Regression-Tree Modeling of Desert Tortoise Habitat in the Central Mojave Desert

Abstract

This paper describes an interdisciplinary study of the habitat requirements of threatened desert tortoises (Gopherus agassizii) on eight 225‐ha study plots in a 14 000 ha study area near the southern boundary of the U.S. Army's National Training Center at Fort Irwin in the central Mojave Desert of southern California. The objective of the study was to produce an empirical, statistical, GIS‐based model of desert tortoise habitat use based on a combination of field data and data derived from various spatial databases, including satellite imagery. A total of 11 primary and secondary data layers constitute the spatial database used for this project. Vegetation and tortoise relative density data were obtained from field surveys. Regression‐tree methods were used to develop the statistical model. The tree has 11 terminal nodes and a residual mean deviance of 1.985. Out of 73 potential predictors in the model specification, only eight were selected by the algorithm to be used in construction of the tree. The model suggests that tortoises tend to occur on southwest exposures and loamy soils, and that they avoid stony soils, north exposure, and areas of very low plant cover. Our results imply that soil composition and parent materials can be important determinants of habitat suitability for desert tortoises, and for burrowing animals in general. Our study also provides an example of how the interdisciplinary integration of biology, earth sciences, GIS, and statistical modeling can reveal important aspects of the habitat requirements of endangered species.

Mitigation-driven translocation effects on temperature, condition, growth, and mortality of Mojave desert tortoise ( Gopherus agassizii ) in the face of solar energy development

“…Our study also suggests the importance of shade structures and monitoring along project fencing in the first month post-release to reduce thermal exposure for individual tortoises attempting to home, particularly for short-distance translocations (Hinderle et al, 2015). To further reduce thermal stress, mitigations should ensure suitable habitat for translocated animals, such as adequate shrub cover and loamy soils, that allow tortoises to most easily find shelter and dig burrows (Andersen et al, 2000). Habitat might be enhanced by providing artificial burrows in a 'soft' rather than 'hard' release.…”

Predicting the post-fire establishment and persistence of an invasive tree species across a complex landscape

“…Although other studies have utilized classification and regression tree (CART) models to successfully predict habitat requirements for rare plants (Engler et al 2004) and animals (Anderson et al 2000;Dettmers and Bart 1999), species response to climate change (Iversen andPrasad 1998), coral distributions (De'ath andFabricius 2000), and the distribution of California oak species (Vayssieres et al 2000), our study is the first to utilize these models to predict invasive species habitat. While other studies have shown the utility of CART models for native species management and research, our work demonstrates the same potential for invasive species.…”

Factors influencing the formation of unburned forest islands within the perimeter of a large forest fire

“…The decision tree algorithm is based on a process called recursive partitioning, which consists of dividing the data into two sets that produce the largest decrease in deviance (i.e. measure of heterogeneity) (Andersen et al, 2000). To create these trees, a randomly elected sample test was selected (n = 1249) (77% of the total cases), while the remaining n = 359 cases were used to test the model.…”