Modeling Raccoon Occupancy And Mesocarnivore Scat Misidentification In Western Maryland
Date
2022-05
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
Non-invasive survey methods are a valuable tool for understanding wildlife populations.
From camera trap data, wildlife researchers can create occupancy models that evaluate or predict
patterns of species distribution based on covariates. In this study, I investigated raccoon
(Procyon lotor) habitat use in western Maryland using occupancy models accounting for average
elevation, distance to water, edge density, distance to buildings, and road density. These models
revealed that within the study area, road density had the strongest overall effect on raccoon
occupancy. Where road density was greater, so was raccoon occupancy; this finding may be the
result of the correlation between roads, human presence and food sources, and the creation of
edge habitat favorable to raccoons. Collecting species data from scat sampling is difficult
because observers often misidentify scat in the field based on its morphological characteristics.
Using independent observers, I compared rates of correct identification for images of
mesocarnivore scat samples collected in western Maryland. I replicated the results of previous
studies which suggested that identifying features assigned to the scat of each species are not
always reliable. Even between mesocarnivores in separate families Canidae and Felidae, I
documented significant overlap in scat appearance. Therefore, genetic analysis is encouraged for
confirmation of samples to species level.
Description
Keywords
Raccoon, Maryland, Scat