- Email: [email protected]
Recent advances in understanding Apiales and a revised classification
Copyright © NISC Pty Ltd
South African Journal of Botany 2004, 70(3): 371–381 Printed in South Africa — All rights reserved
SOUTH AFRICAN JOURNAL OF BOTANY ISSN 0254–6299
Recent advances in understanding Apiales and a revised classification GM Plunkett1*, GT Chandler1,2, PP Lowry II3, SM Pinney1 and TS Sprenkle1 Department of Biology, Virginia Commonwealth University, PO Box 842012, Richmond, Virginia 23284-2012, United States of America 2 Present address: Department of Biology, University of North Carolina, Wilmington, North Carolina 28403-5915, United States of America 3 Missouri Botanical Garden, PO Box 299, St Louis, Missouri 63166-0299, United States of America; Département de Systématique et Evolution, Muséum National d’Histoire Naturelle, Case Postale 39, 57 rue Cuvier, 75231 Paris CEDEX 05, France * Corresponding author, e-mail: [email protected] 1
Received 23 August 2003, accepted in revised form 18 November 2003 Despite the long history of recognising the angiosperm order Apiales as a natural alliance, the circumscription of the order and the relationships among its constituent groups have been troublesome. Recent studies, however, have made great progress in understanding phylogenetic relationships in Apiales. Although much of this recent work has been based on molecular data, the results are congruent with other sources of data, including morphology and geography. A unified picture of relationships has now emerged regarding the delimitation of
Apiales, which includes a core group of four families (Apiaceae, Araliaceae, Myodocarpaceae, Pittosporaceae) to which three small families are also added (Griseliniaceae, Torricelliaceae and Pennantiaceae). After a brief review of recent advances in each of the major groups, a revised classification of the order is presented, which includes the recognition of the new suborder Apiineae (comprising the four core families) and two new subfamilies within Apiaceae (Azorelloideae and Mackinlayoideae).
Introduction Almost all authors have considered the dicot families Apiaceae and Araliaceae closely related (e.g. Bentham and Hooker 1867, Engler and Prantl 1898, Bessey 1915, Dahlgren 1980, Cronquist 1981, 1988, Takhtajan 1987, Thorne 1992, but see Hutchinson 1967, 1973) and have usually treated them in the same order, albeit under various names, including Umbellales, Cornales (when associated with Cornaceae), Araliales, and Apiales (the name we shall use hereafter). The near-ubiquity of this interpretation stands in stark contrast to the diverse treatments both above and below the ordinal level, such as the placement of Apiales among the other orders of dicots, the inclusion of additional families within the order (Cornaceae, Torricelliaceae, Helwingiaceae, Griseliniaceae, etc.), and the many interand infrafamilial treatments of Apiaceae and Araliaceae themselves. As with many higher taxonomic groups, the difficulties in sorting out relationships in Apiales presumably stem from repeated cycles of divergence and migration, followed by convergence and parallelism. In the decades immediately preceding the development of objective approaches to phylogenetic analysis (e.g. parsimony, phenetic, and likelihood methods) and the use of molecular sources of data, few substantial advances were
made in the ordinal classification of Apiales (Dahlgren 1980 being the only notable exception). In the past ten years, however, a unified picture of relationships has rapidly emerged regarding both the circumscription and internal relationships of the order. Similar progress has been made in many other angiosperm orders, and several ongoing efforts are now underway to formalise the results of these advances (APG 1998, 2003, Stevens 2003). Most authors working on Apiales have hesitated to provide a formal recircumscription of the order during this ‘period of discovery’. The recent publication of a new apialean classification by an author with little or no expertise in the group (Doweld 2001), including the description of two new families based on our own work (viz. Plunkett and Lowry 2001), has prompted us to provide both a formal classification of Apiales together with a review of the relevant literature upon which our decisions are based. We hope this effort will also provide a fitting framework from which the other contributions to this symposium volume can be better understood. Several publications already provide comprehensive reviews of the taxonomic history of Apiales and/or their constituent groups (e.g. Constance 1971, Heywood 1978, Plunkett et al. 1996a, 1996b, 1997, Downie et al. 2001,
372
Plunkett and Lowry 2001, Plunkett 2001), and therefore this information is not repeated herein. Instead, we focus on reviewing (or in some cases previewing) results from recent studies (or unpublished work) completed since the publication of the preceding Apiales Special Issue of the Edinburgh Journal of Botany (2001, Volume 58, no. 2). Following this, we provide a formal recircumscription and interfamilial classification of the order. Apiaceae In their contributions to the last symposium volume, both Plunkett (2001) and Downie et al. (2001) explored relationships within Apiaceae using ‘supertree’ and ‘consensus’ approaches, respectively (Downie et al. 2001 also examined relationships within subfamily Apioideae using a ‘combined data’ approach). With the exception of Chandler and Plunkett (2004), most recent studies of Apiaceae have focused on relationships within subfamilies and/or among genera (e.g. Lee et al. 2001, Spalik et al. 2001a, 2001b, Valiejo-Roman et al. 2002, Liu et al. 2003), and relatively few developments have been made in the familial and subfamilial classification of Apiaceae during the past two years. Unfortunately, several different definitions of Apiaceae are currently in use, including the traditional circumscription (Drude 1898 as updated by Pimenov and Leonov 1993), a much more expansive circumscription that includes both Apiaceae (sensu Drude) and Araliaceae (e.g. Judd et al. 2002), and a narrower concept generally referred to as ‘core Apiaceae’ (originally coined by Plunkett et al. 1997). Core Apiaceae are in large part consistent with Drude’s familial and subfamilial concepts, but not his tribes, which rarely correspond to monophyletic groups (see Downie and KatzDownie 1996, Downie et al. 1996, 2000, 2001, Plunkett et al. 1996b, Plunkett and Downie 1999). At the subfamilial level, the greatest taxonomic impact of recent phylogenetic studies has been the recognition that the apiaceous subfamily Hydrocotyloideae is polyphyletic, with no fewer than three unrelated clades of hydrocotyloids spread throughout Apiales (e.g. Plunkett et al. 1997, Plunkett and Lowry 2001, Chandler and Plunkett 2004) (Figure 1). Of those hydrocotyloids sampled to date, a large clade approximates Drude’s original circumscription of this subfamily, including Azorella, Bolax, Bowlesia, Dichosciadium, Dickinsia, Diplapsis, Eremocharis, Gymnophyton, Huanaca, Mulinum, Schizeilema and Spananthe, plus Stilbocarpa (which has been transferred from Araliaceae; see Mitchell et al. 1999) and possibly Klotzchia. This clade, however, cannot retain the name Hydrocotyloideae because it does not contain the type genus, Hydrocotyle, which instead forms a subclade with Trachymene within Araliaceae (a result that is strongly supported in all studies based on molecular data; Plunkett et al. 1996a, 1997, Downie et al. 2000, Plunkett and Lowry 2001, Chandler and Plunkett 2004, but see also Henwood and Hart 2001). Downie et al. (2000) used the informal name ‘Azorella clade’ to describe the ‘hydrocotyloids’ that remain in Apiaceae, but refrained from recognising it formally1, a
1
Plunkett, Chandler, Lowry, Pinney and Sprenkle
step that we now take with the description of Azorelloideae as a new subfamily of Apiaceae in Table 1. A third group of ‘hydrocotyloids’ forms a clade together with two genera traditionally referred to Araliaceae, Mackinlaya and Apiopetalum. This clade is currently recognised either as a tribe (Mackinlayeae, e.g. in Plunkett and Lowry 2001), a family (Mackinlayaceae, Doweld 2001), or simply as the ‘Mackinlaya group’ (Chandler and Plunkett 2004). In addition to Mackinlaya and Apiopetalum, it includes the ‘hydrocotyloid’ genera Centella, Micropleura, Actinotus, Platysace, Xanthosia and possibly other, as yet unsampled, genera (see Plunkett et al. 1997, Downie et al. 2000, Plunkett and Lowry 2001, Chandler and Plunkett 2004). In the study of Chandler and Plunkett (2004), a Bayesian inference tree based on combined 26S rDNA, rbcL, and matK data place the Mackinlaya group as sister to core Apiaceae (Figure 1), in agreement with an earlier study using matK and ITS rDNA data (Plunkett and Lowry 2001). These results are supported by several morphological characters that serve to link the Mackinlaya group to Apiaceae, such as laterally-compressed bicarpellate fruits, valvate petals that are both clawed and inflexed, and sheathing petiole bases. For this reason, we recognise the Mackinlaya group as a distinct subfamily of Apiaceae, which is likewise formally published as Mackinlayoideae in Table 1. Great progress has recently been made in re-circumscribing subfamily Saniculoideae. In their study of fruit structure, Liu et al. (2003) refined the list of carpological features that characterise Saniculoideae, which now includes (with some exceptions) the presence of non-lignified endocarps, mericarp outgrowths, and prominent intrajugal secretory ducts (also called ‘companion canals’ by Tseng 1967), plus the lack of true vittae in the commissures. On the basis of these findings, several taxa can now be added to or removed from the subfamily, and in most cases these transfers are confirmed by molecular data. For instance, both matK data (Plunkett et al. 1996b) and ITS data (Valiejo-Roman et al. 2002) indicate that Lagoecia should be removed from Saniculoideae and placed in Apioideae, and this finding is supported by the presence of commissural vittae in Lagoecia, a feature lacking in all other saniculoids but common among the apioids (Liu et al. 2003). Fruit anatomical characters also suggest that Saniculoideae should be broadened to include three former apioids, Lichtensteinia, Polemanniopsis and Steganotaenia, plus the former ‘hydrocotyloid’ Arctopus, a finding that agrees with ITS sequence data for Polemanniopsis and Steganotaenia (see Downie and Katz-Downie 1999), as well as data sets based on ITS + matK (Plunkett and Lowry 2001) and 26S rDNA + rbcL + matK (Chandler and Plunkett 2004) for Arctopus. Much work remains to be completed in subfamily Apioideae, which includes about 400 genera and over 3 000 species (Pimenov and Leonov 1993). The study of Downie et al. (2001) remains the best overview of the status of intergeneric taxa among the apioids. In recognising only monophyletic groups, Downie et al. (2001) reported ten formal tribes and seven informally-named groups currently accepted on the basis of recent molecular studies. However, these
Downie et al. (2000) indicated that Cerceau-Larrival (1962) invalidly and illegitimately published the name Azorelloideae. However, her use of this name was not accompanied by the designation of type nor by a Latin diagnosis; thus it was not validly published and remains available.
South African Journal of Botany 2004, 70: 371–381
373
91
Apiales
97
90 100 100 98
100
65
100
100
100 100
100
100 100 100 100
100
99
100
100
99
100 100
100 100 100 100
100
100
100
100 100 100
100 100
99
100 100
98
100 100
98 100
91 99 99
86
100
100 100 100
82 86
86 98
100 100 100
86 100 99
86 96
100
98
100
100 100
100
100
100
100 100
96 100
100
70
100
100 100
100
Aegopodium Lagoecia Pimpinella Apium Petroselinum Arracacia Donnellsmithia Angelica Apioideae Endressia Coriandrum Neogoezia Aciphylla Daucus Bupleurum Anginon Heteromorpha Eryngium Sanicula Saniculoideae Astrantia Arctopus Azorella sel. Huanaca Azorella trif. Mulinum chil. Mulinumsp. Diplaspis Azorelloideae Spananthe Bolax Dichosciadium Bowlesia Eremocharis Centella Micropleura Mackinlaya conf. Mackinlaya macr. Mackinlayoideae Apiopetalum Actinotus Platysace Hymenosporum Pittosporum Rhytidosporum Marianthus Sollya Delarbrea Pseudosciadium Myodocarpus Dendropanax Fatsia Trevesia Hedera Kalopanax Eleutherococcus Polyscias Astrotricha Osmoxylon Tetrapanax Tupidanthus Gastonia Reynoldsia Munroidendron Tetraplasandra Arthrophyllum Meryta Pseudopanax Aralia Panax Schefflera Cheirodendron Cussonia Hydrocotyle mod. Hydrocotyle vert. Hydrocotyle bowl. Trachymene Griselinia Melanophylla Torricellia Aralidium Pennantia Helianthus Menyanthes Campanula Dipsacus Valeriana Alangium Cornus Nyssa Schizophragma
Apiaceae
Pittosporaceae Myodocarpaceae
Araliaceae
Griseliniaceae Torricelliaceae Pennantiaceae
Outgroups
Figure 1: Phylogenetic tree showing relationships among 87 representative taxa of Apiales and outgroups based on the analysis of 5 958 characters derived from a combined data set of nuclear 26S rDNA and plastid rbcL + matK sequences using Bayesian inference (with a GTR + I model of nucleotide substitutions). Posterior probability scores are indicated above each node. Redrawn from Chandler and Plunkett 2004
374
17 clades represent only 171 genera (less than half the total number) and just a small fraction of the species currently included in Apioideae. Moreover, many of these genera show evidence of non-monophyly. In short, despite the many advances made in recent years, Apioideae remain both the largest and most vexing group in Apiales, and many years of additional study will be needed to resolve fully relationships within and among the apioid genera. Fortunately, the progress made over the past decade should provide a framework from which more detailed analyses can be undertaken. Araliaceae Overall relationships in Araliaceae are reviewed by Lowry et al. (2004), on the basis of two recent studies (Wen et al. 2001, Plunkett et al. 2004), necessitating only a brief overview here. First, all molecular studies confirm the removal of five genera from Araliaceae: Mackinlaya and Apiopetalum (now placed in Apiaceae subfamily Mackinlayoideae, discussed above), plus Delarbrea, Myodocarpus and Pseudosciadium (placed in the newly recognised family Myodocarpaceae, discussed below). After the additional transfer of Stilbocarpa to Apiaceae (see Mitchell et al. 1999), the remaining araliad genera form a monophyletic group, often referred to as ‘core Araliaceae’. This clade is the basis for our recircumscription of Araliaceae in the revised classification provided in Table 1. Among the major findings within this somewhat narrowed Araliaceae are the non-monophyly of the two largest genera: Schefflera, which is polyphyletic, comprising at least five unrelated clades (confirmed by a much broader study of Schefflera, Plunkett, Lowry, Frodin and Wen, in prep.), and Polyscias, which is rendered paraphyletic by the inclusion of Tetraplasandra, Munroidendron, Reynoldsia, Gastonia, Arthrophyllum and Cuphocarpus in the broad ‘Polyscias sensu lato’ clade (see also Plunkett et al. 2001). Taken as a whole, Araliaceae comprise three major clades plus as many as nine much smaller clades, whose interrelationships are poorly resolved (see figures in Lowry et al. 2004). The three large clades are (1) the Asian palmate group, centered in eastern and southeastern Asia but also including several Neotropical lineages, (including genera such as Eleutherococcus, Kalopanax, Brassaiopsis, Macropanax, Dendropanax, Fatsia, Hedera and Oplopanax, plus two clades of Schefflera (Asian and Neotropical)); (2) the Polyscias–Pseudopanax group, including both Polyscias sensu lato and a subclade uniting Pseudopanax (sensu Mitchell and Wagstaff 1997), Meryta, and most Pacific species of Schefflera; and (3) the Aralia–Panax group, including both of these genera and their former segregates (e.g. Pentapanax, Sciadodendron). Among the poorly resolved clades forming the basal polytomy are two additional clades of Schefflera (the ‘African–Malagasy’ and ‘section Schefflera’ clades), plus seven other clades: Cheirodendron + Raukaua, Cussonia + Seemannaralia, Harmsiopanax, Cephalaralia, Motherwellia, Osmoxylon and Astrotricha. As discussed above, Hydrocotyle and Trachymene form a clade sister to the large clade of araliads, and are herein transferred to Araliaceae (Figure 1, see also Lowry et al. 2004).
Plunkett, Chandler, Lowry, Pinney and Sprenkle
Myodocarpaceae Myodocarpaceae comprise three genera segregated from Araliaceae: Delarbrea, Myodocarpus and Pseudosciadium. The formal recognition of this new family by Doweld (2001) is based largely on our molecular data (Plunkett and Lowry 2001; see also Plunkett et al. 1996a, 1997, Plunkett 2001, Lowry et al. 2001, Chandler and Plunkett 2004), but is also supported by a unique morphological character found in no other member of Apiales (viz. specialised oil vesicles in the endocarp; see Lowry 1986a, 1986b). The recognition of very small families is often considered undesirable (e.g. Cronquist 1981: x–xi), and we agree that Myodocarpaceae (with only three genera and 17 species) should be accepted only if our dual criteria of monophyly and morphological coherence compel their recognition. Our data strongly indicate that Delarbrea, Myodocarpus and Pseudosciadium fall outside the main ‘core Araliaceae’ clade. The only approach that could unite these three genera to the other araliads in a monophyletic family would also necessitate the union of Araliaceae with both Apiaceae and Pittosporaceae. Given the distinctive morphologies of these families, and especially of Pittosporaceae, the union of these clades would yield a highly heterogeneous family. Similarly problematic would be the union of Myodocarpaceae with Pittosporaceae as a single family (see Figure 1), a solution that satisfies the criterion of monophyly but not of morphological coherence. Thus, we contend that Delarbrea, Myodocarpus and Pseudosciadium are best treated as a distinct family. Within Myodocarpaceae, phylogenetic relationships among the genera were first discussed by Lowry (1986a, 1986b), and later tested by Plunkett and Lowry (2001). Additional data are now available from the study of Sprenkle, Plunkett and Lowry (in prep., see also Sprenkle 2002), based on both plastid and nuclear sequence data from samples of all species and subspecies in each genus (Figure 2). Molecular data suggest that Myodocarpaceae comprise two well supported clades, Myodocarpus and Delarbrea + Pseudosciadium. These clades are also well differentiated on the basis of morphology, and in particular fruit types (the fleshy or spongy drupes of Delarbrea and Pseudosciadium, vs the dry, schizocarpic fruits of Myodocarpus). Pseudosciadium, however, is nested within the Delarbrea clade, necessitating the taxonomic transfer of its single species into Delarbrea (Lowry, Plunkett and Sprenkle in prep.). In Myodocarpus, the species with simple leaves and those having pinnately-compound leaves are found in distinct subclades. Plunkett et al. (1996a) postulated that simple leaves were ancestral throughout Apiales (see also Lowry et al. 2001, Plunkett 2001, Chandler and Plunkett 2004), but given the tree topology in Figure 2, and the fact that the species of Delarbrea (including Pseudosciadium) are entirely pinnate, the ancestral state cannot be unequivocally reconstructed in Myodocarpaceae. If we interpret pinnate leaves as ancestral in Myodocarpaceae, there must have been a single gain of this character in the common ancestor (from the simple leaves that characterised the rest of Apiales), followed by a reversal in the simple-leaved subclade of Myodocarpus. Alternatively, pinnate leaves may have evolved twice inde-
South African Journal of Botany 2004, 70: 371–381
pendently, once in the Delarbrea + Pseudosciadium clade, and a second time in the pinnately-compound subclade of Myodocarpus. Because both scenarios require a minimum of two steps, they are equally parsimonious. Studies of leaf development may offer some clues as to the evolution of this character. Biogeographically, Myodocarpaceae are clearly centred on New Caledonia, where all but two of the taxa are endemic (Lowry 1986a, 1986b). Delarbrea paradoxa subsp. paradoxa is more widely distributed across the south-western Pacific, and D. michieana is endemic to Queensland, Australia. The placement of D. paradoxa subsp. paradoxa in the cladogram (nested within an entirely New Caledonian clade) suggests a case of dispersal, but the position of D. michieana as the first-diverging lineage in the Delarbrea + Pseudosciadium clade could indicate a more ancient vicariance event associated with the separation of New Caledonia from Australia (c. 55MY; see Lowry 1998 and references therein). Pittosporaceae Evidence of a close relationship of Pittosporaceae to Apiales has been available for over a century, and includes data from stem structure (particularly the presence of schizogenous secretory canals, Van Tieghem 1884), ovular structure and development (Jurica 1922), cytology (Jay 1969), and phytochemistry (Hegnauer 1971, 1982, Dahlgren 1980, Jensen 1992). Choosing to dismiss these characters, some authors stressed the difference in ovary position (and to a lesser degree leaf shape) in separating Pittosporaceae from Apiales (e.g. Cronquist 1981, Eyde and Tseng 1971). However, the studies of Erbar and Leins (1995, 1996) have demonstrated developmental homology among the ovary positions of Pittosporaceae, Araliaceae and Apiaceae (see also Leins and Erbar 2004). These data, together with a growing body of molecular evidence (e.g. Plunkett et al. 1996a, Plunkett 2001, Chandler and Plunkett 2004, Kårehed 2003), unequivocally place Pittosporaceae within Apiales. However, the relationship of this family to the other clades in the order remains largely unresolved. The study of Chandler and Plunkett (2004) suggests a relationship of Pittosporaceae to Myodocarpaceae, which in turn may be sister to Apiaceae (Figure 1). Within Pittosporaceae, Cayzer et al. (1999a, 1999b, 2000a, 2000b, 2004, see also Cayzer 1998, Cayzer and Crisp 2004) have undertaken phylogenetic analyses based on morphological characters, resulting in several taxonomic changes. For example, all species of Sollya are now treated under Billardiera (Cayzer et al. 2004) and the species of Citriobatus are included within Pittosporum (Cayzer et al. 2000a). In addition, the new genus Auranticarpa has been erected for a clade of species formerly included within Pittosporum (Cayzer et al. 2000b). Altogether, Cayzer and her co-workers currently recognise nine genera in the family: Pittosporum, Billardiera, Bursaria, Marianthus, Auranticarpa, Rhytidosporum, Cheiranthera, Bentleya and Hymenosporum. Molecular data provide independent confirmation of Cayzer’s findings. A preliminary cladogram based on ITS +
375
trnL-trnF data (Chandler, Pinney, Cayzer and Plunkett, in prep.; see also Pinney 2002) is presented in Figure 3. This tree confirms the inclusion of Sollya within Billardiera. Moreover, Pittosporum is monophyletic only if the species previously placed in Citriobatus are included and those of Auranticarpa are excluded. In fact, the species of Auranticarpa are not closely related to Pittosporum, but rather form a clade sister to Bursaria + Rhytidosporum. Molecular data indicate that Marianthus may be paraphyletic (Figure 3), and the nature of this paraphyly may necessitate the union of all species assigned to Marianthus, Bentleya, and Billardiera (+ Sollya) into a single genus. Pittosporum is the largest of the nine genera (~140 species), accounting for nearly half of the species in the family. It is also the only genus with members found outside Australia (except the monotypic Hymenosporum, which occurs in both Australia and nearby New Guinea). To explore biogeographic relationships in the family, we examined trees resulting from two analyses, one based on ITS + trnL-trnF data (Figure 3), and another based on an expanded ITS data set that included sequences from the published study of Gemmill et al. (2002) (tree not shown). In agreement with Crisp et al. (1989), our preliminary results suggest that the family was probably not widely distributed across eastern Gondwanaland before its breakup, but rather was limited to Australia. Of the major clades in Pittosporaceae, only the Pittosporum clade has non-Australian species. Moreover, taxa from both the Pacific and Indian Ocean basins are nested within clades apparently originating in Australia, suggesting more recent dispersal events to relatively young island groups. Due to limited sample size (especially from the Indian Ocean basin) and the lack of strong support for many clades, these results are tentative and speculative, but they do suggest that Pittosporaceae may be a rich source for detailed biogeographic studies. Griselinia, Aralidium, Torricellia, Melanophylla and Pennantia Evidence for placing Araliaceae, Apiaceae, Pittosporaceae and Myodocarpaceae in a single order is now overwhelming, and includes information from morphological, anatomical, developmental, cytological, phytochemical, and molecular data (discussed above). Although relationships among these four families are not yet fully resolved, all phylogenetic studies unite these lineages into a single clade. Molecular data provide further evidence for expanding Apiales to include five additional genera: Griselinia, Aralidium, Torricellia, Melanophylla and Pennantia. The morphological support for their inclusion, however, is less straightforward (reviewed in Plunkett 2001, Chandler and Plunkett 2004, Kårehed 2003). For example, none of these taxa possesses the schizogenous secretory canals that so strikingly characterise the wood (and often other plant parts) of the four core families in Apiales. The basic chromosome number inferred for Apiales (x = 6 or 12; see Yi et al. in press) is shared by Torricellia and possibly Pennantia (n = 25, which may represent an aneuploid increase from n = 24, and thus x = 12), but not with Griselinia (x = 9) or Aralidium (x = 10) (this character remains unknown for Melanophylla) (reviewed in Plunkett
376
Plunkett, Chandler, Lowry, Pinney and Sprenkle
Myodocarpaceae 61
Myodocarpus nervatus Myodocarpus touretteorum
99
Myodocarpus simplicifolius
Myodocarpus crassifolius
50
Myodocarpus gracilis Myodocarpus involucratus 92
Myodocarpus lanceolatus
simple leaves
Myodocarpus simplicifolius
Myodocarpus simplicifolius
Myodocarpus
Myodocarpus vieillardii Myodocarpus vieillardii 70 97
Myodocarpus fraxinifolius Myodocarpus fraxinifolius
71
Myodocarpus fraxinifolius 99
98
Myodocarpus pinnatus Myodocarpus pinnatus Myodocarpus fraxinifolius
100
99 98
compound leaves
100
Delarbrea collina Delarbrea collina Pseudosciadium balansae
100
85
Delarbrea paradoxa ssp. depauperata Delarbrea paradoxa ssp. depauperata
82
100
Delarbrea paradoxa ssp. paradoxa 100
100
Delarbrea + Pseudosciadium
Delarbrea paradoxa ssp. paradoxa Delarbrea harmsii Delarbrea michieana
100
Astrotricha sp. Osmoxylon geelvinkianum Apiopetalum velutinum
Figure 2: Strict consensus of 17 most parsimonious trees based on a combined data set of nuclear rDNA spacers (ITS and partial ETS) and plastid trnL-trnF sequences from 29 taxa of Myodocarpaceae and outgroups. The total data set included 1 804 characters, of which 336 were potentially informative. The shortest trees were 569 steps long, with a consistency index of 0.815 and a retention index of 0.930. Bootstrap percentages are indicated at each node. Redrawn from Sprenkle, Plunkett and Lowry, in prep.
2001, Kårehed 2003). Similarly, both phytochemistry and wood anatomy provide conflicting evidence for the inclusion of these five genera in Apiales (reviewed in Plunkett 2001, Kårehed 2003). Plunkett et al. (1996a) originally hypothesised that bicarpelly was ancestral in Apiales, a result confirmed for the core families (e.g. Lowry et al. 2001, Plunkett 2001, Plunkett and Lowry 2001, Chandler and Plunkett 2004, Kårehed 2003), but not for Griselinia, Aralidium, Torricellia, Melanophylla and Pennantia. In these genera, the number of locules is variable, but the ovaries are clearly derived from three carpels, and in all cases, only a single functional ovule is produced (see Watson and Dallwitz 1992, Dillon and Muñoz-Schick 1993, Plunkett 2001, Kårehed
2003). In Torricellia and Melanophylla, all three carpels persist, yielding three locules (reports of ‘(2–)3’ or ‘3(–4)’ appear to be due to irregularly-shaped locules that may be either overlooked or counted twice, respectively (GM Plunkett, pers. obs.)). In Griselinia, only a single locule is discernable, and in both Aralidium and Pennantia, two abortive locules are located adjacent to the single, fully developed one. Thus, it appears that tricarpelly was ancestral in Apiales, followed by a single reduction to two carpels only after the divergence of these five genera (see also Chandler and Plunkett 2004, Kårehed 2003).
South African Journal of Botany 2004, 70: 371–381
377
Pittosporaceae 69
Billardiera cymosa Billardiera lehmanniana
62 99
Billardiera (= Sollya) heterophylla Billardiera (= Sollya) fusiformis
Billardiera (incl. Sollya)
Billardiera longifolia Billardiera fraseria 60
Marianthus ringens Marianthus bicolor
93
Bentlya spinescens Cheiranthera linearis Hymenosporum flavum 80
Bursaria incana Bursaria spinosa Rhytidosporum alpinum
72 100 56
Bentleya Cheiranthera Hymenosporum Bursaria Rhytidosporum
Auranticarpa papyracea Auranticarpa edentata
100
Marianthus
Auranticarpa
Auranticarpa ilicifolia 60 93
Pittosporum oreillyanum Pittosporum (=Citriobatus) lancifolium Pittosporum (=Citriobatus) spinescens Pittosporum (=Citriobatus) multiflorum
98
Pittosporum trilobium Pittosporum rubignosum
85
Pittosporum suberosum Pittosporum moluccanum
83 80
Pittosporum (incl. Citriobatus)
Pittosporum venulosum Pittosporum ferrigineum
100
Pittosporum angustifolium Pittosporum ligustrifolium Pittosporum wingii Pittosporum eugenioides Pittosporum crassifolium
100
Myodocarpus fraxinifolius Delarbrea collina
Outgroups
Schefflera baillonii
Figure 3: Strict consensus of 116 most parsimonious trees based on a combined data set of ITS nuclear rDNA and plastid trnL-trnF sequences from 35 taxa of Pittosporaceae and outgroups. The total data set included 1 705 characters, of which 299 were potentially informative. The shortest trees were 737 steps long, with a consistency index of 0.581 and a retention index of 0.719. Bootstrap percentages are indicated above each node. Redrawn from Chandler, Pinney, Cayzer and Plunkett, in prep.
378
Revised Classification of the Order Apiales The manner of translating phylogenetic studies into formal systems of classification is currently a subject of great debate (e.g. De Queiroz and Gauthier 1994, Nixon and Carpenter 2000, Langer 2001, Sennblad and Bremer 2002), but systematists appear to be united in espousing the principle of stability. In Apiales, many of the traditional taxonomic groups are now known to be non-monophyletic and are widely acknowledged to be in need of substantial realignment. However, given the size of the order (representing over 5 000 species), most students of Apiales have been reluctant to make formal taxonomic changes based on the sampling of relatively few taxa, especially when the composition of some clades appeared to be unstable from study to study. Nonetheless, there is an obvious need to name clades for the purpose of discussion and as a framework for further investigations. In the past, we have advocated the use of informal clade names (e.g. ‘core Apiaceae’, the ‘Asian palmate clade’, the ‘apioid superclade’, the ‘Mackinlaya group’, etc.) until a unified picture of phylogenetic relationships has emerged. To a large degree, such stability in our understanding of Apiales has now been achieved at the interfamilial level. Moreover, the multiplication of informal names (often with similar wording, but defining slightly different and overlapping groups) has reached the point where it is hindering rather than helping communication. For example, there are at least three informal definitions of Apiaceae in current usage (viz. the ‘traditional Apiaceae’ (sensu Drude), ‘Apiaceae sensu lato’ (including Araliaceae), and ‘core Apiaceae’ (excluding many ‘hydrocotyloids’)), as well as a number of competing notions of both Araliaceae and the entire order. Several authors have provided formal names or recircumscriptions affecting Apiales either at higher levels of classification (e.g. the ordinal system of the APG 1998, 2003) or at the intergeneric level (e.g. Downie et al. 2000 for core Apiaceae), but the time is now ripe to offer an interfamilial classification of Apiales as a whole. In the system that follows, we recognise groups on the basis of two criteria, monophyly (usually determined by phylogenetic studies based on molecular data) and morphological coherence. Using these two criteria, we translate the current concept of ‘core Araliaceae’ into a formal recircumscription of that family, which now excludes six former araliad genera (Myodocarpus, Delarbrea, Pseudosciadium, Apiopetalum, Mackinlaya and Stilbocarpa) and includes two former ‘hydrocotyloid’ genera (Hydrocotyle and Trachymene, leaving open the potential to add additional hydrocotyloids as more taxa are sampled). In a similar fashion, we treat the three major clades of ‘core Apiaceae’ (the apioids, the saniculoids and the ‘Azorella clade’) plus one additional clade (the ‘Mackinlaya group’) as four distinct subfamilies in a recircumscribed Apiaceae. Some authors have chosen to unite Apiaceae and Araliaceae into a single family (e.g. Judd et al. 2002), but our data indicate that this cannot be done without also including Pittosporaceae (which is placed as sister either to Apiaceae or Araliaceae; see Plunkett and Lowry 2001, Chandler and Plunkett 2004). We believe the union of these taxa into a single, broadly defined family yields a group with an inordinate amount of morpho-
Plunkett, Chandler, Lowry, Pinney and Sprenkle
logical heterogeneity, making it difficult to characterise or recognise. Thus, both Pittosporaceae and the small (but well supported and morphologically coherent) Myodocarpaceae are recognised as distinct families. Perhaps the most difficult issue to resolve in the classification of Apiales is also the most fundamental: how broadly to define the order. The four ‘core’ families described above (Apiaceae, Araliaceae, Pittosporaceae and Myodocarpaceae) represent a monophyletic group that shares many features (e.g. anatomy, cytology, phytochemistry). Molecular data agree in placing Griselinia, Aralidium, Torricellia, Melanophylla and Pennantia as successively sister to the four-family clade, but these five genera lack many of the distinctive features uniting the four core families. Each of these five genera has been recognised as its own order, but we believe that monogeneric (and in the case of Aralidium, monotypic) orders are redundant and offer no additional taxonomic information. Thus, we concur with earlier proposals (e.g. APG 1998, 2003, Plunkett 2001, Chandler and Plunkett 2004, Kårehed 2003) that include Griselinia, Aralidium, Torricellia, Melanophylla and Pennantia within Apiales, placed in three families: Pennantiaceae (monogeneric), Griseliniaceae (monogeneric), and Torricelliaceae (broadened to include Torricellia, Aralidium and Melanophylla). However, because the ‘four-family clade’ (comprising Apiaceae, Araliaceae, Pittosporaceae and Myodocarpaceae) is well supported, morphologically coherent, and the focus of most of the research being conducted within the order, we contend that it should be recognised formally. The application of an informal name (such as ‘core Apiales’) might appeal to some, but recent experience (at least in Apiales) suggests that the definitions and circumscriptions of such names are often unstable and can lead to considerable confusion. We therefore provide a formal name (with an explicit circumscription) in recognising a new suborder, Apiineae. Despite suggestions to the contrary, the principle of exhaustive subsidiary taxa is not a requirement of the ICBN (Greuter et al. 2000), and thus subordinal names are neither required nor provided for the other three families in the order. A synopsis of the revised classification is presented in Table 1. Acknowledgements — The authors gratefully acknowledge Prof. Ben-Erik van Wyk and Dr Patricia Tilney for organising the 4th International Apiales Symposium (Pretoria) and this proceedings volume, and for kindly inviting our contributions to both; thanks are also extended to Dr Gordon McPherson for proofreading the Latin diagnoses of the new taxa. Support was provided by grants from the National Science Foundation (DEB-9981641) and the National Geographic Society (5793-96) to GMP and PPL.
References APG (Angiosperm Phylogeny Group) (1998) An ordinal classification for families of flowering plants. Annals of the Missouri Botanical Garden 85: 531–553 APG (Angiosperm Phylogeny Group) (2003) An update of the Angiosperm Phylogeny Group classification for the orders and families of flowering plants: APG II. Botanical Journal of the Linnean Society 141: 399–436 Bentham G, Hooker JD (1867) Genera Plantarum. A Black, W Pamplin, Lovell Reeve and Co., London
South African Journal of Botany 2004, 70: 371–381
379
Table 1: Revised classification of the angiosperm order Apiales Order : APIALES Nakai 1930 Family: TORRICELLIACEAE (Wang.) Hu 1934 (only Torricellia, Melanophylla, Aralidium) Family: GRISELINIACEAE J.R. Forst. & G. Forst. 1839 (only Griselinia) Family: PENNANTIACEAE J. Agardh 1858 (only Pennantia) Suborder : APIINEAE, Plunkett & Lowry, subord. nov. Lignum canalibus secretoriis schizogenis; gynoecium oriunum a duo carpellis. T: Apium L. Family: APIACEAE Lindl. 1836 (nom. alt., Umbelliferae Durande 1782) Subfamily: APIOIDEAE Seem. 1866 Subfamily: SANICULOIDEAE Burnett 1835 Subfamily: AZORELLOIDEAE Plunkett & Lowry, subfam. nov. Fructus teres vel a dorso compressus, sine vittis; canales intra juga praesentes; endocarpium induratum; stratum interius mesocarpii lignescens vel parenchymatum, crystallis rhombeis (sed crystallis compositis in strato exteriore Azorellae); carpella secedentes in fructu maturo.T: Azorella Lam. (also incl. Bolax, Bowlesia, Dichosciadium, Dickinsia, Diplapsis, Eremocharis, Gymnophyton, Huanaca, Mulinum, Schizeilema, Spananthe, Stilbocarpa). Syn.: Hydrocotyloideae Link 1829, pro parte, excl. Hydrocotyle Subfamily: MACKINLAYOIDEAE Plunkett & Lowry, subfam. nov. Fructus a latere compressus (vel teres in Apiopetalo), sine vittis; canales intra juga minuti vel obsoleti; endocarpium sclerescens; stratum interius mesocarpii lignescens vel parenchymatum, crystallis rhombeis; carpella non secedentes in fructu maturo. T: Mackinlaya F. Muell. (also incl. Apiopetalum, Centella, Micropleura, Actinotus, Platysace, Xanthosia). Syn.: Mackinlayeae Hook.f. 1873.; Mackinlayaceae Doweld 2001 Family: ARALIACEAE Durande 1782 (incl. Hydrocotyle, Trachymene; excl. Apiopetalum, Delarbrea, Mackinlaya, Myodocarpus, Pseudosciadium, Stilbocarpa) Family: MYODOCARPACEAE Doweld 2001 Syn: Myodocarpeae R. Vig. 1906 Family: PITTOSPORACEAE R. Br. 1814
Bessey CE (1915) Phylogenetic taxonomy of flowering plants. Annals of the Missouri Botanical Garden 2: 109–164 Cayzer LW (1998) Revision of the Family Pittosporaceae in Australia. PhD Thesis, Australian National University, Canberra, Australia Cayzer LW, Crisp MD (2004) Reinstatement and revision of the genus Marianthus (Pittosporaceae). Australian Systematic Botany 17: 127–144 Cayzer LW, Crisp MD, Telford IRH (1999a) Bursaria (Pittosporaceae): a morphometric analysis and revision. Australian Systematic Botany 12: 117–143 Cayzer LW, Crisp MD, Telford IRH (1999b) Revision of Rhytidosporum (Pittosporaceae). Australian Systematic Botany 12: 689–708 Cayzer LW, Crisp MD, Telford IRH (2000a) Revision of Pittosporum (Pittosporaceae) in Australia. Australian Systematic Botany 13: 845–902 Cayzer LW, Crisp MD, Telford IRH (2000b) Auranticarpa, a new genus of Pittosporaceae from northern Australia. Australian Systematic Botany 13: 903–917 Cayzer LW, Crisp MD, Telford IRH (2004) Cladistic analysis and revision of Billardiera (Pittosporaceae). Australian Systematic Botany 17: 83–125 Cerceau-Larrival MT (1962) Plantules et pollens d’Ombellifères: leur intérêt systématique et phylogénetique. Mémoires du Muséum National d’Histoire Naturelle, Série B, Botanique 14: 1–164 Chandler GT, Plunkett GM (2004) Evolution in Apiales: nuclear and chloroplast markers together in (almost) perfect harmony. Botanical Journal of the Linnean Society 144: 123–147 Constance L (1971) History of the classification of Umbelliferae (Apiaceae). In: Heywood VH (ed) The biology and chemistry of the Umbelliferae. Botanical Journal of the Linnean Society (Supplement 1) 64: 1–8 Crisp MD, Taylor JM, Bennett EM (1989) Pittosporaceae — an old east Gondwanan family? Australian Systematic Botany Society
Newsletter 60: 20 Cronquist A (1981) An Integrated System of Classification of Flowering Plants. Columbia University Press, New York Cronquist A (1988) The Evolution and Classification of Flowering Plants (2nd edn). New York Botanical Garden, Bronx, New York Dahlgren RMT (1980) A revised system of classification of the angiosperms. Botanical Journal of the Linnean Society 80: 91–124 De Queiroz K, Gauthier J (1994) Toward a phylogenetic system of biological nomenclature. Trends in Ecology and Evolution 9: 27–31 Dillon MO, Muñoz-Schick M (1993) A revision of the dioecious genus Griselinia (Griseliniaceae), including a new species from the coastal Atacama Desert of northern Chile. Brittonia 45: 261–274 Doweld A (2001) Prosyllabus Tracheophytorum, Tentamen Systematis Plantarum Vascularium (Tracheophyta) (An attempted system of the vascular plants). Pensoft, Moscow Downie SR, Katz-Downie DS (1996) A molecular phylogeny of Apiaceae subfamily Apioideae: evidence from nuclear ribosomal DNA internal transcribed spacer sequences. American Journal of Botany 83: 234–251 Downie SR, Katz-Downie DS (1999) Phylogenetic analysis of chloroplast rps16 intron sequences reveals relationships within the woody southern African Apiaceae subfamily Apioideae. Canadian Journal of Botany 77: 1120–1135 Downie SR, Katz-Downie DS, Cho K-J (1996) Phylogenetic analysis of Apiaceae subfamily Apioideae using nucleotide sequences from the chloroplast rpoC1 intron. Molecular Phylogenetics and Evolution 6: 1–18 Downie SR, Katz-Downie DS, Watson MF (2000) A phylogeny of the flowering plant family Apiaceae based on chloroplast DNA rpl16 and rpoC1 intron sequences: towards a suprageneric classification of subfamily Apioideae. American Journal of Botany 87: 273–292
380
Downie SR, Plunkett GM, Watson MF, Spalik K, Katz-Downie DS, Valiejo-Roman CM, Terentieva EI, Troitsky AV, Lee B-Y, Lahham J, El-Oqlah A (2001) Tribes and clades within Apiaceae subfamily Apioideae: the contribution of molecular data. Edinburgh Journal of Botany 58: 301–330 Drude CGO (1898) Umbelliferae. In: Engler A, Prantl K (eds) Die Natürlichen Pflanzenfamilien, Vol 3. Wilhelm Engelmann, Leipzig, pp 63–250 Engler A, Prantl K (1898) Die Natürlichen Planzenfamilien. Engelmann, Leipzig Erbar C, Leins P (1995) An analysis of the early floral development of Pittosporum tobira (Thunb.) Aiton and some remarks on the systematic position of the family Pittosporaceae. Feddes Repertorium 106: 463–473 Erbar C, Leins P (1996) Distribution of the character states “early sympetaly” and “late sympetaly” within the “Sympetalae Tetracyclicae” and presumably allied groups. Botanica Acta 109: 427–440 Eyde RH, Tseng CC (1971) What is the primitive floral structure of Araliaceae? Journal of the Arnold Arboretum 52: 205–239 Gemmill CEC, Allan GJ, Wagner WL, Zimmer EA (2002) Evolution of insular Pacific Pittosporum (Pittosporaceae): origin of the hawaiian radiation. Molecular Phylogenetics and Evolution 22: 31–42 Greuter W, McNeill J, Barrie FR, Burdet H-M, Demoulin V, Filgueiras TS, Nicolson DH, Silva PC, Skog JE, Trehane P, Turland NJ, Hawksworth DL (2000) International Code of Botanical Nomenclature (St Louis Code). Regnum Vegetabile 138. Koeltz Scientific Books, Königstein Hegnauer R (1971) Chemical patterns and relationships of Umbelliferae. In: Heywood VH (ed) The Biology and Chemistry of the Umbelliferae. Botanical Journal of the Linnean Society (Supplement 1) 64: 267–277 Hegnauer R (1982) Phytochemie und Klassifikation der Umbelliferen, eine Neubewertung im Lichte der Siet 1972 bekannt gewordenen phytochemischen Tatsachen. In: Cauwet-Marc AM, Carbonnier J (eds) Actes du 2e Symposium International sur les Ombellifères, Contributions Pluridisciplinaires à la Systématique. Missouri Botanical Garden, St Louis, pp 335–364 Henwood MJ, Hart JM (2001) Towards an understanding of the phylogenetic relationships of Australian Hydrocotyloideae (Apiaceae). Edinburgh Journal of Botany 58: 269–289 Heywood VH (1978) Introduction to the taxonomy of the Umbelliferae. In: Cauwet-Marc A-M, Carbonnier J (eds) Actes du 2e Symposium International sur les Ombellifères, Contributions Pluridisciplinaires à la Systématique. Missouri Botanical Garden, St Louis, pp 107–112 Hutchinson J (1967) The Genera of Flowering Plants, Vol. II. Oxford University Press, London Hutchinson J (1973) The Families of Flowering Plants. Oxford University Press, London Jay M (1969) Chemotaxonomic researches on vascular plants. XIX. Flavonoid distribution in the Pittosporaceae. Botanical Journal of the Linnean Society 62: 423–429 Jensen SR (1992) Systematic implications of the distribution of iridoids and other chemical compounds in the Loganiaceae and other families of the Asteridae. Annals of the Missouri Botanical Garden 79: 284–302 Judd WS, Campbell CS, Kellogg EA, Stevens PF, Donoghue MJ (2002) Plant systematics: A Phylogenetic Approach (2nd edn). Sinauer Associates, Sunderland, Massachusetts Jurica HS (1922) A morphological study of the Umbelliferae. Botanical Gazette 74: 292–307 Kårehed J (2003) The family Pennantiaceae and its relationships to Apiales. Botanical Journal of the Linnean Society 141: 1–24 Langer MC (2001) Linnaeus and the PhyloCode: where are the differences? Taxon 50: 1091–1096
Plunkett, Chandler, Lowry, Pinney and Sprenkle
Lee B-Y, Levin GA, Downie SR (2001) Relationships within the spiny-fruited umbellifers (Scandiceae subtribes Daucinae and Torilidinae) as assessed by phylogenetic analysis of morphological characters. Systematic Botany 26: 622–642 Leins P, Erbar C (2004) Organ sequences in Apiales (Apiaceae, Araliaceae, Pittosporaceae). South African Journal of Botany 70: 468–474 Liu M, Van Wyk B-E, Tilney PM (2003) The taxonomic value of fruit structure in the subfamily Saniculoideae and related African genera (Apiaceae). Taxon 52: 261–270 Lowry PP II (1986a) A Systematic Study of Three Genera of Araliaceae Endemic to or Centered on New Caledonia: Delarbrea, Myodocarpus, and Pseudosciadium. PhD Thesis, Washington University, St Louis, Missouri, USA Lowry PP II (1986b) A systematic study of Delarbrea Vieill. (Araliaceae). Allertonia 4: 169–201 Lowry PP II (1998) Diversity, endemism, and extinction in the flora of New Caledonia: a review. In: Peng C-I, Lowry PP II (eds) Proceedings of the International Symposium on Rare, Threatened, and Endangered Floras of Asia and the Pacific. Academia Sinica, Taipei, Taiwan, pp 181–206 Lowry PP II, Plunkett GM, Oskolski AA (2001) Early lineages in Apiales: insights from morphology, wood anatomy and molecular data. Edinburgh Journal of Botany 58: 207–220 Lowry PP II, Plunkett GM, Wen J (2004) Generic relationships in Araliaceae: looking into the crystal ball. South African Journal of Botany 70: 382–392 Mitchell AD, Wagstaff SJ (1997) Phylogenetic relationships of Pseudopanax species inferred from parsimony analysis of rDNA sequence data and morphology. Plant Systematics and Evolution 208: 121–138 Mitchell AD, Meurk CD, Wagstaff SJ (1999) Evolution of Stilbocarpa, a megaherb from New Zealand’s sub-antarctic islands. New Zealand Journal of Botany 37: 205–211 Nixon KC, Carpenter JM (2000) On the other “Phylogenetic Systematics”. Cladistics 16: 298–318 Pimenov MG, Leonov MV (1993) The Genera of the Umbelliferae: A Nomenclator. Royal Botanic Gardens, Kew Pinney SM (2002) A Molecular Systematic Study of Pittosporaceae. MSc Thesis, Virginia Commonwealth University, Richmond, Virginia, USA Plunkett GM (2001) Relationships of the order Apiales to subclass Asteridae: a reevaluation of morphological characters based on insights from molecular data. Edinburgh Journal of Botany 58: 183–200 Plunkett GM, Downie SR (1999) Major lineages within Apiaceae subfamily Apioideae: a comparison of chloroplast restriction site and DNA sequence data. American Journal of Botany 86: 1014–1026 Plunkett GM, Lowry PP II (2001) Relationships among “ancient araliads” and their significance for the systematics of Apiales. Molecular Phylogenetics and Evolution 19: 259–276 Plunkett GM, Lowry PP II, Burke MK (2001) The phylogenetic status of Polyscias (Araliaceae) based on nuclear ITS sequence data. Annals of the Missouri Botanical Garden 88: 213–230 Plunkett GM, Soltis DE, Soltis PS (1996a) Higher level relationships of Apiales (Apiaceae and Araliaceae) based on phylogenetic analysis of rbcL sequences. American Journal of Botany 83: 499–515 Plunkett GM, Soltis DE, Soltis PS (1996b) Evolutionary patterns in Apiaceae: inferences based on matK sequence data. Systematic Botany 21: 477–495 Plunkett GM, Soltis DE, Soltis PS (1997) Clarification of the relationship between Apiaceae and Araliaceae based on matK and rbcL sequence data. American Journal of Botany 84: 565–580 Plunkett GM, Wen J, Lowry PP II (2004) Infrafamilial classifications
South African Journal of Botany 2004, 70: 371–381
and characters in Araliaceae: insights from the phylogenetic analysis of nuclear (ITS) and plastid (trnL-trnF) sequence data. Plant Systematics and Evolution 245: 1–39 Sennblad B, Bremer B (2002) Classification of Apocynaceae s.l. according to a new approach combining Linnaean and phylogenetic taxonomy. Systematic Biology 51: 389–409 Spalik K, Wojewódzka A, Downie SR (2001a) The evolution of fruit in Scandiceae subtribe Scandicinae (Apiaceae). Canadian Journal of Botany 79: 1358–1374 Spalik K, Wojewódzka A, Downie SR (2001b) Delimitation of genera in Apiaceae with examples from Scandiceae subtribe Scandicinae. Edinburgh Journal of Botany 58: 331–346 Sprenkle TS (2002) Molecular Systematics of tribe Myodocarpeae (Araliaceae s.l.). MSc Thesis, Virginia Commonwealth University, Richmond, Virginia, USA Stevens PF (2003) Angiosperm Phylogeny Website, Version 4. University of Missouri — St Louis and Missouri Botanical Garden. Available at: http://www.mobot.org/MOBOT/ Research/APweb/ Takhtajan A (1987) System of Magnoliophyta. Academy of Sciences, Leningrad
Edited by B-E van Wyk and PM Tilney
381
Thorne RF (1992) Classification and geography of flowering plants. Botanical Review 58: 225–348 Tseng CC (1967) Anatomical studies of flower and fruit in the Hydrocotyloideae (Umbelliferae). University of California Publications in Botany 30: 1–79 Valiejo-Roman CM, Terentieva EI, Samigullin TH, Pimenov MC (2002) Relationships among genera in Saniculoideae and selected Apioideae (Umbelliferae) inferred from nrITS sequences. Taxon 51: 91–101 Van Tieghem P (1884) Sur la structure et les affinités des Pittosporées. Bulletin de la Société Botanique de France 31: 384–385 Watson L, Dallwitz MJ (1992 and onwards) The families of flowering plants: descriptions, illustrations, identification, and information retrieval. Available at: http://biodiversity.uno.edu/delta/angio/ Wen J, Plunkett GM, Mitchell AD, Wagstaff SJ (2001) The evolution of Araliaceae: a phylogenetic analysis based on ITS sequences of nuclear ribosomal DNA. Systematic Botany 26: 144–167 Yi T, Lowry PP II, Plunkett GM, Wen J (in press) Chromosomal Evolution in Araliaceae s. lat. Taxon
South African Journal of Botany 2004, 70(3): 371–381 Printed in South Africa — All rights reserved
SOUTH AFRICAN JOURNAL OF BOTANY ISSN 0254–6299
Recent advances in understanding Apiales and a revised classification GM Plunkett1*, GT Chandler1,2, PP Lowry II3, SM Pinney1 and TS Sprenkle1 Department of Biology, Virginia Commonwealth University, PO Box 842012, Richmond, Virginia 23284-2012, United States of America 2 Present address: Department of Biology, University of North Carolina, Wilmington, North Carolina 28403-5915, United States of America 3 Missouri Botanical Garden, PO Box 299, St Louis, Missouri 63166-0299, United States of America; Département de Systématique et Evolution, Muséum National d’Histoire Naturelle, Case Postale 39, 57 rue Cuvier, 75231 Paris CEDEX 05, France * Corresponding author, e-mail: [email protected] 1
Received 23 August 2003, accepted in revised form 18 November 2003 Despite the long history of recognising the angiosperm order Apiales as a natural alliance, the circumscription of the order and the relationships among its constituent groups have been troublesome. Recent studies, however, have made great progress in understanding phylogenetic relationships in Apiales. Although much of this recent work has been based on molecular data, the results are congruent with other sources of data, including morphology and geography. A unified picture of relationships has now emerged regarding the delimitation of
Apiales, which includes a core group of four families (Apiaceae, Araliaceae, Myodocarpaceae, Pittosporaceae) to which three small families are also added (Griseliniaceae, Torricelliaceae and Pennantiaceae). After a brief review of recent advances in each of the major groups, a revised classification of the order is presented, which includes the recognition of the new suborder Apiineae (comprising the four core families) and two new subfamilies within Apiaceae (Azorelloideae and Mackinlayoideae).
Introduction Almost all authors have considered the dicot families Apiaceae and Araliaceae closely related (e.g. Bentham and Hooker 1867, Engler and Prantl 1898, Bessey 1915, Dahlgren 1980, Cronquist 1981, 1988, Takhtajan 1987, Thorne 1992, but see Hutchinson 1967, 1973) and have usually treated them in the same order, albeit under various names, including Umbellales, Cornales (when associated with Cornaceae), Araliales, and Apiales (the name we shall use hereafter). The near-ubiquity of this interpretation stands in stark contrast to the diverse treatments both above and below the ordinal level, such as the placement of Apiales among the other orders of dicots, the inclusion of additional families within the order (Cornaceae, Torricelliaceae, Helwingiaceae, Griseliniaceae, etc.), and the many interand infrafamilial treatments of Apiaceae and Araliaceae themselves. As with many higher taxonomic groups, the difficulties in sorting out relationships in Apiales presumably stem from repeated cycles of divergence and migration, followed by convergence and parallelism. In the decades immediately preceding the development of objective approaches to phylogenetic analysis (e.g. parsimony, phenetic, and likelihood methods) and the use of molecular sources of data, few substantial advances were
made in the ordinal classification of Apiales (Dahlgren 1980 being the only notable exception). In the past ten years, however, a unified picture of relationships has rapidly emerged regarding both the circumscription and internal relationships of the order. Similar progress has been made in many other angiosperm orders, and several ongoing efforts are now underway to formalise the results of these advances (APG 1998, 2003, Stevens 2003). Most authors working on Apiales have hesitated to provide a formal recircumscription of the order during this ‘period of discovery’. The recent publication of a new apialean classification by an author with little or no expertise in the group (Doweld 2001), including the description of two new families based on our own work (viz. Plunkett and Lowry 2001), has prompted us to provide both a formal classification of Apiales together with a review of the relevant literature upon which our decisions are based. We hope this effort will also provide a fitting framework from which the other contributions to this symposium volume can be better understood. Several publications already provide comprehensive reviews of the taxonomic history of Apiales and/or their constituent groups (e.g. Constance 1971, Heywood 1978, Plunkett et al. 1996a, 1996b, 1997, Downie et al. 2001,
372
Plunkett and Lowry 2001, Plunkett 2001), and therefore this information is not repeated herein. Instead, we focus on reviewing (or in some cases previewing) results from recent studies (or unpublished work) completed since the publication of the preceding Apiales Special Issue of the Edinburgh Journal of Botany (2001, Volume 58, no. 2). Following this, we provide a formal recircumscription and interfamilial classification of the order. Apiaceae In their contributions to the last symposium volume, both Plunkett (2001) and Downie et al. (2001) explored relationships within Apiaceae using ‘supertree’ and ‘consensus’ approaches, respectively (Downie et al. 2001 also examined relationships within subfamily Apioideae using a ‘combined data’ approach). With the exception of Chandler and Plunkett (2004), most recent studies of Apiaceae have focused on relationships within subfamilies and/or among genera (e.g. Lee et al. 2001, Spalik et al. 2001a, 2001b, Valiejo-Roman et al. 2002, Liu et al. 2003), and relatively few developments have been made in the familial and subfamilial classification of Apiaceae during the past two years. Unfortunately, several different definitions of Apiaceae are currently in use, including the traditional circumscription (Drude 1898 as updated by Pimenov and Leonov 1993), a much more expansive circumscription that includes both Apiaceae (sensu Drude) and Araliaceae (e.g. Judd et al. 2002), and a narrower concept generally referred to as ‘core Apiaceae’ (originally coined by Plunkett et al. 1997). Core Apiaceae are in large part consistent with Drude’s familial and subfamilial concepts, but not his tribes, which rarely correspond to monophyletic groups (see Downie and KatzDownie 1996, Downie et al. 1996, 2000, 2001, Plunkett et al. 1996b, Plunkett and Downie 1999). At the subfamilial level, the greatest taxonomic impact of recent phylogenetic studies has been the recognition that the apiaceous subfamily Hydrocotyloideae is polyphyletic, with no fewer than three unrelated clades of hydrocotyloids spread throughout Apiales (e.g. Plunkett et al. 1997, Plunkett and Lowry 2001, Chandler and Plunkett 2004) (Figure 1). Of those hydrocotyloids sampled to date, a large clade approximates Drude’s original circumscription of this subfamily, including Azorella, Bolax, Bowlesia, Dichosciadium, Dickinsia, Diplapsis, Eremocharis, Gymnophyton, Huanaca, Mulinum, Schizeilema and Spananthe, plus Stilbocarpa (which has been transferred from Araliaceae; see Mitchell et al. 1999) and possibly Klotzchia. This clade, however, cannot retain the name Hydrocotyloideae because it does not contain the type genus, Hydrocotyle, which instead forms a subclade with Trachymene within Araliaceae (a result that is strongly supported in all studies based on molecular data; Plunkett et al. 1996a, 1997, Downie et al. 2000, Plunkett and Lowry 2001, Chandler and Plunkett 2004, but see also Henwood and Hart 2001). Downie et al. (2000) used the informal name ‘Azorella clade’ to describe the ‘hydrocotyloids’ that remain in Apiaceae, but refrained from recognising it formally1, a
1
Plunkett, Chandler, Lowry, Pinney and Sprenkle
step that we now take with the description of Azorelloideae as a new subfamily of Apiaceae in Table 1. A third group of ‘hydrocotyloids’ forms a clade together with two genera traditionally referred to Araliaceae, Mackinlaya and Apiopetalum. This clade is currently recognised either as a tribe (Mackinlayeae, e.g. in Plunkett and Lowry 2001), a family (Mackinlayaceae, Doweld 2001), or simply as the ‘Mackinlaya group’ (Chandler and Plunkett 2004). In addition to Mackinlaya and Apiopetalum, it includes the ‘hydrocotyloid’ genera Centella, Micropleura, Actinotus, Platysace, Xanthosia and possibly other, as yet unsampled, genera (see Plunkett et al. 1997, Downie et al. 2000, Plunkett and Lowry 2001, Chandler and Plunkett 2004). In the study of Chandler and Plunkett (2004), a Bayesian inference tree based on combined 26S rDNA, rbcL, and matK data place the Mackinlaya group as sister to core Apiaceae (Figure 1), in agreement with an earlier study using matK and ITS rDNA data (Plunkett and Lowry 2001). These results are supported by several morphological characters that serve to link the Mackinlaya group to Apiaceae, such as laterally-compressed bicarpellate fruits, valvate petals that are both clawed and inflexed, and sheathing petiole bases. For this reason, we recognise the Mackinlaya group as a distinct subfamily of Apiaceae, which is likewise formally published as Mackinlayoideae in Table 1. Great progress has recently been made in re-circumscribing subfamily Saniculoideae. In their study of fruit structure, Liu et al. (2003) refined the list of carpological features that characterise Saniculoideae, which now includes (with some exceptions) the presence of non-lignified endocarps, mericarp outgrowths, and prominent intrajugal secretory ducts (also called ‘companion canals’ by Tseng 1967), plus the lack of true vittae in the commissures. On the basis of these findings, several taxa can now be added to or removed from the subfamily, and in most cases these transfers are confirmed by molecular data. For instance, both matK data (Plunkett et al. 1996b) and ITS data (Valiejo-Roman et al. 2002) indicate that Lagoecia should be removed from Saniculoideae and placed in Apioideae, and this finding is supported by the presence of commissural vittae in Lagoecia, a feature lacking in all other saniculoids but common among the apioids (Liu et al. 2003). Fruit anatomical characters also suggest that Saniculoideae should be broadened to include three former apioids, Lichtensteinia, Polemanniopsis and Steganotaenia, plus the former ‘hydrocotyloid’ Arctopus, a finding that agrees with ITS sequence data for Polemanniopsis and Steganotaenia (see Downie and Katz-Downie 1999), as well as data sets based on ITS + matK (Plunkett and Lowry 2001) and 26S rDNA + rbcL + matK (Chandler and Plunkett 2004) for Arctopus. Much work remains to be completed in subfamily Apioideae, which includes about 400 genera and over 3 000 species (Pimenov and Leonov 1993). The study of Downie et al. (2001) remains the best overview of the status of intergeneric taxa among the apioids. In recognising only monophyletic groups, Downie et al. (2001) reported ten formal tribes and seven informally-named groups currently accepted on the basis of recent molecular studies. However, these
Downie et al. (2000) indicated that Cerceau-Larrival (1962) invalidly and illegitimately published the name Azorelloideae. However, her use of this name was not accompanied by the designation of type nor by a Latin diagnosis; thus it was not validly published and remains available.
South African Journal of Botany 2004, 70: 371–381
373
91
Apiales
97
90 100 100 98
100
65
100
100
100 100
100
100 100 100 100
100
99
100
100
99
100 100
100 100 100 100
100
100
100
100 100 100
100 100
99
100 100
98
100 100
98 100
91 99 99
86
100
100 100 100
82 86
86 98
100 100 100
86 100 99
86 96
100
98
100
100 100
100
100
100
100 100
96 100
100
70
100
100 100
100
Aegopodium Lagoecia Pimpinella Apium Petroselinum Arracacia Donnellsmithia Angelica Apioideae Endressia Coriandrum Neogoezia Aciphylla Daucus Bupleurum Anginon Heteromorpha Eryngium Sanicula Saniculoideae Astrantia Arctopus Azorella sel. Huanaca Azorella trif. Mulinum chil. Mulinumsp. Diplaspis Azorelloideae Spananthe Bolax Dichosciadium Bowlesia Eremocharis Centella Micropleura Mackinlaya conf. Mackinlaya macr. Mackinlayoideae Apiopetalum Actinotus Platysace Hymenosporum Pittosporum Rhytidosporum Marianthus Sollya Delarbrea Pseudosciadium Myodocarpus Dendropanax Fatsia Trevesia Hedera Kalopanax Eleutherococcus Polyscias Astrotricha Osmoxylon Tetrapanax Tupidanthus Gastonia Reynoldsia Munroidendron Tetraplasandra Arthrophyllum Meryta Pseudopanax Aralia Panax Schefflera Cheirodendron Cussonia Hydrocotyle mod. Hydrocotyle vert. Hydrocotyle bowl. Trachymene Griselinia Melanophylla Torricellia Aralidium Pennantia Helianthus Menyanthes Campanula Dipsacus Valeriana Alangium Cornus Nyssa Schizophragma
Apiaceae
Pittosporaceae Myodocarpaceae
Araliaceae
Griseliniaceae Torricelliaceae Pennantiaceae
Outgroups
Figure 1: Phylogenetic tree showing relationships among 87 representative taxa of Apiales and outgroups based on the analysis of 5 958 characters derived from a combined data set of nuclear 26S rDNA and plastid rbcL + matK sequences using Bayesian inference (with a GTR + I model of nucleotide substitutions). Posterior probability scores are indicated above each node. Redrawn from Chandler and Plunkett 2004
374
17 clades represent only 171 genera (less than half the total number) and just a small fraction of the species currently included in Apioideae. Moreover, many of these genera show evidence of non-monophyly. In short, despite the many advances made in recent years, Apioideae remain both the largest and most vexing group in Apiales, and many years of additional study will be needed to resolve fully relationships within and among the apioid genera. Fortunately, the progress made over the past decade should provide a framework from which more detailed analyses can be undertaken. Araliaceae Overall relationships in Araliaceae are reviewed by Lowry et al. (2004), on the basis of two recent studies (Wen et al. 2001, Plunkett et al. 2004), necessitating only a brief overview here. First, all molecular studies confirm the removal of five genera from Araliaceae: Mackinlaya and Apiopetalum (now placed in Apiaceae subfamily Mackinlayoideae, discussed above), plus Delarbrea, Myodocarpus and Pseudosciadium (placed in the newly recognised family Myodocarpaceae, discussed below). After the additional transfer of Stilbocarpa to Apiaceae (see Mitchell et al. 1999), the remaining araliad genera form a monophyletic group, often referred to as ‘core Araliaceae’. This clade is the basis for our recircumscription of Araliaceae in the revised classification provided in Table 1. Among the major findings within this somewhat narrowed Araliaceae are the non-monophyly of the two largest genera: Schefflera, which is polyphyletic, comprising at least five unrelated clades (confirmed by a much broader study of Schefflera, Plunkett, Lowry, Frodin and Wen, in prep.), and Polyscias, which is rendered paraphyletic by the inclusion of Tetraplasandra, Munroidendron, Reynoldsia, Gastonia, Arthrophyllum and Cuphocarpus in the broad ‘Polyscias sensu lato’ clade (see also Plunkett et al. 2001). Taken as a whole, Araliaceae comprise three major clades plus as many as nine much smaller clades, whose interrelationships are poorly resolved (see figures in Lowry et al. 2004). The three large clades are (1) the Asian palmate group, centered in eastern and southeastern Asia but also including several Neotropical lineages, (including genera such as Eleutherococcus, Kalopanax, Brassaiopsis, Macropanax, Dendropanax, Fatsia, Hedera and Oplopanax, plus two clades of Schefflera (Asian and Neotropical)); (2) the Polyscias–Pseudopanax group, including both Polyscias sensu lato and a subclade uniting Pseudopanax (sensu Mitchell and Wagstaff 1997), Meryta, and most Pacific species of Schefflera; and (3) the Aralia–Panax group, including both of these genera and their former segregates (e.g. Pentapanax, Sciadodendron). Among the poorly resolved clades forming the basal polytomy are two additional clades of Schefflera (the ‘African–Malagasy’ and ‘section Schefflera’ clades), plus seven other clades: Cheirodendron + Raukaua, Cussonia + Seemannaralia, Harmsiopanax, Cephalaralia, Motherwellia, Osmoxylon and Astrotricha. As discussed above, Hydrocotyle and Trachymene form a clade sister to the large clade of araliads, and are herein transferred to Araliaceae (Figure 1, see also Lowry et al. 2004).
Plunkett, Chandler, Lowry, Pinney and Sprenkle
Myodocarpaceae Myodocarpaceae comprise three genera segregated from Araliaceae: Delarbrea, Myodocarpus and Pseudosciadium. The formal recognition of this new family by Doweld (2001) is based largely on our molecular data (Plunkett and Lowry 2001; see also Plunkett et al. 1996a, 1997, Plunkett 2001, Lowry et al. 2001, Chandler and Plunkett 2004), but is also supported by a unique morphological character found in no other member of Apiales (viz. specialised oil vesicles in the endocarp; see Lowry 1986a, 1986b). The recognition of very small families is often considered undesirable (e.g. Cronquist 1981: x–xi), and we agree that Myodocarpaceae (with only three genera and 17 species) should be accepted only if our dual criteria of monophyly and morphological coherence compel their recognition. Our data strongly indicate that Delarbrea, Myodocarpus and Pseudosciadium fall outside the main ‘core Araliaceae’ clade. The only approach that could unite these three genera to the other araliads in a monophyletic family would also necessitate the union of Araliaceae with both Apiaceae and Pittosporaceae. Given the distinctive morphologies of these families, and especially of Pittosporaceae, the union of these clades would yield a highly heterogeneous family. Similarly problematic would be the union of Myodocarpaceae with Pittosporaceae as a single family (see Figure 1), a solution that satisfies the criterion of monophyly but not of morphological coherence. Thus, we contend that Delarbrea, Myodocarpus and Pseudosciadium are best treated as a distinct family. Within Myodocarpaceae, phylogenetic relationships among the genera were first discussed by Lowry (1986a, 1986b), and later tested by Plunkett and Lowry (2001). Additional data are now available from the study of Sprenkle, Plunkett and Lowry (in prep., see also Sprenkle 2002), based on both plastid and nuclear sequence data from samples of all species and subspecies in each genus (Figure 2). Molecular data suggest that Myodocarpaceae comprise two well supported clades, Myodocarpus and Delarbrea + Pseudosciadium. These clades are also well differentiated on the basis of morphology, and in particular fruit types (the fleshy or spongy drupes of Delarbrea and Pseudosciadium, vs the dry, schizocarpic fruits of Myodocarpus). Pseudosciadium, however, is nested within the Delarbrea clade, necessitating the taxonomic transfer of its single species into Delarbrea (Lowry, Plunkett and Sprenkle in prep.). In Myodocarpus, the species with simple leaves and those having pinnately-compound leaves are found in distinct subclades. Plunkett et al. (1996a) postulated that simple leaves were ancestral throughout Apiales (see also Lowry et al. 2001, Plunkett 2001, Chandler and Plunkett 2004), but given the tree topology in Figure 2, and the fact that the species of Delarbrea (including Pseudosciadium) are entirely pinnate, the ancestral state cannot be unequivocally reconstructed in Myodocarpaceae. If we interpret pinnate leaves as ancestral in Myodocarpaceae, there must have been a single gain of this character in the common ancestor (from the simple leaves that characterised the rest of Apiales), followed by a reversal in the simple-leaved subclade of Myodocarpus. Alternatively, pinnate leaves may have evolved twice inde-
South African Journal of Botany 2004, 70: 371–381
pendently, once in the Delarbrea + Pseudosciadium clade, and a second time in the pinnately-compound subclade of Myodocarpus. Because both scenarios require a minimum of two steps, they are equally parsimonious. Studies of leaf development may offer some clues as to the evolution of this character. Biogeographically, Myodocarpaceae are clearly centred on New Caledonia, where all but two of the taxa are endemic (Lowry 1986a, 1986b). Delarbrea paradoxa subsp. paradoxa is more widely distributed across the south-western Pacific, and D. michieana is endemic to Queensland, Australia. The placement of D. paradoxa subsp. paradoxa in the cladogram (nested within an entirely New Caledonian clade) suggests a case of dispersal, but the position of D. michieana as the first-diverging lineage in the Delarbrea + Pseudosciadium clade could indicate a more ancient vicariance event associated with the separation of New Caledonia from Australia (c. 55MY; see Lowry 1998 and references therein). Pittosporaceae Evidence of a close relationship of Pittosporaceae to Apiales has been available for over a century, and includes data from stem structure (particularly the presence of schizogenous secretory canals, Van Tieghem 1884), ovular structure and development (Jurica 1922), cytology (Jay 1969), and phytochemistry (Hegnauer 1971, 1982, Dahlgren 1980, Jensen 1992). Choosing to dismiss these characters, some authors stressed the difference in ovary position (and to a lesser degree leaf shape) in separating Pittosporaceae from Apiales (e.g. Cronquist 1981, Eyde and Tseng 1971). However, the studies of Erbar and Leins (1995, 1996) have demonstrated developmental homology among the ovary positions of Pittosporaceae, Araliaceae and Apiaceae (see also Leins and Erbar 2004). These data, together with a growing body of molecular evidence (e.g. Plunkett et al. 1996a, Plunkett 2001, Chandler and Plunkett 2004, Kårehed 2003), unequivocally place Pittosporaceae within Apiales. However, the relationship of this family to the other clades in the order remains largely unresolved. The study of Chandler and Plunkett (2004) suggests a relationship of Pittosporaceae to Myodocarpaceae, which in turn may be sister to Apiaceae (Figure 1). Within Pittosporaceae, Cayzer et al. (1999a, 1999b, 2000a, 2000b, 2004, see also Cayzer 1998, Cayzer and Crisp 2004) have undertaken phylogenetic analyses based on morphological characters, resulting in several taxonomic changes. For example, all species of Sollya are now treated under Billardiera (Cayzer et al. 2004) and the species of Citriobatus are included within Pittosporum (Cayzer et al. 2000a). In addition, the new genus Auranticarpa has been erected for a clade of species formerly included within Pittosporum (Cayzer et al. 2000b). Altogether, Cayzer and her co-workers currently recognise nine genera in the family: Pittosporum, Billardiera, Bursaria, Marianthus, Auranticarpa, Rhytidosporum, Cheiranthera, Bentleya and Hymenosporum. Molecular data provide independent confirmation of Cayzer’s findings. A preliminary cladogram based on ITS +
375
trnL-trnF data (Chandler, Pinney, Cayzer and Plunkett, in prep.; see also Pinney 2002) is presented in Figure 3. This tree confirms the inclusion of Sollya within Billardiera. Moreover, Pittosporum is monophyletic only if the species previously placed in Citriobatus are included and those of Auranticarpa are excluded. In fact, the species of Auranticarpa are not closely related to Pittosporum, but rather form a clade sister to Bursaria + Rhytidosporum. Molecular data indicate that Marianthus may be paraphyletic (Figure 3), and the nature of this paraphyly may necessitate the union of all species assigned to Marianthus, Bentleya, and Billardiera (+ Sollya) into a single genus. Pittosporum is the largest of the nine genera (~140 species), accounting for nearly half of the species in the family. It is also the only genus with members found outside Australia (except the monotypic Hymenosporum, which occurs in both Australia and nearby New Guinea). To explore biogeographic relationships in the family, we examined trees resulting from two analyses, one based on ITS + trnL-trnF data (Figure 3), and another based on an expanded ITS data set that included sequences from the published study of Gemmill et al. (2002) (tree not shown). In agreement with Crisp et al. (1989), our preliminary results suggest that the family was probably not widely distributed across eastern Gondwanaland before its breakup, but rather was limited to Australia. Of the major clades in Pittosporaceae, only the Pittosporum clade has non-Australian species. Moreover, taxa from both the Pacific and Indian Ocean basins are nested within clades apparently originating in Australia, suggesting more recent dispersal events to relatively young island groups. Due to limited sample size (especially from the Indian Ocean basin) and the lack of strong support for many clades, these results are tentative and speculative, but they do suggest that Pittosporaceae may be a rich source for detailed biogeographic studies. Griselinia, Aralidium, Torricellia, Melanophylla and Pennantia Evidence for placing Araliaceae, Apiaceae, Pittosporaceae and Myodocarpaceae in a single order is now overwhelming, and includes information from morphological, anatomical, developmental, cytological, phytochemical, and molecular data (discussed above). Although relationships among these four families are not yet fully resolved, all phylogenetic studies unite these lineages into a single clade. Molecular data provide further evidence for expanding Apiales to include five additional genera: Griselinia, Aralidium, Torricellia, Melanophylla and Pennantia. The morphological support for their inclusion, however, is less straightforward (reviewed in Plunkett 2001, Chandler and Plunkett 2004, Kårehed 2003). For example, none of these taxa possesses the schizogenous secretory canals that so strikingly characterise the wood (and often other plant parts) of the four core families in Apiales. The basic chromosome number inferred for Apiales (x = 6 or 12; see Yi et al. in press) is shared by Torricellia and possibly Pennantia (n = 25, which may represent an aneuploid increase from n = 24, and thus x = 12), but not with Griselinia (x = 9) or Aralidium (x = 10) (this character remains unknown for Melanophylla) (reviewed in Plunkett
376
Plunkett, Chandler, Lowry, Pinney and Sprenkle
Myodocarpaceae 61
Myodocarpus nervatus Myodocarpus touretteorum
99
Myodocarpus simplicifolius
Myodocarpus crassifolius
50
Myodocarpus gracilis Myodocarpus involucratus 92
Myodocarpus lanceolatus
simple leaves
Myodocarpus simplicifolius
Myodocarpus simplicifolius
Myodocarpus
Myodocarpus vieillardii Myodocarpus vieillardii 70 97
Myodocarpus fraxinifolius Myodocarpus fraxinifolius
71
Myodocarpus fraxinifolius 99
98
Myodocarpus pinnatus Myodocarpus pinnatus Myodocarpus fraxinifolius
100
99 98
compound leaves
100
Delarbrea collina Delarbrea collina Pseudosciadium balansae
100
85
Delarbrea paradoxa ssp. depauperata Delarbrea paradoxa ssp. depauperata
82
100
Delarbrea paradoxa ssp. paradoxa 100
100
Delarbrea + Pseudosciadium
Delarbrea paradoxa ssp. paradoxa Delarbrea harmsii Delarbrea michieana
100
Astrotricha sp. Osmoxylon geelvinkianum Apiopetalum velutinum
Figure 2: Strict consensus of 17 most parsimonious trees based on a combined data set of nuclear rDNA spacers (ITS and partial ETS) and plastid trnL-trnF sequences from 29 taxa of Myodocarpaceae and outgroups. The total data set included 1 804 characters, of which 336 were potentially informative. The shortest trees were 569 steps long, with a consistency index of 0.815 and a retention index of 0.930. Bootstrap percentages are indicated at each node. Redrawn from Sprenkle, Plunkett and Lowry, in prep.
2001, Kårehed 2003). Similarly, both phytochemistry and wood anatomy provide conflicting evidence for the inclusion of these five genera in Apiales (reviewed in Plunkett 2001, Kårehed 2003). Plunkett et al. (1996a) originally hypothesised that bicarpelly was ancestral in Apiales, a result confirmed for the core families (e.g. Lowry et al. 2001, Plunkett 2001, Plunkett and Lowry 2001, Chandler and Plunkett 2004, Kårehed 2003), but not for Griselinia, Aralidium, Torricellia, Melanophylla and Pennantia. In these genera, the number of locules is variable, but the ovaries are clearly derived from three carpels, and in all cases, only a single functional ovule is produced (see Watson and Dallwitz 1992, Dillon and Muñoz-Schick 1993, Plunkett 2001, Kårehed
2003). In Torricellia and Melanophylla, all three carpels persist, yielding three locules (reports of ‘(2–)3’ or ‘3(–4)’ appear to be due to irregularly-shaped locules that may be either overlooked or counted twice, respectively (GM Plunkett, pers. obs.)). In Griselinia, only a single locule is discernable, and in both Aralidium and Pennantia, two abortive locules are located adjacent to the single, fully developed one. Thus, it appears that tricarpelly was ancestral in Apiales, followed by a single reduction to two carpels only after the divergence of these five genera (see also Chandler and Plunkett 2004, Kårehed 2003).
South African Journal of Botany 2004, 70: 371–381
377
Pittosporaceae 69
Billardiera cymosa Billardiera lehmanniana
62 99
Billardiera (= Sollya) heterophylla Billardiera (= Sollya) fusiformis
Billardiera (incl. Sollya)
Billardiera longifolia Billardiera fraseria 60
Marianthus ringens Marianthus bicolor
93
Bentlya spinescens Cheiranthera linearis Hymenosporum flavum 80
Bursaria incana Bursaria spinosa Rhytidosporum alpinum
72 100 56
Bentleya Cheiranthera Hymenosporum Bursaria Rhytidosporum
Auranticarpa papyracea Auranticarpa edentata
100
Marianthus
Auranticarpa
Auranticarpa ilicifolia 60 93
Pittosporum oreillyanum Pittosporum (=Citriobatus) lancifolium Pittosporum (=Citriobatus) spinescens Pittosporum (=Citriobatus) multiflorum
98
Pittosporum trilobium Pittosporum rubignosum
85
Pittosporum suberosum Pittosporum moluccanum
83 80
Pittosporum (incl. Citriobatus)
Pittosporum venulosum Pittosporum ferrigineum
100
Pittosporum angustifolium Pittosporum ligustrifolium Pittosporum wingii Pittosporum eugenioides Pittosporum crassifolium
100
Myodocarpus fraxinifolius Delarbrea collina
Outgroups
Schefflera baillonii
Figure 3: Strict consensus of 116 most parsimonious trees based on a combined data set of ITS nuclear rDNA and plastid trnL-trnF sequences from 35 taxa of Pittosporaceae and outgroups. The total data set included 1 705 characters, of which 299 were potentially informative. The shortest trees were 737 steps long, with a consistency index of 0.581 and a retention index of 0.719. Bootstrap percentages are indicated above each node. Redrawn from Chandler, Pinney, Cayzer and Plunkett, in prep.
378
Revised Classification of the Order Apiales The manner of translating phylogenetic studies into formal systems of classification is currently a subject of great debate (e.g. De Queiroz and Gauthier 1994, Nixon and Carpenter 2000, Langer 2001, Sennblad and Bremer 2002), but systematists appear to be united in espousing the principle of stability. In Apiales, many of the traditional taxonomic groups are now known to be non-monophyletic and are widely acknowledged to be in need of substantial realignment. However, given the size of the order (representing over 5 000 species), most students of Apiales have been reluctant to make formal taxonomic changes based on the sampling of relatively few taxa, especially when the composition of some clades appeared to be unstable from study to study. Nonetheless, there is an obvious need to name clades for the purpose of discussion and as a framework for further investigations. In the past, we have advocated the use of informal clade names (e.g. ‘core Apiaceae’, the ‘Asian palmate clade’, the ‘apioid superclade’, the ‘Mackinlaya group’, etc.) until a unified picture of phylogenetic relationships has emerged. To a large degree, such stability in our understanding of Apiales has now been achieved at the interfamilial level. Moreover, the multiplication of informal names (often with similar wording, but defining slightly different and overlapping groups) has reached the point where it is hindering rather than helping communication. For example, there are at least three informal definitions of Apiaceae in current usage (viz. the ‘traditional Apiaceae’ (sensu Drude), ‘Apiaceae sensu lato’ (including Araliaceae), and ‘core Apiaceae’ (excluding many ‘hydrocotyloids’)), as well as a number of competing notions of both Araliaceae and the entire order. Several authors have provided formal names or recircumscriptions affecting Apiales either at higher levels of classification (e.g. the ordinal system of the APG 1998, 2003) or at the intergeneric level (e.g. Downie et al. 2000 for core Apiaceae), but the time is now ripe to offer an interfamilial classification of Apiales as a whole. In the system that follows, we recognise groups on the basis of two criteria, monophyly (usually determined by phylogenetic studies based on molecular data) and morphological coherence. Using these two criteria, we translate the current concept of ‘core Araliaceae’ into a formal recircumscription of that family, which now excludes six former araliad genera (Myodocarpus, Delarbrea, Pseudosciadium, Apiopetalum, Mackinlaya and Stilbocarpa) and includes two former ‘hydrocotyloid’ genera (Hydrocotyle and Trachymene, leaving open the potential to add additional hydrocotyloids as more taxa are sampled). In a similar fashion, we treat the three major clades of ‘core Apiaceae’ (the apioids, the saniculoids and the ‘Azorella clade’) plus one additional clade (the ‘Mackinlaya group’) as four distinct subfamilies in a recircumscribed Apiaceae. Some authors have chosen to unite Apiaceae and Araliaceae into a single family (e.g. Judd et al. 2002), but our data indicate that this cannot be done without also including Pittosporaceae (which is placed as sister either to Apiaceae or Araliaceae; see Plunkett and Lowry 2001, Chandler and Plunkett 2004). We believe the union of these taxa into a single, broadly defined family yields a group with an inordinate amount of morpho-
Plunkett, Chandler, Lowry, Pinney and Sprenkle
logical heterogeneity, making it difficult to characterise or recognise. Thus, both Pittosporaceae and the small (but well supported and morphologically coherent) Myodocarpaceae are recognised as distinct families. Perhaps the most difficult issue to resolve in the classification of Apiales is also the most fundamental: how broadly to define the order. The four ‘core’ families described above (Apiaceae, Araliaceae, Pittosporaceae and Myodocarpaceae) represent a monophyletic group that shares many features (e.g. anatomy, cytology, phytochemistry). Molecular data agree in placing Griselinia, Aralidium, Torricellia, Melanophylla and Pennantia as successively sister to the four-family clade, but these five genera lack many of the distinctive features uniting the four core families. Each of these five genera has been recognised as its own order, but we believe that monogeneric (and in the case of Aralidium, monotypic) orders are redundant and offer no additional taxonomic information. Thus, we concur with earlier proposals (e.g. APG 1998, 2003, Plunkett 2001, Chandler and Plunkett 2004, Kårehed 2003) that include Griselinia, Aralidium, Torricellia, Melanophylla and Pennantia within Apiales, placed in three families: Pennantiaceae (monogeneric), Griseliniaceae (monogeneric), and Torricelliaceae (broadened to include Torricellia, Aralidium and Melanophylla). However, because the ‘four-family clade’ (comprising Apiaceae, Araliaceae, Pittosporaceae and Myodocarpaceae) is well supported, morphologically coherent, and the focus of most of the research being conducted within the order, we contend that it should be recognised formally. The application of an informal name (such as ‘core Apiales’) might appeal to some, but recent experience (at least in Apiales) suggests that the definitions and circumscriptions of such names are often unstable and can lead to considerable confusion. We therefore provide a formal name (with an explicit circumscription) in recognising a new suborder, Apiineae. Despite suggestions to the contrary, the principle of exhaustive subsidiary taxa is not a requirement of the ICBN (Greuter et al. 2000), and thus subordinal names are neither required nor provided for the other three families in the order. A synopsis of the revised classification is presented in Table 1. Acknowledgements — The authors gratefully acknowledge Prof. Ben-Erik van Wyk and Dr Patricia Tilney for organising the 4th International Apiales Symposium (Pretoria) and this proceedings volume, and for kindly inviting our contributions to both; thanks are also extended to Dr Gordon McPherson for proofreading the Latin diagnoses of the new taxa. Support was provided by grants from the National Science Foundation (DEB-9981641) and the National Geographic Society (5793-96) to GMP and PPL.
References APG (Angiosperm Phylogeny Group) (1998) An ordinal classification for families of flowering plants. Annals of the Missouri Botanical Garden 85: 531–553 APG (Angiosperm Phylogeny Group) (2003) An update of the Angiosperm Phylogeny Group classification for the orders and families of flowering plants: APG II. Botanical Journal of the Linnean Society 141: 399–436 Bentham G, Hooker JD (1867) Genera Plantarum. A Black, W Pamplin, Lovell Reeve and Co., London
South African Journal of Botany 2004, 70: 371–381
379
Table 1: Revised classification of the angiosperm order Apiales Order : APIALES Nakai 1930 Family: TORRICELLIACEAE (Wang.) Hu 1934 (only Torricellia, Melanophylla, Aralidium) Family: GRISELINIACEAE J.R. Forst. & G. Forst. 1839 (only Griselinia) Family: PENNANTIACEAE J. Agardh 1858 (only Pennantia) Suborder : APIINEAE, Plunkett & Lowry, subord. nov. Lignum canalibus secretoriis schizogenis; gynoecium oriunum a duo carpellis. T: Apium L. Family: APIACEAE Lindl. 1836 (nom. alt., Umbelliferae Durande 1782) Subfamily: APIOIDEAE Seem. 1866 Subfamily: SANICULOIDEAE Burnett 1835 Subfamily: AZORELLOIDEAE Plunkett & Lowry, subfam. nov. Fructus teres vel a dorso compressus, sine vittis; canales intra juga praesentes; endocarpium induratum; stratum interius mesocarpii lignescens vel parenchymatum, crystallis rhombeis (sed crystallis compositis in strato exteriore Azorellae); carpella secedentes in fructu maturo.T: Azorella Lam. (also incl. Bolax, Bowlesia, Dichosciadium, Dickinsia, Diplapsis, Eremocharis, Gymnophyton, Huanaca, Mulinum, Schizeilema, Spananthe, Stilbocarpa). Syn.: Hydrocotyloideae Link 1829, pro parte, excl. Hydrocotyle Subfamily: MACKINLAYOIDEAE Plunkett & Lowry, subfam. nov. Fructus a latere compressus (vel teres in Apiopetalo), sine vittis; canales intra juga minuti vel obsoleti; endocarpium sclerescens; stratum interius mesocarpii lignescens vel parenchymatum, crystallis rhombeis; carpella non secedentes in fructu maturo. T: Mackinlaya F. Muell. (also incl. Apiopetalum, Centella, Micropleura, Actinotus, Platysace, Xanthosia). Syn.: Mackinlayeae Hook.f. 1873.; Mackinlayaceae Doweld 2001 Family: ARALIACEAE Durande 1782 (incl. Hydrocotyle, Trachymene; excl. Apiopetalum, Delarbrea, Mackinlaya, Myodocarpus, Pseudosciadium, Stilbocarpa) Family: MYODOCARPACEAE Doweld 2001 Syn: Myodocarpeae R. Vig. 1906 Family: PITTOSPORACEAE R. Br. 1814
Bessey CE (1915) Phylogenetic taxonomy of flowering plants. Annals of the Missouri Botanical Garden 2: 109–164 Cayzer LW (1998) Revision of the Family Pittosporaceae in Australia. PhD Thesis, Australian National University, Canberra, Australia Cayzer LW, Crisp MD (2004) Reinstatement and revision of the genus Marianthus (Pittosporaceae). Australian Systematic Botany 17: 127–144 Cayzer LW, Crisp MD, Telford IRH (1999a) Bursaria (Pittosporaceae): a morphometric analysis and revision. Australian Systematic Botany 12: 117–143 Cayzer LW, Crisp MD, Telford IRH (1999b) Revision of Rhytidosporum (Pittosporaceae). Australian Systematic Botany 12: 689–708 Cayzer LW, Crisp MD, Telford IRH (2000a) Revision of Pittosporum (Pittosporaceae) in Australia. Australian Systematic Botany 13: 845–902 Cayzer LW, Crisp MD, Telford IRH (2000b) Auranticarpa, a new genus of Pittosporaceae from northern Australia. Australian Systematic Botany 13: 903–917 Cayzer LW, Crisp MD, Telford IRH (2004) Cladistic analysis and revision of Billardiera (Pittosporaceae). Australian Systematic Botany 17: 83–125 Cerceau-Larrival MT (1962) Plantules et pollens d’Ombellifères: leur intérêt systématique et phylogénetique. Mémoires du Muséum National d’Histoire Naturelle, Série B, Botanique 14: 1–164 Chandler GT, Plunkett GM (2004) Evolution in Apiales: nuclear and chloroplast markers together in (almost) perfect harmony. Botanical Journal of the Linnean Society 144: 123–147 Constance L (1971) History of the classification of Umbelliferae (Apiaceae). In: Heywood VH (ed) The biology and chemistry of the Umbelliferae. Botanical Journal of the Linnean Society (Supplement 1) 64: 1–8 Crisp MD, Taylor JM, Bennett EM (1989) Pittosporaceae — an old east Gondwanan family? Australian Systematic Botany Society
Newsletter 60: 20 Cronquist A (1981) An Integrated System of Classification of Flowering Plants. Columbia University Press, New York Cronquist A (1988) The Evolution and Classification of Flowering Plants (2nd edn). New York Botanical Garden, Bronx, New York Dahlgren RMT (1980) A revised system of classification of the angiosperms. Botanical Journal of the Linnean Society 80: 91–124 De Queiroz K, Gauthier J (1994) Toward a phylogenetic system of biological nomenclature. Trends in Ecology and Evolution 9: 27–31 Dillon MO, Muñoz-Schick M (1993) A revision of the dioecious genus Griselinia (Griseliniaceae), including a new species from the coastal Atacama Desert of northern Chile. Brittonia 45: 261–274 Doweld A (2001) Prosyllabus Tracheophytorum, Tentamen Systematis Plantarum Vascularium (Tracheophyta) (An attempted system of the vascular plants). Pensoft, Moscow Downie SR, Katz-Downie DS (1996) A molecular phylogeny of Apiaceae subfamily Apioideae: evidence from nuclear ribosomal DNA internal transcribed spacer sequences. American Journal of Botany 83: 234–251 Downie SR, Katz-Downie DS (1999) Phylogenetic analysis of chloroplast rps16 intron sequences reveals relationships within the woody southern African Apiaceae subfamily Apioideae. Canadian Journal of Botany 77: 1120–1135 Downie SR, Katz-Downie DS, Cho K-J (1996) Phylogenetic analysis of Apiaceae subfamily Apioideae using nucleotide sequences from the chloroplast rpoC1 intron. Molecular Phylogenetics and Evolution 6: 1–18 Downie SR, Katz-Downie DS, Watson MF (2000) A phylogeny of the flowering plant family Apiaceae based on chloroplast DNA rpl16 and rpoC1 intron sequences: towards a suprageneric classification of subfamily Apioideae. American Journal of Botany 87: 273–292
380
Downie SR, Plunkett GM, Watson MF, Spalik K, Katz-Downie DS, Valiejo-Roman CM, Terentieva EI, Troitsky AV, Lee B-Y, Lahham J, El-Oqlah A (2001) Tribes and clades within Apiaceae subfamily Apioideae: the contribution of molecular data. Edinburgh Journal of Botany 58: 301–330 Drude CGO (1898) Umbelliferae. In: Engler A, Prantl K (eds) Die Natürlichen Pflanzenfamilien, Vol 3. Wilhelm Engelmann, Leipzig, pp 63–250 Engler A, Prantl K (1898) Die Natürlichen Planzenfamilien. Engelmann, Leipzig Erbar C, Leins P (1995) An analysis of the early floral development of Pittosporum tobira (Thunb.) Aiton and some remarks on the systematic position of the family Pittosporaceae. Feddes Repertorium 106: 463–473 Erbar C, Leins P (1996) Distribution of the character states “early sympetaly” and “late sympetaly” within the “Sympetalae Tetracyclicae” and presumably allied groups. Botanica Acta 109: 427–440 Eyde RH, Tseng CC (1971) What is the primitive floral structure of Araliaceae? Journal of the Arnold Arboretum 52: 205–239 Gemmill CEC, Allan GJ, Wagner WL, Zimmer EA (2002) Evolution of insular Pacific Pittosporum (Pittosporaceae): origin of the hawaiian radiation. Molecular Phylogenetics and Evolution 22: 31–42 Greuter W, McNeill J, Barrie FR, Burdet H-M, Demoulin V, Filgueiras TS, Nicolson DH, Silva PC, Skog JE, Trehane P, Turland NJ, Hawksworth DL (2000) International Code of Botanical Nomenclature (St Louis Code). Regnum Vegetabile 138. Koeltz Scientific Books, Königstein Hegnauer R (1971) Chemical patterns and relationships of Umbelliferae. In: Heywood VH (ed) The Biology and Chemistry of the Umbelliferae. Botanical Journal of the Linnean Society (Supplement 1) 64: 267–277 Hegnauer R (1982) Phytochemie und Klassifikation der Umbelliferen, eine Neubewertung im Lichte der Siet 1972 bekannt gewordenen phytochemischen Tatsachen. In: Cauwet-Marc AM, Carbonnier J (eds) Actes du 2e Symposium International sur les Ombellifères, Contributions Pluridisciplinaires à la Systématique. Missouri Botanical Garden, St Louis, pp 335–364 Henwood MJ, Hart JM (2001) Towards an understanding of the phylogenetic relationships of Australian Hydrocotyloideae (Apiaceae). Edinburgh Journal of Botany 58: 269–289 Heywood VH (1978) Introduction to the taxonomy of the Umbelliferae. In: Cauwet-Marc A-M, Carbonnier J (eds) Actes du 2e Symposium International sur les Ombellifères, Contributions Pluridisciplinaires à la Systématique. Missouri Botanical Garden, St Louis, pp 107–112 Hutchinson J (1967) The Genera of Flowering Plants, Vol. II. Oxford University Press, London Hutchinson J (1973) The Families of Flowering Plants. Oxford University Press, London Jay M (1969) Chemotaxonomic researches on vascular plants. XIX. Flavonoid distribution in the Pittosporaceae. Botanical Journal of the Linnean Society 62: 423–429 Jensen SR (1992) Systematic implications of the distribution of iridoids and other chemical compounds in the Loganiaceae and other families of the Asteridae. Annals of the Missouri Botanical Garden 79: 284–302 Judd WS, Campbell CS, Kellogg EA, Stevens PF, Donoghue MJ (2002) Plant systematics: A Phylogenetic Approach (2nd edn). Sinauer Associates, Sunderland, Massachusetts Jurica HS (1922) A morphological study of the Umbelliferae. Botanical Gazette 74: 292–307 Kårehed J (2003) The family Pennantiaceae and its relationships to Apiales. Botanical Journal of the Linnean Society 141: 1–24 Langer MC (2001) Linnaeus and the PhyloCode: where are the differences? Taxon 50: 1091–1096
Plunkett, Chandler, Lowry, Pinney and Sprenkle
Lee B-Y, Levin GA, Downie SR (2001) Relationships within the spiny-fruited umbellifers (Scandiceae subtribes Daucinae and Torilidinae) as assessed by phylogenetic analysis of morphological characters. Systematic Botany 26: 622–642 Leins P, Erbar C (2004) Organ sequences in Apiales (Apiaceae, Araliaceae, Pittosporaceae). South African Journal of Botany 70: 468–474 Liu M, Van Wyk B-E, Tilney PM (2003) The taxonomic value of fruit structure in the subfamily Saniculoideae and related African genera (Apiaceae). Taxon 52: 261–270 Lowry PP II (1986a) A Systematic Study of Three Genera of Araliaceae Endemic to or Centered on New Caledonia: Delarbrea, Myodocarpus, and Pseudosciadium. PhD Thesis, Washington University, St Louis, Missouri, USA Lowry PP II (1986b) A systematic study of Delarbrea Vieill. (Araliaceae). Allertonia 4: 169–201 Lowry PP II (1998) Diversity, endemism, and extinction in the flora of New Caledonia: a review. In: Peng C-I, Lowry PP II (eds) Proceedings of the International Symposium on Rare, Threatened, and Endangered Floras of Asia and the Pacific. Academia Sinica, Taipei, Taiwan, pp 181–206 Lowry PP II, Plunkett GM, Oskolski AA (2001) Early lineages in Apiales: insights from morphology, wood anatomy and molecular data. Edinburgh Journal of Botany 58: 207–220 Lowry PP II, Plunkett GM, Wen J (2004) Generic relationships in Araliaceae: looking into the crystal ball. South African Journal of Botany 70: 382–392 Mitchell AD, Wagstaff SJ (1997) Phylogenetic relationships of Pseudopanax species inferred from parsimony analysis of rDNA sequence data and morphology. Plant Systematics and Evolution 208: 121–138 Mitchell AD, Meurk CD, Wagstaff SJ (1999) Evolution of Stilbocarpa, a megaherb from New Zealand’s sub-antarctic islands. New Zealand Journal of Botany 37: 205–211 Nixon KC, Carpenter JM (2000) On the other “Phylogenetic Systematics”. Cladistics 16: 298–318 Pimenov MG, Leonov MV (1993) The Genera of the Umbelliferae: A Nomenclator. Royal Botanic Gardens, Kew Pinney SM (2002) A Molecular Systematic Study of Pittosporaceae. MSc Thesis, Virginia Commonwealth University, Richmond, Virginia, USA Plunkett GM (2001) Relationships of the order Apiales to subclass Asteridae: a reevaluation of morphological characters based on insights from molecular data. Edinburgh Journal of Botany 58: 183–200 Plunkett GM, Downie SR (1999) Major lineages within Apiaceae subfamily Apioideae: a comparison of chloroplast restriction site and DNA sequence data. American Journal of Botany 86: 1014–1026 Plunkett GM, Lowry PP II (2001) Relationships among “ancient araliads” and their significance for the systematics of Apiales. Molecular Phylogenetics and Evolution 19: 259–276 Plunkett GM, Lowry PP II, Burke MK (2001) The phylogenetic status of Polyscias (Araliaceae) based on nuclear ITS sequence data. Annals of the Missouri Botanical Garden 88: 213–230 Plunkett GM, Soltis DE, Soltis PS (1996a) Higher level relationships of Apiales (Apiaceae and Araliaceae) based on phylogenetic analysis of rbcL sequences. American Journal of Botany 83: 499–515 Plunkett GM, Soltis DE, Soltis PS (1996b) Evolutionary patterns in Apiaceae: inferences based on matK sequence data. Systematic Botany 21: 477–495 Plunkett GM, Soltis DE, Soltis PS (1997) Clarification of the relationship between Apiaceae and Araliaceae based on matK and rbcL sequence data. American Journal of Botany 84: 565–580 Plunkett GM, Wen J, Lowry PP II (2004) Infrafamilial classifications
South African Journal of Botany 2004, 70: 371–381
and characters in Araliaceae: insights from the phylogenetic analysis of nuclear (ITS) and plastid (trnL-trnF) sequence data. Plant Systematics and Evolution 245: 1–39 Sennblad B, Bremer B (2002) Classification of Apocynaceae s.l. according to a new approach combining Linnaean and phylogenetic taxonomy. Systematic Biology 51: 389–409 Spalik K, Wojewódzka A, Downie SR (2001a) The evolution of fruit in Scandiceae subtribe Scandicinae (Apiaceae). Canadian Journal of Botany 79: 1358–1374 Spalik K, Wojewódzka A, Downie SR (2001b) Delimitation of genera in Apiaceae with examples from Scandiceae subtribe Scandicinae. Edinburgh Journal of Botany 58: 331–346 Sprenkle TS (2002) Molecular Systematics of tribe Myodocarpeae (Araliaceae s.l.). MSc Thesis, Virginia Commonwealth University, Richmond, Virginia, USA Stevens PF (2003) Angiosperm Phylogeny Website, Version 4. University of Missouri — St Louis and Missouri Botanical Garden. Available at: http://www.mobot.org/MOBOT/ Research/APweb/ Takhtajan A (1987) System of Magnoliophyta. Academy of Sciences, Leningrad
Edited by B-E van Wyk and PM Tilney
381
Thorne RF (1992) Classification and geography of flowering plants. Botanical Review 58: 225–348 Tseng CC (1967) Anatomical studies of flower and fruit in the Hydrocotyloideae (Umbelliferae). University of California Publications in Botany 30: 1–79 Valiejo-Roman CM, Terentieva EI, Samigullin TH, Pimenov MC (2002) Relationships among genera in Saniculoideae and selected Apioideae (Umbelliferae) inferred from nrITS sequences. Taxon 51: 91–101 Van Tieghem P (1884) Sur la structure et les affinités des Pittosporées. Bulletin de la Société Botanique de France 31: 384–385 Watson L, Dallwitz MJ (1992 and onwards) The families of flowering plants: descriptions, illustrations, identification, and information retrieval. Available at: http://biodiversity.uno.edu/delta/angio/ Wen J, Plunkett GM, Mitchell AD, Wagstaff SJ (2001) The evolution of Araliaceae: a phylogenetic analysis based on ITS sequences of nuclear ribosomal DNA. Systematic Botany 26: 144–167 Yi T, Lowry PP II, Plunkett GM, Wen J (in press) Chromosomal Evolution in Araliaceae s. lat. Taxon