Economic and management value of weed maps at harvest in semi-arid cropping systems of the US Pacific Northwest

Weed maps created late in the growing season are potentially useful in regions where late maturing weeds are problematic and need to be controlled before they produce seed. The objectives of this study were to (1) spatially characterize the population dynamics of predominant weed species and apply this information into quantifying the effect of treated and untreated weed infestations on wheat (Triticum aestivum L.) yields and (2) evaluate potential herbicide savings with post-harvest site-specific treatments. Multi-year grain yield and weed data were acquired at harvest in each of four years (2015-18) within a dryland production field (9.2 ha) in eastern Oregon. Abundance of weed species (2015) and percent cover of weed species (2016, 2017 and 2018) were visually estimated on a square grid based on dividing the field into 7-m(2) cells. Spatial patterns in the weed community were subject to rapid change and depended on year, crop and weed control strategy. While tumble mustard (Sisymbrium altissimum) was the most predominant and competitive species, the spatial distribution of this weed and that of other species varied each year. Tumble mustard and prickly lettuce (Lactuca serriola) were equally problematic in spring wheat and winter wheat whereas Russian thistle (Salsola tragus) was problematic in spring wheat and downy brome (Bromus tectorum) in winter wheat. Potential savings from site-specific herbicide application varied from 10 to 95% based on percentage of field infested. Weed maps at harvest are useful for studying weed dynamics, identifying potentially herbicide-resistant weeds and planning site-specific weed management. Combined with yield maps, weed maps at harvest are also useful for explaining crop yield variability that is associated with weed competition and weed control in furtherance of integrated weed management strategies.