Page Banner

United States Department of Agriculture

Agricultural Research Service

Title: GIS Analysis of Spatial Clustering and Temporal Change in Weeds of Grass Seed Crops)

item Mueller warrant, George
item Whittaker, Gerald - Jerry
item Young, William

Submitted to: Weed Science
Publication Type: Peer reviewed journal
Publication Acceptance Date: 4/21/2008
Publication Date: 10/31/2008
Citation: Mueller Warrant, G.W., Whittaker, G.W., Young, W.C. 2008. GIS Analysis of Spatial Clustering and Temporal Change in Weeds of Grass Seed Crops. Weed Science. 56:647-669.

Interpretive Summary: Statistical analysis of spatially-referenced data can provide us with far better understanding/appreciation of complex phenomena than more traditional, nonspatial approaches normally achieve. Numerous technical hurdles were overcome in the process of transforming an extremely large, nonspatial database of 10 years of OSU Seed Certification field inspection reports into visual displays of weed severity without compromising the confidentiality of individual growers. Weed species varied greatly in their tendency to appear in clusters as opposed to random patterns, with German velvetgrass, field bindweed, roughstalk bluegrass, and annual bluegrass clustering most strongly. Weeds tended to appear in groups based on crop species for the major crops grown in the area, and were also linked to soil type and soil chemical and physical properties. Distances over which individual weed species clustered varied from a 5-kilometer scale for bentgrass to a 59-kilometer scale for mayweed chamomile. Crop management implications include the possibility/desirability of tailoring weed control practices by geographic location based on likeliness of individual weed species occurring and a better understanding of reasons behind recent increases in weeds such as wild carrot. The public release of weed hot spot maps should help growers, consultants, and other interested parties deal with these weeds.

Technical Abstract: Ten years of Oregon Seed Certification Service pre-harvest field inspection reports previously converted from a non-spatial database to a GIS were used to analyze spatial patterns in distribution of severity of the 36 most commonly occurring weeds. Moran's I spatial autocorrelation of maximum weed severity observed over the 10-year period using simple inverse distance weighting was extremely sensitive to changes in field boundaries over time because these caused distances between a small number of field centroids to be much smaller than average distances from centroids to edges. Addition of arbitrary offsets in the range from 6 to 6437 m to distances between centroids in inverse distance weighting matrices removed this sensitivity to spatially overlapping fields, allowing fixed distance and inverse distance weighting methods to produce similar classifications of weeds possessing strongest and weakest spatial autocorrelation. Clustering was significant for 10-year maximum severity for all 43 weeds and in 78% of single-year analyses. For the seven weeds also grown as crops, clustering was stronger when fields growing those crops were included in the analyses than when they were excluded, with exception of Kentucky bluegrass and roughstalk bluegrass, species grown in only a few western Oregon fields. In decreasing order, weeds with strongest spatial autocorrelation were German velvetgrass, field bindweed, roughstalk bluegrass, annual bluegrass, orchardgrass, common velvetgrass, Italian ryegrass, Agrostis spp., perennial ryegrass, and quackgrass. Of these 10 weeds, distance for peak spatial autocorrelation ranged from 2 km for Agrostis spp. to 34 km for common velvetgrass.

Last Modified: 8/24/2016
Footer Content Back to Top of Page