Italian ryegrass growth and production from positionally selective applications and sublethal rates of a nonselective herbicide were compared to growth and production when treated with diuron applied preemergence as a broadcast treatment and safened by activated carbon applied over the crop row. In the positionally selective treatments, herbicide spray was directed to leave an untreated zone over the crop row at planting. Growth analysis indicated no detectable differences when Italian ryegrass safened by directed spray was compared to Italian ryegrass safened by activated carbon. The major difference between systems was greater weed control in the crop row for carbon-safened treatments. The non-safened application superimposed sublethal diuron rates over Italian ryegrass and rattail fescue grown in varying densities and proportiors. Growth analysis of monoculture stands indicated differences due to planted density, species, and herbicide, whereas growth analysis of plants grown as space-planted individuals indicated difference due to species only. Diuron at the rates applied did not affect seed yield or above ground dry weight. Soil samples were taken in crop rows where diuron was applied as directed spray or broadcast spray safened by carbon. Samples were assayed and soil profile concentrations mapped. To assist future investigators' understanding of the role rainfall plays in herbicide movement from directed applications, elementary rainfall depth and occurrence models were examined. The Markov and mixed-exponential models adequately described rainfall occurrence and depth patterns for Corvallis, Oregon. / Graduation date: 1992