Agricultural practices rely on the use of monocultures to achieve high yields. Current breeding methods may help crops adapt to diverse stresses whilst maintaining productivity in conventional farming, but these methods rely heavily on chemical inputs such as fungicides, pesticides and herbicides, the use of which will be tightly restricted in the future. Mixtures of different crop varieties within a single field may provide a response to the challenge of increasing yield stability in the increasingly variable environments predicted to result from climate change, without high dependence on chemical inputs. The greater yield stability of varietal mixtures has been attributed to higher levels of within-field genetic diversity. Our understanding of the specific mechanisms underlying this enhanced yield stability, however, remains limited. We have investigated the suitability of Arabidopsis thaliana as an ecological model for varietal mixtures. Experiments were designed to imitate growing conditions of arable farming, with plants grown at high densities to generate high competition for resources. Multiple replications of a large scale glasshouse experiment were conducted at different times of year, leading to wide variation in glasshouse environments between replicate experiments. An additional nutrient stress was applied to some plants to increase stress and competition levels further.
Results/Conclusions
Combined data indicate a positive effect of mixtures in stabilising yield across environments. Yield stability could be attributed to the performance of the most competitive genotype which overyielded in mixtures thereby compensating for loss of yield associated with less competitive genotypes. Phenotypic characteristics such as large rosette and longer flowering time increased the competitive ability of genotypes in mixtures. To test this hypothesis about stabilisation of yield through competition, a second set of Arabidopsis accessions with different phenotypes was selected; in mixture experiments, their phenotypes proved to be a fairly accurate predictor of competitive ability. These results indicate that the mixing ability of crop varieties can be predicted from their phenotypes, and suggests that an initial phenotypic screen can accelerate the selection of suitable components for varietal mixtures. Our study also highlights the usefulness of Arabidopsis as a model for ecological studies.