|Mueller warrant, George|
Submitted to: Meeting Abstract
Publication Type: Other
Publication Acceptance Date: 7/31/2003
Publication Date: 7/31/2003
Citation: MUELLER WARRANT, G.W., SCHWEITZER, L.R., COOK, R.L. 2003. MONITORING TRENDS IN WEEDS OF GRASS SEED CROPS USING GIS TOOLS. MAP #736. ESRI USERS CONFERENCE. JULY 25-30, 2003. Interpretive Summary: Oregon's grass seed industry produces the majority of cool season forage and turfgrass seed grown in the United States. Reasons for Oregon's success include both favorable climate and a well developed infrastructure. A key component of that infrastructure is the Oregon Seed Certification System of Oregon Seed Services, which acts in concert with growers, inspectors, seed conditioners, researchers and consultants to insure that standards are met for varietal purity and freedom from weed seed contamination. In excess of 58,000 pre-harvest field inspections have been conducted during the past 10 years, with data then entered into a standardized database for generating reports to individual growers and summaries for the entire industry. To enhance value of such summaries, a GIS approach has been taken, and a copy of the data has been migrated into ARC-GIS 8.3. Objectives included desires to: (1) visually display location of weeds across the landscape, (2) determine how those distributions change over time, (3) identify weeds with similar distribution patterns, and (4) identify associations between weed distribution patterns and factors such as crop management, field history, soil type, rainfall, and elevation.
Technical Abstract: Geographic Information System (GIS) software contains powerful tools for displaying and interpreting spatial data, but information must first be georeferenced. Oregon Seed Services routinely collects information on production practices of certified grass seed crops, including pre-harvest inspection reports on weed presence within fields. Access to data was granted under the stipulation that grower confidentiality be maintained in any public releases of maps or other summaries. A total of 71 grassy weeds and 136 broadleaves, sedges, and other types were found in 10 years of inspections of an average of 5,801 fields per year. The primary obstacle to importing data into ArcGIS was that fields were only localized to township/range/section (TRS) position, with an average of 4.2 and a maximum of 23 unique production fields per TRS. Somewhat arbitrary latitude/longitude values were assigned to each field using a procedure maximizing distances between fields within each TRS. Raster maps were generated with Inverse Distance Weighted (IDW) methods using weed abundance values of 0, 1, 10 or 100 for ratings of absent, trace, many, or excessive numbers of each species. Subtracting rasters for one year from the next identified regions in which weeds such as Poa trivialis were changing in prevalence over time.