Incorporating drone imagery and field-data to assess geographic and social drivers of agroforestry biodiversity in western Ecuador

Rapid deforestation for agricultural commodities continues to threaten tropical forests, which harbor unparalleled biodiversity and provide essential carbon sequestration. Agroforestry offers a restorative solution to forest loss by integrating trees into agricultural landscapes, allowing farmers to profit from tree products while maintaining ecosystem health. Despite its benefits, the widespread adoption of agroforestry by smallholder farmers is often hindered by high establishment costs. Payment for Ecosystem Service (PES) programs present a promising opportunity to support agroforestry expansion, yet limited monitoring strategies constrain efforts to scale up restoration and evaluate factors influencing farm biodiversity. Organizations promoting biodiverse agricultural landscapes face challenges in assessing diversity due to restricted farm access, dangerous terrain, lack of time, and financial constraints. In collaboration with a local organization funding agroforestry adoption in Ecuador's highly degraded Pacific Forest, we acquired high-resolution RGB drone imagery and field data with help from local botanical experts to identify individual trees in drone imagery. We then calculated farm-level diversity metrics in order to analyze their relationships with social and geographic characteristics of 30 farms in the region. Results indicated that elevation was the main factor driving tree species richness, with little effect of farm size, farm range of elevation, or farm steward gender. Understanding how environmental factors relate to variation in tree diversity across farms can inform targeted assistance for farmers and enhance the effectiveness of community-led efforts to restore these critical forest ecosystems.