Date of Award

1989

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

First Advisor

Robert Costanza

Abstract

The goals of this research were: to develop a methodology for the study of landscape process and pattern; to determine the spatial pattern of land loss in coastal Louisiana and relate it to causal processes; and to assess the relative importance of natural versus human-induced processes in causing land loss. Land loss was analyzed for a 22 year period at three sites, using a Geographic Information System composed of digital habitat maps. The most significant finding of this research was that land loss is not a simple, spatially homogeneous phenomenon, that occurs uniformly. Instead, large clumps of high loss, or "hot spots," were observed. Hot spots comprised 9-18% of the original study area, but contained 50-73% of all loss. There is evidence that hot spots occur coastwide, and are not limited to these three study sites. The cause of the hot spots could not be determined from the current study. Simple models containing factors such as site, habitat type, salinity change, and proximity to canal and spoil or natural channel could not account for all the land loss in these areas, although higher order interactions were not included. Mechanisms with a short response time may be responsible, since those areas examined were not enlarging. Analyses were also performed to assess the impact of saltwater intrusion and canals and associated spoil banks on land loss. The conclusion of this study is that saltwater intrusion is not a major cause of loss at the three sites, although this could not be discounted for some of the loss at the Cameron site. Dredging of new canals accounts for a significant percentage of loss at the Terrebonne and Lafourche study sites (14 and 15%, respectively), but is not a major factor at Cameron (3%). More difficult to assess are indirect effects, due to possible impacts on hydrology and sedimentation. Analyses of patch size, patch abundance, and results from statistical models indicate that the indirect effects of canal and spoil may be less than has been suggested by previous research, possibly because these studies were too spatially aggregated.

Pages

316

Share

COinS