Analysis of spatiotemporal patterns of home range use and habitat selection to inform management of feral pigs on Auckland Island

Abstract

: Feral pigs ( Sus scrofa ) were introduced to Auckland Island in subantarctic New Zealand in 1807. They established and became invasive, subsequently causing substantial unwanted impacts on native biodiversity. Understanding pig movement behaviour and habitat selection can lead to focused, efficient, and effective management efforts, especially during initial knockdown of the population. Here we used location data from ARGOS telemetry collars deployed on 15 Auckland Island pigs from 2007 to 2008 to estimate seasonal home-range sizes and habitat selection. The annual home-range sizes of pigs ranged from 1.26 to 36.4 km 2 , with a mean of 14.1 km 2 (SD = 11.3 km 2 ). The variation in home-range sizes between winter and summer was generally low. We found that the pig population selected areas of their home ranges that were on average closer to the coast (though this was not necessarily selection for beaches or strandlines per se) during the spring and summer months. Pigs also selected areas closer to the coast in winter, albeit with weaker strength of selection than in spring and summer. We had insufficient data to differentiate between different coastal habitats (i.e. cliffs versus beaches and strandlines). We also found that pigs showed annual selection for tussock, hills with north-facing aspects, and rivers and streams. They avoided areas of bog-swamp and forest-scrub. Importantly, we do not know what food resources or other factors were driving habitat selection. Our results highlight areas selected by pigs that could be targeted for initial knockdown; however, staff will have to search all habitats on the island to achieve rapid eradication or a protracted mop-up of survivors. Simply targeting key accessible areas like strandlines and coastal forests will not achieve eradication. Based on these home-range sizes, eradication would require a hunter path spacing of no more than 1.4 km to ensure every pig home range was searched at least once.