The impact of sown grass strips on the spatial distribution of weed species in adjacent boundaries and arable fields

Surveys were performed along paired transects running from the boundary to the field core and with and without sown grass strips. Their presence affected weed assemblages in the boundaries and over the first metre of the field margins as well as the patterns of weed species distribution. Sown grass strip effect always significantly explained the weed species assemblage in boundaries (23.6%), crop edges (28.4%) and field margins (8.1%) but not in field cores. Sown grass strips enlarged the habitat of boundary species (Plantago lanceolata, Arrhenatherum elatius, Carex sp.). They limited the spread of some species into the field (e.g. Sonchus asper, Anagallis arvensis). Some species were favoured by the establishment of sown grass strips (e.g. Cirsium arvense, Bromus sp., Elytrigia repens). However, sown grass strips acted as a buffer zone for arable species (Alopecurus myosuroides, Viola tricolor, Capsella bursa-pastoris), preventing them to extend their habitat from the fields to the boundaries. With short-term consideration, the establishment of sown grass strips puts areas of arable land aside, but offers an opportunity for many plant species to enlarge their habitat without increasing weediness.

► Sown grass strips limited the spread of many boundary species into the field.
► Sown grass strips acted as a filter for many non competitive annual species.
► Sowing 5 m wide grass strips along boundary causes a loss of diversity in the field.