Hatching distribution, abundance, and losses to freshwater diversions of longfin smelt inferred using hydrodynamic and particle-tracking models

The distribution of larval fishes is influenced by where they hatch and their movements after hatching. In the San Francisco Estuary, the threatened longfin smelt Spirinchus thaleichthys spawns adhesive eggs in fresh to brackish water. Attached eggs hatch and the larvae disperse seaward toward higher-salinity water. Actual locations of spawning and hatching are unknown, and cannot reliably be inferred from distributions of larvae because intense tidal mixing obscures the history of movement. Human interventions such as manipulations and diversions of freshwater flow may contribute to the ongoing decline of this species, and these effects depend on where the fish hatch. We combined movement estimated using hydrodynamic and particle-tracking modeling with trawl data using a Bayesian model to estimate the location and timing of hatching, as well as natural mortality of larvae and losses to freshwater diversions. Results indicate that longfin smelt may hatch further seaward than previously believed, and that estimated direct losses due to diversions were small relative to other sources of mortality. Abundance of this species varies similar to 100-fold with freshwater flow, but our results suggest that proportional entrainment was practically zero during an extreme wet year and relatively low (2 %) in a moderately dry year. This method could be applied to other estuarine and coastal systems where strong mixing reduces the ability of simpler models to predict hatching locations and larval movements.