Genetic non-invasive sampling (gNIS) as a cost-effective tool for monitoring elusive small mammals

Genetic non-invasive sampling (gNIS) may provide valuable information for population monitoring, as it allows inferences of population density and key behavioural traits such as dispersal, kinship and reproduction. Despite its enormous potential, gNIS has rarely been applied to small mammals, for which live-trapping is still the most commonly used sampling method. Here we evaluated the applicability and cost-effectiveness of gNIS comparedwith live-trapping, to monitor ametapopulation of an Iberian endemic and elusive rodent: the Cabrera vole (Microtus cabrerae). We compared the genetic diversity, kinship and dispersal movements inferred using both methods. For that, we optimised microsatellite markers for individual identification of M. cabrerae, using both tissue (n = 31) and faecal samples (n = 323) collected from a metapopulation in south-western Iberia. An initial set of 20 loci was optimised for tissue samples, from which 11 were selected to amplify in faecal samples. Overall, gNIS revealed a higher number of identified individuals (65) than live-trapping (31), and the estimated genetic diversity was similar using data from tissues and gNIS. Kinship analysis showed a higher number of inferred relationships and dispersal events when including gNIS, and indicated absence of sex-biased dispersal. The total cost (fieldwork and genetic analysis) of each genotype obtained through live-trapping was three times greater than for gNIS. Our data strongly supports the high potential and costeffectiveness of gNIS for monitoring populations of elusive and/or threatened small mammals.We also illustrate how this genetic tool can be logistically feasible in conservation.