Spatial Point Pattern Analysis of Plants

Abstract

Plants, especially terrestrial long-lived perennials such as trees, do not usually move once established. Spatial patterns of sessile organisms can suggest or reveal ecological processes affecting the population or community in the present or the past – dispersal, establishment, competition, mortality, facilitation, growth – and as such, patterns of plants motivated early developments in spatial statistics (Pielou, 1977; Diggle, 1983). Specifically, it is intuitive to treat individual plants (or other sessile organisms) as discrete events on a plane whose locations are known and generated by point pattern processes (Ripley, 1981; Diggle, 1983; Fortin and Dale, 2005). Second-order point pattern statistics are used to measure their spatial pattern. Arthur Getis (Getis and Franklin, 1987) introduced ecologists to the application of local spatial statistics, specifically neighborhood second-order point pattern analysis, to maps of organisms. As Wiegand and Moloney (2004) noted in their review paper, second-order global statistics based on the distribution of distances between pairs of points, especially Ripley’s K-function (Ripley, 1976, 1977) derived from distances between all pairs, have been widely used in plant ecology. However, their review does not mention neighborhood analysis or local measures of spatial association (Anselin, 1995) at all. This chapter revisits the impact of the Getis and Franklin paper on the practice of spatial point pattern analysis in plant ecology, and specifically aims to determine if local statistics are being used and how.