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.dataset 1dataset 1, Savannah Sparrow (biological variable) x Grassland (landscape variable)data1.txtdataset 2dataset 2, Bobolink (biological variable) x Grassland (landscape variable)data2.txtdataset 3dataset 3, Bobolink (biological variable) x Forest (landscape variable)data3.txtdataset 4dataset 4, Pterostichus Melanarius (biological variable) x Corn (landscape variable)data4.txtR scriptR script of the functions used in the methods of the paperMiguetEtAl_Script.RReadMe fileThe ReadMe file provides details about the content of the other files