Molecular Phylogenetics and Evolution 92 (2015) 25–44 Contents lists available at ScienceDirect Molecular Phylogenetics and Evolution journal homepage: www.elsevier.com/locate/ympev Evolution of Asparagus L. (Asparagaceae): Out-of-South-Africa and multiple origins of sexual dimorphism q Maria F. Norup a, Gitte Petersen a, Sandie Burrows b, Yanis Bouchenak-Khelladi c, Jim Leebens-Mack d, J. Chris Pires e, H. Peter Linder c, Ole Seberg a,⇑ a Natural History Museum of Denmark, Sølvgade 83, Opg. S, K-1307 Copenhagen K, Denmark Buffelskloof Nature Reserve Herbarium, P.O. Box 710, Lydenburg 1120, South Africa Institute of Systematic Botany, University of Zurich, Zollikerstrasse 107, 8008 Zurich, Switzerland d Department of Plant Biology, University of Georgia, Athens, GA 30602, USA e Division of Biological Sciences, 371 B Life Sciences Center, 1201 Rollins Road, University of Missouri-Columbia, Columbia, MO 65211-7310, USA b c a r t i c l e i n f o Article history: Received 26 May 2014 Revised 3 June 2015 Accepted 4 June 2015 Available online 14 June 2015 Keywords: cpDNA Sexual dimorphism Infrageneric structure Phylogeny PHYC Species complexes a b s t r a c t In the most comprehensive study to date we explored the phylogeny and evolution of the genus Asparagus, with emphasis on the southern African species. We included 211 accessions, representing 77 (92%) of the southern African, 6 (17%) of the tropical African, 10 (56%) of the strictly European and 6 (9%) of the Eurasian species. We analyzed DNA sequences from three plastid regions (trnH-psbA, trnD-T, ndhF) and from the nuclear region phytochrome C (PHYC) with parsimony and maximum likelihood methods, and recovered a monophyletic Asparagus. The phylogeny conflicts with all previous infra-generic classifications. It has many strongly supported clades, corroborated by morphological characters, which may provide a basis for a revised taxonomy. Additionally, the phylogeny indicates that many of the current species delimitations are problematic. Using biogeographic analyses that account for phylogenetic uncertainty (S-DIVA) and take into account relative branch lengths (Lagrange) we confirm the origin of Asparagus in southern Africa, and find no evidence that the dispersal of Asparagus follow the Rand flora pattern. We find that all truly dioecious species of Asparagus share a common origin, but that sexual dimorphism has arisen independently several times. Ó 2015 Elsevier Inc. All rights reserved. 1. Introduction Although the few commercially important species of the monocot genus Asparagus L. (Asparagaceae) (edible, e.g. A. officinalis, A. albus; ornamental or medicinal, e.g. A. asparagoides, A. falcatus, A. setaceus, A. scandens) are well known, the remaining species are poorly studied. The exact number of species in the genus is uncertain with estimates between 120 and 300+ spp. (Kubitzki and Rudall, 1998; Mabberley, 2008; World Checklist of Selected Plant Families, 2009). Many of these are listed as threatened or vulnerable, but the status of even more species is not sufficiently known (Raimondo et al., 2009; Red List of Threatened Species of Japan, 2007; Sánchez Gómez et al., 2009; Santana et al., 2004). A few species are even a general threat to biodiversity, as they have become invasive weeds in e.g. Australia, New Zealand and Lord Howe Island (Auld and Hutton, 2004). q This paper was edited by the Associate Editor Timothy Evans. ⇑ Corresponding author. E-mail address: Oles@snm.ku.dk (O. Seberg). http://dx.doi.org/10.1016/j.ympev.2015.06.002 1055-7903/Ó 2015 Elsevier Inc. All rights reserved. Asparagaceae subfamily Asparagoideae as currently delimited contains the two genera, Asparagus and Hemiphylacus S. Watson (Chase et al., 2009; Fay et al., 2000; APG III, 2009; Pires et al., 2006; Seberg et al., 2012), with the large and morphologically variable Asparagus encompassing the vast majority of species. Asparagus is strictly Old World with a diversification hotspot in South Africa (Fig. 1), and occurs in regions with semiarid to arid and Mediterranean-type climates. Hemiphylacus, with only five known species, is found exclusively in Mexico (Raimondo et al., 2009). All species of Hemiphylacus are herbaceous, with characters reminiscent of e.g. Chlorogalum (Lindl.) Kunth, Hastingsia S. Watson, Schoenolirion Durand, Hesperocallis A. Gray and Leucocrinum Nutt. ex. A. Gray (Hernandez, 1995). Asparagus, however, contains both herbaceous and sub-shrubby members, as well as scramblers/climbers, and is generally recognized by the presence of reduced, scale-like leaves, subtending the usually needle-like, fascicled cladodes. The plants often bear spines and the flowers are small (2–12 mm diameter), white or whitish, and the fruit is a berry or rarely a nutlet. The African species of 26 M.F. Norup et al. / Molecular Phylogenetics and Evolution 92 (2015) 25–44 Fig. 1. Distribution map of the 209 currently accepted Asparagus species (Govaerts et al., 2009). Borders represent TDWG regions (Biodiversity Information Standards; www. tdwg.org), and regions are coloured (see colour bars on left side of figure) according to the number of species in the area. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) Asparagus generally have hermaphroditic flowers, whereas the majority of Eurasian species are dioecious. There are two competing hypotheses on the evolutionary pathway to dioecy: either dioecy evolved via monoecy through fixation of the male/female ratio (Renner and Ricklefs, 1995), or it first included the coexistence of male-sterile plants (females) in a hermaphroditic population (gynodioecy) followed by female sterility in the remaining hermaphrodites (Darwin, 1877; Charlesworth and Charlesworth, 1978). A recent study on dioecy in the monocots suggests that dioecy in this group evolves more often from hermaphroditism than from monoecy (Weiblen et al., 2000). Despite lack of direct observations, it has been suggested that the pathway would possibly include a gynodioecious phase, which is more plausible than a direct transition (involving an unlikely simultaneous suppression of both male and female development) or androgynoecy (extremely rare), but should be focus of further studies (Weiblen et al., 2000). The infrageneric classification in Asparagus has been contentious since Willdenow (1808) described the genus Myrsiphyllum, that included all known bisexual Asparagus species with flattened cladodes and axillary flowers. Since then, authors have either divided the genus into two or more genera (Dahlgren et al., 1985; Huber, 1969; Kunth, 1850; Obermeyer, 1983, 1984; Willdenow, 1808), or recognized a single genus (Baker, 1875; Malcomber and Demissew, 1993; Roemer and Schultes, 1829), based on different opinions on the importance of various morphological characters (e.g. spines, flowers, cladodes, or inflorescence structure). Dividing the genus into subgenera was another approach, i.e. Baker (1875) based his divisions on lowering the rank of the genera described by Kunth (1850) (Asparagus (L.) Kunth, Asparagopsis Kunth and Myrsiphyllum), thus resulting in three subgenera (Eu)asparagus Baker (= Asparagus; the dioecious species), Asparagopsis (Kunth) Baker (hermaphroditic species with linear cladodes) and Myrsiphyllum (Willd.) Baker (hermaphroditic species with flattened, leaf-like cladodes). Jessop (1966) challenged Baker’s infrageneric classification in the first complete revision of a major part of the species in which he recognized 40 southern African species in one genus (Asparagus). He rejected the split between Asparagopsis and Asparagus (sensu Kunth (1850) and Baker (1875)), at least as applicable to the South African species, as none of the species which Kunth (1850) placed in subgenus Asparagus were in fact dioecious. Jessop instead subdivided Asparagus into eight sections, based on what he regarded as natural groupings, and with Myrsiphyllum retained as one of the sections. Obermeyer (1983, 1984) advocated a split into three genera sensu Kunth: Asparagus, Protasparagus Oberm. and Myrsiphyllum. Thus, Asparagus again comprising the dioecious (Eurasian) species; Myrsiphyllum, the hermaphroditic, southern African species with flattened, leaf-like cladodes, no spines, and usually connivent filaments and tepals; and Protasparagus containing the remaining hermaphroditic species in southern Africa. This circumscription was retained in the M.F. Norup et al. / Molecular Phylogenetics and Evolution 92 (2015) 25–44 treatment for the Flora of southern Africa (Obermeyer and Immelman, 1992), which remains the most recent taxonomic treatment of the genus in the region. Malcomber and Demissew (1993) disputed the validity of the subdivision at a generic level, with only flower sexuality differing between Asparagus and Protasparagus. Instead they proposed a unified genus Asparagus with only two subgenera, Asparagus and Myrsiphyllum (sensu Obermeyer, but including only the species with connivent filaments). Subgenus Asparagus thus contained species with unisexual or hermaphrodite flowers with free and spreading filaments, and linear to filiform cladodes. However, taking into account the species not included in the treatment by Malcomber and Demissew (1993), the morphological characters used to distinguish subgenus Myrsiphyllum become less clear-cut, and consequently the subdivision was rejected by Fellingham and Meyer (1995), at least for the southern African species. As a result the last publication on the subject recognized an undivided genus, Asparagus, without an infrageneric classification. Considering the difficulties in delimiting distinct recognizable groupings based on morphological characters alone, reconstructing the phylogeny of the genus based on molecular data is the next logical step. So far, only a few attempts have been made, and most have a very limited number of species. Thus, both Lee et al. (1997) and Stajner et al. (2002) included 10 species in their studies based on Restriction Fragment Length Polymorphism (RFLP), with only three species overlapping between the studies. Fukuda et al. (2005) included 24 species in a study using the chloroplast (cpDNA) loci petD and rpoA. Wiegand (2006) used the ribosomal ITS and included 53 species. However, our study, using a considerably denser taxon sampling, indicates that ITS is highly problematic for phylogenetic purposes in Asparagus. Despite the widespread use of ITS in phylogenetic reconstructions, non-uniform concerted evolution leads to the existence of several incompletely homogenized rRNA copies, and results in an array of orthologs and paralogs, which confound the phylogenetic signal (reviewed e.g. in Álvarez and Wendel, 2003; Feliner and Rosselló, 2007) and render ITS of limited use. The most recent phylogenetic study of Asparagus (Kubota et al., 2012) used 5 non-coding regions of cpDNA but only included 23 species. Both Fukuda et al. (2005) and Kubota et al. (2012) indicate that the species previously placed in subgenus Myrsiphyllum do not form a monophyletic group, but find the species included in subgenus Asparagus to be monophyletic. The historical biogeography of Asparagus has received little attention. However, the pattern of an Old World distribution with a massing of species in southern Africa is not unusual, and is also found, inter alia, in Aloe L., Gladiolus L., Scrophulariaceae and Thesium L. Fukuda et al. (2005) and Kubota et al. (2012) suggest that Asparagus initially radiated and subsequently dispersed out of southern Africa, a pattern also proposed for Thesium (Moore et al., 2010), Scrophulariaceae (Oxelman et al., 2005), Crassulaceae (Mort et al., 2001) and Amaryllidoideae (Rønsted et al., 2012). Although the delimitation of most species in Asparagus appears to be largely unproblematic, there are several difficult species complexes. Consequently the two most recent revisions, by Jessop (1966) and Obermeyer and Immelman (1992), recognized very different numbers of species. The differences appear to be based on different interpretations of character variation, with Jessop interpreting many characters as ecologically plastic, while Obermeyer and Immelman used all variation to define taxa. While molecular data might not be suitable to determine species limits, they may still be helpful in circumscribing species. Using a much broader taxon sampling, the present paper focuses on the 81 species occurring in southern Africa (Burrows and Burrows, 2008; Fellingham and Meyer, 1995; National Red 27 List of South African plants, 2009). The species in this region exhibit substantial ecological variation, with both shade and sun tolerant species, shrubs and climbers, as well as the occasional, introduced European species (A. officinalis) that dies back each winter. Furthermore, there is a great diversity of growth forms and rooting systems, as well as an interesting variation in leaf (cladode) morphology (Jessop, 1966; Obermeyer and Immelman, 1992; personal observations). With data sampled from both cpDNA (trnH-psbA, trnD-trnT and 30 ndhF) and nuclear DNA (phytochrome C (PHYC)) we (1) reconstruct the phylogeny of Asparagus with focus on southern African taxa, (2) make an assessment of the infrageneric delimitation and species circumscription in some larger species complexes in Asparagus, (3) explore the evolution of dioecy and (4) test the ‘‘out of Africa’’ hypothesis of Fukuda et al. (2005) and Kubota et al. (2012). 2. Materials and methods 2.1. Taxon sampling A total of 211 accessions of Asparagus were sampled and used for DNA sequencing (Table 1), representing 77 southern African species (92% of all currently accepted in the region Obermeyer and Immelman, 1992), six (of 36) other African species, ten (of 18) European species (including the Canary Islands) and six (of 67) species from temperate and tropical Asia. Species with known species delimitation problems were represented by multiple samples in order to investigate their delimitation. Table 1 also include the author abbreviations. Several new and as yet undescribed species were also included in the analysis; these are given provisional names in quotation marks. Outgroup species consisted of a number of near-relatives of Asparagus, all selected on the basis of previous molecular phylogenetic studies of the order Asparagales (Fay et al., 2000; Pires et al., 2006; Seberg et al., 2012). Due to sampling issues it was not always possible to obtain sequences for the same species for all regions. Following the procedure of Campbell and Lapointe (2009), we combined the sequences building composite taxa to represent some of the outgroup genera. This is readily discernible from Table 1. 2.2. DNA extraction, amplification and sequencing Material was collected in the field and immediately dried in silica. Total genomic DNA was extracted from the silica-dried samples using the DNeasy Plant Mini Kit (Qiagen Nordic, Copenhagen, Denmark). The primers used for amplification and sequencing are presented in Table 2. For PCR amplification of the cpDNA regions, reactions (per 50 lL) included the following: 5 lL buffer, 10 lL GATC, 1 lL BSA, 0.25 lL taq polymerase, 4 lL Mg2+, 25.75 lL H2O, 1.5 lL of each of the two primers and 1 lL DNA. For PHYC we included 1 lL DMSO and only 24.75 lL H2O. The products were purified using the QIAquick PCR purification kit (Qiagen) according to the manufacturer’s instructions. PCR reactions were run on a Bio-Rad C1000 Thermal Cycler (Bio-Rad Laboratories, Copenhagen, Denmark) according to the manufacturer’s protocol. Thermocycler conditions employed for trnH-psbA amplification were an initial 3 min at 94 °C, followed by 32 cycles of 94 °C for 1 min, 53 °C for 1 min, 72 °C for 1 min, and a final extension at 72 °C for 7 min. Amplification of 30 ndhF was conducted using conditions similar to the above, except for an annealing temperature of 47 °C. Amplification of trnD-trnT was similar, except we ran 34 cycles and the final extension at 72 °C was only 5 min. Thermocycler conditions for PHYC were 5 min at 94 °C followed 28 M.F. Norup et al. / Molecular Phylogenetics and Evolution 92 (2015) 25–44 Table 1 Specimen and voucher information for the taxa included in this study. Herbarium acronyms are in accordance with Index Herbariorum (http://sciweb.nybg.org/science2/ IndexHerbariorum.asp). Voucher information for accessions in GenBank is not indicated. Note: All LPA numbers refer to numbers in Banco ADN at Jardin Canario ‘‘Viera y Clavijo’’. Species Agavoideae Agave sp. Agave parviflora Torr. Agave striata Zucc. Lomandroideae Arthropodium cirrhartum (G.Forst.) R.Br. Cordyline stricta (Sims) Endl. Lomandra hastilis (R.Br.) Ewart Lomandra sp. Thysanotus spiniger Brittan Thysanotus sp. Voucher M.W. Chase s. n. Herbarium acronym K GenBank accession no. trnH-psbA trnD-trnT 30 ndhF PHYC – – AM884850 – AF508399 – JX574227 – – – JX574446 – – – – – JX574655 – – – – AY191184 JX574027 – AY225024 – AY225026 – EU850207 – – JX574447 – JX574448 – EU850212 JX574449 – – – – – JX574656 – JX574657 – – JX574658 – – AF508404 – AY191186 JX574028 – EU850060 EU850041 JX574029 AY225018 JX574230 – JX574231 – – JX574659 – JX574233 – – – – M.W. Chase 651 Frederiksen et al. C206 M.W. Chase 2215 K C K P.J. Rudall K M.W. Chase 496 K J.C. Pires s. n. MU M.W. Chase 2053 P. Goldblatt 9463 G. Petersen C421 M.W. Chase 492 K Unknown C K H. Meusel 379 Unknown Brodiaeoideae Brodiaea coronaria (Salisb.) Jeps. J.C. Pires 96-044 Unknown – – AF508356 JX574229 Asparagoideae Hemiphylacus latifolius S. Watson Hemiphylacus alatostylus L. Asparagus acocksii Jessop Asparagus acutifolius L. Asparagus aethiopicus L. Asparagus aethiopicus L. Asparagus aethiopicus L. Asparagus africanus Lam. Asparagus africanus Lam. Asparagus aggregatus (Oberm.) Fellingham & N.L. Mey. Asparagus albus L. Asparagus altissimus Munby Asparagus angusticladus (Jessop) J.-P. Lebrun & Stork Asparagus aphyllus L. ssp. orientalis (Baker) P.H. Davis Asparagus arborescens Willd. ex Schult. & Schult.f. Asparagus arborescens Willd. ex Schult. & Schult.f. Asparagus aridicola Sebsebe Asparagus asparagoides (L.) W. Wight Asparagus asparagoides (L.) W. Wight Asparagus asparagoides (L.) W. Wight Asparagus asparagoides (L.) W. Wight Asparagus asparagoides (L.) W. Wight? Asparagus aspergillus Jessop Asparagus biflorus (Oberm.) Fellingham & N.L. Mey. Asparagus buchananii Baker Asparagus buchananii Baker Asparagus burchellii Baker Asparagus burchellii Baker Asparagus capensis var. capensis L. Asparagus capensis var. capensis L. Asparagus capensis var. capensis L. Asparagus capensis var. capensis L. Asparagus capensis var. capensis L. Asparagus capensis var. capensis L. Asparagus capensis var. litoralis Suess. & Karling Asparagus capensis var. litoralis Suess. & Karling Asparagus cf. aethiopicus L. Asparagus cf. alopecurus (Oberm.) Malcomber & Sebsebe Asparagus cf. alopecurus (Oberm.) Malcomber & Sebsebe Asparagus cf. bechuanicus Baker Asparagus cf. bechuanicus Baker Asparagus cf. concinnus (Baker) Kies Asparagus cf. cooperi Baker Asparagus cf. cooperi Baker M.W. Chase 668 K.H. Hertweck S.S. Burrows 9379 A. Marrero s. n. M.V. Norup 17 M.V. Norup 111 M.V. Norup 103 M.V. Norup 84 S. Burrows 9445 M.V. Norup 141 A. Marrero s.n. A. Marrero s.n. S. Burrows 7771 G.&J. Petersen 06-11 J. Caujape et al. s.n. Acevedo & Siverio s.n. S. Demissew s.n. M.V. Norup 55 M.V. Norup 81 M.V. Norup 83 S. Burrows 7762 S. Burrows 8184 M.V. Norup 148 Edwards & Craig s.n. S. Burrows 9455 M.V. Norup 140 S. Burrows 8209 M.V. Norup 73 M.V. Norup 26 S. Burrows 9422 M.V. Norup 25 M.V. Norup 89 M.V. Norup 52 S. Burrows 9404 M.V. Norup 48 M.V. Norup 49 M.V. Norup 7 M.V. Norup 54 M.V. Norup 60 S. Burrows 9342 M.V. Norup 136 M.V. Norup 110 S. Burrows 9715 S. Burrows 9675 K Unknown BNRH LPA 1037 C C C C BNRH C LPA 1040 LPA 1060 BNRH C LPA 1735 LPA 1714 ETH C C C BNRH BNRH C No Voucher BNRH C BNRH C C BNRH C C C BNRH C C C C C BNRH C C BNRH BNRH JX574445 – JX574290 JX574271 JX574294 JX574311 JX574312 JX574336 JX574354 JX574327 JX574276 JX574267 JX574326 JX574256 JX574272 JX574273 JX574285 JX574236 JX574237 JX574238 JX574239 JX574240 JX574277 JX574283 JX574321 JX574322 JX574397 JX574398 JX574387 JX574389 JX574390 JX574391 JX574395 JX574396 JX574393 JX574394 JX574298 JX574241 JX574242 JX574344 JX574350 JX574367 JX574337 JX574338 JX574654 – – JX574486 JX574507 JX574524 JX574525 JX574548 JX574565 JX574539 JX574491 JX574482 – JX574471 JX574487 JX574488 JX574500 JX574451 JX574452 JX574453 JX574454 JX574455 JX574492 JX574498 JX574534 JX574535 JX574608 JX574609 JX574598 JX574600 JX574601 JX574602 JX574606 JX574607 JX574604 JX574605 JX574511 JX574456 JX574457 JX574555 JX574561 JX574578 JX574549 – AY225020 – JX573880 JX573863 JX573884 JX573900 JX573901 JX573921 JX573939 JX573912 JX573867 JX573859 JX573911 JX573849 – JX573864 JX573875 JX573829 JX573830 JX573831 JX573832 JX573833 JX573868 – JX573907 JX573908 JX573980 JX573981 JX573970 JX573972 JX573973 JX573974 JX573978 JX573979 JX573976 JX573977 JX573888 JX573834 JX573835 JX573929 JX573935 JX573952 JX573922 JX573923 – JX574660 JX574079 JX574062 JX574083 JX574100 JX574101 JX574124 JX574141 JX574115 JX574066 JX574059 JX574114 JX574051 – JX574063 JX574074 JX574031 JX574032 JX574033 JX574034 JX574035 JX574067 JX574072 – JX574110 JX574181 JX574182 JX574171 JX574173 JX574174 JX574175 JX574179 JX574180 JX574177 JX574178 JX574087 JX574036 JX574037 JX574132 JX574138 – JX574125 JX574126 Nolinoideae Convallaria keiskei Miq. Convallaria majalis L. Dracaena reflexa Lam. Dracaena aubryana Brongn. ex Morren Eriospermum parvifolium Jacq. Eriospermum cernuum Baker Maianthemum bifolium (L.) F.W. Schmidt Polygonatum hookeri Baker Polygonatum cirrhifolium (Wall.) Royle Ruscus hypophyllum L. Ruscus aculeatus L. JX574228 – JX574232 JX574234 29 M.F. Norup et al. / Molecular Phylogenetics and Evolution 92 (2015) 25–44 Table 1 (continued) Species Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Asparagus Voucher cf. cooperi Baker cf. cooperi Baker cf. cooperi Baker cf. cooperi Baker cf. cooperi Baker cf. cooperi Baker cf. cooperi Baker cf. cooperi Baker cf. cooperi Baker cf. cooperi Baker cf. cooperi Baker cooperi Baker cooperi Baker cooperi Baker ’forma’ cooperi Baker sensu stricto cooperi Baker sensu sticto cooperi Baker sensu sticto cooperi Baker sensu stricto cf. juniperoides Engl. cf. laricinus Burch. cf. suaveolens Burch. cf. suaveolens Burch. cf. suaveolens Burch. cf. suaveolens Burch. cf. suaveolens Burch. cf. suaveolens Burch. cf. suaveolens Burch. cf. suaveolens Burch. cf. suaveolens Burch. cf. suaveolens Burch. cf. suaveolens Burch. cf. suaveolens Burch. cf. suaveolens Burch. clareae (Oberm.) Fellingham & N.L. Mey. cochinchinensis (Lour.) Merr. coddii (Oberm.) Fellingham & N.L. Mey. concinnus (Baker) Kies confertus K. Krause cf. bechuanicus crassicladus Jessop crassicladus Jessop declinatus L. declinatus L. declinatus L. densiflorus (Kunth) Jessop densiflorus (Kunth) Jessop densiflorus (Kunth) Jessop densiflorus ‘‘Licuati’’ denudatus (Kunth) Baker devenishii (Oberm.) Fellingham & N.L. Mey. divaricatus (Oberm.) Fellingham & N.L. Mey. edulis (Oberm.) J.-P. Lebrun & Stork elephantinus S.M.S. Burrows exsertus (Oberm.) Fellingham & N.L. Mey. exsertus (Oberm.) Fellingham & N.L. Mey. exuvialis Burch. exuvialis Burch. falcatus L. falcatus L. fasciculatus Thunb. filicinus Buch.-Ham. ex D.Don filicladus (Oberm.) Fellingham & N.L. Mey. flagellaris (Kunth) Baker flagellaris (Kunth) Baker flavicaulis (Oberm.) Fellingham & N.L. Mey. fouriei (Oberm.) Fellingham & N.L. Mey. fractiflexus (Oberm.) Fellingham & N.L. Mey. graniticus (Oberm.) Fellingham & N.L. Mey. hirsutus S.M.S. Burrows humilis Engl. inderiensis Blume ex Ledeb. sp. indet. 2 (‘‘mueda’’) intricatus (Oberm.) Fellingham & N.L. Mey. kiusianus Makino S. Burrows 9151 S. Burrows 9526 S. Burrows 9569 M.V. Norup 139 M.V. Norup 144 M.V. Norup 151 S. Burrows 9446 M.V. Norup 135 S. Burrows 9527 M.V. Norup 119 M.V. Norup 122 M.V. Norup 27 S. Burrows 9501 S. Burrows 7790b S. Burrows 7792 S. Burrows 9529 S. Burrows 9553 M.V. Norup 127 M.V. Norup 56 M.V. Norup 137 M.V. Norup 88 S. Burrows 9420 M.V. Norup 112 S. Burrows 9429 M.V. Norup 11 M.V. Norup 12 M.V. Norup 65 M.V. Norup 19 M.V. Norup 97 M.V. Norup 18 M.V. Norup 23 S. Burrows 9418 M.V. Norup 104 M.V. Norup 40 S. Burrows 7922 S. Burrows 9572 S. Burrows 9353 S. Burrows 9539 S. Burrows 8370 M.V. Norup 114 M.V. Norup 61 S. Burrows 8465 S. Burrows 9413 M.V. Norup 113b S. Burrows 7823 M.V. Norup 80 S. Burrows 8155 S. Burrows 8392 M.V. Norup 134 M.V. Norup 41 S. Burrows 7757 S. Burrows 8781 M.V. Norup 85 M.V. Norup 79 M.V. Norup 143 M.V. Norup 22 S. Burrows 8153 M.V. Norup 133 M.V. Norup 67 M. Suzuki et al. s.n. S. Burrows 9451 S. Burrows 9708 S. Demissew s.n. S. Burrows 9003 S. Burrows 8784 S. Burrows 9511 M.V. Norup 50 S. Burrows 5352 S. Burrows 9755 Sagalev s. n. S. Burrows 9786 M.V. Norup 29 Watanabe 93 Herbarium acronym BNRH BNRH BNRH C C C BNRH C BNRH C C C BNRH BNRH BNRH C BNRH C C C C BNRH C BNRH C C C C C C C BNRH C C No voucher BNRH BNRH BNRH BNRH BNRH C C BNRH BNRH C BNRH C BNRH BNRH C C BNRH BNRH C C C C BNRH C C TUS BNRH BNRH ETH BNRH BNRH BNRH C BNRH BNRH Unknown BNRH C Unknown GenBank accession no. trnH-psbA trnD-trnT 30 ndhF PHYC JX574339 JX574345 JX574346 JX574347 JX574348 JX574349 JX574355 JX574356 JX574357 JX574358 JX574359 JX574342 JX574343 JX574341 JX574360 JX574433 JX574434 JX574435 JX574243 JX574370 JX574401 JX574402 JX574412 JX574413 JX574421 JX574413 JX574414 JX574426 JX574427 JX574429 JX574430 JX574431 JX574432 JX574299 JX574259 JX574325 JX574368 JX574300 JX574351 JX574291 JX574292 JX574246 JX574247 JX574248 JX574308 JX574309 JX574310 JX574441 JX574269 JX574340 JX574380 JX574352 JX574287 JX574386 JX574388 JX574318 JX574319 JX574288 JX574289 JX574249 JX574257 JX574293 JX574252 JX574253 JX574414 JX574281 JX574268 JX574314 JX574323 JX574275 JX574263 JX574328 JX574378 JX574260 JX574550 JX574556 JX574557 JX574558 JX574559 JX574560 JX574566 JX574567 JX574568 JX574569 JX574570 JX574553 JX574554 JX574552 JX574571 JX574642 JX574643 JX574644 JX574458 JX574581 JX574612 JX574613 JX574621 JX574622 JX574630 JX574631 JX574632 JX574635 JX574636 JX574638 JX574639 JX574640 JX574641 JX574512 JX574474 JX574538 JX574579 JX574513 JX574562 JX574504 JX574505 JX574461 JX574462 JX574463 JX574521 JX574522 JX574523 JX574650 JX574484 JX574551 JX574591 JX574563 JX574502 JX574597 JX574599 JX574531 JX574532 – JX574503 JX574464 JX574472 JX574506 JX574467 JX574468 JX574623 JX574496 JX574483 JX574527 JX574536 JX574490 JX574478 JX574540 JX574589 JX574475 JX573924 JX573930 JX573931 JX573932 JX573933 JX573934 JX573940 JX573941 JX573942 JX573943 JX573944 JX573927 JX573928 JX573926 JX573945 JX574015 JX574016 JX574017 JX573836 JX573955 JX573984 JX573985 JX573995 JX573996 JX574004 JX574005 JX574006 JX574009 JX574010 JX574011 JX574012 JX574013 JX574014 JX573889 JX573852 JX573910 JX573953 JX573890 JX573936 JX573881 JX573882 JX573839 JX573840 JX573841 JX573897 JX573898 JX573899 JX574023 JX573861 JX573925 JX573964 JX573937 JX573877 JX573969 JX573971 JX573905 JX573906 JX573878 JX573879 JX573842 JX573850 JX573883 JX573845 JX573846 JX573997 JX573872 JX573860 JX573902 – JX573866 JX573855 JX573913 JX573962 JX573853 JX574127 JX574133 JX574134 JX574135 JX574136 JX574137 JX574142 JX574143 JX574144 JX574145 JX574146 JX574130 JX574131 JX574129 JX574147 JX574216 JX574217 JX574218 JX574038 JX574156 JX574185 JX574186 JX574196 JX574197 JX574204 JX574205 JX574206 JX574209 JX574210 JX574212 JX574213 JX574214 JX574215 JX574088 JX574053 JX574113 JX574154 JX574089 JX574139 JX574080 JX574081 JX574041 JX574042 JX574043 JX574097 JX574098 JX574099 JX574223 JX574061 JX574128 JX574166 JX574140 JX574076 JX574170 JX574172 JX574107 JX574108 JX574077 JX574078 JX574044 – JX574082 JX574047 JX574048 JX574198 JX574070 JX574060 JX574103 JX574111 JX574065 – JX574116 JX574164 JX574054 (continued on next page) 30 M.F. Norup et al. / Molecular Phylogenetics and Evolution 92 (2015) 25–44 Table 1 (continued) Species Asparagus krebsianus (Kunth) Jessop Asparagus krebsianus (Kunth) Jessop Asparagus krebsianus or confertus Asparagus laricinus Burch. Asparagus lignosus Burm.f. Asparagus lignosus Burm.f. Asparagus lignosus or laricinus Asparagus lynettae (Oberm.) Fellingham & N.L. Mey. Asparagus macowanii Baker Asparagus mariae (Oberm.) Fellingham & N.L. Mey. Asparagus maritimus (L.) Mill. Asparagus microraphis (Kunth) Baker Asparagus microraphis (Kunth) Baker Asparagus minutiflorus (Kunth) Baker Asparagus sp. indet. Asparagus mucronatus Jessop Asparagus mucronatus Jessop Asparagus multiflorus Baker Asparagus myriocladus Baker Asparagus natalensis (Baker) J.-P. Lebrun & Stork Asparagus natalensis (Baker) J.-P. Lebrun & Stork Asparagus nelsii Schinz Asparagus nelsii Schinz Asparagus nesiotes ssp. purpuriensis Marrero Rodr. & A. Ramos Asparagus pauli-guilelemi Solms Asparagus officinalis L. Asparagus oligoclonos Maxim. Asparagus ovatus T.M. Salter Asparagus oxyacanthus Baker Asparagus oxyacanthus Baker Asparagus petersianus Kunth Asparagus cf. racemosus Willd. Asparagus racemosus Willd. Asparagus radiatus Sebsebe Asparagus ramosissimus Baker Asparagus recurvispinus (Oberm.) Fellingham & N.L. Mey. Asparagus recurvispinus (Oberm.) Fellingham & N.L. Mey. Asparagus retrofractus L. Asparagus rigidus Jessop Asparagus rubicundus P.J. Bergius Asparagus scandens Thunb. Asparagus schoberioides Kunth Asparagus sekukuniensis (Oberm.) Fellingham & N.L. Mey. Asparagus setaceus (Kunth) Jessop Asparagus setaceus (Kunth) Jessop Asparagus sp. D Asparagus spinescens Steud. ex Schult. & Schult.f. Asparagus spinescens Steud. ex Schult. & Schult.f. Asparagus spinescens Steud. ex Schult. & Schult.f. Asparagus stellatus Baker Asparagus stipulaceus Lam. Asparagus stipularis Forssk. Asparagus striatus (L.f.) Thunb. Asparagus suaveolens Burch. Asparagus suaveolens Burch. Asparagus suaveolens Burch. Asparagus suaveolens Burch. Asparagus suaveolens Burch. Asparagus suaveolens Burch. Asparagus suaveolens Burch. Asparagus suaveolens Burch. Asparagus suaveolens Burch. Asparagus suaveolens Burch. Asparagus suaveolens Burch. Asparagus suaveolens Burch. Asparagus suaveolens Burch. Asparagus subulatus Thunb. Asparagus sylvicola S.M.S. Burrows Asparagus transvaalensis (Oberm.) Fellingham & N.L. Mey. Asparagus umbellatus ssp. umbellatus Link Asparagus virgatus Baker Asparagus virgatus Baker Asparagus virgatus Baker Voucher Herbarium acronym GenBank accession no. trnH-psbA trnD-trnT 30 ndhF PHYC M.V. Norup 108 S. Burrows 9403 M.V. Norup 100 M.V. Norup 28 S. Burrows 8761 M.V. Norup 93 M.V. Norup 86 M.V. Norup 36 S. Burrows 8159 S. Burrows 9423 Fukuda 041402 S. Burrows 9442 M.V. Norup 123 S. Burrows 7885 S. Burrows 7817 M.V. Norup 9 M.V. Norup 69 M.V. Norup 20 S. Burrows 10090 S. Burrows 8081 S. Burrows 9481 M.V. Norup 142 M.V. Norup 35 Scholz & Jaen s.n. C BNRH C C BNRH C C C BNRH BNRH C BNRH C BNRH BNRH C C C BNRH BNRH BNRH C C LPA 2285 JX574295 JX574301 JX574297 JX574369 JX574365 JX574366 JX574364 JX574307 JX574331 JX574424 JX574261 JX574372 JX574373 JX574284 JX574437 JX574332 JX574333 JX574335 JX574306 JX574279 JX574280 JX574315 JX574316 JX574270 JX574508 JX574514 JX574510 JX574580 JX574576 JX574577 JX574575 JX574520 JX574543 JX574633 JX574476 JX574583 JX574584 JX574499 JX574646 JX574544 JX574545 JX574547 JX574519 JX574494 JX574495 JX574528 JX574529 JX574485 JX573885 JX573891 JX573887 JX573954 JX573950 JX573951 JX573949 JX573896 JX573916 JX574007 JX573854 JX573956 JX573957 JX573874 JX574019 JX573917 JX573918 JX573920 – JX573870 JX573871 JX573903 JX573904 JX573862 JX574084 JX574090 JX574086 JX574155 JX574152 JX574153 JX574151 JX574096 JX574119 JX574207 JX574055 JX574158 JX574159 JX574073 JX574220 JX574120 JX574121 JX574123 JX574095 – JX574069 JX574104 JX574105 – S. Burrows 9774 G. Petersen & O. Seberg C471 Deguchi 8090 S. Burrows 9424 S. Burrows 9433 M.V. Norup 115 S. Burrows 8852 S. Burrows 8872 Sebsebe s.n. S. Burrows 9653 S. Burrows 7759 S. Burrows 9365 M.V. Norup 99 M.V. Norup 66 S. Burrows 7791 M.V. Norup 64 M.V. Norup 68 M. Maki s.n. M.V. Norup 34 M.V. Norup 101 S. Burrows 9371 S. Burrows 7761 S. Burrows 9430 S. Burrows 9434 S. Burrows 9449 S. Burrows 10073 M.V. Norup 82 A. Marrero s.n. M.V. Norup 107 S. Burrows 7884 S. Burrows 7744 S. Burrows 8197 S. Burrows 8527 S. Burrows 9346 M.V. Norup 131 M.V. Norup 138 M.V. Norup 145 M.V. Norup 31 M.V. Norup 43 M.V. Norup 46 S. Burrows 9408 M.V. Norup 128 M.V. Norup 105 S. Burrows 7895 S. Burrows 8926 Scholz s.n. S. Burrows 7760 M.V. Norup 130 M.V. Norup 150 BNRH C Unknown BNRH BNRH C BNRH BNRH JX574255 JX574262 JX574264 JX574235 JX574329 JX574330 JX574254 JX574320 JX574317 JX574286 JX574250 JX574383 JX574384 JX574362 JX574376 JX574374 JX574251 JX574265 JX574282 JX574438 JX574439 JX574361 JX574415 JX574416 JX574428 JX574371 JX574392 JX574258 JX574377 JX574399 JX574403 JX574400 JX574404 JX574405 JX574406 JX574407 JX574408 JX574409 JX574410 JX574411 JX574420 JX574425 JX574379 JX574440 JX574324 JX574274 JX574442 JX574443 JX574444 JX574470 JX574477 JX574479 JX574450 JX574541 JX574542 JX574469 JX574533 JX574530 JX574501 JX574465 JX574594 JX574595 JX574573 JX574587 JX574585 JX574466 JX574480 JX574497 JX574647 JX574648 JX574572 JX574624 JX574625 JX574637 JX574582 JX574603 JX574473 JX574588 JX574610 JX574614 JX574611 JX574615 JX574616 JX574617 JX574618 JX574619 JX574620 – – JX574629 JX574634 JX574590 JX574669 JX574537 JX574489 JX574651 JX574652 JX574653 JX573849 – JX573856 JX573828 JX573914 JX573915 JX573847 – – JX573876 JX573843 JX573966 JX573967 JX573947 JX573960 JX573958 JX573844 JX573857 JX573873 JX574020 JX574021 JX573946 JX573998 JX573999 – – JX573975 JX573851 JX573961 JX573982 JX573986 JX573983 JX573987 JX573988 JX573989 JX573990 JX573991 JX573992 JX573993 JX573994 JX574003 JX574008 JX573963 JX574022 – JX573865 JX574024 JX574025 JX574026 JX574050 JX574056 – JX574030 JX574117 JX574118 JX574049 JX574109 JX574106 JX574075 JX574045 JX574167 JX574168 JX574149 JX574162 JX574160 JX574046 JX574057 JX574071 JX574221 JX574222 JX574148 JX574199 JX574200 JX574211 JX574157 JX574176 JX574052 JX574163 JX574183 JX574187 JX574184 JX574188 JX574189 JX574190 JX574191 JX574192 JX574193 JX574194 JX574195 JX574203 JX574208 JX574165 – JX574112 JX574064 JX574224 JX574225 JX574226 BNRH BRNH BNRH C C BNRH C C TUS C C BNRH BNRH BNRH BNRH BNRH BNRH C LPA 1038 C BNRH BNRH BNRH BNRH BNRH C C C C C C BNRH C C BNRH BNRH LPA 1731 BNRH C C (continued on next page) 31 M.F. Norup et al. / Molecular Phylogenetics and Evolution 92 (2015) 25–44 Table 1 (continued) Species Voucher Herbarium acronym GenBank accession no. trnH-psbA trnD-trnT 30 ndhF PHYC Asparagus volubilis (L.f.) Thunb. Asparagus volubilis (L.f.) Thunb. S. Burrows 9425 M.V. Norup 95 BNRH C JX574244 JX574245 JX574459 JX574460 JX573837 JX573838 JX574039 JX574040 Manuscript names Asparagus ‘‘arenosus’’ Asparagus ‘‘barbertonicus’’ Asparagus ‘‘candelus’’ Asparagus ‘‘ferox’’ Asparagus ‘‘iinoinatus’’ Asparagus ‘‘jessopii’’ Asparagus ‘‘karooicus’’ Asparagus ‘‘longissimus’’ Asparagus ‘‘macrocarpa’’ Asparagus ‘‘muedaensis’’ Asparagus ‘‘odoratus’’ Asparagus ‘‘petiolatus’’ Asparagus plocamoides Webb ex Svent. Asparagus ‘‘pongolanus’’ Asparagus ‘‘praetermissus’’ Asparagus ‘‘pseudoconfertus’’ Asparagus ‘‘vulpicaudatus’’ Asparagus ‘‘vulpicaudatus’’ Asparagus ‘‘Zululand’’ M.V. Norup 47 S. Burrows 8542 M.V. Norup 30 S. Burrows 9439 S. Burrows 9762 S. Burrows 10304 S. Burrows 9428 S. Burrows 7846 S. Burrows 10060 S. Burrows 9778 S. Burrows 8991 Winston Park ex Hort. Bot. Acevedo & Siverio s.n. S. Burrows 9530 S. Burrows 9360 S. Burrows 9363 S. Burrows 9440 S. Burrows 7820 S. Burrows 9537 C BNRH C BNRH BNRH BNRH BNRH BNRH BNRH BNRH BNRH No voucher LPA 1163 BNRH BNRH BNRH BNRH BNRH BNRH JX574363 JX574334 JX574417 JX574418 JX574278 JX574302 JX574385 JX574419 JX574303 JX574353 JX574382 JX574381 JX574266 JX574375 JX574313 JX574296 JX574304 JX574305 JX574436 JX574574 JX574546 JX574626 JX574627 JX574493 JX574515 JX574596 JX574628 JX574516 JX574564 JX574593 JX574592 JX574481 JX574586 JX574526 JX574509 JX574517 JX574518 JX574645 JX573948 JX573919 JX574000 JX574001 JX573869 JX573892 JX573968 JX574002 JX573893 JX573938 JX573965 – JX573858 JX573959 – JX573886 JX573894 JX573895 JX574018 JX574150 JX574122 JX574201 JX574202 JX574068 JX574091 JX574169 – JX574092 – – – JX574059 JX574161 JX574102 JX574085 JX574093 JX574094 JX574219 Table 2 Primer regions used for PCR amplification and sequencing. Marker trnH-psbA 30 ndhF trnD-trnT PHYC (degenerate) PHYC (Asparagus specific) Primer name GUG /psbA trnH 1318F/2110R GUC trnD /trnTGGU Asp_PhyC_F1/Asp_PhyC_R1 PHYC-F1/PHYC-R1 by 39 cycles of 94 °C for 1 min, 55 °C for 1 min, and 72 °C for 1.5 min followed by a final extension at 72 °C for 7 min. Cycle sequencing was performed using the ABI PRISM BigDye Terminator Cycle Sequencing Ready Reaction Kit (Applied Biosystems, Foster City, CA, USA), and the products were purified using the DyeEX 96 kit (Qiagen) according to the manufacturer’s instructions. The cycle sequencing reactions were run on a Bio-Rad DNA-Engine Thermal Cycler (Bio-Rad Laboratories). The protocol consisted of 30 s at 96 °C followed by 29 cycles of 10 s at 96 °C, 15 s at 50 °C and 4 min at 60 °C. Cleaned cycle-sequencing products were analyzed using an automatic sequencer 3130xl Genetic Analyzer (Applied Biosystems). Both strands were analyzed for each region for all taxa, however, despite repeated attempts, some accessions failed to amplify for some regions. As all of these accessions did amplify for at least one region, we judged this sufficient to place them in a phylogenetic context. 2.3. Alignment and phylogenetic analysis Sequences were edited and assembled using Sequencer 4.9 (Gene Codes, Ann Arbor, Michigan, USA). All sequences were initially aligned using ClustalX ver.2 (Larkin et al., 2007) using default settings, and followed by manual adjustment in MacClade 4.08 (Maddison and Maddison, 2005). Ambiguous parts of the alignment (e.g., some rather long homopolymer regions) and parsimony uninformative characters were excluded from the analysis. The four regions were analyzed separately to identify phylogenetic conflicts prior to performing a combined analysis. In addition, the data were also partitioned into cpDNA and nrDNA datasets. A ‘‘hard’’ incongruence test was performed by directly comparing Reference/sequence Shaw et al. (2005) Olmstead and Sweere (1994) Shaw et al. (2005) Hertweck et al. (2015) 50 CAG TTA ACC CTG CTG ATG TAC C 30 ; 50 ACC TCG CCA CTT TAC AAC CT 30 topologies from the separate analyses (i.e., looking for incongruent clades with bootstrap percentages (BP) > 85 (Sheahan and Chase, 2000; Wiens, 1998)). Only one clade was found to be incongruent, and consequently the data partitions were combined. We searched for the optimal cladogram for each dataset using parsimony as implemented in PAUP⁄ 4.10b (Swofford, 2002). Due to the large size of the datasets we used the parsimony ratchet (Nixon, 1999) as implemented in PAUPRat ver.1 (Sikes and Lewis, 2001). All character changes were treated as unordered and weighted equally (Fitch, 1971), and indels treated as missing data. We performed ten parsimony ratchet searches of 200 iterations each with 15% of the characters perturbed. The trees from these searches were imported into PAUP⁄ and filtered for the best trees, and these were then used to create a strict consensus tree (Forest et al., 2007). Bootstrap values (Felsenstein, 1985) were calculated in PAUP⁄ using 1000 replicates of simple taxon addition, TBR swapping and retaining a maximum of 10 trees per replicate. The combined dataset was also analyzed under maximum likelihood (ML). A best fitting model of molecular evolution was selected using jModeltest 0.1.1 (Guindon and Gascuel, 2003; Posada, 2008). The GTR + I + G model thus selected was then used in an ML analysis using GARLI v. 0.96 (Zwickl, 2006). As recommended by Zwickl (2006), multiple runs were performed to test whether the results were consistent. The log-likelihood values of each run were retained in order to compare the individual runs, and the tree corresponding to the best score was chosen as the ML tree. ML bootstrap values were estimated from 100 bootstrap replicates in GARLI. We refer to nodes with 50–74% bootstrap support as weakly supported, 75–84% as moderately supported and 85–100% as strongly supported (Chase et al., 2000). 32 M.F. Norup et al. / Molecular Phylogenetics and Evolution 92 (2015) 25–44 2.4. Mapping plant reproductive mode The reproductive mode of the included Asparagaceae species were scored from own observations in the field, from personal comments (Julia Pérez de Paz, Jardín Botánico Canario ‘‘Viera y Clavijo’’), from observations in C, and from the literature (Baker, 1875; Demissew et al., 2008; Feinbrun-Dothan, 1986; Hernandez, 1995; Jessop, 1966; Kunth, 1850; Malcomber and Demissew, 1993; Obermeyer, 1983, 1984; Obermeyer and Immelman, 1992; Roemer and Schultes, 1829; Sánchez Gómez et al., 2009; Santana et al., 2004; Valdés, 1979, 1980). Species were scored for the following character states: (1) bisexual, (2) dioecious, (3) gynodioecious or (4) having an ‘‘unidentified sexual dimorphism’’ (for species in which dimorphism was noted but not specified). Using Mesquite version 2.72 (Maddison and Maddison, 2009) character states were optimized onto one of the most parsimonious trees from the combined analysis. However, due to the consistency in topology between the parsimony tree topology and the best scoring ML tree (see Section 3.2), the result will be shown on the ML tree. 2.5. Biogeographic analysis In order to reconstruct the historical biogeography for Asparagus despite phylogenetic uncertainty, we used the program S-DIVA 1.1b (Yu, 2009). This implements the same methodology as Bayes-DIVA (Nylander et al., 2008). In S-DIVA the ancestral reconstructions are averaged over all trees and each alternative ancestral area at each node is weighted by the probability (S-DIVA; Yu et al., 2010), thus generating credibility support values for alternative biogeographical scenarios. S-DIVA has the same computational constraints as DIVA 1.2 (Ronquist, 1997), and can include only trees with fewer than 128 terminals and maximally 15 areas. The number of terminals on the MP trees from the combined dataset were therefore trimmed by (1) reducing monophyletic species to one accession each (this does not change the distribution pattern), (2) deleting undescribed taxa without provisional names (the distribution areas of these species were unknown), (3) deleting terminals that had not been referred to a particular species (e.g. the sample referred to as ‘‘krebsianus’’ or ‘‘confertus Norup 100’’) (these have the same distribution pattern, and the deletion would not change the overall pattern in the clades), (4) reducing monophyletic groups of accessions of the same species or ‘‘cf. species’’ (also within the polyphyletic species) to one accession (these all have the same distribution pattern, and the reduction to one accession would not change the overall pattern in the clades), 5) by retaining only a single accession of each ‘‘cf. species’’ (see above) and 6) combining sister taxa with same distribution (e.g. cf. alopecurus/cf. juniperoides). S-DIVA requires fully resolved phylogenies, and as the trees from the combined analysis included a few non-resolved nodes, e.g. in the heavily sampled suaveolens and cooperi clades (however, see below), the trees were dichotomized using R, ver. 2.10 (Team, 2009). The 15 areas (northern Europe, southern Europe, North Africa, Macaronesia, West Africa, East Africa, South Central Africa, Madagascar, southern Africa, Siberia, Middle East and western Asia, Far East, India to the Philippines, Australasia and the Americas) were simplified from the areas from which Asparagaceae is recorded by Govaerts et al. (2009) by combining areas with similar species compositions. Analyses allowing an unconstrained number of ancestral area combinations resulted in uninterpretable results (i.e. > eight ancestral areas) for some nodes, thus we constrained the maximum number of areas in ancestral distributions to two (see also del Hoyo et al., 2009; Guo and Wang, 2007; Nylander et al., 2008). As only 18 of the 93 investigated species are found in more than two areas, a constraint of two areas for ancestral nodes is reasonable. Species ranges were recorded from Govaerts et al. (2009). These maps are biased against the widespread species, which also occur north of southern Africa. Because S-DIVA does not take branch lengths into account when inferring ancestral ranges (Clark et al., 2008), we also used the dispersal-extinction-cladogenesis (DEC) analysis (Ree and Smith, 2008a,b). It is a continuous-time model for geographic range evolution in which dispersal events cause range expansion, local extinction events cause range contraction and the probability of each event along a phylogenetic internode is proportional to evolutionary time (Clark et al., 2008). To obtain branch lengths and ultrametricize the phylogeny, we used BEAST v 1.7.5 (Drummond et al., 2006, 2012; Drummond and Rambaut, 2007). Because we do not have an easily justifiable primary calibration point, and because the biogeography reconstruction only need relative branch lengths, we constrained the root node to 1.0 under a normal distribution. We performed one run of 100,000,000 chains, sampling every 1000 generations. After removing 30,000,000 burn-in samples, we summarized the results using a Maximum Clade Credibility (MCC) tree. DEC analysis was performed using LAGRANGE (Ree and Smith, 2008a) on the MCC tree. We wanted to test the ‘‘out of Africa’’ hypothesis, and DEC requires relatively few areas, which are occupied by numerous species, to obtain robust results. Consequently the 15 areas used in the S-Diva analysis were reduced to five as follows: (A) Northern and southern Europe, (B) North Africa and Macaronesia, (C) Western, Eastern and Central Africa including Madagascar, (D) southern Africa and (E) Siberia, Asia, southern Asia, Australasia and the Americas. Maximum range sizes and dispersals were unrestricted. Area optimizations were reported and considered significant only if the fraction of the global likelihood at each split exceeded 0.5 (Clark et al., 2008; Drummond and Rambaut, 2007; Drummond et al., 2006, 2012; Ree and Smith, 2008a,b). 3. Results 3.1. Phylogeny – infrageneric delimitation and species circumscription The phylogenetic hypotheses (the strict consensus trees from the sets of bootstrapped trees) for the individual cpDNA data sets were largely congruent, except for a single strongly supported conflict. In the ndhF data set, one accession of doubtful identity, here tentatively assigned to A. asparagoides (Burrows 8184) is placed as sister to another accession of A. asparagoides (Burrows 7762) with strong bootstrap support (88 BP%), whereas in the trnD-T data set it is even stronger supported (98 BP) as sister to A. ovatus (Burrows 9424). Combining the cpDNA datasets resulted in an increase in the resolution and the number of moderately to strongly supported nodes (P75 BP), compared to each of the individual analyses (bootstrap consensus shown in Fig. S1, Suppl. Mat.; Table 3). Compared to the cpDNA sequence data, the PHYC data (bootstrap consensus shown in Fig. S2, Suppl. Mat.) included less than half the number of characters, but the percentage of variable characters is 2.3 times larger, and in the most parsimonious tree there are 32 (rather than 22) nodes with more than 75 BP (Table 3). There was no strongly supported conflict between the bootstrap consensus trees from the combined cpDNA data set and from the PHYC-dataset. The matrix of the combined cpDNA and nrDNA included a total of 3969 characters. Of these, 1080 were variable and 633 (18.1%) were parsimony informative. The combined MP analysis yielded 1904 shortest trees of length 1828, with a CI = 0.49 and an RI = 0.81 (Table 3, Figs. S3a–3c, Suppl. Mat.). The resolution and number of highly supported nodes in the combined tree increased 33 M.F. Norup et al. / Molecular Phylogenetics and Evolution 92 (2015) 25–44 Table 3 Statistics of the included datasets (variable and PI characters calculated after exclusion of ambiguous data). trnD-trnT trnH-psbA ndhF Combined cpDNA PHYC cpDNA + PHYC Aligned matrix length Included characters Variable characters Parsimony informative characters (PICs) CI RI Length of shortest tree Number of nodes with P75BP 1314 677 803 2794 1175 3969 1222 585 731 2538 962 3500 231 97 306 634 446 1080 126 (10.3%) 41 (7%) 158 (21.6%) 325 (12.8%) 308 (32%) 633 (18.1%) 0.783 0.838 0.700 0.609 0.458 0.489 0.913 0.916 0.860 0.870 0.786 0.812 302 130 523 768 1080 1828 9 4 14 22 32 52 significantly compared to both the combined cpDNA analysis and the PHYC analysis. The parsimony tree topology is congruent with the best scoring ML tree ( ln = 21976.467), with the latter being better resolved. We therefore base the following discussion on the ML tree (Figs. 2a–2c, with MP bootstrap support indicated above the branches). The combined analysis strongly supports the sister group relationship between Hemiphylacus and Asparagus (100 BP, Fig. 2a) and Asparagus itself is clearly monophyletic (100 BP), too. The Setaceus clade is weakly supported as sister to the remaining species of Asparagus (70 BP). Several other major, moderately to strongly supported clades are resolved: the Africani-Capenses clade (93 BP), the Asparagus clade (82 BP), and the Myrsiphyllum clade (82 BP), as well as weakly supported clades: the Racemose clade (64 BP) (sister to a clade of A. macowanii and A. mucronatus), and the Lignosus clade (66 BP), sister to A. scandens. The overall relationships among these major clades within Asparagus are not strongly supported. Within the clades, the internal relationships receive varying support. The internal structure within the Africani-Capenses clade is generally strongly supported. Thus within the Africani clade (65 BP, Fig. 2c) the Cooperi clade (as defined here the Cooperi clade excluded three collections (Burrows 9529, Burrows 9553 and Norup 127) actually found in the Setaceus clade), the Sympodioidi clade and the Canary Island clade (which includes the gynodioecious Canary Islands species A. plocamoides and A. acutifolius) are all strongly supported (90, 99 and 95 BP, respectively). Asparagus multiflorus is resolved in an unsupported weak sister-group relationship (<50 BP) to the Cooperi clade. Within the Capenses clade (73 BP; Fig. 2c), the Recurvispinus clade (95 BP) consisting of A. recurvispinus and the undescribed A. ‘‘karooicus’’ is sister to the remaining members of the clade. The relationships within rest of the Capenses clade are not strongly supported. The Asparagus clade (Fig. 2a) is composed of a moderately to strongly supported group (82 BP) including bisexual African species (with distributions from southern to East Africa) plus a few Mediterranean species, and a monophyletic group of all the dioecious Eurasian species included in the study (Dioecy clade, 60 BP) with moderate to strongly supported internal relationships. The Racemose clade (Fig. 2b) consists of the Racemose 1 clade (including the Falcatus clade; 80 BP), the Racemose 2 clade that includes A. racemosus and allies (with distributions in Mozambique, Ethiopia and NE South Africa) as well as the Mediterranean species A. albus. Sister to the rest of the racemose species is a strongly supported clade of A. graniticus and A. nelsii (98 BP). bootstrap support in only three cases: A. asparagoides, A. africanus and A. cooperi (however, see above). Paraphyletic species are significantly more likely to have wide distribution ranges, than to be geographically restricted (Mann–Whitney U = 19.00, Z = 2.649, p = 0.007). A few collection of A. cooperi occur in the Setacues clade making A. cooperi polyphyletic. 3.3. Analysis of plant reproductive mode Sexual dimorphism has evolved several times in Asparagus (Fig. 3). True dioecy seems to have originated twice in the Asparagus clade, within both the Dioecy clade and in a clade consisting of A. stipularis and A. aphyllus ssp. orientalis. Gynodioecy has also evolved twice, once in the Setaceus clade and once in the Canary Islands clade; the latter includes two additional taxa which hae been scored as having ‘‘unidentified sexual dimorphism’’. However, Valdés (1980) considers A. acutifolius from the Canary Island clade, dioecious, but based on our own observation in herbarium C and observations done by Julia Pérez de Paz (Jardín Botánico Canario ‘‘Viera y Clavijo’’), A. acutifolius has here been scored as gynodioecious. This is supported by Feinbrun-Dothan (1986) who considers all four Asparagus species in Flora Palaestina, excluding A. stipularis, but including A. acutifolius, polygamous. 3.4. Biogeographic analysis The S-DIVA analysis (Fig. 4) gives a 100% probability that the most recent common ancestor of Asparagus is southern African (I), and that the common ancestor of Asparagaceae subfamily Asparagoideae (sensu APG III) originated in the New World (O), more specifically in Mexico. There were at least five range expansions from southern into tropical Africa, one expansion into North Africa and Macaronesia. Eurasia was occupied at least twice from Africa. The dioecious Asian and European species included in the study originated from a single common ancestor in southern or South Central Africa. The DEC analysis (Fig. 5S, Suppl. Mat.) gives similar results, albeit with a relative probability of the basal Asparagus nodes being southern Africa of between 0.35 and 0.5, probably because several species in the first two diverging clades are more widespread in Africa. DEC finds at least five range expansions into North Africa or Macaronesia, 10 range expansions into Eurasia and numerous expansions into tropical Africa. Many of these expansions are by widespread species. 4. Discussion 3.2. Species delimitation 4.1. Phylogeny – infrageneric delimitation and species circumscription Of the almost 100 taxa included in the analyses, 69 were represented by single accessions. Hence, the species delimitation of these could not be evaluated. Of the 30 taxa with more than one accession, 21 were grouped in the same clade, and nine were largely paraphyletic. However, their paraphyly received significant 4.1.1. Relationship of Asparagus As in previous analyses of Asparagales (Fay et al., 2000; Pires et al., 2006; Seberg et al., 2012) we find the Asparagus and Hemiphylacus to be closely related and their relationship to be 34 M.F. Norup et al. / Molecular Phylogenetics and Evolution 92 (2015) 25–44 98 91 95 90 91 90 94 86 92 100 100 88 99 100 100 100 100 100 100 79 100 74 82 81 79 81 66 71 60 71 62 61 62 100 98 100 94 75 56 60 69 81 89 82 89 Asparagus 62 88 83 67 88 97 64 60 72 91 70 68 100 100 65 67 64 65 100 100 95 92 93 93 100 100 Myrsiphyllum clade Lignosus clade Asparagus clade Racemose Clade 90 87 Asparagaceae ovatus Burrows 9424 asparagoides winter rainfall Burrows 8184 asparagoides summer rainfall Burrows 7762 asparagoides Norup 55 alopecurus or juniperoides Norup 54 cf. alopecurus Norup 60 cf. juniperoides Norup 56 asparagoides Norup 81 asparagoides Norup 83 volubilis Burrows 9425 volubilis Norup 95 declinatus Norup 61 declinatus Burrows 8465 declinatus Burrows 9413 ramosissimus Burrows 7759 fasciculatus Norup 67 scandens Norup 68 lignosus or laricinus Norup 86 lignosus Norup 93 lignosus Burrows 8761 cf. concinnus Norup 110 concinnus Burrows 9572 stellatus Burrows 10073 microraphis Burrows 9442 microraphis Norup 123 laricinus Norup 28 cf. laricinus Norup 137 rubicundus Norup 64 flagellaris Burrows 9708 flagellaris Sebsebe s.n. petersianus Burrows 8852 aphyllus ssp. orientalis stipularis pauli-guilelemi Burrows 9774 filicinus schoberioides kiusianus oligoclonos Dioecy maritimus clade inderiensis officinalis cochinchinensis exuvialis Limpopo Prov. Norup 143 exuvialis W. Cape Norup 22 macowanii Burrows 8159 mucronatus Norup 9 mucronatus Norup 69 Africani-Capenses Clade 97 98 retrofractus Norup 66 ”arenosus” Norup 47 90 89 63 88 67 88 54 73 67 52 67 68 64 82 70 67 86 75 52 80 92 85 99 100 99 100 56 89 59 52 100 62 100 arborescens Gran C. umbellatus ssp. umbellatus arborescens Tenerife humilis Burrows 9755 cooperi s. str. Burrows 9529 cooperi s. str. Burrows 9553 cooperi s. str. Norup 127 "Zululand” Burrows 9537 mollis Burrows 7817 ”barbertonicus” Burrows 8542 setaceus Norup 101 setaceus Burrows 9371 sylvicola Burrows 7895 densiflorus "Licuati" Burrows 8155 virgatus Burrows 7760 virgatus Norup 130 virgatus Norup 150 Hemiphylacus Cordyline Arthropodium Thysanotus Lomandra Eriospermum Maianthemum Outgroup Convallaria Setaceus clade Ruscus Polygonatum Dracaena Agave Brodiaea Fig. 2a. The GARLI ML tree with the best log likelihood score. Bootstrap values from the analyses are indicated above (MP) and below (ML) the branches, respectively. Nodes supported by one of the analyses only have just a single value above or below the branch. Undescribed species with manuscript names are indicated as e.g. ‘‘bamboosicolus’’. Name and number following species epithets in the Southern African Asparagus is the collection name and number. Clade designations in the figure are discussed in the text. strongly supported (100 BP; Fig. 2a). Apart from the molecular support, there are certain shared morphological and embryological characters that support this sister group relationship: Rudall et al. (1998) found that the shape and histology of the fertilized ovule in Hemiphylacus resembled that of Asparagus. Furthermore, both genera possess true rhizomes and all species of Hemiphylacus bear underground tubers, as do several Asparagus species (Hernandez, 1995; Obermeyer and Immelman, 1992). M.F. Norup et al. / Molecular Phylogenetics and Evolution 92 (2015) 25–44 35 Fig. 2b. The Racemose clade from the ML analysis shown in Fig. 2a. Bootstrap values from the analyses are indicated above (MP) and below (ML) the branches, respectively. Nodes supported by one of the analyses only have just a single value above or below the branch. Undescribed species with manuscript names are indicated as e.g. ‘‘biflorus’’. Name and number following species epithets in the Southern African Asparagus is the collection name and number. Clade designations in the figure are discussed in the text. However, most of the characters shared by the two genera are plesiomorphic (Rudall et al., 1998). There are several very noticeable morphological differences which corroborate the generic distinction between Asparagus and Hemiphylacus. Asparagus species vary from herbaceous climbers to woody shrubs, with flowers placed singly, a few together, or in raceme-like inflorescences, the fruit is a berry and the leaves are reduced to scales (the plants instead bearing needle-like cladodes). In contrast, Hemiphylacus species are more reminiscent of species of Asparagaceae subfamily Agavoideae (sensu APG III) in being erect, with normal leaves arranged in a basal rosette and bearing 1–2 reduced, woody stems each producing a 0.5–1.5 m naked, racemose or panicle-like inflorescence (Hernandez, 1995). 4.1.2. Infrageneric delimitation The species of Asparagus are resolved as a strongly supported monophyletic group (100 BP). A subdivision into three separate genera (Asparagus, Protasparagus and Myrsiphyllum) sensu Obermeyer (Obermeyer, 1983,1984; Obermeyer and Immelman, 1992) is clearly not warranted, as Asparagus and Myrsiphyllum then become nested within Protasparagus (Fig. 2). Either a very large number of genera should be recognized, breaking up the paraphyletic Protasparagus, or a single genus, Asparagus, has to be maintained. In these circumstances the current taxonomic trend is to maintain a single, large genus (Humphreys and Linder, 2009). This facilitates recognition and should lead to more nomenclatural stability. Due to the high number of currently described Asparagus species, a division of the genus into series or other subgeneric divisions, as previously done (Jessop, 1966; Obermeyer and Immelman, 1992) is merited. We tested the validity of the subgeneric divisions described in the latest treatment of Asparagus (Obermeyer and Immelman, 1992) by mapping these on a species level phylogeny (Fig. 2), and our results indicate that the majority of the previous divisions cannot be readily adopted. Rather than modify the previous classification by transferring species between the sections to make them monophyletic, it might be better to use the main clades in our results to provide the backbone of a new set of subgeneric groups within Asparagus. It is possible that the southern African bias in our sampling could impact the internal structure of the phylogeny, and thus the subgeneric division based on this. However, based on morphology all non-sampled species can be readily assigned to the proposed clades. This is consistent with the observation that the sampled non-South African species are all included within strongly supported southern African clades, rather than forming separate clades. We suggest that the Asparagus species might all be accommodated in six clades. Most of these molecular defined clades can also be diagnosed morphologically. The formal reclassification will be done in a separate publication. Clade 1: Asparagus clade: This moderately supported clade (82 BP, Fig. 2a) consists of (a) the bisexual, semi-woody African species, A. exuvialis, A. flagellaris, A. petersianus and A. ‘‘pauli-guilelemi’’, (b) the black-fruited, semi-woody Mediterranean species, A. aphyllus 36 M.F. Norup et al. / Molecular Phylogenetics and Evolution 92 (2015) 25–44 Fig. 2c. The Africani-Capenses clade from the ML analysis shown in Fig. 2a. Bootstrap values from the analyses are indicated above (MP) and below (ML) the branches, respectively. Nodes supported by one of the analyses only have just a single value above or below the branch. Undescribed species with manuscript names are indicated as e.g. ‘‘candelus’’. Name and number following species epithets in the Southern African Asparagus is the collection name and number. Clade designations in the figure are discussed in the text. 37 M.F. Norup et al. / Molecular Phylogenetics and Evolution 92 (2015) 25–44 bisexual dioecious Myrsiphyllum clade scandens Norup 68 gynodioecious Lignosus clade Asparagus flagellaris Burrows 9708 flagellaris Sebsebe s.n. petersianus Burrows 8852 aphyllus ssp. orientalis stipularis pauli-guilelemi Burrows 9774 filicinus schoberioides kiusianus oligoclonos Dioecy maritimus clade inderiensis officinalis cochinchinensis exuvialis Limpopo Prov. Norup 143 exuvialis W. Cape Norup 22 Asparagus clade Racemose clade macowanii Burrows 8159 mucronatus Norup 9 mucronatus Norup 69 Asparagaceae plocamoides altissimus nesiotes ssp. purpuriensis acutifolius fractiflexus Burrows 9511 denudatus Burrows 8392 multiflorus Norup 20 Cooperi clade Sympodioidi clade Canary Island clade Africani clade AfricaniCapenses clade Capenses clade retrofractus Norup 66 ”arenosus” Norup 47 umbellatus ssp. umbellatus arborescens Gran C. arborescens Tenerife humilis Burrows 9755 cooperi s. str. Burrows 9529 cooperi s. str. Burrows 9553 cooperi s. str. Norup 127 setaceus bushveld form Burrows 9537 mollis Burrows 7817 ”Zululand” Burrows 8542 setaceus Norup 101 setaceus TOPO Burrows 9371 sylvicola Burrows 7895 densiflorus "Licuati" Burrows 8155 virgatus Burrows 7760 virgatus Norup 130 virgatus Norup 150 Hemiphylacus Outgroup Setaceus clade Fig. 3. The ML tree from Fig. 2a–c redrawn to show evolution of reproductive mode in Asparagus. Clades in which no shifts have occurred (or are known) have been collapsed. See the text for further explanation. ssp. orientalis and A. stipularis (both of which express sexual dimorphisms as they have been described as diclinous (Roemer and Schultes, 1829) and functionally unisexual (Valdés, 1979)), respectively, and (c) the truly dioecious, mainly herbaceous Eurasian species (Dioecy clade; 60 BP), belonging to subgenus Asparagus. The majority of the species in this clade are spineless; however, the tropical A. flagellaris and A. petersianus (possibly also A. ‘‘pauli-guilelemi’’) do possess leaf-derived spines. These three species are morphologically similar scramblers with orange fruits, possibly part of a species complex (to be reviewed by Burrows et al., in preparation). The sampling in the distribution range of this clade is sparse; in total it might include 4–5 species. Clade 2: Racemose clade: The Racemose clade does not receive high bootstrap support (64 BP; Fig. 2b), but includes all species with raceme-like inflorescences, many of them being scramblers or climbers. It also includes the majority of the Asparagus species with ‘‘normal’’ spines (hard, woody spines modified from leaves supporting the side-branches). Obermeyer dispersed the species from this clade in four series: Globosi p.p., Protasparagus, Racemosi and Exuviali p.p. The species within the Racemose clade also differ from the rest of the known Asparagus species in their rooting structures. Within the clade, two unequally-sized, weakly supported subclades (Racemose 1 and Racemose 2 clade, supported by 55 and 57 BP respectively; Fig. 2b) and one strongly 38 M.F. Norup et al. / Molecular Phylogenetics and Evolution 92 (2015) 25–44 Fig. 4. Phylogeny of the combined analysis showing the result of the biogeographic analysis. The trees used for the S-DIVA analysis were the 1720 MP trees. Node charts are the results of averaging ancestral reconstructions over all trees, with each alternative ancestral area at a node weighted by its probability. The range for each accession/group of accessions are given after the accession names and is from Govaerts et al. (2009). Clade designation follows Fig. 2. Areas used in the biogeographic analysis are based on the TDWG regions used in Govaerts et al. (2009): A = Northern Europe + Western Europe + Eastern Europe; B = South-western Europe + South-eastern Europe; C = Northern Africa; D = Macaronesia (Canary Islands only); E = West Tropical Africa + West-Central Tropical Africa; F = East Tropical Africa + Northeast Tropical Africa; G = South Tropical Africa; H = Western Indian Ocean (Madagascar only); I = Southern Africa; J = Siberia + Russian Far East; K = Arabian Peninsula + Middle Asia + Caucasus + Western Asia; L = China + Mongolia + Eastern Asia; M = Indian Subcontinent + Indo-China + Malesia (Philippines only); N = Australia + New Zealand; O = New World (all New World regions combined). Grey circles represent area optimizations = ABCDEFGHIJKLM. M.F. Norup et al. / Molecular Phylogenetics and Evolution 92 (2015) 25–44 supported subclade (Graniticus clade, 98 BP) can be recognized, all of which have specialized root structures. The Racemose 1 subclade, which has tubers on the lateral roots, is strongly supported in the PHYC dataset (91 BP; Fig. S2), though not so in the combined analysis. This is probably due to incongruence between the cpDNA dataset and the PHYC dataset with respect to the position of A. ‘‘inopinatus’’. The Racemose 2 subclade and the Graniticus clade lack tubers, but instead have swollen main roots (although in A. graniticus the main roots are thick and woody). Similar fleshy roots are also found in the sister group to the Racemose clade (A. mucronatus and A. macowanii), however, the roots in A. mucronatus bear tubers (absent in A. macowanii) and the flowers in these species are solitary among the cladode fascicles, not in compound inflorescences. Clade 3: The Lignosus clade (66 BP; Fig. 2a) includes six morphologically similar species with erect, woody stems, leaf-derived spines, orange anthers and usually 10–15 cladodes. Remarkably, this clade brings together species from four of Obermeyer’s series: Protasparagus, Myrsiphyllum, Globosi and Africani. Clade 4: Africani-Capenses clade: The species within this clade (93 BP; Fig. 2c) do not possess the woody leaf-derived spines that are often associated with Asparagus. In the Africani clade, the leaves are either modified into hard, woody scales (Sympodioidi clade) or forming brown, papery spine-structures (Cooperi clade), whereas in the Capenses clade the leaves are present as soft, papery scales. The Africani clade (65 BP) includes three distinctive elements: (a) the Cooperi clade (90 BP), which includes species from Obermeyer’s series Retrofracti and Africani and mainly grows in exposed habitats (i.e. to sun, wind and weather impacts), and often is thicket-forming, with papery spine-structures (sometimes suppressed) and red berries; (b) the Sympodioidi clade (99 BP), which corresponds to Obermeyer’s series Sympodioidi (Obermeyer and Immelman, 1992) and is easily identified by combing divaricating stems with hard, woody scales and umbellate inflorescences and (c) the Canary Island clade (95 BP) which includes semi-woody and woody species from southern Africa (A. denudatus and A. fractiflexus), West and North Africa (A. altissimus) and the Canary Islands (the gynodioecious species A. nesiotes, A. plocamoides and A. acutifolius). We have not been able to identify morphological characters supporting this, by molecular data strongly supported, clade. The Capenses clade is a molecularly weakly (73 BP; Fig. 2c), but morphologically strongly, supported clade, congruent with series Suaveolens (Obermeyer and Immelman, 1992). The species in the Suaveolens clade are low, woody, erect shrubs, and are easy to distinguish by their branch-derived spines (branch-tips modified into spines) and compound cladodes (deciduous ultimate, cladode-bearing branches). There is an unsupported division between species with a main distribution area in the Western Cape Province of South Africa (A. capensis and allies), and the widespread A. suaveolens plus allies distributed in the eastern and northern parts of South Africa. The Capenses clade includes a distinct Recurvispinus clade (95 BP) in which the branch-spines are recurved when young (compared to straight in the rest of the Capenses clade), reminiscent of the leaf-derived spines seen in the rest of the spine-bearing Asparagus. Also, the fruit of A. recurvispinus and A. ‘‘karooicus’’ is a nutlet, compared to the black berry found in the rest of the Capenses species. Clade 5: Myrsiphyllum clade: This moderately supported monophyletic clade (82 BP) includes almost all species which Obermeyer placed in Myrsiphyllum, except A. scandens, which is resolved in an unsupported relationship to the Lignosus clade. However, the lack of bootstrap support for the position of A. scandens still leaves the possibility that Myrsiphyllum sensu Obermeyer (1984) might be monophyletic. Flattened cladodes and tubers, 39 which have often been used as key characters of Myrsiphyllum, are homoplasious, and are also found within the Racemose and Asparagus clades (although the tubers here are usually found on lateral roots, compared to on the main roots in the Myrsiphyllum clade). Obermeyer (1984) used connivent filaments together with connivent perianth to separate Myrsiphyllum from Protasparagus, but in A. scandens and A. ramosissimus the tepals are spreading rather than tube-forming (Malcomber and Demissew, 1993), supporting the exclusion of at least A. scandens from the Myrsiphyllum clade. Our results do place A. ramosissimus within the Myrsiphyllum clade together with A. declinatus and A. fasciculatus, as sisters to the strongly supported group of species with connivent filaments within the Myrsiphyllum clade (99 BP). We were unable to include some of the species previously placed in Myrsiphyllum (A. multituberosus, A. kraussianus and A. undulatus) in our analysis. However, due to their great morphological similarity to A. asparagoides and A. volubilis we anticipate that they will fall within the Myrsiphyllum clade, most likely within the group with connivent filaments. Clade 6: Setaceus clade: This clade (70 BP; Fig. 2a) is mainly composed of shade-loving, non-spiny, semi-woody climbers and shrubs and includes species from Obermeyer’s series Penduli and Globosi. Embedded in the Setaceus clade is a strongly supported group of non-South African species (90 BP) composed of the East African A. humilis plus a group (88 BP) of three Macaronesian accessions (A. arborescens ‘‘Tenerife’’, A. arborescens ‘‘Gran Canaria’’ and A. umbellatus ssp. umbellatus). Additionally it includes a weakly supported clade of three species assigned to A. cooperi (see above). A few species fall outside these major clades. These include (1) A. macowanii and A. mucronatus (strongly supported as sister species by 90 BP), which are most closely related to the Racemose clade; (2) A. retrofractus which, together with the Setaceus Clade, is resolved in a weak sistergroup relationship to the remaining Asparagus species; and (3) Asparagus scandens, sister to the Lignosus clade, which is morphologically very similar, but its phylogenetic position is still uncertain. 4.1.3. Species circumscription in larger complexes We tested 33% of all species for their monophyly by including at least two specimens in the phylogenetic analysis. At least a third (27) are not monophyletic, but only in three cases do those groups receive significant support. These results may be expected, as reviews by Crisp and Chandler (1996) and Funk and Omland (2003) suggest that paraphyletic species may be quite common and indeed their surveys produced rather similar percentages of paraphyletic species (23% of all tested species being paraphyletic). In some cases, such as the genus Kniphofia, all tested species have been shown to be paraphyletic (Ramdhani et al., 2009). The potential explanations for the occurrence of paraphyletic species have attracted much attention (Meimberg et al., 2010; Ramdhani et al., 2009; Rieseberg and Brouillet, 1994). A simple approach is to argue that only monophyletic species should be recognized (Baum and Donoghue, 1995; Mishler and Theriot, 2000), or that the concepts of mono- or paraphyly do not apply to species (Davis and Nixon, 1992). However, this cannot negate the observation that often specimens from the same taxonomic species do not form clades on the optimal phylogeny. Here we group the potential explanations into three sets: genetic, taxonomic and phylogenetic. Genetic explanations are most often invoked – these assume that the species delimitations (taxonomy) are correct, and that the species are reciprocally monophyletic, but that the molecular data are misleading. Two mechanisms are commonly proposed: post-speciation gene flow (e.g., hybridization) and/or lineage sorting (Meimberg et al., 2010; Ramdhani et al., 2009). There is evidence of polyploidy in Asparagus (Jessop, 1966), and although hybrids can be made between some closely related species in the 40 M.F. Norup et al. / Molecular Phylogenetics and Evolution 92 (2015) 25–44 genus, there appear to be sterility barriers among many species (Falavigna et al., 2008; Ito et al., 2008). Furthermore, there are no field observations of hybrids in southern Africa. The absence of robust conflict between the nuclear and plastid partitions in the phylogenetic analyses is not a good indicator of the absence of hybridization, since the resolution of these partial phylogenies is not robust enough to reveal conflict with confidence. Consequently, we currently have no convincing indications of the presence of natural gene flow between closely related species. Lineage sorting is most commonly observed in recently diverged clades, and is often interpreted as a signal for recent radiation (Ramdhani et al., 2009). This might apply in some clades in Asparagus, where the resolution within the species complexes (e.g. the asparagoides-, the cooperi- or suaveolens-species complexes) is poor. Taxonomic explanations are rarely invoked; these assume that the phylogenetic signal is correct, but the species delimitation is incorrect. For the southern African species of Asparagus two species level taxonomies are available. The earlier taxonomy, of Jessop (1966), uses a broad species concept, while that of Obermeyer and Immelman (1992) uses a very narrow species concept. These different taxonomic philosophies are most evident in the three large species complexes mentioned above (Table 4). For the delimitation of A. asparagoides, A. cooperi and A. suaveolens, Jessop (1966) includes several taxa (separated as species by Obermeyer), which he found to be embedded within these three widespread species. Jessop argues that the characters used to separate the segregates are ecologically labile, and that this in particular applies to the widespread species, where geographically separated populations are responding differently to different habitats. However, he also notes that considerable variation may be found within populations. The segregates are generally separated based on vegetative characters, such as cladode shape and size, branching patterns, and the habit in general. If variable within populations, these characters are of limited use in separating species (Davis, 1997; Du Rietz, 1930). However, we have no quantified data supporting this pattern. Jessop (1966) also notes that certain characters change gradually geographically, with transition zones (e.g. the growth form differences between A. burchellii and A. suaveolens). Such gradual clines may define ecotypes (Turesson, 1922), but not species. This sort of explanation might well apply to the Cooperi clade (Fig. 2c), in which the relationships between species are unclear, and most likely reflect that most of these species are common and widespread throughout southern Africa, and usually identified by overlapping character sets. From the variety of small subclades within the Cooperi clade, it is evident that there is a lot of morphological variation to account for within the complex. The Table 4 Summary of the species included in the three largest species complexes. Clade Jessop (1966) Obermeyer and Immelman (1992) Range asparagoides asparagoides asparagoides asparagoides asparagoides absent asparagoides kraussianus asparagoides ovatus multituberosus alopecuroides juniperoides 8 47 9 8 5 5 cooperi cooperi cooperi cooperi africanus cooperi bechuanicus 20 79 17 suaveolens suaveolens suaveolens suaveolens absent absent suaveolens burchellii spinescens mariae flavicaulis 76 17 3 4 10 many accessions, which we could only refer to as ‘‘cf. cooperi’’ and the three cooperi accessions found in the Setaceus clade (see above), makes it evident that species delimitations within this group are strongly in need of revision. Separating traits, which vary within populations from characters suitable to delimit species, will require substantial fieldwork, in order to provide a solid information base on which to base species delimitations. Phylogenetic explanations are based on the assumption that both the phylogenetic signal and species delimitations are correct and that the species are genuinely phylogenetically nested. Rieseberg and Brouillet (1994) presented convincing arguments that paraphyletic species are a necessary consequence of speciation, as most speciation is allopatric, by divergence of isolated populations, thus leaving a paraphyletic rest and only following differentiation of the ‘‘parent species’’ will both species become monophyletic. Numerous recent population-level investigations have demonstrated this predicted pattern (e.g. Albach et al., 2004; Funk and Omland, 2003; Patton and Smith, 1994). This sort of pattern could well apply to the A. suaveolens-complex, with A. mariae restricted to limestone near the southern tip of Africa, A. burchellii and A. spinescens found along the all-year rainfall south-eastern part of the subcontinent (Schulze, 1997) and A. flavicaulis on rocky outcrops. Following this interpretation A. spinescens and A. burchellii (represented by more than two accessions) are both nested within A. suaveolens, which is the paraphyletic ‘‘ancestor’’ widespread in the bushveld. 4.2. The evolution of sexual dimorphism in Asparagus The most plausible evolutionary pathway to dioecy in Asparagus has been from hermaphroditism via gynodioecy. The presence of female flowers as well as bisexual flowers has been observed in several Canary Islands species (e.g., A. umbellatus, A. nesiotes, A. plocamoides and A. acutifolius) (Julia Pérez de Paz, Jardín Botánico Canario ‘‘Viera y Clavijo’’, personal comments; Baker, 1875; Kunth, 1850, checked on material in herbarium C). Furthermore, these species originated within separate Asparagus clades (Fig. 3), indicating several independent origins of gynodioecy correlated with the dispersal from Africa to the Canary Islands. Multiple origins of gynodioecy in a genus is documented in dicots (Fritsch, 2003; Navajas-Perez et al., 2005; Weller et al., 1998) although not commonly described and is so far almost undocumented in monocots (Connor, 1973). Some sources indicated that A. acutifolius is dioecious (e.g., Sica et al., 2005; Kubota et al., 2012), but this is apparently not based on observations and cannot be verified as no vouchers are indicated. It appears evident that the transitions from hermaphroditism to sexual dimorphism (dioecy or gynodioecy) are closely related to the dispersal out of southern Africa. This is consistent with the theory that the evolution of dioecy in plants may be the result of selection for outcrossing and that this mechanism often evolve during bottlenecks; e.g., inbreeding depression due to long-distance dispersal and colonization of new habitats) (see e.g. Barrett, 2002; Charlesworth and Charlesworth, 1978; Pannell, 2006; Weiblen et al., 2000). It has furthermore been proposed that a link exists between sexual dimorphism and polyploidization, as seen in e.g. Lycium L. (Solanaceae), Rubus L. and Fragaria L. (Rosaceae) and Astilbe Buch.-Ham. (Saxifragaceae), where chromosome doubling is correlated with the evolution of self-compatibility in otherwise self-incompatible plants, giving rise to polyploid species with gender dimorphism (Barrett, 2002; Miller and Venable, 2000). This could be a possible pathway for the evolution of sexual dimorphism in Asparagus, as in general Eurasian species are reported to have larger genomes than southern African species (Kar and Sen, 1985; Kuhl et al., 2005; Moreno et al., 2008), although a few M.F. Norup et al. / Molecular Phylogenetics and Evolution 92 (2015) 25–44 widespread southern and south central African species are reported to be tetraploid or hexaploid (e.g. A. falcatus, A. cooperi: 2n = 40; Kar and Sen (1985) and A. densiflorus: 2n = 60; Jessop (1966)). 4.3. Out-of-Africa – the origin of Asparagus and dispersal to Eurasia The genus Asparagus is widely distributed in the Old World, growing in semiarid to arid areas in Africa and Eurasia. However, due to insufficient analyses or sampling, most recent treatments of the genus have reserved judgement as to its area of origin (e.g. Dahlgren et al., 1985; Kubitzki and Rudall, 1998). The S-DIVA (Fig. 4) and DEC analyses (Fig. S5, Suppl. Mat.) unequivocally infer a southern African ancestral distribution of Asparagus (area I) and thus confirm the hypothesis presented by Fukuda et al. (2005). We do not believe that our sampling imbalance (towards southern Africa, which is by far the most species rich region) had a significant influence in placing southern Africa as the area of origin, as all included non-southern African species are found to have common ancestors within southern Africa. These non-southern African species are shown to be the result of not one, but multiple independent dispersal events out of southern Africa. Our results demonstrate that all truly dioecious species of Asparagus have a single common ancestral area – which the S-DIVA analysis infers as area L (China, Mongolia and Eastern Asia) – with later dispersal back to Europe (areas A and B), a lineage from which the widespread A. officinalis and the more local Mediterranean species A. maritimus evolved. Confidence in this result is weakened by the poor sampling from the non-African areas and the low bootstrap support of the Dioecy clade (Fig. 2). 41 The Dioecy clade originated within a group of hermaphroditic species from southern and South Tropical Africa (areas G and I), East and Northeastern Tropical Africa (area F), North Africa (area C), the Canary Islands (area D), the Mediterranean areas (area B; Southwestern + Southeastern Europe) and the Arabian Peninsula (area K; Arabian Peninsula + Middle Asia + Caucasus + Western Asia). Looking at the entire clade including the Dioecy clade, we conclude that the dispersal of Asparagus seems to have followed a route (1) from southern Africa through a tropical East-Northeastern African corridor westwards to North Africa, Macaronesia and Mediterranean Europe (reflected in the clade including A. flagellaris, A. petersianus, A. ‘‘pauli-guilelemi’’, A. aphyllus and A. stipularis) and (2) eastwards to Eurasia (the Dioecy clade) (see Fig. 5). Part of this dispersal pattern is repeated in the Setaceus clade (Figs. 2a and 3), where S-DIVA infers the distribution of the most recent common ancestor of the Canary Island species A. arborescens, A. umbellatus ssp. umbellatus plus A. humilis (East Africa) to have been widespread through South Central Africa (area G), East and Northeastern Tropical Africa (area F) to the Canary Islands (area D). The distribution pattern of Asparagus does not fit the ‘‘Rand Flora Pattern’’ which is defined as an old xerophylic tropical African flora (Quezel, 1978), which shows a distribution pattern also described as the ‘‘arid track’’ (de Winter, 1971; Jürgens, 1997; Verdcourt, 1969). The Rand flora was supposed to have originally diversified in south-western Africa, but also to link the arid elements of the Canary Islands with those of north-east Africa (Marrero et al., 1998). Asparagus largely avoids the dry areas and is centred on the more mesic habitats in southern Africa, from the fynbos flora to the montane grassland and subtropical Fig. 5. A map of the Old World showing the main dispersal routes of Asparagus (arrows) derived from the DIVA analysis (Fig. 4), including known radiation centres (stars) (the major radiation centre still being South Africa). 42 M.F. Norup et al. / Molecular Phylogenetics and Evolution 92 (2015) 25–44 savannas. However, Sanmartin et al. (2010) used a broader definition of the Rand flora, which encompasses all open country lineages, and as such would also include Asparagus. Asparagus fits the model of a southern radiation with a northward dispersal. A radiation pattern like this may be congruent with, inter alia, Aloe (Holland, 1978), Crassula L. (Mort et al., 2001), Scrophulariaceae (Oxelman et al., 2005), Amaryllideae (Meerow et al., 1999; Motomi et al., 1999; Rønsted et al., 2012), Gladiolus (e.g. Goldblatt and Manning, 1998) and Lotononis (DC) Eckl. & Zeyh. (Linder et al., 1992). These clades all have their ancestral areas in mesic, open habitats in southern Africa (in the heathlands of the Cape flora, or the grasslands of the central southern African plateau) and achieved global or near-global distributions. Although they are currently also found in the semi-arid Succulent Karoo, this biome is of Late Miocene origin (Dupont et al., 2011), younger than the original radiations of these clades. 5. Conclusion Asparagus is a complex genus that shows a notable discrepancy between the molecular phylogeny and the morphologically defined species in the existing taxonomy, with e.g. several species belonging to larger species complexes, most of which are paraphyletic, though a few are polyphyletic. Despite the fact that our study is the most complete and well supported to date, the lack of an up-to-date taxonomy makes the interpretation of patterns at species level difficult. We provide the phylogenetic basis for a new infrageneric classification of Asparagus and conclude that species delimitation needs to be based on both molecular and morphological data. Incongruence of sequence and morphological data indicates problems with species delimitation that need further investigation, but large-scale morphological and biogeographical patterns are evident (e.g. dioecy). Our analysis provides no support for the classical interpretation of the evolution of dioecy through gynodioecy. Acknowledgments The authors thank Akira Kanno and Tatsuya Fukuda for providing DNA material for the Asian species; Juli Caujapé-Castells and Ruth Jaén Molina at Jardin Botanico Canario ‘‘Viera y Clavijo’’, Las Palmas, Gran Canaria, for help with samples and information; Sebsebe Demissew for samples; Terry Trinder-Smith for help in the Bolus Herbarium; Michael S. Kinney for sharing information on PHY-C primer optimization and outgroup sequences; Charlotte Hansen for assistance in the laboratory; Nina Rønsted and Alexandre Antonelli for assistance and discussions; Justin Moat for assistance with the map; Reto Nyfeller and Melanie Ranft for help with GenBank submission. This research was supported by grants to the first author from the EDIT Women Scientists Grant, the Christian and Ottilia Brorsons Grant, the Augustinus Foundation, the Oticon Foundation and the foundation of Svend G. Fiedler and wife. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.ympev.2015.06. 002. References Albach, D.C., Martínez-Ortega, M.M., Fischer, M.A., Chase, M.W., 2004. A new classification of the tribe Veroniceae: problems and a possible solution. Taxon 53, 429–452. Álvarez, I., Wendel, J.F., 2003. Ribosomal ITS sequences and plant phylogenetic inference. Mol. Phylogenet. Evol. 29, 417–434. Angiosperm Phylogeny Group, 2009. An update of the Angiosperm Phylogeny Group classification for the orders and families of flowering plants: APG III. Bot. J. Linn. Soc. 161, 105–121. Auld, T.D., Hutton, I., 2004. Conservation issues for the vascular flora of Lord Howe Island. Cunninghamia 8, 490–500. Baker, J.G., 1875. Revision of the genera and species of Asparagaceae. J. Linn. Soc., Bot. 14, 508–632. Barrett, S.C.H., 2002. The evolution of plant sexual diversity. Nat. Rev. Genet. 3, 274– 284. Baum, D.A., Donoghue, M.J., 1995. Choosing among alternative ‘‘phylogenetic’’ species concepts. Syst. Bot. 20, 560–573. Burrows, S.M., Burrows, J.E., 2008. Three new species of Asparagus (Asparagaceae) from South Africa, with notes on other taxa. Bothalia 38, 23–29. Campbell, V., Lapointe, F.J., 2009. The use and validity of composite taxa in phylogenetic analysis. Syst. Biol. 58, 560–572. Charlesworth, B., Charlesworth, D., 1978. A model for the evolution of dioecy and gynodioecy. Am. Nat. 112, 975–997. Chase, M.W., Soltis, D.E., Soltis, P.S., Rudall, P.J., Fay, M.F., Hahn, W.J., Sullivan, S., Joseph, J., Molvray, M., Kores, P.J., Givnish, T.J., Sytsma, K.J., Pires, J.C., 2000. Higher-level systematics of the monocotyledons: an assessment of current knowledge and a new classification. In: Wilson, K.L., Morrison, D.A. (Eds.), Monocots: Systematics and Evolution. CSIRO, Collingwood, Australia, pp. 3–16. Chase, M.W., Reveal, J.L., Fay, M.F., 2009. A subfamilial classification for the expanded asparagalean families Amaryllidaceae, Asparagaceae and Xanthorroeaceae. Bot. J. Linn. Soc. 161, 132–136. Clark, J.R., Ree, R.H., Alfaro, M.E., King, M.G., Wagner, W.L., Roalson, E.H., 2008. A comparative study in ancestral range reconstruction methods: retracing the uncertain histories of insular lineages. Syst. Biol. 57, 693–707. Connor, H.E., 1973. Breeding systems in Cortaderia (Gramineae). Evolution 27, 663– 678. Crisp, M.D., Chandler, G.T., 1996. Paraphyletic species. Telopea 6, 813–844. Dahlgren, R.M.T., Clifford, H.T., Yeo, P.F., 1985. The Families of the Monocotyledons. Springer-Verlag, Berlin, pp. 140–142. Darwin, C., 1877. The Different Forms of Flowers on Plants of the Same Species. J. Murray, London. Davis, J.I., 1997. Evolution, evidence, and the role of species concepts in phylogenetics. Syst. Bot. 22, 373–403. Davis, J.I., Nixon, K.C., 1992. Populations, genetic variation, and the delimitation of phylogenetic species. Syst. Biol. 41, 421–435. de Winter, B., 1971. Floristic relationships between the northern and southern arid areas in Africa. Mitt. Bot. Staatssamml., München 10, 424–437. del Hoyo, A., GarcÌa-MarÌn, J.L., Pedrola-Monfort, J., 2009. Temporal and spatial diversification of the African disjunct genus Androcymbium (Colchicaceae). Mol. Phylogenet. Evol. 53, 848–861. Demissew, S., 2008. Four new species of Asparagus (Asparagaceae) from the Flora Zambesiaca area. Kew Bull. 63, 269–275. Drummond, A.J., Rambaut, A., 2007. BEAST: bayesian evolutionary analysis by sampling trees. BMC Evol. Biol. 7, 214. Drummond, A.J., Ho, S.Y.W., Phillips, M.J., Rambaut, A., 2006. Relaxed phylogenetic and dating with confidence. PLoS Biol. 4, e88. Drummond, A.J., Suchard, M.A., Xie, D., Rambaut, A., 2012. Bayesian phylogenetics with BEAUTi and the BEAST 1.7. Mol. Biol. Evol. 29, 1969–1973. Du Rietz, G.E., 1930. The fundamental units of biological taxonomy. Sven. Bot. Tidskr. 24, 333–428. Dupont, L.M., Linder, H.P., Rommerskirchen, F., Schefuss, E., 2011. Climate-driven rampant speciation of the Cape flora. J. Biogeogr. 38, 1059–1068. Falavigna, A., Alberti, P., Casali, P.E., Toppino, L., Wang, H.S., Mennella, G., 2008. Interspecific hybridization for asparagus breeding in Italy. In: Mulder, J.H., Lavrijsen, P.J.M. (Eds.), Proceedings of the XIth International Asparagus Symposium. International Society Horticultural Science, Leuven 1, pp. 291–297. Fay, M.F., Rudall, P.J., Sullivan, S., Stobart, K.L., de Bruijn, A.Y., Reeves, G., QamaruzZaman, F., Hong, W.-P., Joseph, J., Hahn, W.J., Conran, J.G., Chase, M.W., 2000. Phylogenetic studies of Asparagales based on four plastid DNA regions. In: Wilson, K.L., Morrison, D.A. (Eds.), Monocots: Systematics and Evolution. CSIRO Publishing, Melbourne, pp. 360–371. Feinbrun-Dothan, N., 1986. Asparagus. In: Feinbrun-Dothan, N. (Ed.), Flora Palaestina, vol. 4. Text. Jerusalem. Israel Academy of Sciences and Humanities, pp. 99–101. Feliner, G.N., Rosselló, J.A., 2007. Better the devil you know? Guidelines for insightful utilization of nrDNA ITS in species-level evolutionary studies in plants. Mol. Phylogent. Evol. 44, 911–919. Fellingham, A.C., Meyer, N.L., 1995. New combinations and a complete list of Asparagus species in southern Africa (Asparagaceae). Bothalia 25, 205–209. Felsenstein, J., 1985. Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39, 783–791. Fitch, W.M., 1971. Toward defining the course of evolution: minimum change for a specific tree topology. Syst. Zool. 20, 406–416. Forest, F., Grenyer, R., Rouget, M., Davies, T.J., Cowling, R.M., Faith, D.P., Balmford, A., Manning, J.C., Proches, S., van der Bank, M., Reeves, G., Hedderson, T.A.J., Savolainen, V., 2007. Preserving the evolutionary potential of floras in biodiversity hotspots. Nature 445, 757–760. Fritsch, P.W., 2003. Multiple geographic origins of Antillean Styrax. Syst. Bot. 28, 421–430. M.F. Norup et al. / Molecular Phylogenetics and Evolution 92 (2015) 25–44 Fukuda, T., Ashizawa, H., Suzuki, R., Ochiai, T., Nakamura, T., Kanno, A., Kameya, T., Yokoyama, Y., 2005. Molecular phylogeny of the genus Asparagus (Asparagaceae) inferred from plastid petB intron and petD-rpoA intergenic spacer sequences. Plant Species Biol. 20, 121–132. Funk, D.J., Omland, K.E., 2003. Species-level paraphyly and polyphyly: frequency, causes, and consequences, with insights from animal mitochondrial DNA. Annu. Rev. Ecol. Evol. Syst. 34, 397–423. Goldblatt, P., Manning, J.C., 1998. Gladiolus in Southern Africa: Systematics, Biology and Evolution. Fernwood Press, Cape Town. Govaerts, R., Zonneveld, B.J.M., Zona, S.A., 2009. World Checklist of Asparagaceae. The Board of Trustees of the Royal Botanic Gardens, Kew. Published on the Internet; <http://www.kew.org/wcsp/> (accessed 22.10.09). Guindon, S., Gascuel, O., 2003. A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood. Syst. Biol. 52, 696–704. Guo, X., Wang, Y., 2007. Partitioned Bayesian analyses, dispersal-vicariance analysis, and the biogeography of Chinese toad-headed lizards (Agamidae: Phrynocephalus): a re-evaluation. Mol. Phylogenet. Evol. 45, 643–662. Hernandez, L.S., 1995. Taxonomic study of the Mexican genus Hemiphylacus (Hyacinthaceae). Syst. Bot. 20, 546–554. Hertweck, K.L., Kinney, M.S., Stuart, S.A., Maturin, O., Mathews, S., Chase, M.W., Gandolfo, M.A., Pires, J.C., 2015. Phylogenetics, divergence times and diversification from three genome partitions in monocots. Bot. J. Linn. Soc. 178, 375–393. Holland, P.G., 1978. An evolutionary biogeography of the genus Aloe. J. Biogeogr. 5, 213–226. Huber, H., 1969. Die Samenmarkmale und Verwandtschaftsverhältnisse der Liliifloren. Mitt. Bot. Staatssamml., München 8, 219–538. Humphreys, A.M., Linder, H.P., 2009. Concept versus data in the delimitation of plant genera. Taxon 33, 1054–1074. Ito, T., Ochiai, T., Fukuda, T., Ashizawa, H., Kanno, A., Kameya, T., Sonoda, T., 2008. Potential of interspecific hybrids in the genus Asparagus. In: Mulder, J.H., Lavrijsen, P.J.M. (Eds.), Proceedings of the XIth International Asparagus Symposium. International Society Horticultural Science, Leuven 1, pp. 279–284. Jessop, J.P., 1966. The genus Asparagus in Southern Africa. Bothalia 9, 31–96. Jürgens, N., 1997. Floristic biodiversity and history of African arid regions. Biodivers. Conserv. 6, 495–514. Kar, D.K., Sen, S., 1985. Chromosome characteristics of Asparagus – sapogenin yielding plant. Cytologia 50, 147–155. Kubitzki, K., Rudall, P.J., 1998. Asparagaceae. In: Kubitzki, K. (Ed.), The Families and Genera of Vascular Plants, Monocotyledons, vol. 3. Springer-Verlag, Berlin, Heidelberg, pp. 125–129. Kubota, S., Konno, I., Kanno, A., 2012. Molecular phylogeny of the genus Asparagus (Asparagaceae) explains interspecific crossability between the garden asparagus (A. officinalis) and other Asparagus species. Theor. Appl. Genet. 124, 345–354. Kuhl, J.C., Havey, M.J., Martin, W.J., Cheung, F., Yuan, Q., Landherr, L., Hu, Y., LeebensMack, J., Town, C.D., Sink, K.C., 2005. Comparative genomic analyses in Asparagus. Genome 48, 1052–1060. Kunth, K.S., 1850. Enumeratio plantarum omnium hucusque cognitarum, secundum familias naturales disposita, adjectis characteribus, differentiis et synonymis. J.G. Cotta, Stuttgart, pp. 57–110. Larkin, M.A., Blackshields, G., Brown, N.P., Chenna, R., McGettigan, P.A., McWilliam, H., Valentin, F., Wallace, I.M., Wilm, A., Lopez, R., Thompson, J.D., Gibson, T.J., Higgins, D.G., 2007. Clustal W and Clustal X version 2.0. Bioinformatics 23, 2947–2948. Lee, Y.-O., Kanno, A., Kameya, T., 1997. Phylogenetic relationships in the genus Asparagus based on the restriction enzyme analysis of the chloroplast DNA. Breed. Sci. 47, 4. Linder, H.P., Meadows, M.E., Cowling, R.M., 1992. History of the Cape flora. In: Cowling, R.M. (Ed.), The Ecology of Fynbos: Nutrients, Fire and Diversity. Oxford University Press, Cape Town, pp. 113–134. Mabberley, D.J., 2008. Mabberley’s Plant-book: A Portable Dictionary of Plants, their Classifications, and Uses. Cambridge University Press. Maddison, W.P., Maddison, D.R., 2005. MacClade: Analysis of Phylogeny and Character Evolution. Sinauer Associates Inc., Sunderland, Massachusetts. Maddison, W.P., Maddison, D.R., 2009. Mesquite: A Modular System for Evolutionary Analysis. Version 2.72. <http://mesquiteproject.org>. Malcomber, S.T., Demissew, S., 1993. The status of Protasparagus and Myrsiphyllum in the Asparagaceae. Kew Bull. 48, 63–78. Marrero, A., Almeida, R.S., González-Martín, M., 1998. A new species of the wild dragon tree, Dracaena (Dracaenaceae) from Gran Canaria and its taxonomic and biogeographic implications. Bot. J. Linn. Soc. 128, 291–314. Meerow, A.W., Fay, M.F., Guy, C.L., Li, Q.-B., Zaman, F.Q., Chase, M.W., 1999. Systematics of Amaryllidaceae based on cladistic analysis of plastid sequence data. Am. J. Bot. 86, 1325–1345. Meimberg, H., Rischer, H., Turini, F.G., Chamchumroon, V., Dreyer, M., Sommaro, M., Bringmann, G., Heubl, G., 2010. Evidence for species differentiation within the Ancistrocladus tectorius complex (Ancistrocladaceae) in Southeast Asia: a molecular approach. Plant Syst. Evol. 284, 77–98. Miller, J.S., Venable, D.L., 2000. Polyploidy and the evolution of gender dimorphism in plants. Science 289, 2335–2338. Mishler, B.D., Theriot, E.C., 2000. The phylogenetic species concept (sensu Mischler and Theroid): monophyly, apomorphy, and phylogenetic species concepts. In: Wheeler, Q.D., Meier, R. (Eds.), Species Concepts and Phylogenetic Theory. Columbia University Press, New York, pp. 44–54. 43 Moore, T.E., Verboom, G.A., Forest, F., 2010. Phylogenetics and biogeography of the parasitic genus Thesium L. (Santalaceae), with an emphasis on the Cape of South Africa. Bot. J. Linn. Soc. 162, 435–452. Moreno, R., Espejo, J.A., Cabrera, A., Gil, J., 2008. Origin of tetraploid cultivated asparagus landraces inferred from nuclear ribosomal DNA internal transcribed spacers’ polymorphisms. Ann. Appl. Biol. 153, 233–241. Mort, M.E., Soltis, D.E., Soltis, P.S., Francisco-Ortega, J., Santos-Guerra, A., 2001. Phylogenetic relationships and evolution of Crassulaceae inferred from matK sequence data. Am. J. Bot. 88, 76–91. Motomi, I., Kawamoto, A., Kita, Y., Yukawa, T., Kurita, S., 1999. Phylogenetic relationships of Amaryllidaceae based on matK sequence data. J. Plant. Res. 112, 207–216. Navajas-Perez, R., de la Herran, R., Lopez Gonzalez, G., Jamilena, M., Lozano, R., Ruiz Rejon, C., Ruiz Rejon, M., Garrido-Ramos, M.A., 2005. The evolution of reproductive systems and sex-determining mechanisms within Rumex (Polygonaceae) inferred from nuclear and chloroplastidial sequence data. Mol. Biol. Evol. 22, 1929–1939. Nixon, K.C., 1999. The parsimony ratchet, a new method for rapid parsimony analysis. Cladistics 15, 407–414. Nylander, J.A.A., Olsson, U., Alström, P., Sanmartín, I., 2008. Accounting for phylogenetic uncertainty in biogeography: a Bayesian approach to dispersalvicariance analysis of the thrushes (Aves: Turdus). Syst. Biol. 57, 257–268. Obermeyer, A.A., 1983. Protasparagus Oberm., nom.nov.: new combinations. S. Afr. J. Bot. 2, 243–244. Obermeyer, A.A., 1984. Revision of the genus Myrsiphyllum Willd. Bothalia 15, 77– 88. Obermeyer, A.A., Immelman, K.L., 1992. Asparagaceae. In: Leistner, O.A., du Plessis, E. (Eds.), Flora of Southern Africa. National Botanical Institure, Pretoria, pp. 11– 82. Olmstead, R.G., Sweere, J.A., 1994. Combining data in phylogenetic systematics – an empirical approach using 3 molecular data sets in the Solanaceae. Syst. Biol. 43, 467–481. Oxelman, B., Kornhall, P., Olmstead, R.G., Bremer, B., 2005. Further disintegration of Scrophulariaceae. Taxon 54, 411–425. Pannell, J.R., 2006. Effects of colonization and metapopulation dynamics on the evolution of plant sexual systems. In: Harder, L.D., Barrett, S.C.H. (Eds.), Ecology and Evolution of Flowers. Oxford University Press, Oxford, NY, pp. 223–238. Patton, J.L., Smith, M.F., 1994. Paraphyly, polyphyly, and the nature of species boundaries in pocket gophers (genus Thomomys). Syst. Biol. 43, 11–26. Pires, J.C., Maureira, I.J., Givnish, T.J., Sytsma, K.J., Seberg, O., Petersen, G., Davis, J.I., Stevenson, D.W., Rudall, P.J., Fay, M.F., Chase, M.W., 2006. Phylogeny, genome size, and chromosome evolution of Asparagales. Aliso 22, 278–304. Posada, D., 2008. jModelTest: phylogenetic model averaging. Mol. Biol. Evol. 25, 1253–1256. Quezel, P., 1978. Analysis of the flora of Mediterranean and Saharan Africa. Ann. Miss. Bot. Gard. 65, 479–534. Raimondo, D., Von Staden, L., Foden, W., Victor, J.E., Helme, N.A., Turner, R.C., Kamundi, D.A., Manyama, P.A.E. (Eds.), 2009. Red List of South African Plants. South African National Biodiversity Institute, Pretoria. Ramdhani, S., Barker, N.P., Baijnath, H., 2009. Rampant non-monophyly of species in Kniphofia Moench (Asphodelaceae) suggests a recent Afromontane radiation. Taxon 58, 1141–1152. Red List of Threatened Species of Japan. 2007. Japanese Ministry of the Environment, <http://www.biodic.go.jp/english/rdb/rdb_f.html>. Ree, R.H., Smith, S., 2008a. LAGRANGE: Likelihood Analysis of Geographic Range Evolution. Version 2.0. Ree, R.H., Smith, S., 2008b. Maximum-likelihood inference of geographic range evolution by dispersal, local extinction, and cladogenesis. Syst. Biol. 57, 4–14. Renner, S.S., Ricklefs, R.E., 1995. Dioecy and its correlates in the flowering plants. Am. J. Bot. 82, 596–606. Rieseberg, L.H., Brouillet, L., 1994. Are many plant species paraphyletic? Taxon 43, 21–32. Roemer, J.J., Schultes, J.A., 1829. Caroli a Linné, equitis, Systema Vegetabilium 7 (1– 2). J.G. Cotta, Stuttgart. Ronquist, F., 1997. Dispersal-vicariance analysis: a new approach to the quantification of historical biogeography. Syst. Biol. 46, 195–203. Rønsted, N., Symonds, M.R.E., Birkholm, T., Christensen, S.B., Meerow, A.W., Molander, M., Mølgaard, P., Petersen, G., Rasmussen, N., van Staden, J., Stafford, G.I., Jäger, A.K., 2012. Can phylogeny predict chemical diversity and potential medicinal activity of plants? A case study of Amaryllidaceae. BMC Evol. 12, 182. Rudall, P.J., Engelman, E.M., Hanson, L., Chase, M.W., 1998. Systematics of Hemiphylacus, Asparagus and Anemarrhena (Asparagales). Plant Syst. Evol. 211, 181–199. Sánchez Gómez, P., Vera Péréz, J.B., Jiménéz Martínez, J.F., 2009. Asparagus maritimus. In: Bañares, Á., Blanca, G., Güemes, J., Moreno, J.C., Ortiz, S. (Eds.), Atlas y Libro Rojo de la Flora Vascular Amenazada de España – Adenda 2008. Dirección General de Medio Natural y Política Forestal y Sociedad Española de Biología de la Conservación de Plantas, Madrid. Sanmartin, I., Anderson, L.C., Alarcon, M., Ronquist, F., Aldasoro, J.J., 2010. Bayesian island biogeography in a continental setting: the Rand Flora case. Biol. Lett. 6, 703–707. Santana, I., Scholz, S., Naranjo, J., 2004. Asparagus nesiotes subsp. purpuriensis. In: Bañares, Á., Blanca, G., Güemes, J., Moreno, J.C., Ortiz, S. (Eds.), Atlas y Libro Rojo de la Flora Vascular Amenazada de España. Dirección General de Conservación de la Naturaleza, Madrid. 44 M.F. Norup et al. / Molecular Phylogenetics and Evolution 92 (2015) 25–44 Schulze, R.E., 1997. South African atlas of Agrohydrology and -Climatology. Water Research Commission, Pretoria. Seberg, O., Petersen, G., Davis, J.I., Pires, J.C., Stevenson, D.W., Chase, M.W., Fay, M.F., Devey, D.S., Jørgensen, T., Sytsma, K.J., Pillon, Y., 2012. Phylogeny of the Asparagales based on three plastid and two mitochondrial genes. Am. J. Bot. 99, 875–889. Shaw, J., Lickey, E.B., Beck, J.T., Farmer, S.B., Liu, W., Sirpun, K.C., Winder, C.T., Schilling, E.E., Small, R.L., 2005. The tortoise and the hare II: Relative utility of 21 noncoding chloroplast DNA sequences for phylogenetic analysis. Am. J. Bot. 92, 142–166. Sheahan, M.C., Chase, M.W., 2000. Phylogenetic relationships within Zygophylaceae based on DNA sequences of three plastid regions, with special emphasis on Zygophylloideae. Syst. Bot. 25, 371–384. Sica, M., Gamba, G., Montieri, G.S., Gaudio, L., Aceto, S., 2005. ISSR markers show differentiation among Italian populations of Asparagus acutifolius L. BMC Genet. 6, 17. Sikes, D.S., Lewis, P.O., 2001. Beta software. Version 1. PAUPRat⁄ Implementation of the Parsimony Ratchet. Distributed by the Authors, Department of Ecology and Evolutionary Biology, University of Connecticut, Storrs. Stajner, N., Bohanec, B., Javornik, B., 2002. Genetic variability of economically important Asparagus species as revealed by genome size analysis and rDNA ITS polymorphisms. Plant Sci. 162, 931–937. Swofford, D.L., 2002. PAUP⁄. Phylogenetic Analysis Using Parsimony (⁄and other methods). Version 4. Sinauer Associates, Sunderland, Massachusetts. Team, R.D.C., 2009. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna. <http://www.R-project.org>. Turesson, G., 1922. The genotypical response of the plant species to the habitat. Hereditas 3, 211–350. Valdés, B., 1979. Revision del genero Asparagus (Liliaceae) en Macaronesia. Lagascalia 9, 65–107. Valdés, B., 1980. Asparagus. In: Tutin, T.G., Heywood, V.H., Burges, N.A., Moore, D.M., Valentine, D.H., Walters, S.M., Webb, D.A. (Eds.), Flora Europea, Alismataceae to Orchidaceae (Monocotyledones), vol. 5. Cambridge University Press, Cambridge, pp. 71–73. Verdcourt, B., 1969. The arid corridor between the northeast and southwest areas of Africa. Palaeoecol. Africa 4, 140–144. Weiblen, G.D., Oyama, R.K., Donoghue, M.J., 2000. Phylogenetic analysis of dioecy in monocotyledons. Am. Nat. 155, 13. Weller, S.G., Sakai, A.K., Rankin, A.E., Golonka, A., Kutcher, B., Ashby, K.E., 1998. Dioecy and the evolution of pollination systems in Schiedea and Alsinidendron (Caryophyllaceae: Alsinoideae) in the Hawaiian Islands. Am. J. Bot. 85, 1377– 1388. Wiegand, C., 2006. Asparagus (Asparagaceae) nrDNA ITS Phylogeny of South Africa. M.A. Thesis, Department of Ecology and Evolutionary Biology, University of Kansas, Kansas. Wiens, J.J., 1998. Combining data sets with different phylogenetic histories. Syst. Biol. 47, 568–581. Willdenow, C.L., 1808. Nähere Bestimmung einiger Lilienwächse, besonders aber derjenigen, die zur sechsten Klasse und dritten Ordnung des Linnéschen Systems gehören. Gesellschaft naturforschender Freunde zu Berlin Magazin für die neuesten Entdeckungen in der Gesammten Naturkunde, vol. 2, pp. 25–26. World Checklist of Selected Plant Families, 2009. The Board of Trustees of the Royal Botanic Gardens, Kew. Published on the Internet; <http://www.kew.org/wcsp/> (accessed 22.10.09). Yu, Y., 2009. S-DIVA 1.1b. Statistical Dispersal-Vicariance Analysis. Distributed freely by Yan Yu, College of Life Sciences, Sichuan University, Chengdu, China. <http://mnh.scu.edu.cn/s-diva/>. Yu, Y., Harris, A.J., He, X., 2010. S-DIVA (Statistical Dispersal-Vicariance Analysis): a tool for inferring biogeographic histories. Mol. Phylogenet. Evol. 56, 848–850. Zwickl, D.J., 2006. Genetic Algorithm Approaches for the Phylogenetic Analysis of Large Biological Sequence Datasets under the Maximum Likelihood Criterion. Ph.D. Dissertation, University of Texas, Austin.