The diversity of agroforestry systems across regions. From left to right, a coffee system in Tanzania, a coffee system in Costa Rica, and a cocoa system in Ghana.
Photo credit: Sigrun Wagner

By Clément Rigal, Sigrun Wagner, Mai Phuong Nguyen, Laurence Jassogne, and Philippe Vaast

Read the full paper here.

Farmers looking to include shade trees in their fields, either establishing new or modifying existing agroforestry systems, are constantly confronted with the question; what tree species? Tree species selection should aim to maximize the provision of ecosystem services while minimizing negative impacts. Considering the complexity of these systems, selecting tree species should account for multiple aspects including local agroecological conditions and farmers’ priorities or constraints.

The ShadeTreeAdvice methodology was developed to support this selection process using farmers’ local ecological knowledge. It provides the steps to rapidly identify tree species and evaluate their impacts on a range of locally important ecosystem services. Results are uploaded to a decision support tool to tailor tree species recommendations to individual farmers’ needs (www.shadetreeadvice.org). The tool can be used by governmental agencies, extension services, private sector, or NGOs; end-users are farmers and agroforesters. The database currently comprises information on more than 160 tree species and more than 10 ecosystem services and disservices, with studies spanning coffee and cocoa growing regions in Africa, Asia, and Central America.

In this review, we examine the eight studies conducted during the first five years after the methodology was developed (2016-2020). When synthesizing the findings, we identified some similarities across studies, especially on the importance of fruit trees in these systems and the use of leguminous trees for the provision of ecosystem services. We further evaluated the methodology to identify strengths and weaknesses and develop improvements. The method can successfully evaluate tree species’ impacts on soil and climate regulation, crop production, and economic outcomes. It can also identify links between local ecological knowledge and socio-economic groups or environmental factors. Limitations are met for evaluating impacts related to the incidence of pests and diseases, often associated with knowledge gaps.

Our suggestions for future studies using this tool include (i) broadening the scope beyond ecosystem services to include tree species’ impact on farming practices; (ii) including a non-shaded scenario as comparison; (iii) providing a clear pathway for validating results; and (iv) using tree species’ functional traits to generalize the results.