Management implications of habitat selection by whooping cranes (Grus americana) on the Texas coast

By: , and 



Effective habitat management for rare and endangered species requires a thorough understanding of their specific habitat requirements. Although machine learning models have been increasingly used in the analyses of habitat use by wildlife, the primary focus of these models has been on generating spatial predictions. In this study, we used machine learning models in combination with simulated management actions to guide planning and inform managers. We used data from 61 whooping cranes (Grus americana) tagged with GPS telemetry collars between 2009 and 2018 near Aransas National Wildlife Refuge in coastal Texas. We included variables based on topography, land use classification, vegetation height, plant phenology, drought, storm surge events, and both wild and prescribed fires. We then built models at multiple scales: population level, home range level, and roosting and daytime within home range level. We simulated responses to the two primary management actions used to enhance whooping crane habitat on Aransas National Wildlife Refuge: prescribed fire and removal of woody vegetation. At the population and home range scales, land use classification variables had the highest importance values, whereas the combined elevation and bathymetry layer was the most important predictor at both roosting and daytime within home range scales. Our findings revealed that the effects of fire, although generally modest, varied spatially. Areas dominated by estuarine wetlands exhibited higher predicted use within the first months after a fire, whereas those dominated by palustrine wetlands were more likely to be avoided in the immediate postfire years. Our simulation of vegetation removal identified the areas on Aransas National Wildlife Refuge where whooping cranes were predicted to benefit the most if vegetation were removed. These techniques can be used by other researchers wanting to examine and predict the effects of potential management actions on target species habitat.

Study Area

Publication type Article
Publication Subtype Journal Article
Title Management implications of habitat selection by whooping cranes (Grus americana) on the Texas coast
Series title Ecosphere
DOI 10.1002/ecs2.4820
Volume 15
Issue 4
Year Published 2024
Language English
Publisher Wiley
Contributing office(s) Northern Prairie Wildlife Research Center
Description e4820, 19 p.
Country United States
State Texas
Google Analytic Metrics Metrics page
Additional publication details