WHY SCALE MATTERS IN SPATIAL ANALYSIS: AN EXAMPLE USING NEOISOROPHUSELLA LANEI EDRIOASTEROID-ENCRUSTED SLABS FROM THE CHESTERIAN (UPPER MISSISSIPPIAN) KINCAID LIMESTONE OF SOUTHERN ILLINOIS

Spatial distribution patterns can tell us much about population dynamics in addition to a variety of habitat characteristics, such as suitability of the substrate for encrustation, resource availability, and territoriality. This study examines two limestone slab surfaces bearing 242 specimens of the Late Mississippian (Chesterian) edrioasteroid, Neoisorophusella lanei, collected as float material from a single obrution horizon in the Kincaid Formation of southern Illinois. Clusters of large individuals are found attached to internal molds of Promytilus bivalve shells with juveniles located within interstices or along the edge of the shell. The relationship of each edrioasteroid occurrence on the substrate was evaluated using two tests for space utilization: (1) Nearest Neighbor Analysis (NNA), a first-order statistic, compares the observed mean distance between nearest individuals to the expected mean distance between randomly distributed individuals; and, (2) Ripley’s K Analysis, a second-order statistic, assesses the complete distribution of all distances in the pattern. Spatial distribution was examined at both large scale (encrustation across the entire slab) and small scale (individual clusters of edrioasteroids). Large scale NNA and Ripley’s K distribution scores show clustered distributions, where distance between individuals is minimized; however, NNA and Ripley’s K statistics performed at the local scale show that the edrioasteroids are dispersed, where space between neighboring edrioasteroids is maximized. We attribute these results to increased fitness where edrioasteroids, who participated in broadcast spawning, increased fertilization success with clustered distributions of individuals, yet dispersion of individuals at the smaller scale promoted a sufficient degree of genetic variability, important for continued success of the population. Although distributional patterns by themselves do not demonstrate the nature of the underlying processes, careful examination of the results can inform the nature of the interactions. Therefore, analyses at multiple spatial scales is important as a means to glean valuable information about underlying distribution patterns that help elucidate the factors and processes that determine these distributions.