CIFOR and its partners recently completed a near three-year review process that examines how forests and trees can support, or even inhibit, agricultural production.
As the study’s co-authors, we sought evidence from across the tropics, where food production often occurs within complex land use mosaics that are increasingly subject to concomitant climatic and anthropogenic pressures.
Our review focused on forests and trees, both within the farm and across the wider landscape, and was concerned with the indirect provisioning of ecosystem services, as opposed to the direct supply of food (fruits, nuts etc.) from trees themselves.
After screening almost 12,500 documents for relevance, we were left with a sample of 74 studies that we included for analysis in the final review. These studies were distributed across the tropics, with India, Kenya, Brazil, Indonesia, and Malawi as the most commonly represented geographical regions.
Our findings were both varied and surprising. They reveal significant potential for integrated tree-crop systems, but also highlight considerable gaps in our understanding of the broader landscape implications.
POSITIVE YET VARYING RESULTS
Captured studies evaluated the effect of single or multiple ecosystem services (i.e. pollination, pest control, nutrient cycling) on one or more stated outcomes (i.e. crop yield, income, soil fertility). This variability in the types of studies screened made comparative analysis problematic, and yet in most cases, trees were found overall to have a positive effect on food yields.
While this is an encouraging result, our findings were not spatially consistent: Africa, Australia and the Americas exhibit a positive relationship between tree cover and agricultural production while in Asia, trees were found to have an aggregate negative impact on yields. This finding represents a first gap in our understanding— Why is that trees on farms are not providing similar returns in Asia compared to elsewhere? —and certainly warrants further consideration.
We further investigated the effects of trees on agriculture by examining how they contribute more broadly to rural livelihoods (i.e. through the provision or sale of fuelwood, or via positive environmental impacts such as improved soil fertility). Once again, the overall findings were encouraging. We found a positive effect of trees on livelihood outcomes across the tropics. Even in Asia, where yields were reduced, the system-wide effect was often positive as yield losses were compensated for by the broader net contributions of trees to livelihoods.
MORE RESEARCH REQUIRED
A key objective of this review was to better understand how forests and trees within the landscape contribute to food production systems. However, a lack of evidence prevented us from exploring this question in any considerable detail. Of the 74 studies, only 12 focused on off-farm trees, and 11 of these were examining the effect of forest distance gradients on a single ecosystem service: pollination.
While such studies are useful, there is an urgent requirement for studies that examine the multiple ecosystem service effects of forests and trees to agro-ecosystems.
A further concern was highlighted by the fact that the vast majority of studies were conducted over very short time period: one to two years in most cases. There is good evidence of the need to consider ecosystem processes over time, and therefore the short timeframes of the reviewed studies limits our understanding of the temporal contribution of forest or tree-based ecosystem services.
The findings of this review suggest that when incorporating forests and trees within an appropriate and contextualized natural resource management strategy, agricultural yields can be maintained or enhanced comparable to intensive monoculture systems. Furthermore, this review has illustrated the potential of achieving net positive livelihood gains through the integration of trees on farms, providing rural farmers with additional and diverse income sources and greater resilience strategies to adapt to both environmental and economic shocks.
However, the limited spatial and temporal scale of the current evidence suggests a degree of caution must be applied. A future research agenda that considers the effect of forests and trees within the landscape over longer time periods would greatly enhance our understanding of how forests and trees interact with agricultural systems. Developing this evidence base further is essential as we grapple with delivering sustainable development that utilizes, rather than depletes the natural resource base.