Species-specific responses of mammal activity to exurbanization in New Hampshire, USA

Abstract

Urbanization and habitat fragmentation can disrupt wildlife behavior and cause declines in biodiversity and ecosystem function. Most urban wildlife research has compared highly urbanized regions with rural areas. However, human development is also rapidly occurring in exurban areas, which consist of a matrix of lower-density housing and natural patches. Thus, although such “exurbanization” is intensifying, little research has examined how mammals respond to exurban development. To address this knowledge gap, we evaluated the activity of 12 species using 104 camera traps in exurban and rural areas across southeastern New Hampshire, USA, during summer 2021 and winter 2021–2. We quantified species’ activity levels (overall portion of daily activity) and patterns (variation of diel activity period) to test hypotheses regarding how species’ space requirements and nocturnality modulated their responses to exurban development. We found mixed support for our hypotheses. Two species with large space requirements (bobcats Lynx rufus and white-tailed deer Odocoileus virginianus) reduced activity levels in exurban areas, following hypothesized predictions, while other species (e.g., coyote Canis latrans) did not. As predicted, nocturnal species were less likely to shift activity patterns, but this varied across species and seasons. We also found evidence for a coupled predator–prey response among bobcats and lagomorphs in summer, with similarly altered activity in exurban areas. These results suggest that wildlife modify activity in response to exurban development with substantial species and season-specific variation within the mammal community, highlighting the complex ways wildlife adapt to urbanization and the potential consequences thereof for mammal communities.

Generalized nonlinearity in animal ecology: Research, review, and recommendations

“…The remaining 8% is comprised of a combination of agriculture, grasslands, shrublands, and sand beaches (Dewitz & U.S. Geological Survey, 2021 ). We selected 75 of the camera sites using a generalized random tessellation design with sites separated by at least 1 km (Stevens & Olsen, 2003 ), and we selected the remaining 33 sites using a fully random design within properties managed by the University of New Hampshire with sites separated by at least 250 m (Moll et al., 2023 ; Poisson et al., 2023 ). We selected sites using QGIS (QGIS Association, 2021 ), then made minor adjustments in the field to locate a suitable location to affix the camera on a tree at approximately 50 cm above ground, facing north, and not conspicuously placed to avoid tampering.…”