Weeds and the soil environment - Ecological Management of Agricultural Weeds - page 223

It is clear that intercom, allocate, and other models can be useful
tools for examining the outcome of interactions among a range of weed-crop
combinations grown in different soil environments, and for predicting
morphological or physiological traits that can contribute to a crop's ability to
tolerate or suppress weeds. However, as the following discussion indicates,
consideration of several additional types of information would add to the
value of models in developing weed management strategies. Of particular
importance are the potential effects of spatial and temporal variation in nutri-
ent availability, and qualitative differences between different nutrient
sources.
Spatial variation in nutrient availability
Placement of fertilizer in bands close to crop rows is a method of con-
centrating nutrients for use by crops; it may also reduce nutrient availability
to weeds not growing near the bands. Banding fertilizers within the row of
such crops as bean, soybean, peanut, wheat, alfalfa, and rice has been shown
not only to increase crop yield compared to broadcast applications, but also to
reduce weed density and biomass (DiTomaso, 1995). Advantages of fertilizer
banding are generally more pronounced for nutrients applied in a deep band
(e.g., 5-7 cm below seed level) than for nutrients applied in a band on the soil
surface (DiTomaso, 1995), perhaps because crop seedlings tend to emerge
from deeper in the soil profile than do weed seedlings.
The success of fertilizer banding as a weed management strategy may
depend on background levels of soil fertility. For example, in field experi-
ments conducted for three years on two soil types in Denmark, Rasmussen,
Rasmussen & Petersen (1996) observed that banding N fertilizer 5 cm below
spring barley rows decreased weed biomass by an average of 55% and
increased barley grain yield by an average of 28%, compared with broadcast
fertilizer application. Comparison of results from the two soil types indicated
that reductions in weed growth and increases in crop yield due to fertilizer
banding were greater on a low-fertility coarse sand than on a more fertile
sandy loam. In field experiments conducted on a silt loam in eastern
Washington (USA), banding N fertilizer 5 cm below winter wheat rows gener-
ally increased wheat grain yield, but had little or no effect on biomass produc-
tion by the dominant grass weeds ( Bromus tectorum and Aegilops cylindrica )
infesting the crop (Cochran, Morrow & Schirman, 1990). Cochran, Morrow &
Schirman (1990) noted that background fertility levels at the site may have
been sufficient to prevent major N deficits for weeds and that weeds growing
close to fertilizer bands may have had access to the applied N.The researchers
suggested that greater weed suppression would be expected when a lack of