Application of Weed Seedbank Ecology to Low-Input Crop Management

Ecol Appl. 1993 Feb;3(1):74-83. doi: 10.2307/1941793.

Abstract

Pesticide use in the United States is concentrated in the Corn Belt, mainly in the form of herbicides on corn (Zea mays) and soybean (Glycine max). If this chemical load to the environment is to be reduced, without appreciably affecting crop yields, an intimate understanding of weed ecology is necessary. For annual weeds in the U.S. Corn Belt, critical ecological information includes seedbank density, seed dormancy, seedling emergence, and environmental variables that regulate these factors. For three weeds (Setaria spp., Amaranthus retroflexus, and Chenopodium album) in Minnesota, this information is summarized and its application to weed management is illustrated. If weed seedbank densities (0-10 cm depth) are <100 seeds/m2 prior to spring tillage operations, subsequent seedling populations are too low (<40 seedlings/m2 ) to require control. Seedbanks ranging from 100-1000 seeds/m2 produce seedling populations (<400 seedlings/m2 ) that can be controlled mechanically. In contrast, pre-tillage seedbanks >1000 seeds/m2 typically result in such high seedling populations that nonchemical control measures alone cannot prevent large crop yield losses. About one-half of all arable fields harbor seedbanks with densities >1000 seeds/m2 . The proportion of buried seeds giving rise to seedlings is controlled by seed dormancy, which in turn is governed partially by April temperatures (in Minnesota), whereas the timing of subsequent seedling emergence is determined by soil temperature and soil water. EMERG models simulate daily percentages of emerging weed seedlings for April through June, when most Corn Belt crops are sown. If seedbeds of fields with potential seedling densities of 100 seedlings/m2 are prepared and sown when simulated emergence exceeds °80% for C. album, 85% for A. retroflexus, and 99% for Setaria spp., then mechanical weed control sufficiently reduces weed populations and crop yield losses are negligible. Accordingly, mechanical weed control may be substituted economically for chemical control, but only when accompanied by adequate understanding of the ecology of pertinent weed species.