Spatial Analysis in Ecology

Abstract The first step in understanding ecological processes is to identify their spatial patterns. Ecological data are usually characterized by spatial structures and as such, they are said to be spatially autocorrelated. Spatial autocorrelation refers to the pattern where the values of a quantitative variable are more similar at nearby locations than expected by chance alone. Most ecological data exhibit some degree of spatial autocorrelation that is modulated by the spatial sampling design used to record the data and the method used to analyze them. Furthermore, ecological data can be the end result of several processes operating at different spatial scales. In such cases, ecological data are a composite of large‐scale trends at the macroscale, gradients and patchiness at mesoscale, and random patterns at local and microscales. Hence, to estimate the magnitude and the extent of spatial autocorrelation, various spatial statistics can be used. Here, we review the spatial statistics most commonly used by ecologists.