The Effects of Crop Rotation and Topography On the Weed Seed Bank In the Palouse Wheat Region of Washington State.

Managing the weed seed bank is key to reducing weed seed accumulation and overall weed populations. Understanding how crop rotation, soil characteristics, and terrain influence the weed seed bank may help identify field-related factors that contribute to greater weed pressure. A multi-year cropping systems study in no-tillage, initiated in 1999, was conducted on a 37 ha field of the Cook Agronomy Farm near Pullman, WA. One objective for this study was to better understand how conservation tillage and crop rotation practices affect weed species composition and distribution across the landscape. The cropping rotations included six different three year rotations of spring wheat – winter wheat – alternative crop (spring or winter plantings of barley, canola, lentil, or pea). Each rotation was represented by a farm-scale plot. Soil cores were taken in 1999 and 2007 from 369 geo-referenced locations across the farm to analyze the weed seed bank. Samples were exhaustively germinated and germination was recorded weekly by species over the course of the study. Every four weeks samples were re-randomized on the greenhouse benches. Topographic variables and cropping systems were assessed as predictors of viable weed seed levels within the seed bank. Topographic variables were calculated from a 2 m-resolution Digital Elevation Model (DEM). The data were analyzed using Poisson generalized linear model (GLM) and zero-inflated Poisson regression model. The weed species that were analyzed were mayweed chamomile (Anthemis cotula L.), wild oat (Avena fatua L.), common lambsquarters (Chenopodium album L.), prickly lettuce (Lactuca serriola L.), and Italian ryegrass [Lolium perenne L. ssp. multiflorum (Lam.) Husnot]. In 1999, all landscape characteristics that were analyzed using the Poisson GLM were significant for all weed species. Field and strip were also significant for all weed species with the exception of prickly lettuce. In 2007, all landscape characteristics were significant for Italian ryegrass, prickly lettuce, and mayweed chamomile.  All landscape characteristics except global irradiation were significant for wild oat and all landscape characteristics except the wetness index were significant for common lambsquarters. Field was significant for mayweed chamomile and wild oat. Only Field C was significant for prickly lettuce and Field B for Italian ryegrass. Crop rotations were also significant for each of the weed species. In 1999, wetness index and elevation were significant for mayweed chamomile, slope and elevation were significant for wild oat and common lambsquarters, and slope and transformed aspect were significant for Italian ryegrass when analyzed using the zero-inflated Poisson regression model. In 2007, elevation was significant for mayweed chamomile, wild oat, and Italian ryegrass. Transformed aspect was also significant for Italian ryegrass. Topography and cropping systems do effect the viable weed species distribution within the weed seed bank.