Intensive agriculture in Germany is not only highly productive but has also led to detrimental effects in the environment. Crop diversification together with new field arrangements considering soil heterogeneities can be an alternative to improve resource use efficiency (RUE), ecosystem services (ESS), and biodiversity. Agroecosystem models are tools that help us to understand and design diversified new field arrangements. The main goal of this study was to review the extent to which agroecosystem models have been used for crop diversification design at field and landscape scale by considering soil heterogeneities and to understand the model requirements for this purpose. We found several agroecosystem models available for simulating spatiotemporal crop diversification at the field scale. For spatial crop diversification, simplified modelling approaches consider crop interactions for light, water, and nutrients, but they offer restricted crop combinations. For temporal crop diversification, agroecosystem models include the major crops (e.g., cereals, legumes, and tuber crops). However, crop parameterization is limited for marginal crops and soil carbon and nitrogen (N). At the landscape scale, decision-making frameworks are commonly used to design diversified cropping systems. Within-field soil heterogeneities are rarely considered in field or landscape design studies. Combining static frameworks with dynamic agroecosystems models can be useful for the design and evaluation of trade-offs for ESS delivery and biodiversity. To enhance modeling capabilities to simulate diversified cropping systems in new field arrangements, it will be necessary to improve the representation of crop interactions, the inclusion of more crop species options, soil legacy effects, and biodiversity estimations. Newly diversified field arrangement design also requires higher data resolution, which can be generated via remote sensing and field sensors. We propose the implementation of a framework that combines static approaches and process-based models for new optimized field arrangement design and propose respective experiments for testing the combined framework.

德国的集约化农业不仅生产力高，而且对环境造成了不利影响。作物多样化以及考虑土壤异质性的新田间安排可以成为提高资源利用效率 (RUE)、生态系统服务 (ESS) 和生物多样性的替代方案。农业生态系统模型是帮助我们理解和设计多样化的新田间布置的工具。本研究的主要目的是通过考虑土壤异质性来审查农业生态系统模型在田间和景观尺度上用于作物多样化设计的程度，并了解为此目的的模型要求。我们发现了几种农业生态系统模型可用于模拟田间规模的时空作物多样化。对于空间作物多样化，简化的建模方法考虑了作物对光、水和养分的相互作用，但它们提供了有限的作物组合。对于时间作物多样化，农业生态系统模型包括主要作物（例如谷物、豆类和块茎作物）。然而，作物参数化仅限于边缘作物和土壤碳和氮 (N)。在景观尺度上，决策框架通常用于设计多样化的种植系统。在田间或景观设计研究中很少考虑田间土壤异质性。将静态框架与动态农业生态系统模型相结合，有助于设计和评估 ESS 交付和生物多样性的权衡取舍。为了增强建模能力，以模拟新田间布置中的多样化种植系统，有必要改进作物相互作用的表现，纳入更多作物品种选择、土壤遗留效应和生物多样性估计。新的多样化场布置设计也需要更高的数据分辨率，可以通过遥感和场传感器生成。我们建议实施一个框架，该框架结合了静态方法和基于过程的模型，用于新的优化场布置设计，并提出了测试组合框架的相应实验。