* Many ecological questions could gain insight from linking variation in spatial structure to explanatory variables. Although methods that make these links have been developed in the statistical literature, they have been used rarely in ecology. * We present an approach that considers point patterns from different sources (e.g. multiple plots or species) as replicates of the spatial processes of interest. Using an integrated two-step procedure, the method uses a linear mixed-effects modelling framework to explain variation in the spatial structure of the replicated patterns with categorical and continuous predictors. * Inferences on the strength of the relationships between the predictors and spatial structure are made on the basis of a semi-parametric bootstrap procedure that accounts for the dependency among spatial scales and, if required, among replicates. * We illustrate the approach with three simulated examples. These demonstrate that replicated analyses can increase the performance of spatial point pattern analyses, allow analysis of sparse patterns (e.g. rare species) and help identify the drivers of variation among spatial point patterns. * The paucity of ecological studies that utilize replicated point pattern analysis may be due partly to a lack of familiarity with the methods. By highlighting the options available, we hope to encourage greater use of these methods by ecologists in the future.
- spatial modelling, second-order spatial structure, covariate, K-function, mixed-effects models, bootstrap, rare species, Thomas process, Strauss process