Publication Date

Summer 2026

Degree Type

Thesis

Degree Name

Master of Science (MS)

Department

Environmental Studies

Advisor

Rachel O’Malley; Karine Tokatlian; Rick Kos

Abstract

The California newt (Taricha torosa) is listed as a State Species of Special Concern across its southern range due to habitat degradation and loss, and limited studies indicate it may be experiencing these threats in its northern range as well. This project assesses where the California newt may be encountering roads as migration barriers and experiencing high mortality in western Santa Clara County, California, in an effort to inform future conservation efforts. I conducted surveys of California newt roadkill on roads bordering two reservoirs used as breeding habitat and accessed roadkill data collected by the Lexington Reservoir Newt Patrol. Roadkill data were used to train presence-only prediction models that tested how three different resolutions of explanatory raster layers, two spatial scales, and the effect of spatially thinning systematically-collected data impact explanatory variables. I also assessed how raster resolution impacted the prediction of opportunistically collected data from iNaturalist. Regional low-resolution rasters incorporated explanatory variables from both local survey extents and identified several new locations with high probabilities of roadkill occurrences, but failed to predict the majority of opportunistic observations. While these regional models trained with systematically collected data provide guidance for future surveys, the inclusion of opportunistic observations in training data would likely help to train models to recognize more locations as having a high probability of roadkill presence.

Download

Share

COinS