Date of Submission


Document Type


Degree Name

Master of Science in Environmental Science


Biology and Environmental Sciences


Roman Zajac, Ph.D.


Seafloor habitats, Spatial statistics tools, ArcGIS, Benthic infaunal community, Taxonomic, Spatial patterns


Geographic Information Systems


Marine sciences--Geographic information systems, Benthic ecology, Long Island Sound (N.Y. and Conn.), Spatial ecology, Environmental impact analysis


Geographic Information Systems (GIS) are now regularly used in a variety of sectors including public health, city planning, and emergency management planning. Environmental scientists and managers also utilize GIS as a framework for gathering, managing, and analyzing terrestrial ecological data. More recently, GIS has been used for characterizing seafloor habitats and the communities within those habitats. Less known and used among marine ecologists is the Spatial Statistics toolbox within ESRI ArcGIS software. These tools can be used to determine spatial patterns and their statistical significance for several types of data associated within benthic communities. The goal of this study was threefold. Firstly, to evaluate the effectiveness of a variety of spatial statistics tools including Getis-Ord Gi Hot spot, Cluster-Outlier analysis (Anselin Local Morans I), Spatial Autocorrelation (Global Moran's I), Incremental Spatial Autocorrelation and Grouping Analysis, for understanding benthic communities in the eastern area of Long Island Sound (LIS); secondly, to identify which analyses provided the most useful information on the spatial patterns and scales of benthic infaunal community characteristics; and to use the analyses to inform what factors could be driving the spatial patterns identified. The results show that, in general, as the distance grouping classes among sample sites increased, the number of statistically significant hot spots and cluster-outliers also increased. For taxonomic richness and total abundance, analyses indicated that environmental and ecological factors that vary on a spatial scale of 1,500 and 2,500 m may be driving infaunal communities in the eastern portion of Long Island Sound, from the Connecticut River through Fishers Island Sound. Analyses of diversity data suggested that spatial statistical routines (in ArcGIS and otherwise) might not be sensitive enough to pick up spatially significant patterns due to the v narrow range of values in these data. Results from the grouping (clustering) analysis for benthic communities indicated that community types were spatial variable but that large- scale trends were identified, with generally differing community types in the western, central and most eastern portions of the study area. However, the relative similarity of the communities depended on the number of groups designated in the analysis parameters. The spatial statistics toolbox in ArcGIS provides a powerful set of analyses that can help researchers gain additional insights into seafloor habitats and their associated benthic communities that compliment more regularly employed analysis procedures, and it turn inform management and conservation efforts.

Off campus users: To download campus access theses or dissertations, log in to proxy server.