River otters (Lontra canadensis) were extirpated from New Mexico by the 1950s. A limited reintroduction occurred during 2008–2010 in which 33 otters sourced from Washington (WA) were translocated to the Upper Rio Grande Basin (URG) of New Mexico. We conducted a noninvasive genetic capture-recapture survey during the winter of 2018 by collecting fecal DNA samples from river otter scats found at latrines in the URG dendritic network of perennial waterways. Our objectives were to: 1) estimate genetic diversity and effective population size; 2) genetic divergence from the WA source population and potential connectivity with regionally proximal populations; 3) spatially explicit population density and size; and 4) population growth rate since the founder event. Between February and April 2018, we collected 1,184 fecal DNA samples from 622 individual scats at 20 latrines; genotyping was attempted at 10 otter-specific microsatellite loci for a subsample of 543 samples. A bottlenecking founder effect was strongly supported, which, combined with genetic drift, reduced genetic diversity and effective population size by 20–26% and 106–170%, respectively, compared with the WA source population. Estimated population density from spatial capture-recapture models was 0.23–0.28 otter/km of waterway, or 1 otter/3.57–4.35 km of waterway, corresponding to a total population size of 83–100 otters across 359 km of the perennial dendritic network from La Mesilla, New Mexico to Alamosa National Wildlife Refuge, Colorado. Estimated average annual population growth rate since the founder event was 1.12–1.15/year. Despite successful population establishment, the URG river otter population remains small, is genetically degraded, and does not yet meet the criteria for long-term reintroduction success. Projections suggested that the population could reach the recommended minimum viable population size of ≥400 otters by the years 2030–2033, though sufficient habitat may not exist in the URG Basin to support that many otters. 

Scat samples were collected from a reintroduced population of North American river otters (Lontra canadensis) via a capture-recapture survey design in the Upper Rio Grande River Basin of the southwestern, USA. Otter latrine sites were surveyed for 8 consecutive occasions, each of which were 7-10 days in duration, during February-April, 2018. Additional tissue samples were obtained from the reintroduction source population in Washington, USA, to investigate genetic changes since the founder event.

All collected samples were processed at the Laboratory for Ecological, Evolutionary and Conservation Genetics at University of Idaho (Moscow, USA) for DNA extraction, PCR amplification, and microsatellite genotyping. DNA was extracted from tissue and fecal samples using DNeasy Blood and Tissue Kits and QIAmp Fast DNA Stool Mini Kits (Qiagen, Inc.), respectively. Fecal samples were extracted in a separate laboratory dedicated to low quality, low quantity DNA sources, and one negative control was included in each extraction to monitor reagent contamination. Twelve candidate otter-specific microsatellite loci were evaluated (Dallas & Piertney, 1998; Beheler et al., 2004, 2005; Mowry et al., 2011); however, locus RIO03 was monomorphic in the tissue samples and locus RIO11 failed to amplify. Therefore, we used a 10-locus multiplex to obtain genotypes: RIO01, RIO02, RIO04, RIO06, RIO07, RIO08, RIO12, RIO13, RIO16, and Lut453, as well as the SRY2 sex marker (Dallas et al., 2000). Four to six replicate PCRs were performed for the fecal samples that consistently amplified after an initial screening step of two amplifications; the tissue samples were amplified in duplicate. PCR products were visualized using a 3130xl DNA Sequencer and allele sizes were scored using Genemapper 5.0 (Applied Biosystems, Foster City, USA). Sample quality assessment and genotype screening methods followed those described by Stenglein et al. (2010). In short, we developed consensus genotypes for each sample by requiring an allele be detected in two independent PCRs to confirm a heterozygote and an allele be detected in three independent PCRs to confirm a homozygote. The SRY2 sex marker amplifies a fragment in males but not females (Dallas et al., 2000); we required 3–6 replicates for sex determination using this marker. If ≥2 replicates detected the Y chromosome, the sample was classified as a confirmed male, whereas if one replicate amplified the Y chromosome, we classified the sample as unconfirmed male because the Y chromosome amplification result was not confirmed. If no replicates amplified the Y chromosome, we classified the sample as female because 8–10 loci worked, thereby indicating sufficient DNA in the sample to avoid allelic dropout of the Y chromosome across 3–6 replicates.

Probability of identity for siblings (PI_sibs) was calculated separately for tissue and scat genotypes using GenAlEx v6.5 (Peakall & Smouse, 2012). PI_sibs ≤ 0.01 was used as a cutoff for the number of loci required to distinguish among unique genotypes; to determine the number of unique genotypes, matching analysis was conducted using GenAlEx (Stenglein et al., 2010). Two fecal samples were conservatively considered as originating from the same individual if their locus-specific alleles matched across ≥8 loci, and also if two fecal samples matched at only eight loci but the mismatches at locus nine or 10 were likely due to allelic dropout. If the 8–10 locus consensus genotype matched another sample, but the sex results differed, we conservatively retained the samples as a match and the sex of the individual as male. We calculated genotyping error rates from the first two PCR replicates of fecal samples that had consensus genotypes at 8–10 loci, following the methods of Broquet & Petit (2004).

This dataset contains the genotypes, genotyping error rates, spatial capture-recapture detection histories, and latrine site locations (geographical coordinates) for the study by Murphy et al., "Evaluating otter reintroduction outcomes using genetic spatial capture-recapture modified for dendritic networks."