Modelling yields of multispecies grassland mixtures across a climatic gradient: results from an international biodiversity experiment
James O'Malley, Trinity College Dublin.
2nd of October 2025
Abstract:
Multispecies grassland mixtures have been proposed as one pathway to more sustainable farming practices in the context of the global climate and biodiversity crises. These mixtures also have the potential to deliver higher yields than single species communities through the synergistic interactions between different plant functional groups. To date however, much of the research in this area has been conducted in single site studies with varying experimental designs. To address this issue the LegacyNet international network of experiments implemented a common experimental design at 26 sites across 15 countries. This talk will present the results from the analysis of the grassland yields in this multi-site experiment. We fit a linear mixed-effects Diversity-Interactions model with random coefficients to model the between- and within-site variation in the data, and to quantify the magnitude of species’ identity and interaction effects. Using this model, we predict the yields of the best performing mixtures on average across sites, and how these compare to the yields of conventional farming practices across a range of different climates. Through these predictions we identify the compositions of a wide range of multispecies mixtures that allow more targeted design of grassland communities to suit differing farming systems, without compromising yield. These results may inform on the potential of multispecies mixtures as a practical farm management tool on an international scale, and may encourage farmers to adopt more environmentally, and financially, sustainable practices.