SITE SPECIFIC MANAGEMENT OF AVENA STERILIS L. POPULATIONS PRESENT IN CEREAL CROPS IN CENTRAL SPAIN.

Field experiments were conducted to compare various detection methods for Avena sterilis patches, to determine the temporal and spatial patch dynamics of this species and to evaluate a 3-year site specific weed management (SSWM) program. Simulation models were used to study the economics of these programs.
Although the four detection methods tested (1. Scoring panicle density from the ground; 2. Scoring panicle density from a combine; 3. Counting panicle contacts; and 4. Counting seed rain on the ground) provided similar descriptions of the spatial distribution of A. sterilis, the cost of collecting the data differed widely (from 9 euro/ha to 750 euro/ha). The sampling cost was found to be critical in detemmining the herbicide application strategy and the optimum sampling method. Defining the location of weed patches from the combine was found to be a relatively inexpensive and reliable method for the creation of weed management maps to be used for patch spraying in the following season.
The spatial distribution of A. sterilis was monitored for five and for three years in two continuous bariey fields annually sprayed with half rate of imazamethabenz. was applied annually to the fields. Wild oat distribution was found persistent but not stable in both fields. Population decrease was faster when initial infestation was high than when it was low. In one of the experiments heavy weed patches (up to 1,200 seedlings/m²) were reduced down to relatively safe levels (18 seedlings/m²) in a 5-year period. However, under adverse environmental conditions, half rates of the herbicide failed to control adequately weed populations.
A 3-year SSWM program for wild oat was compared with two broadcast application strategies. Although the results vaned slightly within the years, the highest control levels were generally obtained with the broadcast application of graminicides at full dose. Using half doses reduced A. sterilis control. This effect was mainly attributed to the poor performance of low doses with high weed densities. The site specific approach resulted in the highest variability in the control, with some control failures in areas of low weed density.
The profitability of SSWM programs for wild oat was examined by a simulation model. Net returns of site specific A. sterilis management were found to be influenced by the interaction among the percentage of field infested, the pattern distribution and the weed sampling and/or herbicide spraying resolution. Net returns increased when both the proportion of field infested and the number of patches decreased. The highest net returns were obtained with the coarsest resolutions (12 m) due to the high current technology costs.