For the sanitary control of shellfish produced and sold for human consumption, the U.S. Food and Drug Administration (FDA) and the Interstate Shellfish Sanitation Conference (ISSC) administer the National Shellfish Sanitation Program (NSSP). The NSSP Model Ordinance requires that each shellfish growing area should be classified as approved, conditionally approved, restricted, conditionally restricted, or prohibited. The state of Alabama uses the adverse pollution condition strategy of the NSSP as its water quality management strategy, wherein most of the oyster growing and harvesting areas in Alabama coastal waters are classified as conditionally approved. As part of this strategy, oyster harvesting is closed in adverse pollution conditions that are triggered by an environmental variable. In Alabama, oyster harvesting is closed when the river stage at the Barry Steam Plant (BSP) exceeds eight feet. Under the adverse pollution condition strategy, oyster harvesting is reopened when the river stage recedes below eight feet and water sampling shows that the pollution event has ended. As part of this ongoing research, it was shown that using the river stage at the BSP is suboptimal for both closing and reopening harvesting. A Bayesian theory based Dynamic Linear Model (ALOHA BAM) was developed that demonstrated its merit in predicting water quality conditions in the Mobile Bay and could be effectively implemented to predict adverse pollution conditions [1].

This research investigated another water quality management strategy as outlined in the NSSP guidelines, called the systematic sampling. Every station currently sampled by the state of Alabama with at least thirty historical samples (as required by the NSSP guidelines for systematic sampling) was subjected to a retrospective study for qualification under the systematic sampling strategy. Most of the stations currently sampled in Alabama qualified under the systematic sampling, and can be open year round for oyster harvesting. For stations where water quality is sensitive to the effects of local events such as high rainfall or sewage spill, a modified Bayesian theory based Dynamic Linear Model (modified ALOHA BAM) is proposed, where the posterior model is developed by factoring in the likelihood of water contamination by the effects of local events. A combination of systematic sampling and the modified ALOHA BAM will be optimal for management of Alabama coastal waters in the context of oyster harvesting.

References: 1. M. Colomb, M. Misra and K. Park "A Bayesian-Probability Based Model of Fecal Coliform Bacteria for Closing and Reopening of Oyster Harvesting Areas in Alabama Coastal Waters", Env. Sci. Tech., (to be submitted, May 2007)