Using satellite data to assess spatial drivers of bird diversity.

Area of Science:

• Ecology
• Conservation Biology
• Remote Sensing

Background:

• Global biodiversity is declining, with birds serving as crucial indicators of ecosystem health.
• Understanding the spatial drivers influencing bird distribution is essential for effective conservation strategies.

Purpose of the Study:

• To assess the importance of satellite-derived habitat productivity, heterogeneity, and landscape structure for modeling bird distributions.
• To identify key spatial drivers influencing woodland and farmland bird diversity and richness across Great Britain.

Main Methods:

• Utilized Random Forest (RF) regression to model bird diversity and richness using satellite-derived covariates.
• Employed feature contribution analysis to understand the relationships between spatial drivers and bird species richness.
• Incorporated monthly habitat-specific productivity values and landscape structure metrics (FRAGSTATS).

Main Results:

• Habitat productivity and landscape structure metrics explained a significant portion of bird distribution variation (R² 0.64-0.77).
• Identified high-productivity grassland in spring as a key driver for farmland birds.
• Highlighted woodland patch edge length as an important factor for woodland birds.
• Revealed specific relationships, such as the positive effect of high NDVI in broadleaved woodlands on bird richness.

Conclusions:

• Satellite-derived spatial data effectively models bird diversity and richness.
• Specific habitat characteristics (e.g., spring grassland productivity, woodland edge) are critical spatial drivers for different bird groups.
• The identified relationships offer a valuable tool for analyzing and comparing the impact of spatial drivers on bird populations.