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]