Annual Postdispersal Weed Seed Predation in Contrasting Field Environments

Abstract

Interest in weed seed predation as an ecological weed management tactic has led to a growing number of investigations of agronomic and environmental effects on predation rates. Whereas the measurements in most of these studies have taken place at very short timescales, from days to weeks, measurements at longer timescales (from several months to a year) have greater relevance to the demographic impact of weed seed predation and potential contributions from this process to ecological weed management. Our aim was to quantify the impact of crop phase, within a corn–soybean–wheat crop sequence, on quarterly and annual seed predation rates of giant foxtail, giant ragweed, and velvetleaf. The study took place in areas of the northern U.S. Corn Belt contrasting in dominant land use: Savoy, IL (2005–2007), where corn and soybean production predominates, and East Lansing, MI (2005–2008), where crop production occurs within an old field/forest landscape matrix. Mean annual rates of weed seed predation by the combined action of invertebrate and vertebrate predators were 31 ± 1.6% for giant ragweed, 37 ± 1.4% for velvetleaf, and 53 ± 1.4% for giant foxtail. Crop phase had negligible effects upon long-term seed predation rates, accounting for less than 2% of observed variation. Weed species and site-year, in contrast, contributed 35% and 40%, respectively, of the variation in cumulative annual seed predation. These results are consistent with the spatial variability in best management practices seen at spatial scales greater than the county level: weed seed predation appears to be an inherently site-specific phenomenon. New developments in managing weed seed predation as an ecosystem service are therefore likely to have local recommendation domains or to be driven by stochastic annual variation related to weather or granivore demography.