Unequal Greening and Urbanization Trajectories in Santiago de Chile, and Their Consequences for an Urban-Adapted Woodpecker

Urban green infrastructure and its potential biodiversity benefits are unevenly distributed, but species-level consequences remain poorly characterized in Latin American cities. We combined Sentinel-2 and Dynamic World time series (2015–2025) with effort-filtered eBird records to model the encounter probability of the Striped Woodpecker or Carpinterito (Veniliornis lignarius) within Santiago de Chile’s fixed 2017 urban extent. Generalized additive models were fitted to 4,419 checklists aggregated on a 100-ha hexagonal grid, and the selected habitat relationship was applied to annual NDVI, tree-cover, and built-cover rasters to generate standardized model-based projections. Encounter probability was nonlinearly associated with NDVI (edf = 2.826, $p<0.001$), with a curve compatible with diminishing gains at higher greenness. Model performance was stable across repeated thinning realizations (median geographically clustered AUC = 0.698), while forward temporal AUC ranged from 0.713 to 0.735 and calibration improved when recent years were included. Study-area mean projected encounter probability fluctuated between 0.107 and 0.135, whereas 2015–2025 endpoint change was positive in 70.7% of hexagons and negative in 29.3%. After false-discovery-rate correction, only 10 of 941 hexagons retained significant monotonic trends. Signed endpoint change was strongly spatially autocorrelated (Moran’s $I=0.603$, $p=0.001$), and gains and losses were unevenly concentrated (Gini = 0.485 and 0.584, respectively). Commune-level income and multidimensional poverty were negatively associated with local clustering statistics, providing descriptive socioeconomic context. Thus, Santiago’s annual environmental configurations produced spatially uneven model projections. Urban-forest planning should complement greenness targets with retention and monitoring of large, senescent, and cavity-bearing trees, particularly in under-canopied and disadvantaged areas.