Designing a sampling scheme to reveal correlations between weeds and soil properties at multiple spatial scales

Area of Science:

• Agricultural Science
• Ecology
• Soil Science

Background:

• Weed distribution in agricultural fields is often patchy.
• Soil property variations are a key driver of weed patchiness.
• The influence of soil properties on weed density is scale-dependent.

Purpose of the Study:

• To develop and validate a general method for exploring scale-dependent relationships between soil properties and weed density.
• To quantify the impact of spatial scale on weed patch dynamics.
• To optimize sampling strategies for mapping weed distribution and guiding targeted interventions.

Main Methods:

• Utilized novel within-field nested sampling and residual maximum-likelihood (REML) estimation.
• Partitioned variance and covariance into scale-specific components.
• Employed variograms for spatial structure quantification and kriging for mapping.

Main Results:

• Successfully captured scale-dependent effects of edaphic drivers on weed patchiness.
• Identified stronger correlations between Alopecurus myosuroides and soil organic matter/clay content at scales >50 m, masked by overall weak correlations.
• Optimized sampling design by focusing effort on scales contributing most to total variance.

Conclusions:

• The developed methodology effectively reveals scale-dependent soil-weed relationships.
• Optimized sampling strategies can improve the efficiency of weed mapping and management.
• Findings support targeted weed spraying by identifying vulnerable field areas.