A simple model relating yield loss to weed density

TLDR

The analysis is contextualized by comparison to the popular sigmoidal model for yield loss. The study develops a simple model to describe crop yield loss as a function of weed density. The authors fit a rectangular hyperbola with two agronomically meaningful parameters to literature data and compared it to 17 other models across 22 datasets. Two‑parameter yield‑loss models were generally sufficient, with the hyperbolic model providing the best fit, though only marginally better than three other models.

Abstract

SUMMARY A simple model is developed to describe crop yield loss as a function of weed density. The model, a rectangular hyperbola, has two agronomically meaningful parameters which together can be used as indices of competitiveness. Examples are given where the model is fitted to experiments from the literature. Seventeen other models are compared through application to 22 data sets. It was found that for the data available two‐parameter yield loss models were generally sufficient. The hyperbolic model gave the best description of data, but this was only marginally better than three other models. This conclusion is discussed with reference to the popular sigmoidal model.