Bayesian network model that predicts the probability of habitat availability (days) per winter month for age-0 Gulf Sturgeon at a 30-m pixel scale in Apalachicola Bay, FL

Publication Date

2021-01-20

Time Period

2019

Cronin, J.P., Dale, L.L., Brink, V.L., Tirpak, B.E., and Tirpak, J.M., 2021, Data for Gulf Sturgeon Bayesian Network Model: U.S. Geological Survey data release, https://doi.org/10.5066/P9KNSKMT.

Type: Related Primary Publication

Data were collected to create a spatially explicit Bayesian network model for Gulf Sturgeon that predicts habitat availability (days) for 75 alternative physiological and habitat scenarios, which were the unique combination of river discharge, winter month, and month of arrival to the estuary. The tabular data associated with this metadata were a combination of empirical data available in the literature, expert elicitation, and simplifying assumptions. These tabular data were used to populate conditional probability tables for predicted variables. Of the 75 possible model outputs, geospatial datasets could not be created for 28 scenarios due to missing data and 12 scenarios were excluded due to incongruous regression results. When entering evidence directly into the Bayesian network, the 40 omitted scenarios have equal prior probabilities for each acceptable water condition (days) class.

• Gulf Coast Prairie Landscape Conservation Cooperative
• LC MAP - Landscape Conservation Management and Analysis Portal

Data source

Input directly