Year of Award

2011

Document Type

Thesis

Degree Type

Master of Science (MS)

Degree Name

Environmental Studies

Department or School/College

Environmental Studies Program

Committee Chair

Vicki Watson

Commitee Members

Len Broberg, Linda Vance, Winsor Lowe

Keywords

Akaike's Information Criterion, breeding site selection, habitat models

Publisher

University of Montana

Abstract

Amphibians have a complex life history that requires a mosaic of habitats, including breeding, foraging, and over-wintering areas. Historically, regulators have focused on wetland breeding habitat quality to explain amphibian presence. Recently, other habitat requirements including landscape level factors have been examined. Data collected from amphibian surveys in Montana were used to determine if wetland quality factors or landscape level factors were better determinants of amphibian occurrence at breeding sites. Twenty-six habitat models were constructed a priori for eight species of amphibians in Montana. This included five models containing parameters associated with wetland quality, ten landscape level models, and ten models that combined both local and landscape covariates. Logistic regression analysis with an information theoretic approach was used to select the best approximating model. Results indicate that habitat models including only wetland variables were not good predictors of presence for most amphibians. The landscape scale at which habitat models were best supported varied among species and was consistent with differing life history traits. The presence of Ambystoma macrodactylum, the western population of Ambystoma tigrinum, Bufo boreas, and Rana luteiventris was best predicted by landscape covariates. Models with a combination of local and landscape covariates were best supported for Rana pipiens, Bufo woodhousii, Pseudacris maculata, and Pseudacris regilla. The probability of Ambystoma macrodactylum presence is highest at breeding sites that are surrounded by forested areas. The western population of Ambystoma tigrinum was positively associated with an increased distance to forest and a higher density of wetlands around a breeding site. The eastern population of Ambystoma tigrinum was negatively associated with higher elevations. Bufo boreas and Rana luteiventris were positively associated with increased forest within 1,000 m, and negatively associated with increased distance to forest and aquatic sites in an agricultural landscape. Bufo woodhousii and Pseudacris maculata were positively associated with open landscapes dominated by natural grasses. The presence of Rana pipiens was positively associated with open landscapes dominated by natural grasses. This project highlights the importance of maintaining intact landscapes around amphibian breeding ponds in order to meet the habitat requirements of amphibians during all stages of their life cycle.

Download

Share

COinS
 

© Copyright 2011 Catherine Anne McIntyre