Changes in salt marsh vegetation, Phragmites australis, and nekton in response to increased tidal flushing in a New England salt marsh

This study examined the response of Argilla Marsh in Ipswich, Massachusetts, USA to increased tidal flushing instituted to restore a salt marsh invaded by Phragmites australis. In late fall 1998, we replaced the old 0.9-m-diameter culvert feeding this marsh with a 2.4 × 1.5-m box culvert, thus increasing both the volume of tidal exchange and porewater salinity. We carried out yearly sampling of vegetation for two years pre-restoration and for four years after restoration. Analysis of Similarities (ANOSIM) showed that the plant community on the restored marsh had changed after restoration but that on the reference marshes had not. Over 80% of the change in the restored marsh was attributed to an increase in Spartina alterniflora cover and decreases in the cover of Phragmites australis, Typha angustifolia, and Solidago sempervirens. The two brackish species, P. australis and T. angustifolia showed an immediate negative response to the increased flooding and salinity. Surviving P. australis culms in the restored marsh were shorter in stature than they were prior to restoration, suggesting that the increased flooding and porewater salinities had lowered the productivity of this species. The increase in S. alterniflora post-restoration fit an exponential curve, indicating that there was a lag in its response initially, but then it expanded rapidly and was still in a very rapid expansion phase after four years. Despite an overall decline of P. australis on the scale of the whole marsh, there was a great deal of variation in responses of individual patches of P. australis to the restoration. Some declined, some were unchanged, and some even increased. The response of nekton to the restoration was less obvious than that of vegetation. Before restoration, the creek system in the tidally restricted marsh functioned like an impoundment that was only marginally connected to the larger salt marsh ecosystem. At that time, seining indicated that more species of nekton occurred at greater densities in creeks in the tidally restricted marsh than in the downstream reference. Increasing the tidal amplitude in the restored marsh resulted in an overall decline in the catch per unit effort there. In contrast to the creeks, the Spartinadominated section of the flooded marsh surface in the restored marsh did harbor more Fundulus heteroclitus, particularly the smaller size class, than did the downstream reference marsh or a P. australis patch in the restored marsh. Our analysis of vegetation and nekton suggests that Argilla Marsh was still adjusting to hydrologic changes four years after restoration.