Abstract Reducing pesticide use in agriculture is essential but involves shifting towards more complex agroecosystems. Plant canopies are expected to be more heterogeneous because a more abundant weed flora is likely to remain in low-herbicide fields, and because intercropping (i.e. mixing various crops species or varieties) is a promising option to reduce chemical inputs. Simulation models are useful to understand and design complex agroecological cropping systems, but they rarely represent the root systems of plants. However, belowground processes, such as competition for soil resources or infection by root-parasitic plants, are key determinants of the structure of plant communities. The aim of our study was to develop a model that simulates heterogeneous 3D individual-based crop-weed canopies from cropping system and pedoclimate and that will ultimately be used to design agroecological cropping systems. Therefore, we (1) connected a root system model (RSCone) to a weed dynamics model ( FlorSys ) in order to include both above and belowground parts of plants, (2) evaluated the prediction quality of our model, and (3) analysed the influence of species parameters on potential soil-resource uptake and root infection by parasitic plants. We used the well-known allometric relationship between root and total plant biomass to connect RSCone and FlorSys , and we created new formalisms to model the effect of soil compaction on root growth. Our model was shown to correctly predict long-term weed dynamics in various cropping systems. From step 3), we characterized crops and weed communities that are potentially competitive for soil resources and most likely to be infected by parasitic plants, and we deduced agronomic recommendations. For example, species emerging and occupying the soil quickly were the most likely to relay broomrape infestation and control of such species should take precedence. Although we focussed on crop-weed competition, our approach can be applied to other heterogeneous canopies, for designing crop mixtures for example.

中文翻译：

---

在多年尺度上对异质植物冠层中的根系进行基于个体的 3D 建模。杂草动力学模型的案例研究

摘要 减少农业中农药的使用至关重要，但需要转向更复杂的农业生态系统。预计植物冠层将更加异质，因为更丰富的杂草植物群可能保留在低除草剂的田地中，并且因为间作（即混合各种作物物种或品种）是减少化学投入的有希望的选择。模拟模型有助于理解和设计复杂的农业生态种植系统，但它们很少代表植物的根系。然而，地下过程，例如对土壤资源的竞争或根寄生植物的感染，是植物群落结构的关键决定因素。我们研究的目的是开发一个模型，该模型模拟来自种植系统和土壤气候的基于个体的异质 3D 作物杂草冠层，并最终用于设计农业生态种植系统。因此，我们 (1) 将根系模型 (RSCone) 连接到杂草动力学模型 (FlorSys) 以包括植物的地上部分和地下部分，(2) 评估我们模型的预测质量，以及 (3) 分析物种参数对潜在土壤资源吸收和寄生植物根部感染的影响。我们使用根和植物总生物量之间众所周知的异速生长关系来连接 RSCone 和 FlorSys，我们创建了新的形式来模拟土壤压实对根生长的影响。我们的模型被证明可以正确预测各种种植系统中的长期杂草动态。从步骤 3) 中，我们表征了对土壤资源具有潜在竞争力且最有可能被寄生植物感染的作物和杂草群落，并推导出了农艺建议。例如，迅速出现并占据土壤的物种最有可能传播扫帚菜侵扰，因此应优先控制此类物种。尽管我们专注于作物与杂草的竞争，但我们的方法可以应用于其他异质树冠，例如设计作物混合物。迅速出现并占据土壤的物种最有可能传播扫帚菜侵扰，因此应优先控制此类物种。尽管我们专注于作物与杂草的竞争，但我们的方法可以应用于其他异质树冠，例如设计作物混合物。迅速出现并占据土壤的物种最有可能传播扫帚菜侵扰，因此应优先控制此类物种。尽管我们专注于作物与杂草的竞争，但我们的方法可以应用于其他异质树冠，例如设计作物混合物。