Polyphyly of the spring-parsleys ( Cymopterus ): molecular and morphological evidence suggests complex relationships among the perennial endemic genera of western North American Apiaceae

Deborah Katz-downie; Stephen R.  Downie

Determination of phylogenetic affinity among the members of Apiaceae especially the subfamily Apioideae is much debatable. The morphological evidences on inflorescence, fruit and seed characters showed little concomitance with those derived from molecular parameters as far as groupings in clades are concerned. In the present study involving few selected Apiaceae members, a Dendrogram was computed and a putative evolutionary trend was outlined based on leaf margin insertion pattern and cotyledonary leaf features, general plant morphology as well as leaf surface micromorphology by SEM. The members of Apiaceae with broad-ovate pinnatifid to pinnatipartite leaves and ovate –lanceolate cotyledonary leaves were included in a distinct cluster but the members with pinnatisect – decompounds leaves and linear cotyledonary leaves were included in a separate cluster. Long Coriander (Eryngium foetidum) included in the subfamily Saniculoideae may represent the most primitive condition having simple entire leaf with spiny margin. The member like Long coriander might have served as progenitor stock from which Coriander, Celery, Parsley (all with ovate pinnatifid to pinnatipartite leaves) and Cumin, Fennel, Dill (all having pinnatisect to decompounds leaves with linear segments) might have originated as evidenced by gradual increase in marginal insertion. Similarly, gradual narrowing of ovate-lanceolate cotyledonary leaves might have given rise to narrow linear cotyledonary leaves. Leaf-margin insertion pattern and cotyledonary leaf features appeared to be promising in tracing phylogeny in Apiaceae especially in Apioideae. Simple entire leaf represents the primitive condition and evolution in Apiaceae supposed to have progressed towards gradual increase in marginal insertion leading to pinnately dissected, pinnatipartite and ultimately to decompound leaves with narrow linear segments. Dendrogram computed from morphological features showed concomitance with that phylogenetic trend as far as grouping and interrelationships of members are concerned. In the present study leaf margin insertion pattern and seedling morphology appeared to be instrumental in tracing the phylogeny in Apiaceae specially subfamily Apioideae while molecular systematics are not conclusive.

Evolutionary relationships among 116 representatives (80 genera) ofApiaceae (Umbelliferae) subfam.Apioideae were investigated by comparative sequencing of the two internal transcribed spacers of the 18S–26S nuclear ribosomal DNA repeat. The resultant phylogenies, inferred using maximum parsimony and neighbor-joining methods, clarified the relationships of several genera whose phylogenetic placements have heretofore been problematic. Comparisons between the phylogenies inferred and the distribution of

1295 Can. J. Bot. Downloaded from www.nrcresearchpress.com by Canadian Science Publishing on 06/08/15 For personal use only. Polyphyly of the spring-parsleys (Cymopterus): molecular and morphological evidence suggests complex relationships among the perennial endemic genera of western North American Apiaceae Stephen R. Downie, Ronald L. Hartman, Feng-Jie Sun, and Deborah S. Katz-Downie Abstract: Cladistic analyses of DNA sequences from the nuclear rDNA internal transcribed spacer region and cpDNA rps16 intron and, for a subset of taxa, the cpDNA trnF-trnL-trnT locus were carried out to evaluate the monophyly of Cymopterus and to ascertain its phylogenetic placement among the other perennial genera of Apiaceae (Umbelliferae) subfamily Apioideae endemic to western North America. To elucidate patterns in the evolution of specific fruit characters and to evaluate their utility in circumscribing genera unambiguously, additional evidence was procured from cross-sections of mature fruits and the results of cladistic analysis of 25 morphological characters. Analyses of the partitioned data sets resulted in weakly supported and largely unresolved phylogenetic hypotheses, possibly due to the rapid radiation of the group, whereas the combined analysis of all molecular evidence resulted in a well-resolved phylogeny with higher bootstrap support. The traditionally used fruit characters of wing shape and composition and orientation of mericarp compression are highly variable. The results of these analyses reveal that Cymopterus and Lomatium, the two largest genera of western North American Apiaceae, are polyphyletic, and that their species are inextricably linked with those of other endemic perennial genera of the region (such as, Aletes, Musineon, Oreoxis, Pseudocymopterus, Pteryxia, and Tauschia), many of which are also not monophyletic. Prior emphasis on characters of the fruit in all systems of classification of the group has led to highly artificial assemblages of species. A complete reassessment of generic limits of all western endemic Apiaceae is required, as is further systematic study of this intractable group. Key words: Apiaceae, Cymopterus, phylogeny, ITS, rps16 intron, morphology. Résumé : Pour évaluer la monophylie du genre Cymopterus et pour s’assurer de sa position phylogénétique parmi les autres genres pérennes des Apiaceae (Umbelliferae) sous famille Apioideae endémiques à l’ouest nord-américain, les auteurs ont conduit des analyses cladistiques en utilisant des séquences d’ADN provenant de la région de l’espaceur interne transcrit du rADN nucléique et de l’intron cpADN rps16, ainsi que du lieu cpADN trnF-trnL-trnT pour un sous ensemble de taxons. Afin d’élucider les patrons dans l’évolution de caractères spécifiques du fruit et d’évaluer leur utilité pour circonscrire les genres de façon non ambiguë, ils ont obtenu des preuves supplémentaires à partir de sections transverses de fruits matures et de résultats d’analyses cladistiques portant sur 25 caractères morphologiques. L’analyse des ensembles de données réparties conduit à des hypothèses phylogénétiques faiblement supportées et largement irrésolues, possiblement dû à la rapide radiation de ce groupe, alors que les analyses combinées de toute la preuve moléculaire conduit à une phylogénie bien définie avec un fort support en lacet. Les caractères traditionnellement utilisés du fruit tel que la forme de l’aile et la composition ainsi que l’orientation de la compression du méricarpe sont fortement variables. Les résultats de ces analyses révèlent que les genres Cymopterus et Lomatium, les deux plus grands genres d’Apiaceae nord-américaines, sont polyphylétiques, et que leurs espèces sont inextricablement liées avec celles de d’autres genres endémiques et pérennes de la région (tels que Aletes, Musineon, Oreoxis, Pseudocymopterus, Pteryxia et Tauschia) dont plusieurs ne sont également pas monophylétiques. L’emphase placées jusqu’ici sur les caractères du fruit dans tous les systèmes de classification du groupe à conduit à des assemblages très artificiels d’espèces. On doit revoir complètement les limites génériques de toutes les Apiaceae nord-américaines endémiques afin de poursuivre l’étude systématique de ce groupe récalcitrant. Mots clés : Apiaceae, Cymopterus, phylogénie, ITS, rps16 intron, morphologie. [Traduit par la Rédaction] Downie et al. 1324 Received 23 July 2002. Published on the NRC Research Press Web site at http://canjbot.nrc.ca on 24 January 2003. S.R. Downie,1 F.-J. Sun, and D.S. Katz-Downie. Department of Plant Biology, University of Illinois at Urbana-Champaign, Urbana, IL 61801, U.S.A. R.L. Hartman. Department of Botany, University of Wyoming, P.O. Box 3165, Laramie, WY 82071, U.S.A. 1 Corresponding author (e-mail: sdownie@life.uiuc.edu). Can. J. Bot. 80: 1295–1324 (2002) DOI: 10.1139/B02-119 © 2002 NRC Canada 1296 Can. J. Bot. Downloaded from www.nrcresearchpress.com by Canadian Science Publishing on 06/08/15 For personal use only. Introduction Considerable confusion exists with regard to the proper delimitation of and relationships among the perennial endemic genera of western North American Apiaceae (Umbelliferae) subfamily Apioideae. This confusion is particularly evident in those taxa surrounding Cymopterus Raf. (the spring-parsleys). Historical treatments range from the recognition of many small, generically distinct elements (such as Aulospermum J.M. Coult. & Rose, Glehnia F. Schmidt ex Miq., Oreoxis Raf., Phellopterus (Nutt. ex Torr. & A. Gray) J.M. Coult. & Rose, Pseudocymopterus J.M. Coult. & Rose, Pteryxia (Nutt. ex Torr. & A. Gray) J.M. Coult. & Rose, and Rhysopterus J.M. Coult. & Rose; Coulter and Rose 1900; Mathias 1930; Mathias and Constance 1944–1945) to different sections and subgroups within the highly variable and expanded genus Cymopterus (Jones 1908). Contemporary treatments recognize Oreoxis, Pseudocymopterus, and Pteryxia as distinct genera (e.g., Kartesz 1994), or include them within a broadly circumscribed Cymopterus (Cronquist 1997; Table 1). Putatively related to Cymopterus sensu lato are the genera Aletes J.M. Coult. & Rose, Harbouria J.M. Coult. & Rose, Lomatium Raf., Musineon Raf., Neoparrya Mathias, Oreonana Jeps., Orogenia S. Watson, Podistera S. Watson, Shoshonea Evert & Constance, and Tauschia Schltdl. (Mathias 1930; Evert and Constance 1982; Sun et al. 2000). Many of these genera have a xerophytic or semixerophytic habit. They occur practically without exception in the dry, sandy, or alkaline regions of western North America (NA) and usually in montane or alpine habitats (Mathias 1930). Many species are narrowly distributed and have strict edaphic requirements. They are all herbaceous perennials and are frequently low-growing and acaulescent. Traditionally, classification of Apiaceae has been based on anatomical and morphological features of the mature fruit, sometimes to the exclusion of all other characters. Many of these features are apparent only after detailed examination and sectioning. In most umbellifers, the dry schizocarp splits down a broad commissure into two one-seeded mericarps that are typically joined by a central stalk (carpophore). In some species, the carpophore may be obsolete by adnation of its halves to the commissural faces of the mericarps. The fruit may be compressed laterally, at right angles to the commissural plane, or dorsally, parallel to the commissural plane, if it is compressed at all. Each mericarp commonly bears five primary, longitudinal ribs or ridges that contain the vascular bundles: three dorsal and two marginal (or lateral), with the ribs filiform to broadly winged. Oil canals (vittae) are commonly present in the intervals between the primary ridges, with additional vittae occurring on the commissural face. In the absence of mature fruits, many perennial species of Apiaceae endemic to western NA are essentially indistinguishable. Indeed, when considered collectively, these plants present such a confusing integration of characters that generic delimitation is made exceedingly difficult. The genus Cymopterus, as currently treated, consists of some 35–45 species, with Utah, Nevada, Idaho, and California holding the greatest diversity (Kartesz 1994; Cronquist 1997; Table 1). Its name is derived from the Greek kyma, a wave, and pteron, a wing, referring to the often undulate Can. J. Bot. Vol. 80, 2002 wings of the fruit, for the marginal and usually one or more of the dorsal ribs are conspicuously winged. However, these ribs and wings vary greatly in shape and composition, as does the orientation of fruit compression. The ribs may appear as inconspicuous lines or be highly prominent. The shape of the wings in cross section may be short or extended into linear projections of various forms, and their composition may vary from thin and scarious to thick and corky. Loss of the carpophore occurs in nearly half of the species, presumed to have happened several times independently during the evolution of the genus (Hartman and Constance 1985; Cronquist 1997; Hartman 2000). The number of vittae in the intervals varies from 3 to 5, but in some species there may be only one. Most species are caespitose, with the taproot surmounted by a branching, surficial caudex, while others have pseudoscapes arising from the subterranean crown of the taproot (Cronquist 1997). Cymopterus occurs in a wide variety of, and often very restricted, habitat types, and its concomitant variation in growth forms and fruit types makes any taxonomic definition of the genus difficult and precludes inferences of infrageneric relationships. Phylogenetic studies of these endemic and largely cordilleran genera are few, and have focused almost exclusively upon Lomatium (Schlessman 1984; Simmons 1985; Mastrogiuseppe et al. 1985; Gilmartin and Simmons 1987; Soltis and Kuzoff 1993; Soltis and Novak 1997; Soltis et al. 1995; Hardig and Soltis 1999). To ascertain what genera might be most closely related to Lomatium, Gilmartin and Simmons (1987) carried out phenetic analyses and, for one of the phenetic groups they delimited, a cladistic analysis was conducted using morphological data. Their examination of 88 NA genera using combinations of character states for three binary characters revealed 7 phenetic alliances, with one group (the “Lomatium alliance”) comprising Lomatium, Cymopterus, Glehnia, Polytaenia DC., Prionosciadium S. Watson, Pseudocymopterus, and Pteryxia. This group was the closest phenetically to two other alliances, represented by such genera as Aletes, Donnellsmithia J.M. Coult. & Rose, Harbouria, Musineon, Neoparrya, Orogenia, Taenidia (Torr. & A. Gray) Drude, Tauschia, Thaspium Nutt., and Zizia W.D.J. Koch. The results of their cladistic analyses were equivocal in suggesting a clear sister group to Lomatium, but did highlight the possible paraphyletic nature of Cymopterus (and Pteryxia). Monophyly of all remaining taxa was tacitly assumed. Herein, we present results of a phylogenetic study of Cymopterus and its allies based on molecular and morphological evidence. Our first objective is to evaluate the monophyly of Cymopterus. However, given the complex and overlapping patterns of morphological character variation observed, both within the genus and among its putative allies, we hypothesize a priori that the genus is not monophyletic. Thus, our second objective is to determine the phylogenetic relationships of the elements that currently comprise Cymopterus with other perennial, endemic umbellifers of western NA. Patterns in the evolution of individual morphological characters and their usefulness in clade determination will be assessed, as will the comparative utility of DNA sequence data for several chloroplast and nuclear loci in their ability to resolve relationships among these taxa. The results obtained will eventually enable us to © 2002 NRC Canada Downie et al. 1297 Table 1. A comparison of taxonomic treatments for Cymopterus sensu lato and selected allies. Can. J. Bot. Downloaded from www.nrcresearchpress.com by Canadian Science Publishing on 06/08/15 For personal use only. Mathias (1930) Musineon divaricatum (Pursh) Nutt. ex Torr. & A. Gray Musineon divaricatum var. hookeri (Torr. & A. Gray) Mathias Musineon vaginatum Rydb. Musineon lineare (Rydb.) Mathias Musineon tenuifolium Nutt. ex Torr. & A. Gray Mathias and Constance (1944–1945) Weber (1984, 1991); Weber and Wittmann (1992) Kartesz (1994) Cronquist (1997) Musineon divaricatum Musineon divaricatum Musineon divaricatum Musineon divaricatum var. hookeri Musineon divaricatum var. hookeri Musineon divaricatum Musineon vaginatum Musineon lineare Musineon vaginatum Musineon lineare Musineon vaginatum Musineon lineare Musineon tenuifolium Aletes tenuifolius (Nutt. ex Torr. & A. Gray) W.A. Weber Musineon tenuifolium Rhysopterus plurijugus J.M. Coult. & Rose Rhysopterus plurijugus Cymopterus corrugatus Neoparrya lithophila Mathias Neoparrya lithophila Aletes lithophilus (Mathias) W.A. Weber Neoparrya lithophila Aletes acaulis (Torr.) J.M. Coult. & Rose Aletes humilis J.M. Coult. & Rose Aletes acaulis Aletes acaulis Aletes acaulis Aletes humilis Aletes humilis Aletes humilis Aletes sessiliflorus W.L. Theob. & C.C. Tseng Aletes eastwoodiae (J.M. Coult. & Rose) W.A. Weber Aletes juncea (Barneby & N.H. Holmgren) W.A. Weber Aletes latilobus (Rydb.) W.A. Weber Aletes minima (Mathias) W.A. Weber Aletes nuttallii (A. Gray) W.A. Weber Aletes parryi (S. Watson) W.A. Weber Aletes scabra (J.M. Coult. & Rose) W.A. Weber Aletes sessiliflorus Cymopterus corrugatus Aletes filifolius Mathias, Constance & W.L. Theob. Oreoxis alpina (A. Gray) J.M. Coult. & Rose Oreoxis humilis Raf. Oreoxis bakeri J.M. Coult. & Rose Oreoxis alpina Oreoxis alpina Oreoxis humilis Oreoxis bakeri Oreoxis alpina subsp. puberulenta W.A. Weber Oreoxis humilis Oreoxis bakeri Cymopterus alpinus A. Gray Cymopterus bakeri (J.M. Coult. & Rose) M.E. Jones © 2002 NRC Canada 1298 Can. J. Bot. Vol. 80, 2002 Table 1 (continued). Can. J. Bot. Downloaded from www.nrcresearchpress.com by Canadian Science Publishing on 06/08/15 For personal use only. Mathias (1930) Oreoxis macdougalii (J.M. Coult. & Rose) Rydb. Mathias and Constance (1944–1945) Aletes macdougalii J.M. Coult. & Rose Weber (1984, 1991); Weber and Wittmann (1992) Aletes macdougalii Aletes macdougalii subsp. breviradiatus W.L. Theob. & C.C. Tseng Pseudocymopterus montanus (A. Gray) J.M. Coult. & Rose Pseudocymopterus davidsonii (J.M. Coult. & Rose) Mathias Pseudocymopterus anisatus (A. Gray) J.M. Coult. & Rose Pseudocymopterus humboldtensis (M.E. Jones) Mathias Pseudocymopterus bipinnatus (S. Watson) J.M. Coult. & Rose Pseudocymopterus nivalis (S. Watson) Mathias Pseudocymopterus hendersonii J.M. Coult. & Rose Kartesz (1994) Aletes macdougalii Aletes macdougalii subsp. breviradiatus Cronquist (1997) Cymopterus macdougalii (J.M. Coult. & Rose) Tidestr. Cymopterus macdougalii Oreoxis trotteri S.L. Welsh & S. Goodrich Cymopterus trotteri (S.L. Welsh & S. Goodrich) Cronquist Pseudocymopterus montanus Pseudocymopterus montanus Pteryxia davidsonii (J.M. Coult. & Rose) Mathias & Constance Pteryxia anisata (A. Gray) Mathias & Constance Cymopterus humboldtensis M.E. Jones Cymopterus bipinnatus S. Watson Pteryxia davidsonii Cymopterus lemmonii (J.M. Coult. & Rose) Dorn Pseudocymopterus davidsonii Cymopterus nivalis S. Watson Pteryxia hendersonii (J.M. Coult. & Rose) Mathias & Constance Aletes anisatus (A. Gray) W.L. Theob. & C.C. Tseng Aletes anisatus Cymopterus nivalis Cymopterus nivalis Aletes bipinnata (S. Watson) W.A. Weber Aletes nivalis (S. Watson) W.A. Weber Aletes hendersonii (J.M. Coult. & Rose) W.A. Weber Cymopterus nivalis Cymopterus nivalis Cymopterus nivalis Cymopterus nivalis Pteryxia hendersonii Aletes longiloba (Rydb.) W.A. Weber Pteryxia hendersonii Cymopterus hendersonii (J.M. Coult. & Rose) Cronquist Cymopterus hendersonii Pseudocymopterus longiradiatus Mathias, Constance & W.L. Theob. Pteryxia terebinthina (Hook.) J.M. Coult. & Rose Pteryxia terebinthina Pteryxia terebinthina Pteryxia terebinthina var. foeniculacea (Nutt. ex Torr. & A. Gray) Mathias Pteryxia terebinthina var. foeniculacea Pteryxia terebinthina var. foeniculacea Pteryxia terebinthina var. calcarea (M.E. Jones) Mathias Pteryxia terebinthina var. calcarea Pteryxia terebinthina var. albiflora Cymopterus terebinthinus (Hook.) Torr. & A. Gray Cymopterus terebinthinus var. foeniculaceus (Nutt. ex Torr. & A. Gray) Cronquist Cymopterus terebinthinus var. albiflorus (Nutt. ex Torr. & A. Gray) M.E. Jones © 2002 NRC Canada Downie et al. 1299 Can. J. Bot. Downloaded from www.nrcresearchpress.com by Canadian Science Publishing on 06/08/15 For personal use only. Table 1 (continued). Mathias (1930) Pteryxia terebinthina var. californica (J.M. Coult. & Rose) Mathias Pteryxia terebinthina var. albiflora (Nutt. ex Torr. & A. Gray) Mathias Pteryxia petraea (M.E. Jones) J.M. Coult. & Rose Aulospermum longipes (S. Watson) J.M. Coult. & Rose Mathias and Constance (1944–1945) Pteryxia terebinthina var. californica Weber (1984, 1991); Weber and Wittmann (1992) Pteryxia terebinthina var. albiflora Pteryxia petraea Cymopterus longipes S. Watson Kartesz (1994) Pteryxia terebinthina var. californica Pteryxia terebinthina var. albiflora Aletes petraeus (M.E. Jones) W.A. Weber Pteryxia petraea Cronquist (1997) Cymopterus terebinthinus var. albiflorus Cymopterus terebinthinus var. albiflorus Cymopterus petraeus M.E. Jones Cymopterus longipes Cymopterus longipes Cymopterus lapidosus (M.E. Jones) M.E. Jones Cymopterus planosus Cymopterus longipes Cymopterus ibapensis Cymopterus longipes var. ibapensis (M.E. Jones) Cronquist Aulospermum planosum Osterh. Aulospermum ibapense (M.E. Jones) J.M. Coult. & Rose Cymopterus planosus (Osterh.) Mathias Cymopterus ibapensis M.E. Jones Aulospermum glaucum (Nutt.) J.M. Coult. & Rose Aulospermum watsonii J.M. Coult. & Rose Cymopterus glaucus Nutt. Cymopterus glaucus Cymopterus watsonii (J.M. Coult. & Rose) M.E. Jones Cymopterus aboriginum M.E. Jones Cymopterus minimus (Mathias) Mathias Cymopterus basalticus M.E. Jones Cymopterus rosei (M.E. Jones ex J.M. Coult. & Rose) M.E. Jones Cymopterus duchesnensis M.E. Jones Cymopterus purpureus S. Watson Cymopterus ibapensis Cymopterus longipes var. ibapensis Cymopterus aboriginum Cymopterus aboriginum Cymopterus minimus Aulospermum aboriginum (M.E. Jones) Mathias Aulospermum minimum Mathias Aulospermum basalticum (M.E. Jones) Tidestr. Aulospermum rosei M.E. Jones ex J.M. Coult. & Rose Aulospermum duchesnense (M.E. Jones) Tidestr. Aulospermum purpureum (S. Watson) J.M. Coult. & Rose Aulospermum jonesii (J.M. Coult. & Rose) J.M. Coult. & Rose Aulospermum panamintense (J.M. Coult. & Rose) J.M. Coult. & Rose Aulospermum panamintense var. acutifolium J.M. Coult. & Rose Cymopterus minimus Cymopterus basalticus Cymopterus rosei Cymopterus basalticus Cymopterus rosei Cymopterus duchesnensis Cymopterus duchesnensis Cymopterus purpureus Cymopterus purpureus Cymopterus jonesii J.M. Coult. & Rose Cymopterus jonesii Cymopterus jonesii Cymopterus panamintensis J.M. Coult. & Rose Cymopterus panamintensis var. acutifolius (J.M. Coult. & Rose) Munz Cymopterus panamintensis Cymopterus panamintensis var. acutifolius © 2002 NRC Canada 1300 Can. J. Bot. Vol. 80, 2002 Table 1 (continued). Can. J. Bot. Downloaded from www.nrcresearchpress.com by Canadian Science Publishing on 06/08/15 For personal use only. Mathias (1930) Phellopterus montanus Nutt. ex. Torr. & A. Gray Phellopterus macrorhizus (Buckley) J.M. Coult. & Rose Phellopterus bulbosus (A. Nelson) J.M. Coult. & Rose Phellopterus purpurascens (A. Gray) J.M. Coult. & Rose Phellopterus multinervatus J.M. Coult. & Rose Cymopterus cinerarius A. Gray Cymopterus megacephalus M.E. Jones Cymopterus deserticola Brandegee Cymopterus globosus (S. Watson) S. Watson Cymopterus coulteri (M.E. Jones) Mathias Cymopterus corrugatus M.E. Jones Cymopterus acaulis (Pursh) Raf. Cymopterus fendleri A. Gray Cymopterus newberryi (S. Watson) M.E. Jones Mathias and Constance (1944–1945) Cymopterus montanus Nutt. ex Torr. & A. Gray Cymopterus macrorhizus Buckley Weber (1984, 1991); Weber and Wittmann (1992) Kartesz (1994) Cymopterus montanus Cronquist (1997) Cymopterus macrorhizus Cymopterus bulbosus A. Nelson Cymopterus bulbosus Cymopterus bulbosus Cymopterus purpurascens (A. Gray) M.E. Jones Cymopterus multinervatus (J.M. Coult. & Rose) Tidestr. Cymopterus purpurascens Cymopterus purpurascens Cymopterus multinervatus Cymopterus multinervatus Cymopterus cinerarius Cymopterus cinerarius Cymopterus megacephalus Cymopterus deserticola Cymopterus megacephalus Cymopterus deserticola Cymopterus cinerarius Cymopterus megacephalus Cymopterus globosus Cymopterus globosus Cymopterus globosus Cymopterus coulteri Cymopterus coulteri Cymopterus coulteri Cymopterus corrugatus Cymopterus corrugatus Cymopterus acaulis Cymopterus acaulis Cymopterus corrugatus Cymopterus acaulis Cymopterus fendleri Cymopterus acaulis var. fendleri (A. Gray) S. Goodrich Cymopterus acaulis var. greeleyorum J.W. Grimes & P.L. Packard Cymopterus acaulis var. higginsii (S.L. Welsh) S. Goodrich Cymopterus acaulis var. parvus S. Goodrich Cymopterus newberryi Cymopterus newberryi Cymopterus ripleyi Barneby Cymopterus gilmanii C. Morton Cymopterus ripleyi Cymopterus acaulis var. fendleri Cymopterus acaulis var. greeleyorum Cymopterus acaulis var. fendleri Cymopterus acaulis var. greeleyorum Cymopterus newberryi Cymopterus ripleyi Cymopterus gilmanii Cymopterus beckii S.L. Welsh & S. Goodrich Cymopterus davisii R.L. Hartm. Cymopterus evertii R.L. Hartm. & R.S. Kirkp. Cymopterus beckii Cymopterus davisii Cymopterus evertii © 2002 NRC Canada Downie et al. 1301 Table 1 (concluded). Can. J. Bot. Downloaded from www.nrcresearchpress.com by Canadian Science Publishing on 06/08/15 For personal use only. Mathias (1930) Mathias and Constance (1944–1945) Weber (1984, 1991); Weber and Wittmann (1992) Kartesz (1994) Cymopterus goodrichii S.L. Welsh & Neese Cymopterus douglassii R.L. Hartm. & Constance Cymopterus williamsii R.L. Hartm. & Constance Cronquist (1997) Cymopterus goodrichii Note: Authors of plant names are standardized according to Brummitt and Powell (1992). achieve our broader goals, which are to define and delimit the various generic elements that have been confused with Cymopterus and to produce a modern classification of the group that reflects its evolutionary history. Materials and methods Molecular tools For phylogenetic inference, we have exploited variation in the two nuclear ribosomal DNA (rDNA) internal transcribed spacers (ITS) and the chloroplast DNA (cpDNA) rps16 intron. Previous studies have demonstrated the utility of these loci for estimating infrafamilial relationships in other Apiaceae (Downie and Katz-Downie 1996, 1999; Downie et al. 1998, 2000a, 2000c; Lee and Downie 1999, 2000), as well as in angiosperms in general (Baldwin et al. 1995; Lidén et al. 1997; Oxelman et al. 1997). For a subset of taxa, we also examined variation from the cpDNA trnF-trnL-trnT (trnF-L-T) locus. This region, comprising two intergenic spacers and the trnL intron, has not been used for phylogenetic study of Apiaceae, although it has been used successfully in other groups at comparable taxonomic levels (Taberlet et al. 1991; Gielly and Taberlet 1994). Congruence of relationships derived from independent lines of evidence is necessary to examine the robustness of the phylogenetic hypothesis and to identify discrepant organismal and gene phylogenies. Accessions examined One hundred and fifty accessions representing 148 species in 73 genera of Apiaceae subfamily Apioideae were examined for ITS, rps16 intron, and (or) trnF-L-T sequence variation (Table 2). Complete ITS sequences for 66 taxa are reported here for the first time; combining these with 82 previously published or available ITS sequences yielded a matrix of 148 taxa for a global analysis. For two Lomatium species, data for only ITS-1 were available (Soltis and Kuzoff 1993). Fifty-six complete rps16 intron sequences (plus a portion of its flanking 3′ exon region) were procured as part of this study and combined with 29 previously published sequences for a matrix of 85 taxa. Twenty-seven complete trnF-L-T sequences were also obtained, representing the trnF-trnL and trnL-trnT intergenic spacer regions, gene trnL with its intron, and portions of genes trnF and trnT. Eighty-three accessions were included in both the ITS and rps16 intron analyses; 27 species were common to all three molecular data sets, including the analysis of morphological data. Kartesz (1994), whose checklist of Apiaceae was influenced by the work of Lincoln Constance, recognized 78 genera of subfamily Apioideae in NA (north of Mexico). Of these, we have sampled 61 plus nine meso-American genera (Arracacia Bancr., Coaxana J.M. Coult. & Rose, Coulterophytum B.L. Rob., Dahliaphyllum Constance & Breedlove, Enantiophylla J.M. Coult. & Rose, Mathiasella Constance & C.L. Hitchc., Myrrhidendron J.M. Coult. & Rose, Prionosciadium, and Rhodosciadium S. Watson). These meso-American genera (plus Donnellsmithia) are endemic to the highlands of Mexico and neighboring Central America (Mathias 1965), exhibit a large number of paleopolyploid members (Bell and Constance 1966; Moore 1971), and have been provisionally recognized as the Arracacia clade (Downie et al. 2000b, 2001). A previous study had suggested a possible affinity of Prionosciadium and Donnellsmithia with Cymopterus (Gilmartin and Simmons 1987); the fruits of many of these meso-American taxa are morphologically similar to those of Angelica and some Lomatium. In this paper, we follow the nomenclature of Kartesz (1994) with one exception — the name Helosciadium nodiflorum (L.) W.D.J. Koch replaces Apium nodiflorum (L.) Lag. (Downie et al. 2000b, 2000c). Emphasis was placed on sampling the perennial endemic members of western NA Apioideae, with the selection of species influenced primarily by material availability. We examined 17 of the 35 species of Cymopterus recognized by Kartesz (1994). Also examined were five of six species of Aletes, 28 of 78 species of Lomatium, three of four species of Musineon, three of four species of Oreoxis, and two of four species of Pteryxia. The Eurasian genera Physospermum Cusson and Pleurospermum Hoffm. were used to root all trees in the global analyses; their selection as outgroups is based on previous higher-level studies (summarized in Downie et al. 2001). Experimental strategy Leaf material for DNA extraction was obtained either directly from the field, from plants cultivated from seed in the greenhouse, from herbarium specimens, or from the personal collections of Lincoln Constance (University of California Botanical Garden, Berkeley, Calif.). Vouchers and their depositions are indicated in Table 2. Details of DNA extraction, polymerase chain reaction (PCR) primer construction © 2002 NRC Canada 1302 Can. J. Bot. Vol. 80, 2002 Table 2. Species of Apiaceae subfamily Apioideae examined for nuclear rDNA ITS (148 taxa), cpDNA rps16 intron (85 taxa), and (or) trnF-L-T (27 species) sequence variation. Species Can. J. Bot. Downloaded from www.nrcresearchpress.com by Canadian Science Publishing on 06/08/15 For personal use only. Aegopodium podagraria L. Aethusa cynapium L. Aletes acaulis (Torr.) J.M. Coult. & Rose Aletes anisatus (A. Gray) W.L. Theob. & C.C. Tseng Aletes humilis J.M. Coult. & Rose Aletes macdougalii J.M. Coult. & Rose subsp. breviradiatus W.L. Theob. & C.C. Tseng Aletes sessiliflorus W.L. Theob. & C.C. Tseng Ammi majus L. Anethum graveolens L. Angelica ampla A. Nelson Angelica archangelica L. subsp. archangelica Angelica arguta Nutt. ex Torr. & A. Gray Angelica breweri A. Gray Angelica pinnata S. Watson Angelica roseana L.F. Hend. Angelica sylvestris L. Anthriscus caucalis M. Bieb. Apium graveolens L. Arracacia aegopodioides (Kunth) J.M. Coult. & Rose Arracacia bracteata J.M. Coult. & Rose Arracacia brandegei J.M. Coult. & Rose Arracacia nelsonii J.M. Coult. & Rose Arracacia tolucensis (Kunth) Hemsl. var. tolucensis Arracacia tolucensis var. multifida (S. Watson) Mathias & Constance Berula erecta (Huds.) Coville Bifora radians M. Bieb. Carum carvi L. Caucalis platycarpos L. Chaerophyllum tainturieri Hook. Source and voucher Downie et al. 1998 Downie et al. 1998 U.S.A., Colorado, Larimer Co., Canyon of the Big Thompson, 15 July 1989, Hartman 24386 (RM) U.S.A., Colorado, Park Co., Corral Creek, 6 August 1995, Chumley 2807 (RM) Downie et al. 1998 U.S.A., New Mexico, San Juan Co., Blanco, 2 May 1982, Hartman 13963 (RM) GenBank accession No. ITS-1, ITS-2 rps16 intron U30536, U30537 U30582, U30583 AF358461, AF358528 AF110539 AF358595 AF358462, AF358529 AF358596 U78401, U78461 AF358463, AF358530 AF358597 U.S.A., New Mexico, Rio Arriba Co., NW of Embudo, 1 May 1992, Hartman 13954 (RM) Downie et al. 1998 Downie et al. 1998 Downie et al. 1998 Downie et al. 1998 U78386, U30550, U79597, U30576, Downie et al. 1998 U79599, U79600 Downie et al. 1998 U.S.A., Wyoming, Lincoln Co., Commissary Ridge, 22 July 1993, Hartman 41500 (RM) U.S.A., Wyoming, Teton Co., Blue Miner Lake, 25 August 1994, Hartman 50090 (RM) Downie et al. 1998 Downie et al. 1998 Downie et al. 1998 Cult. UC Berkeley; Mexico, Oaxaca, Breedlove 72231 (CAS), L. Constance pers. coll. C-2408 Cult. UC Berkeley; Mexico, Oaxaca, Breedlove 72536 (CAS), L. Constance pers. coll. C-2412 Downie et al. 1998 U78396, U78456 AF358465, AF358532 Downie et al. 1998 U30556, U30557 Cult. UC Berkeley; Mexico, Querétaro, Cerro Zamorano, 16 December 1978, Ornduff 8560 (UC), L. Constance pers. coll. C -2124 Cult. UC Berkeley; Mexico, UNAM 88, L. Constance pers. coll. C-2355 AF358469, AF358536 Downie Downie Downie Downie Downie et et et et et al. al. al. al. al. 1998 1998 1998 1998 2000a trnF-L-T AF444008 AF358464, AF358531 U78446 U30551 U79598 U30577 AF164814 AF110542 AF358598 AF110536 AF444007 AF358599 AF358600 AF358466, AF358533 U78414, U78474 U79601, U79602 U30552, U30553 AF358467, AF358534 AF110549 AF110545 AF358468, AF358535 U30570, U30571 AF358470, AF358537 U79605, U79606 U78408, U78468 U78377, U78437 U78364, U78424 AF073647, AF073648 AF164819 AF164809 AF123745 © 2002 NRC Canada Downie et al. 1303 Can. J. Bot. Downloaded from www.nrcresearchpress.com by Canadian Science Publishing on 06/08/15 For personal use only. Table 2 (continued). Species Source and voucher Ciclospermum leptophyllum (Pers.) Sprague ex Britton & E.H. Wilson Cicuta maculata L. var. angustifolia Hook. Coaxana purpurea J.M. Coult. & Rose Conioselinum chinense (L.) Britton, Stern & Poggenb. Conioselinum scopulorum (A. Gray) J.M. Coult. & Rose Conium maculatum L. Coriandrum sativum L. Coulterophytum jaliscense McVaugh U.S.A., Oklahoma, Pittsburg Co., 9 May 1991, Seigler et al. 13245 (ILL) Coulterophytum laxum B.L. Rob. Cryptotaenia canadensis (L.) DC. Cuminum cyminum L. Cymopterus acaulis (Pursh) Raf. var. acaulis Cymopterus acaulis var. fendleri (A. Gray) S. Goodrich Cymopterus basalticus M.E. Jones Cymopterus bulbosus A. Nelson Cymopterus duchesnensis M.E. Jones Cymopterus evertii R.L. Hartm. & R.S. Kirkp. Cymopterus globosus (S. Watson) S. Watson Cymopterus ibapensis M.E. Jones Cymopterus jonesii J.M. Coult. & Rose Cymopterus longipes S. Watson Cymopterus montanus Nutt. ex Torr. & A. Gray Cymopterus multinervatus (J.M. Coult. & Rose) Tidestr. Cymopterus nivalis S. Watson Cymopterus panamintensis J.M. Coult. & Rose var. panamintensis GenBank accession No. ITS-1, ITS-2 AF358471, AF358538 rps16 intron AF358472, AF358539 AF358601 U.S.A., Wyoming, Goshen Co., Bear Creek, 4 August 1994, Nelson et al. 33517 (RM) Downie et al. 1998 U30572, U30573 Downie et al. 1998 U78374, U78434 Katz-Downie et al. 1999 AF008634, AF009113 Downie et al. 1998 Downie et al. 1998 Cult. UC Berkeley; Mexico, Jalisco, Zarzamora (Las Joyas), Sierra de Manantlán, Iltis et al. 1299 (UC), L. Constance pers. coll. C-2236 Downie et al. 1998 Downie et al. 1998 Downie et al. 1998 U.S.A., Colorado, Garfield Co., Grand Hogback, Burning Mtn., 27 May 1991, Vanderhorst 2236 (RM) U.S.A., Utah, Emery Co., S of Price River, 14 May 1979, Hartman 8674 (RM) U.S.A., Utah, Millard Co., Desert Range Experiment Station, 22 May 1982, Fonken 1611 (RM) U.S.A., Utah, Uintah Co., ESE of Vernal, 18 April 1982, Hartman 13951 (RM) U.S.A., Utah, Uintah Co., Tridell, 28 May 1982, Hartman 13984 (RM) U.S.A., Wyoming, Hot Springs Co., SE of Meeteetse, 30 May 1985, Hartman 20097 and Haines (RM) Downie et al. 1998 U30588, U30589 U30586, U30587 AF358473, AF358540 U.S.A., Utah, Sevier Co., UT 4, 26 May 1982, Hartman 13978 (RM) U.S.A., Utah, Washington Co., road to Apex Mine, 20 May 1981, Fonken 1195 (RM) U.S.A., Wyoming, Lincoln Co., Grade Canyon Creek, 22 May 1993, Hartman 37464 (RM) U.S.A., Colorado, El Paso Co., Rockrimmon Road, 18 May 1982, Hartman 13968 (RM) U.S.A., Arizona, Mohave Co., Mt. Trumbull, 31 March 1983, Hartman 14098 (RM) U.S.A., Wyoming, Teton Co., E of Crystal Creek, 24 June 1994, Hartman 46444 and Cramer (RM) U.S.A., California, Inyo Co., Argus Range NNE of Ridgecrest, Ertter 7043 (UC) U30560, U30561 U79613, U79614 U78362, U78422 AF358474, AF358541 AF358475, AF358542 trnF-L-T AF110546 AF358602 AF358603 AF358604 AF358476, AF358543 AF358477, AF358544 AF358478, AF358545 AF358605 AF358479, AF358546 AF358606 U78398, U78458 AF358607 AF444009 AF358480, AF358547 AF358481, AF358548 AF358608 AF358483, AF358550 AF358609 AF358484, AF358551 AF110534 AF444010 AF358485, AF358552 AF358610 AF444011 AF358486, AF358553 AF358611 AF444012 AF358487, AF358554 © 2002 NRC Canada 1304 Can. J. Bot. Vol. 80, 2002 Table 2 (continued). Species Source and voucher Cymopterus planosus (Osterh.) Mathias U.S.A., Colorado, Routt Co., base of Dunckley Flat Tops, 16 June 1991, Vanderhorst 2592 (RM) U.S.A., Arizona, Mohave Co., NE of Peach Springs, 30 March 1983, Hartman 14096 (RM) U.S.A., Colorado, Garfield Co., Grand Hogback, 25 May 1991, Vanderhorst 2166a (RM) U.S.A., Wyoming, Natrona Co., along Baker Cabin, 23 May 1994, Nelson 30642 (RM) U.S.A., Illinois, Jackson Co., Shawnee Natl. Forest, 27 May 1993, Phillippe 21886 (ILLS) Downie et al. 1998 U78395, U78455 Lee and Downie 1999 Downie et al. 1998 AF077788, AF077103 U30558, U30559 AF123729 Katz-Downie et al. 1999 Downie et al. 1998 Downie et al. 1998 U.S.A., New Mexico, Colfax Co., Philmont Scout Ranch, 24 June 1991, Embry 56 (RM) Downie et al. 2000c AF008636, AF009115 U78378, U78438 U78385, U78445 AF358493, AF358560 AF110554 AF164823, AF164848 AF164820 Downie et al. 1998 Downie et al. 1998 Katz-Downie et al. 1999 Downie et al. 1998 U30544, U30545 U78389, U78449 AF008635, AF009114 U78375, U78435 AF164800 Downie et al. 1998 Petersen et al. 2002 U78357, U78417 AF466276 AF123756 U.S.A., Wyoming, Sublette Co., Packsaddle Ridge, 21 June 1993, Nelson 26111 and Nelson (RM) Hardig and Soltis 1999 AF358494, AF358561 AF358616 AF444016 AF011803, AF011820 Downie et al. 1998 U78397, U78457 AF358617 AF444017 Hardig and Soltis 1999 AF011804, AF011821 U.S.A., Wyoming, Sublette Co., Palmer Peak, 5 August 1994, Hartman 49374 (RM) Downie et al. 1998 AF358495, AF358562 AF358618 U30580, U30581 AF358619 Hardig and Soltis 1999 AF011809, AF011826 U.S.A., Wyoming, Converse Co., Southern Powder River Basin, 12 May 1994, Nelson 30083 (RM) AF358496, AF358563 Can. J. Bot. Downloaded from www.nrcresearchpress.com by Canadian Science Publishing on 06/08/15 For personal use only. Cymopterus purpurascens (A. Gray) M.E. Jones Cymopterus purpureus S. Watson Cymopterus williamsii R.L. Hartm. & Constance Cynosciadium digitatum DC. Dahliaphyllum almedae Constance & Breedlove Daucus pusillus Michx. Enantiophylla heydeana J.M. Coult. & Rose Erigenia bulbosa (Michx.) Nutt. Falcaria vulgaris Bernh. Foeniculum vulgare Mill. Harbouria trachypleura (A. Gray) J.M. Coult. & Rose Helosciadium nodiflorum (L.) W.D.J. Koch (as Apium nodiflorum (L.) Lag.) Heracleum sphondylium L. Levisticum officinale W.D.J. Koch Ligusticum canadense (L.) Britton Ligusticum porteri J.M. Coult. & Rose var. porteri Ligusticum scoticum L. Lilaeopsis carolinensis J.M. Coult. & Rose Lomatium bicolor (S. Watson) J.M. Coult. & Rose var. bicolor Lomatium brandegei (J.M. Coult. & Rose) J.F. Macbr. Lomatium californicum (Nutt.) Mathias & Constance Lomatium concinnum (Osterh.) Mathias Lomatium cous (S. Watson) J.M. Coult. & Rose Lomatium dasycarpum (Torr. & A. Gray) J.M. Coult. and Rose subsp. dasycarpum Lomatium dissectum (Nutt.) Mathias & Constance var. dissectum Lomatium foeniculaceum (Nutt.) J.M. Coult. & Rose subsp. foeniculaceum GenBank accession No. ITS-1, ITS-2 AF358488, AF358555 rps16 intron AF358612 trnF-L-T AF358490, AF358557 AF358613 AF444013 AF358491, AF358558 AF358614 AF444014 AF358489, AF358556 AF358492, AF358559 AF110543 AF358615 AF444015 AF444018 © 2002 NRC Canada Downie et al. 1305 Can. J. Bot. Downloaded from www.nrcresearchpress.com by Canadian Science Publishing on 06/08/15 For personal use only. Table 2 (continued). Species Source and voucher Lomatium graveolens (S. Watson) Dorn & R.L. Hartm. var. graveolens Lomatium grayi (J.M. Coult. & Rose) J.M. Coult. & Rose Lomatium greenmanii Mathias Lomatium howellii (S. Watson) Jeps. Lomatium idahoense Mathias & Constance Lomatium junceum Barneby & N.H. Holmgren Lomatium juniperinum (M.E. Jones) J.M. Coult. & Rose Lomatium laevigatum (Nutt.) J.M. Coult. & Rose Lomatium latilobum (Rydb.) Mathias Lomatium lucidum (Nutt. ex. Torr. & A. Gray) Jeps. Lomatium macrocarpum (Nutt. ex Torr. & A. Gray) J.M. Coult. & Rose Lomatium nudicaule (Pursh) J.M. Coult. & Rose Lomatium nuttallii (A. Gray) J.F. Macbr. Lomatium orientale J.M. Coult. & Rose U.S.A., Wyoming, Sublette Co., Packsaddle Ridge, 21 June 1993, Nelson 26101 and Nelson (RM) Soltis and Kuzoff 1993 (ITS-1 only) (not in GenBank) Hardig and Soltis 1999 Hardig and Soltis 1999 AF011805, AF011822 AF011800, AF011817 Hardig and Soltis 1999 AF011806, AF011823 U.S.A., Utah, Emery Co., NE of Emery, 18 June 1982, Fonken 1962 (RM) U.S.A., Utah, Cache Co., Bear River Range, 12 August 1980, Hartman 11885 (RM) Soltis and Kuzoff 1993 (ITS-1 only) Lomatium parvifolium (Hook. & Arn.) Jeps. Lomatium repostum (Jeps.) Mathias Lomatium rigidum (M.E. Jones) Jeps. Lomatium scabrum (J.M. Coult. & Rose) Mathias var. scabrum Lomatium shevockii R.L. Hartm. & Constance Lomatium triternatum (Pursh) J.M. Coult. & Rose subsp. platycarpum (Torr.) Cronquist Mathiasella bupleuroides Constance & C.L. Hitchc. Musineon divaricatum (Pursh) Nutt. ex Torr. & A. Gray var. divaricatum Musineon tenuifolium Nutt. ex Torr. & A. Gray Musineon vaginatum Rydb. Myrrhidendron donnell-smithii J.M. Coult. & Rose Myrrhis odorata (L.) Scop. Neoparrya lithophila Mathias U.S.A., Utah, Grand Co., SE of Moab, 13 April 1995, Tuby 3772 (RM) Hardig and Soltis 1999 U.S.A., Wyoming, Lincoln Co., Dempsey Ridge, 25 June 1993, Nelson 26537 and Nelson (RM) U.S.A., Nevada, Elko Co., S. of Hot Creek, 16 May 1979, Hartman 8736 (RM) Hardig and Soltis 1999 GenBank accession No. ITS-1, ITS-2 AF358497, AF358564 rps16 intron AF358620 trnF-L-T AF444019 AF358498, AF358565 AF358621 AF444020 AF358499, AF358566 AF358622 AF444021 (not in GenBank) AF358500, AF358567 AF011799, AF011816 AF358501, AF358568 AF358624 AF444022 AF358502, AF358569 AF358625 AF444023 AF011811, AF011828 U.S.A., Wyoming, Natrona Co., along Notches, 23 May 1994, Nelson 30536 (RM) Hardig and Soltis 1999 AF011801, AF011818 Hardig and Soltis 1999 Hardig and Soltis 1999 AF011802, AF011819 AF011797, AF011814 U.S.A., Utah, Millard Co., S. of Ganison, 16 May 1981, Fonken 1168 (RM) Hardig and Soltis 1999 AF358504, AF358571 U.S.A., Wyoming, Lincoln Co., Boulder Ridge, 22 May 1993, Hartman 37526 (RM) Downie et al. 1998 U.S.A., Wyoming, Platte Co., NW of Chugwater, 26 May 1994, Nelson 30905 (RM) U.S.A., Wyoming, Niobrara Co., Hat Creek Breaks, 17 May 1994, Nelson 30335 (RM) U.S.A., Wyoming, Sheridan Co., Big Horn Mtns., 18 June 1979, Hartman 9020 (RM) Downie et al. 1998 (Grantham and Parsons 0433–90) Downie et al. 1998 U.S.A., Colorado, Saguache Co., Upper Saguache Forest Service Station, 18 September 1983, Hartman 17360 (RM) AF358623 AF358503, AF358570 AF358626 AF011798, AF011815 AF358505, AF358572 AF358627 U78394, U78454 AF358506, AF358573 AF358628 AF444024 AF358507, AF358574 AF358629 AF444025 AF358508, AF358575 AF358630 U30554, U30555 U30530, U30531 AF358509, AF358576 AF123755 AF358631 AF444026 © 2002 NRC Canada 1306 Can. J. Bot. Vol. 80, 2002 Can. J. Bot. Downloaded from www.nrcresearchpress.com by Canadian Science Publishing on 06/08/15 For personal use only. Table 2 (continued). Species Source and voucher Oenanthe pimpinelloides L. Oreoxis alpina (A. Gray) J.M. Coult. & Rose subsp. alpina Oreoxis bakeri J.M. Coult. & Rose Downie et al. 1998 U.S.A., Colorado, Rio Blanco Co., Pyramid Peak, 27 June 1991, Vanderhorst 2806 (RM) U.S.A., New Mexico, Santa Fe Co., Lake Peak, 19 June 1980, Hartman 11725 (RM) U.S.A., Colorado, Teller Co., Pikes Peak Road, 17 June 1980, Hartman 11718 (RM) U.S.A., Wyoming, Lincoln Co., Hams Fork Plateau, 23 May 1993, Hartman 37557 (RM) Downie et al. 1998 U.S.A., Illinois, Vermilion Co., Windfall Hill Prairie Nature Reserve, 17 July 1991, Phillippe et al. 19411 (ILLS) Downie et al. 1998 Downie et al. 1998 Oreoxis humilis Raf. Orogenia linearifolia S. Watson Osmorhiza longistylis (Torr.) DC. Oxypolis rigidior (L.) Raf. Pastinaca sativa L. Perideridia kelloggii (A. Gray) Mathias Petroselinum crispum (Mill.) A.W. Hill Physospermum cornubiense (L.) DC. Pimpinella saxifraga L. Pleurospermum foetens Franch. Podistera eastwoodiae (J.M. Coult. & Rose) Mathias & Constance Polytaenia nuttallii DC. Polytaenia texana (J.M. Coult. & Rose) Mathias & Constance Prionosciadium acuminatum B.L. Rob. Prionosciadium simplex Mathias & Constance Prionosciadium turneri Constance & Affolter Prionosciadium watsonii J.M. Coult. & Rose Pseudocymopterus montanus (A. Gray) J.M. Coult. & Rose Pteryxia hendersonii (J.M. Coult. & Rose) Mathias & Constance Pteryxia terebinthina (Hook.) J.M. Coult. & Rose var. albiflora (Nutt. ex Torr. & A. Gray) Mathias Rhodosciadium argutum (Rose) Mathias & Constance Scandix pecten-veneris L. Shoshonea pulvinata Evert & Constance GenBank accession No. ITS-1, ITS-2 U78371, U78431 AF358510, AF358577 rps16 intron AF110553 AF358511, AF358578 AF358632 AF358512, AF358579 AF358633 AF358513, AF358580 AF358634 U79617, U79618 AF358514, AF358581 AF123754 U30546, U30547 U78373, U78433 AF110538 AF358635 Downie et al. 1998 U78387, U78447 AF110544 Downie et al. 1998 U78382, U78442 AF110556 Downie et al. 1998 Katz-Downie et al. 1999 U.S.A., Colorado, Garfield Co., Edge Lake, 3 July 1991, Vanderhorst 3016 (RM) U.S.A., Illinois, Rock Island Co., N of Cordova, 19 June 1973, Evers 110464 (ILLS) U.S.A., Texas, Burnet Co., E of Briggs, 25 May 1985, Barrie 1403 (RM) Cult. UC Berkeley; Mexico, Sinaloa, 3 km NE of Palmito, Breedlove 36448 (CAS), L. Constance pers. coll. C-1871 Cult. UC Berkeley; Mexico, Tamaulipas, Breedlove 63487 (CAS), L. Constance pers. coll. C-2341 Downie et al. 1998 (as Constance pers. coll. C-2053) Cult. UC Berkeley; Mexico, Durango, Breedlove 61338 (CAS), L. Constance pers. coll. C-2330 U.S.A., Colorado, Rio Blanco Co., Dunckley Flat Tops, 17 June 1991, Vanderhorst 2637 (RM) U.S.A., Montana, Ravalli Co., Bitterroot Wilderness, 6 August 1981, Hartman 13889 (RM) U.S.A., Wyoming, Lincoln Co., Twin Creek, 23 May 1993, Hartman 37616 (RM) U30590, U30591 AF008639, AF009118 AF358515, AF358582 AF110559 AF358636 AF358516, AF358583 AF358637 Downie et al. 1998 U30566, U30567 Downie et al. 1998 Downie et al. 1998 U30538, U30539 U78400, U78460 trnF-L-T AF444027 AF358638 AF358517, AF358584 AF358518, AF358585 U30568, U30569 AF358519, AF358586 AF358520, AF358587 AF358639 AF358521, AF358588 AF358640 AF444028 AF358522, AF358589 AF358641 AF444029 AF123753 AF358642 AF444030 © 2002 NRC Canada Downie et al. 1307 Can. J. Bot. Downloaded from www.nrcresearchpress.com by Canadian Science Publishing on 06/08/15 For personal use only. Table 2 (concluded). Species Source and voucher Sium suave Walter Cult. UIUC from seeds obtained from Jardin botanique de Montréal, Canada, Downie 12 (ILL) Katz-Downie et al. 1999 AF008602, AF009081 Katz-Downie et al. 1999 AF008600, AF009079 AF358644 Downie et al. 1998 U.S.A., California, Trinity Co., SE of Burnt Ranch, 11 July 1990, Spellenberg 10254 (RM) U.S.A., California, San Bernardino Co., 12 April 1986, Boyd 1762 (RM) U.S.A., Texas, Gonzales Co., 22 February 1986, Barrie 1435 (RM) U.S.A., Illinois, Champaign Co., N of Tolono, 12 June 1990, Ulaszek 1484 (ILLS) Downie et al. 1998 U78399, U78459 AF358645 AF358646 AF358524, AF358591 AF358647 AF358525, AF358592 AF358648 Downie et al. 1998 U78410, U78470 AF358649 Downie et al. 2000c Downie et al. 1998 AF164844, AF164869 U78380, U78440 AF110548 Lee and Downie 1999 Lee and Downie 1999 AF077810, AF077125 AF077806, AF077121 AF123743 AF123742 U.S.A., Wyoming, Teton Co., road to Granite Falls, 29 May 1994, Hartman 45748 (RM) Downie et al. 1998 AF358527, AF358594 AF358650 AF444032 U30574, U30575 AF110535 AF444033 Spermolepis inermis (Nutt. ex DC.) Mathias & Constance Sphenosciadium capitellatum A. Gray Taenidia integerrima (L.) Drude Tauschia glauca (J.M. Coult. & Rose) Mathias & Constance Tauschia parishii (J.M. Coult. & Rose) J.F. Macbr. Tauschia texana A. Gray Thaspium barbinode (Michx.) Nutt. Thaspium pinnatifidum (Buckley) A. Gray Thaspium trifoliatum (L.) A. Gray var. trifoliatum Torilis arvensis (Huds.) Link Trachyspermum copticum (L.) Link (as T. ammi (L.) Sprague in Turrill) Turgenia latifolia (L.) Hoffm. Yabea microcarpa (Hook. & Arn.) Koso-Pol. Zizia aptera (A. Gray) Fernald Zizia aurea (L.) W.D.J. Koch GenBank accession No. ITS-1, ITS-2 AF358523, AF358590 rps16 intron AF358643 trnF-L-T AF358526, AF358593 U78410, U78470 AF444031 Note: Reference citations indicate source and voucher information for previously published DNA data. ITS data have been deposited with GenBank as separate ITS-1 and ITS-2 sequences. Species nomenclature follows Kartesz (1994); standardized authors names according to Brummitt and Powell (1992); herbarium acronyms according to Holmgren et al. (1990). and amplification, and template purification and sequencing for both ITS and rps16 intron loci are the same as described previously (Downie and Katz-Downie 1996, 1999). Similar procedures were used for the trnF-L-T study, using the primers and PCR amplification protocols of Taberlet et al. (1991). Both manual and automated sequencing methods were used. Simultaneous consideration of both DNA strands across all sequenced regions permitted unambiguous base determination in nearly all cases. Sequence analysis All newly procured sequences were aligned manually in the data editor of PAUP* version 4.0 (Swofford 1998), with gaps positioned to minimize nucleotide mismatches. When alignment was ambiguous because of, for example, tracts of poly-As, -Gs, or -Ts or indirect duplications of adjacent elements in two or more taxa, these positions were eliminated from the analysis. The determination of boundary sequences for the six conserved structural domains of the rps16 group II intron was based on similar boundary sequences inferred for tobacco and mustard (Michel et al. 1989; Neuhaus et al. 1989) and other Apiaceae (Downie et al. 2000c). A similar breakdown of the trnL intron was not done, given its 50% smaller size relative to rps16. Uncorrected pairwise distances (p) were calculated by PAUP*, as they are commonly provided in other angiosperm ITS analyses (Baldwin et al. 1995). All sequence data have been deposited in GenBank (Table 2); aligned data in PAUP* nexus files are available upon request. Phylogenetic analysis of molecular data Initially, a maximum parsimony analysis of ITS data for all 148 taxa was carried out to confirm the placements of the Arracacia clade and Cymopterus sensu lato (including Oreoxis, Pseudocymopterus, and Pteryxia) and its allies within a broader phylogeny. Based on the results of this global analysis and in an effort to increase resolution by reducing the number of excluded positions because of the greater ambiguity involved in aligning sequences from more distantly related taxa, the clade comprising Cymopterus and allies was isolated for subsequent and more comprehensive phylogenetic analysis. For this smaller (local) set of taxa, a © 2002 NRC Canada Can. J. Bot. Downloaded from www.nrcresearchpress.com by Canadian Science Publishing on 06/08/15 For personal use only. 1308 maximum likelihood analysis (Felsenstein 1981) was also performed. All trees in the local analyses were rooted by Aethusa cynapium, as suggested by the global results. Separate analysis of each spacer region was not done. Previous studies, in Apioideae and other angiosperms, have indicated the high complementarity of spacer data and the greater phylogenetic resolution and internal support achieved in trees when both spacers are considered together than when either spacer is treated alone (Baldwin et al. 1995; Downie and Katz-Downie 1996). Data from the rps16 intron and 3′ exon portion were analyzed using maximum parsimony separately (85 species) and, for 83 taxa, in combination with ITS data. The trnF-L-T region was analyzed in its entirety, as well as partitioned into the two intergenic spacers and intron region for separate consideration. The 27 species common to the ITS, rps16 intron, and trnF-L-T data sets were also analyzed separately and in combination using maximum parsimony, with the trees rooted by positioning the root along the branch connecting Angelica archangelica to the rest of the network. To examine the extent of conflict among separate data sets, the incongruence length difference test of Farris et al. (1995) was conducted using the partition homogeneity test of PAUP*. One hundred replicates were considered for each partition, using simple addition sequence of taxa and tree bisection reconnection (TBR) branch swapping. Incongruence among data sets is identified if the additive tree lengths taken from resampled matrices are greater than the sum of the tree lengths from the original data. Prior to carrying out a maximum likelihood analysis of the 85-taxon ITS data set, the program Modeltest vers. 3.06 (Posada and Crandall 1998) was used to select an evolutionary model of nucleotide substitution (among 56 possible models) that best fits these data. The settings appropriate for the chosen model (base frequencies and among-site rate variation) were inputted into PAUP* and a heuristic search performed using ten random addition sequence replicates and TBR branch swapping under maximum likelihood optimization. Analyses of all but the smallest data sets were carried out initially using equally weighted maximum parsimony and the following protocol. One thousand heuristic searches were initiated using random addition starting trees, with TBR branch swapping and multrees selected, but saving no more than five of the shortest trees from each search. These trees were subsequently used as starting trees for further TBR branch swapping. The maximum number of saved trees was set at 20 000 and these trees were permitted to swap to completion. The strict consensus of these 20 000 minimal length trees was then used as a topological constraint in another round of 500–1000 random-addition replicate analyses but, in this case, only those trees that did not fit the constraint tree were saved (Catalán et al. 1997). No additional trees were found at the length of the initial shortest trees, which suggests that the strict consensus tree adequately summarizes the available evidence, even though the exact number of trees at that length is not known. Bootstrap values (Felsenstein 1985) were calculated from 100 000 replicate analyses using “fast” stepwise-addition of taxa; only those values compatible with the 50% majority-rule consensus tree were recorded. For all small (i.e., 27-taxon) data sets except that of ITS, a finite number of shortest trees was obtained Can. J. Bot. Vol. 80, 2002 using 500 random-addition replicate searches and TBR branch swapping. Bootstrap values were calculated from 100 replicate analyses, simple-addition sequence of taxa, and TBR branch swapping. For the 27-taxon ITS matrix, the strategy used for the larger data sets was employed, with the exception that a maxtree limit of 500 trees was set for each of 100 bootstrap replicates. The number of additional steps required to force particular taxa into a monophyletic group was examined using the constraint option of PAUP*. In all maximum parsimony analyses, gap states were treated as either missing data or a fifth base (“new state”), or were excluded. Alignment gaps were incorporated into parsimony analyses by scoring each unambiguous insertion or deletion as a separate presence–absence (i.e., binary) character, while maintaining gap states as missing data. The resultant topology was compared to one inferred when alignment gaps were omitted as additional characters. The “Character Steps/ etc.” charting option of MacClade vers. 3.08 (Maddison and Maddison 1992), under the assumption of Fitch parsimony, was used to calculate the number of steps of each gap character across all maximally parsimonious topologies. For two Lomatium species, two compressed regions of undetermined length were reported in ITS-1 (Soltis and Kuzoff 1993). In several other Lomatium species (Hardig and Soltis 1999), runs of ambiguous bases or missing data are evident within the same regions, suggestive that compressions were a problem here too. Given the discrepancies in length between these sequences and those sequenced by us for other Lomatium taxa, gaps were not scored as additional characters in the analyses of ITS data. Morphology Characters of the fruit have been important traditionally in delimiting taxa within the group, but have yet to be analyzed cladistically across a wide spectrum of species. Thus, in the absence of a phylogenetic estimate, patterns in the evolution of these characters and their utility in circumscribing monophyletic groups could not be properly assessed. Moreover, we have observed that published fruit transections are occasionally interpreted incorrectly, because they were based on immature material or the species were misidentified. Microscope slides of mature fruit cross-sections were prepared for two or more populations of nearly all species of Cymopterus and several related genera (except Lomatium). Prior to sectioning, fruits were softened by treating them for several minutes in Pohl’s Softening Agent (Radford et al. 1974). Free-hand sections through the middle of the mature mericarps were made using a razor blade and preserved using Hoyer’s Mounting Medium (Radford et al. 1974). These sections were examined for orientation of fruit and seed compression, features of the ribs, wings, and commissure, and the number, position, and size of vittae. All microscope slides have been deposited at the Rocky Mountain Herbarium (RM). Our preparations of mature fruit cross-sections and an examination of abundant representative material stored at RM uncovered 25 qualitative morphological characters that are potentially parsimony informative (see Table 5). About half of these characters were obtained from the fruits, the remainder from plant habit, inflorescence, and flowers. For © 2002 NRC Canada Can. J. Bot. Downloaded from www.nrcresearchpress.com by Canadian Science Publishing on 06/08/15 For personal use only. Downie et al. each of 27 species, 30–50 herbarium specimens were examined from throughout its range; these species represented the same ones as used in the cladistic analyses of trnF-L-T data. Character polymorphisms were recorded. Because the number of states differed among characters (ranging from two to five), all characters were weighted in inverse proportion to their number of steps using the scale option of PAUP*, hence fractional weights were employed. Heuristic searches were conducted with 500 random-addition replicate searches and TBR branch swapping. All character states were assumed unordered, and the options multrees, collapse, and acctran optimization were selected. Bootstrap values were calculated from 500 replicate analyses, simple-addition sequence of taxa, and TBR branch swapping. The pattern of evolution of each morphological character across all minimal length trees was assessed using MacClade, with the goal of finding those characters most useful in the delimitation of major clades and genera. Results ITS Alignment of all 148 ITS-1 and ITS-2 sequences resulted in a matrix of 490 positions. Thirty-one positions from ITS-1 and 50 positions from ITS-2 were eliminated from subsequent analyses because of confounding interpretations of homology; these included several small autapomorphic insertions, as well as length mutations of varying sizes in two or more taxa. These 81 positions represented 29 excluded regions, with 25 of them only 1 or 2 bp in size. The largest excluded region, encompassing 33 positions in ITS-2 near gene 5.8S, was characterized by highly variable sequences in all members of tribes Scandiceae Spreng. and Oenantheae Dumort. Characteristics of the included positions are presented in Table 3. Both spacer regions contributed comparable numbers of informative characters to the phylogenetic analysis. Measures of pairwise sequence divergence across both ITS-1 and ITS-2 ranged from identity (between the two varieties of Cymopterus acaulis and among the three species of Thaspium) to 34.6% of nucleotides (between Daucus pusillus and Cynosciadium digitatum). Excluding all Lomatium species, 13 unambiguous alignment gaps were parsimony informative; these ranged from 1 to 4 bp in size. Numerous autapomorphic deletions of a single bp were prevalent throughout the alignment. The two largest length mutations, each of 14 bp in size, represent deletions in Myrrhidendron and Ligusticum scoticum ITS-1 sequences relative to outgroups Physospermum and Pleurospermum. No evidence of divergent paralogous rDNA copy types was found in any of the species investigated. Maximum parsimony analysis of 148 ITS-1 and ITS-2 sequences, with gap states treated as missing data, resulted in over 20 000 minimal length trees. The strict consensus of these trees, rooted with Physospermum and Pleurospermum, is presented in Fig. 1A. Those tribes and major clades outlined previously in subfamily Apioideae (Downie et al. 2001) are maintained, whereas resolution of relationships among Cymopterus sensu lato (including Oreoxis, Pseudocymopterus, and Pteryxia) and allies is quite poor, with only a few clades supported strongly in a large polytomy. In accordance with previous studies, the Arracacia group is 1309 monophyletic but supported weakly (with < 50% bootstrap value); its sister group is not realized. The genera Arracacia, Coulterophytum, and Prionosciadium are each not monophyletic. The same can be said for western NA genera Aletes, Cymopterus, Lomatium, Musineon, Oreoxis, and Pteryxia, and several other genera represented by more than one species. In contrast, Thaspium and Zizia are each monophyletic. Given the large polytomy with many weakly supported clades, such an analysis is unsatisfactory in resolving relationships among Cymopterus and its allies. However, this global analysis does suggest that Aethusa cynapium or members of tribe Coriandreae W.D.J. Koch may be appropriate outgroups for further local analyses of Cymopterus and relatives. Upon reduction of the global ITS matrix to include only those members comprising the large polytomy (except those of the Arracacia clade, which were also removed to facilitate analysis), a heuristic search was repeated using Aethusa cynapium as a functional outgroup. Alignment of 85 ITS sequences resulted in a matrix of 454 positions, with none excluded (Table 3). Maximum pairwise sequence divergence estimates approached 10.5% over both spacers, and 117 positions were parsimony informative (representing an increase of six positions relative to the same subset of taxa in the global analysis). Parsimony analysis of these sequence data, with gap states treated as missing, resulted in 20 000 minimal length trees whose strict consensus is presented in Fig. 1B. Once more, little resolution of relationships is achieved, with the results highly comparable to those obtained by the global analysis. However, in the local analysis, six of seven species of Angelica (plus Sphenosciadium) arise as a clade sister to all other ingroup taxa; Angelica sylvestris, however, is basal to this group. Rooting the trees with either Bifora radians or Coriandrum sativum (both of tribe Coriandreae) resulted in trees (not shown) consistent to those when Aethusa cynapium is used to root the network, the only exception being that all seven species of Angelica (plus Sphenosciadium) formed a monophyletic group arising from a large, basal polytomy. Constraining the 17 species (18 taxa) of Cymopterus to monophyly resulted in trees 21 steps longer than those most parsimonious. A monophyletic Lomatium resulted in trees 30 steps longer, whereas constraining the six narrowly endemic species comprising the Euryptera group of Lomatium (Lomatium howellii, Lomatium lucidum, Lomatium parvifolium, Lomatium repostum, Lomatium rigidum, and Lomatium shevockii) to monophyly, one of the few natural assemblages within the genus (Hardig and Soltis 1999), resulted in trees five steps longer than those most parsimonious. Repeating the local analysis with gap states treated as a fifth base (“new state” in PAUP*) resulted in 20 000 minimal length trees, each of 621 steps. Their strict consensus is presented in Fig. 2A. Differences from the previous analyses include an increased resolution of relationship (albeit with many clades still supported weakly), the placements of Pseudocymopterus montanus and the clade of Spermolepis and Ciclospermum as successive sister groups to a large polytomous clade of NA Apiaceae, the union of Thaspium, Zizia, and Polytaenia, and a monophyletic Euryptera species group (with 83% bootstrap support). Excluding gapped positions from the analysis resulted in a strict consensus tree of identical topol© 2002 NRC Canada 1310 Can. J. Bot. Vol. 80, 2002 Table 3. Sequence characteristics of the nuclear rDNA ITS and cpDNA rps16 intron and 3′ exon regions, separately and combined, used in the phylogenetic analyses of Apiaceae subfamily Apioideae. Can. J. Bot. Downloaded from www.nrcresearchpress.com by Canadian Science Publishing on 06/08/15 For personal use only. Region Length variation (bp) No. of aligned positions No. of positions eliminated Nuclear rDNA ITS Global analysis (n = 148) ITS-1 202–221 244 31 ITS-2 207–231 246 50 ITS-1 and ITS-2 417–443 490 81 Local analysis (n = 85) ITS-1 207–218 224 0 ITS-2 221–226 230 0 ITS-1 and ITS-2 427–441 454 0 Chloroplast DNA rps16 intron and 3′ exon (n = 85) Intron Domain I 476–506 575 118 Domain II 67–108 123 64 Domain III 64–76 78 23 Domain IV 85–153 192 82 Domain V 34–34 34 0 Domain VI 35–35 35 1 Entire intron 801–884 1059 288 3′ exon (partial) 110–110 110 1 Intron and 3′ exon 911–994 1169 289 Combined ITS and rps16 intron and 3′ exon (n = 83) Entire matrix 1349–1432 1645 356 No. of unambiguous gaps informative Maximum pairwise sequence divergence (%) No. of positions constant No. of positions informative No. of positions autapomorphic 35 29 64 151 143 294 27 24 51 40.7 36.4 34.6 108 122 230 65 52 117 51 56 107 12.1 10.9 10.5 313 32 39 57 32 27 518 99 617 68 15 8 25 2 3 123 5 128 76 12 8 28 0 4 130 5 135 12 3 1 3 0 0 19 0 19 7.9 11.9 12.7 15.5 5.9 5.9 6.6 5.5 6.2 703 377 209 19 13.9 Note: Alignment gaps were not scored as additional characters in the analyses of ITS data. ogy to that produced when gap states were treated as missing. The lack of resolution in the ITS trees derived from maximum parsimony precludes unambiguous hypotheses of relationship, but does show clearly that many NA genera, where resolved, are not monophyletic. The two largest genera within the complex, Cymopterus and Lomatium, are each highly polyphyletic. Based on the hierarchical likelihood ratio test statistic, Modeltest selected the TrN + G model (Tamura and Nei 1993) as fitting the ITS data best (base frequencies: 0.2501, A; 0.2416, C; 0.2474, G; 0.2609, T; estimates of substitution rates: A:C, 1; A:G, 2.2111; A:T, 1; C:G, 1; C:T, 4.7219; G:T, 1; proportion of invariable sites = 0; gamma distribution shape parameter = 0.5188). Using these parameters, a single most-likely tree was recovered in PAUP*, with a –Ln likelihood score of 3968.51683; this tree is presented in Fig. 2B. The relationships suggested by this phylogram include the monophyly of all Angelica species (plus Sphenosciadium), and the position of this clade, along with that comprising Spermolepis and Ciclospermum, as sister to all NA endemic species. The major clades inferred are similar to those presented by the maximum parsimony tree when gaps are treated as new character states (Fig. 2A), with no resolution among them. Many genera are not monophyletic (Aletes, Musineon, Oreoxis, Pteryxia, and Tauschia), and Cymopterus and Lomatium are grossly polyphyletic. Members of the Euryptera species group are closely allied, but do not form a clade. Rps16 intron Among the 85 species examined for rps16 intron sequence variation, the length of the intron varied from 801 to 884 bp. Juxtaposed was 110 bp of sequence from the 3′ exon. Alignment of these intron and flanking exon data resulted in a matrix of 1169 positions, of which 289 were excluded from subsequent analyses because of alignment ambiguities. These ambiguous regions ranged from 1 to 43 bp in size, averaging 11 positions each. Characteristics of all unambiguous positions, including the number of constant, autapomorphic, and parsimony informative sites, are Fig. 1A. Strict consensus of 20 000 minimal length 2242-step trees derived from equally weighted maximum parsimony analysis of 148 taxa and 409 unambiguously aligned ITS-1 and ITS-2 nucleotide positions, with gap states treated as missing data (CIs = 0.3189 and 0.2979, with and without uninformative characters, respectively; RI = 0.6335; RC = 0.2020). Numbers at nodes are bootstrap estimates for 100 000 replicate analyses using “fast” stepwise-addition; values ≤ 50% are not indicated. Brackets indicate major clades of Apioideae (Downie et al. 2001). Fig. 1B. Strict consensus of 20 000 minimal length 567-step trees derived from equally weighted maximum parsimony analysis of 85 nuclear rDNA ITS-1 and ITS-2 sequences, with gap states treated as missing data (CIs = 0.5485 and 0.4298, with and without uninformative characters, respectively; RI = 0.6139; RC = 0.3367). Numbers at nodes are bootstrap estimates for 100 000 replicate analyses using “fast” stepwise-addition; values ≤ 50% are not indicated. Complete taxon names, including ranks of infraspecific taxa which have been omitted for brevity, are provided in Table 2. © 2002 NRC Canada Can. J. Bot. Downloaded from www.nrcresearchpress.com by Canadian Science Publishing on 06/08/15 For personal use only. Downie et al. 1311 © 2002 NRC Canada 1312 Can. J. Bot. Vol. 80, 2002 Can. J. Bot. Downloaded from www.nrcresearchpress.com by Canadian Science Publishing on 06/08/15 For personal use only. Fig. 2A. Strict consensus tree of 20 000 minimal length 621-step trees derived from equally weighted maximum parsimony analysis of 85 nuclear rDNA ITS-1 and ITS-2 sequences, with gap states treated as a fifth base (“new character state”; CIs = 0.5588 and 0.4408, with and without uninformative characters, respectively; RI = 0.6432; RC = 0.3594). Numbers at nodes are bootstrap estimates for 100 000 replicate analyses using “fast” stepwise-addition; values ≤ 50% are not indicated. Fig. 2B. Most-likely tree derived from maximum likelihood analysis of ITS sequence data, based on the TrN + G model of nucleotide substitution (–Ln likelihood = 3968.51683). Complete taxon names are provided in Table 2. presented in Table 3. Nineteen unambiguous alignment gaps were potentially parsimony informative, ranging from 1 to 11 bp in size. Pairwise sequence divergence ranged from identity (between three pairs of sequences) to 6.2% of nucleotides (between the Lomatium bicolor – Lomatium californicum pair and Scandix). The secondary structure of the rps16 intron, like other plastid group II introns, is characterized by six major domains. For each domain and across all 85 species compared, features of the aligned sequences are presented in Table 3. Domain I is the largest, ranging between 476 and 506 bp in size, whereas domains V and VI are the smallest, each rang- ing between 34 and 35 bp in size. Domains V and VI are also the most conserved, with five informative positions collectively, low nucleotide sequence divergence, and no inferred gaps. These two small domains provide as much information to the phylogenetic analysis as does the 3′ exon portion. Relative to their size, domains II and IV provide the most phylogenetic information, with some 23 to 25% of all included positions parsimony informative. Maximum parsimony analysis of 880 unambiguously aligned rps16 intron and 3′ exon nucleotide positions plus 19 binary-scored informative gaps, with gap states treated as missing data, resulted in over 20 000 minimal length trees, © 2002 NRC Canada Can. J. Bot. Downloaded from www.nrcresearchpress.com by Canadian Science Publishing on 06/08/15 For personal use only. Downie et al. of which their strict consensus (with accompanying bootstrap values) is shown in Fig. 3A. Here, the positions of all alignment gap changes are indicated, with the 14 nonhomoplastic changes shown by solid circles and homoplasies (reversals and parallel gains of gaps 1 to 9 bp in size) by open circles. Repeating the analysis without the 19 scored gaps resulted in trees 27 steps shorter and collapse of the four branches indicated by asterisks in Fig. 3A. Treating gap states as a fifth base (“new state”) or excluding them altogether made no appreciable difference to the resultant consensus tree topologies. However, regardless of analysis, resolution of relationships among Cymopterus and its allies is poor. Moreover, the genera Heracleum and Pastinaca (the Heracleum clade) and Aethusa cynapium fall within a large, polytomous group. The four included species of Angelica and Sphenosciadium do not form a clade, but arise within the same lineage as two species of Lomatium, Aletes anisatus, Cymopterus globosus, and Shoshonea pulvinata. While this lineage is weakly supported (with or without scored gaps) it does suggest that Lomatium, Aletes, and Cymopterus may each not be monophyletic. Upon consideration of binary-scored gaps as additional characters, Tauschia and Pteryxia may also not be monophyletic. Pairwise sequence divergence estimates among the 58 species comprising the large polytomy barely exceed 3.0% of nucleotides. Within this same group, the number of characters potentially informative for parsimony analysis is 47 (excluding scored gaps), and their distribution is inadequate to resolve more than only a few clades. Clearly, these intron and exon data are insufficient by themselves to resolve relationships among the perennial, endemic, apioid umbellifers of western NA. However, basal resolutions in the tree are generally strongly supported, with the major clades identified similar to those resolved in the global analysis of ITS data (Fig. 1A). ITS and rps16 intron combined ITS and rps16 intron data for the same set of 83 taxa were combined for simultaneous consideration, as separate analyses of these data failed to provide adequate resolution among Cymopterus and its allies. Given the lack of resolution and poorly supported nodes in the rps16 intron-derived trees, a test of incongruence was considered unnecessary. Details of the alignment are provided in Table 3, with 377 characters potentially informative. Maximum parsimony analysis of both data partitions including the 19 informative intron gaps, with gap states treated as missing data, resulted in 20 000 minimal length trees; their strict consensus is presented in Fig. 3B. Identical results were obtained when gap states were treated as a fifth base; slightly less resolution among the western NA endemics was achieved when gap positions were excluded from the analysis. The consensus tree (Fig. 3B) shows more resolution than either of the separate analyses and, in general, greater bootstrap support for many clades. Nevertheless, a large polytomy of western NA taxa is maintained. Greater resolution is achieved, but many genera are still not monophyletic. Cymopterus (with 13 included species in the combined analysis) is highly polyphyletic, comprising 10 separate lineages occurring in all major branches of the polytomy. Lomatium (12 species) is also polyphyletic, but in this case more than half of its species 1313 are found in just one major branch. Constraining Cymopterus to monophyly resulted in trees 24 steps longer than those produced without the constraint; trees of similar length resulted when Lomatium was constrained as monophyletic. Aletes, Musineon, Oreoxis, Pteryxia, and Tauschia are also each not monophyletic. Angelica and Sphenosciadium form a clade alongside the same five species as in the separate analysis of intron data. Other noteworthy results include the close association of Zizia, Thaspium, and Polytaenia; the sister relationship between Aethusa cynapium and the large polytomous clade of western NA taxa; and the highly resolved and strongly supported relationships among the basal elements of the phylogeny. trnF-trnL-trnT Length variation of the entire trnF-L-T region for the 27 species studied ranged from 1693 to 1816 bp, and alignment of these data resulted in a matrix of 1884 positions. Forty-six positions, representing tracts of poly-A’s or insertions of dubious homology in two or more taxa, were eliminated. Characteristics of these unambiguously aligned data, including partitions representing the two intergenic spacers and the trnL intron, are presented in Table 4. No length variation was exhibited by the trnL exons (50 and 35 bp for the 3′ and 5′ exons, respectively) and from gene portions trnF and trnT (39 and 17 bp, respectively). The trnL intron ranged in size from 456 to 508 bp. The proportion of nucleotide differences ranged from identity to 2.5% for the trnF-L spacer and from identity to 2.9% for the trnL-T spacer; however, the trnL intron, intermediate in size between the two intergenic spacers, was more conserved, with a maximum pairwise sequence divergence of 1.6%. Overall, 42 positions were potentially informative, with over half of these coming from the trnL-T spacer. Twenty-seven gaps, ranging in length from 1 to 46 bp, were required to facilitate alignment; these represented 15 insertions (1–17 bp) and 12 deletions (2–46 bp) relative to the Angelica archangelica sequence. Seven of these gaps were potentially informative for parsimony analysis (size range 2–40 bp; representing two insertions and five deletions). Maximum parsimony analysis of the entire trnF-L-T region, with gap states treated as missing, resulted in 220 minimal length trees, each of 145 steps (see Table 4 for measures of character fit); the strict consensus of these trees is shown in Fig. 4A. Angelica archangelica was used to root these trees, as suggested by the local analyses of ITS data (Figs. 1B, 2A, and 2B). The strict consensus reveals three major clades, two of which are largely unresolved. One major clade (with 91% bootstrap support) consists exclusively of Lomatium species. Cymopterus and Pteryxia are each divided between the two remaining major clades. Additionally, Harbouria, Musineon, Thaspium, and Zizia are placed in one clade (with 73% bootstrap support), and Aletes, Lomatium, Neoparrya, Podistera, and Shoshonea occur in the other (with 82% bootstrap support). The latter is sister to the Lomatium clade. Cymopterus, Lomatium, and Pteryxia are each not monophyletic, and the results are equivocal in establishing the monophyly of Musineon and Zizia. Phylogenetic analyses of the three trnF-L-T data partitions yielded trees (not shown) highly consistent with respect to their major groups, and results of a partition © 2002 NRC Canada Can. J. Bot. Downloaded from www.nrcresearchpress.com by Canadian Science Publishing on 06/08/15 For personal use only. 1314 Can. J. Bot. Vol. 80, 2002 Fig. 3A. Strict consensus of 20 000 minimal length 474-step trees derived from equally weighted maximum parsimony analysis of 85 cpDNA rps16 intron and 3′ exon sequences plus 19 binary-scored alignment gaps, with gap states treated as missing data (CIs = 0.6983 and 0.5600, with and without uninformative characters, respectively; RI = 0.8281; RC = 0.5783). Numbers at nodes are bootstrap estimates for 100 000 replicate analyses using “fast” stepwise-addition; values ≤ 50% are not indicated. The positions of all state changes for the 19 informative gaps are indicated: solid circles indicate nonhomoplastic changes; open circles indicate homoplastic changes. Asterisks indicate branches that collapse when the 19 informative gaps are excluded and the analysis rerun (tree length = 447 steps; CIs = 0.6980 and 0.5470, with and without uninformative characters, respectively; RI = 0.8183; RC = 0.5712). Fig. 3B. Strict consensus of 20 000 minimal length 1907-step trees derived from equally weighted maximum parsimony analysis of combined ITS and rps16 intron and 3′ exon data (1289 unambiguously aligned positions and 19 informative gaps) for 83 taxa (CIs = 0.4803 and 0.4059, with and without uninformative characters, respectively; RI = 0.6875; RC = 3302). Treating gap states as either missing data or a fifth base resulted in identical topologies. Numbers at nodes are bootstrap estimates for 100 000 replicate analyses using “fast” stepwise-addition; values ≤ 50% are not indicated. Brackets indicate major clades of Apioideae (Downie et al. 2001). Complete taxon names are provided in Table 2. © 2002 NRC Canada Downie et al. 1315 Table 4. Characteristics of the trnF-L-T, ITS, and rps16 intron regions, separately and combined, used in the maximum parsimony analyses of 27 species of western North American Apiaceae subfamily Apioideae. Can. J. Bot. Downloaded from www.nrcresearchpress.com by Canadian Science Publishing on 06/08/15 For personal use only. trnF-L-T No. of total characters Length variation (bp) No. of eliminated characters No. of constant characters No. of autapomorphic characters No. of informative characters % informative charactersa % divergence (range) No. of unambiguous gaps No. of unambiguous gaps parsimony informative No. of minimal length trees Length of shortest trees Consistency indexb Retention index Rescaled consistency index ITS rps16 intron Combined (trnF-L-T, ITS, and rps16 intron) Entire region trnF-L intergenic spacer trnL intron trnL-T intergenic spacer 1884 1693–1816 46 1721 75 42 2.3 0.1–1.7 27 7 382 330–368 15 343 18 6 1.6 0–2.5 8 4 511 456–508 0 487 15 9 1.8 0–1.6 6 0 850 733–809 31 753 39 27 3.3 0–2.9 13 3 444 437–439 7 321 74 42 9.6 0.2–6.7 6 3 997 911–976 20 902 41 34 3.5 0.2–3.2 13 4 3325 3096–3229 73 2944 190 118 3.6 0.3–2.6 46 14 220 145 0.6769 0.8727 0.7463 4 28 0.8571 0.9565 0.9224 27 27 0.7500 0.9000 0.8000 30 81 0.7250 0.9018 0.7793 > 20 000 196 0.4955 0.5912 0.4223 6 94 0.6604 0.8583 0.6939 16 478 0.4926 0.6783 0.4825 a No. of informative characters / (no. of total characters – no. of eliminated characters). Excluding uninformative characters. b homogeneity test showed that these data sets do not yield significantly different phylogenetic estimates. Greatest resolution of relationships was obtained with trnL-T data, given its higher number of informative characters. Poorest resolution was achieved using trnL intron data, with only four clades resolved within a large, basal polytomy. Comparative analysis of molecular data The trnF-L-T results were compared to those obtained using ITS and rps16 intron data for the same set of 27 taxa (Table 4). The proportion of nucleotide differences in the ITS partition was two to three times higher than either the trnF-L-T or intron partition, and relative to its size the ITS region contributed the greatest percentage of informative characters to the analysis (9.6%). The strict consensus trees resulting from separate analyses of these molecular data are shown in Figs. 4A–4C. Phylogenetic resolution within the ITS tree (Fig. 4B) is poor, with all clades but one (Zizia aptera + Zizia aurea, with 96% bootstrap value) supported weakly. Somewhat better resolution is achieved in the rps16 intron tree (Fig. 4C), but several of its basal branches are also very weakly supported. The trnF-L-T tree (Fig. 4A) offers a comparable level of resolution to that of intron tree, but with generally higher bootstrap values. Despite a four-fold greater size, the entire trnF-L-T region yielded almost exactly the same numbers of autapomorphic and potentially informative characters as did the ITS region, yet the ITS tree was less resolved, possessed lower consistency index (CI), retention index (RI), and rescaled consistency (RC) values, and showed lower bootstrap support overall than did those trees derived from trnF-L-T (or intron) data. Visual inspection of the trees derived from these three data partitions indicates discordance among them, largely attributable to poorly supported nodes, and results of a parti- tion homogeneity test indicate significant incongruence. However, by collapsing those branches with bootstrap values < 80%, the trees become highly consistent with respect to their major groupings. Some disagreement persists between the trnF-L-T and rps16 intron trees, but since these loci are both found on the chloroplast genome and are inherited as a single linkage group, the differences seen are not likely the result of, for example, hybridization and (or) introgression, but rather weaknesses of the data themselves. Maximum parsimony analysis of the combined data (using 118 potentially informative characters and treating gap states as missing) resulted in 16 minimal length trees, each of 478 steps; their strict consensus, with accompanying bootstrap support, is shown in Fig. 4D. Bootstrap estimates ranged between 22 and 100%, with 6 of 19 nodes supported by values > 80%. A similar tree was obtained (not shown) when 14 informative gaps were included, with only one node collapsing upon addition of these gap data. Seven gaps are synapomorphic, and support relationships based on nucleotide substitutions alone. The remaining gaps each required two to three steps to explain their distribution across all minimal length trees, as determined by MacClade. The inclusion of gaps did little to increase resolution or bolster bootstrap support. A single, arbitrarily selected, maximally parsimonious tree (Fig. 4E) illustrates that most character state changes occur at the tips of the branches, with many internal branches of short length. Repeating the analysis with gap states treated as a fifth base resulted in a consensus tree of similar topology to that when gaps were treated as missing, with only minor shuffling of the most distal branches in the tree. Compared to the results of the partitioned analyses, consideration of combined molecular data yielded a strict consensus tree of greatest resolution. Three major clades are recognized, in which the composition of each is identical to © 2002 NRC Canada 1316 Can. J. Bot. Vol. 80, 2002 Can. J. Bot. Downloaded from www.nrcresearchpress.com by Canadian Science Publishing on 06/08/15 For personal use only. Fig. 4. Trees resulting from equally-weighted maximum parsimony analyses of (A) trnF-L-T, (B) ITS, (C) rps16 intron, (D and E) combined molecular, and (F) morphological data sets for 27 members of Apiaceae subfamily Apioideae. Trees A–D and F represent strict consensus trees; tree E represents one of 16 minimal length trees. Measures of character fit for the molecular data sets are presented in Table 4; those for the morphological data set are presented in the text. Complete taxon names are provided in Table 2. Bootstrap values are indicated at the nodes. The asterisk in tree D indicates the one branch that collapses when the 14 scored gaps are included in the analysis (tree length = 502 steps; CIs = 0.7052 and 0.5000, with and without uninformative characters, respectively; RI = 0.6831; RC = 0.4817). Bootstrap values, for analyses without and with scored informative gaps, are presented above and below the branches, respectively. © 2002 NRC Canada Downie et al. Can. J. Bot. Downloaded from www.nrcresearchpress.com by Canadian Science Publishing on 06/08/15 For personal use only. that inferred by the separate analysis of trnF-L-T data. Once more, Cymopterus, Lomatium, Musineon, and Pteryxia are each not monophyletic. Cymopterus is highly polyphyletic, and constraining its six species to monophyly resulted in trees 25 steps longer (excluding scored gaps) than those without the constraint. Forcing each of Lomatium (eight species), Musineon (two species), and Pteryxia (two species) to monophyly resulted in trees seven, four, and 20 steps longer, respectively. Morphology The characters and states considered in the cladistic analysis of morphological data are presented in Table 5; the data matrix is presented in Table 6. Cladistic analysis of 25 morphological characters, using fractional weights, revealed 19 most parsimonious trees each of 49.91667 steps (CI = 0.5008; RI = 0.6843; RC = 0.3427). The strict consensus of these trees (Fig. 4F) shows much resolution, but only two clades, Zizia aptera + Zizia aurea and Zizia + Thaspium, are well supported, with bootstrap values ≥ 86%. Of the five genera represented by at least two species, only Zizia is monophyletic. Cymopterus is monophyletic upon the exclusion of Cymopterus williamsii; this species is unique among the six members of the genus examined in that its dorsal mericarp ribs are prominent, rounded, and corky rather than winged and its fruits are terete (to subterete) in outline rather than compressed dorsally. Lomatium is paraphyletic, albeit with very weak bootstrap support. Constraining Lomatium to monophyly requires trees of 50.16667 steps, whereas Cymopterus, Musineon, and Pteryxia are each monophyletic at 50.41667 steps. In trees of 51.6667 steps, all genera occur as monophyletic. The results of the analysis of morphological data yield trees that are not at all congruent to those achieved through separate or combined analysis of molecular data, nor are the relationships proposed in agreement with any historical or contemporary treatment of the group. Across all 19 minimal length trees, seven characters occur without homoplasy (Nos. 6, 7, 10, 13, 15, 22, and 24; CI = 1.00; Table 5); however, only two of these (Nos. 22, terete seed compression, and 24, constricted commissure) support clades consisting of three or more species: the clade of Neoparrya, Podistera, Musineon, Harbouria, Zizia, and Thaspium; and the clade consisting of only the first four of these genera. Filiform fruit ribs (No. 18; CI = 0.500) occur in all taxa from Neoparrya through Lomatium junceum, but are absent in Harbouria (where the ribs are instead obtuse and corky). The presence of sepals > 0.6 mm (No. 14; CI = 0.500) occurs in the Aletes anisatus – Pteryxia terebinthina clade, as well as in Neoparrya and Podistera. Dorsally compressed fruits (No. 21; CI = 0.500) bearing conspicuous dorsal and marginal wings (No. 16; CI = 0.400) and the absence of a carpophore (No. 23; CI = 0.400) are each homoplastic. Characters exhibiting the highest levels of homoplasy include the presence of a conspicuously sheathing leaf (No. 8; CI = 0.250), the occurrence and type of peduncle pubescence (No. 5; CI = 0.286), the presence of a pseudoscape and rosette of leaves (No. 4; CI = 0.333), and habit (No. 1; CI = 0.333). A prominent conical stylopodium is absent in all taxa but Podistera and the outgroup Angelica. No unique character supports the monophyly of Cymopterus, either with or without Cymopterus williamsii. 1317 Discussion Historical accounts of taxonomic confusion Torrey and Gray (1840) provided the first treatment of Cymopterus, recognizing a heterogeneous assemblage of eight species in four sections, with the names of three of these sections based on unpublished genera of Nuttall: Leptocnemia Nutt. ex Torr. & A. Gray, Phellopterus Nutt. ex Torr. & A. Gray, and Pteryxia Nutt. ex Torr. & A. Gray. These sections differed by subtleties in calyx teeth development, pericarp composition, the number of vittae in the commissure, and persistence of a carpophore. Coulter and Rose (1888), summarizing the accounts of the previous four decades in their Revision of North American Umbelliferae, recognized 13 species in Cymopterus and erected the genera Coloptera J.M. Coult. & Rose and Pseudocymopterus for plants either similar to or previously referable to Cymopterus. These genera were distinguished from Cymopterus by their strongly dorsally compressed fruits with broad, thick (and occasionally corky), lateral wings. Cymopterus was restricted to those plants with five generally broad, thin, and equal wings and fruits not at all dorsally flattened. Oreoxis, Podistera, and Phellopterus Benth. (= Glehnia), each containing species that had previously been described under Cymopterus, and the monotypic Aletes and Harbouria, were also listed in their revision. Lomatium, comprising species then referred to the Eurasiatic genus Peucedanum L., was separated from Cymopterus and allies by having fruits with narrowly winged or wingless dorsal ribs and broad, thin lateral wings. Twelve years later and during a time of much botanical exploration in NA, Coulter and Rose (1900) reduced Coloptera to synonymy under Cymopterus and transferred the NA species of Peucedanum to Lomatium. Eight species were recognized in Cymopterus, but its composition was vastly different from that they had circumscribed earlier. Many previously described Cymopterus species were instead placed in Aulospermum, Glehnia, Oreoxis, Phellopterus, Rhysopterus, Pteryxia, Podistera, and Pseudocymopterus. Characters such as the degree and direction of fruit compression, the shape of the endosperm, and features of the carpophore, vittae and mericarp ribs were again stressed, in addition to leaf habit. Great variation was evident in Cymopterus with regard to the development of its dorsal wings, with stark differences apparent even on the same plant. Subsequently, Jones (1908) reduced the genera Aulospermum, Oreoxis, Phellopterus, Rhysopterus, Pteryxia, and Pseudocymopterus to sectional ranks under Cymopterus, recognizing seven sections, 34 species, and 12 varieties within the genus. The composition of each section, however, was not always equivalent to its generic counterpart (for instance, species of Aulospermum were placed into three sections). Species bridging the sections were numerous, leading Jones to comment that it was futile to divide these species into separate genera. Mathias (1930) followed the treatment of Coulter and Rose (1900) in her monograph of Cymopterus and allies by treating Aulospermum, Glehnia, Oreoxis, Phellopterus, Rhysopterus, Pteryxia, and Pseudocymopterus as generically distinct. Characters distinguishing the genera included the orientation of fruit compression, the occurrence and relative © 2002 NRC Canada 1318 Can. J. Bot. Vol. 80, 2002 Can. J. Bot. Downloaded from www.nrcresearchpress.com by Canadian Science Publishing on 06/08/15 For personal use only. Table 5. Morphological characters and states used in the phylogenetic analysis of western North American Apiaceae subfamily Apioideae. Character no. Character States 1 2 3 Habit Habit Roots 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Scape Peduncle Peduncle Leaf margin Sheath Bracts Bractlets Bractlets Flower color Central flowers Calyx in fruit Style in fruit Fruit ribs Fruit wings Fruit ribs Fruit apex Fruit surface Fruit compression Seed compression Carpophore Commissure Stylopodium 0 = caulescent; 1 = subcaulescent to subacaulescent; 2 = acaulescent 0 = stems one-few, tufted; 1 = stems cushion-forming 0 = tap, slender; 1 = tap, thickened; 2 = tap, globose; 3 = tap, branched woody caudex; 4 = fibrous, fascicled 0 = no pseudoscape or rosette of leaves; 1 = pseudoscape and rosette present 0 = glabrous; 1 = generally pubescent; 2 = hirtellous/scabrous at summit 0 = not swollen at summit; 1 = swollen at summit 0 = variously toothed or entire; 1 = evenly serrate or dentate 0 = not or slightly ampliate; 1 = conspicuously sheathing 0 = present; 1 = absent 0 = present; 1 = absent 0 = herbaceous; 1 = herbaceous with thin scarious margins; 2 = mostly scarious 0 = white; 1 = purplish; 2 = yellow; 3 = greenish 0 = pedicellate; 1 = subsessile 0 = >0.6 mm; 1 = <0.6 mm 0 = more or less erect; 1 = widely spreading 0 = all ribs winged; 1 = lateral ribs winged only; 2 = none winged 0 = chartaceous; 1 = thick, corky 0 = filiform; 1 = rounded, corky 0 = normal; 1 = constricted 0 = glabrous; 1 = pubescent; 2 = granulose; 3 = scabrose 0 = dorsally compressed; 1 = terete; 2 = laterally compressed 0 = dorsally compressed; 1 = terete 0 = present; 1 = present, falling with mericarp; 2 = absent 0 = not constricted; 1 = constricted 0 = absent; 1 = present development of lateral and (or) dorsal wings, the shape of the wing in cross-section, and features of the involucre and involucel. She included nine species in Cymopterus (Table 1). Mathias and Constance (1944–1945) subsequently placed Phellopterus and Aulospermum into synonymy under Cymopterus, recognizing 32 species within the genus (Table 1). Pseudocymopterus was considered monotypic, with other species previously referable to this genus transferred to Cymopterus or Pteryxia. The lack of substantial distinguishing characters separating genera — for example, Pteryxia differs from Cymopterus mainly in its development of conspicuous calyx teeth — prompted Cronquist (1961) to expand the limits of Cymopterus to include Pteryxia and Pseudocymopterus, as well as other segregates included by Mathias and Constance (1944–1945). With the additional transfer of Oreoxis to Cymopterus, this system was maintained by Cronquist (1997) in his treatment of the group for Intermountain Flora (Table 1). The genus Aletes, as initially described (Coulter and Rose 1888), was characterized as having a single large vitta in the broad intervals between its filiform ribs, two vittae on the commissural side of the fruit, and a small one in each rib. Based on the presence of a single vitta in most of its fruit intervals, Theobald et al. (1963) transferred Pteryxia anisata (A. Gray) Mathias & Constance into Aletes. Weber expanded the concept of Aletes by permitting considerable variation in flower color, the number, size, and disposition of vittae, and the compression and development of the lateral and dorsal wings of the mericarps, such that species of Cymopterus, Lomatium, Pteryxia, and Neoparrya were all brought into the genus (Weber 1984; Weber and Wittmann 1992; Table 1). Emphasizing a similarity in acaulescent habit, Weber (1991) also placed Musineon tenuifolium in Aletes. Cronquist (1997) has reported that the distinction between some species of Aletes and Musineon is nothing more than the number of oil tubes in the intervals between the ribs (one in Aletes, and two or more in Musineon) and, as such, submerged Aletes into Musineon. Polyphyly of Cymopterus and their relationships among the endemic perennial genera of Apiaceae (north of Mexico) Contemporary treatments of Cymopterus include some 35–45 species (Kartesz 1994; Cronquist 1997), with no formally recognized infrageneric taxa. The results of phylogenetic analyses of molecular and morphological data indicate that Cymopterus, sensu Kartesz or Cronquist, is clearly polyphyletic, and in the molecular analyses, trees of much greater length than those most parsimonious are required to invoke monophyly of the genus. Moreover, no unique morphological synapomorphy supports the monophyly of Cymopterus, and the characters used traditionally to delimit the genus show overlapping patterns of variation with those of many other endemic, perennial apioid umbellifers of western NA. Cymopterus is inextricably linked with Aletes, Harbouria, Lomatium, Musineon, Neoparrya, Oreoxis, Orogenia, Podistera, Pseudocymopterus, Pteryxia, Shoshonea, and Tauschia. As such, the genera Lomatium, Musineon, and Pteryxia (and perhaps Aletes, Oreoxis, and © 2002 NRC Canada Downie et al. 1319 Can. J. Bot. Downloaded from www.nrcresearchpress.com by Canadian Science Publishing on 06/08/15 For personal use only. Table 6. Matrix of morphological data. Taxon Morphological characters Aletes anisatus Angelica archangelica Cymopterus globosus Cymopterus montanus Cymopterus multinervatus Cymopterus nivalis Cymopterus purpureus Cymopterus williamsii Harbouria trachypleura Lomatium bicolor Lomatium californicum Lomatium dasycarpum Lomatium graveolens Lomatium junceum Lomatium juniperinum Lomatium macrocarpum Lomatium nudicaule Musineon divaricatum Musineon tenuifolium Neoparrya lithophila Podistera eastwoodiae Pteryxia hendersonii Pteryxia terebinthina Shoshonea pulvinata Thaspium trifoliatum Zizia aptera Zizia aurea 1–5 2 0000 0 0102 {12}0110 {12}0112 {12}0110 2 0300 {12}0110 2 0300 0 0302 1 0202 0 0100 {12}0001 2 0300 2 1310 {12}0001 1 0101 2 0100 0 0111 2 0302 2 0300 2 0100 2 0300 {01}0000 2 1300 0 0402 0 0402 0 0402 6–10 00010 00110 00010 00000 00000 00010 00010 00010 00010 00010 10111 00110 00010 00110 00110 00010 10011 00010 00010 00010 00010 00010 00110 00010 01010 01000 01000 11–15 0 2001 0 0011 1 0011 2{01}011 2 1011 1 0011 1{12}011 1 2011 1 2011 1 2011 ? 2011 1 3011 1 2011 1 2011 1 2011 1{01}011 ? 2011 1 2011 1 2011 1 2001 0 3001 0 2001 1 2001 0 2001 1 1010 1 2110 1 2110 16–20 21–25 10100 11100 01?00 00?00 00?00 00?00 00?00 11100 2?102 10010 10000 10001 10000 10000 10000 10010 10000 2?012 2?012 2?010 2?010 00?00 00?00 2?103 00?00 2?000 2?000 00000 00001 00200 00000 00200 00000 00000 10200 21010 00000 00000 00000 00000 00000 00000 00000 00000 21010 21010 21010 21011 00000 00000 10100 11200 21000 21000 Note: Characters and states are described in Table 5. Question marks denote inapplicable data; polymorphisms are scored in parentheses. Tauschia as well) are each not monophyletic either. Affinity also extends to four other indigenous perennial genera of primarily central to eastern North American distribution (i.e., Polytaenia, Taenidia (including Pseudotaenidia; Cronquist 1982), Thaspium, and Zizia), and while it is evident that all of these NA genera are undoubtedly closely related, our results suggest an evolutionary history of the group much more complicated than previously considered. The species of Cymopterus examined herein permeate many of the clades resolved in the trees derived from molecular data, and thus further study of all of these taxa (which should include a re-evaluation of the generic limits of many) is necessary to properly circumscribe Cymopterus and to ascertain its phylogenetic position within the group. Monophyly of the endemic perennial apioid genera of NA The restricted distribution of many of our indigenous apioid genera to dry habitats in western NA, their shared life history and general habit, and overlapping patterns of fruit variation suggest that this group of umbellifers (with the addition of Polytaenia, Taenidia, Thaspium, and Zizia) is monophyletic. An obsolete stylopodium in all genera save Podistera (where the stylopodium is well-developed (conical), as it is in most other umbellifers (Mathias and Constance 1944–1945)) is a synapomorphy adding credence to this hypothesis. Further support for the monophyly of the group comes from the shared presence of a protogynous breeding system, atypical in a family where floral protandry prevails (Lindsey and Bell 1980; Lindsey 1982; Barrie and Schlessman 1987; Schlessman et al. 1990; Schlessman and Graceffa 2002). In contrast, the NA representatives of the perennial circumboreal genera Angelica, Seseli L., Selinum L., and Peucedanum all have a breeding system that is protandrous (Barrie and Schlessman 1987). Other differences between the protogynous and protandrous groups of NA Apioideae include flowering time, habitat preference, nectary morphology, and patterns of variation in sex expression (Schlessman and Barrie 2003). Many species of Cymopterus, Lomatium, and Pteryxia have also been reported as hosts for a morphologically distinct species group of larvae of the holarctic moth genus Depressaria, often with a single species of Depressaria seemingly restricted to a single (or rarely few) species of Cymopterus or Lomatium (Clarke 1952; Hodges 1974; Thompson 1983; McKenna 2000). In western NA, there has been a striking radiation of Apiaceae-feeding Depressaria, and the host specificity of these insects, coupled with the rich diversity of substituted coumarins occurring in the host plants serving a defensive role (Murray et al. 1982), might imply an association consistent with reciprocal coevolutionary interactions (McKenna 2000; Berenbaum 2001) and, thus, monophyly of each of their interacting groups. © 2002 NRC Canada Can. J. Bot. Downloaded from www.nrcresearchpress.com by Canadian Science Publishing on 06/08/15 For personal use only. 1320 However, in the present study, the assumption of monophyly of these indigenous NA genera is confounded by the phylogenetic placements of the perennial circumboreal genus Angelica (with included Sphenosciadium) and members of the meso-American Arracacia group. (Protogyny has also been reported for Myrrhidendron; Webb 1984.) All of these taxa, including the examined NA endemics, occur within the Angelica and Arracacia clades of the apioid superclade, the latter a heterogeneous assemblage of both New and Old World genera (Downie et al. 2001). Included within this assemblage are the circumboreal genera Seseli, Selinum, and Peucedanum, whose distribution in NA is restricted to the eastern U.S.A. (Kartesz 1994), and prior analyses of this superclade using ITS and cpDNA sequences, albeit with very limited sampling of the NA endemic species, show little resolution of relationships among these taxa (Downie and Katz-Downie 1996; Downie et al. 1996, 1998, 2000b; Plunkett et al. 1996). Cladistic analysis of cpDNA restriction sites, however, does provide weak support for a monophyletic group of NA apioid taxa, but with the inclusion of the Arracacia clade (Plunkett and Downie 1999). Confirmation of monophyly of the endemic perennial apioid genera of NA must therefore await further study that includes additional representation of these and other Old World genera of circumboreal distribution. The phylogenetic position of Spermolepis and Ciclospermum is not fully resolved. Both genera unite as a strongly supported clade, but only in the trees derived from analyses using maximum parsimony (with gap states treated as a fifth base; Fig. 2A) or maximum likelihood (Fig. 2B) do they fall outside of the large, polytomous clade of NA umbellifers. Spermolepis and Ciclospermum are taprooted annuals possessing threadlike to linear leaf segments and are widely distributed throughout the southern U.S.A., and other warm, temperate areas. Their placement away from the clade of perennial, endemic, apioid umbellifers is consistent with their unique life history and overall general habit. Their putative sister-group relationship to the latter, as seen in Fig. 2A, needs confirmation through further study with greater outgroup representation. Fruit characters as indicators of phylogeny Our examination of relevant herbarium material, our observations of mature fruit cross-sections of nearly all species of Cymopterus and many related genera (see also Hartman 1983, 1985), and the results of the cladistic analysis of morphological data presented herein confirm that characters of the fruit can be quite variable and, thus, poor indicators of phylogeny. As examples, both Cymopterus and Lomatium have well-developed lateral wings. In most Cymopterus species, one or more (and often all three) of the dorsal ribs bear wings, with the dorsal wings often narrower than the lateral ones, whereas in Lomatium, the dorsal ribs are generally filiform and wingless or occasionally very narrowly winged. However, some Cymopteri lack dorsal wings. In Cymopterus newberryi (S. Watson) M.E. Jones and Cymopterus megacephalus M.E. Jones, the one or two wings on the dorsal surface of each mericarp vary from nearly as large as the lateral ones to often narrower and irregularly developed, or are more frequently obsolete. Fruits with scarcely developed or obsolete dorsal wings are also seen in Cymopterus Can. J. Bot. Vol. 80, 2002 deserticola Brandegee, Cymopterus douglassii R.L. Hartm. & Constance, Cymopterus ripleyi Barneby, and Cymopterus williamsii, as well as in some species of Pteryxia (e.g., Pteryxia terebinthina and Pteryxia hendersonii) and Pseudocymopterus (e.g., Pseudocymopterus montanus), and all show similarities to fruits of typical Lomatium. Throughout most of its range, Cymopterus longipes has saliently winged fruits, but in populations from southwestern Wyoming and adjacent Utah the dorsal wings are reduced to narrow ridges (Hartman and Constance 1985). These latter populations have been referred to as Cymopterus lapidosus (M.E. Jones) M.E. Jones (Hartman 1986), and their fruits superficially resemble those of some species of Lomatium. In Cymopterus corrugatus M.E. Jones, the fruits have strongly corrugated narrow wings when young, but at maturity the ribs are merely raised and thickened, with or without an irregular vestige of a wing (Hitchcock and Cronquist 1961). Thus, the distinction among some species of Cymopterus, Lomatium, Pteryxia, and Pseudocymopterus based on characters of the fruit wing is subject to failure, and their differentiation is not improved upon consideration of other morphological data. The genus Pseudocymopterus has been described as “one of the most complex situations in the family” (Mathias 1930), reflecting its great morphological variability and uncertain generic position, with the only character separating it from most species of Cymopterus being the characteristic short stiff pubescence at the top of the peduncle (Cronquist 1997). Populations of Pseudocymopterus montanus (= Cymopterus lemmonii (J.M. Coult. & Rose) Dorn, Pseudocymopterus tidestromii J.M. Coult. & Rose) from higher elevations throughout much of Utah have fruits with wings that are equally well-developed. Conversely, populations elsewhere exhibit fruits with dorsal wings reduced to low ridges (New Mexico, Colorado, Wyoming) or are completely absent (Arizona; Hartman and Constance 1985). In other words, populations of Pseudocymopterus montanus from Utah are indistinguishable from Cymopterus and those from Arizona look like Lomatium (and, thus, have been described as Lomatium lemmonii (J.M. Coult. & Rose) J.M. Coult. & Rose). Additional study of this polymorphic genus is currently being carried out (Sun et al. 2000; F.-J. Sun, data not included). We observed that other characters of the fruit are also highly variable. While definite laterally or dorsally compressed fruits are readily distinguishable in Cymopterus, there are numerous intermediate stages such that “the interpretation [of orientation of fruit compression] depends on the individuals point of view” (Mathias 1930). Fruit cross-sections reveal a complex series, from fruits that are subterete to somewhat compressed laterally (e.g., Cymopterus davisii R.L. Hartm., Cymopterus douglassii, Cymopterus jonesii, Cymopterus longipes, Cymopterus nivalis, and Cymopterus panamintensis) to fruits that are markedly compressed dorsally (e.g., Cymopterus deserticola and Cymopterus newberryi). In Cymopterus, loss of the carpophore (through adnation of its halves to the mericarps) has been independently achieved several times. Its absence has been reported from nearly half the species in the genus (Hartman and Constance 1985; Cronquist 1997; Hartman 2000), and in our cladistic analysis of morphological data at © 2002 NRC Canada Can. J. Bot. Downloaded from www.nrcresearchpress.com by Canadian Science Publishing on 06/08/15 For personal use only. Downie et al. least two losses and one reversal must be postulated to explain the distribution of this character over the six species of Cymopterus examined. The number, size, and position of vittae may also be quite variable, sometimes even within a genus (Mathias 1930). Aletes is characterized by usually a single vitta in the interval between the ribs, whereas most other genera have between two and six vittae per interval lying in a uniform row around the seed. In Cymopterus, the number of vittae in the intervals varies from 3 to 5, but in some species there may be only one. However, in Neoparrya, the oil tubes are numerous and are scattered throughout the pericarp. Some or all species of Aletes, Musineon, Neoparrya, Oreoxis, Podistera, and Shoshonea have fruits that are either subterete or slightly compressed laterally. With the exception of Oreoxis, with its very thick, corky-winged ribs, the aforementioned genera all have fruits that are not obviously winged and instead have ribs that may or may not be well developed. As examples, Musineon and Shoshonea have conspicuously ribbed fruits, Aletes and Podistera have ribs that may be inconspicuous or prominent and variously corky-winged, whereas Neoparrya shows practically no development of ribs at all. The variation exhibited by fruit morphology and anatomy among these western NA umbellifers severely limits their utility in delimiting genera unambiguously. The repeated occurrences of dorsal flattening and wing formation in Cymopterus and its allies are undoubtedly adaptations for various modes of seed dispersal (Theobald 1971; Heywood 1986) and, therefore, are susceptible to convergence. Patterns of development leading to similar dorsal flattening and gross morphology of umbellifer fruits can also be quite different (Theobald 1971). In contrast, putative sister species which are otherwise indistinguishable (Taenidia integerrima and Taenidia montana (Mack.) Cronquist) can differ substantially in their orientation of fruit compression (Cronquist 1982). The number and arrangement of resin-filled vittae (containing active compounds that are toxic to insects; Berenbaum 1981) and the presence of thick, corky ribs may confer protection to the endosperm (Spalik et al. 2001a), and it is not unrealistic to presume that these characters too may be susceptible to homoplasy. Interspecific hybridization may also obscure generic limits, but such hybridization among NA umbellifers is rare (Mathias and Constance 1959; Brehm and French 1966; Schlessman 1984; Cronquist 1997), as it is in the family in general (Bell and Constance 1957; Heywood 1982). Postmating isolating mechanisms in Lomatium, Thaspium, and Zizia are strong (Lindsey 1982; Schlessman 1984), and polyploids are rarely found, with the few reported cases known for Oreoxis alpina, Pteryxia terebinthina, and some species of Lomatium (Bell and Constance 1957, 1960, 1966; Moore 1971; Crawford and Hartman 1972; Schlessman 1984). In summary, our study confirms that fruit characters are of limited value for delimiting taxa and estimating phylogenetic relationships in this group of western NA umbellifers. Such a conclusion is not surprising, given the common dissatisfaction among systematists in using these characters to circumscribe higher-level taxa within the family (e.g., Heywood 1971; Theobald 1971; Davis 1972; Cronquist 1982). Indeed, the results of numerous molecular systematic investigations provide very little support for all but a few suprageneric taxa 1321 erected on the basis of anatomical and morphological features of the mature fruit (summarized in Downie et al. 2001). In contrast, and unlike the results presented herein, fruit morphology may be quite useful at lower taxonomic levels. For example, in Apiaceae tribes Scandiceae and Oenantheae, whose members are also well represented in NA, the distribution of fruit characteristics is highly consistent with ITS-derived trees and cladistic analyses of both morphological and molecular data corroborate the monophyly of nearly every genus within these tribes (Spalik and Downie 2001; Spalik et al. 2001a; S. Downie, data not included). However, in other groups, such as the Angelica clade and the apioid superclade, many species-rich genera are polyphyletic (Downie et al. 2000b, 2000c; Spalik et al. 2001b). Additional study is required to define and delimit the various generic elements which have been confused with Cymopterus, and to circumscribe Cymopterus itself. Whether or not we will eventually find morphological synapomorphies delimiting each of these genera remains to be seen. Phylogenetic utility of molecular data Separate analyses of ITS, rps16 intron, and trnF-L-T sequences failed to resolve relationships among the perennial, endemic genera of NA Apiaceae. Several clades were delimited in each of these analyses, but were not always reproduced by the different data sets, nor were many supported strongly. Data from the ITS region were most variable and yielded trees with the least resolution and highest homoplasy. Differential resolution between the plastid-derived rps16 intron and trnF-L-T trees was apparent, largely attributable to poorly supported nodes. Greatest resolution of relationship was achieved by including all molecular data in a simultaneous analysis, yet divergence estimates were still low, approaching 2.6% of nucleotides, and very few nodes were supported by high bootstrap values. As additional molecular data become available, perhaps from a more rapidly-evolving locus, greater resolution of relationships may be achieved and regions of discordance, if any, more rigorously addressed. The limited ability of these nuclear and organellar sequences to resolve relationships among the western NA apioid umbellifers might also reflect a real biological phenomenon — the rapid evolutionary radiation of this lineage. Such a hypothesis is consistent with trees exhibiting short internal branches and (or) a large basal polytomy comprising several distinct lineages (Futuyma 1997). Many species of western NA umbellifers are narrowly distributed and have strict edaphic requirements (Mathias 1930), and all exhibit low levels of sequence divergence. This, coupled with the prevalent and pronounced intergradation of morphological characters making species and generic delimitation difficult, suggests a recent origin and rapid diversification of these genera. This pattern of rapid radiation has been proposed for Lomatium (Soltis et al. 1995; Hardig and Soltis 1999), and has been suggested for other genera of western NA distribution (e.g., Hershkovitz and Zimmer 2000). Given the intercalation of members of Lomatium among other western NA Apiaceae, this pattern may very well be prevalent for the entire group. However, to evaluate the hypothesis of recent, rapid diversification, it is necessary to compare the degree of © 2002 NRC Canada Can. J. Bot. Downloaded from www.nrcresearchpress.com by Canadian Science Publishing on 06/08/15 For personal use only. 1322 evolutionary divergence within a clade with the degree of divergence within its sister clade (Jensen 1990). Pending further study, information on a definitive sister group is lacking (although the data presented herein suggest that Angelica may be a likely candidate). Clearly, a more resolved phylogeny confirming sister group relationships is in order before hypotheses of evolutionary success can be tested. We are continuing our systematic investigation of the perennial endemic genera of western NA with the goal of uncovering morphological synapomorphies useful for generic determination. If such synapomorphies cannot be identified, we would have to accept that the task of reclassifying this group is to be accomplished on the basis of molecular evidence rather than on traditional taxonomic data. If future studies support the conclusions presented herein, and if further resolution of relationships can be achieved, radical changes to the prevailing classification of western NA Apioideae will be required. Acknowledgements The authors thank Christine Desfeux, Jonathan Luttrell, and Erica Rogers for assistance in the laboratory; Lincoln Constance and Tim Chumley for providing leaf material; and Mark Schlessman and two anonymous reviewers for comments on the manuscript. This work was supported by National Science Foundation grants DEB 9407712 and DEB 0089452. References Baldwin, B.G., Sanderson, M.J., Porter, J.M., Wojciechowski, M.F., Campbell, C.S., and Donoghue, M.J. 1995. The ITS region of nuclear ribosomal DNA: a valuable source of evidence on angiosperm phylogeny. Ann. Mo. Bot. Gard. 82: 247–277. Barrie, F.R., and Schlessman, M.A. 1987. The incidence of protogyny in the North American species of Apioideae (Apiaceae). Am. J. Bot. 74: 646. (Abstract). Bell, C.R., and Constance, L. 1957. Chromosome numbers in Umbelliferae. Am. J. Bot. 44: 565–572. Bell, C.R., and Constance, L. 1960. Chromosome numbers in Umbelliferae. II. Am. J. Bot. 47: 24–32. Bell, C.R., and Constance, L. 1966. Chromosome numbers in Umbelliferae. III. Am. J. Bot. 53: 512–520. Berenbaum, M.R. 1981. Evolution of specialization in insect–umbellifer associations. Annu. Rev. Entomol. 35: 319–343. Berenbaum, M.R. 2001. Chemical mediation of coevolution: phylogenetic evidence for Apiaceae and associates. Ann. Mo. Bot. Gard. 88: 45–59. Brehm, B.G., and French, D.H. 1966. Chemical evidence for genetic exchange in Lomatium (Umbelliferae). Am. J. Bot. 53: 637. (Abstract). Brummitt, R.K., and Powell, C.E. 1992. Authors of plant names. Royal Botanic Gardens, Kew, U.K. Catalán, M.P., Kellogg, E.A., and Olmstead, R.G. 1997. Phylogeny of Poaceae subfamily Pooideae based on chloroplast ndhF gene sequences. Mol. Phylogenet. Evol. 8: 150–166. Clarke, J.F.G. 1952. Host relationships of moths of the genera Depressaria and Agonopteryx, with descriptions of new species. Smithson. Misc. Collect. 117(7): 1–20. Coulter, J.M., and Rose, J.N. 1888. Revision of North American Umbelliferae. pp. 144. Wabash College, Crawfordsville, Ind. Can. J. Bot. Vol. 80, 2002 Coulter, J.M., and Rose, J.N. 1900. Monograph of the North American Umbelliferae. Contrib. U.S. Natl. Herb. 7: 1–256. Crawford, D.J., and Hartman, R.L. 1972. Chromosome numbers and taxonomic notes for Rocky Mountain Umbelliferae. Am. J. Bot. 59: 386–392. Cronquist, A. 1961. Umbelliferae. In Vascular plants of the Pacific Northwest. Edited by C.L. Hitchcock, A. Cronquist, M. Ownbey, and J.W. Thompson. Univ. Wash. Publ. Biol. 17(3): 506–586. Cronquist, A. 1982. Reduction of Pseudotaenidia to Taenidia (Apiaceae). Brittonia, 34: 365–367. Cronquist, A. 1997. Apiaceae. In Intermountain flora: vascular plants of the Intermountain West, U.S.A. Vol. 3, part A. Edited by A. Cronquist, N. H. Holmgren, and P. K. Holmgren. New York Botanical Garden, Bronx, N.Y. Davis, P.H. 1972. Umbelliferae. In Flora of Turkey and the East Aegean Islands. Vol. 4. University Press, Edinburgh. pp. 265–538. Downie, S.R., and Katz-Downie, D.S. 1996. A molecular phylogeny of Apiaceae subfamily Apioideae: evidence from nuclear ribosomal DNA internal transcribed spacer sequences. Am. J. Bot. 83: 234–251. Downie, S.R., and Katz-Downie, D.S. 1999. Phylogenetic analysis of chloroplast rps16 intron sequences reveals relationships within the woody southern African Apiaceae subfamily Apioideae. Can. J. Bot. 77: 1120–1135. Downie, S.R., Katz-Downie, D.S., and Cho, K.-J. 1996. Phylogenetic analysis of Apiaceae subfamily Apioideae using nucleotide sequences from the chloroplast rpoC1 intron. Mol. Phylogenet. Evol. 6: 1–18. Downie, S.R., Katz-Downie, D.S., and Spalik, K. 2000a. A phylogeny of Apiaceae tribe Scandiceae: evidence from nuclear ribosomal DNA internal transcribed spacer sequences. Am. J. Bot. 87: 76–95. Downie, S.R., Katz-Downie, D.S., and Watson, M.F. 2000b. A phylogeny of the flowering plant family Apiaceae based on chloroplast DNA rpl16 and rpoC1 intron sequences: towards a suprageneric classification of subfamily Apioideae. Am. J. Bot. 87: 273–292. Downie, S.R., Watson, M.F., Spalik, K., and Katz-Downie, D.S. 2000c. Molecular systematics of Old World Apioideae (Apiaceae): relationships among some members of tribe Peucedaneae sensu lato, the placement of several island-endemic species, and resolution within the apioid superclade. Can. J. Bot. 78: 506–528. Downie, S.R., Ramanath, S., Katz-Downie, D.S., and Llanas, E. 1998. Molecular systematics of Apiaceae subfamily Apioideae: phylogenetic analyses of nuclear ribosomal DNA internal transcribed spacer and plastid rpoC1 intron sequences. Am. J. Bot. 85: 563–591. Downie, S.R., Plunkett, G.M., Watson, M.F., Spalik, K., Katz-Downie, D.S., Valiejo-Roman, C.M., Terentieva, E.I., Troitsky, A.V., Lee, B.-Y., Lahham, J., and El-Oqlah, A. 2001. Tribes and clades within Apiaceae subfamily Apioideae: the contribution of molecular data. Edinb. J. Bot. 58: 301–330. Evert, E.F., and Constance, L. 1982. Shoshonea pulvinata, a new genus and species of Umbelliferae from Wyoming. Syst. Bot. 7: 471–475. Farris, J.S., Källersjö, M., Kluge, A.G., and Bult, C. 1995. Testing significance of incongruence. Cladistics, 10: 315–319. Felsenstein, J. 1981. Evolutionary trees from DNA sequences: a maximum likelihood approach. J. Mol. Evol. 17: 368–376. Felsenstein, J. 1985. Confidence limits on phylogenies: an approach using the bootstrap. Evolution, 39: 783–791. © 2002 NRC Canada Can. J. Bot. Downloaded from www.nrcresearchpress.com by Canadian Science Publishing on 06/08/15 For personal use only. Downie et al. Futuyma, D.J. 1997. Evolutionary biology, 3rd ed. Sinauer, Sunderland, Mass. Gielly, L., and Taberlet, P. 1994. The use of chloroplast DNA to resolve plant phylogenies: noncoding versus rbcL sequences. Mol. Biol. Evol. 11: 769–777. Gilmartin, A.J., and Simmons, K.S. 1987. Relationships of Lomatium among other genera of Apiaceae. Plant Syst. Evol. 157: 95–103. Hardig, T.M., and Soltis, P.S. 1999. An ITS-based phylogenetic analysis of the Euryptera species group in Lomatium (Apiaceae). Plant Syst. Evol. 219: 65–78. Hartman, R.L. 1983. Two novelties in the Umbelliferae from the central Rocky Mountains. Am. J. Bot. 70(6): 116. (Abstract). Hartman, R.L. 1985. A new species of Cymopterus (Umbelliferae) from southern Idaho. Brittonia, 37: 102–105. Hartman, R.L. 1986. Studies in Rocky Mountain species of Cymopterus (Apiaceae). Am. J. Bot. 73: 765–766. (Abstract). Hartman, R.L. 2000. A new species of Cymopterus (Apiaceae) from the Rocky Mountain region, U.S.A. Brittonia, 52: 136–141. Hartman, R.L., and Constance, L. 1985. Two new species of Cymopterus (Umbelliferae) from western North America. Brittonia, 37: 88–95. Hershkovitz, M.A., and Zimmer, E.A. 2000. Ribosomal DNA evidence and disjunctions of western American Portulacaceae. Mol. Phylogenet. Evol. 15: 419–439. Heywood, V.H. 1971. Systematic survey of Old World Umbelliferae. In The biology and chemistry of the Umbelliferae. Edited by V.H. Heywood. Academic Press, New York. pp. 31–42. Heywood, V.H. 1982. General introduction to the taxonomy of the Umbelliferae. In Les Ombellifères: contributions pluridisciplinaires à la systématique. Actes du 2ème Symposium International sur les Ombellifères. Edited by A.-M. Cauwet-Marc and J. Carbonnier. Braun-Brumfield, Inc., Ann Arbor, Mich. Monogr. Syst. Bot. Mo. Bot. Gard. No. 6. pp. 107–112. Heywood, V.H. 1986. The Umbelliferae: An impossible family? Acta Univ. Ups. Symb. Bot. Ups. 26: 173–180. Hitchcock, C.L., and Cronquist, A. 1961. Vascular plants of the Pacific Northwest, Part 3: Saxifragaceae to Ericaceae. Univ. Wash. Publ. Biol., Vol. 17. pp. 506–587. Hodges, R.W. 1974. Gelechoidea: Oecophoridae. In The moths of America north of Mexico. Edited by R.B. Dominick, D.C. Ferguson, J.G. Franclemont, R.W. Hodges, and E.G. Munroe. E. W. Classey Ltd., London. Holmgren, P.K., Holmgren, N.H., and Barnett, L.C. 1990. Index herbariorum. New York Botanical Garden, New York. Jensen, J.S. 1990. Plausibility and testability: assessing the consequences of evolutionary innovation. In Evolutionary novelties. Edited by M.N. Nitecki. University of Chicago Press, Chicago, Ill. pp. 171–190. Jones, M.E. 1908. New species and notes. Contrib. West. Bot. 12: 1–81. Kartesz, J.T. 1994. A synonymized checklist of the vascular flora of the United States, Canada, and Greenland, 2nd ed. Biota of the North American Program of the North Carolina Botanical Garden. Timber Press, Portland, Oreg. Katz-Downie, D.S., Valiejo-Roman, C.M., Terentieva, E.I., Troitsky, A.V., Pimenov, M.G., Lee, B., and Downie, S.R. 1999. Towards a molecular phylogeny of Apiaceae subfamily Apioideae: additional information from nuclear ribosomal DNA ITS sequences. Plant Syst. Evol. 216: 167–195. Lee, B.-Y., and Downie, S.R. 1999. A molecular phylogeny of Apiaceae tribe Caucalideae and related taxa: inferences based on ITS sequence data. Syst. Bot. 24: 461–479. 1323 Lee, B.-Y., and Downie, S.R. 2000. Phylogenetic analysis of cpDNA restriction sites and rps16 intron sequences reveals relationships among Apiaceae tribes Caucalideae, Scandiceae and related taxa. Plant Syst. Evol. 221: 35–60. Lidén, M., Fukuhara, T., Rylander, J., and Oxelman, B. 1997. Phylogeny and classification of Fumariaceae, with emphasis on Dicentra s. l., based on the plastid gene rps16 intron. Plant Syst. Evol. 206: 411–420. Lindsey, A.H. 1982. Floral phenology patterns and breeding systems in Thaspium and Zizia (Apiaceae). Syst. Bot. 7: 1–12. Lindsey, A.H., and Bell, C.R. 1980. Protogyny and associated reproductive characters in Apiaceae. Proceedings of the Second International Congress of Systematic and Evolutionary Biology, July 17–24, 1980, The University of British Columbia, Vancouver, B.C. p. 269. (Abstract). Maddison, W.P., and Maddison, D.R. 1992. MacClade: Analysis of phylogeny and character evolution, version 3 edition. Sinauer Associates, Sunderland, Mass. Mastrogiuseppe, J.D., Gill, S.J., Simmons, K.S., and Brown, G.K. 1985. Morphologic and cytotaxonomic evaluation of Lomatium tuberosum (Apiaceae). Brittonia, 37: 252–260. Mathias, M.E. 1930. Studies in the Umbelliferae. III. A monograph of Cymopterus including a critical study of related genera. Ann. Mo. Bot. Gard. 17: 213–476. Mathias, M.E. 1965. Distribution patterns of certain Umbelliferae. Ann. Mo. Bot. Gard. 52: 387–398. Mathias, M.E., and Constance, L. 1944–1945. Umbelliferae. In North American flora, Vol. 28B. The New York Botanical Garden, New York. pp. 43–295. Mathias, M.E., and Constance, L. 1959. New North American Umbelliferae—III. Bull. Torrey Bot. Club, 86: 374–382. McKenna, D.D. 2000. Ecology, phylogeny, and host-usage of North American Depressaria (Lepidoptera: Elachistidae). M.Sc. thesis, Department of Entomology, University of Illinois at Urbana-Champaign, Ill. Michel, F., Umesono, K., and Ozeki, H. 1989. Comparative and functional anatomy of group II catalytic introns—a review. Gene (Amst.), 82: 5–30. Moore, D.M. 1971. Chromosome studies in the Umbelliferae. In The biology and chemistry of the Umbelliferae. Edited by V.H. Heywood. Academic Press, New York. pp. 233–255. Murray, R.D.H., Méndez, J., and Brown, S.A. 1982. The natural coumarins: occurrence, chemistry and biochemistry. Wiley and Sons, New York. Neuhaus, H., Scholz, A., and Link, G. 1989. Structure and expression of a split chloroplast gene from mustard (Sinapsis alba): ribosomal protein gene rps16 reveals unusual transcriptional features and complex RNA maturation. Curr. Genet. 15: 63–70. Oxelman, B., Lidén, M., and Berglund, D. 1997. Chloroplast rps16 intron phylogeny of the tribe Sileneae (Caryophyllaceae). Plant Syst. Evol. 206: 393–410. Petersen, G., Seberg, O., and Larsen, S. 2002. The phylogenetic and taxonomic position of Lilaeopsis (Apiaceae), with notes on the applicability of ITS sequence data for phylogenetic reconstruction. Aust. Syst. Bot. 15: 181–191. Plunkett, G.M., and Downie, S.R. 1999. Major lineages within Apiaceae subfamily Apioideae: a comparison of chloroplast restriction site and DNA sequence data. Am. J. Bot. 86: 1014–1026. Plunkett, G.M., Soltis, D.E., and Soltis, P.S. 1996. Evolutionary patterns in Apiaceae: inferences based on matK sequence data. Syst. Bot. 21: 477–495. Posada, D., and Crandall, K.A. 1998. Modeltest: testing the model of DNA substitution. Bioinformatics, 14: 817–818. © 2002 NRC Canada Can. J. Bot. Downloaded from www.nrcresearchpress.com by Canadian Science Publishing on 06/08/15 For personal use only. 1324 Radford, A.E., Dickison, W.C., Massey, J.R., and Bell, C.R. 1974. Vascular plant systematics. Harper and Row, New York. Schlessman, M.A. 1984. Systematics of tuberous lomatiums (Umbelliferae). Sys. Bot. Monogr. 4: 55. Schlessman, M.A., and Barrie, F.R. 2003. Dichogamy and sex expression of North American apioid Apiaceae (Apiaceae, Apioideae). In Apiales 2003: 4th International Apiales Symposium, held in conjunction with the 29th Conference of the South African Association of Botanists, Pretoria, South Africa January 6-7, 2003. (Abstract). Schlessman, M.A., and Graceffa, L.M. 2002. Protogyny, pollination, and sex expression of andromonoecious Pseudocymopterus montanus (Apiaceae, Apioideae). Int. J. Plant Sci. 163: 409–417. Schlessman, M.A., Lloyd, D.G., and Lowry II, P.P. 1990. Evolution of sexual systems in New Caledonian Apiaceae. Mem. N. Y. Bot. Gard. 55: 105–117. Simmons, K. S. 1985. Systematic studies in Lomatium. Ph.D. thesis, Washington State University, Pullman, Wash. Soltis, P.S., and Kuzoff, R.K. 1993. ITS sequence variation within and among populations of Lomatium grayi and L. laevigatum (Umbelliferae). Mol. Phylogenet. Evol. 2: 166–170. Soltis, P.S., and Novak, S.J. 1997. Phylogeny of the tuberous Lomatiums (Apiaceae): cpDNA evidence for morphological convergence. Syst. Bot. 22: 99–112. Soltis, P.S., Novak, S.J., Hardig, T.M., and Soltis, D.E. 1995. Phylogenetic relationships and rapid radiation in Lomatium (Apiaceae). Am. J. Bot. 82 (6): 163 (Abstract). Spalik, K., and Downie, S.R. 2001. The utility of morphological characters for inferring phylogeny in Scandiceae subtribe Scandicinae (Apiaceae). Ann. Mo. Bot. Gard. 88: 270–301. Spalik, K., Wojewódzka, A., and Downie, S.R. 2001a. The evolution of fruit in Scandiceae subtribe Scandicinae (Apiaceae). Can. J. Bot. 79: 1358–1374. Spalik, K., Wojewódzka, A., and Downie, S.R. 2001b. Delimitation of genera in Apiaceae with examples from Scandiceae subtribe Scandicinae. Edinb. J. Bot. 58: 331–346. Can. J. Bot. Vol. 80, 2002 Sun, F.-J., Downie, S.R., and Hartman, R.L. 2000. A phylogenetic study of Cymopterus and related genera. Am. J. Bot. 87 (6): 161–162 (Abstract). Swofford, D.L. 1998. PAUP*. Phylogenetic analysis using parsimony (*and other methods), version 4 edition. Sinauer Associates, Sunderland, Mass. Taberlet, P., Gielly, L., Pautou, G., and Bouvet, J. 1991. Universal primers for amplification of three non-coding regions of chloroplast DNA. Plant Mol. Biol. 17: 1105–1109. Tamura, K., and Nei, M. 1993. Estimation of the number of nucleotide substitutions in the control region of mitochondrial DNA in humans and chimpanzees. Mol. Biol. Evol. 10: 512–526. Theobald, W.L. 1971. Comparative anatomical and developmental studies in the Umbelliferae. In The biology and chemistry of the Umbelliferae. Edited by V. H. Heywood. Academic Press, New York. pp. 177–197. Theobald, W.L., Tseng, C.C., and Mathias, M.E. 1963. A revision of Aletes and Neoparrya (Umbelliferae). Brittonia, 16: 296–315. Thompson, J.N. 1983. The use of ephemeral plant parts on small hostplants: How Depressaria leptotaeniae (Lepidoptera: Oecophoridae) feeds on Lomatium dissectum (Umbelliferae). J. Anim. Ecol. 52: 281–291. Torrey, J., and Gray, A. 1840. A flora of North America. Vol. 1. Hafner Publ. Co., New York. pp. 598–645. Webb, C.J. 1984. Pollination specialization and protogyny in Myrrhidendron donnellsmithii (Umbelliferae). Syst. Bot. 9: 240–246. Weber, W.A. 1984. New names and combinations, principally in the Rocky Mountain flora. IV. Phytologia, 55: 1–11. Weber, W.A. 1991. New names and combinations, principally in the Rocky Mountain flora. VIII. Phytologia, 70: 231–233. Weber, W.A., and Wittmann, R.C. 1992. Catalog of the Colorado flora: a biodiversity baseline. University Press of Colorado, Boulder, Colo. © 2002 NRC Canada Can. J. Bot. Downloaded from www.nrcresearchpress.com by Canadian Science Publishing on 06/08/15 For personal use only. This article has been cited by: 1. Jinxin Liu, Linchun Shi, Jianping Han, Geng Li, Heng Lu, Jingyi Hou, Xiaoteng Zhou, Fanyun Meng, Stephen R. Downie. 2014. Identification of species in the angiosperm family Apiaceae using DNA barcodes. Molecular Ecology Resources 14:10.1111/ men.2014.14.issue-6, 1231-1238. [CrossRef] 2. Vasu Dev, Wayne H. Whaley, Sarah R. Bailey, Eric Chea, Jeannie G. Dimaano, Dhara K. Jogani, Bill Ly, Dennis Eggett. 2010. Essential oil composition of nine Apiaceae species from western United States that attract the female Indra Swallowtail butterfly (Papilio indra). Biochemical Systematics and Ecology 38, 538-547. [CrossRef] 3. Na Sun, Xing-Jin He, Song-Dong Zhou. 2010. Epidermal Morphology of Ligusticum (Apiaceae) from China. Annales Botanici Fennici 47, 261-279. [CrossRef] 4. Feng-Jie Sun, Stephen R. Downie. 2010. Phylogenetic analyses of morphological and molecular data reveal major clades within the perennial, endemic western North American Apiaceae subfamily Apioideae 1. The Journal of the Torrey Botanical Society 137, 133-156. [CrossRef] 5. Philip S. Beauchamp, Eric Chea, Jeannie G. Dimaano, Vasu Dev, Bill Ly, A. E. Miranda, Wayne H. Whaley. 2009. Essential Oil Composition of Six Lomatium Species Attractive to Indra Swallowtail Butterfly ( Papilio indra ): Principal Component Analysis Against Essential Oil Composition of Lomatium dissectum var. multifidum. Journal of Essential Oil Research 21, 535-542. [CrossRef] 6. Tu Feng, Stephen R. Downie, Yan Yu, Xuemei Zhang, Weiwei Chen, Xingjin He, Shuang Liu. 2009. Molecular systematics of Angelica and allied genera (Apiaceae) from the Hengduan Mountains of China based on nrDNA ITS sequences: phylogenetic affinities and biogeographic implications. Journal of Plant Research 122, 403-414. [CrossRef] 7. Na Sun, Xing-Jin He, Song-Dong Zhou. 2008. Morphological cladistic analysis of Ligusticum (Umbelliferae) in China. Nordic Journal of Botany 26:10.1111/njb.2008.26.issue-1-2, 118-128. [CrossRef] 8. Renata Kurzyna-Młynik, Alexei A. Oskolski, Stephen R. Downie, Rafał Kopacz, Aneta Wojewódzka, Krzysztof Spalik. 2008. Phylogenetic position of the genus Ferula (Apiaceae) and its placement in tribe Scandiceae as inferred from nrDNA ITS sequence variation. Plant Systematics and Evolution 274, 47-66. [CrossRef] 9. Feng-Jie Sun, Geoffrey A. Levin, Stephen R. Downie. 2008. A multivariate analysis of Pteryxia terebinthina (Apiaceae) 1. The Journal of the Torrey Botanical Society 135, 81-93. [CrossRef] 10. Feng-Jie Sun, Geoffrey A. Levin, Stephen R. Downie. 2006. A multivariate analysis of Pseudocymopterus (Apiaceae) 1. The Journal of the Torrey Botanical Society 133, 499-512. [CrossRef] 11. Feng-Jie Sun, Geoffrey A. Levin, Stephen R. Downie. 2005. A multivariate analysis of Cymopterus glomeratus, formerly known as C. acaulis (Apiaceae). Rhodora 107, 359-385. [CrossRef] 12. Craig S. Smith, Lance M. McNees. 2005. Cymopterus bulbosus and Prehistoric Foragers: Patch Size, Plant Density, and Return Rates. Journal of Ethnobiology 25, 1-23. [CrossRef] 13. M.A. Schlessman, F.R. Barrie, B.-E. van Wyk, P.M. Tilney. 2004. Protogyny in Apiaceae, subfamily Apioideae: systematic and geographic distributions, associated traits, and evolutionary hypotheses. South African Journal of Botany 70, 475-487. [CrossRef] 14. T.M. Hardway, K. Spalik, M.F. Watson, D.S. Katz-Downie, S.R. Downie, B.-E. van Wyk, P.M. Tilney. 2004. Circumscription of Apiaceae tribe Oenantheae. South African Journal of Botany 70, 393-406. [CrossRef] 15. F.-J. Sun, S.R. Downie, B.-E. van Wyk, P.M. Tilney. 2004. A molecular systematic investigation of Cymopterus and its allies (Apiaceae) based on phylogenetic analyses of nuclear (ITS) and plastid (rps16 intron) DNA sequences. South African Journal of Botany 70, 407-416. [CrossRef] 16. MARK A. SCHLESSMAN, NORA, TIM WATKINS, LAURA M. GRACEFFA, DIANE CORDRAY. 2004. Functions of staminate flowers in andromonoecious Pseudocymopterus montanus (Apiaceae, Apioideae). Plant Species Biology 19:10.1111/ psb.2004.19.issue-1, 1-12. [CrossRef]

RELATED PAPERS

RELATED TOPICS

Log In

Polyphyly of the spring-parsleys ( Cymopterus ): molecular and morphological evidence suggests complex relationships among the perennial endemic genera of western North American Apiaceae

Polyphyly of the spring-parsleys ( Cymopterus ): molecular and morphological evidence suggests complex relationships among the perennial endemic genera of western North American Apiaceae

Related Papers

RELATED PAPERS

RELATED TOPICS