Abstract
We examined the 5-yr (1996-2000) response of subtidal marsh creek fishes (2,793 trawls, 47 species, 30,719 fish) to a large marsh restoration project in the upper Delaware Bay, and found that the salinity gradient covaried along with marsh surface vegetation type among two treated and one untreated reference sites, confounding direct comparison of fish utilization. Examination of environmental correlates with monthly and yearly trends highlighted differences between potential mechanisms driving assemblage dynamics either intrinsic or extrinsic to the marshes. Within-site and among-site differences in fish assemblage, as described by principal components analysis, correlated poorly with marsh vegetation on both seasonal and interannual scales and appeared to driven by larval supply. Assemblage dynamics could be explained in part by the occurrence of juveniles of transient marine fishes along a salinity gradient (0-15.2‰ range in monthly site mean), but were largely determined by fluctuations in the distribution of two transient species: young-of-the-year bay anchovy Anchoa mitchilli and Atlantic croaker Micropogonias undulatus. A minor mode in variance, driven by locally spawned species, was moderately correlated with environmental parameters. Analysis of marshes on an individual basis did not discern additional important gradients. Our findings are in contrast to those in systems dominated by resident species, probably because transient fishes, which often dominate the system, are more plastic to the nature of ecological services or are affected as much by environment outside of the marsh as by that in the marsh.