Submitted to: European Weed Research Society Symposium Proceedings
Publication Type: Abstract only
Publication Acceptance Date: 2/25/2011
Publication Date: 3/3/2011
Citation: Davis, A.S. 2011. New approaches to understanding weed seed predation in agroecosystems. Proceedings of the European Weed Research Society Symposium on Biotic Diversity. 4:42. Interpretive Summary:
Technical Abstract: Postdispersal predation of weed seeds in arable systems can be a valuable ecosystem service, with the potential to support ecological approaches to weed management by reducing inputs to the soil seed bank. Scientific understanding of factors regulating weed seed predation rates is still insufficient, however, to enable producers to consistently manipulate predator activity with the aim of increasing postdispersal seed losses. Two recent methodological advances highlight key considerations for future studies of weed seed predation. First, whereas most measurements of weed seed predation are made at very short (hours to days) time scales due to rapid weed seed removal by granivores, the ultimate impact of seed predation on weed population dynamics is realised at a longer (monthly to annual) time scale. It is now possible to scale sequential estimates of seed predation rates made at a daily time scale to seasonal or annual rates using a simple function, verified through empirical tests in Europe and the USA. Iterative resampling indicates that for maximum predictive power, experimental units should be allocated primarily towards adequate temporal representation within the desired study period, with a secondary focus on spatial replication and comparisons of management treatments. Second, empirical studies weighing the relative importance of abiotic and biotic influences on postdispersal weed seed predation suggest that both bottom-up (intrinsic to seed predator) and top-down (extrinsic to seed predator) processes must be considered to fully understand the ecological context for seed predation. In particular, evidence of top-down control of both invertebrate and vertebrate granivores by their predators indicates that granivory should be studied within the broader scope of agricultural food webs.