Explaining advection: do larval bay anchovy (Anchoa mitchilli) show selective tidal-stream transport?

Bay anchovy (Anchoa mitchilli) have been reported in several systems to display net up-estuary larval movements, against the mean flow. However, the means by which such transport occurs is poorly understood. We address how estuarine circulation and larval behaviors serve to transport larvae of the bay anchovy. In two successive summer seasons, we conducted multiple near-synoptic samples of larval distributions and water column structure along a 45-km section of the middle Hudson River estuary. The analysis focuses on patterns in the vertical distribution of larvae that may help explain transport, and the along-river distribution of different ontogenetic stages. The prediction that post-flexion larvae induce selective tidal-stream transport (STST) by vertically migrating in conjunction with tidal or diel cycles was tested via harmonic regression. Larval concentrations and average larval depths often varied with tidal stage. Maximum concentrations tended to occur at times of slack water, and larvae were often closer to the surface during slack tides as well. These patterns may be the result of tidal movements of horizontal abundance gradients, rather than vertical migrations. The prediction that larval transport is facilitated by a preference for deep water was addressed via analysis of variance, testing for depth effects on time-averaged concentration estimates. At some sites, larvae were most concentrated at intermediate depths, which would promote retention (no net horizontal movement) or slow up-river transport. However, in 1996, larvae were found most concentrated at the surface at two sites, suggesting down-river advection. With respect to along-river distribution, we tested the prediction that ontogenetic stages differed in their distribution in a manner consistent with up-river transport. In 1995, pre-flexion larvae were distributed further up river than eggs, and post-flexion larvae were slightly up river of pre-flexion larvae. Along-river distributions were perturbed in 1996 by a storm that caused high run-off and forced larvae down river. Following this event, along-river position did not vary with ontogenetic stage. The study design that combined analysis of larval depth distribution and along-river distribution enabled us to make and test predictions regarding transport processes. A portion of our depth distribution data implied stasis or weak up-river advection, and we found evidence that this was the case.