East Vs. West: Monophyletic Clades Within the Paraphyletic Carex Acuta Complex, Section Phacocystis (Cyperaceae) Julie A
Total Page:16
File Type:pdf, Size:1020Kb
10.Dragon 11/12/08 12:36 PM Page 215 Chapter 10 East vs. West: Monophyletic Clades Within the Paraphyletic Carex acuta Complex, Section Phacocystis (Cyperaceae) Julie A. Dragon and David S. Barrington ABSTRACT The Carex acuta L. complex, as defined by Standley, comprises 12 species from section Phacocystis Dumort., one of the largest sections within the genus Carex L. The species of the complex were originally distinguished by their nerved, stipitate perigynia with torulose bases adnate to iridescent achenes. Internal transcribed spacer (ITS) and external transcribed spacer fragment (ETS 1f) sequence data were used to test the phylogenetic integrity of the C. acuta complex, examine infraspecific variation within two of its polymorphic species, C. nigra (L.) Reichard and C. lenticularis Michx., and identify biogeographic patterns indicated by the phylogeny. The C. acuta complex was found to be paraphyletic as circumscribed. The newly discerned monophyletic groups comprise four clades: (1) an Austral–East Asian group; (2) a Eurasian clade; (3) a C. bigelowii Torr. ex Schwein.– C. scopulorum T. Holm–C. stricta Lam. clade; and (4) a clade of North American species, plus C. aquatilis Wahlenb. from Finland. Within this latter clade, western and eastern North American taxa are further divided into monophyletic groups. The two western varieties of C. lenticularis examined were more closely related to two Mexican taxa and the circum- boreal C.aquatilis than they were to the eastern variety of C. lenticularis, which formed a clade with the amphi- Atlantic C.rufina Drejer. In contrast, C. nigra was found to be monophyletic, with little sequence divergence among American and European accessions of C. nigra or between them and their European allies. High infra- specific variation was found within C. eleusinoides Turcz. ex Kunth and C. stricta. Poor species circumscription and hybridization were considered as possible causes of this pattern. KEY WORDS Biogeography, Carex acuta complex, ETS, ITS, phylogeny, polymorphic species, reticulate evolution, section Phacocystis. 215 10.Dragon 11/12/08 12:36 PM Page 216 216 Sedges: Uses, Diversity, and Systematics of the Cyperaceae he genus Carex L. is one of the most wide- zation are possible explanations for the taxonomic spread and ecologically important of all plant complexity of these species. Tgenera (Reznicek, 1990). The diversity repre- The goal of the research described here was to sented by this genus of approximately 2000 species use cladistic analysis of the ITS region and an exter- has challenged systematists for over a century. nal transcribed spacer fragment (ETS 1f) of riboso- Convergence and homoplasy, likely a consequence mal DNA (rDNA) to elucidate the phylogeny of the of extreme floral reduction, have hampered phyloge- species placed in the Carex acuta complex. Our spe- netic analysis and have complicated taxonomic cir- cific objectives were to (1) test the monophyly of the cumscription. Given the challenges presented by complex, (2) assess the infraspecific variation within morphological characters in Carex, systematists the highly polymorphic species C. nigra and C. have increasingly turned to molecular data for lenticularis, and (3) interpret biogeographic history increased phylogenetic resolution. The few molecu- in the light of the retrieved phylogeny. lar analyses completed to date have shed new light on subgeneric, sectional, and species-level relation- MATERIALS AND METHODS ships within Carex (Starr et al., 1999; Yen & Olmstead, 2000; Roalson et al., 2001). This research SAMPLING addresses problems of species phylogeny and delim- Forty-one accessions representing 23 species itation in a widespread and problematic group, the were included in the analysis. Voucher data and C. acuta L. complex of section Phacocystis Dumort. GenBank accession numbers for all accessions are The Carex acuta complex was distinguished by provided in Table 1. All the members of Standley’s Standley (1987a, b) based on the shared characters Carex acuta complex were sampled, with the excep- of conspicuously nerved stipitate perigynia with tion of C. decidua Boott from South America and C. swollen torulose bases adnate to whitish iridescent cuchumatanensis Standl. & Steyerm. from Guate- achenes and chromosome numbers ranging from n = mala, for a total of 10 species. 42 to 46, unique within section Phacocystis. Roalson Because the choice of outgroup taxa was prob- et al. (2001), in their analysis of sections and subgen- lematic, we chose a diversity of taxa that might be era of Carex using internal transcribed spacer (ITS) expected to be sister groups to the Carex acuta com- and chloroplast DNA (cpDNA) sequences, included plex. The outgroup taxa we sampled fall into three nine taxa of Phacocystis, three of which were mem- groups. First, we sampled five taxa from section bers of the C. acuta complex. Although they found Phacocystis. Two additional unnamed Mexican taxa section Phacocystis to be paraphyletic, they provid- that share morphological features with the members ed preliminary molecular evidence that the C. acuta of section Phacocystis and the C. acuta complex complex was monophyletic. were also included in this first group. Second, fol- One of the key barriers to developing strong lowing the suggestion of Standley (1987a) that the phylogenetic hypotheses for carices is the problem taxa of section Bicolores Tuck. were sister to those of species circumscription. In the Carex acuta com- of the C. acuta complex, C. bicolor All. and C. aurea plex, two species are prominent in this regard: C. Nutt. were included as outgroup taxa. Third, taxa of lenticularis Michx. and C. nigra (L.) Reichard. Both other sections that share morphological features with taxa are highly variable morphologically depending members of section Phacocystis were also sampled. on their habitat, elevation, and latitude. Where These include (1) C. lemanniana Boott and C. don- morphs overlap, intermediate populations are often nell-smithii L. H. Bailey (Central America) of sec- found. While no confirmed hybrids involving C. tion Fecundae Kük., and (2) the North American C. lenticularis have been identified, C. nigra has been bella L. H. Bailey of section Racemosae G. Don and shown to produce fertile hybrids with a number of the amphi-Pacific C. podocarpa R. Br. of section species of section Phacocystis, including C. acuta, Scitae Kük. Carex podocarpa was especially impor- C. elata All., C. bigelowii Torr. ex Schwein., C. stric- tant to include in this study because Roalson et al. ta Lam., and C. aquatilis Wahlenb. (Faulkner, 1972, (2001) found it to be inserted within section 1973). Thus both phenotypic variation and hybridi- Phacocystis. 10.Dragon 11/12/08 12:36 PM Page 217 Monophyletic Clades Within the Carex acuta Complex, Section Phacocystis (Cyperaceae) 217 DNA EXTRACTION AND SEQUENCING used to elute the DNA from the gels. Dye-terminator Total genomic DNA was extracted from fresh reaction mixtures of 20 µl were made with the puri- material using the DNEasy Plant Mini Kit (Qiagen, fied DNA templates. The reactions were pretreated Valencia, California), and from silica-dried material in the thermocycler for five minutes at 96°C. Cycle with a CTAB buffer following the protocol of Doyle sequencing of the reactions comprised 25 cycles of and Doyle (1987). DNA denaturation at 96°C for 10 seconds, primer The polymerase chain reaction (PCR) was used annealing at 50°C for five seconds, and DNA strand to amplify ITS and 5.8S rDNA using a 1:1 ratio of extension by Taq DNA polymerase at 60°C for four primers ITS-4 (White et al., 1990) and AB101 (Sun minutes. Cycle sequencing of ITS 1 used the exter- et al., 1994) (Table 3). Each 25-µl reaction mixture nal sequencing primers ITS-L and internal sequenc- contained 2.5 µl of 10× reaction buffer, 5 µl of 5× Q- ing primer ITS-2, while cycle sequencing of ITS 2 solution (Qiagen, Valencia, California), 0.5 µl used external sequencing primer ITS-4 and ITS-3 dNTPs (10 mM each), 1.25 µl of both primers (10 (Table 3) (White et al., 1990; Hsiao et al., 1994). µM concentrations), 0.125 µl Taq (5 units/µl), 23.75 Cycle sequencing of ETS 1f used the same primers used in PCR. Automated sequencing was used to µl ddH2O, and 10–50 ng of DNA template. Double- stranded PCR products were produced in an process the amplified templates (Vermont Cancer Amplitron II Thermolyne thermocycler (Barnstead, Center, Burlington, Vermont). Sequence chro- Dubuque, Iowa). The DNA was pretreated at 96°C matograms were proofed by inspection and edited for two minutes followed by 32 cycles of DNA using Sequence Navigator 1.0 (Perkin-Elmer, denaturation at 96°C for one minute, primer anneal- Wellesley, Massachusetts). Sequences were initially ing at 50°C for one minute, and DNA strand exten- aligned by Sequence Navigator, followed by addi- sion by Taq DNA polymerase at 72°C for two min- tional alignment by inspection. The aligned utes. The PCR was terminated by a final extension of sequence data were analyzed for biases and ambigu- 72°C for seven minutes. ities. Indels were coded as present or absent and incorporated into the sequence data. The poly-A tail PCR was also used to amplify the 3' end of the at the 3 end of ITS 1 was excluded from pairwise ETS-1f using a 1:1 ratio of primers ETS-1F and 18S- ' sequence-divergence and phylogenetic analysis. A R (Starr et al., 2003; see Table 3). Each 25-µl reac- single potentially informative site from the 5 end of tion mixture contained 1.25 µl of 100% glycerol, ' 5.8S was included in the analyses. Ambiguous 11.5 µl of 100 mM KCl, 2.3 µl of 100 mM Tris HCl, nucleotides were scored as such in the data set. 2.3 µl of 25 mM MgCl2, 3.8 µl of 5× Q-solution (Qiagen, Valencia, California), 0.575 µl dNTPs (10 mM each), 0.575 µl of both primers (20 µM concen- PHYLOGENETIC ANALYSIS trations), 0.125 µl Taq (5 units/µl), and 10–50 ng of Maximum parsimony was used to conduct a DNA template.