Use of camera traps to examine the mesopredator release hypothesis in a fragmented midwestern landscape
Abstract
The mesopredator release hypothesis (MRH) has been suggested as a reason why many mammalian generalist mesopredators flourish and become abundant. However, the MRH has only been examined in a limited number of field studies. Some studies have argued that coyotes (Canis latrans) act as top predators in fragmented forest systems and coyote presence has a positive effect on song bird diversity and abundance by controlling mesopredator abundance. We integrated camera trap data and occupancy modeling to determine the factors that affect coyote detection probability, and habitat use in a fragmented suburban landscape in central Missouri. We then examined the influence of coyote presence and other habitat variables on mesopredator detection probability and habitat use in the same system. Coyote detection was negatively related to increasing forest cover, whereas red fox (Vulpes vulpes) detection was positively related to increasing urbanization. Coyote occurrence models suggested little habitat selection, while the mesopredator occurrence models suggested an affinity for urbanization. Although there was a slight negative effect of coyote presence on site use by other mesopredators, we suggest that the smaller species are better adapted to coexisting with humans and thus have increased in abundance.