Spatial explicity capture recapture model apply to camera trapping and genetic scat survey on the wildcat from Sicily

Recent development of spatial explicit capture recapture (SECR) models have override the problem of how to calculate the study area of which the expected density refers. The future use of such models allows for a great flexibility of the study design (e.g. allowing individual recaptures into the same occasion or even to limit monitoring to just one sample occasion): given that a great revolution in census methods is ongoing. We used this new generation of model, specifically the software Density 4.4 which, trough a Maximum Likelihood based model selection, generated density estimates in a spatial-explicit capture-recapture (SECR) framework. We assessed density of wildcat in Sicily using data from a study where camera trapping and genetic scat survey where conducted simultaneously on the Etna Volcano. Genetic scat survey was conducted in order to provide both molecular data on the taxonomic status of this population and an independent estimation of the density. In 2010 we monitored two consecutive and adjacent trapping lines from 14 May to 11 September for a total of 18 camera stations (camera traps were placed in pair at each stations). 1080 trap-days gained 67 events from 16 stations. The resulting pictures were compared between them and with those collected during previous camera trapping monitoring undertaken in the same study area. We identified at least 14 wildcats (excluding two kittens) and the rate of capture success was 1/16 trap-nights. Scat collection was done during the camera trapping monitor and the transects were designed in order to search near the camera stations, therefore genetic captures from scats were then assigned to the nearest camera trapping station. 4 transects (23 Km) were repeated weekly for 17 times (391 km) and 11 out of 39 fresh wildcat scats provided, using genetic molecular tool analyses, the identification of at least 7 individuals (2 individuals were located three times whilst the remains once). Genotype were identified using a panel of 12 autosomal microsatellites, and assignment procedures to wildcat, domestic cat or hybrid reference sample groups. The software analyses performed on the two independent data set of spatial captures produced the follow results (wildcat/100 ha): Dct= 0.31 + 0.1 (s.e), g0= 0.17410, σ=978, with 95% confidence interval of 0.17-0.54 and Dg= 0.76 + 0.4 (s.e), g0= 0.09, σ=280 with 95% confidence interval of 0.28-2. For comparison, traditional capture-recapture model associated with FMMDM buffer results are: Dct= 0.41 + 0.35 (s.e); Dg= 0.29 + 1.15 (s.e). Genetic scat density is greater than that of camera trapping, probably as consequence of smallest sample size, however the confidence intervals of the two monitoring slightly overlapped. The goodness of camera trapping to assess the wildcat’s density is reasserted, and to our knowledge, our study is the first that successfully performed genetic molecular analyses on the wildcat scats. Future developments will be focus on the maximization of genetic analyses success as this non-invasive method provides crucial data on the taxonomic status of the wildcat populations. Camera trapping became a common tools for wildlife management and the scientific progress into the analyses and implementation of spatial capture recapture models promise a wider range of field application. Finally we strong suggest, when camera trapping monitors detect the presence of wildcats, to engage in the collection of the scats and successive molecular analyses: the combination and integration of this two non invasive methods surveys will produce a wider and more complete view on the wildcat biology, which is an essential condition to formulate concrete and successfully conservation measures.