Carex L. is a taxonomically challenging, cosmopolitan genus comprising approximately 2000 species (Reznicek, 1990), many of which possess unusually small (Nishikawa et al., 1984) but labile genomes (Lipnerová et al., 2013). This complexity presents challenges at all taxonomic levels. Carex sect. Albae (Asch. & Graebn.) Kük., like most Carex sections, has no microsatellite markers developed to address evolutionary dynamics among recently diverged species, where many taxonomic issues occur. One small but challenging group is the C. eburnea–C. mckittrickensis complex. Species boundaries between C. eburnea Boott and C. mckittrickensis P. W. Ball are unclear based on randomly amplified inter-simple sequence repeat (ISSR) markers (Gillespie, 2005) and on trnS^(GCU)-trnG^(UUC) and 3′trnV^(UAC)-ndhC chloroplast intergenic spacer data (E. Gillespie, Marshall University, unpublished data). Additionally, morphological characters vary continuously (Ball, 1998) across the two species, making this taxon an excellent target for microsatellite marker development.

Carex eburnea is a diploid species (Löve, 1981) that occurs across North America, from Alaska to Newfoundland and southward into the Ozark Mountains, the Cumberland Plateau, and the southern Appalachian Mountains. Disjunct populations occur in the southern Appalachian Mountains and in the Sierra Madre Mountains in Mexico. Based on herbarium specimens and fieldwork (by E.L.G.), C. eburnea occurs nearly exclusively on limestone and exists on rock outcrops, in cedar glades and bogs, and in treeless habitats such as alvar and tundra. Co-occurring dominant tree species include spruce (Picea A. Dietr. spp.) in the American Northwest and northern white cedar (Thuja occidentalis L.) in the upper Midwest and eastern North America. In the southwestern United States and in Mexico, C. eburnea co-occurs with junipers (Juniperus L. spp.) and oaks (Quercus L. spp.). The closest relative of C. eburnea is C. mckittrickensis, which occurs at a single locality in the Guadalupe Mountains National Park (Culberson County, Texas, USA). Two Eurasian species (C. alba Scop. and C. ussuriensis Kom.) are the only other members of Carex sect. Albae. Development of microsatellite markers will be helpful in clarifying the species boundaries and evolutionary history of this recently diverged, widespread, limestone-limited lineage and could be useful within the two Eurasian members of Carex sect. Albae.

METHODS AND RESULTS

DNA was extracted from one individual of C. eburnea using a QIAGEN Plant Mini Kit (QIAGEN, Valencia, California, USA) (Appendix 1). A microsatellite sequencing library (MiSeq v2 protocol) was constructed and 2 × 250 paired-end sequencing was performed on an Illumina MiSeq at the Cornell Life Sciences Sequencing and Genotyping Facility (Ithaca, New York, USA). A total of 2,093,696 raw sequence reads (GenBank Short Read Archive accession SRA557216) were trimmed to remove vectors and low-quality sequence. The resulting reads were queried by MSATCOMMANDER version 1.0.8 (Faircloth, 2008) with default settings, except that mononucleotide repeats were not included in the search, minimum primer size was set at 20 bp, maximum primer GC content was limited to 50%, and a PIG-tail sequence (GTTT) (Brownstein et al., 1996) was added to one primer. Out of 312,744 identified microsatellites, unique DNA suitable for primer design flanked 89,413.

Table 1.

Characteristics of 16 microsatellite primer pairs developed for Carex eburnea.

Forty-eight primer pairs were selected and screened in seven C. eburnea individuals (Appendix 1), prioritizing motif diversity and melting temperature difference ≤1°C. PCRs were prepared in a 10-µL reaction consisting of 1× Go Taq Flexi Buffer, 2.5 mM MgCl₂, 800 µM dNTPs, 0.5 µM each primer, 0.5 units Go Taq Flexi DNA Polymerase (Promega Corporation, Madison, Wisconsin, USA), and ∼20 ng DNA. PCR was completed using a touchdown thermal cycling program on an Eppendorf Mastercycler (Eppendorf, Hauppauge, New York, USA) or an MJ Mini Thermal Cycler (Bio-Rad, Hercules, California, USA) with annealing temperatures ranging from 68°C to 55°C. Initial denaturation was 94°C for 5 min, followed by 13 cycles (45 s at 94°C, 2 min at touchdown temperature, and 1 min at 72°C), followed by 24 cycles (45 s at 94°C, 1 min at 55°C, and 1 min at 72°C), followed by 5 min at 72°C. PCR products were examined on a 1% agarose gel in 1× TBE and scored for the presence or absence of an appropriately sized PCR product and uniform amplification. Sixteen primer pairs produced repeatable amplicons across all seven individuals. These 16 pairs were screened for polymorphisms in 68 individuals from three populations (Appendix 1).

Table 2.

Descriptive statistics for 14 polymorphic microsatellite loci in Carex eburnea.^a

Fig. 1.

Principal coordinates analysis (PCoA) of 68 Carex eburnea individuals and 14 microsatellite loci. Axis loadings for coordinates 1, 2, and 3 were 16.6%, 15.5%, and 7.3%, respectively. ♦ = Johnson Co., Tennessee, USA; ▪ = southeast Fairbanks Co., Alaska, USA; ▴ = Rockbridge Co., Virginia, USA.

PCR reaction conditions for screening polymorphisms were the same as above, except that the forward primer concentration was reduced to 0.25 µM and replaced with 0.25 µM M13 primer (5′-CACGACGTTGTAAAACGAC-3′), labeled with 6-FAM, VIC, NED, or PET (Life Technologies, Grand Island, New York, USA). PCR products labeled with different fluorescent dyes were pooled in equal amounts, and 2 µL of the pooled reactions were submitted along with a GeneScan 500 LIZ Size Standard (Life Technologies) for genotyping on an ABI 3730xl DNA Analyzer at the Georgia Genomics Facility (Athens, Georgia, USA). Resulting chromatograms were scored using Geneious 9.1.5 (Kearse et al., 2012; Biomatters Ltd., Auckland, New Zealand). Genotypic data were analyzed using GenAlEx version 6.503 (Peakall and Smouse, 2006, 2012) to obtain standard descriptive statistics, to test for deviations from Hardy-Weinberg equilibrium (HWE) assumptions, to examine the utility of the markers to distinguish among populations, and to evaluate the level of clonality within each population.

Table 3.

Cross-amplification of 14 primer pairs in additional representatives from Carex section Albae.^a

Of the 16 primer pairs, 14 loci revealed chromatograms that were consistent with a diploid taxon (Table 1), and two markers (CEB016 and CEB024) did not amplify consistently across all populations. The number of alleles per locus ranged from one to seven with an average of 3.071 across all three populations (Table 2). Observed heterozygosity ranged from 0.0 to 0.952 (mean 0.202). Twelve (86%) loci failed to meet the expectations of HWE in at least one population. Of these, four (29%) loci failed to meet HWE assumptions in all three populations. In almost all cases, excess homozygosity is evident, which may indicate inbreeding or genetic drift. Genetic distance followed by principal coordinates analysis (Orloci, 1978) (Fig. 1) demonstrated that the 14 loci distinguish among the populations, with the first three axes explaining 39.4% of the variation. A multilocus match analysis (Peakall and Smouse, 2006, 2012) revealed no identical individuals across all 14 loci within or among populations.

Cross-amplification of 14 primer pairs was conducted on three additional C. eburnea population representatives from across the range (Arkansas, USA; Ontario, Canada; and Querétaro, Mexico), five C. mckittrickensis individuals (all from the only known locality in Texas), and single representatives of C. alba and C. ussuriensis (Table 3). Twelve primer pairs amplified well in all three additional C. eburnea representatives (the remaining two pairs failed in two different C. eburnea individuals). All but two individual reactions were successful in the C. mckittrickensis individuals. Eight and 10 primer pairs cross-amplified successfully in the more distantly related C. alba and C. ussuriensis, respectively.

CONCLUSIONS

The markers reported here will likely be useful in population studies within C. eburnea; despite elevated levels of homozygosity generally, these markers discriminated among three populations (including two from the same physiographic region). Cross-amplification experiments confirmed that these markers should be applicable in the C. eburnea–C. mckittrickensis species complex and potentially in additional members of Carex sect. Albae, providing a novel population genetic tool in Carex.