Search for a command to run...
Cereal-legume intercrops are developed mainly in low input or organic farming systems because of the overyielding and numerous ecosystem services they provide. For this management, little advice is available for varietal choice and there are almost no specific breeding programs. Our study aimed to evaluate the mixing ability of a panel of bread wheat genotypes in intercropping and to assess the impact of environment and legume tester choice on this ability. We used partial land equivalent ratios (LERs) to assess the mixing ability of a genotype defined as the combination of its ability to maintain its own yield in intercropping (producer effect, LERw) and to let the mixed species produce (associate effect, LERl). Eight wheat genotypes and 5 legume testers (3 pea and 2 faba bean varieties) were grown in sole crop and in all possible binary intercrops in nine contrasting environments. A mixed model was used to evaluate the effects of wheat genotypes, legume testers, environments, and all the interactions among these 3 factors on LERw and LERl. The chosen wheat genotypes presented contrasting mixing ability, either in terms of producer effect (LERw) or associate effect (LERl). A strong negative correlation was observed between these two components of genotype mixing ability, with an increase in producer effect being generally associated with similar decrease in associate effect, except for three genotypes. The impact of environment on the producer and associate effects was limited and similar between genotypes. Legume tester had a significant effect on both LERw and LERl, making the choice of tester a major issue to reveal the producer or associate effects of wheat genotype. Although the 5 testers showed no significant differences in wheat genotype order for both producer or associate effects, they showed different competitiveness and ability to discriminate genotypes: faba bean was very competitive, resulting in low LERt and low capacity to discriminate wheat genotypes for their mixing ability. On the contrary, pea was less competitive, resulting in higher LERt and better capacity to discriminate wheat genotypes. In particular, the Hr varieties (Geronimo and Spencer) discriminated best the wheat genotypes. Consequences on the implementation of breeding programs for wheat varieties adapted to intercropping are discussed.