Vegetation cover may affect weed seed predation by modifying the habitat quality for predatory organisms. Post-dispersal weed seed predation was measured by placing 'seed cards' in two perennial crops (alfalfa, cocksfoot) with and without crop cutting and in plots with bare soil. Each treatment was repeated four times in a randomized complete block design. Vegetation cover was measured by canopy light interception. Predation trials lasted two weeks and were repeated three times. Seed predation rates varied among three weed species (highest for Viola arvensis, intermediate for Alopecurus myosuroides, lowest for Sinapis arvensis). Vertebrate exclusion cages (12 mm x 12 mm openings) strongly reduced seed predation rates. Positive relationships were observed between vegetation cover and seed predation rates by both vertebrates and invertebrates for all weed species and trials, except when overall predation rates were very low. Predation rates were highest in uncut alfalfa, lowest on bare soil, but 16-64% of this variation could equally be explained by vegetation cover. The factorial design indicated that cutting had a stronger impact than crop species (legume or grass). Results suggest that weed seed predation may be enhanced by maintaining a high and temporally extended vegetation cover. (C) 2010 Elsevier B.V. All rights reserved.
