Assessing innovative sowing patterns for integrated weed management with a 3D crop:weed competition model

Abstract Weeddynamics models are needed to design innovative weedmanagement strategies. Here, we developed a 3D individual-based model called FlorSys predicting growth and development of annual weedsand cropsas a function of daily weather and croppingpractices: (1) cropemergence is driven by temperature, and emerged plantsare placed onto the 3D field map, depending on sowingpattern, density, and emergence rate; plants are described as cylinders with their leaf area distributed according to height; (2) weedemergence is predicted by an existing submodel, emerged weedseedlings are placed randomly; (3) plant phenology depends on temperature; (4) a previously developed submodel predicts available lightin each voxel of the canopy; after emergence, plantgrowth is driven by temperature; when shaded, biomass accumulation results from the difference between photosynthesis and respiration; shading causes etiolation; (5) frost reduces biomass and destroys plants, (6) at plant maturity, the newly produced seeds are added to the soil seed bank. The model was used to test different sowingscenarios in an oilseed rape/winter wheat/winter barley rotation with sixteen weedannuals, showing that (1) cropyield loss was negatively correlated to weedbiomass averaged over the croppingseason; (2) weedbiomass was decreased by scenarios allowing early and homogenous cropcanopy closure (e.g. reduced interrows, increased sowingdensity, associated or undersown crops), increased summer fatal weedseed germination (e.g. delayed sowing) or, to a lesser degree, cleaner fields at cash crop sowing(e.g. sowinga temporary cover cropfor “catching” nitrogen); (3) the scenario effect depended on weedspecies (e.g. climbing species were little affected by increased cropcompetition), and the result thus varied with the initial weedcommunity (e.g. communities dominated by small weedspecies were hindered by the faster emergence of broadcast-sown cropswhereas taller species profited by the more frequent gap canopies); (4) the effect on weedbiomass of sowingscenarios applied to one year was still visible up to ten years later, and the beneficial effect during the test year could be followed by detrimental effects later (e.g. the changed tillage dates accompanying catch cropsreduced weedemergence in the immediately following cash cropbut increased seed survival and thus infestation of the subsequent crops). This simulation showed FlorSys to predict realistic potential cropyields, and the simulated impact of cropscenarios was consistent with literature reports.