PARENT SESSION
Oral Session 62: Fish Ecology II: Salmon.
Presiding: J Power
Wednesday, August 4, 8:00 AM to 11:30 AM, Meeting Room B 117.

Estimating the strength and shape of size-dependent survival in stream salmonids.

Letcher, Ben^*,1, Horton, Gregg¹, Dubreuil, Todd¹, O'Donnell, Matthew¹, ¹ Conte Anadromous Fish Research Center, Turners Falls, MA, USA

ABSTRACT- The strength and shape of size selection can significantly influence population dynamics and rates of evolution, but size selection can be difficult to quantify in the field. Difficulties include bias resulting from heterogeneity in probability of capture and emigration among individuals. Using simulations designed to test the power of detecting size-dependent survival using the Cormack-Jolly-Seber mark-recapture model, we determined the level of model complexity required to effectively uncover a range of shapes and intensities of size selection for various probabilities of survival and capture. We then applied these results to field data on individually-tagged Atlantic salmon (Salmo salar) from a long-term study site. The study site was bounded by PIT tag antennas allowing estimation of emigration out of the study site. We also accounted for size-dependent variation in probability in the mark-recapture models. The very wide size ranges (often > 100% of the mean) exhibited by the salmon in our study stream suggested ample opportunity for size selection. Simulation results indicated that population size and probability of capture had much smaller effects on bias in detecting directional, stabilizing and disruptive forms of size selection than did overall survival and strength of selection. The simulation also suggested that size selection in the field should be identifiable with little bias for representative sample sizes (>500), probabilities of capture (0.6 – 0.8) and survival (0.5 – 0.8). Across two cohorts, we found evidence of size selection for three out of 12 sampling intervals. In two of these, the form of selection was disruptive and in one the form was directional (negative). In a system with a high likelihood of detecting size selection (good power based on simulation results, and both emigration and heterogeneity in capture success were accounted for in the field study), we found no strong patterns in size selection. The lack of strong size selection we observed may be related to the high level of habitat heterogeneity in streams which offers shelter for a wide range of fish sizes.

Key words: mark-recapture models, stream, Size selection, Atlatic salmon