Submitted to: Aspects of Applied Biology
Publication Type: Proceedings
Publication Acceptance Date: August 1, 2003
Publication Date: September 17, 2003
Citation: Teasdale, J.R., Mangum, R.A., Radhakrishnan, J., Cavigelli, M.A. 2003. Factors influencing annual fluctuations of the weed seedbank at the long-term beltsville farming systems project. Aspects of Applied Biology. 69:93-99. Interpretive Summary: Weeds are an important limitation to crop production worldwide. One factor that contributes to the perpetuation of weed problems is the persistence of weed seeds in agricultural fields for many years. An understanding of the changes in weed seed populations in the soil is needed in order to improve management of agricultural weeds. This paper reports on changes in the soil weed seed population over six years in a long-term cropping systems experiment at the USDA-ARS Beltsville Agricultural Research Center. The analysis focused on identifying factors that influence the annual changes in these seed populations. A model was produced showing that the majority of these changes could be explained by two factors, one, the abundance of weeds leading to seed production during the year, and, two, the rate that viable seed were lost from the soil. This research will help scientists develop models for better understanding weed seed changes in soil and management strategies for reducing weed populations in agricultural fields.
Technical Abstract: The weed seedbank will change annually according to variation in management and weather factors. This paper explored the relationship between above and below ground weed populations in a long-term Farming Systems Project that included recommended, reduced input, and three organic cropping systems. There was a high degree of year-to-year variation in both weed seedbank and aboveground weed cover but no trend toward increasing seedbank over time. Multiple regression analysis showed that much of the annual change in the seedbank of Amaranthus hybridus L., Chenopodium album L., and annual grasses could be described as a function of two variables, weed cover and initial seedbank, with an R-squared of 0.63, 0.61, and 0.81, respectively. Weed cover served as a proxy for seed input whereas the rate of seed loss was proportional to the initial seedbank. The balance of these opposing processes defined an equilibrium around which seedbank populations oscillated.