Discovery of 20,000 RAD-SNPs and development of a 52-SNP array for monitoring river otters TECHNICAL NOTE Discovery of 20,000 RAD–SNPs and development of a 52-SNP array for monitoring river otters Jeffrey B. Stetz1 • Seth smith2 • Michael A. Sawaya1 • Alan B. Ramsey3 • Stephen J. Amish2 • Michael K. Schwartz4 • Gordon Luikart2,5 Received: 28 July 2015 / Accepted: 16 June 2016 � Springer Science+Business Media Dordrecht 2016 Abstract Many North American river otter (Lontra canadensis) populations are threatened or recovering but are difficult to study because they occur at low densities, it is difficult to visually identify individuals, and they inhabit aquatic environments that accelerate degradation of bio- logical samples. Single nucleotide polymorphisms (SNPs) can improve our ability to monitor demographic and genetic parameters of difficult to study species. We used restriction site associated DNA (RAD) sequencing to dis- cover 20,772 SNPs present in Montana, USA, river otter populations, including 14,512 loci that were also variable in at least one other population range-wide. After applying careful filtering criteria meant to minimize ascertainment bias and identify high quality, highly heterozygous (Ho = 0.2–0.50) SNPs, we developed and tested 52 inde- pendent SNP qPCR genotyping assays, including 41 that performed well with diluted DNA. The 41 loci provided high power for population assignment tests with only 1 misassignment (1.6 %) between closely neighboring pop- ulations. Our SNPs showed high power to differentiate individuals and assign them to population of origin, as well as strong concordance of genotypes from high and diluted concentrations of DNA, and between original RAD and the SNP qPCR array. Keywords Conservation genomics � RAD � Next generation sequencing � River otter � SNP � Population monitoring � Noninvasive genetic tagging Like other wide-ranging, elusive species, monitoring river otter populations remains a challenge for management agencies (Melquist et al. 2003). Individual otters are dif- ficult to distinguish visually and are known to travel con- siderable distances (Melquist and Hornocker 1983; Newton 2012). Radio-tagging studies, which require surgically implanting transmitters, are expensive and logistically difficult, and typically produce small samples sizes. For these and other reasons, studies aimed at estimating abundance, population growth rates, and connectivity have been difficult. Interest in ensuring long-term population persistence in the face of harvest and habitat loss has led to the devel- opment of molecular tools, primarily microsatellite mark- ers, to monitor otter population dynamics (e.g., Mowry et al. 2011). Advancements in discovery and genotyping of single nucleotide polymorphisms (SNPs), with concurrent reduction in costs, present an opportunity to vastly improve our ability to monitor wildlife populations. Relative to Data available from the Dryad Digital Repository: 10.5061/dryad. 96m9r]. Electronic supplementary material The online version of this article (doi:10.1007/s12686-016-0558-3) contains supplementary material, which is available to authorized users. & Jeffrey B. Stetz jeff.stetz@gmail.com 1 Sinopah Wildlife Research Associates, 127 N. Higgins, Suite 310, Missoula, MT 59802, USA 2 Fish and Wildlife Genomics Group, Division of Biological Sciences, University of Montana, Missoula, MT 59812, USA 3 MPG Ranch, Missoula, MT 59803, USA 4 National Genomics Center for Wildlife and Fish Conservation, Rocky Mountain Research Station, United States Forest Service, Missoula, MT 59801, USA 5 Flathead Lake Biological Station, University of Montana, Polson, MT 59860, USA 123 Conservation Genet Resour DOI 10.1007/s12686-016-0558-3 http://orcid.org/0000-0003-2876-6895 http://dx.doi.org/10.5061/dryad.96m9r http://dx.doi.org/10.5061/dryad.96m9r http://dx.doi.org/10.1007/s12686-016-0558-3 http://crossmark.crossref.org/dialog/?doi=10.1007/s12686-016-0558-3&domain=pdf http://crossmark.crossref.org/dialog/?doi=10.1007/s12686-016-0558-3&domain=pdf microsatellites, SNPs are less prone to genotyping errors, easier to transfer and analyze consistently among labora- tories, genotyping samples is faster and cheaper, and SNPs can include neutral markers or those linked to regions under selection (Morin et al. 2004; Allendorf et al. 2010; Helyar et al. 2011; Fabbri et al. 2012). One of the strengths of using genetic methods for monitoring river otter populations is the ease of collecting fecal samples from shared latrine sites along water bodies. Latrines are easy to locate and provide an opportunity to collect fecal material from multiple individuals at one site. Likely because of degradation of DNA due to moisture and ultraviolet light (Vynne et al. 2011; Stetz et al. 2015), these efforts have been largely unsuccessful due to poor genotyping success rates. For example, a study in Mon- tana was able to obtain complete multilocus genotypes for just 6 % of otter scat samples using microsatellites (Newton 2012). As SNPs are considerably shorter than microsatellites, higher genotyping success rates (i.e., lower rates of allelic drop out and false amplifications) are likely, even with poor quality samples such as scat (Morin and McCarthy 2007; Fabbri et al. 2012; Fitak et al. 2015). We therefore set out to develop a SNP array for river otters that would improve our ability to assess and monitor otter populations in Montana and across the species’ range. Although our emphasis was on otter populations in Montana, we used muscle tissue samples from a large geographic area to minimize issues of ascertainment bias and to ensure usefulness range-wide (Fig. 1; Allendorf et al. 2010). Sampling two populations that are somewhat close geographically (i.e., NB and QE) strengthened our test of marker power to assign individuals to population of origin. DNA was extracted from tissue samples using the Qia- gen DNeasy protocol then quantified using the Quant-iT TM PicoGreen � dsDNA assay to ensure DNA concentrations [5 ng/ll, needed for producing restriction-site-associated (RAD) sequencing libraries (Etter et al. 2011). RAD libraries were sequenced on the Illumina HiSeq 2000 platform using 150 base pair paired-end reads. Following Amish et al. (2012), we selected for infor- mative loci while applying strict data quality and assay design filters. Samples were excluded from downstream analysis if [ 50 % of their genotyped loci had\5 reads or [fivefold read count difference between alleles. To facil- itate SNP PCR genotyping assay design SNP loci had to be located between 40–70 nucleotides from the end, and we allowed only 1 SNP per RAD locus to avoid RAD loci assembled from paralogs and to avoid physically linked SNPs. We then excluded RAD loci where C2 samples had \5 reads or [fivefold read count difference between 2 alleles. We also required that observed heterozygosity (both range-wide and within Montana) be 0.2–0.6, that C1 sample from outside Montana was heterozygous at each RAD locus, and that each locus was successfully geno- typed in [80 % of Montana samples. These criteria produced 100 candidate SNPs, from which we selected the 96 with the highest expected heterozygosity for KASP-by-Design Fluidigm Assays (LGC Genomics � ). We tested these 96 assays on 73 samples from across otter range on a Fluidigm microfluidic SNP-chip. After excluding 7 loci due to high linkage dis- equilibrium (p \ 0.01), we identified 52 loci in Hardy– Weinberg proportions, with expected and observed heterozygosity [0.2 for Montana samples, and with B1 instance where the initial RAD genotype did not match the SNP-chip genotype (Table S1). Within SNP-chip geno- types, each sample was run at least three times, and there could be B1 instance of replicate genotypes not matching. We next required that each of the three possible genotypes from each locus was observed at least once on each of 2 SNP-chips. We set a call rate threshold of 90 % for each SNP-chip (i.e., 90 % of individuals and replicates yielded quality genotypes), and mean genotype confidence had to be [90 % for both chips. We then identified a subset of 41 loci for use with low quantity DNA samples by testing loci on samples where we reduced the original DNA concentration (C50 ng/ll) by half (Table S1). We excluded loci where C1 instance of normal and low concentration genotypes did not match, and where B1 instance of low concentration duplicate genotypes did not match. We used probability of identity statistics and population assignment tests to examine the SNP panel’s power to answer questions of sample identity or population of origin. We used GenAlEx 6.5 (Peakall and Smouse 2006) to cal- culate probability of identity, and Geneclass 2.0 (Piry et al. 2004; Rannala and Mountain 1997) to test how well indi- vidual otters assigned to populations (Paetkau et al. 1995) using our two classes of SNPs. Both sets of loci showed acceptable power to differentiate even closely related individuals within and among populations (Fig. 1). Using 52 loci produced 3 misassignments (5.1 %) compared to just 1 observed in the reduced set of 41 loci (1.6 %). All 3 putatively misassigned samples, however, originated in the closest neighboring population. For example, the 2 indi- viduals from New Brunswick that assigned to Quebec had 46–49 % assignment scores to their native region, sug- gesting that these may have been related to recent migra- tion events. The SNPs we report here represent a new tool for monitoring demographic and genetic status and changes in river otter populations across North America. SNPs may be particularly well suited to studying otter populations given the observed increase in genotyping success of fecal Conservation Genet Resour 123 samples relative to microsatellite markers in other species (e.g., Campbell and Narum 2009; Fabbri et al. 2012; Fitak et al. 2015). Further, this SNP array may be a powerful tool to explore genetic structure and evolutionary potential of otter populations while taking advantage of noninvasive sampling techniques. Such information is particularly valuable for reintroduction efforts and general questions on river otter ecology. Necessary next steps include optimiz- ing sampling and preservation methods to maximize SNP performance in spraints, and to directly compare perfor- mance of SNPs to microsatellites. Acknowledgments We are grateful to MPG Ranch for generously funding this research and to the University of Montana for in-kind support, including laboratory facilities, equipment, and protocols. We thank the following for providing river otter tissue samples or source DNA: E. Bunting (Cornell University), C. Brown (Rhode Island Division of Fish and Wildlife), C. Bernier (Vermont Fish & Wildlife Department), J. Cormier and L. Elson (New Brunswick Department of Environment), P. Canac-Marquis (Quebec Ministry of Energy and Natural Resources), J. Gude and N. Anderson (Montana Department of Fish, Wildlife, and Park), C. Nelson (British Columbia Forests, Lands and Natural Resource Operations), J. Harms (Environment Yukon), S. Talbot (U.S. Geological Survey), J. Hayden and Z. Lockyer (Idaho Department of Fish and Game), and K. Pilgrim (U.S. Forest Service Rocky Mountain Research Station). Fig. 1 Locations of tissue samples collected from North American river otter for SNP development; population- specific SNP statistics are reported within the figure Conservation Genet Resour 123 References Allendorf FW, Hohenlohe P, Luikart G (2010) Genomics and the future of conservation, invited review. Nat Rev Genet 11:697–709 Amish SJ, Hohenlohe PA, Painter S, Leary RF, Muhlfeld C, Allendorf FW, Luikart G (2012) RAD sequencing yields a high success rate for westslope cutthroat and rainbow trout species-diagnostic SNP assays. Mol Ecol Resour 12:653–660 Campbell NR, Narum SR (2009) Quantitative PCR assessment of microsatellite and SNP genotyping with variable quality DNA extracts. Conserv Genet. doi:10.1007/s10592-008-9661-7 Etter PD, Bassham S, Hohenlohe PA, Johnson EA, Cresko WA (2011) SNP discovery and genotyping for evolutionary genetics using RAD sequencing. In: Orgogozo V, Rockman MV (eds) Molecular methods for evolutionary genetics. Humana Press, New York, pp 157–178 Fabbri E, Caniglia R, Mucci N, Thomsen HP, Krag K, Pertoldi C, Loeschcke V, Randi E (2012) Comparison of single nucleotide polymorphisms and microsatellites in non-invasive genetic monitoring of a wolf population. Arch Biol Sci Belgrade 64:321–335 Fitak RR, Naidu A, Thompson RW, Culver M (2015) A new panel of SNP markers for the individual identification of North American pumas. J Fish Wildl Manag 7:13–27. doi:10.3996/112014- JFWM-080 Helyar SJ, Hemmer-Hansen J, Bekkevold D, Taylor MI, Ogden R, Limborg MT, Cariani A, Maes GE, Diopere E, Carvalho GR, Nielsen EE (2011) Application of SNPs for population genetics of nonmodel organisms: new opportunities and challenges. Mol Ecol Resour 11:123–136 Melquist W, Hornocker M (1983) Ecology of river otters in west central Idaho. Wildl Monogr 83:3–60 Melquist WE, Polechla PJ, Toweill D (2003) River otter: lontra canadensis. In: Feldhamer GA, Thompson BC, Chapman JA (eds) Wild mammals of North America: biology, management, and conservation. Johns Hopkins University Press, Baltimore, pp 708–734 Morin PA, McCarthy M (2007) Highly accurate SNP genotyping from historical and low-quality samples. Mol Ecol Notes 7:937–946 Morin PA, Luikart G, Wayne RK (2004) SNPs in ecology, evolution and conservation. Trends Ecol Evol 19:208–216 Mowry RA, Gompper ME, Beringer J, Eggert LS (2011) River otter population size estimation using noninvasive latrine surveys. J Wildl Manag 75:1625–1636 Newton DE (2012) Northern river otter population assessment and connectivity in western Montana. Thesis, University of Montana, Missoula, MT Paetkau D, Calvert W, Stirling I, Strobeck C (1995) Microsatellite analysis of population structure in Canadian polar bears. Mol Ecol 4:347–354 Peakall R, Smouse PE (2006) GENALEX 6, genetic analysis in Excel. Population genetic software for teaching and research. Mol Ecol Notes 6:288–295 Piry S, Alapetite A, Cornuet J-M, Paetkau D, Baudouin L, Estoup A (2004) GENECLASS2: a software for genetic assignment and first-generation migrant detection. J Hered 95:536–539 Rannala B, Mountain JL (1997) Detecting immigration by using multilocus genotypes. Proc Natl Acad Sci 94:9197–9201 Stetz JB, Seitz T, Sawaya MA (2015) Effects of exposure on genotyping success rates of hair samples from grizzly and American black bears. J Fish Wildl Manag 6:191–198 Vynne C, Baker MR, Breuer ZL, Wasser SK (2011) Factors influencing degradation of DNA and hormones in maned wolf scat. Anim Conserv 15:194 Conservation Genet Resour 123 http://dx.doi.org/10.1007/s10592-008-9661-7 http://dx.doi.org/10.3996/112014-JFWM-080 http://dx.doi.org/10.3996/112014-JFWM-080 Discovery of 20,000 RAD--SNPs and development of a 52-SNP array for monitoring river otters Abstract Acknowledgments References