Aquaculture 2025

March 6 - 10, 2025

New Orleans, Louisiana USA

Add To Calendar 09/03/2025 12:15:0009/03/2025 12:35:00America/ChicagoAquaculture 2025A BAYESIAN NETWORK ANALYSIS OF OYSTER HABITAT SUITABILITYSalon EThe World Aquaculture Societyjohnc@was.orgfalseDD/MM/YYYYanrl65yqlzh3g1q0dme13067

A BAYESIAN NETWORK ANALYSIS OF OYSTER HABITAT SUITABILITY

William S. Fisher* and John F. Carriger

 

Unaffiliated

165 Evergreen Parkway

 DeFuniak Springs FL 32435

 



The life cycle of the oyster Crassostrea virginica includes four distinct life-stages; gametes, larvae, spat (juveniles) and adults. The success of each life stage is determined partially by the success of preceding stages and partially by environmental water and habitat characteristics. Water and habitat qualities change seasonally during the oyster life cycle, and suitable ranges for oyster survival can differ for each stage. As an example, gametogenesis in the spring may be maximized at temperatures around 21oC, but spawning in early summer may require temperatures of 24oC. A critical step in the life cycle of the oyster is spat settlement, where larvae attach to substrate and metamorphose into the juvenile and adult forms.

A Bayesian network, which is a probabilistic graphical model that incorporates multiple variables, was constructed to examine environmental suitability for spat settlement. Bayesian networks can accommodate categorical and continuous variables and can combine empirical data with expert judgement. If the variables are related to one another causally in the network, scenarios with possible combinations of causal factors can be used to estimate the probability of different outcomes on variables of interest.

To estimate spat abundance, ranges of water quality factors (salinity and temperature), habitat characteristics (cultch and turbidity) and success of preceding life stages are incorporated into a Bayesian network. Prior (unconditional) probabilities are based on available literature, and conditional probabilities consider the uncertainty across the range of consequences for different environmental scenarios.