South African Journal of Botany 89 (2013) 150–155

Contents lists available at ScienceDirect

South African Journal of Botany

journal homepage: www.elsevier.com/locate/sajb

Phylogenetic position of the southern African members of the tribe based on molecular and morphological data

Meshack N. Dludlu ⁎, Charles H. Stirton, Samson B.M. Chimphango, Abubakar Bello, A. Muthama Muasya

Bolus Herbarium, Department of Biological Sciences, University of Cape Town, Private Bag X3, Rondebosch 7701, Cape Town, South article info abstract

Available online 18 July 2013 The tribe Psoraleeae (Leguminosae subfamily Papilionoideae) comprises 185 in nine genera that have a nearly worldwide distribution, occurring predominantly in Mediterranean regions. About 60% of the species Edited by B-E Van Wyk belong to the genera, C.H.Stirt. and L., which have a centre of diversity in the Cape Floristic Region of South Africa. Since previous molecular studies have sampled only a few species of the Keywords: tribe from this region, this study sought to determine the phylogenetic position of the southern African genera and to test whether they are monophyletic. Phylogenetic relationships were reconstructed using Monophyly African Psoraleeae DNA sequence data (trnL-F, rpoB-trnC and ITS) and seven morphological characters, which diagnose the Phylogenetics two southern African genera. The data were analysed using the parsimony method. There was strong support Hallia for the Psoraleeae as a clade, but most of the nodes within the large genera were poorly supported. The Otholobium southern African species of Psoralea and Otholobium together formed a strongly supported clade. This clade Psoralea was sister to the Rydb., but without support. However, the Psoralea species were nested within the southern African Otholobium. Additionally, some South American species that are currently recognised as Otholobium were resolved in a clade distinct from the southern African species, making Otholobium polyphyletic. Morphological characters that separate Otholobium and Psoralea are discussed. Finally, the southern African gen- era as currently circumscribed are not monophyletic. However, further investigations using more informative DNA loci are required to validate this observation. Furthermore, the taxonomic placement of the South American species needs to be reviewed. © 2013 SAAB. Published by Elsevier B.V. All rights reserved.

1. Introduction is currently being revised by our research group and the work indicates that there could be up to 72 species, many of which are known only by The tribe Psoraleeae Benth. (Leguminosae, Papilionoideae) has a manuscript names (e.g. Goldblatt and Manning, 2000). nearly worldwide distribution. It comprises 185 species in nine genera, Rydberg (1919, 1928) was the first to clearly define the circum- with about 60% of the diversity endemic to southern Africa and scriptions of the new world genera (Hoita, , , contained in Otholobium (53 species) and Psoralea (50 species; excluding ) and demonstrated that these were distinct from the old undescribed species) (Stirton, 2005). The rest of the diversity is world genera ( and Psoralea), and suggested that the name geographically restricted to Mediterranean and subtropical biomes in Psoralea should be used in a narrower sense for African taxa. Forbes Australia (Cullen, 34 species, with 4 species in Arabia and Africa), North (1930), following on the seminal studies of Harvey (I862), adopted America (Pediomelum, Orbexilum, , Hoita, Ladeania; 38 species), this narrower concept by including most southern African taxa into in Eurasia (;2species)andinSouthAmerica(Otholobium; Psoralea, excluding some Psoralea species recognised as a segregate 8 species). Recent studies on taxonomy and phylogenetics have focussed genus Hallia Thunb. Salter (1939) amalgamated Hallia into Psoralea, on new world taxa (Grimes, 1990; Egan and Crandall, 2008a; Turner, arguing that the gross morphology of Hallia species did not differ from 2008; Egan and Reveal, 2009) and the Australian members of the tribe that of Psoralea. Stirton (1981) split Psoralea sensu Forbes into Psoralea (Grimes, 1997). Neither has the phylogenetic position of the southern and a new genus Otholobium,andfurthersplitPsoralea sensu lato, rein- African members relative to the rest of the tribe nor the monophyly of stating the previously described genera Cullen and Bituminaria.Anum- the genera been tested. Taxonomically, the generic and species bound- ber of species have been published since Forbes' revision (e.g. Stirton, aries between and within the genera, Otholobium and Psoralea are still 1981, 1995; Stirton et al., 2011). not well resolved. The latest taxonomic treatment is a revision of the Stirton (2005) restricted the name Psoralea to about 50 species genus Otholobium by Stirton (1989) in an unpublished thesis. Psoralea endemic to the Cape Floristic Region (CFR) of South Africa and a few others that extend their range to as far north as the Northern Limpopo ⁎ Corresponding author. Tel.: +27 21 650 3684; fax: +27 21 650 4041. province in South Africa. These species share a unique structure, the E-mail address: [email protected] (M.N. Dludlu). cupulum, a term coined by Tucker and Stirton (1991) which refers to

0254-6299/$ – see front matter © 2013 SAAB. Published by Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.sajb.2013.06.019 M.N. Dludlu et al. / South African Journal of Botany 89 (2013) 150–155 151 a cup-shaped structure (resulting from the fusion of intercalary ) 2.2. Morphological character scoring found at the base of each flower pedicel. Although its function is not clear, Tucker and Stirton (1991) postulated that it protects the young A total of 7 morphological characters (Table 1)werescoredforallthe flower bud by encapsulating the true . Hallia species, considered sampled taxa. They are characters that have been used to delimit genera by some authors (e.g. Harvey, I862; Forbes, 1930) to be distinct from or some sections of the genera in the southern African Psoraleeae Psoralea share the presence of a cupulum. Unlike Psoralea, which has (Stirton, 1981, 1989). The character set consisted of both binary and compound and filiform , Hallia species have simple and broader multi-state characters, and character states were not ordered (Table 1). leaves. Nonetheless, further studies (Tucker and Stirton, 1991; Crow The primary sources of data included: specimens collected as part of et al., 1997) have failed to find sufficient characters to distinguish Hallia this study, herbarium specimens (BOL, PRE, and NBG) and Stirton's from Psoralea,andconfirmed Salter's (1939) amalgamation of Hallia (1989) revision of the genus Otholobium. Secondary sources of data with Psoralea. In his new genus Otholobium, occurring mainly in southern included published literature such as Bean and Johns (2005), Goldblatt Africa, with a few extending their range as far as eastern Africa, Stirton and Manning (2000), Grimes (1990), Grimes (1997), Stirton (2005), (1981) separated out 20 species of Psoralea sensu lato. His diagnostic Trinder-Smith (2006), Turner (2008) and Egan and Reveal (2009). characters for this genus were the occurrence of entire recurved mucronate-obovate to oblanceolate leaflets and the occurrence of 2.3. DNA extraction, amplification and sequencing inflorescences with bracteate triplets of flowers in which each triplet is subtended by a single . Further studies (Stirton, 1989) DNA was extracted using the CTAB technique according to Doyle recognised 53 species in the genus Otholobium. and Doyle (1987). Polymerase chain reaction (PCR) was performed Grimes (1990) presented a morphology based hypothesis of using an Applied Biosystems GeneAmp 2700 thermal cycler (Applied phylogenetic relationships among the Psoraleoid genera. He postulated Biosystems, Foster City, CA, USA). Target DNA regions were amplified that Psoralea and Otholobium were sister genera which together formed and sequenced using standard primers for the internal transcribed a clade sister to the rest of the Psoraleeae. However, in his revision of the spacer (ITS; White et al., 1990), the trnL-F intron (Taberlet et al., 1991) American Psoraleeae Grimes (1990) placed eight South American and rpoB-trnC (Shaw et al., 2005). Thirty microliters of PCR volumes fi species into Otholobium despite that they do not tintoStirton's were made by mixing 3 μl buffer, 3 μlMgCl2,1.2 μldNTP,1 μlofforward (1981) circumscription of Otholobium. Egan and Crandall (2008a) primer, 1 μlofreverseprimer,0.2μl Taq polymerase, 2 μl of template conducted a molecular phylogenetic study of the American Psoraleeae, DNA, and 18.6 μl of sterile distilled water. For the ITS the forward and including three South African Otholobium species. They found that the reverse primer sequences in the 5 -3 orientation were GGAAGTAA South African Otholobium species formed a clade sister to the rest of AAGTCGTAACAAGG and TCCTCCGCTTATTGATATGC, respectively and the American Psoraleeae, whereas the South American Otholobium spe- reaction conditions followed those of White et al. (1990).Amplification cies were embedded within the American clade as sister to the genus of the trnL-F intron followed the reaction conditions of Taberlet et al. Orbexilum.ThissuggeststhatOtholobium as currently circumscribed (1991) using the primer sequences CGAAATCGGTAGACGCTACG (for- is polyphyletic. However, in the absence of molecular studies with ward) and ATTTGAACTGGTGACACGAG (reverse). Finally, the rpoB-trnC detailed sampling of the African genera, Egan and Crandall (2008a) was amplified following the reaction conditions of Shaw et al. (2005) were unable to make formal taxonomic changes. and the forward and reverse primer sequences were CACCCRGAT The current study sought to test whether the splitting of Psoralea TYGAACTGGGG and CKACAAAAYCCYTCRAATTG, respectively. PCR sensu lato by Stirton (1981) and the hypothesis of phylogenetic rela- products were run on an electrophoresis tank containing 0.5× TBE tionships presented by Grimes (1990) are supported by a phylogenetic using 1% agarose gel and visualised under UV light to check if the ampli- analysis of the genera. We employ a total evidence approach, whereby fication process had been successful. Amplified products were purified we analyse morphology and molecular data as this has been shown to using GE Healthcare DNA purification kits and sent to MacroGen reflect evolutionary relationships that could be obscured if only one (http://www.macrogen.com) in Korea or the University of Stellenbosch kind of data was used (Murrel et al., 2001; Schuh et al., 2009; Wheeler DNA sequencing facility for sequencing using the same primers that et al., 2013). The specific objectives were (i) to determine the phyloge- were used in the PCR. netic position of the African Psoraleoid genera Otholobium and Psoralea relative to the other members of the tribe and (ii) to test if they are 2.4. Sequence alignment and phylogenetic analyses monophyletic. Sequence contigs were assembled and edited using Staden package 2. Material and methods version 1.6.0 (Staden et al., 1998) and aligned using the ClustalW multiple alignment tool in Bioedit version 7.0 (Hall, 1999). Misaligned 2.1. Taxon sampling sections were adjusted manually in Bioedit. The DNA data sets were first analysed separately, and then concatenated to form a combined In this study we sampled the DNA of 74 taxa from silica dried field DNA data matrix. Taxa that were absent in some partitions were collected material in South Africa representing 37 species of coded as missing for those partitions. An additional matrix comprised Otholobium and 37 species of Psoralea (voucher specimens deposited at the Bolus Herbarium). A majority of the species were represented by Table 1 a single sample, with only a few that had samples from two or three Morphological characters and their states as used in the phylogenetic analysis. Where a character is not applicable, it was coded as “?”. different collections. For the other Psoraleoid genera outside southern Africa (i.e. Pediomelum,extra-AfricanCullen, Hoita, Rupertia, Ladeania, Character States Orbexilum and Bituminaria) sequences were downloaded from GenBank, Secondary loss of leaves 0 = absent; 1 = present most of which were from the study by Egan and Crandall (2008a). Leaf type 0 = pinnate; 1 = palmate taxa were chosen based on the phylogenetic analysis of Maximum number of leaflets 0 = one; 1 = three; fi North American Psoraleeae by Egan and Crandall (2008a).Thesewere 2= ve; 3 = more than 5 Leaflets filiform 0 = no; 1 = yes Abrus precatorius (tribe Abreae), Apios americana (tribe ), Recurved mucro on leaf apex 0 = absent; 1 = straight; Amphicarpaea bracteata, Cologania pallida, Glycine canescens, Glycine 2 = recurved microphylla (all from Phaseoleae, sub-tribe Glycininae) and Desmodium Cupulum 0 = absent; 1 = present floridanum from the tribe Desmodieae. A list of the sampled taxa is borne in triplets subtended by a bract, 0 = absent; 1 = present where each flower is further subtended by a bract provided in Appendix A. 152 M.N. Dludlu et al. / South African Journal of Botany 89 (2013) 150–155 the combined DNA and character codes for the seven morphological 4. Discussion characters. Phylogeny reconstruction was done using parsimony in the programme PAUP (Phylogenetic Analysis Using Parsimony) version 4.1. Phylogenetic relationships 4.0b10 (Swofford, 2002). The searches were done heuristically, with 10 000 random addition replicates, bisection-reconnection (TBR) The low levels of DNA sequence variation in the sampled loci branch swapping and multrees in effect, holding 20 at each step. (Table 2) led to poor node support on trees and the formation of For evaluating support, bootstrap analyses were done based on 1000 polytomies in the strict consensus trees. However, there was strong replicates, each involving a heuristic search, with simple addition se- support for the tribe Psoraleeae, the African Psoraleeae as a clade quence and TBR branch swapping.

3. Results 98 3.1. Data matrices

The aligned matrices of the three gene regions comprised 148 taxa, seven of which formed the outgroup and the remainder constituted the ingroup. The combined DNA and morphology matrix had 3852 characters, comprising ITS (937 characters), rpoB-trnC (1618), trnL-F (1290) and seven morphological characters (Table 2). Trees from the analyses of the individual gene regions were congruent, but very poorly resolved and mostly unsupported, with a majority of the branches forming polytomies in the strict consensus trees. However, the analysis 60 of the combined DNA matrix gave fewer and better resolved trees, at least at the generic level (tree provided as Supplementary material). The combined DNA and morphology matrix had a similar topology and resolution to that of the combined DNA matrix, but with improved support for some nodes. The strict consensus tree of the combined DNA and morphology matrix, with bootstrap support (BS) values is presented (Fig. 1).

3.2. Generic relationships within the tribe Psoraleeae

The monophyly of the ingroup, representing the tribe Psoraleeae was well supported in both the analysis of combined DNA alone and 52 73 that of DNA and morphology (BS = 100%). A majority of nodes at 57 the backbone of the tribe were retained in the strict consensus tree although they lacked bootstrap support. Within the genera only a few 67 nodes were retained in the strict consensus tree as indicated by the 65 polytomies in Fig. 1.ThegeneraPsoralea and the southern African Otholobium formed a well-supported clade (BS = 99%), sister to the 94 genus Hoita (but this relationship was not well-supported). However, 77 species of the genus Psoralea together with two Otholobium species (Otholobium swartbergense and Otholobium wilmsii)werepartofa moderately supported grade (BS = 60%) that was nested within the southern African clade. The two South American species of Otholobium (Otholobium caliginis and Otholobium mexicanum) were resolved as a well-supported clade (BS = 100%) sister to the genus Bituminaria, 99 a relationship that had 51% bootstrap support. The genera Orbexilum, Pediomelum, Ladeania, Cullen and Rupertia were all monophyletic with very strong support (BS = 100%, 96%, 100%, 99%, and 100% respectively).

Table 2 Summary of data matrices for alignment and parsimony analysis: vc = variable parsimony un-informative characters, pic = parsimony informative characters, CI = consistency index, RI = retention index. Values in parentheses next to DNA regions are the number of taxa sampled. All taxa were scored for morphology.

Matrix Aligned No. (%) vc No. (%) pic Tree CI RI No. 100 length length trees

ITS (117) 937 156 (16.65) 353 (37.67) 974 0.77 0.86 2630 trnL-F (118) 1290 256 (19.84) 232 (17.98) 1622 0.53 0.80 3426 rpoB-trnC (110) 1618 666 (41.16) 358 (22.13) 1604 0.85 0.83 4320 Fig. 1. Strict consensus tree obtained from the analysis of the combined DNA and Combined DNA 3845 753 (19.58) 912 (23.72) 3630 0.65 0.81 60 morphology matrix. Numbers above branches are bootstrap support values (%) greater Combined 3852 861 (22.35) 954 (24.77) 4166 0.67 0.80 90 than 50%. Taxon labels: P = Psoralea,O=Otholobium,Or=Orbexilum,H=Hoita, DNA and B=Bituminaria,Pe=Pediomelum,La=Ladeania,Ru=Rupertia,Cu=Cullen,A= morphology Apios,Ab= Abrus,Am= Amphicarpaea,Co= Cologania,De= Desmodium,G = Glycine. M.N. Dludlu et al. / South African Journal of Botany 89 (2013) 150–155 153

study of Egan and Crandall (2008a) found better resolution by sequenc- 100 ing up to eight molecular markers, sampling more DNA loci for the African clade might improve the resolution. In addition, the inclusion 100 of morphological data in the phylogenetic analysis slightly improved 51 support for some nodes, which supports the total evidence approach to cladistics, an approach that has proven useful for several groups of 81 organisms (Murrel et al., 2001; Schuh et al., 2009; Lopardo et al., 2011; Wheeler et al., 2013). 100 100 The study sought to determine the phylogenetic position of the southern African genera and to test their monophyly. With regards to the American Psoraleeae our results corroborate those of Egan and Crandall (2008a) in showing that except for Otholobium,thegenera 98 are monophyletic and thus their circumscriptions represent natural groupings. Our results also support Egan and Crandall (2008a) in refuting the placement of South American species under Otholobium by Grimes (1990) as this renders it polyphyletic. However, given the low support of the relationship between these species and the genus Bituminaria,it 55 is not clear whether they should be assigned to this genus or to a genus of their own. Moreover, since only two of the eight species were included in both our study and that of Egan and Crandall (2008a), further studies, 59 including all species might resolve this. The phylogenetic position of the southern African genera could 78 not be unequivocally established because the sister relationship between Hoita and the southern African clade was not supported 100 76 (Fig. 1). On the other hand the study supports Stirton's (1981) applica- tion of the name Otholobium only to the African species, since the South 54 American species were resolved as a distinct clade that does not share the most recent ancestor with the African clade. However, it does not 100 support the recognition of Psoralea and Otholobium as distinct genera. This is because the Hennigian principle of monophyly, on which classi- fications are based recognises only monophyletic groups as natural and 100 thus requires that supraspecific taxa refer only to such groups (Sosef, 1997). However, given the low resolution and lack of support within the southern African clade we desist from making any taxonomic changes until a more robust phylogenetic hypothesis is available. A robust phylogeny would also allow for testing of the evolution of the 100 morphological characters that are used to separate the genera (as discussed below) and clarify whether the placement of two Otholobium species among Psoralea species is due to the low resolution in the data or a reflection of their true phylogenetic position. 96 4.2. Morphological characters separating Otholobium and Psoralea 74 100 While the results of the phylogenetic analysis do not support the 65 recognition of Psoralea and Otholobium as distinct genera, it is impor- 100 tant to note that the two genera vary extensively in their morphology. Some of the seven characters that were included in the study separate 100 the two genera and others are unique to sections within the genera. 100 These characters relate to the leaves and flowers of the species as 100 illustrated in Fig. 2. The loss of leaves is restricted to a group of about 10 species in the genus Psoralea. Whether this state has occurred as a 100 single evolutionary event or arisen multiple independent times is not known. Both genera have a few species that have unifoliolate leaves. In Psoralea, these are the species of the former genus Hallia (Psoralea imbricata, Psoralea laxa and Psoralea monophylla), and in Otholobium, Fig. 1 (continued). these are dwarf species that have capitate inflorescences subtended by long peduncles (i.e. Otholobium thomii, Otholobium rotundifolium (Otholobium/Psoralea), the American Psoraleeae clades (Ladeania/ and Otholobium lanceolatum). An analysis of the evolutionary history Rupertia/Pediomelum) and for the other genera (Cullen, Orbexilum, and of this character could shed light on whether or not it is a synapomorphy Hoita). Except for the southern African clade, which is being presented of specific clades. Apart from the unifoliolate species, all other for the first time (with greater sampling) in this study, these results Otholobium species have trifoliolate leaves, while in Psoralea the number are consistent with those of Egan and Crandall (2008a). Given that the of leaflets is highly variable ranging from three to seventeen. The pos- tribe Psoraleeae diversified within the last 6.3 million years (Egan and session of filiform leaflets, mostly with rounded and grooved leaves Crandall, 2008b), perhaps there has not been a sufficient accumulation (rarely flattened or conduplicate), is restricted to members of the of mutations at shallow nodes especially among the African clade, genus Psoralea,whileOtholobium species have broader leaves. Stirton hence the lack of resolution within the genera. However, since the (1981) used the occurrence of recurved mucronate leaf apices as one 154 M.N. Dludlu et al. / South African Journal of Botany 89 (2013) 150–155

A D

C

B E

Fig. 2. Illustrations of some of the various leaf types and the cupulum found in the African Psoraleeae. A = Cupulum, indicated by the arrow, B = Pinnate and filiform leaves as found in , C = Unifoliolate leaf, found in Psoralea monophylla (=Hallia monophylla), D and E = Trifoliolate and recurved mucronate leaves as found in Otholobium fruticans and O. swartbergense, respectively. Photographs by C.H. Stirton. of the diagnostic characters for the genus Otholobium. However, the Grimes, 1990) polyphyletic. More work is needed to determine if all unifoliolate members of the genus have straight or arching mucros and eight South American species form a clade before they are assigned to one species of Psoralea, Psoralea aculeata, also has recurved mucronate a genus of their own. Additional sampling and adding more informative leaf apices. The homology of this character requires further investigation molecular markers, such as done previously by Egan and Crandall through ancestral state reconstruction. (2008a), might help resolve the issues of phylogenetic support and The main character that separates Otholobium from Psoralea is the resolution and allow us to assess to what extent the generic segregation presence of the cupulum in Psoralea (Fig. 2). The cupulum furnishes a between Otholobium and Psoralea is valid, and whether key morpholog- number of useful morphological and anatomical characters (number, ical characters currently considered as diagnostic or potentially diag- size and shapes of lobes, vasculature, glandulosity and vestiture). This is nostic have taxonomic value. also the character state upon which the sinking of Hallia into Psoralea Supplementary data to this article can be found online at http:// (Salter, 1939; Stirton, 1989; Crow et al., 1997) was based. dx.doi.org/10.1016/j.sajb.2013.06.019. Another character that separates the two genera is the inflorescence structure of Otholobium, in which flowers are borne in triplets, with Acknowledgements each triplet subtended by a single bract and then each flower having an ovate or oblong bract subtending it. In Psoralea there are two minute We acknowledge the curators of the Bolus (BOL), Pretoria (PRE) free filiform bracts subtending the base of each flower pedicel. In and Compton (NBG) herbaria for allowing us access to Psoraleeae Otholobium the main florescence is the first to flower followed by specimens. Funding for the research was provided by the University sequential basipetal flowering of the paracladia. In Psoralea the terminal of Cape Town, SANBI Threatened Species Programme (TSP) and the shoots flower first, followed by the lateral shoots (Stirton, 1989). Less South African National Research Foundation (NRF-SABI), to whom we clear cut is that proliferation, or alternating flowering and vegetative are very grateful. phases on the same axis, does not occur in Otholobium but is quite com- mon in Psoralea. Pseudoracemes, sensu Tucker and Stirton (1991) occur References in Psoralea but are absent in Otholobium. It is important to note that a preliminary phytochemical survey of Bean, A., Johns, A., 2005. Stellenbosch to Hermanus South African Wildflower Guide 5, second ed. Botanical Society of South Africa, Cape Town. the Psoraleeae (Boardley et al., 1986) and anatomical studies (Crow Boardley, M., Stirton, C.H., Harborne, J.B., 1986. A chemotaxonomic study of the tribe et al., 1997) suggested that there are other characters (i.e. type of Psoraleeae in Africa. Systematic Evolution 14 (6), 603–613. secretory cavity, multicellular gland type, and essential oils) that Crow, E., Stirton, C.H., Cutler, D.F., 1997. Leaf anatomy of the genus Psoralea sensu stricto (Psoraleeae, Papilionoideae, Leguminosae). Botanical Journal of the Linnean Society separate the two genera. However, more species need to be studied to 124, 156–182. test their applicability at the generic level. Doyle, J.J., Doyle, J.L., 1987. A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochemical Bulletin of the Botanical Society of America 19, 11–15. 5. Conclusions Egan, A.N., Crandall, K.A., 2008a. Incorporating gaps as phylogenetic characters across eight DNA regions: ramifications for North American Psoraleeae (Leguminosae). Molecular Phylogenetics and Evolution 46, 532–546. The study has shown that the southern African Psoraleeae form a Egan, A.N., Crandall, K.A., 2008b. Divergence and diversification in North American – monophyletic group, but Psoralea is nested within Otholobium. How- Psoraleeae, due to climate change. BMC Biology 6, 55 68. Egan, A.N., Reveal, J.L., 2009. A new combination in Pediomelum and a new genus, Ladeania, ever, because of the poor resolution in the phylogenetic analysis of from western North America (Fabaceae; Psoraleeae). Novon 19, 310–314. the three DNA markers and morphology, generic delimitations within Forbes, H.M.L., 1930. The genus Psoralea Linn. Bothalia 3, 1–35. the Psoraleoid genera and within the African clade are still uncertain. Goldblatt, P., Manning, J., 2000. Cape : A Conspectus of the Cape Flora of South Africa, National Botanical. Institute of South Africa, Strelitzia 9, Pretoria. Nonetheless, the current data indicate (with strong support) that the Grimes, J.W., 1990. A revision of the New World species of Psoralea (Leguminosae: two South American species render the genus Otholobium (sensu Papilionoideae). Memoirs of the New York Botanical Gardens 61, 1–114. M.N. Dludlu et al. / South African Journal of Botany 89 (2013) 150–155 155

Grimes, J.W., 1997. A revision of Cullen (Leguminosae: Papilionoideae). Australian Staden, R., Beal, K.F., Bonfield, J.K., 1998. The Staden Package. Bioinformatics Methods Systematic Botany 10, 565–648. and Protocols. Hall, T.A., 1999. BioEdit: a user friendly biological sequence alignment editor and analysis Stirton, C.H., 1981. Psoraleeae. In: Polhill, R.M., Raven, P.H. (Eds.), Advances in Legume program for windows 95/98/NT. Nucleic Acids Symposium. University Press, Oxford. Systematic Part 1. Royal Botanic Gardens, Kew, pp. 337–343. Harvey, W.H., 1862. Leguminosae. In: Harvey, W.H., Sonder, O.W. (Eds.), Flora Capensis Stirton, C.H., 1989. A revision of Otholobium C.H. Stirton (Papilionoideae; Leguminosae). volume II Leguminosae to Loranthaceae. Lovell Reeve & Co., London, pp. 143–157. University of Cape Town. Lopardo, L., Giribet, G., Hormiga, G., 2011. Morphology to the rescue: molecular data Stirton, C.H., 1995. AnewspeciesofPsoralea (Leguminosae; Papilionoidea) from eastern and the signal of morphological characters in combined phylogenetic analyses—a South Africa. Kew Bulletin 50 (3), 619–622. case study from mysmenid spiders (Araneae, Mysmenidae), with comments on Stirton, C.H., 2005. Tribe Psoraleeae. In: Lewis, G., Schrire, B., Mackinder, B., Lock, M. the evolution of web architecture. Cladistics 27, 278–330. (Eds.), Legumes of the World. Royal Botanic Gardens, Kew, pp. 447–451. Murrel, A., Campbell, N.J.H., Barker, S.C., 2001. A total evidence phylogeny of ticks Stirton, C.H., Clark, V.R., Barker, N.P., Muasya, A.M., 2011. Psoralea margaretiflora provides insight into the evolution of life cycles and biogeography. Molecular (Psoraleeae, Fabaceae): a new species from the Sneeuberg centre of floristic Phylogenetics and Evolution 21 (2), 244–258. endemism, Eastern Cape, South Africa. Phytokeys 5, 31–38. Rydberg, P.A., 1919. Psoraleae. North American Flora 24, 1–24. Swofford, D.L., 2002. PAUP*: Phylogenetic Analysis Using Parsimony, Version 4. Sinauer Rydberg, P.A., 1928. Genera of North American Fabaceae III: tribe Psoraleae. American Associates, Sunderland, Massachusetts. Journal of Botany 15, 195–203. Taberlet,P.,Gielly,L.,Pautou,G.,Bouvet,J.,1991.Universal primers for amplification of three Salter, T.M., 1939. Amalgamation of the genus Hallia Thunb. with Psoralea L. South non-coding regions of chloroplast DNA. Molecular Biology 17, 1105–1109. African Journal of Botany 5, 45–46. Trinder-Smith, T.H., 2006. Wild flowers of Table Mountain National Park. Botanical Schuh, R.T., Wierauch, C., Wheeler, W.C., 2009. Phylogenetic relationships within the Society of South Africa, Cape Town. Cimicomorpha (Hemiptera: Heteroptera): a total-evidence analysis. Systematic Tucker, S., Stirton, C.H., 1991. Development of the cymose inflorescence, cupulum and Entomology 34 (1), 15–48. flower of Psoralea pinnata (Leguminosae: Papilionoideae: Psoraleeae). Botanical Shaw, J., Lickey, E.B., Beck, J.T., Farmer, S.B., Liu, W., Miller, J., Siripun, K.C., Winder, C.T., Journal of the Linnean Society 106, 206–227. Schilling, E.E., Small, R.L., 2005. The tortoise and the hare II: relative utility of 21 Turner, B.L., 2008. Revision of the genus Orbexilum (Fabaceae: Psoraleeae). Lundellia 11, 1–7. noncoding chloroplast DNA sequences for phylogenetic analysis. American Journal Wheeler, Q., Assis, L., Rieppel, O., 2013. Heed the father of cladistics. Nature 496, of Botany 92, 142–166. 295–296. Sosef, M.S.M., 1997. Hierarchical models, reticulate evolution and the inevitability of White, T.J., Bruns, T., Lee, S., Taylor, J. (Eds.), 1990. Amplification and Direct Sequencing paraphyletic supraspecific taxa. Taxon 46, 75–85. of Fungal Ribosomal RNA Genes for Phylogenies. Academic Press, San Diego.