Date of Award

8-2010

Degree Type

Masters Thesis

Degree Name

Master of Science (MS)

Department

Biological Sciences

Committee Chair

Patricia Biesiot

Committee Chair Department

Biological Sciences

Committee Member 2

Ray Bauer

Committee Member 3

Robb Diehl

Committee Member 4

Carl Qualls

Abstract

The present study focused on factors influencing Alpheus distribution in the Gulf of Mexico and the northwestern Atlantic Ocean. The objectives were to perform a biogeographical analysis of the currently known Alpheus species distribution, develop a spatially explicit predictive model, and experimentally verify the importance of substrate as a key modeling variable through a laboratory-based substrate choice study. Significant predictor variables included in the final model for the Gulf of Mexico were Shepard code, seabed class, the quadratic artificial reef term, the quadratic shore term, the distance to shore, the indirect predictors of longitude and latitude, and three interaction terms (longitude and latitude, quadratic reef term and quadratic shore terms, and Shepard code and distance to shore). The predicted probabilities of snapping shrimp absence were comparable to those found in the training data set for the Gulf of Mexico. The AUC indicated that the best model built on the Gulf of Mexico training data set was not able to predict snapping shrimp distribution in the northwestern Atlantic better than a random prediction. The laboratory study verified that the snapping shrimp Alpheus heterochaelis exhibit a significant preference for substrates in a manner consistent with the results of the biogeographical analysis. The present study elucidates relationships among environmental characteristics and genus level spatial distribution of snapping shrimp in the Gulf of Mexico, but not the northwestern Atlantic. Further refinement of this model will enable more accurate prediction of snapping shrimp distribution.

Share

COinS