2018 ESA Annual Meeting (August 5 -- 10)

PS 13-12 - The dynamics of cohabitation: Using camera traps to estimate important wildlife occupancy covariates in an urban matrix

Tuesday, August 7, 2018
ESA Exhibit Hall, New Orleans Ernest N. Morial Convention Center
Stephen Kovari1, Colin Peros2, Julia Martinez Franks1, Ev Cheng1, Keri L. VanCamp1 and Lynn Christenson1,2, (1)Biology, Vassar College, Poughkeepsie, NY, (2)Environmental Research Institute, Vassar College
Background/Question/Methods

Vassar College is a one thousand-acre campus nestled within an urban matrix. It has a varied landscape; comprised of multiple ponds, streams, forest patches, and a 527.5-acre farm and ecological preserve. The campus therefore has the potential to provide ample habitat for many mammalian species within the heavily developed area. However, with 2,500 students, many buildings and roads, as well as a golf course, the campus has multiple features that could act as barrier to the occupancy of certain species. Using twenty-one motion-activated cameras distributed equally in 50 acre grid cells, we analyze the effect that anthropogenic land-use factors (main campus vs ecological preserve, distance to nearest structure, patch size, naturalized area) as well as natural factors (habitat type, distance to water) have on occupancy of four key mammal species: red fox (Vulpes vulpes), eastern coyote (Canis latrans), bobcat (Lynx rufus), and northern raccoon (Procyon lotor). We also analyze the effect that human impact has on daily activity patterns for each species. Occupancy models in the USGS software “Presence” will be run to estimate the most important covariates for all species throughout the winter and spring seasons of 2018.

Results/Conclusions

Of the four target species (red fox, coyote, northern raccoon, bobcat), we have detected all but bobcats across both landscape types (preserve and campus). Bobcats have only been detected at the preserve. All species except raccoons were active throughout the day, regardless of landscape type, suggesting that human activity on the Vassar campus does not alter animal activity. Using percent naturalized area as a metric for disturbance (higher naturalized = lower disturbance), we found that red fox and northern raccoon showed little aversion to the most disturbed areas. Coyotes also utilized the most disturbed locations, but avoided areas that were particularly fragmented (camera sites in small patch sizes), suggesting habitat connectivity is more important than amount of disturbance. These preliminary results highlight the importance of larger tracts of ‘naturalized’ areas even when bounded by development, because they could provide bobcats habitat that more disturbed areas do not. Occupancy modeling in “Presence” in the spring will bring new insights into what drives occupancy probability for these species across the landscape and allow us to map occupancy of all four species on Vassar’s campus.