Disentangling farmers' preferences for conditional incentives to upscale biodiversity-friendly agroforestry in the Amazon

Although agroforestry is a promising approach to reconcile biodiversity conservation and local development, uptake has proven challenging, justifying the introduction of incentives to promote adoption. We firstly conducted a discrete choice experiment (DCE) over a sample comprised by 223 cacao farmers from the province Padre Abad in the Peruvian Amazon to gain insights into their preferences for four attributes of a proposed biodiversity-oriented agroforestry- Payment for Ecosystem Services (PES) scheme: (1) requirement for tree canopy cover in enrolled plots, (2) modalities for a biodiversity connectivity bonus, (3) farmers' participation in monitoring, and (4) the compensatory cash amount. Although we found that a PES scheme would have a wide acceptance among eligible farmers, their preferences are highly heterogeneous regarding assessed attributes. For disentangling such differences, we perform a latent class (LC) model and identify three farmer classes: first, the larger one (56% of the sample) comprised by farmers that are more likely to enroll in the proposed PES as the cash compensation for accepting both high required canopy level and collective bonus rises, and they are allowed to participate in monitoring tasks. Meanwhile, preferences of the second group (34% of the sample) seemed to be less sensitive to changes in levels of the assessed PES attributes. Finally, the third farmers group (10% of the sample) features, in average, a willingness to accept (WTA) lower than the first farmers group for required high canopy level and a larger preference for individual rather than collective connectivity bonus. These findings highlight the necessity of tailoring incentives to address differentiated farmer preferences for increasing scheme effectiveness and equity.