Payments for ecosystem services (PES) have been developed as a policy instrument to help safeguard the contributions of ecosystems to human well-being. A critical measure of a programme's effectiveness is whether it is generating an additional supply of ecosystem services (ES). So far, there has been limited analysis of PES programmes based on the actual supply of ES. In line with ecosystem accounting principles, we spatially quantified three ES recognised by Costa Rica's Pago de Servicios Ambientales (PSA) programme: carbon storage, soil erosion control and habitat suitability for biodiversity as a cultural ES. We used the machine learning algorithm random forest to model carbon storage, the Revised Universal Soil Loss Equation (RUSLE) to model soil erosion control and Maxent to model habitat suitability. The additional effect of the PSA programme on carbon storage was examined using linear regression. Forested land was found to store 235.3 Mt of carbon, control for 148 Mt yr−1 of soil erosion and contain 762,891 ha of suitable habitat for three iconic but threatened species. PSA areas enrolled in the programme in both 2011 and 2013 were found to store an additional 9 tonC ha−1 on average. As well as enabling a direct quantification of additionality, spatial distribution analysis can help administrators target high-value areas, confirm the conditional supply of ES and support the monetary valuation of ES. Ultimately, this can help improve the social efficiency of payments by enabling a comparison of societal costs and benefits.
- Carbon storage
- Ecosystem accounting
- Ecosystem services
- Machine learning
- Payments for ecosystem services