Detecting spatial hot spots in landscape ecology

摘要: Hot spots are typically locations of abundant phenomena. In ecology, hot often detected with a spatially global threshold, where value for given observation is compared all values in data set. When spatial relationships important, local definitions - those that compare the vicinity, or neighbourhood provide more explicit consideration space. Here we outline methods spot detection: kernel estimation and measures autocorrelation. To demonstrate these approaches, landscape level on magnitude mountain pine beetle infestations. Using estimators, explore how selection size (τ) threshold impact detection. We found as T increases, larger fewer; become plentiful will reflect coarser scale processes. The delineation identified autocorrelation was also investigated. general, used analysis, area, greater number areas, spots.