World Aquaculture Society Meetings

Add To Calendar 24/02/2016 15:15:0024/02/2016 15:35:00America/ChicagoAquaculture 2016EMPIRICAL MODELING OF Vibrio parahaemolyticus CONCENTRATIONS AT THE ESTUARY SCALE   BurgundyThe World Aquaculture Societyjohnc@was.orgfalseanrl65yqlzh3g1q0dme13067DD/MM/YYYY

EMPIRICAL MODELING OF Vibrio parahaemolyticus CONCENTRATIONS AT THE ESTUARY SCALE  

Steve Jones, Erin Urquhart, Meghan Hartwick, Cheryl Whistler, Vaughn Cooper
 
Email: stephen.jones@unh.edu
 

Vibrio parahaemolyticus (Vp) has become an increasingly important public health threat to shellfish consumers in the Northeast US. This increase has occurred mainly over the past five years during which significant regional sea surface temperatures have also increased.  The illnesses have mainly occurred south of the Gulf of Maine, however an increased frequency of cases has also occurred in association with shellfish harvested from Massachusetts Bay, while shellfish-borne illnesses caused by Vp are much less frequent in New Hampshire and Maine.

There is region-wide interest in forecasting risk conditions for Vp in shellfish harvesting waters, thus, it is critical to know what significant environmental and climatic factors can provide the capacity for accurately capturing the temporal changes in Vp concentrations. With such varying conditions in coastal and estuarine ecosystem conditions and reported regional variations in Vibrio concentrations, the scale at which modeling of Vp levels can be accurate is a significant consideration.

Long-term surveillance of Vp levels in the Great Bay estuary of New Hampshire and Maine provides data that span this time period (2007-15). The surveillance included measures of Vp concentrations in oysters by MPN-PCR detection, water temperature, salinity, dissolved oxygen, pH, turbidity, chlorophyll a, dissolved nutrients and rainfall. Several predictive modeling approaches capable of estimating Vp concentrations in oysters from April to December, 2007 to 2013, were developed and compared. Multiple regression analysis revealed that temperature and salinity resulted in the best fit to Vp concentration data and were the only parameters used for the models. Differences in these parameters between the two main surveillance sites resulted in modeling of data for the site most representative of the estuary and not for the tributary site. Temperature and salinity had seasonally varying influences on Vp levels, so a model of the full, undifferentiated dataset and one reflecting seasonal data were tested. Both models provided relatively accurate fits to summertime Vp levels, but the seasonal model better reflected Vp levels in spring and fall, resulting in lower mean error and root mean square error values. A published seasonally segmented model (Froelich et al. 2015) developed for the Neuse River in NC, with a wider and higher range of Vp concentrations and higher salinity and temperature values over a single year, provided a less accurate fit to the Great Bay Vp data, especially during warm summer months. This illustrates the variation in estuarine conditions in different regions, the need for caution in adopting models from areas differing in environmental conditions, and the benefit of local-scale modeling for accurate risk forecasting.

Copyright © 2001-2019 World Aquaculture Society All Rights Reserved.