Date of Award

Spring 2014

Degree Name

Master of Science (MS)

Department

Biology

Advisor

Dr. Rob Channell

Abstract

Global climate change is a serious threat to global biodiversity (IPCC, 2001). Due to their limited dispersal ability, reptiles and amphibians might be more vulnerable to rapid climate change than are other taxonomic groups (Gibbons et al., 2000). Herpetofauna in south-central Kansas was sampled from May through August in 2012 and 2013. Seven study sites spanning Meade, Clark, Comanche, and Barber counties were sampled. Drift fence and cover-board traps were arranged in transects at each site to capture reptiles and amphibians. Species were also sampled through surveys on all-terrain vehicles and on foot. Two thousand nine hundred and forty five individuals, belonging to 44 species were captured. The effect of climate change on the distribution of Blanchard’s Cricket Frog (Acris blanchardi), the Six-lined Racerunner (Aspidoscelis sexlineata), the Eastern Collared Lizard (Crotaphytus collaris), the Texas Horned Lizard (Phrynosoma cornutum), the Prairie Lizard (Sceloporus consobrinus), the Coachwhip (Coluber flagellum), the Ground Snake (Sonora semiannulata), and the Ornate Box Turtle (Terrapene ornata) were predicted using maximum entropy (MaxEnt) species distribution modeling. These species were selected because of their relatively large number of captures. The distribution of each of these species was modeled for historical climate conditions and then that distribution was projected for 2050 and 2080 using high and low emission scenarios. MaxEnt also was used to model the historical Kansas distribution for each of these species by using remote sensing data from Landsat 8. Occupancy models were constructed for each of the focal species to determine if habitat variables affect each species’ detectability and occupancy along the transects. In general, the distributions of the eight focal species responded to climate change by shifting north and fragmenting, although the magnitude of these distributional changes varied by species. The MaxEnt models generated with the Landsat 8 images did poorly at distinguishing species habitat from non-habitat. Most of the best supported occupancy models suggested that occupancy did not change across the landscape, meaning that either the focal species were ubiquitous or none of the covariates in this study distinguished habitat from non-habitat. The best supported occupancy models that contained one or more covariates that affected occupancy could be considered for conservation planning. Climate change threatens biodiversity worldwide. The MaxEnt climate models generated in this study differentiate between areas where species have high probabilities of occurrence or low probabilities of occurrence based on their climatic limits. Areas that were predicted to contain many species, contain species of high conservation priority, or become habitat corridors that species could use for northern dispersal prompted by future climate change could then be conserved. The occupancy models could be applied within each species’ historical distribution for conservation planning on a spatial scale appropriate for management. If the models generated in this study are used appropriately, this study could guide conservation planning to minimize the negative effects of climate change on the herpetofauna of Kansas.

Rights

Copyright 2014 Lisa M. Prowant

Library Call Number

LD2652 .T5 B5 P769 2014

Comments

Notice: This material may be protected by copyright law (Title 17 U.S. Code).

Included in

Biology Commons

Share

COinS