Abstract

Aaron O. Benjamin, David Taylor Gibb and Hugh Tom Johnson

It is essential that the net effect of biodiversity on crop yields is determined; particularly in developing nations, where both increasing food security and reducing biodiversity losses are of high importance. This study modelled the abundance of pests, pollinators and pest-control animals and determined their impact on crop yield within agroecosystems in four rural villages in Malawi. Data on the habitat area, survivorship, fecundity, birthing month and effect on crop yield for 14 animal functional groups were collated through a focused meta-analysis. Using this data, models were created to determine the abundance of each functional group using land cover as the sole input variable; with Participatory Rural Appraisals (PRA) utilised to validate the model prediction. Bees, birds and insects always improved crop yield, whereas monkeys, rodents and large herbivores always result in losses. Three out of four villages experienced a net benefit to crop yield from the animal biodiversity present. We conclude that models derived from metaanalyses appear useful for broadly predicting the local-scale abundance of functional groups and their qualitative impact on crop yield. However, long-term field observations should be conducted to ensure that the PRA values in this study correlate with direct observation.