Data from: How to quantify a distance-dependent landscape effect on a biological response

To quantify the effect of the surrounding landscape context on a biological response at a site, most studies measure landscape variables within discs centred on this biological response (threshold-based method). This implicitly assumes that the effect of a unit area of the landscape is consistent up to a threshold distance beyond which it drops to zero. However, it seems more likely that the landscape effect declines with increasing distance from the biological response point. Here we develop a method to quantify landscape context effects by weighting the landscape variables by functions that decrease with distance. We illustrate the method using abundance data on birds and insects, and compare the results to the threshold approach. We defined distance weighting functions by the function family (e.g. negative exponential, Gaussian…) and by the parameters for this function. We developed a method to simultaneously estimate the parameters characterizing the effect of the landscape variables and the parameters of the best weighting functions. For each test dataset, we determined which weighting function (family and parameters) had the most support, by optimizing the model AIC. The distance-weighted method improved model support over the threshold-based method in three of four datasets, with the exponential power function selected as the best weighing function in all three cases. The observed differences between estimations of landscape context effects by the distance-weighted and the threshold methods have significant implications for landscape management. For example, the distance-weighted method suggests that managing a landscape for 90% of its effect on a focal population requires an area over five times larger than the area estimated by the threshold method, a situation that might apply for priority conservation of few remnant populations of a severely endangered species. In contrast, management for 30% of the landscape effect requires only about half the area estimated using the threshold method, a situation that might apply to a management situation with limited resources or low political/societal support. The distance-weighted method is applicable to any species-habitat relationship. More comparisons are needed to determine the situations in which distance-weighted estimation of landscape context effects is warranted over the simpler threshold method.为量化周边景观背景对某样点生物响应的影响，多数研究以该生物响应点为圆心，在圆形缓冲区（阈值法（threshold-based method））内测算景观变量。该方法默认预设：在阈值距离内，单位面积景观的影响保持恒定，超出该阈值距离后影响骤降为零。但更合理的推测是，景观影响随距生物响应点的距离增加而逐渐衰减。为此，本研究提出一种新方法：通过随距离递减的函数对景观变量赋予权重，以此量化景观背景效应。我们利用鸟类和昆虫的种群多度数据对该方法进行演示，并将结果与阈值法进行对比。 我们通过函数族（如负指数函数（negative exponential）、高斯函数（Gaussian）等）及其参数定义距离权重函数。本研究开发了一种可同时估算景观变量效应特征参数与最优权重函数参数的方法。针对每个测试数据集，我们通过优化模型AIC来筛选支持度最高的权重函数（含其函数族与参数）。 在四组测试数据集中，距离权重法相较阈值法提升了模型支持度，且这三组数据集均选择指数幂函数作为最优权重函数。 距离权重法与阈值法对景观背景效应的估算结果存在显著差异，这一差异对景观管理具有重要启示。例如，若要覆盖目标种群90%的景观效应，距离权重法所需的管理面积较阈值法估算结果大五倍以上，该场景可适用于极度濒危物种少量残存种群的优先保护工作。与之相反，若仅需覆盖30%的景观效应，距离权重法所需的管理面积仅为阈值法估算结果的一半左右，该场景可适用于资源有限或政治/社会支持度较低的管理情境。 距离权重法可推广应用于所有物种-生境关系研究。未来仍需开展更多对比研究，以明确相较于简易的阈值法，何时更适宜采用距离权重法估算景观背景效应。