Project Number: 3611-12220-008-02
Start Date: Sep 22, 2011
End Date: Aug 31, 2012
The first objective will be achieved by collecting accessions (seeds from individual female plants) from multiple locations and screening these accessions for resistance in the greenhouse. Molecular assays for resistance currently being developed will be used for follow-up studies. We propose to sample two to four fields in each of 60 to 90 counties throughout Illinois, specifically targeting any fields where glyphosate resistance is suspected. From this component of the research we will be able to map where glyphosate-resistant waterhemp is present and most prevalent. To achieve the second objective, we will focus on a 16 x 16 km area (100 sections) in south-central Illinois where the presence of glyphosate-resistance waterhemp is confirmed. We will work with a local agrichemical retail dealer/applicator to obtain current and historical data on agronomic management (including planting information, tillage practices, and chemical inputs) as well as current estimates of levels of waterhemp infestation and glyphosate-resistant status in fields comprising approx. 15,000 ha (37,000 acres). A subset of these fields will be selected that comprise a range of management practices, landscape positions, and waterhemp infestation levels (including presence/absence of glyphosate resistance). Within this subset of fields, data will be collected on: a) waterhemp plant and seed population density within the field and in field margins, b) prevalence of glyphosate-resistant waterhemp plants and seeds in field and field margins, and c) geospatial coordinates. Greenhouse bioassays will be used to determine whether sampled plants are glyphosate resistant. Soil textural data for the fields will be extracted from a NRCS geographic information system (GIS) layer. Field landscape position, relative to features of interest (e.g. barriers and dispersal corridors within the landscape matrix such as forests, riparian areas, railroads, and roads), will also be extracted from an NRCS GIS layer. Using fields as the experimental unit, presence/absence of mature, glyphosate-resistant waterhemp as he response variable, and the environmental and management variables described above as predictor variables, multiple logistic regression models will be fit using restricted maximum likelihood methods. Candidate models will be considered that include management variables, environmental variables, and a combination thereof. The most parsimonious models (i.e., those models that offer the best fit with fewest parameters) will be used to identify factors associated with low risk of glyphosate-resistance waterhemp establishment in production fields.