Systematic Botany (2007), 32(3): pp. 671–681 # Copyright 2007 by the American Society of Plant Taxonomists

Inclusion of the South Pacific Alpine Genus Oreomyrrhis (Apiaceae) in Chaerophyllum Based on Nuclear and Chloroplast DNA Sequences KUO-FANG CHUNG Department of Biology, Washington University, St. Louis, Missouri 63130 U.S.A. Present address: Research Center for Biodiversity, Academia Sinica, Nankang, Taipei 115, Taiwan ([email protected]) Communicating Editor: Andrea Schwarzbach ABSTRACT. To test the hypotheses that Oreomyrrhis species form the sister clade of North American Chaerophyllum and thus render the mainly Eurasian Chaerophyllum paraphyletic, two chloroplast intergenic spacers (atpB-rbcL and trnS-trnG) were analyzed. Phylogenetic estimates using maximum parsimony, maximum likelihood, and Bayesian inference of separate and combined matrices strongly support the monophyly of Oreomyrrhis and its sister-group relationship with North American Chaerophyllum. Chaerophyllum temulum, the type species of Chaerophyllum, is the sister taxon to the clade composed of Oreomyrrhis and North American Chaerophyllum. Relationships among other major clades of Chaerophyllum are congruent with previous studies. Based on these phylogenetic estimates, all currently recognized taxa of Oreomyrrhis are synonymised with Chaerophyllum. The unranked names, North American clade and Oreomyrrhis clade, are advocated to designate the two well-supported clades within Chaerophyllum sect. Chaerophyllum. The proposed nomenclatural changes include three new names, Chaerophyllum australianum, Chaerophyllum guatemalense, and Chaerophyllum novae-zelandiae, and 26 new combinations, Chaerophyllum andicola, Chaerophyllum argentum, Chaerophyllum azorellaceum, Chaerophyllum basicola, Chaerophyllum borneense, Chaerophyllum brevipes, Chaerophyllum buwaldianum, Chaerophyllum colensoi, Chaerophyllum colensoi var. delicatulum, Chaerophyllum colensoi var. hispidum, Chaerophyllum colensoi var. multifidum, Chaerophyllum daucoides, Chaerophyllum eriopodum, Chaerophyllum gunnii, Chaerophyllum involucratum, Chaerophyllum lineare, Chaerophyllum nanhuense, Chaerophyllum orizabae, Chaerophyllum papuanum, Chaerophyllum plicatum, Chaerophyllum pulvinificum, Chaerophyllum pumilum, Chaerophyllum ramosum, Chaerophyllum sessiliflorum, Chaerophyllum taiwanianum, and Chaerophyllum tolucanum. KEYWORDS:

atpB-rbcL, monophyly, nrITS, phylogeny, taxonomy, trnS-trnG.

Plants of the genus Oreomyrrhis Endl. are small herbs characterized by a highly reduced inflorescence consisting of a simple umbel borne terminally on penduncles that elongate from the bases of sheathed and often rosette-forming leaves (Mathias and Constance 1955). In one of the most remarkable New Guinean species, Oreomyrrhis azorellacea Buwalda, the inflorescence is further reduced to a solitary flower on a short peduncle ascending from the side branches of a densely matted, mosslike cushion (Mathias and Constance 1955). Within the subfamily Apioideae, which are known for their compound umbels, the simple umbel is a rare trait that is consistently found only in a small number of genera, such as Lilaeopsis Green, Neogoezia Hemsl., and Oreomyrrhis (Constance 1987). Composed of ca. 26 species, Oreomyrrhis inhabits alpine, subalpine, and sub-Antarctic habitats around the South Pacific Basin in Mesoamerica, central to northern Andes, southern Patagonia, Tierra del Fuego, the Falklands, New Zealand, southeastern Australia, Tasmania, New Guinea, Borneo, and Taiwan (Mathias and Constance 1955). As one of a few Apioideae genera found mainly in the Southern Hemisphere (Mathias 1965), this remarkable disjunct distribution in the South Pacific has fascinated and puzzled generations of phytogeographers (e.g., Hooker 1853; Hayata 1911;

Mathias and Constance 1955; van Steenis 1964; Raven and Axelrod 1974; Melville 1981). In addition to its remarkable geographic pattern and unusual morphology, Oreomyrrhis is characterized by a base chromosome number of n 5 6 (Moore 1971; Raven 1975), a number that is found in only about 4% of apioid species (Pimenov et al. 2003). The combination of its unusual morphological, geographic, and cytological features made Oreomyrrhis one of the most enigmatic genera in Apiaceae (Mathias and Constance 1955). In a recent study that examined the taxonomic and biogeographic hypotheses surrounding Oreomyrrhis based on nuclear internal transcribed spacer (ITS) DNA sequence data (Chung et al. 2005), Oreomyrrhis was unambiguously placed within Scandiceae subtribe Scandicinae, supporting the taxonomic treatments by de Candolle (1830), Bentham and Hooker (1867), and Koso-Poljansky (1916). Within Scandicinae, Oreomyrrhis was nested within Chaerophyllum. Specifically, Oreomyrrhis and the North American Chaerophyllum [C. procumbens (L.) Crantz and C. tainturieri Hook. & Arn.] formed a strongly supported clade that was sister to C. temulum L., the type species of the genus. The ITS data also suggested that the disjunct distribution of Oreomyrrhis in the South Pacific is better explained by a more recent (most likely late-Tertiary to Quater-

671

672

[Volume 32

SYSTEMATIC BOTANY

nary) range expansion via long-distance dispersal (Chung et al. 2005). Based on the phylogenetic placement of Oreomyrrhis in Chung et al. (2005), both the genus Chaerophyllum and section Chaerophyllum are paraphyletic. As recently re-circumscribed by Spalik and Downie (2001) and Spalik et al. (2001a) using molecular (ITS) and morphological data, Chaerophyllum is the most species-rich and ecologically/ morphologically diverse genus in Scandicinae. This primarily Eurasian (including Northern Africa) genus is most diversified in the Mediterranean and Caucasus regions, with two species native to eastern North America (Spalik and Downie 2001). Molecular data revealed four wellsupported clades within Chaerophyllum (Downie et al. 2000). However, because they lack apparent morphological synapomorphies, Spalik and Downie (2001) treated these clades as sections: sect. Chaerophyllum, sect. Chrysocarpum Spalik & S. R. Downie, sect. Dasypetalon Neilr., and sect. Physocaulis DC. Spalik and Downie (2001) and Spalik et al. (2001a) explicitly applied monophyly as a fundamental criterion to delimit and re-circumscribe taxa within Scandicinae, and thus the unexpected placement of Oreomyrrhis within the Chaerophyllum clade requires that the circumscription of Chaerophyllum be amended to maintain the integrity of their taxonomic efforts. However, it has become widely acknowledged that the unique evolutionary features of the ITS region can occasionally lead to ´ lvarez and incorrect phylogenetic inferences (A Wendel 2003). Results drawn from a single molecular data set, especially ITS, ideally should be tested with data from other DNA regions before making formal nomenclatural changes. Moreover, support for the monophyly of Oreomyrrhis was weak in Chung et al. (2005), an uncertainty that could further undermine the biological inferences of that study. To test the phylogenetic hypotheses raised by the ITS data (Chung et al. 2005), the author surveyed and sequenced several chloroplast DNA (cpDNA) markers reported to contain adequate sequence variation [e.g. trnL-trnF (Taberlet et al. 1991), atpBrbcL (Chiang et al. 1998), psaI-accD (Small et al. 1998), trnS-trnG (Hamilton 1999), and rpl16 (Seelanan et al. 1999)]. Two markers, trnS-trnG and atpBrbcL, that showed substantial variation and were easy to sequence were chosen in this study. The other regions surveyed were either too conserved and/or difficult to provide quality data. In a recent study that compared 21 noncoding chloroplast DNA sequences, Shaw et al. (2005) also reported that trnS-trnG region was among the most variable

region for infragenric investigations. The atpB-rbcL was excluded from their analysis as Shaw et al. (2005) considered it too conserved for infrageneric analyses. Nevertheless, my preliminary study indicated that atpB-rbcL was potentially informative in untangling relationship between Oreomyrrhis and the North American Chaerophyllum that remained unresolved using ITS data (Chung et al. 2005). In this study, separate and combined data sets of cpDNA and ITS sequences are analyzed to address the following questions: (1) Is Oreomyrrhis monophyletic? (2) Is Oreomyrrhis the sister taxon to North American Chaerophyllum, rendering Chaerophyllum paraphyletic? (3) Are the phylogenetic relationships revealed by cpDNA data congruent with the ITS phylogeny? MATERIALS AND METHODS Taxon Sampling and DNA Isolation. Thirty-six species of ingroup and outgroup taxa, with complete sequences for all three loci (ITS, atpB-rbcL, and trnS-trnG), were included in the phylogenetic analyses (Appendix 1). Twenty-three out of the 26 described species of Oreomyrrhis were sampled, comprising representatives from all major geographic areas. The species not sampled in this study include O. buwaldiana Mathias & Constance and O. plicata Mathias & Constance, both rare species known only by a few collections from New Guinean, and the Mexican highland endemic O. tolucana for which only atpB-rbcL sequence has been collected successfully. For Chaerophyllum, at least one species for each section circumscribed by Spalik and Downie (2001) was sampled, including all three species of sect. Chaerophyllum, one species of sect. Dasypetalon (3 species in total), five species of sect. Chrysocarpum (ca. 14 species), one species of sect. Physocaulis (one species), and two species not yet subjected to molecular analysis (C. humile Stev. and C. roseum M. Bieb.). Myrrhis odorata (L.) Scop. and Osmorhiza longistylis (Torr.) DC. were chosen as outgroups based on recent molecular phylogenetic studies (Downie et al. 2000). DNA was isolated from both silica-dried and herbarium material. For silica-dried material, total genomic DNA extraction was performed in a 1.5 ml tube following the CTAB protocol outlined in Doyle and Doyle (1987), with 1.5 ml of 20 mg/ml RNase A added during the 60–65uC incubation stage. For herbarium material, the aqueous phase from a chloroform-isoamyl alcohol (24:1) extraction stage was further purified using the QIAquickH PCR purification kit (Qiagen, Valencia, California). The aqueous phase was first mixed with 5 volumes of PB buffer. The sample was then cleaned and DNA was eluted using a Qiagen kit following manufacturer’s protocol (Stefanovic´ et al. 2002). PCR Amplifications and Sequencing. Polymerase Chain Reaction (PCR) amplifications were performed in a 30 ml reactions using conditions outlined in Chung et al. (2005). The chloroplast intergenic spacer between atpB (ATP synthase beta subunit) and rbcL (RuBisCO large subunit) was amplified using the primers ‘‘atpB’’ and ‘‘rbcL’’ (Table 1) designed based on tobacco (Nicotiana tabacum L.) and lettuce (Lactuca sativa L.) chloroplast genome sequences. Amplifications from DNAs isolated from herbarium material required additional external and internal primers (Table 1). The following cycling conditions were used: 94uC (5 min); 30 (silica-dried material) or 40 (herbarium material) cycles of 94uC (30 sec), 54uC (1 min), 68uC (30 sec); 68uC (7 min) and then 25uC (1 min). The chloroplast intergenic spacer between

2007]

CHUNG: THE OREOMYRRHIS CLADE OF CHAEROPHYLLUM

TABLE 1. Primers designed for amplifying and sequencing atpB-rbcL. Primer name

Sequence

atpB atpB-rbcL-F3 atpB-rbcL-F4 atpB-rbcL-R4 atpB-rbcL-R3 rbcL

59–CTACATCCAGTACCGGACCA–39 59–CCATATATATGAAAGAGTATAC–39 59–GATTTAYATATACAACATAGAC–39 59–GTATACTCTTTCATATATATGG–39 59–GTCTATGTTGTATATGTAAATCC–39 59–AACACCAGCTTTGAATCCAA–39

the trnS [tRNA-Ser(GCU)] and trnG [tRNA-Gly(UCC)] genes was amplified using the primers trn S (GCU) and trn G (UCC) (Hamilton 1999) with the following cycling conditions: 93uC (2 min); 30 (silica-dried materials) or 40 (herbarium materials) cycles of 93uC (30 sec), 53uC (30 sec), 68uC (35 sec); 68uC (7 min) and then 25uC (1 min). To remove excess dNTP’s and unincorporated primers, 15 ml of PCR product were treated with 1 ml of exonuclease I (1 U/ml, New England BioLabs, Beverly, Massachusetts) and 1.5 ml of shrimp alkaline phosphatase (1 U/ml; Promega, Madison, Wisconsin) at 37uC for 15 min and 85uC for 15 min. Cycle sequencing (30 cycles, 15 sec denaturation at 95uC, 10 sec annealing at 50uC, 3.5 min extension at 60uC) using dye terminators (ABI BigDye Terminator v3.1 Cycle Sequencing kit) was performed in a 4-ml volume. In addition to the external PCR primers, internal primers (Table 1) were used for the atpB-rbcL spacer to ensure sequence quality. Sequence reactions were cleaned with a 96-well PVDF filter plate (Innovative Microplate, Chicopee, Massachusetts) containing G-50 Fine Sephadex (Amersham Biosciences AB, Uppsala, Sweden). Sequence reaction products were separated and visualized using an ABI Prism 3130 16-capillary automated sequencer. Contigs were assembled using DNA STAR-SeqMan II version 5.00 (Lasergene Navigator, Madison, Wisconsin). All newly acquired sequences have been archived in GenBank (Appendix 1). DNA Sequence Alignment, Phylogenetic Analyses, and Congruence Tests. Boundaries of the two chloroplast intergenic spacers were established by comparisons to the published tobacco and lettuce chloroplast genomes. Multiple sequence alignments were performed by ClustalW and subsequently manually adjusted using the Alignment Explorer in MEGA 3.1 (Kumar et al. 2004). Because chloroplast loci are linked on a nonrecombinant chromosome and inherited as a single unit, we combined atpB-rbcL and trnStrnG sequences into a single cpDNA data set for phylogenetic analyses. Regions of ambiguous alignment in the chloroplast data set were excluded in all phylogenetic analyses. The combined matrix is available in TreeBASE (study number 5 S1571). All three data sets (ITS, cpDNA, and combined matrix) were analyzed using maximum parsimony (MP) and maximum likelihood (ML) optimality criteria in PAUP* (Swofford 2002) and the Bayesian Inference of phylogeny (BI) in MrBayes v3.1.2 (Ronquist and Huelsenbeck 2003). For MP analyses, alignment gaps were treated as missing data, all characters were treated as unordered, and all character transformations were weighted equally. One heuristic search was performed for each data set with 1,000,000 random addition replicates, tree bisection-reconnection (TBR) branch swapping, and MulTrees and Steepest descent option in effect. For parsimony bootstrap (PB) analysis, heuristic searches of the 1,000 pseudo-replicates were conducted, with 10 random sequence additions, MaxTrees 5 1,000 without increasing, TBR swapping, Multrees and Steepest descent options in effect. Decay

673

indices (DI) were calculated with the aid of the program TreeRot.v2 (Sorenson 1999). For the ITS data set, heuristic searches by enforcing a topological constraint (monophyletic Oreomyrrhis) were also conducted with TBR swapping and 100,000 random addition replicates. For ML analysis, heuristic searches for each data set with 10 random replicates and TBR swapping were conducted using the model of nucleotide substitution and parameters (Table 2) selected by Akaike information criterion (AIC) in the program Modeltest 3.7 (Posada and Crandall 1998). Likelihood bootstrap (LB) analyses of 100 pseudo-replicates were performed to assess the node support, using as-is sequence addition and TBR swapping, based on the same parameters inferred by Modeltest. Bayesian inferences of phylogeny of the three data sets were performed in MrBayes by Metropolis-coupled Markov chain Monte Carlo, based on the model and parameters (Table 2) selected by AIC in the program MrModeltest version 2.0 (Nylander 2004). For the combined data set, BI analyses were conducted using mixed models specified to each data partition. A separate BI analysis was also performed based on the single model selected for the entire data set. The topology of the Bayesian phylogeny estimate from the simple model was identical to the one from the mixed-model analysis but with a lower likelihood score (Table 2). We only report the results from the mixed model analysis. Two independent searches (four independent runs) were conducted for each data set (3,000,000 generations, sampling every 100 generations). The burn-in for each run was determined by plotting the likelihood values against the generations. After discarding the burn-in, trees sampled in all four runs were combined and the posterior probabilities (PP) were obtained from the 50% majority-rule consensus tree using PAUP*. To assess the congruence among different data partitions, we performed the incongruence length difference test (ILD test or partition-homogeneity test in PAUP*; Farris et al. 1994). One thousand homogeneity replicates of heuristic searches were conducted with 100 random sequence additions and MulTrees not in effect.

RESULTS Information regarding sequence variability, phylogenetic inferences, and model choices of different data partitions is summarized in Table 2. Seventy characters of uncertain alignment of cpDNA sequences (aptB-rbcL, aligned positions 65–67 and 542–558; trnS-trnG, aligned position 934–937, 969– 974, and 1190–1229), characterized by long stretches of consecutive As or Ts, were excluded from all analyses. Phylogenetic Analysis of ITS. Results of phylogenetic analyses of the ITS data set are summarized in Fig. 1A. Topologies of the majority-rule consensus of the equally parsimonious trees (6,717 trees of 181 steps), the strict consensus of 12 ML trees, and the BI phylogeny were highly concordant (Fig. 1A). In all analyses, Oreomyrrhis and the two North American Chaerophyllum species form a moderately supported clade (PB 5 80; DI 5 2; PP 5 83; LB 5 55) that is the sister taxon to C. temulum with good support (PB 5 86; DI 5 2; PP 5 100; LB 5 96), a relationship that was noted in Chung et al. (2005). Among the 6,717 equally parsimonious

GTR+I + SYM+G (GTR+I+G) 24682.9166.62 (24913.4065.61) 24683.3366.57 (24913.2865.69) 24683.0866.64 (24913.2065.65) 24683.2766.681287307 (24913.2665.67) SYM+G 22159.7265.72 22072.85679.58 22147.74621.88 22159.6265.76 Model(s) for Bayesian inference MrBayes run 1 log likelihood (lnL) MrBayes run 2 log likelihood (lnL) MrBayes run 3 log likelihood (lnL) MrBayes run 4 log likelihood (lnL)

(burn-in (burn-in (burn-in (burn-in

5 5 5 5

the the the the

first first first first

100 100 100 100

trees) trees) trees) trees)

GTR+I 22569.8869.10 22569.6466.94 22570.2767.27 22570.0067.12

1 24624.89666 1 22506.30689 No. tree found log likelihood (lnL) (highest if multiple trees present)

12 21934.81217

53558 306 0.85/0.9/0.77 K81uf+I+G 6717 181 0.87/0.91/0.79 SYM+G 96996 117 0.87/0.94/0.82 TIM+I No. MPT Tree length CI/RI/RC Model used for ML analysis

Combined ITS (ITS-1/5.8S/ITS-2)

598–670 (210–220/163–164/222–226) 616 (224/164/228) 0 74 (12) [29 (12.9)/5 (3.1)/40 (17.5)] 1252–1292 (733–757/510–541) 1342 (776/566) 70 (5.2) [20 (2.8)/50 (8.8)] 47 (3.5) [19 (2.5)/28 (4.9)]

cpDNA (atpB-rbcL/trnSG)

1857–1898 1958 70 (3.6) 121 (6.2)

SYSTEMATIC BOTANY

Length variation (bp) No. of aligned position (bp) No. characters excluded (%) Parsimony informative characters (%)

TABLE 2. DNA sequence characteristics and phylogenetic statistics of each data partition. Models used for ML analysis were selected by Modeltest. Models used for Bayesian inference were selected by MrModeltest. MPT 5 most parsimonious trees. In the combined data set, Bayesian analyses were performed on both mixed and single (in parentheses) models.

674

[Volume 32

trees (MPTs; 181 steps), the monophyly of Oreomyrrhis was recovered in 4.4% (297 trees) of the tree topologies. Therefore, the monophyly of Oreomyrrhis cannot be rejected based on the ITS data. One relationship not seen in previous analyses but present in all 12 ML trees, 505 out of 6,727 (7.5%) MPTs, and the Bayesian phylogeny (PP 5 60%) is the grouping of two New Guinea species (O. pumila and O. papuana) and two North American Chaerophyllum species (Fig. 1A), however, this relationship has low bootstrap support (,50%) and a decay index of 0. Chaerophyllum humile and C. roseum, which have not been analyzed using molecular data (Spalik et al. 2001a), were both placed in sect. Chrysocarpum with strong support (PB 5 100; DI 5 9; PP 5 100; LB 5 99; Fig. 1A). Relationships among the four major clades of Chaerophyllum, sect. Chaerophyllum (including Oreomyrrhis), sect. Chrysocarpum, sect. Dasypetalon, and sect. Physocaulis, remain unresolved. Phylogenetic Analysis of cpDNA. Compared to the ITS data set, the cpDNA matrix exhibits fewer parsimony-informative characters (Table 2) but provides more phylogenetic resolution (Fig. 1B). Based on cpDNA data, Oreomyrrhis species form a strongly supported clade (PB: 90; DI: 3; PP: 100; LP: 94) that is the sister taxon to the North American Chaerophyllum clade with moderate support (PB: 60; DI: 1; PP: 92: LP: 60) in all three analyses (Fig. 1B). However, C. temulum, the type species of Chaerophyllum does not appear to be allied with the North American Chaerophyllum and Oreomyrrhis, as suggested in the ITS phylogeny (Fig. 1A). Instead, its phylogenetic position in the cpDNA data remains unresolved (Fig. 1B). Within Oreomyrrhis, New Guinean and Bornean species form basal grade that is sister to the rest of Oreomyrrhis with moderate support (PB: 65; DI: 1: PP: 97; LP: 64). The monophyly of sect. Chrysocarpum is strongly supported in all analyses (PB: 97; DI: 7; PP: 100; LB: 98). Chaerophyllum humile and C. roseum are both placed in sect. Chrysocarpum, as indicated by the ITS data set (Fig. 1). Additionally, the cpDNA data set indicates that sect. Chrysocarpum and sect. Physocaulis (C. nodosum) are sister taxa to each other with good support (PB: 88; DI: 3; PP: 99; LB: 86). One apparent difference between the ITS and cpDNA phylogenies is the placement of C. azoricum Trel., an endemic in the Azores Islands. While ITS data indicate that C. azoricum is closely related to C. aromaticum L. (Fig. 1A; Downie et al. 2000), the cpDNA data set strongly supports its basal position in sect. Chrysocarpum (PB 5 99; DI 5 7; PP 5 100; LB 5 98). Congruence Tests and Phylogenetic Analysis of Combined Data set. The result of the ILD test (P 5

2007]

CHUNG: THE OREOMYRRHIS CLADE OF CHAEROPHYLLUM

675

FIG. 1. Bayesian consensus cladograms. A. ITS data set. B. chloroplast DNA data set. Numbers above clades indicate parsimony bootstrap support (.50%) and decay index; numbers below clades indicate Bayesian posterior probabilities and likelihood bootstrap support (.50%). A dash indicates clade that was not present in the strict consensus tree of maximum parsimony analysis.

0.002) indicates significant incongruence between cpDNA and ITS data sets. Comparing the two topologies in Fig. 1, the different placement of Chaerophyllum azoricum between ITS and cpDNA phylogenies could be the source of the conflicting signal in the ILD test. This conjecture was corroborated by an ILD test in which the Azorean endemic was excluded (P 5 0.485). To further evaluate if the inclusion of C. azoricum in the combined matrix would result in different phylogenetic relationships among major clades of Chaerophyllum, maximum parsimony analyses were also performed on matrices with and without this Azorean Chaerophyllum. Except for the inclusion of C. azoricum, however, the strict consensus trees of the two parsimony analyses are identical. Given that the relationships within Chaerophyllum were not affected by the inclusion of Azorean species, the cpDNA and nrITS data matrices were combined without excluding C. azoricum. Phylogenetic analyses of the combined matrix based on the three methods resulted in similar topologies (Fig. 2). The monophyly of Oreomyrrhis (PB: 85; DI: 2; PP: 100; LB: 70) and its sister-group

relationship (PB: 95; DI: 3; PP: 100; LB: 94) to the North American Chaerophyllum species are well supported. The combined matrix also indicated the placement of C. temulum as the sister taxon to the clade composed of Oreomyrrhis and the North American Chaerophyllum clade with strong support (PB: 89; DI: 4; PP: 100; LB: 98). Chaerophyllum nodosum (sect. Physocaulis) appears as the sister taxon to the strongly supported sect. Chrysocarpum (PB: 100; DI: 20; PP: 100; LB: 100) with good support (PB: 83; DI: 3; PP: 95; LB: 83). DISCUSSION Although Apiaceae were the first plant family to be generally recognized morphologically (Constance 1971), for centuries the taxonomy of the family has been among the most contentious in the flowering plants. Traditionally, Apiaceae taxonomists have relied heavily on mericarp morphology to construct infrafamiliar taxonomy. However, it has become evident that mericarp characters are often associated with dispersal mechanisms and are prone to convergent evolution, making them

676

SYSTEMATIC BOTANY

[Volume 32

FIG. 2. Bayesian consensus cladogram of the combined data set and proposed taxonomy for Chaerophyllum. Numbers above clades indicate parsimony bootstrap support (.50%) and decay index; number below clades indicate Bayesian posterior probability and likelihood bootstrap support (.50%). The dash lines in the phylogeny denote clade that was not present in the strict consensus tree of maximum parsimony analysis. The infrageneric taxonomy of Chaerophyllum proposed by Spalik and Downie (2001) is shown by the brackets immediately after the taxon name of the phylogeny.

highly homoplasious (Spalik et al. 2001b). Recent advances in molecular systematics have greatly improved our knowledge of the phylogenetic relationships within Apiaceae (Downie et al. 2000), and many of these studies have been translated into formal classification (e.g., Spalik and Downie 2001). The well-supported phylogenetic relationships presented in this study provide strong evidence for the paraphyletic nature of Chaerophyllum and its sect. Chaerophyllum (sensu Spalik and Downie 2001). Additionally, the monotypic sect. Physocaulis (C. nodosum) appears to be the sister taxon to sect. Chrysocarupum, based on the combined data set (Fig. 2). Although the sampling of Chaerophyllum is incomplete, an ongoing phylogenetic study of Chaerophyllum with comprehensive sampling of the group supports the recognition of the four sections (sensu Spalik and Downie 2001), with Oreomyrrhis nested within sect. Chaerophyllum (Piwczynski and Spalik 2005; Spalik, pers. comm.). Given this well-established relationship, it is imperative to translate these relationships into

a formal taxonomy, a move that will facilitate the transmission of evolutionary insights derived from this study (Franz 2005). While the status of paraphyletic taxa in the Linnaean taxonomic system is currently the subject of contentious debate (e.g., Nordal and Stedje 2005; Williams et al. 2005; Brummitt 2006; Stevens 2006), a full discussion of this issue is beyond the scope of this paper. I take the position of Stevens (2006) and advocate his view to achieve a universal and stable naming system by naming only monophyletic groups using a Linnaean nomenclature. Such a position is also in agreement with the criteria implemented in Spalik and Downie (2001) and Spalik et al. (2001a) to circumscribe taxa in the apioid subtribe Scandicinae. Based on the topology inferred from the combined data set (Fig. 2), a monophyletic Chaerophyllum can be achieved by including Oreomyrrhis. Alternatively, retaining the generic status of Oreomyrrhis would require the recognition of at least three additional genera (corresponding to sect. Dasypetalon, sect. Physocaulis plus sect. Chry-

2007]

CHUNG: THE OREOMYRRHIS CLADE OF CHAEROPHYLLUM

socarpum, and the North American Chaerophyllum clade), leaving Chaerophyllum a monotypic genus composed of only its type species C. temulum. Although the total number of specific name changes is similar (29 for Oreomyrrhis vs. ca. 30 for Chaerophyllum) in the two alternative classifications, subsuming Oreomyrrhis under Chaerophyllum does not require the recognition of any new generic names. Additionally, splitting Chaerophyllum would result in clades that lack apparent morphological synapomorphies (Spalik and Downie 2001). Moreover, the short branch length leading to Oreomyrrhis (Fig. 2) suggests that this genus might be a young (Pliocene-Pleistocene) lineage recently derived from within Chaerophyllum (Chung et al. 2005). Morphologically, members of Chaerophyllum (sensu Spalik and Downie 2001) can be identified by a suite of fruit characters, including oblong, straw-yellow to brown mericarps with thin cuticles, broad primary ridges, and solitary oil-tubes (vittae) in mericarp furrows (valleculae) that remain extant at fruit maturity (Spalik et al. 2001a). Although none of these characters is unique to Chaerophyllum (Spalik et al. 2001a), these features also characterize the fruit of Oreomyrrhis (Mathias and Constance 1955), with some minor exceptions in the number of vallecular vittae (Chen and Wang 2001), and thus supporting the taxonomy proposed here. With the inclusion of Oreomyrrhis, Chaerophyllum will comprise more than 60 species and, thus, a new infrageneric classification will be useful. However, recognizing ‘‘section Oreomyrrhis’’ will again render sect. Chaerophyllum paraphyletic and would mandate the recognition of the North American Chaerophyllum clade as a sect. nov. (Fig. 2), leaving sect. Chaerophyllum a monotypic section containing only C. temulum. As suggested by Stevens (2006), one way to avoid the cascades of name changes introduced by rank alternations is to use unranked and informal names. Since the International Code of Botanical Nomenclature (Greuter et al. 2000) does not prohibit the use of unranked names, designating clades by unranked names seems a practical alternative (Stevens 2006) and has been employed with increasing frequencies (e.g. Wojciechowski et al. 2004). Additionally, generic and specific names are the ones used most often to communicate information on biodiversity and are the only names every species is required to have (Moore et al. 2004). Here, I propose the use of ‘‘Oreomyrrhis clade’’ and ‘‘North American clade’’ (Fig. 2) for the two well supported clades within sect. Chaerophyllum. Compared to their Eurasian congeners, flowers in the North American clade and the Oreomyrrhis clade are all hermaphroditic

677

with smaller corollas, traits that are often associated with selfing (Spalik and Downie 2001; Chung et al. 2005). The evolution from outcrossing to selfing may have played a pivotal role in the successful colonizing and occupation of the cold and alpine environments by the Oreomyrrhis clade (Chung et al. 2005). TAXONOMIC TREATMENT Taxa of Oreomyrrhis Endl. circumscribed and described by Mathias and Constance (1955) and novelties subsequently published by Allan (1961), Mathias and Constance (1977), Chen & Wang (2001), and Heenan and Molloy (2006) are transferred to Chaerophyllum sect. Chaerophyllum (Spalik & Downie, 2001; Spalik et al. 2001). Heterotypic synonyms were enumerated in these previous works and are not repeated here. CHAEROPHYLLUM L., Sp. Pl. 258. 1753; Spalik & Downie, Ann. Missouri Bot. Gard. 88: 290. 2001.—LECTOTYPE: Chaerophyllum temulum L.; designated by Reduron & Jarvis, Taxon 41: 560. 1992. CHAEROPHYLLUM sect. CHAEROPHYLLUM CHAEROPHYLLUM TEMULUM L., Sp. Pl. 258. 1753. North American clade 1. CHAEROPHYLLUM PROCUMBENS (L.) Crantz, Cl. Umbel. Emend. 77. 1767. 2. CHAEROPHYLLUM TAINTURIERI Hook. & Arn. in Hook. Comp. Bot. Mag. 1: 47. 1835. Oreomyrrhis clade 1. Chaerophyllum andicola (Kunth) K. F. Chung, comb. nov. Myrrhis andicola Kunth in Humboldt, Bopland and Kunth,, Nov. Gen. Sp. 5: 13, tab. 419. 1821. Caldasia andicola (Kunth) Lag. ex DC., Coll. Me´m. 5: 60, pl. 2(J4–J5). 1829. Oreomyrrhis andicola (Kunth) Endl. ex Hook. f., Bot. Antarct. Voy., Vol. 1, Fl. Antarct. 2: 288, pl. 101. 1846.— TYPE: ECUADOR. Crescit locis alpinis regni Quitensis, in herbida planitis Antisanae, alt. 2100 hex., (fr), Bonpland 2258 (holotype: P-Bonpl, IDC microfiche 6209.115:I.5!). 2. Chaerophyllum argentum (Hook. f.) K. F. Chung, comb. nov. Caldasia argentea Hook. f., Icon. Pl. 3: tab. 300. 1840. Oreomyrrhis argentea (Hook. f.) Hook. f., Bot. Antarct. Voy., Vol. 3, Fl. Tasman. 1: 162. 1856.—TYPE: AUSTRALIA. Tasmania: Middlesex Plains, Van Dieman’s Land, 14 Feb 1837 (fr), R. Gunn 823 (holotype: K!). 3. Chaerophyllum australianum K. F. Chung, nom. nov. Oreomyrrhis ciliata Hook. f., London

678

SYSTEMATIC BOTANY

J. Bot. 6: 471. 1847 (non Chaerophyllum ciliatum Kit. ex Kanitz, Reliq. Kit. 103. 1863.).—TYPE: AUSTRALIA. Tasmania: Arthur’s Lake, 17 Jan 1845, Gunn 824 (holotype: K). Note. The new name is required because of the existence of Chaerophyllum ciliatum Kit. ex Kanitz. The specific epithet ‘‘australianum’’ is proposed for this distinct and second commonest Australian Oreomyrrhis species. 4. Chaerophyllum azorellaceum (Buwalda) K. F. Chung, comb. nov. Oreomyrrhis azorellacea Buwalda, Fl. Males., Ser. 1, Spermat. 4: 130, figs. 7–8. 1949.—TYPE: PAPUA NEW GUINEA. Central Division: Mount Albert Edward, 21 Jun 1933 (fr), Brass 4306 (holotype: A!; isotype: NY!). 5. Chaerophyllum basicola (Heenan & Molloy) K. F. Chung, comb. nov. Oreomyrrhis basicola Heenan & Molloy, NZ J. Bot. 44: 100, fig. 3A. 2006.—TYPE: NEW ZEALAND. Otago: Waitaki River Valley, Awahokomo, limestone outcrop, 8 Nov 2001, Heenan CHR 546301 (holotype: CHR). 6. Chaerophyllum borneense (Merr.) K. F. Chung, comb. nov. Oreomyrrhis borneensis Merr., Amer. J. Bot. 5: 514, pl. 36. 1918.—TYPE: MALAYSIA. Borneo: Mount Kinabalu, Low’s peak, 13 Nov 1915 (fr), Clemens 10622 (holotype: UC!). 7. Chaerophyllum brevipes (Mathias & Constance) K. F. Chung, comb. nov. Oreomyrrhis brevipes Mathias & Constance, Univ. Calif. Publ. Bot. 27: 390, fig. 14(d–f). 1955.—TYPE: AUSTRALIA. New South Wales: Snowy Mountains, Mar 1890 (fr), Bauerlen s.n. (holotype: NSW!; isotype: US!). 8. Chaerophyllum buwaldianum (Mathias & Constance) K. F. Chung, comb. nov. Oreomyrrhis buwaldiana Mathias & Constance, Univ. Calif. Publ. Bot. 27: 401, fig. 18(d–f). 1955.—TYPE: INDONESIA. Irian Jaya: northern slopes of Mount Wilhelmina, Sep 1938 (fr), Brass & Meyer-Dress 10082 (holotype: A!; isotype: L!). 9. Chaerophyllum colensoi (Hook. f.) K. F. Chung, comb. nov. Oreomyrrhis colensoi Hook. f., Bot. Antarct. Voy., Vol. 2, Fl. Nov.-Zel. 1: 92. 1852. Oreomyrrhis andicola var. colensoi Kirk, Stud. Fl. New Zealand 198. 1899.—TYPE: NEW ZEALAND. North Island: Mountainous place on the east coast and in the interior, 1849 (fr), Colenso 1698 (lectotype: K!, designated by Mathias and Constance 1955). 9b. Chaerophyllum colensoi var. delicatulum (Allan) K. F. Chung, comb. nov. Oreomyrrhis colensoi var. delicatula Allan, Fl. New Zealand

[Volume 32

1: 973, fig. 17. 1961.—TYPE: NEW ZEALAND. North Island: Hauhangaroa Range, Jan 1952, Druce s.n. (holotype: CHR; isotype: UC!). 9c. Chaerophyllum colensoi var. hispidum (Allan) K. F. Chung, comb. nov. Oreomyrrhis colensoi var. hispida Allan, Fl. New Zealand 1: 973, fig. 17. 1961.—TYPE: NEW ZEALAND. North Island: Mount Hikurangi, Raukumara Range, Jan 1897, Petrie s.n. (holotype: CHR). 9d. Chaerophyllum colensoi var. multifidum (Allan) K. F. Chung, comb. nov. Oreomyrrhis colensoi var. multifida Allan, Fl. New Zealand 1: 973, fig. 17. 1961.—TYPE: NEW ZEALAND. North Island: Hutt River, ca. 5 miles N from Upper Hutt, ca. 500 ft, 9 Feb 1947, Druce s.n. (holotype: CHR). 10. Chaerophyllum daucoides (d’Urv.) K. F. Chung, comb. nov. Azorella daucoides d’Urv., Me´m. Soc. Linn. Paris 4: 613. 1826 (non Oreomyrrhis daucoides Urban, Linnaea 43: 303. 1882.).—Oreomyrrhis hookeri Mathias & Constance, Univ. Calif. Publ. Bot. 27: 369, fig. 4(a–b). 1955.—TYPE: FALKLAND ISLANDS (ISLAS MALVINAS). Berkeley Sound, Soledad, 1825 (fr), d’Urville 85 (holotype: P!; isotype: W!). 11. Chaerophyllum eriopodum (DC.) K. F. Chung, comb. nov. Caldasia eriopoda DC., Coll. Me´m. 5: 60, pl. II (J4–J5). 1829. Oreomyrrhis eriopoda (DC.) Hook. f., Bot. Antarct. Voy., Vol. 3, Fl. Tasman. 1: 162. 1856.—TYPE: AUSTRALIA. ‘‘Nov. Holl.,’’ 1830, Gaudichaud (lectotype: G, designated by Mathias and Constance 1955). 12. Chaerophyllum guatemalense K. F. Chung, nom. nov. Oreomyrrhis daucifolia I. M. Johnst., J. Arnold Arbor. 19: 125. 1938 (non Chaerophyllum daucifolium Desf., Cat. Pl. Horti Paris. 200, 405. 1829.).—TYPE: GUATEMALA. Dept. of Huehuetenango, ‘‘Charcol,’’ Sierra Cuchumatanes, 15 Sep 1934 (fr), Skutch 1263 (holotype: GH!; isotype: US!). Note. The existence of the name Chaerophyllum daucifolia Desf. prohibits the direct transfer of Oreomyrrhis daucifolia into Chaerophyllum. The species epithet ‘‘guatemalense’’ is selected for this Guatemalan highland endemic. 13. Chaerophyllum gunnii (Mathias & Constance) K. F. Chung, comb. nov. Oreomyrrhis gunnii Mathias & Constance, Univ. Calif. Publ. Bot. 27: 395, fig. 16(f–g). 1955.—TYPE: AUSTRALIA. Tasmania: Limestone Cliffs, Gordon River, Macquarie Harbour, 14 Dec 1846 (fr), Milligan 703 (holotype: K!; isotype: HO-52100!, HO-23568!, K!, NY!, W!). 14. Chaerophyllum involucratum (Hayata) K. F. Chung, comb. nov. Oreomyrrhis involucrata

2007]

CHUNG: THE OREOMYRRHIS CLADE OF CHAEROPHYLLUM

Hayata, J. Coll. Sci. Imp. Univ. Tokyo 30: 128. 1911.—TYPE: TAIWAN. Monte Morrison [Yushan], 20 Oct 1906, Kawakami and Mori 2249 (lectotype: TI!, designated by Chen & Wang 2001; isolectotype: TAIF!). Note. In the recent treatment of Flora of China, Pan and Watson (2005) recognized Oreomyrrhis involucrata as the only species in Taiwan. They questioned Chen and Wang’s (2001) new species Oreomyrrhis nanhuensis and invalidated their reinstatement of the species status of O. taiwaniana. Based on extensive field investigations and herbarium work, I confirm that all three species recognized by Chen and Wang are easily distinguishable in the field as well as herbarium specimens. Additionally, the associated natural habitats described by Chen and Wang (2001) are consistent with my observation. 15. Chaerophyllum lineare (Hemsl.) K. F. Chung, comb. nov. Oreomyrrhis linearis Hemsl., Hooker’s Icon. Pl. 26: pl. 2590. 1899.—TYPE: PAPUA NEW GUINEA. Wharton Range, 1897 (fr), Giulianetti s.n. (holotype: K!). 16. Chaerophyllum nanhuense (Chih H. Chen & J. C. Wang) K. F. Chung, comb. nov. Oreomyrrhis nanhuensis Chih H. Chen & J. C. Wang, Bot. Bull. Acad. Sin. 42: 308, figs. 4, 5. 2001.—TYPE: TAIWAN. Ilan Hsien: Tatung Hsiang, Taroko National Park, Nanhutashan cirque, 23 Jun 1998 (fl & fr), Chen et al. 2402 (holotype: TNU!). 17. Chaerophyllum novae-zelandiae K. F. Chung, nom. nov. Oreomyrrhis andicola var. rigida Kirk, Stud. Fl. New Zealand p. 198. 1899. Oreomyrrhis rigida (Kirk) Allan ex Mathias & Constance, Univ. Calif. Publ. Bot. 27: 379, fig. 10(a– d). 1955 (non Chaerophyllum rigidum Huet ex Nyman, Consp. Fl. Eur. 2: 300. 1879.).—TYPE: NEW ZEALAND. South Island: Nelson, Rotoiti Jan 1875, Kirk 589 (lectotype: WELT, designated by Mathias & Constance 1955). Note. A new name is mandatory because of the existence of Chaerophyllum rigidum Huet ex Nym. The species epithet ‘‘novae-zelandiae’’ is chosen for this distinct New Zealand Oreomyrrhis species. 18. Chaerophyllum orizabae (I. M. Johnst.) K. F. Chung, comb. nov. Oreomyrrhis orizabae I. M. Johnst., J. Arnold Arbor. 19: 126. 1938.—TYPE: MEXICO. State of Puebla: Open glades of Mt. Orizaba, 12000 ft, 26 Jul 1901 (fr), Pringle 8546 (holotype: GH!; isotype: G!, F!, MO!, NSW!, NY!, P!, UC!, US!, W!). 19. Chaerophyllum papuanum (Buwalda ex Steenis) K. F. Chung, comb. nov. Oreomyrrhis papuana Buwalda ex Steenis, Bull. Jard. Bot.

679

Buitenzorg III. 13: 255. 1934.—TYPE: INDONESIA. Irian Jaya: Doorman Top, Lam 1694 (holotype: BO). 20. Chaerophyllum plicatum (Mathias & Constance) K. F. Chung, comb. nov. Oreomyrrhis plicata Mathias & Constance, J. Arnold Arbor. 58: 190, figs. 1–6. 1977.—TYPE: PAPUA NEW GUINEA. Raba Raba Subdist., Milne Bay Dist. Summit of Goe¨ Dendeniwa, Goropu Mountains (Mt. Suckling complex), 26 Jun 1972 (fr), Stevens & Veldkamp LAE 54276 (holotype: L!; isotype: A!, CANB!, LAE!, NSW!). 21. Chaerophyllum pulvinificum (F. Muell.) K. F. Chung, comb. nov. Oreomyrrhis pulvinifica F. Muell., Fragm. 8: 185. 1874.—TYPE: AUSTRALIA. New South Walses: Ad fontes glaciales alpium Australiae a monte Kosciusko usque ad Mount Buller, Jan 1874 (fr), Mueller s.n. (holotype: K!; isotype: MEL, NSW!). 22. Chaerophyllum pumilum (Ridley) K. F. Chung, comb. nov. Oreomyrrhis pumila Ridley, Trans. Linn. Soc. II. 9: 63. 1916.—TYPE: INDONESIA. Irian Jaya: (Mt Carstensz), Camp XIII–XIV, (3200–3810), Kloss s.n. (holotype: K). 23. Chaerophyllum ramosum (Hook. f.) K. F. Chung, comb. nov. Oreomyrrhis ramosa Hook. f., Handb. New Zeal. Fl. 91. 1864.—TYPE: NEW ZEALAND. South Island: Otaga, Lake District, 1863 (fr), Hector & Buchanan 5 (holotype: K!; isotype: K!). 24. Chaerophyllum sessiliflorum (Hook. f.) K. F. Chung, comb. nov. Oreomyrrhis sessiliflora Hook. f., London J. Bot. 6: 471. 1847.—TYPE: AUSTRALIA. Tasmania: Ben Lomond Apr 1896 (fl & fr), Gunn 326 (lectotype: K!, designated by Mathias and Constance 1955). Note. R. Gunn 326 is a mixed collection from two localities in Tasmania, Ben Lomond and Mount Olympus. The part collected from Ben Lomond was designated as lectotype by Mathias and Constance (1955). 25. Chaerophyllum taiwanianum (Masam.) K. F. Chung, comb. nov. Oreomyrrhis taiwaniana Masam., Trans. Nat. Hist. Soc. Formosa 28: 139. 1938.—TYPE: TAIWAN. Mount Nankotaizan [Nanhutashan], 17 Jul 1937 (fr), Suzuki s.n. (holotype: TAI!). 26. Chaerophyllum tolucanum (I. M. Johnst.) K. F. Chung, comb. nov. Oreomyrrhis tolucana I. M. Johnst., J. Arnold Arbor. 19: 127. 1938.—TYPE: MEXICO. Mexico: growing in close mats in the bottom of the crater of the Nevado de Toluca, 13,500 ft, 1 Sep 1892 (fr), Pringle 4236 (holotype: A!; isotype: G!, F!, K!, MO!, NY!, P!, UC!, US!, W!).

680

SYSTEMATIC BOTANY

ACKNOWLEDGEMENTS. This paper presents part of the doctoral dissertation of the author supervised by P. H. Raven and B. A. Schaal, M. Richardson, and A. Larson. Valuable comments and suggestions by J. Beck, S. Downie, P. Hoch, B. Oberle, K. Spalik, P. Stevens, B. Torke, the associate editor A. Schwarzbach, and two anonymous reviewers were greatly appreciated. This work would not be possible without the following: assistance and logistical support of individuals at MO, QCNE, HAST, NSW, UNE, CANB, MEL, HO, CHR, UVAL, SI, CORD, HIP, Museo Acatushu´n, and the Falkland Conservation; permission to sample herbarium material from A, CANB, HAST, HO, MO, and UC; loans from A, B, CANB, F, G, GH, HO, K, L, LA, LAE, NSW, NY, P, UC, US, VCU, W, and WELT; A.-L. Anderberg for silica-dried leaf material; research and collecting permission from the conservation authorities of Ecuador, Taiwan, Australia, New Zealand, Guatemala, and the Falklands. This project was supported by a U.S. National Science Foundation Doctoral Dissertation Improvement Grant (DEB-0408105), Botanical Society of America Karling Award, American Society of Plant Taxonomists Graduate Student Research Grant, and a Mellon Foundation grant through the Missouri Botanical Garden and the Division of Biology and Biomedical Sciences at Washington University in St. Louis.

LITERATURE CITED ALLAN, H. H. 1961. Flora of New Zealand, vol. 1. Wellington: Government Printer. ´ LVAREZ, I. and J. F. WENDEL. 2003. Ribosomal ITS sequences A and plant phylogenetic inference. Molecular Phylogenetics and Evolution 29: 417–434. BENTHAM, G. and J. D. HOOKER. 1867. Umbelliferae. Pp. 859– 931 in Genera plantarum, vol. 1. London: Lovell Reeve. BRUMMITT, R. K. 2006. Am I a bony fish? Taxon 51: 31–41. CANDOLLE, A. P. DE. 1830. Umbelliferae. Pp. 55–250 in Prodromus systematis naturalis regni vegetabilis vol. 4, ed. A. P. de Candolle. Paris: Treuttel et Wu¨rtz. CHEN, C.-H. and J.-C. WANG. 2001. Revision of the genus Oreomyrrhis Endl. (Apiaceae) in Taiwan. Botanical Bulletin of Academia Sinica 42: 303–312. CHIANG, T.-Y., B. A. SCHAAL, and C.-I. PENG. 1998. Universal primers for amplification and sequencing a noncoding spacer between the atpB and rbcL genes of chloroplast DNA. Botanical Bulletin of Academia Sinica 39: 245–250. CHUNG, K.-F., C.-I PENG, S. R. DOWNIE, K. SPALIK, and B. A. SCHAAL. 2005. Molecular systematics of the trans-Pacific alpine genus Oreomyrrhis (Apiaceae)—phylogenetic affinities and biogeographic implications. American Journal of Botany 92: 2054–2071. CONSTANCE, L. 1971. History of the classification of Umbelliferae (Apiaceae). Pp. 1–8 in The Biology and Chemistry of the Umbelliferae, ed. V. H. Heywood. New York: Academic Press. ———. 1987. Neogoezia (Apiaceae), a very distinct and elegant genus of Mexican Umbelliferae. Opera Botanica 92: 59–71. DOWNIE, S. R., D. S. KATZ-DOWNIE, and K. SPALIK. 2000. A phylogeny of Apiaceae tribe Scandiceae: evidence from nuclear ribosomal DNA internal transcribed spacer sequences. American Journal of Botany 87: 76–95. DOYLE, J. J. and J. L. DOYLE. 1987. A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochemical Bulletin 19: 11–15. FARRIS, J. S., M. KA¨LLERSJO¨, A. G. KLUGE, and C. BULT. 1994. Testing significance of incongruence. Cladistic 10: 315–319. FRANZ, N. M. 2005. On the lack of good scientific reasons for

[Volume 32

the growing phylogeny/classification gap. Cladistics 21: 495–500. GREUTER, W., J. MCNEILL, F. R. BARRIE, H. M. BURDET, V. DEMOULIN, T. S. FILGUEIRAS, D. H. NICOLSON, P. C. SILVA, J. E. SKOG, P. TREHANE, N. J. TURLAND, and D. L. HAWKSWORTH (eds.). 2000. International Code of Botanical Nomenclature (Saint Louis Code). Regnum Vegetabile 138: 1–474. HAMILTON, M. B. 1999. Four primer pairs for the amplification of chloroplast intergenic regions with intraspecific variation. Molecular Ecology 8: 521–523. HAYATA, B. 1911. Materials for a flora of Formosa. Journal of the College of Science, Imperial University of Tokyo 30: 1–471. HEENAN, P. B. and B. P. J. MOLLOY. 2006. A new species of Oreomyrrhis (Apiaceae) from southern South Island, New Zealand, and comparison of its limestone and ultramafic habitats. New Zealand Journal of Botany 44: 99–106. HOOKER, J. D. 1853. The botany of the Antarctic voyage of H. M. Discovery Ships Erebus and Terror in the year 1839–1843. II. Flora Novae-Zelandiae. London: Lovell-Reeve. KOSO-POLJANSKY, B. M. 1916. Sciadophytorum systematis lineamenta. Bulletin de la Socie´te´ impe´riale des Naturalistes (Moscou) 29: 93–222. KUMAR, S., K. TAMURA, and M. NEI. 2004. MEGA3: integrated software for molecular evolutionary genetics analysis and sequence alignment. Briefings in Bioinformatics 5: 150–163. MATHIAS, M. E. 1965. Distribution patterns of certain Umbelliferae. Annals of the Missouri Botanical Garden 52: 387–398. ——— and L. CONSTANCE. 1955. The genus Oreomyrrhis (Umbelliferae), a problem in South Pacific distribution. University of California Publications in Botany 27: 347–416. ——— and ———. 1977. A new species of Oreomyrrhis (Umbelliferae, Apiaceae) from New Guinea. Journal of the Arnold Arboretum 58: 190–192. MELVILLE, R. 1981. Vicarious plant distributions and paleogeography of the Pacific region. Pp. 238–274 in Vicariance biogeography: a critique, eds. G. Nelson and D. E. Rosen. New York: Columbia University Press. MOORE, D. M. 1971. Chromosome studies in Umbelliferae. Pp. 233–255 in The biology and chemistry of the Umbelliferae, ed. V. H. Heywood. New York: Academic Press. MOORE, G., T. M. BARKLEY, P. DEPRIEST, V. FUNK, R. W. KIGER, W. J. KRESS, D. H. NICOLSON, D. W. STEVENSON, and Q. D. WHEELER. 2004. (065–067) Proposals to amend Article 3.1, Article 22.3, and Article 26.3. Taxon 53: 214. NORDAL, I. and B. STEDJE. 2005. Paraphyletic taxa should be accepted. Taxon 54: 5–6. NYLANDER, J. A. A. 2004. MrModeltest 2.0. Program distributed by the author. Evolutionary Biology Centre, Uppsala University, Available at website, http:// www.ebc.uu.se/systzoo/staff/nylander.html/. PAN, Z. and M. F. WATSON. 2005. Oreomyrrhis. Pp. 30–31 in Flora of China, vol 14: Apiaceae through Ericaceae, eds. Z. Wu and P. H. Raven. Beijing: Science Press: and St. Louis: Missouri Botanical Garden Press. PIMENOV, M. G., VASIL’EVA, M. V. LEONOV, and J. V. DAUSHKEVICH. 2003. Karyotaxonomical analysis in the Umbelliferae. Enfield: Science Publishers. PIWCZYNSKI, M. and K. SPALIK. 2005. Phylogenetic relationships within Chaerophyllum (Apiaceae) as inferred from nuclear rDNA ITS sequence variation [abstract]. XVII International Botanical Congress: Vienna, Austria. POSADA, D. and K. A. CRANDALL. 1998. Modeltest: testing the model of DNA substitution. Bioinformatics 14: 817–818. RAVEN, P. H. 1975. The bases of angiosperm phylogeny:

2007]

CHUNG: THE OREOMYRRHIS CLADE OF CHAEROPHYLLUM

cytology. Annals of the Missouri Botanical Garden 62: 724–764. ——— and D. I. AXELROD. 1974. Angiosperm biogeography and past continental movements. Annals of the Missouri Botanical Garden 61: 539–673. RONQUIST, F. and J. P. HUELSENBECK. 2003. MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics 19: 1572–1574. SEELANAN, T., C. L. BRUBAKER, J. M. STEWART, L. A. CRAVEN, and J. F. WENDEL. 1999. Molecular systematics of Australian Gossypium section Grandicalyx (Malvaceae). Systematic Botany 24: 183–208. SHAW, J., E. B. LICKEY, J. T. BECK, S. B. FARMER, W. LIU, J. MILLER, K. C. SIRIPUN, C. T. WINDER, E. E. SCHILLING, and R. L. SMALL. 2005. The tortoise and the hare II: relative utility of 21 noncoding chloroplast DNA sequences for phylogenetic analysis. American Journal of Botany 92: 142–166. SMALL, R. L., J. A. RYBURN, R. C. CRONN, T. SEELANAN, and J. F. WENDEL. 1998. The tortoise and the hare: choosing between noncoding plastome and nuclear Adh sequences for phylogenetic reconstruction in a recently diverged plant group. American Journal of Botany 85: 1301–1315. SORENSON, M. D. 1999. TreeRot, version 2. Boston: Boston University, Available at website, http://people.bu.edu/ msoren/TreeRot.html/. SPALIK, K., A. WOJEWO´DZKA, and S. R. DOWNIE. 2001a. Delimitation of genera in Apiaceae with examples from Scandiceae subtribe Scandicinae. Edinburgh Journal of Botany 58: 331–346. ———, ———, and ———. 2001b. The evolution of fruit in Scandiceae subtribe Scandicinae (Apiaceae). Canadian Journal of Botany 79: 1358–1374. ——— and S. R. DOWNIE. 2001. The utility of morphological characters for inferring phylogeny in Scandiceae subtribe Scandicinae (Apiaceae). Annals of the Missouri Botanical Garden 88: 270–301. STEFANOVIC´, S., L. KRUEGER, and R. G. OLMSTEAD. 2002. Monophyly of the Convolvulaceae and circumscription of their major lineages based on DNA sequences of multiple chloroplast loci. American Journal of Botany 89: 1510–1522. STEVENS, P. F. 2006. An end to all things? — plants and their names. Australian Systematic Botany 19: 115–133. SWOFFORD, D. L. 2002. PAUP* Phylogenetic analysis using parsimony (*and other methods), v. 4.0 beta 10. Sunderland: Sinauer Associates. TABERLET, P., L. GIELLY, G. PAUTOU, and J. BOUVET. 1991. Universal primers for amplification of three non-coding regions of chloroplast DNA. Plant Molecular Biology 17: 1105–1109. VAN STEENIS, C. G. G. J. 1964. Plant geography of the mountain flora of Mt. Kinabalu. Proceedings of the Royal Society of London, B, Biological Sciences 161: 7–38. WILLIAMS, D. M., M. C. EBACH, and Q. WHEELER. 2005. 150 reasons for paraphyly: a response. Taxon 54: 858. WOJCIECHOWSKI, M. F., M. LAVIN, and M. J. SANDERSON. 2004. A phylogeny of legumes (Leguminosae) based on analysis of the plastid matK gene resolves many wellsupported subclades within the family. American Journal of Botany 91: 1846–1862. APPENDIX 1. Taxa, voucher information [collector (herbarium acronym), country of origins (locality)], and GenBank/ EMBL accession numbers (atpB-rbcL, trnS-trnG, ITS). Oreomyrrhis Endl.: O. andicola, Chung 1493 (MO), Ecuador (Pichinca, Volcan Pinchinca), DQ829703, DQ829739, AM284398; O. argentea, Chung 1609 (MO), Australia (Tasma-

681

nia, Great Lake), DQ829704, DQ829740, AJ854310; O. azorellacea, Vink 16141 (CANB), Papua New Guinea (Western Highlands, Kubor Range), DQ829705, DQ829741, AM284399; O. basicola, Chung 1638 (MO), New Zealand (Otago, Waitaki Valley), DQ829706, DQ829742, AM284403; O. borneensis, Clemens 29809 (A), Malaysia (Borneo, Mt Kinabalu), DQ829707, DQ829743, AJ854312; O. brevipes, Chung 1571 (MO), Australia (New South Wales, Kosciuszko National Park), DQ829708, DQ829744, AJ854313; O. ciliata, Chung 1555 (MO), Australia (New South Wales, Barrintontops National Park), DQ829709, DQ829745, AJ854315; O. colensoi, Chung 1641 (MO), New Zealand (Taupo, Rangiariki River), DQ829710, DQ829746, AJ854318; O. daucifolia, Chung 1650 (MO), Guatemala (Huehuetenango, Sierra de los Cuchumatane), DQ829711, DQ829747, AJ854320; O. eriopoda, Chung 1553 (MO), Australia (New South Wales, Ebor Falls), DQ829712, DQ829748, AM284404; O. hookeri, Chung & Wu 1692 (MO), Falkland Islands (East Falkland, Stanley), DQ829713, DQ829749, AM284405; O. involucrata, Juan 53 (HAST), Taiwan (Taitung, Mt Hsiangyang), DQ829714, DQ829750, AM284406; O. linearis, Walker ANU 5024 (CANB), Papua New Guinea (Eastern Highlands, Mt. Wilhelm), DQ829715, DQ829751, AM284400; O. nanhuensis, Chung 1511 (MO), Taiwan (Ilan, Mt. Nanhu), DQ829716, DQ829752, AJ854332; O. orizabae, Avendan˜o R. 5349 (MO), Mexico (Veracruz, Cofre de Perote), DQ829717, DQ829753, AJ854333; O. papuana, Schodde 1945 (CANB), Papua New Guinea (Southern Highlands, Mt. Giluwe), DQ829718, DQ829754, AM284401; O. pulvinifica, Chung 1572 (MO), Australia (New South Wales, Kosciuszko National Park), DQ829719, DQ829755, AJ854335; O. pumila, Hoogland 9858 (CANB), Papua New Guinea (Morobe District, Huon Penisula), DQ829720, DQ829756, AM284402; O. ramosa, Chung 1617 (MO), New Zealand (Canterbury, Porter’s Pass), DQ829721, DQ829757, AJ854337; O. rigida, Chung 1635 (MO), New Zealand (Otago, Lake Hawea), DQ829722, DQ829758, AJ854341; O. sessiliflora, Chung & Song 1592 (MO), Australia (Tasmania, Ben Lomond National Park), DQ829723, DQ829759, AJ854342; O. taiwaniana, Chung 1525 (MO), Taiwan (Chiayi, Mt Jade), DQ829724, DQ829760, AM284407; Chaerophyllum sect. Chaerophyllum: C. procumbens, Chung 1551 (MO), USA (MO, St. Louis), DQ829734, DQ829770, AJ854344; C. tainturieri, Chung 1495 (MO), USA (MO, St. Louis), DQ829735, DQ829771, AJ854345; C. temulum, Anderberg s.n. (MO), Sweden (Uppland, Stockholm), DQ829736, DQ829772, AM284417; Sect. Physocaulis DC.: C. nodosum (L.) Crantz, Tabrizhjan s.n. (MO), Armenia (Megrinskii region, Mt. Tiumarants), DQ829733, DQ829769, AM284416; Sect. Dasypetalon Neilr.: C. hirsutum L., Mitka & Sztyler 450 (MO), Poland (Regio Subcarpatica, districtus Brzesko), DQ829732, DQ829768, AM284415; Sect. Chrysocarpum Spalik & S.R. Downie: C. aromaticum L., s.n. (MO), USSR (Region Kalininskaja, Turchinovsky), DQ829725, DQ829761, AM284408; C. aureum L., Anderberg s.n. (MO), Sweden (Uppland, Solna, Ra˚stasjo¨n), DQ829726, DQ829762, AM284409; C. azoricum Trel., Sjogren 6796 (MO), Portugal (Azores), DQ829731, DQ829767, AM284414; C. caucasicum (Fisch.) Schischk., Gagnidze 751 (MO), Georgia Republic (Minor Caucasus, Kartli, Bakuriani), DQ829727, DQ829763, AM284410; C. crinitum Boiss., Rechinger s.n. (MO), Iran (Persia, Mt Khali Kuh), DQ829730, DQ829766, AM284413; unclassified in Spalik and Downie (2001): C. humile Stev., Schneeweib 4416 (MO), Georgia Republic (Chewi, Kazbek), DQ829728, DQ829764, AM284411; C. roseum M.Bieb., MO-5548949, Georgia Republic, DQ829729, DQ829765, AM284412; Outgroup: Myrris odorata (L.) Scop., Anderberg s.n. (MO), Sweden (Uppland, Spa˚nga parish, Va¨llingby), DQ829737, DQ829773, AM284418; Osmorhiza longistylis (Torr.) DC., Chung 1552 (MO), USA (MO, St. Louis), DQ829738, DQ829774, AM284419.