INTRODUCTION

The Catspaw, Epioblasma obliquata, was listed as endangered under the U.S. Endangered Species Act in 1990; at that time, two isolated nonreproducing populations were known, one in the Green River in Kentucky and the other in the Cumberland River in Tennessee (USFWS 1990). These two populations now are considered extirpated. However, in 1994, a population of reproducing E. obliquata was discovered in a short reach of Killbuck Creek, a tributary of the Walhonding River in the Muskingum River watershed in Coshocton County, Ohio (Hoggarth et al. 1995). State and federal agencies are using this population as a source of broodstock for captive propagation in an attempt to recover the species.

Given the single-source population, genetic variation in hatchery progeny is a concern. Potential genetic threats to survival of the species include loss of within-population genetic variation from nonrepresentative sampling or low numbers of broodstock and family-size variation in the hatchery (Jones et al. 2006; Cooper et al. 2009). Microsatellites, or simple sequence repeats, are tandemly repeated motifs of multiple bases of nuclear DNA found in all eukaryotic genomes (Zane et al. 2002). Microsatellites are highly polymorphic loci that are ideally suited for genetic monitoring of wild and captive populations. The goal of this study was to develop and evaluate a set of microsatellite DNA PCR primers to analyze the genetic variation of the small source population in Killbuck Creek and any progeny produced in hatcheries.

METHODS

We obtained DNA samples from 44 adult Epioblasma obliquata that originally were collected from Killbuck Creek, Coshocton County, Ohio. These adults represented all individuals found at multiple sites and during multiple visits to the creek to collect broodstock in 2016–17. Adults were transported to and held at the Kentucky Department of Fish and Wildlife Resources' Minor E. Clark Fish Hatchery as part of the recovery program for the species. We nonlethally sampled these 44 individuals from the hatchery in the fall of 2018 by gently opening each mussel and vigorously swabbing the foot with a buccal swab (Kit DDK-50, Isohelix, Harrietsham, UK). From the buccal swab, DNA was isolated and extracted using an Isohelix DNA isolation kit, and its concentration and purity were assessed by using a lLite PC spectrophotometer (Biodrop, Cambridge, UK). In addition to morphological identification, the identification of all individuals as E. obliquata was confirmed using the mitochondrial DNA sequence from the first subunit of NADH dehydrogenase (ND1), a protein-encoding gene amplified by PCR using primers and conditions reported by Serb et al. (2003).

The Savannah River Ecology Laboratory at the University of Georgia developed a microsatellite library. Genomic DNA used to isolate the microsatellite loci was extracted from two individuals collected from the wild in 2016, utilizing a DNEasy Blood and Tissue Kit (Qiagen, Germantown, MD, USA). A genomic library was prepared with inserts size-selected to range from 300 to 600 bp. Paired-end reads were sequenced on an Illumina HiSeq sequencer. Using the program MSATCOMMANDER (Faircloth 2008), 463,713 reads containing 3–6 bp repeat motifs were identified. Primer3 (Untergasser et al. 2012) was used for PCR primer design. Initially, we screened 60 primer pairs on a panel of eight E. obliquata individuals and narrowed our evaluation to a set of 14 microsatellite polymorphic primer pairs. The criteria used to select these primer pairs were polymorphism of the loci amplified (i.e., observation of more than one allele), tri- or tetranucleotide repeat motif, and annealing temperature close to 598C for use in subsequent multiplexing. Forward primers were labelled with fluorescent markers as noted in Table 1. Four sets of loci were coamplified in multiplex PCR—Eoo11 and Eoo20; Eoo16 and Eoo19; Eoo22 and Eoo24; Eoo8, Eoo9, and Eoo10; other loci were amplified individually. PCR conditions consisted of H₂O, 5× PCR buffer (Promega, Madison, WI, USA), 2.5 mM MgCl₂ (Promega), 2.5 mM deoxynucleotide triphosphate (dNTPs) (ThermoFisher Scientific, Waltham, MA, USA), 1 mg/mL bovine serum albumin (BSA) (ThermoFisher Scientific), 5 lM of each primer, 0.1 lL GoTaq Polymerase (New England Biolabs, Ipswich, MA, USA), and 1 lL of genomic DNA at 50 ng/lL, in a total reaction volume of 22 lL. PCR thermal cycling conditions were as follows: 948C for 3 min, followed by 35 cycles of 948C for 40 s, 598C for 40 s, and 728C for 1 min; a final extension at 728C for 5 min; and a hold at 48C. Amplification of PCR products was verified by visualization under ultraviolet light in an ethidium bromide-stained agarose gel. PCR products were sent to the Institute of Biotechnology at Cornell University, Ithaca, New York, for DNA fragment-size analysis. Microsatellites were scored for length using Genemarker (SoftGenetics, State College, PA, USA). Arlequin v3.0 (Excoffier et al. 2005) was used to assess heterozygosity, number of observed alleles per locus, conformance to Hardy–Weinberg equilibrium (HWE), and linkage disequilibrium (LD). Testing for HWE and LD used Arlequin and a critical type I error rate = 0.05. Evidence for a bottleneck at each locus was tested using the Garza–Williamson index (M-ratio, the ratio of the number of alleles observed to the number of alleles possible within the observed range in allele sizes) using Arlequin; values of M below 0.7 suggest the occurrence of a bottleneck (Garza and Williamson 2001). MICROCHECKER 2.2.3 (Van Oosterhout et al. 2004) was used to assess the possibility of segregation of null alleles.

RESULTS AND DISCUSSION

Allelic diversity ranged from 2 to 15 alleles per locus and averaged 7.6 alleles per locus, while observed heterozygosity per locus ranged from 0.091 to 1.000 and averaged 0.674 (Table 1). Significant deviations from HWE, showing deficiencies in heterozygotes, were observed at 6 of the 14 loci, and LD was observed at 9 (∼10%) of the 91 pairwise comparisons among loci and involved 12 of the 14 total loci sampled (Eoo9 and Eoo19; Eoo11 and Eoo19; Eoo9 and Eoo22; Eoo20 and Eoo22; Eoo16 and Eoo24; Eoo11 and Eoo31; Eoo8 and Eoo44; Eoo31 and Eoo38; Eoo11 and Eoo60). The M-ratios for six loci were below 0.70, suggesting recent loss of allelic diversity at these loci. Possible segregation of null alleles was detected at loci Eoo16, Eoo20, Eoo22, and Eoo38. Because of the small size of the population sampled, deviations from HWE and LD could result from loci being closely located on the same chromosome, segregation of null alleles, family structure, population bottlenecks, inbreeding, hermaphroditic reproduction (van der Schalie 1970), or some combination of these factors. Appendix A1 lists individual genotypes at the 14 loci.

These primer pairs are the third set of microsatellite primers developed for the genus Epioblasma. The first set of primers was developed for Epioblasma capsaeformis (Jones et al. 2004) and the second for Epioblasma rangiana (Zanatta and Murphy 2006). We did not test primers developed for E. capsaeformis and E. rangiana on E. obliquata, but allelic diversity of E. obliquata was lower than in those two species. For the 10 loci developed for E. capsaeformis (n = 20 individuals assessed/locus), allelic diversity ranged from 5 to 17 alleles/locus and averaged 9.7 alleles/locus. For the six loci developed for E. rangiana (n = 73–86 individuals/locus), allelic diversity ranged from 12 to 28 alleles/locus and averaged 19.3 alleles/locus. After careful screening for null alleles, HWE, and LD, some of our microsatellite loci developed for E. obliquata may prove useful for cross-species amplification in other species, especially other Epioblasma. Likewise, future studies could screen the microsatellite loci developed by Jones et al. (2004) and Zanatta and Murphy (2006) to determine whether additional loci are suitable for cross-species amplification in E. obliquata.

Table 1.

Characteristics of 14 microsatellite loci developed using DNA obtained in 2016 and 2017 from 44 individuals of the Catspaw (Epioblasma obliquata) from Killbuck Creek, Coshocton County, Ohio. H_O and H_E are observed and expected heterozygosity, respectively. Statistically significant deviations from Hardy–Weinberg Equilibrium (HWE) are denoted by an asterisk (*). M-ratio is the Garza–Williamson index. Individual genotypes at the 14 loci are reported in Appendix A1.

Sampling more individuals of E. obliquata for further population genetic analysis would benefit conservation management. The screening of more wild individuals and any other populations that may be found could provide insight into the population genetic diversity and natural history of this species. Given the isolation and small size of the remaining known population of E. obliquata, these microsatellite loci and other genetic markers will be valuable for monitoring the effects of propagation and management practices seeking to maintain or increase genetic diversity in hatchery stocks and wild populations receiving stocked individuals. For example, if hatchery technology improves to allow for the long-term holding, spawning, and fertilization of broodstock in captivity, the loci developed in this study will be useful for monitoring genetic diversity and inbreeding in parental stocks and progeny, which will be critical for maintaining healthy captive and wild populations of E. obliquata (Jones et al. 2020).