Modelling crop:weed interactions in wheat with ALMANAC

Abstract

ALMANAC is a dynamic model for plant growth, water balance and soil nitrogen dynamics that can simulate on a daily basis two or more competing species. The simulation of competition for light is based on Beer's law, allowing a different extinction coefficient (k) for each species. Light is partitioned between species based on k-values, leaf area index and plant heights. Total biomass is simulated with radiation use efficiency and grain yield with a harvest index approach, sensitive to water stress. The model simulates competition for water and nutrients based on each species' current rooting zone and demand by each species. The effect of crop management on the competition issue can be simulated. The model was evaluated in Dijon (France) using 4 years of experimental data on wheat:oat mixtures, differing in oat (Avena sativa L.) densities, the period of oat emergence, the date of weed suppression by herbicides and the wheat (Triticum aestivum L.) genotype. Additional data on oilseed rape (Brassica napus L.) and vetch (Vicia sativa L.) competition in spring wheat were also used. The wheat grain yield was reasonably simulated with a root mean square error (RMSE) of 0.10-0.35 t ha-1. Corresponding values for oats were 0.10-0.55 t ha-1. The competitiveness of oats, oilseed rape and vetch was correctly simulated. The model appears as a reasonable tool for estimating damage thresholds in integrated weed control programmes.