Bat Community Structure and Habitat Selection Across an Urban-Agricultural Landscape

Galen Burrell, Purdue University

Abstract

Bats serve important ecological and economic roles in their communities. However, due to anthropogenic land use and human-introduced disease, bat populations in North America are facing unprecedented declines. To better inform conservation efforts for bat species in northeastern Indiana, I studied two aspects of bat ecology: (1) the effects of urbanization and agriculture on bat community composition in northeastern Indiana and (2) the roosting behavior of a population of state endangered/federally threatened northern long-eared bats (Myotis septentrionalis; hereafter northern myotis) in a restored mesic forest adjacent to a moderately sized city (Fort Wayne, IN). To study the first aspect, I deployed acoustic detectors in green spaces throughout Fort Wayne and the surrounding rural areas of Allen County. For each detector site, I compared species occupancy rates with site-specific characteristics at the plot scale (e.g., % canopy cover, midstory density) and at multiple landscape scales (e.g., % impervious cover within 1 km). Across 429 survey nights, acoustic detectors recorded calls from eight unique bat species, of which six species were abundant enough to conduct occupancy modeling. In four of the six species, measures of the amount of forest and forest edge in the landscape were included in one or more of the top models. The top models for the two other bat species, tricolored bats (Perimyotis subflavus) and silver-haired bats (Lasionycteris noctivagans), included measures of urban land cover and revealed a negative relationship between probability of occupancy and the proportion of high-density urban land in the area. The effect of habitat scale also differed between species. For example, models in the confidence set for eastern red bats (Lasiurus borealis) included variables associated with the plots surrounding detectors and with landscape features within 100 m of detectors. In contrast, the top models for hoary bats (Lasiurus cinereus) included landscape features at larger scales, within 500m and 1 km of detector sites. These findings suggest that both generalist and forest-obligate bat species in this study area selected spaces with greater levels of forested habitat. Furthermore, heavily urbanized areas were less likely to support the same levels of bat diversity as areas with forests and other green space. To accomplish the second aspect of my project, I captured bats in Fox Island County Park (Fort Wayne, IN) using mist-nets and affixed temperature-sensitive radio transmitters to four northern myotis individuals. I tracked these individuals back to day roosts, where I recorded roost characteristics (e.g., tree height, # of roost trees within 0.1 ha) and monitored bat skin temperatures (Tsk). I compared the characteristics of selected roosts to those of randomly assigned available trees in the same landscape to determine trends in roost selection preferences. Northern myotis in this study strongly preferred standing dead trees within a 31-ha patch of flooded forest on the northern border of Fox Island. These trees were highly exposed to solar radiation and were consistently warmer than ambient weather conditions, which suggests they may provide important thermoregulatory benefits to reproductive females and other members of the population. My research offers valuable information regarding resource use by bat communities in a landscape dominated by anthropogenic development. Urban areas containing large stretches of forests with trees in various stages of decay will be more likely to meet the needs of bats that would otherwise struggle in developed landscapes. The results of this study can be used to inform conservation efforts aimed at protecting populations of bats throughout Indiana and the Midwestern United States.

Degree

M.Sc.

Advisors

Jordan, Purdue University.

Subject Area

Agriculture|Ecology|Forestry|Recreation|Zoology

COinS