Skip to main content

Developing a biological network approach to quantify indirect costs and benefits of natural ecosystems to tropical agriculture

Final Report Summary - NETWORK (Developing a biological network approach to quantify indirect costs and benefits of natural ecosystems to tropical agriculture)

Ecosystem services (ES) have risen to prominence as a motivation for conserving biodiversity. The role of biodiversity in ES is, however, complex and depends on the intricate ways in which the components of ecosystems interact. Advances in network theory, and empirical studies, are now providing insight into how complex ecological systems of interacting species operate, and how influencing one part of a network can have consequences for the entire system. This fundamental ecological science is now at a stage where its potential for application to practical problems is ready to be tested. NETWORK brings together scientists with expertise in the empirical (UK), theoretical (France) and applied (South Africa) aspects that are required to make the transition from concept to practice possible. It will focus on a proof-of-concept case study on the interactions between mango plantations and surrounding natural ecosystems. Natural vegetation near farmland can sustain populations of natural enemies of pest insects and of pollinators but can also be a source of the pest insects themselves, thus providing both ecosystem services (ES) and disservices. A dynamic ecological network approach will be employed in this project to quantify these opposing effects in an important tropical crop (mango) under different management conditions. The project is one of the first studies to take an in-depth look at the trade-offs between ES and the costs of natural systems, giving insight into the quantitative effect of the natural ecosystem on crop production, and the development of methods for testing alternative management strategies in silico to aid informed decision making.
The objectives of this joint exchange programme are threefold:
1. To initiate a research programme and forge international links by exploiting the participants’ complementary expertise to transfer knowledge and skills, whilst conducting research that is relevant to global agriculture.
2. To use a network approach to quantify the costs, as a source of pest insects, and benefits, as a source of beneficial insects, of natural ecosystems to adjacent agricultural systems, thus investigating the link between biodiversity and the provision of ecosystem services.
3. To develop a predictive modelling framework for predicting effects of natural environment management strategies on crop production.

The research started off with an initial meeting at SANBI in Cape Town, South Africa during which the details of the field data required for the modelling work (obj. 3) were discussed and agreed upon by all project partners. This meeting was followed by a visit to proposed field location by the same team where contact was made with farmers, with whom the research was further discussed to identify the key economically important pest species to include in the research. Sampling transects in crops and natural vegetation were established and initial data gathered.
Following this initial campaign by the project leaders in late 2012 and until November 2015, there has been a near constant presence of groups of early stage researchers from the three partner institutions to gather year-round data on the network of interactions between crop, natural vegetation, pollinating insects, pest insects and the latter’s natural enemies. In addition to gathering the large network data set, the early career researchers have worked on a number of smaller projects aimed at testing the strength of indirect effects from natural vegetation and how strongly these effects are dispersal-limited. Understanding these aspects is important for informing the spatial scale (e.g. single field vs landscape) at which management of the vegetation will affect crop yield. The key result from these studies is that there are clear relationships between distance from natural vegetation and densities of pollinating insects, pest insects/pathogens in the crop. This indicates that the ecosystem services provided by the natural vegetation via insects acts at a sufficiently local scale so that they can be manipulated by farm-scale management decisions and do not require unpractical regional-level management to implement potential recommendations resulting from our research. Further evidence for this comes from our analysis of yield data (provided by farmers) in which we found a correlation between management practices that affect within-crop natural vegetation and yield, such that higher densities of natural vegetation are associated with higher yields (when controlling for other factors such as soil type, crop cultivar etc.). This research has so far led to three published papers (in the journals: Journal of Applied Ecology; Landscape Ecology; Agriculture, Ecosystems & Environment) and one that is in press (in the journal: Biocontrol). All these publications have an early stage researcher as lead author and are co-authored by researchers from all participating institutions.
The field collection of network data has provided a very large data set that shows clear indirect links between crop and natural vegetation via insects. Three South African early stage researchers have spent a period at the University of Exeter, learning how to analyse these network data. They have each lead-authored a manuscript on different aspects of the analysis and all three of these papers are currently undergoing peer-review.
One established researcher and four early stage researchers from South Africa have each spent a period at Universite Pierre et Marie Curie in Paris. Here they have learned aspects of network modelling and produced the first building blocks of a simulation model of the system. The assembly and validation of a full model is an ongoing project under UPMC leadership.
In July 2013, researchers from the UK and France held a workshop at the South African partner institute on applying and analysing networks in ecological research. This was attended by experienced and early stage researchers from universities, government institutions and industry (see attached participant photo).
Throughout the project, researchers have held meetings with the stakeholders (farmers) to relay results from aspects of the research and to discuss ongoing research plans. Input from the farmers helped us to ensure that the research produced information that was relevant to them and led us to extend the research to include a major crop pathogen. In the final year, the established researchers from all three institutions spent a month at the field site to discuss potential practical applications of the research with the stakeholders and to identify further research that was required and potential sources of funding for this.