Winners and losers in a transforming landscape: assessing the impact of dry forest conversion on high-risk species in Michoacán, Mexico (1986–2025)

Michoacán possesses extensive forest cover yet has undergone rapid and sustained land use transformations driven by illegal logging and the expansion of the ‘avocado strip’. This study utilized Random Forest (RF) classification of Landsat imagery (1986–2025) to analyze landscape dynamics, identifying Near-Infrared (NIR) and Shortwave Infrared 2 (SWIR 2) bands as the most critical predictors. Change detection analysis revealed severe deforestation: dry forest declined by 4,344 km2 and evergreen forest by 1,698 km2, fueling a corresponding increase in exposed soils (3,270 km2) and croplands (2,158 km2). To assess ecological impacts, MaxEnt was employed to model the distribution of 35 vulnerable species (16 Near Threatened, 10 Vulnerable, and 9 Endangered) with Nighttime Land Surface Temperature in December emerging as the most influential variable. The biodiversity assessment (2000–2025) indicated that the Reddish Egret (Egretta rufescens) suffered the greatest loss of potential distribution area (exceeding 50%), while the Grey Plover (Pluvialis squatarola) and the Mexican Ground Pit Viper (Agkistrodon bilineatus) exhibited the greatest gains (approximately 50% increase). Given that these threatened ecosystems are poorly represented within current protected areas, these findings underscore an urgent need to integrate conservation strategies into regions undergoing intense agricultural conversion to mitigate further biodiversity loss.