Evaluating the predictive power of field variables for species and individual molecular identification on wolf noninvasive samples

Live-trapping elusive animals is often challenging, hampering the achievement of reasonable sample sizes for molecular studies. In such cases, the use of noninvasive samples (NIS) is critical in many research fields, mostly related to ecology, management and conservation of wild species. We analysed the influence of several variables potentially associated with the quality of wolf NIS—season, weather conditions, and in situ collected site and sample characteristics—on the success rates of species and individual identification performed using mtDNA and 13 microsatellites, respectively. NIS included scats, urine and saliva collected from two areas in Portugal. Scat samples exhibited the highest success rate for both species (81%) and individual identification (59%), compared with urine (63 and 30%, respectively) or saliva samples (48 and 36%, respectively). The success rate of species identification of scats was better explained by season of collection, the presence of mucous, moisture and odour. For samples with successful species identification analysis, individual identification success was best predicted by the presence of odour. Performing a preliminary selection of scat samples with the best characteristics can increase up to 13% the success rates of molecular analysis. Urine collected on snow had a higher success rate of species identification than that collected on vegetation. To our knowledge, this was the first time that wolf urine on vegetation near ground-scratching marks is used as DNA source. Saliva samples collected with different substrate types can also be used for species identification. These results contribute to optimising noninvasive sampling procedures, maximising the success of molecular ecology studies, and ultimately minimising sampling efforts and costs.