Abstract
European organic production systems are often concerned by a "conventionalization" of their management methods, with a strong dependence on external inputs. Introducing or promoting a selected plant biodiversity in perennial crops could be an effective lever for moving towards a low-input agriculture, especially on fertilization and crop protection aspects. The six institutes involved in the DOMINO project worked on this research hypothesis from 2018 to 2021, in several organic apple and apricot orchards and vineyards. With experiments performed on nine locations, covering a wide range of climatic conditions (oceanic, Mediterranean and continental climatic areas) and various topographic situations (plain, low and high hills), this work focused on the use of perennial or temporary herbaceous cover crops, settled along the tree-rows or on the interrows, and used as components of a global production system, for the complementary services they can provide: substitution to the standard weed management practices, internal source of nitrogen supply, or additional income for the farmer by the use of secondary cash crops. 44 herbaceous species were tested to be used on the planted rows, with the primary goal to provide a sustainable alternative to the practice of row weeding by tillage. To limit or compensate their potential negative impact on the main crop's yield, some species were chosen for their low-growing habit, or their potential as nitrogen internal source. Officinal or edible crops were also tested, with a marketing perspective, to compensate competition with the perennial crop by adding value. To assess the performance of these plant species when implanted on the rows, their covering percentage was measured over the three years, and compared with those of weeds and the rate of bare soil. To evaluate their ecoservices for the perennial crop, their impact on some indicators of the physical, chemical, and micro-or macro-biological properties of the soil was also observed, while beneficial mites' populations were monitored on the crop's foliage. Trunk cross section areas, leaves nutrient contents, and commercial yields were also measured on the main crop to evaluate competition effects. Results showed that none of the herbaceous species that were tested can be proposed as a " turnkey solution " to the farmer, due to a great variability of adaptation to the ecosystem of the planted row. The use of species taken from local biodiversity provides significant advantages in terms of plant resilience and soil cover. Mentha x piperita or M. spicata seem to be rather ubiquitous plants, while strawberry species (wild or selected clones) and some clover cultivars appeared very dependent from rain (or irrigated) conditions. Overall, without additional weed control measures during the first two years of establishment, most of these species have great difficulty to compete on their own with weeds (especially summer weeds), even if they were chosen for expected allelopathic properties. In suitable local situations, strawberry plants, officinal plants and vegetable crops, such as pumpkin, resulted in interesting options as species able to provide an additional income, as they can give a significant production each year, while controlling weeds, and helping regulating aphids, mites, or nematodes. However, their use as cash crops is compromised in case of pesticide sprays to the main crop, even with PPPs authorized for organic systems. 14 legume species (sown alone or as mixtures with Poaceae) were tested to be used as green fertilizers, either as crops on the interrows, or as combined strategies on the rows and the interrows.