Skip to main content
ARS Home » Research » Publications at this Location » Publication #79715

Title: MULTIVARIATE APPROACHES FOR LINKING FIELD-SCALE VARIABILITY OF SOIL PROPERTIES

Author
item DIELEMAN, J - UNIVERSITY OF NEBRASKA
item MORTENSEN, DAVID - UNIVERSITY OF NEBRASKA
item Buhler, Douglas - Doug

Submitted to: Weed Science Society of America Meeting Abstracts
Publication Type: Abstract Only
Publication Acceptance Date: 2/6/1997
Publication Date: N/A
Citation: N/A

Interpretive Summary:

Technical Abstract: Is the patchy nature of weed populations related to underlying soil variability? The objective of this study was to assess the relationship between soil properties and observed weed populations using a non-spatial statistical technique, canonical correlation analysis. Data were collected for two years, 1994-95 at Shelton, NE, and 1995-96 at Alda, NE, on a systematic grid. Both sites are research and demonstration plots on working farms. Measured soil properties included N, P, K, pH, organic matter, and Zn. Other features included elevation, depth to sand (Shelton) or texture (Alda). Weed seedling counts were within a 0.38 square m quadrant placed between the crop rows. Six to eight species occurred at sufficiently high densities to be included in the analysis. Canonical correlation analysis derives a linear combination for each set of variables (canonical variates, SOIL or WEED) and then finds the largest possible correlation (canonical correlation coefficient) between the two sets. Coefficients of the first canonical correlation were 0.23 and 0.55 at Shelton '94 and Alda '95, respectively. Elevation and organic matter were the largest contributors to SOIL at both sites and were negatively related. WEED was dominated by annual grasses and positively correlated to elevation. Broadleaf species were found in areas of high organic matter. In these fields, higher elevations describe regions with high sand content and low organic matter. Distributions of soil properties and weed populations appear to be linked. However, our further hypothesis is that weed management applications interact with soil properties and reduce persistence or efficacy of herbicidal products, resulting in patchy weed populations.