Date of Award

January 2016

Degree Type

Open Access Thesis

Degree Name

Master of Science (MS)

Department

Biological Sciences

First Advisor

Luke E. Dodd

Department Affiliation

Biological Sciences

Second Advisor

Charles L. Elliott

Department Affiliation

Biological Sciences

Third Advisor

Kelly Watson

Department Affiliation

Geosciences

Abstract

Habitat use of bats may shift following population level impacts of White-nose Syndrome (WNS). Multiple bat species have experienced unprecedented population declines due to WNS, including federally listed Myotis sodalis (Indiana bat) and Myotis septentrionalis (northern long-eared bat). Specifically, the effect of WNS across forest landscapes is unclear in relation to prescribed fire. Mammoth Cave National Park (MACA) has employed a prescribed fire regime since 2002 and WNS was detected on MACA in 2013. Bat activity was monitored across burned and unburned sites at MACA before (2010-2012) and after the detection of WNS (2013-2016) using transects of acoustic detectors (Anabat II and Songmeter SM3). Recordings were classified to phonic groups (low, mid, Myotis) and species using automated classifiers (Bat Call Id v.2.7c and Kaleidoscope Pro v.3.1.4B). Subsequent analyses were conducted using bat passes with 5 ≤ pulses, with a 95% or 70% confidence interval for species and phonic group classification, respectively. Insect traps (blacklight and malaise) were deployed concurrent with acoustic transects and insects were identified to order. There was a significant interaction between WNS and prescribed fire for the Myotis phonic group (P < 0.01), with the greatest activity shifting from unburned areas before WNS to burned areas after WNS. Total insect abundance was greater after WNS (P < 0.01). Abundance was greater after WNS and in unburned areas for Lepidoptera (P < 0.01) and burned areas Coleoptera (P < 0.05). Diptera abundance did not change over the course of the study (P ≥ 0.05). These data indicate substantial changes in both predator and prey community composition at MACA. I used multiple linear regression in conjunction with Akaike’s Information Criterion to determine the most the most parsimonious model for predicting M. sodalis and M. septentrionalis activity in a post-WNS landscape. The only significant model developed considered landscape attributes (P < 0.05); this model provided the best fit for both M. sodalis and M. septentrionalis response variables. Parameter estimates were significant for aspect in the landscape model for M. sodalis (P < 0.05), suggesting a negative relationship with this physiographic variable. However, these data were exploratory and demonstrate a need to further investigate the habitat use of these two imperiled species. In summary, bat activity across MACA has been altered as a result of WNS and prescribed fire. Species impacted by WNS have declined across the landscape, with activity increasing for species not susceptible to WNS. Burned areas across MACA experienced greater levels of bat activity after WNS. Thus, forest managers should take prescribed fire and landscape features into consideration when managing for bat population impacted by WNS.

Share

COinS