Integrating Spatial, Field, and Farm Data for Tracking Farmland Biodiversity and Ecosystem Condition

We developed an ecosystem account for a 38,603-hectare farm in southern Amazonia, Brazil, using integrated spatial, field, and farm-supplied data. About half of the property is covered by native forest designated as legal reserve, which provides an opportunity to assess how on-farm forests compare with reference conditions in the same biome. Our assessment focused on three pillars of ecosystem accounting using SEEA UN (System of Environmental Economic Accounting - United Nations) framework: extent, condition, and services.

To evaluate forest condition, we used several indicators, including net primary productivity (NPP) as a measure of forest growth dynamics, the percentage of area burned by fires, and the evolution of aboveground carbon (AGB). For biodiversity specifically, we relied on a single spatially derived indicator: tree species richness, obtained from a regional species-distribution dataset and validated with field observations across the property. Some indicators were benchmarked against values from a nearby pristine forest. Comparisons showed that on-farm forests maintain species richness and forest growth rates close to those of intact forest, suggesting that the legal-reserve area retains high ecological quality.

Still on biodiversity assessment, we used farm-provided management data to build an additional condition indicator: Aggregated Total Applied Toxicity (ATAT). This metric captures the cumulative toxicity of pesticides applied to groups of fauna and flora species. Mapping ATAT over time revealed rising toxicity loads in recent years and highlighted how some of the most affected farm's crop fields are spatially proximate to river catchments, underscoring potential risks to aquatic ecosystems. 

To deepen the assessment of ecosystem services, we conducted a 3-hectare forest inventory. More than 90 different tree species were identified along with their associated ecological functions, including potential pollinators, mycorrhizal fungi, and documented human uses (food, timber, medicine). While this inventory provided qualitative evidence of multiple potential ecosystem services, it also exposed methodological gaps. Current frameworks lack robust, quantitative tools for valuing the diverse services generated by species-rich tropical forests.