This study examines the spatial patterns and key drivers of outdoor recreation participation in the 14 National Forests (NFs) of USDA Forest Service Region 8—covering thirteen southeastern states and El Yunque NF in Puerto Rico (15 forest units in total)—based on data from the USDA National Visitor Use Monitoring (NVUM) 2010–2014 microdata cycle. We postulated that spatial autocorrelation is statistically significant for individual recreation drivers, particularly around forest boundaries and major road networks. We test for spatial autocorrelation using Global Moran’s I and Local Indicators of Spatial Association (LISA) and identify hot spots with Getis–Ord Gi* statistics. Spatial regression models (OLS, spatial lag, and spatial error) are estimated to assess the effects of proximity to major roads and distance from forest boundaries on population-normalised visitation intensity. We find significant spatial autocorrelation in overall visitation intensity (Global Moran’s I = 0.312, p < 0.001), with high clusters observed within 50 km of forest boundaries and along major Interstate highway corridors. At least four of five key recreation drivers are significantly clustered. Our results provide spatially specific, statistically robust evidence to inform NF management.
Jean et al. (2026) studied this question.