Ecological research increasingly considers integrative relationships among phenomena at broad spatial and temporal domains. However, such large‐scale inferences are commonly confounded by changing properties in the processes that govern phenomena (termed nonstationarity), which can violate assumptions underlying standard analytical methods. Changing conditions are fundamental and pervasive features in ecology, but their influence on ecological inference and prediction increases with larger spatial and temporal domains for a host of factors. Fortunately, tools for identifying and accommodating potentially confounding spatial or temporal trends are available, and new methods are being rapidly developed. Here, we provide guidance for gaining a better understanding of nonstationarity, its causes, and how it can be addressed. Acknowledging and addressing non‐constant trends in ecological patterns and processes is key to conducting large‐scale research and effectively translating findings to local policies and practices.

中文翻译：

---

跨时空工作：生态研究和应用中的不稳定

生态学研究越来越多地考虑在广泛的时空范围内现象之间的整合关系。但是，这种大范围的推论通常与控制现象的过程中的属性变化（称为非平稳性）相混淆，这可能会违反标准分析方法的假设。变化的条件是生态学的基本特征和普遍特征，但由于许多因素，它们对生态学推断和预测的影响会随着时空范围的增大而增加。幸运的是，已经有了用于识别和适应潜在混淆的空间或时间趋势的工具，并且新方法正在迅速开发。在这里，我们为更好地了解非平稳性，其原因以及如何解决提供了指导。