How Environment and Urbanization Shape Bird Diversity in Sri Lanka

This study investigates the environmental and anthropogenic drivers of bird diversity across Sri Lanka. By integrating citizen-science observation records from eBird with remote sensing data—including NDVI, land-cover classifications, and Artificial Light At Night (ALAN)—the researchers sought to determine which factors most strongly influence species richness. The team employed a multi-scale spatial analysis (2 km, 5 km, and 10 km grids) and used Negative Binomial Generalized Linear Models (GLMs) to account for overdispersion in the count data, while applying spatial thinning to mitigate observer bias.

The analysis reveals that categorical land-cover types are superior predictors of bird richness compared to continuous variables like NDVI or climate metrics. While NDVI is a common proxy for ecological productivity, the study demonstrates that it fails to capture the structural habitat complexity provided by different land-cover classes, such as evergreen forests versus croplands. Furthermore, the study finds that urbanization (measured by ALAN) has a nuanced impact: it is associated with a decline in community evenness and the dominance of a few generalist species, but it does not significantly reduce overall species richness once observer effort is properly accounted for.

This research provides a scalable, reproducible framework for biodiversity assessment in tropical regions where data is often fragmented. By demonstrating that land-cover structure is a more reliable indicator of avian diversity than simple greenness indices, the study offers actionable insights for conservation planning. It highlights the importance of protecting specific habitat types—such as evergreen broadleaf forests and woody savannas—to maintain high species richness, rather than relying on broad environmental proxies that may mask critical ecological nuances.