Landscape complexity promotes ecosystem services and agricultural productivity, and often encompasses aspects of compositional or configurational land cover diversity across space. However, a key agricultural diversification practice, crop rotation, extends crop land cover complexity concurrently across space and time. Long-term experiments suggest that complex crop rotations can facilitate yield increases in major crops. Using a compiled county-annual panel dataset, we examined whether yield benefits of crop rotational complexity were apparent on a landscape scale in the conterminous US for four major crops between 2008 and 2020. We found that the benefit of rotational complexity was only apparent for cotton and winter wheat, and that the benefit for wheat was driven by one region. Corn exhibited the opposite pattern, wherein higher yields were consistently obtained with lower rotational complexity, while soybean yield appeared relatively insensitive to rotational complexity. Effects of rotational complexity were sometimes influenced by agrochemical usage. Positive effects of rotational complexity were only apparent with high fertilizer for soybean and wheat, and with low fertilizer for cotton. Corn yield in high-complexity, low-yielding counties appeared to improve with high fertilizer inputs. For the overwhelming majority of acres growing these major crops, crop rotation patterns were quite simple, which when combined with the short time span of available data, may explain the apparent discrepancy between long-term experiments and nationwide data. Current demand and incentives that promote highly intensified and specialized agriculture likely hinder realization of the benefits of rotational complexity for production of key crops in the US. Increasing rotational complexity where major crops are grown thus remains an underutilized approach to mitigate landscape simplification and to promote ecosystem services and crop yields.

中文翻译：

---

目前，美国各县轮作的复杂性不足以观察主要作物的产量增长

景观复杂性可促进生态系统服务和农业生产力，并且通常包含跨空间的组成或配置土地覆盖多样性的各个方面。然而，农业多样化的一个关键实践——轮作，同时在空间和时间上扩展了农作物土地覆盖的复杂性。长期实验表明，复杂的轮作可以促进主要作物的产量增加。使用编制的县年度面板数据集，我们研究了 2008 年至 2020 年间，作物轮作复杂性带来的产量效益是否在美国本土四种主要作物的景观尺度上显而易见。我们发现，轮作复杂性的效益仅对棉花而言很明显和冬小麦，小麦的效益是由一个地区驱动的。玉米表现出相反的模式，其中较高的产量始终以较低的旋转复杂性获得，而大豆产量似乎对旋转复杂性相对不敏感。旋转复杂性的影响有时受到农用化学品使用的影响。轮作复杂性的积极影响仅在大豆和小麦高施肥和棉花低施肥时才明显。高复杂性、低产县的玉米产量似乎因高化肥投入而有所提高。对于种植这些主要作物的绝大多数土地来说，作物轮作模式非常简单，与可用数据的短时间跨度相结合，可以解释长期实验与全国数据之间的明显差异。当前促进高度集约化和专业化农业的需求和激励措施可能会阻碍美国关键作物生产轮作复杂性带来的好处的实现。因此，增加主要作物种植区的轮作复杂性仍然是一种未得到充分利用的方法，以减轻景观简化并促进生态系统服务和作物产量。