A Growing Degree‐Day Approach to Estimate Larval Hatching Sites Using Backtracking Simulations Without Larval Age

Abstract Walleye ( Sander vitreus ) are a valuable commercial and sport fishery in the North American Great Lakes, yet the relative contributions of known spawning sites to the Lake Erie population remain poorly understood. Backward‐in‐time Lagrangian particle tracking (backtracking) models have been widely applied to estimate larval origins, but the required simulation duration (i.e., larval age), typically derived from otoliths, often remains unknown. We developed a novel proof‐of‐concept approach to estimate larval age without otolith data, using an observed exponential relationship between the total length of tank‐reared Walleye larvae and growing degree‐days (GDD °C days): , where = 8.34 mm is length at hatch and = 0.0027 (°C days) −1 . GDD for wild larvae was estimated from measured length. Using temperatures experienced by larvae along their paths from backtracking models, we estimated larval age and the origins of observed Walleye larvae in western Lake Erie. Larval age predicted by the GDD‐age model were 72% of those estimated using an age‐length curve from the tank‐reared Walleye and 53% of those estimated using a common GDD‐approach based on mean‐basin temperatures. In all three approaches, larvae were backtracked to the southern shore of the western basin, with 43%–60% originating from hard substrates. We neglected differences in environmental conditions between the tank‐reared and wild‐hatched larvae, which affected the values of and . Sensitivity analyses showed <12% variation in had negligible effects on predicted hatching sites. Our novel GDD approach is an additional tool for estimating larval age and simulating larval origins when otolith data are unavailable.