Data and R code supporting projected climate-warming effects on Athabasca Rainbow Trout occurrence probability

This repository contains the data, R scripts, configuration templates, and selected derived outputs used to reproduce the analyses reported in the manuscript “Projected climate warming reduces occurrence probability of the imperiled Athabasca Rainbow Trout.” The workflow fits and evaluates a gradient boosting machine model of Athabasca Rainbow Trout (Oncorhynchus mykiss) occurrence using HUC10-blocked spatial validation. It includes model tuning and evaluation, relative-influence and interaction analyses, stream-network projections, assessment of projection support and extrapolation, and sensitivity analyses addressing stream-assignment distance, the spatial scale of road density, sampling gear, repeated observations, imperfect detection, and predictor collinearity. Occurrence probabilities were projected under current conditions and future mean August air-temperature surfaces for SSP2-4.5 and SSP5-8.5 in 2055 and 2085. Road density, land cover, crossings of concern, stream order, and brook trout occurrence were held constant across future scenarios. The approximately 50-m stream-network sampling points were used for numerical integration along stream segments and do not represent independently resolved environmental observations at that spatial scale. The package README describes the directory structure, software requirements, input data, execution order, expected outputs, and provenance of the environmental and climate data. The archived scripts reproduce the accepted analyses and manuscript outputs without retuning or replacing the final model.