 96/48 (K) Scirpoides holoschoenu s (L.) S. Africa: Acocks s.n. AF449550 Y12994 AJ295811 s. Africa: Muasya 1205 (K) (Schrad.) sojak Scirpus polystachyu s AY040593 AY040604 (K) sojak Scirpoides thunbergii AY040602 AF449551 AJ404727 AJ295812 Australia: Pullen 4091 K Dulicheae Rchb. ex J. Schultze-Mo tel Blysmus compressus Panz. Afghanistan : Dobson 221 (K) This study AJ404700 AJ295766 48 --,------ lJ r.:::J -- '- -!---- - - - - - - - - - - - -Schoenoplectus lacustris Britain: Muasya 1043 (K) AF449554 Y12943 (L.) Palla AJ295809 Rhynchosporeae Pleurostachys sp. Brazil: Kallunki et al. 513 Y12989 (NY) AY344151 Rhynchospora alba (L.) Vahl Rhynchospora brownii Roem. et Schult. Rhynchospora nervosa (Vahl.) Boeck. Simpson et al. (2003) S. Africa: Verboom 616 (BOL) 00058336 Brazil: Kallunki et al. 512 (NY) DQO 58 3 53 AY344174 DQ058316 Y12977 Schoeneae Dumort. Baumea rubiginosa (Spreng.) Boeck. Australia: Wilson et al. 94 71 (K) This study (Skelton, AY725940 R.) Capeobolus brevicaulis (C. B. Clarke) J. Browning s. Africa: Verboom 646, BOL 00058324 DQ058343 Carpha alpina Wardle et al. (2001); Zhang et al. (2004) Caustis dioica R.Br. Australia: Chase 2225 (K) This study (Skelton, R.) Y12976 Cladium mariscus (L.) R. Br. Locality unknown: MJC 292 00058319 DQ058338 DQ058303 AF307909 AY230012 DQ058298 (K) 50 ' - - --- --- - - - - - - - - - - - -Cyathocoma bachmannii (Kuk.) 00058325 S. Africa: Browning 835 (GENT) EF200590 EF178604 DQ058344 DQ058304 DQ058349 DQ058311 C.Archer Cyathocoma hexandra (Nees) J. Browning Epischoenus quadrangularis (Boeck.) C. B. Clarke Evandra aristata R.Br. S. Africa: Verboom 648, BOL) 00058332 S. Africa: Verboom 636 (BOL) AY725944 Australia: Wilson et al. 8974 (NSW) Gahnia baniensis Benl. Gymnoschoenus sphaerocephalus (R.Br.) Hook.f. Malaysia: Simpson 2737 (K) Aus-tralia: Wilson et al. 9463 (K);Zhang et al. (2004) 00058323 DQ058342 This study (Skelton, R.) AY725945 DQ058302 AY230033 Lepidosperma aff. filiforme Machaerina sp. Australia: Coveny et al. 17470 (K); Roalson et al. (2001) New Guinea: Johns 9195 (K) Mesomelaena pseudostygia Australia: Chase 2226 (K) 00058322 Y12 9 59 DQ058301 S. Africa: Muasya 1214 (K) 00058327 AY725952 DQ058306 Lepidosperma tortuosum F.Muell. ~, AY725950 ~- hv DQ058340 AF285074 .- DQ058300 (Kuk.) K.L.Wilson Neesenbeckia punctoria (Vahl) Levyns Oreobolus kukenthalii Malaysia: Simpson 2659 (K) Steenis Oreobolus oligocephalus Zhang et al. ( 2004) Y12972 EF178536 AY230031 W.M.Curtis 51 -- -- - -·-- -- - - - - - - - - - -Ptilothrix deusta (R. Br.) Zhang et al. (2004) K.L. Wilson Schoenus nigricans L. Saudi Arabia: Edmondson 3382 (K) Tetraria compar (L.) Lestib. S. Africa: Verboom 549, (BOL) Tricostularia pauciflora Australia: Coveny et al. 17484 (K)iZhang et al. (2004) (R.Br.) Benth. This study (Skelton, R.) This study (Skelton, R.) Y12983 This study (Skelton, R.) AY725954 DQ058350 AJ295814 DQ058312 AY230038 Scirpeae Kunth ex Dumort. Malaysia: Simpson 2660 K Amphiscirpus grossus Amphiscirpus nevadensis (S. Argentina: Charpin et al. 20575 (GENT) DQ317926 Watson) Oteng-Yeboa Eriophorum angustifolium Simpson et al. DQ317925 (2003) AY344177 Honckney Eriophorum vaginatum L. Poland: Beyer et al. 2 (K) Oreobolopsis clementis (M.E.Jones) Dhooge & Dhooge (2005) AF449553 Y12951 AJ295769 Goetgh. Phylloscirpus acaulis (Phil.) Goetgh. & Dhooge et al. 2003 AJ811011 AJ575926 Dhooge (2005) D.A.Simpson AJ576029 52 j -- --~ Phylloscirpus deserticola (Phil.) Dhooge & Goetgh. .. - - - -- - - -- -Ecuador: Laegaard et al. 21478 (GENT) This study (Skelton, AJ704785 R.) Scirpus polystachyus F. AJ704786 Yl2974 Australia: Pullen 4091 (K) Muell. Trichophorum alpinum (L.) AJ295813 CANADA: Waterway 2002.95 (GENT) Pers. Zameioscirpus atacamensis (Phil.) Dhooge & Goetgh. Bolivia: Ruthsatz & Budde 10328 (Trier) This study (Skelton, AJ810999 AJ575929 R.) DQ317924 AJ576032 Sclerieae Kunth ex Fenzl Scleria distans Pair. Kenya: Muasya 1023 (EA, K) Scleria foliosa A.Rich. Tanzania: Muasya 939 (EA, K) Yl2986 Scleria terrestris (L.) Malaysia: Simpson 2658 (K) Yl2947 00058320 Y12968 DQ058299 Fassett Trilepideae Goetgh. Afrotrilepis Reynders et al. Coleochloa abyssinica Ethiopia: Vollesen 80/2 (K) (2005) This study (Skelton, R.) (A. Rich.) Gilly This study (Skelton, Y12975 R.) 53 -:' ____ _ -1- ---- - ----·-7 This study (Skelton, AY725951 Microdracoides squamosus Hua Bonn Ace. 150 Trilepis lhotzkiana Nees Bonn Ace. s.n. AY725955 Indonesia: Johns 8725 (K) EF178588 (J.V.Suringar) Kern Chorizandra cymbaria R.Br. Simpson et al. Bremer (2002) AJ419940 Chrysitrix capensis L. S. Africa: Muasya 1242 (K) AY344148 Exocarya sclerioides Simpson et al. AY344145 R.) II. Mapanioideae C.B.Clarke Chrysitricheae Lestib. ex Fenzl Capitularia foliata Uitt. Capitularina involucrata AY344168 (2003) AJ419938 AY344171 (2003) AY344167 (F.Muell.) Benth. Lepironia articulata (Retz.) Y12957 Malaysia: Simpson 1236 (K) Domin. Hypolytreae Presl ex Fenzl Diplasia karatifolia Rich. ex Pers. Hypolytrum nemorum (Vahl) Spreng. Mapania cuspidata (Miq.) Uittien AY344169 Simpson et al. (2003) AY344166 Malaysia: Simpson 1379 (K) AY344142 Y12958 Brunei: Marsh 4 (K) 00058318 Y12955 AJ295816 AJ295817 54 :~ -·~ Scirpodendron bogneri s.s. Malaysia: Simpson 2650 (K) AY344143 Y12946 Hooper AY344164 OUtgroups Juncus effusus L. Prionium serratum Drege Simpson et al. (2003); Chase et al., 1993 S. Africa: Gettliffe Norris, s.n. (NBG) L12681 U49223 AY344156 AY344155 / 55