Modeling Land Cover Dynamics Across Wildlife Management Areas and Surrounding Landscapes to Inform Conservation...

Rapid land-cover changes across managed landscapes threaten biodiversity by reducing habitat quality, fragmenting connectivity, and weakening ecological resilience, which can limit effective conservation. This study applies multi-temporal remote sensing, advanced classification algorithms, and Cellular Automata-Artificial Neural Network modeling to quantify and project land-cover transitions across each wildlife management area (WMA) and its surrounding 3-kilometer buffer in Mississippi, USA: Black Prairie WMA and Charles Ray Nix WMA. Land-cover dynamics were assessed over a decade (2014-2024) using one Landsat 8 image per WMA for each of the start and end years, chosen to capture consistent seasonal conditions and ensure unobstructed land-surface visibility. Predicted changes through 2034 reveal contrasting patterns. Black Prairie WMA shows notable shrubland decline and grassland expansion, reflecting structural homogenization, while Charles Ray Nix WMA maintains greater compositional stability despite some shrubland contraction and forest growth. These findings underscore the value of assessing land-cover dynamics within a broader landscape context that extends beyond core WMA boundaries. Incorporating the surrounding 3-kilometer matrix into a holistic assessment improves predictive capacity for conservation planning and identifies vulnerabilities that static analyses may miss. We recommend adaptive, landscape-scale management that uses spatially explicit projections to prevent habitat degradation, maintain functional connectivity, and support disturbance-dependent species such as the Northern Bobwhite. This approach advances conservation beyond reactive responses and toward proactive strategies that sustain biodiversity in human-dominated ecosystems.