Efficacy of agroecological approaches to ameliorate trade-offs between food production and biodiversity conservation

Halting agricultural expansion into natural ecosystems and scaling up ecological restoration efforts will be critical to revert global biodiversity loss. Concurrently, the nutrition needs of a growing human population must be met, which necessitates increasing, or at the very least maintaining, current agricultural productivity. However, conventional agricultural intensification frequently erodes farmland biodiversity and the land’s capacity to produce food over the long term. Consequently, there has been significant interest in (agro-)ecological intensification approaches, which entail managing agricultural landscapes to enhance ecosystem services that support productivity. However, little is known about the efficacy of these approaches to balance trade-offs and foster co-benefits between different land system objectives, including between food production and biodiversity conservation. In this thesis, I first review counterfactual techniques (which are needed to elucidate causal relationships and assess impacts credibly) and argue for a step-change in their use at the food-biodiversity nexus, including to answer key questions related to (agro-)ecological intensification (Chapter 2). I review the necessary conditions for causal inference and demonstrate how these conditions are easily violated, especially where interventions are strongly mediated by human behaviour change. I provide a comprehensive guide of design-based and methodological solutions, discuss major knowledge gaps that should be addressed by them, and suggest pathways to mainstreaming counterfactual approaches. I then employ some of these counterfactual methods to assess the multidimensional impacts of one of the largest agroecological transitions globally: the Zero Budget Natural Farming (ZBNF) programme (Chapters 3 and 5) and of (semi-)native woody vegetation patches embedded in agricultural landscapes in Andhra Pradesh, South India (Chapter 4). In Chapter 3, I present results from extensive bird surveys I conducted in ZBNF, agrichemical (conventional), and natural forest landscapes, and from interviews of farmers. I show that the ZBNF programme more than doubled farmers’ economic profit and improved bird conservation outcomes without lowering agricultural productivity. ZBNF was also largely able to dampen bird density-yield and density-profit trade-offs in comparison to agrichemical-based farming. Nonetheless, whilst the bird community similarity to natural forests was improved, ZBNF still has very limited value in conserving forest-specialised species. Chapter 4 is based on the same field data, but I use it to examine the relationships between the proportional cover of vegetation patches and yield as well as bird densities. Almost all existing studies examine the effect of such vegetation patches on yield at field rather than landscape scale, thus not accounting for the land taken out of production by the patches. In my system, I find that landscape-level yield was maximised in vegetation-free landscapes. However, increasing cover from 10 to 20% did not result in further yield losses, possibly via enhanced ecosystem services such as pest control where most ecosystem service-provisioning bird species required at least ~10% vegetation cover. Increasing vegetation patch extent also positively affected the presence of forest-dependent birds and increased the community similarity to natural forests. In Chapter 5, I again examine the effect of the ZBNF programme, using a different set of sites. I conducted flower-insect timed counts at ZBNF and agrichemical cashew orchards, and I interviewed farmers. I show that cashew flower visitation rate, and the abundance and species richness of flower visitors was higher for ZBNF than agrichemical farming. However, this did not necessarily translate into higher yields. The effect of ZBNF on cashew yield was neutral, albeit with stronger support for it being positive than negative. Lastly, I organised a workshop on the food-biodiversity nexus in Andhra Pradesh (Chapter 6). Through deliberative dialogue and nominal group techniques, I asked participants to identify the key policy transformations needed and the main knowledge gaps that must be filled in order for Andhra Pradesh’s landscapes to produce enough nutritious food for people whilst protecting and restoring biodiversity. I present the results of this multi-stakeholder exercise. In brief, policy spheres are currently disjunct, with the ZBNF programme not well aligned with state-level conservation strategies. Structural support across value chains, intersectoral collaborations, and landscape-scale land-use planning will be critical. In all, I demonstrate the potential for agroecological approaches (specifically ZBNF and interwoven vegetation patches) to better balance land system trade-offs and to occasionally craft synergies. However, my work also clearly highlights their limitations. Negotiating fair and acceptable trade-offs and compensations will continue to be crucial, where there is a clear need and demand for assessing multidimensional impacts at appropriate scales using counterfactual methods. Moreover, to harness the potential of agroecological approaches and avoid negative repercussions, integrated multi-scalar governance is needed to ensure that these approaches are coupled with effective area-based conservation efforts, and are part of broadly beneficial landscape-wide management regimes and wider food system transformations towards sustainability.

Whitten Studentship