Contemporary land-use change structures carnivore communities in remaining tallgrass prairie

Abstract

The Flint Hills ecoregion in Kansas, USA, represents the largest remaining tract of native tallgrass prairie in North America. Anthropogenic landscape change (e.g., urbanization, agricultural production) is affecting native biodiversity in this threatened ecosystem. Our understanding of how landscape change affects spatial distributions of carnivores (i.e., species included in the Order ‘Carnivora’) in this ecosystem is limited. I investigated the influence of landscape structure and composition on site occupancy dynamics of 3 native carnivores (coyote [Canis latrans]; bobcat [Lynx rufus]; and striped skunk [Mephitis mephitis]) and 1 nonnative carnivore (domestic cat, [Felis catus]) across an urbanization gradient in the Flint Hills during 2016-2017. I also examined how the relative influence of various landscape factors affected native carnivore species richness and diversity. I positioned 74 camera traps across 8 urban-rural transects in the 2 largest cities in the Flint Hills (Manhattan, pop. > 55,000; Junction City, pop. > 31,000) to assess presence/absence of carnivores. Cameras were activated for 28 days in each of 3 seasons (Summer 2016, Fall 2016, Winter 2017) and I used multisession occupancy models and an information-theoretic approach to assess the importance of various landscape factors on carnivore site occupancy dynamics. Based on previous research in other ecosystems, I expected a negative relationship between both coyote and bobcat occurrence with increasing urban development but a positive relationship for domestic cat and skunk occurrence with increasing urban landcover. I also predicted grassland landcover to positively influence site occupancy for all carnivores except domestic cats. I expected that coyotes, the apex predator in this ecoregion, may limit domestic cat distributions through intraguild competition. Thus, I predicted a negative relationship between site occupancy of domestic cats and coyote occupancy probabilities. Because urban development results in habitat loss and fragmentation, I expected native species richness and diversity to decline with increased urban development. Coyotes had lower occupancy and colonization rates in areas with increased urban landcover. Bobcat occupancy was insensitive to urban landcover and colonization rates were greater in grassland landcover and row-crop agriculture fields. Site occupancy of bobcats was highly influence by forested areas and greater edge densities. Contrary to my hypothesis, striped skunk occupancy and colonization rates were negatively related to urban landcover. As expected, domestic cats were more likely to occur in and colonize sites with increased urban development and less likely to occur at sites with high coyote occupancy probabilities. Native carnivore species diversity and richness were negatively related to urban landcover. Occupancy dynamics of carnivores were shown to be influenced by landscape structure and composition as well as intraguild interactions. My results show urban landcover has a strong influence on the spatial distributions of carnivores in the northern extent of the Flint Hills.