Carmela Musto, Jacopo Cerri , Marco Galaverni, Romolo Caniglia, Elena Fabbri, Nadia Mucci, Marco Apollonio, Paolo Bonilauri, Giulia Maioli, Maria Cristina Fontana

Over the last 40 years the gray wolf (Canis lupus) re-colonized its historical range in Italy increasing human-predator interactions. However, temporal and spatial trends in wolf mortality, including direct and indirect persecution, were never summarized. This study aims to fill this gap by focusing on the situation of Tuscany and Emilia-Romagna regions, believed to host a significant proportion of the Italian wolf population, by: (i) identifying the prevalent causes of wolf mortality, (ii) summarizing their temporal and spatial patterns and (iii) applying spatially-explicit Generalized Linear Models to predict wolf persecution. Between October 2005 and February 2021, 212 wolf carcasses were collected and subjected to necropsy, being involved in collisions with vehicles (n = 104), poisoned (n = 45), wounded with gunshot (n = 24) or blunt objects (n = 4) and being hanged (n = 2). The proportion of illegally killed wolves did not increase through time. Most persecution events occurred between October and February. None of our candidate models outperformed a null model and covariates such as the density of sheep farms, number of predations on livestock, or human density were never associated to the probability of having illegally killed wolves, at the municipal scale. Our findings show that conventional correlates of wolf persecution, combined with a supposedly high proportion of non-retrieved carcasses, fail to predict illegal wolf killings in areas where the species have become ubiquitous. The widespread spatial distribution of illegal killings indicates that persecution probably arises from multiple kinds of conflicts with humans, beyond those with husbandry. Wolf conservation in Italy should thus address cryptic wolf killings with multi-disciplinary approaches, such as shared national protocols, socio-ecological studies, the support of experts’ experience and effective sampling schemes for the detection of carcasses.