An interdisciplinary approach to increase wheat within-field diversity and promote agro-ecosystem services

One major challenge for increasing agriculture sustainability is to better mobilize crop genetic diversity, as prone by agroecology. A simple way to increase within-field diversity is to use cultivar mixtures, and this has been successfully applied to a few crops in the past. Despite numerous scientific papers documeting the value of cultivar mixtures in wheat and other cereals, especially to control diseases, their cultivation has remained marginal throughout the world. To understand the origin of this gap between scientific knowledge and agricultural practices, the French project Wheatamix explored the synergies mobilized by cultivar mixtures, their impact on various ecosystem services, and their potential to reinforce the sustainability, resilience, and multi-functionality of agriculture. It focused on the agro-ecological and socio-economic impacts of variety associations at different scales, from the plant level up to the wheat supply chain. The project aims at developing new blending and breeding methods to design performing mixtures. To understand how plant-to-plant interactions shape wheat mixtures performances, Wheatamix has set five objectives: 1) describe the variability of morphological and ecological traits in a panel of 57 varieties; 2) explore variability by blending 16 contrasted varieties from the panel into 72 mixtures, composed of 2, 4, and 8 components; 3) study the ecosystem services provided; 4) assess the technical and economic performances in farmer conditions; 5) evaluate the impact of cultivar mixtures on the wheat supply chain. To achieve these goals, this project has developed an interdisciplinary approach, mobilizing agronomy, ecology, economics, ecophysiology, epidemiology, genetics, and management sciences. The project brought together scientists from 10 labs, as well as agricultural advisers and farmers from 6 French counties. The project first described the functional diversity of 57 varieties, highlighting the effects of modern breeding on trait variability, that lowered variability of traits subject to direct selection, and impacted both plant architecture, physiological traits as nutrient absorption, but also trade-off between traits. Wheatamix then surveyed how variation in mixture diversity impacted wild communities. A first result highlighted the low abundance of macro-organisms in this experiment: no relationship was found between the number of varieties in a mixture and the diversity/abundance of earthworms, weeds, mycorrhizae, springtails, beetles, nematodes. However, a significant effect of mixture diversity on the abundance of some spiders, and on nitrifying bacteria, was observed. Coming to ecosystem services, disease regulation (rust and septoria) has been confirmed as the most strongly and positively affected by varietal associations, raising also the strong effects of architectural variability of the canopy (septoria). Diversity also contributed to higher predation rates on aphids. Lastly, soil nitrification and denitrification activities were significantly affected by mixture diversity on 4 surveyed sites, contributing to a shift in plant nutrition and positive effect of greenhouse gas emission. Co-design of variety mixtures was carried out with farmers, technical advisers, and scientists. For three years, 30 farmers in the Paris basin proposed varietal blends and measured their performance on their farms. This exchange first highlighted that the first goals for farmers was to i) secure their production ii) simplify plot management. Then co-design workshops allowed to propose assembly rules and design mixtures, resulting in a wide diversity of sown mixtures. Field trials revealed that in more than 70% of the cases, the mixture had a higher yield than the mean of its components. This work highlighted farmers needs and resulted in a Multicriteria Evaluation Tool, helping farmers and advisers to design mixtures. The survey of the wheat supply chain finally highlighted the need for a concerted innovation among the various actors. Finally, Wheatamix also developed new statistical method to infer mixing ability, allowing both to blend the best mixers, and also to propose new breeding methods. Coupling various disciplines and approaches, such as ecophysiological modeling of plant competition (FSPM WALTer), field and controlled experiments, theoretical framework in ecology (sampling vs complementarity effects, functional traits and tradeoff), and mixture co-design and surveys with stakeholders, Wheatamix has allowed to understand the interest of cultivar mixtures for farmers. Wheat cultivar mixtures are experiencing an exponential growth: they only represented 2% of bread wheat sown in 2010, and are presently at 8%, raking at the first position on the cultivar list. Wheatamix emphasizes the need for an interdisciplinary approach when addressing agroecological subjects, and illustrates the strong mutual benefices between agronomic and ecological sciences.