The genus Oenothera L. (Onagraceae) has diversified across diverse habitats of North America with conservative shifts in pollinators (primarily between bees and hawkmoths; Raven, 1979) and more dramatic shifts in life history traits (Evans et al., 2009). Oenothera sect. Calylophus (Spach) Torr. & A. Gray (Onagraceae) consists of seven recognized species (13 taxa) divided into subsections Calylophus (Spach) W. L. Wagner & Hoch (O. capillifolia Scheele, O. gayleana B. L. Turner & M. J. Moore, and O. serrulata Nutt.) and Salpingia (Torr. & A. Gray) W. L. Wagner & Hoch (O. hartwegii Benth., O. lavandulifolia Torr. & A. Gray, O. toumeyi (Small) Tidestr., and O. tubicula A. Gray) (Wagner et al., 2007; Turner and Moore, 2014). Ring chromosomes have been documented in all taxa in sect. Calylophus (Towner, 1977), with only O. serrulata exhibiting permanent translocation heterozygosity (Johnson et al., 2014).
Oenothera gayleana and O. hartwegii subsp. filifolia (Eastw.) W. L. Wagner & Hoch are gypsum endemics that often cooccur in eastern New Mexico and western Texas, easily distinguished by floral characteristics associated with bee pollination and hawkmoth pollination, respectively (Towner, 1977; Turner and Moore, 2014). Because bees forage close to nesting sites (Greenleaf et al., 2007) while hawkmoths can travel great distances (Stockhouse, 1973; Alarcón et al., 2008), differentiation between populations is expected to differ between these two plant species (Finger et al., 2014). Here, we characterize 11 nuclear and four plastid microsatellite loci to be used to contrast pollen and seed dispersal patterns in O. gayleana and O. hartwegii subsp. filifolia. We also describe the transferability of these markers to all 11 other taxa in sect. Calylophus.
METHODS AND RESULTS
We tested a combination of nuclear and plastid microsatellite loci. We screened 36 unpublished nuclear microsatellite markers that were originally developed for O. biennis L., using the microsatellite library prepared by Larson et al. (2008) for studies of genotypic identification and herbivory (Agrawal et al., 2012). In addition, the plastid genome of O. elata Kunth subsp. hookeri (Torr. & A. Gray) W. Dietr. & W. L. Wagner (GenBank accession no. AJ271079; Hupfer et al., 2000) was screened for large strings of single nucleotide repeats. The plastid primers were designed for 12 microsatellite regions using the following settings in Primer3: optimum primer size 20 bp, melting temperature 60°C, and product size range of 100–300 bp (Untergasser et al., 2012).
Characteristics of 11 nuclear and four plastid microsatellite loci tested in Oenothera gayleana and O. hartwegii subsp. filifolia.
Both nuclear and plastid microsatellite regions were initially screened using three randomly selected individuals of three species in sect. Calylophus: O. serrulata (Crosbyton, TX), O. lavandulifolia (Iraan, TX), and O. hartwegii subsp. filifolia (Caballo Mountains, NM) (Appendix 1). DNA was extracted from field-collected leaf tissue (Appendix 1) using a modified cetyltrimethylammonium bromide (CTAB) method (Doyle and Doyle, 1987). For nuclear microsatellite marker amplification, we used a 10-µL reaction containing 5 µL MyTaq DNA polymerase (Bioline, London, United Kingdom), plus 0.125 µL bovine serum albumin (BSA; 0.5 ng/µL), 3.375 µL DNase-free water, 1 µL template DNA, and 0.25 µL of both forward and reverse primers. The forward primers were fluorescently labeled with WellRed D2 (black), D3 (green), or D4 (blue) (Sigma-Proligo, St. Louis, Missouri, USA). PCRs were run at 95°C for 2 min, then 30 cycles of 50 s at 95°C, 30 s at 56°C, and 1 min at 72°C, with a 10-min extension at 72°C. The plastid microsatellite primers were not fluorescently labeled but instead were amplified and labeled in two steps (Schuelke, 2000). The first PCR reaction mix was identical to above except that the forward primer was designed with an M13 sequence (5′-CAGGACGTTGTAAAACGAC-3′) added to the 5′ end. The PCR protocol was as follows: 94°C for 3 min, followed by 13 cycles of 40 s at 94°C, 40 s at 52°C, and 2 min at 72°C, with a final extension of 10 min at 72°C. For the second step, an additional 2.5 µL MyTaq DNA polymerase, 2.0 µL DNase-free water, and 0.5 µL of a labeled M13 forward primer (D2, D3, and D4) was added to each reaction to label any PCR products that contained M13 sequences. The second PCR performed another 27 cycles. The resulting PCR products were analyzed and scored using a 400-bp size standard on a CEQ 8000 Genetic Analysis System version 9.0 (Beckman Coulter, Brea, California, USA).
Of the 36 nuclear primer pairs screened, 14 did not amplify (GenBank accession no.: KT762974–KT762987), 10 amplified unreliably (GenBank accession no.: KT62988–KT62997), one was monomorphic (GenBank accession no.: KT762973), and 11 were polymorphic, one of which (Oenbi2diA_C10; Table 1) amplified two regions in O. hartwegii subsp. filifolia. These 11 polymorphic markers were further characterized using three populations of O. gayleana and two populations of O. hartwegii subsp. filifolia (10–30 individuals per population; Table 2). To test for cross-amplification, they were also tested on three to five individuals from one population of each of the remaining 11 taxa in Oenothera sect. Calylophus (Tables 3 and 4, Appendix 1).
For the nuclear microsatellites, we report the following parameters for two to three populations of O. gayleana and O. hartwegii subsp. filifolia: sample size (N), number of alleles (A), number of private alleles (A p), observed heterozygosity (H o), expected heterozygosity (H e), and deviation from Hardy–Weinberg equilibrium (HWE) (Table 2, calculated using GenAlEx; Peakall and Smouse, 2006). Significant deviation from HWE was observed in at least one population for eight primer pairs in O. gayleana and in four primer pairs in both populations of O. hartwegii subsp. filifolia (Table 2). Primer pairs were tested for linkage disequilibrium for each pair of loci within and across all populations using the log likelihood ratio statistic and Fisher's method in GENEPOP (Raymond and Rousset, 1995). No significant linkage disequilibrium (P < 0.01) was detected in either species, except two primer pairs (Oenbi2triA_D3 and Oenbi2triA_F5; Table 1) that share a reverse primer sequence and therefore are likely to be amplifying the same region. For each population, the presence of null alleles at each locus was determined using exact tests in MICRO-CHECKER (van Oosterhout et al., 2004). Any potential null alleles detected in MICROCHECKER corresponded with a primer pair that showed deviation from HWE (e.g., Oenbi2diA_E9). We suspect that these anomalies may be due to the presence of ring chromosomes, documented throughout sect. Calylophus (Towner, 1977), or the small number of samples included.
Of the 12 plastid regions tested, four amplified reliably and were polymorphic in the two focal species (Table 1). One region (OenelCp5) occasionally produced two peaks; this may be due to stutter or because this region is located within the inverted repeat in the plastid genome. The peak pairs were repeatable and consistent across individuals, hence only the largest peak was scored. Across all species, these four primer pairs identified 28 haplotypes, with one to seven haplotypes per population. Most haplotypes were unique to each population with the exception of one shared haplotype between O. lavandulifolia and O. hartwegii subsp. maccartii (Shinners) W. L. Wagner & Hoch and one between two populations of O. gayleana (Yeso 62/180 and Fort Sumner; Tables 3 and 4).
The 11 nuclear and four plastid microsatellite markers were polymorphic and reliable in O. gayleana and O. hartwegii subsp. filifolia and in some populations of the remaining 11 taxa within Oenothera sect. Calylophus. These markers will be used in future studies of genetic differentiation between populations in the bee-pollinated O. gayleana and the hawkmoth-pollinated O. hartwegii subsp. filifolia. In addition, they will be useful for investigations into gene flow within and among other taxa in sect. Calylophus and may help identify populations and species that exhibit translocation heterozygotes in this group.
Results of initial primer screening of 11 polymorphic nuclear microsatellite markers developed in Oenothera gayleana (three populations) and O. hartwegii subsp. filifolia (two populations).
Results of cross-amplification of nuclear microsatellites in the 11 additional taxa within Oenothera sect. Calylophus. Results from O. gayleana and O. hartwegii subsp. filifolia are included for comparison.
Results of cross-amplification of plastid microsatellites in the 11 additional taxa within Oenothera sect. Calylophus. Results from O. gayleana and O. hartwegii subsp. filifolia are included for comparison.
 The authors thank W. Daichendt, J. Medina, J. M. Keller, H. Flores Olvera, H. Ochoterena, and N. Douglas for laboratory and field assistance; the U.S. Bureau of Land Management and the USDA Forest Service for permits; and the Ecological Genetics Core Facility at Cornell University, specifically S. Bogdanowicz. Funding was provided by the National Science Foundation (DEB 1342873 to K.A.S. and J.B.F., DEB 1054539 to M.J.M., and DEB-1513839 to A.A.A.), the National Geographic Society (M.J.M.), Oberlin College, a Botanical Society of America Graduate Student Research Award (E.M.L.), and the Chicago Botanic Garden.