First results of weed evolution in no tillage systems in winter wheat in France

International audienceThe adoption of no-tillage is presently a strong trend in Europe. If this cropping system is to be widely adopted in the 1980s in the North and South America, in Australia, no tillage is used by few farmers in France where agriculture and tillage are considered synonymous by farmers and agronomists. Weed infestation is generally described as to increase when the selection pressure due to soil tillage is reduced. The European strengthened regulations on herbicide used and the ban of transgenic crop in France does not facilitate the development of non-tillage. This communication presents the first results on the evolution of weed communities on farmer’s fields in East of France. The device included 52 fields (50m x 40 m plots) that are in no tillage for periods of 1 to 13 years. The area is dominated by shallow calcareous soils and the major local crop rotation is winter rapeseed, winter wheat, spring barley. Weed surveys were carried out across winter wheat fields in 2011 and weed data was collected at three periods i.e. the cropping period in late- March before foliar herbicide treatment, before harvest in mid-June (potential seed production) and during the cover cropping period in mid-September. Seven environmental variables were collected that were related to cultivation practices and to soil properties. Thirteen species traits were selected as potentially affected by the adoption of no tillage. A RLQ analysis was performed to investigate the relationship between traits of species and environmental fields characteristics. A total of 121 weed taxa were identified but analyses were carried out with a total of 93 weed species (62% of annual species; 94 of eudicot species). Results indicated that grassweeds and perennials weed species were promoted with time since field conversion to no tillage contrary to annuals and broadleaf species. Data indicated that time since conversion to notillage led to a shift in weed community that favored weeds invested more in maintaining their “underground system than in aerial seed production. Furthermore, the duration since conversion was linked to changes in soil parameters. We observed an accumulation of stones on the soil surface and a trend towards a soil acidification in fields converted for several years. The expansion of monocots in no-till systems may be explained by the end of mechanical weed control and the use of foliar-applied- herbicides generally less efficient on monocot species. If geophyte weed species seems to be favoured, we did not observe a clear gradient reflecting ecological community succession and annual species remained present in the fields. One explanation could be that the gradient of time since conversion was too low. We could also hypothesize that weed communities, which may exhibit long-lived seedbank, may respond with a delay to the changes of soil tillage regime. This study provides empirical evidence of weed trait shifts in conservation agriculture and calls for a move in management practices towards a regime avoiding selection of problematic perennial weeds