Freshwater turtles are one of the most threatened groups of vertebrates in the world (Rhodin et al. 2018), with 61% of species listed as globally threatened (Critically Endangered, Endangered, or Vulnerable) by the International Union for Conservation of Nature (Turtle Taxonomy Working Group 2017; Lovich et al. 2018). Declining turtle populations have been linked to a series of threats, including habitat destruction and fragmentation, the commercial food and pet trade, invasive species, and incidental bycatch (Stanford et al. 2020). Increasingly, molecular data are being used to help inform management efforts on species of conservation concern (Sethuraman et al. 2014; Spinks et al. 2014; Pierson et al. 2016; Grueber et al. 2018; Gallardo-Alvarez et al. 2019). Such studies often employ microsatellite loci to characterize population genetic structure, detect genetic bottlenecks, estimate effective population sizes, and establish kinship among individuals (Selkoe and Toonen 2006).

The alligator snapping turtle (Macrochelys temminckii) is the largest freshwater turtle in North America, inhabiting rivers and associated oxbow lakes of the southeastern United States. Significant population declines have been reported as a result of historical commercial harvest (Pritchard 1989; Reed et al. 2002). Currently, M. temminckii is fully protected in all states it inhabits, with the exception of Mississippi and Louisiana (NatureServe 2021), and it has been proposed for Threatened status under the US Endangered Species Act (US Fish and Wildlife Service [USFWS] 2021a). Thomas et al. (2014) described two new species from the Apalachicola River (Apalachicola alligator snapping turtle, M. apalachicolae) and Suwannee River (Suwannee alligator snapping turtle, M. suwanniensis). The latter has been proposed as federally Threatened because of poaching, bycatch on fishing equipment, and nest predation throughout its limited geographic range (USFWS 2021b).

Nine microsatellite loci have been previously published for M. temminckii that were tested using samples from 8 river systems, where Hackler et al. (2007) found an overall moderate level of allelic diversity. However, Echelle et al. (2010) reported that, at the level of individual river systems, these 9 microsatellite loci had few alleles per locus and low levels of heterozygosity. These microsatellite loci demonstrated relatively low allelic variation, so there is a need for additional loci to provide greater power to detect fine-scale population genetic structure and perform kinship analyses. Additionally, better assessment of genetic structure can also be informative when attempting to reintroduce seized turtles to their appropriate drainages (Biello et al. 2021).

In the present study, we screened 60 new microsatellite loci to determine amplification in M. temminckii and cross-amplification in M. suwanniensis and Chelydra serpentina (eastern snapping turtle). We then characterized 45 of these loci for M. temminckii in the hopes of providing the tools necessary to pursue future population genetic studies in M. temminckii and other chelydrid species.

Methods. — From 2017 to 2019, we used partially submerged hoop nets (90-cm and 120-cm diameter) baited with fresh or frozen fish to capture M. temminckii at two adjacent oxbow lakes within the Pascagoula River drainage in George County, Mississippi. Blood (0.5 to 1 ml) was collected from the dorsal coccygeal vein of all captured M. temminckii using a 23-gauge, 2.5-cm needle and stored in SED buffer (Seutin et al. 1991). Total genomic DNA from 2 individuals was extracted using the DNeasy Tissue Kit (QIAGEN Inc) and sent to the Savannah River Ecology Lab Molecular Ecology Lab where they prepared an Illumina pair-end shotgun library for sequencing (Lance et al. 2013). PAL_FINDERv0.02.03 (Castoe et al. 2012) was used to identify di-, tri-, tetra-, and pentanucleotide microsatellite sequences (i.e., nucleotide repeat motifs) among the reads. Primer3 (Rozen and Skaletsky 1999) was used to develop primers for 942 potential microsatellite loci. We screened 60 penta-, tetra-, and trinucleotide loci, focusing on loci whose sequences appeared only 1 or 2 times within the sequence reads, to reduce the likelihood of targeting repetitive loci.

We initially screened the 60 loci on 5 M. temminckii, 3 M. suwanniensis (provided by D. Stevenson, H. Chandler, B. Stegenga, and K. Enge; Alapaha River, Echols County, Georgia), and 2 C. serpentina (Itawamba County, Mississippi) to determine which loci successfully amplified in each of the 3 species. Loci that amplified in each of the respective species were then tested in 53 M. temminckii, 14 M. suwanniensis, and 19 C. serpentina to assess variability in each locus. Genomic DNA was extracted from each individual using a DNeasy Tissue Kit, and amplifications of samples were conducted in a total volume of 12.5 µl using 8.445 µl of dH₂O, 1.25 µl of 10× standard Taq (Mg-free) buffer (New England BioLabs), 0.25 µl of 2 mM dNTPs, 1 µl of 25 mM MgCl₂, 0.08 units Taq polymerase (New England BioLabs), 0.20 µl of 10 mM forward tailed (Boutin-Ganache et al. 2001) and reverse primers, 0.075 µl of 1 µM labeled M-13 primer (Eurofins), and 1 µl of 20–50 ng of DNA template. Polymerase chain reaction (PCR) cycling conditions were as follows: initial denaturation at 94°C for 2 min, 35 cycles for 30 sec at 94°C, 30 sec at 56°C, and 1 min at 72°C, with a final elongation of 10 min at 72°C. Microsatellite alleles were visualized on an acrylamide gel using a LI-COR 4300 DNA Analyzer and allele sizes were determined using GeneProfiler ver. 4.05 (LI-COR Inc, Lincoln, NE). We used the amplified fragments of the Lambda phage as a size standard (Wang et al. 2010).

GenAlEx 6.51 (Peakall and Smouse 2012) was used to calculate number of alleles (N_A) and observed (H_O) and expected (H_E) heterozygosity. Hardy-Weinberg equilibrium and linkage disequilibrium were tested using the probability tests of GENEPOP on the web (Raymond and Rousset 1995; Rousset 2008), while we assessed the presence of null alleles using MICRO-CHECKER v. 2.2.3 (van Oosterhout et al. 2004). A sequential Bonferroni correction (Rice 1989) was used to adjust the alpha values for multiple comparisons.

Results and Discussion. — Sequence data for the loci isolated in the present study can be found in the National Center for Biotechnology Information Sequence Read Archive listed under BioProject accession PRJNA688648 and BioSample number SAMN17185309 (https://www.ncbi.nlm.nih.gov/bioproject). A total of 47 loci amplified for at least 1 of the 3 species (Table 1). Of these, 45 loci amplified for M. temminckii (7 monomorphic), 45 loci amplified for M. suwanniensis (32 monomorphic), and 40 loci amplified for C. serpentina (8 monomorphic). Two loci (Mtem038 and Mtem047) that amplified for C. serpentina and one locus (Mtem060) for M. suwanniensis require additional optimization by future users before they can be scored with confidence; these 3 loci are not considered monomorphic or polymorphic within this article.

Table 1. Characterization of 47 microsatellite loci isolated for the alligator snapping turtle (Macrochelys temminckii), including the repeat motif and forward (F) and reverse (R) primer sequences. Forward primers are reported without the additional M13 tail. Amplification results for M. temminckii (Mtem), M. suwanniensis (Msuw), and Chelydra serpentina (Cser) are indicated as + (amplification) or – (no amplification).

View Fullscreen

Table 1. Continued.

View Fullscreen

Table 2 provides the number of alleles and observed and expected heterozygosity for each locus in each species. The averages across polymorphic loci are as follows: M. temminckii N_A: 3.58 (range, 2–8), H_O: 0.483 (range, 0.029–0.860), and H_E: 0.455 (range, 0.029–0.762); M. suwanniensis N_A: 2.67 (range, 2–5), H_O: 0.220 (range, 0.000–0.385), and H_E: 0.274 (range, 0.069–0.594); C. serpentina N_A: 8.23 (range, 2–18), H_O: 0.607 (range, 0.105–1.000), and H_E: 0.661 (range, 0.100–0.911). Hardy-Weinberg Equilibrium was violated for 2 loci in both M. temminckii (Mtem012 and Mtem054) and C. serpentina (Mtem021 and Mtem055). Six pairs of loci in M. temminckii exhibited linkage disequilibrium after a sequential Bonferroni correction: Mtem006/010, Mtem006/033, Mtem010/033, Mtem048/054, Mtem052/057, and Mtem057/060. No null alleles were detected for M. temminckii, but potential null alleles were identified at Mtem010, Mtem019, Mtem021, and Mtem055 for C. serpentina and Mtem032 for M. suwanniensis.

Table 2. Results of testing the 45 alligator snapping turtle (Macrochelys temminckii) microsatellite loci on 53 M. temminckii, 14 Suwannee alligator snapping turtles (M. suwanniensis), and 19 eastern snapping turtles (Chelydra serpentina). For each locus, we report the number of alleles (N_A), observed (H_O) and expected (H_E) heterozygosity, and size range (in base pairs [bp]), including the 19 base pairs from the M13 tail added to the forward primer. — = no calculated value due to a monomorphic locus or a locus that did not amplify; * = loci that amplified but require additional optimization for confidence in genotyping. Loci that failed to amplify are listed as NA under size range.

View Fullscreen

We are in the process of using the polymorphic loci identified in the present study to assess patterns of population structure within M. temminckii across broad and fine spatial scales in the state of Mississippi. Some of these loci proved to be highly polymorphic in C. serpentina and should be useful in future population genetic studies of this species. Conversely, our loci were mostly monomorphic (71% of the loci that amplified) in M. suwanniensis, which may reflect its limited geographical range.