RESEARCH PAPER
Improving predation risk modelling: prey-specific models matter
 
More details
Hide details
1
Swiss Ornithological Institute - Vogelwarte, Seerose 1, 6204 Sempach, Switzerland.
2
Department of Biology, University of Naples, Via Cintia 26, 80126 Naples, Italy
3
Unit for Conservation Genetics (BIO-CGE), Department for the Monitoring and Protection of the Environment and for Biodiversity Conservation, Italian Institute for Environmental Protection and Research (ISPRA), Via Ca` Fornacetta 9, 40064 Ozzano dell’Emilia, Bologna, Italy.
4
Department of Architecture and Design, “Sapienza” University of Rome, Via Flaminia 72, 00196 Rome, Italy
5
Regione Emilia-Romagna, Direzione Agricoltura, Pianificazione e Osservatorio Faunistico, Viale della Fiera 8 40127 Bologna, Italy.
6
Regione Umbria, Servizio Programmazione Faunistica Venatoria, Osservatorio Faunistico Regionale, Corso Vannucci 96, 06121 Perugia, Italy.
CORRESPONDING AUTHOR
Pietro Milanesi   

Swiss Ornithological Institute - Vogelwarte, Seerose 1, 6204 Sempach, Switzerland.
Online publish date: 2019-11-26
Publish date: 2019-11-26
 
KEYWORDS
TOPICS
ABSTRACT
Globally, large carnivore livestock predations are major causes of conflicts with humans, thus identifying hotspots of carnivore attacks is fundamental to reduce the impact of these, and hence promote coexistence with humans. Species distribution models combining predictor variables with locations of predation events instead of species occurrences (also known as predation risk models) are increasingly used to predict livestock depredation by carnivores, but they are often developed pooling attacks on different livestock species. We identified the main factors related to predation risk on livestock using an extensive dataset of 4,604 locations of verified wolf predation events on livestock collected in northern and central Italy during 2008–2015 and assessed the importance of pooling versus splitting predation events by prey species. We found the best predictors of predation events varied by prey species. Specifically, predation risk increased with altitude especially for cattle, with grasslands especially for cattle and sheep and with distance to human settlements, especially for goats and livestock but only slightly for cattle and sheep. However, predation risk decreased as human population density, human settlements and artificial night-time light brightness increased, especially for cattle. Finally, livestock density was positively related to predation risk when herd exceeds 500 heads for km2. Moreover, prey-specific risk models are better tools to predict wolf predation risk on domestic ungulates. We believe that our approach can be applied worldwide on different predator-prey systems and landscapes to promote human-carnivore coexistence. Actually, while pooling predation events could be primarily used by managers and personnel of wildlife agencies/offices in developing general policies, splitting predation events by prey species could be used at farm-level to better identify livestock owners at risk in high-priority areas and which prevention tools and deterrents (e.g. electric fences, guarding dogs, predator-proof enclosures) should be applied, as the most effective measures differ by species.
eISSN:1825-5272
ISSN:0394-1914