Emigration-corrected seasonal survival of a size-structured fish population in a nursery habitat

We studied how emigration and seasonal dynamics affect apparent survival estimates in a nursery habitat, thereby altering habitat valuation. During a 3 yr study, we marked 1917 juvenile common snook Centropomus undecimalis and resighted 85.7% with a telemetry array in 4 mangrove creeks. We grouped individuals by size class, marking year, and creek, and estimated survival (phi) using the Barker joint-data model. Using telemetry data, we estimated seasonal emigration probabilities and simulated the effect of emigration on phi estimates. We found a minimal effect of emigration on phi estimates, the magnitude of which may be explained by emigration underestimates due to telemetry coverage or high resighting rates reducing bias. We found seasonal and size-based variation in phi estimates. Our approach allowed survival estimation during a severe thermal-disturbance event that reduced survival in 2 creeks and may have had its impact mitigated in a third creek by an anthropogenic habitat alteration serving as a thermal refuge. Significant fine-scale spatial segregation of size classes likely acted to reduce intercohort cannibalism, thereby maintaining the high survival rates. We recommend caution when estimating apparent survival, since the lack of estimated survival bias was likely due to our extensive tele metry data, which buffered against the effects of emigration. We conclude that temporal variability in juvenile habitat use and survival requires seasonal classification of nursery habitat value, the consideration of multi-habitat nursery mosaics, and an exploration of the role of predation on habitat use.