COMPARING THE EFFECTS OF LANDSCAPE HABITAT DATA COLLECTION METHODS IN A SPECIES OCCUPANCY MODEL
Date
2024-05
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
Understanding how species interact with their environment is an essential
component of wildlife research. One widely used source of habitat data is the National
Land Cover Database (NLCD), which provides 30m resolution raster files of land
cover types, created using satellite imagery. This resource is free and easy to use.
However, a 30m resolution may not always provide as much information about a
landscape as we desire, leading many researchers to collect their own data at finer
scales. I investigated how using 30m resolution NLCD data in place of 10m hand digitized data changes the results of species occupancy models for river otter (Lontra
canadensis) and mink (Neovison vison). These semi-aquatic mammals play key roles
in riparian ecosystems. Often consisting of very narrow patches of forest, riparian
areas can easily get overlooked in habitat data measured at a coarser resolution. For
this reason, comparing the results of occupancy models created using different habitat
data resolutions can provide valuable insight into best practices for habitat data
collection. Although there were differences in both land cover and other variables
included in my top models and the top models of Holland et al. (2019), my results did
not indicate that the coarser resolution data from the NLCD made a significant
difference in the results of the species occupancy models for river otter and mink, due
to the similar nature of the variables and overall management implications each may
pose.