Improved Habitat Assessment of the Eastern Oyster in a Gulf Coast Estuary Through Site Occupancy Modeling and Local Ecological Knowledge

Identifying habitat features that support persistent oyster populations is fundamental for effective reef restoration. A single-season site occupancy modeling framework was applied to evaluate drivers of eastern oyster (Crassostrea virginica Gmelin, 1791) presence at 60 subtidal sites in Suwannee Sound, FL. Surveys spanned gradients of substrate type, depth, temperature, and dissolved oxygen and paired randomly selected sites with sites chosen by an experienced commercial harvester to integrate local ecological knowledge (LEK). Substrate type was the strongest predictor of occupancy: shell-bottom sites had high occupancy probability [ψ = 0.857, 95% confidence interval (CI: 0.639–0.953)], whereas mud (ψ = 0.200, 95% CI: 0.086–0.400) and sand (ψ = 0.071, 95% CI: 0.010–0.370) sites were rarely occupied. Models based only on depth, temperature, or dissolved oxygen performed poorly relative to the substrate model (ΔAIC > 16), indicating limited predictive value without substrate information. An observer-specific model estimated higher detection probability at harvester-selected sites (P = 0.952, 95% CI: 0.862–0.985) than at randomly selected sites (P = 0.083, 95% CI: 0.040–0.164), but exploratory models indicated that observer effects operated primarily through LEK-guided site selection rather than detection skill alone. Harvester-selected sites also had consistently higher occupancy probabilities across substrate types, suggesting that LEK captured habitat features beyond broad substrate category, such as cultch density, reef elevation, or persistent shell structure, that were not measured directly. This occupancy framework provides a rapid, empirical tool for prerestoration habitat assessment that complements regulatory site screening and offers a field-validated alternative to habitat suitability index models that are rarely verified with site-specific data.