The snapping turtle (Chelydra serpentina) is a long-lived species found from southern Canada east of the Rocky Mountains southwards to the Gulf Coast and Florida (Ernst and Lovich 2009). The International Union for the Conservation of Nature (IUCN) lists the species as Least Concern. However, its life history makes it vulnerable to over-harvesting (van Dijk 2011). Reliable microsatellite markers for this species may provide valuable information for potential conservation initiatives. We use 454 sequencing to develop novel microsatellite markers, and cross-amplify nine microsatellite markers previously developed for the alligator snapping turtle (Macrochelys temminckii; Hackler et al. 2006).

For microsatellite development, we isolated genomic DNA from snapping turtle blood using phenol–chloroform procedure (Sambrook et al. 1989) and cleaned the DNA using EtOH precipitation. We obtained a partial genomic library by sequencing on a Roche GS Junior (Roche, Branford, CT) at Trent University’s Natural Resources DNA Profiling and Forensics Centre. The run produced 127,778 sequences averaging 423 base pairs, which we searched for tri- and tetra-nucleotide microsatellites using the program MSTACOMMANDER (Faircloth 2008). We examined potential target loci by eye to identify loci with appropriate flanking regions for primer design and designed 40 primer pairs using Primer 3 (Rozen and Skaletsky 2000). We labelled forward primers with a fluorescent 5′ M13 tail and labelled reverse primers with a 5′ pigtail (GTTTCTT; Brownstein et al. 1996) to facilitate adenylation.

We collected blood samples from 127 C. serpentina from across southern Ontario by caudal venipuncture and stored the blood on FTA cards (Whatman, Inc.). We extracted genomic DNA from each card following the method suggested by Smith and Burgoyne (2004) for samples with nucleated erythrocytes.

We ran a temperature gradient with two DNA samples at each locus and used the optimal temperature for all subsequent PCR reactions. Amplification followed the methods of Schuelke (2000), using 4.0 μL of M13-labelled forward primer, 0.66 μL each of pigtailed reverse primer (Eurofins MWG Operon) and a 6-carboxyfluorescein dye (6-FAM; Eurofins MWG Operon) and 1.0 μL of DNA eluate (6–9 ng). We also tested cross-amplification of nine microsatellite loci developed for the alligator snapping turtle (Macrochelys temminckii; Hackler et al. 2006) using a 12.5 μL PCR reaction containing 0.6 μL of forward primer and 1.0 μL each of reverse primer and 6-FAM. Two of the nine alligator snapping turtle loci amplified successfully in the snapping turtle. PCR cycling parameters followed King and Julian (2004) with annealing temperatures adjusted for each locus as summarized in Table 1. We visualized length of the amplified fragments using a 3730 DNA Analyzer (Applied Biosystems) with GS(500) Liz (Applied Biosystems) as a size standard and scored genotypes using GeneMarker (SoftGenetics, State College, PA). One to two homozygous samples were subsequently sequenced at each locus to confirm identity of the amplified fragments, and five percent of our sampled individuals were genotyped twice at each locus to assess genotyping error.

Table 1 Characteristics of ten novel and two cross-amplified microsatellite loci for 127 Chelydra serpentina sampled from across southern Ontario
Full size table

We successfully amplified ten novel loci and cross-amplified two of the nine alligator snapping turtle loci (MteD9 and MteD111). We genotyped 127 C. serpentina from Ontario to characterize these 12 loci. We found no ambiguities in the genotypes of the individuals amplified and genotyped multiple times. Sequencing of homozygotes confirmed the identity and motifs of the amplified fragments.

We used Genalex v6.0 (Peakall and Smouse 2006) to quantify the number of alleles per locus (k), calculate observed and expected heterozygosity (Ho and He) and probability of identity (PI) for each locus. We used Genepop 4.0.10 (Raymond and Rousset 1995; Rousset 2008) to test for linkage disequilibrium and deviations from Hardy–Weinberg equilibrium (HWE).

The number of alleles per locus ranges from 1 to 14. Heterozygosity ranges from 0.157 to 0.850. Table 1 summarizes the primer sequences, amplification conditions and characteristics of the 12 characterized loci. None of the 66 pairwise comparisons between loci show evidence of linkage disequilibrium after Bonferroni correction for multiple comparisons. Two loci (Cs18 and MteD111) show significant deviations from HWE (p < 0.01).

Three alleles at the tri-nucleotide locus Cs16 differ by only one base pair. Sequencing of homozygous individuals and successful replication of these genotypes in independent amplifications both demonstrate these are unique alleles and are not the result of stutter. One locus (Cs14) is monomorphic in our samples. Because they are all from the northern limits of this species’ range low genetic diversity is expected, but it may be variable in southern populations.

The family Chelydridae contains two other species (Phillips et al. 1996): the Central American Snapping Turtle (Chelydra rossignoni), listed by the IUCN as Vulnerable, and the South American snapping turtle (C. acutirostris), which remains to be assessed. Cross-amplification of these markers could facilitate conservation genetic analyses of these two closely related and poorly understood species.