Predicting timber theft based on environmental features – Insights from Humboldt Redwoods State Park, US

Abstract

ObjectiveThere's a small but growing body of evidence that wildlife crime mirrors the spatial concentrations we see with traditional, high-volume urban crimes. Yet, the majority of this literature has focused on the poaching of fauna, not flora. Therefore, it is unclear whether poaching of flora—both plants and trees—is spatially concentrated and can be explained by both the natural and built landscape. This study set out to empirically examine this question as it relates to the illicit removal of downed redwood trees in Humboldt Redwoods State Park (HRSP)—home to the largest contiguous area of old-growth redwood forest on Earth. MethodsGeo-coordinate data of redwood theft was recorded and mapped along with environmental features such as roads, trails, parking points, campgrounds, and redwood old-growth distribution. A spatial econometric model was used to measure the presence or absence of each variable in 499 grid cells (1 sq. km each) that make up HRSP on the outcome of redwood theft. ResultsA Firth's penalized likelihood logistic regression model found that poaching was more likely to occur in areas with closer proximity to roads, presence of parking spots or trails, in areas with old growth redwoods, and that was more distant from campground sites. ImplicationsSeveral spatially explicit interventions are proposed to reduce natural resource theft in HRSP using situational crime prevention as a guide.