ABSTRACT
The motivation to work with spatial data is partly driven by the need to gain a deep understanding of the spatial structure of a range of phenomena such as crime incidents, injuries, diseases, retail, or bird nesting sites that are represented by point features. Such features are amenable to point pattern analysis in which emphasis is placed on the complete set of observations as well as the location of each observation and its distance relative to others in the distribution. Although the analysis of the point distributions does provide us with fundamental clues about the underlying spatial processes and relationships, the main focus is on the examination of any static evidence of spacing. This evidence is normally depicted as a random or nonrandom pattern. If the point pattern is identified as nonrandom, it can be further described as more clustered than random or more dispersed than random. Therefore, three basic pattern structures exist: random, clustered, or dispersed. These patterns are illustrated in Figure 6.1. In the upper panel, the spatial pattern is clustered and has a large variance. The middle panel is a randomly dispersed pattern, has a moderate variance, and is similar to a Poisson distribution. The lower panel is a dispersed/uniform pattern with no or little variance. The data depicted in this figure are based on the simulation of nesting sites of the African black coucals (Centropus grillii) in the Ssezibwa wetlands, north of the town of Kayunga, Uganda. A polygon layer of distribution and habitats for the African black coucals from the IUCN 2012 Red List of Threatened Species database was used to identify the potential nesting sites. Ancillary information compiled from the 2012 aerial and satellite images was used in identifying land cover with open, dense, marshy, or
swampy grassland. The area was delineated into a rectangular shape of 1507 by 1370 m so that all the nest sites were within the boundaries.
