Experimental analysis of the effect of relay intercropped legume cover crops with winter wheat, in organic crop rotations, on crop performance, weed control and nitrogen dynamic ; Analyse expérimentale de l'effet de couverts de légumineuses associés en relais à un blé d'hiver, conduit en agriculture biologique, sur les performances des cultures, la maîtrise des adventices et la dynamique de l'azote

Cereal productivity and quality are subject to two main problems in organic stockless systems: chronic soil nitrogen (N) deficiencies and weed infestation. Legume species as clovers or alfalfas can be used as cover crops and green manures due to their natural ability to fix atmospheric N. Nevertheless, their substitution to cereals in crop rotations is less economically viable without animals to use it. Relay intercropping of legume cover crops (RIL) in winter wheat was expected to simultaneously enrich the soil-plant system in N, cover the soil from the wheat harvest onwards and limit the risk of competition with wheat by delaying the undersowing of legumes in spring. To evaluate the efficiency of RIL, four legume species (Medicago lupulina L., M. sativa L., Trifolium pratense L. and T. repens L.) were undersown at wheat tillering stage, in eight fields organically managed in the Rhône-Alpes region, France. Their effects on weed control, on N enrichment, preservation and restitution in the soil-plant system and on crop performance were observed during the succession of the winter wheat and a spring crop. RIL residues were incorporated in soil in late winter, 9 to 12 weeks before the sowing of spring crops. Our work illustrated the absence of detrimental effect of RIL on wheat grain yield despite the decrease of the grain protein content in one third of the situations. This decrease was mainly observed with M. lupulina and T. pratense as they were the most developed legume species at wheat harvest. Our monitoring of trophic resources (water, N and light) enlightened the competition for soil water and N during the intercropping period. We also showed the efficiency of RIL in the control of weed density from wheat flowering stage onwards and of weed aerial biomass in late autumn. The best weed control was observed with M. lupulina and T. pratense, at wheat harvest, and with T. pratense and T. repens in late autumn, in relation to the highest aerial biomasses observed with these species. Finally, we noted an ...