An 11-year (1990-2001) study of the Ellison Park wetland, a 423-acre, predominantly cattail (Typha glauca) wetland at the mouth of Irondequoit Creek, was conducted to document the effects that flow modifications, including installation of a flow-control structure (FCS) in 1997 and increased diversion of stormflows to the backwater areas of the wetland, would have on the wetland's ability to decrease chemical loads transported by Irondequoit Creek into Irondequoit Bay on Lake Ontario. The FCS was designed to raise the water-surface elevation and thereby increase the dispersal and detention of stormflows in the upstream half of the wetland; this was expected to promote sedimentation and microbial utilization of nutrients, and thereby decrease the loads of certain constituents, primarily phosphorus, that would otherwise be carried into Irondequoit Bay. An ecological monitoring program was established to document changes in the wetland's water levels, biota, sedimentation rates, and chemical quality of water and sediment that might be attributable to the flow modifications.

Water-level increases during storms were mostly confined to the wetland area, within about 5,000 ft upstream from the FCS. Backwater at a point of local concern, about 13,000 ft upstream, was due to local debris jams or constriction of flow by bridges and was not attributable to the FCS.

Plant surveys documented species richness, concentrations of nutrients and metals in cattail tissues, and cattail productivity. Results indicated that observed differences among survey periods and between the areas upstream and downstream from the FCS were due to seasonal changes in water levels—either during the current year or at the end of the previous year's growing season—that reflected the water-surface elevation of Lake Ontario, rather than water-level control by the FCS. Results showed no adverse effects from the naturally high water levels that prevail annually during the spring and summer in the wetland, nor from the short-duration increases in water levels that result from FCS operation. Fish surveys documented the use of the wetland by 44 species, of which 25 to 29 species were found in any given year. Community composition was relatively consistent during the study, but seasonal and year-to-year variations in dominant resident and nonresident species were noted, and probably reflected natural or regional population patterns in Lake Ontario and Irondequoit Bay. The FCS allowed fish passage at all water levels and had no discernible adverse effect on the fish community.

Bird surveys documented the use of the wetland by more than 90 species for breeding, feeding, and migration. Ground-nesting birds were unaffected by the FCS. Seasonally high water levels, rather than short-duration increases caused by the FCS, might have caused the scarcity or absence of certain wetland species by limiting the extent of breeding habitat for some species and the exposure of mud flats that attracted other species. Some noticeably scarce or absent species also were rare or absent elsewhere along the south-central shore of Lake Ontario.

Benthic-macroinvertebrate studies were of minimal use for evaluating the effect of the FCS because no surveys were conducted after FCS installation. The precontrol results allowed assessment of the ecological quality of the wetland on the basis of biotic indices, and generally indicated moderately to severely impaired conditions. Differences between the macroinvertebrate communities in the southern part of the wetland and those in the northern part were attributed to habitat differences, such as substrate composition, water depth, and density of submerged aquatic vegetation.

Sedimentation rates in the areas upstream and downstream from the FCS increased after the flow modifications, more in the area upstream from the FCS than in the downstream area. The concurrent downstream increase and the dynamic patterns of deposition and scour indicated that although the FCS and the other flow modifications undoubtedly were major factors in the postcontrol upstream increase in sedimentation rates, other factors, such as the magnitude, frequency, and the timing (season) of peak flows, might also have contributed.

Periodic analyses of sediment samples from three longterm depositional sites in the wetland documented the concentrations of major and trace elements, polycyclic aromatic hydrocarbons, and organochlorine and organophosphate compounds. The concentrations of most constituents showed no substantial fluctuation or consistent upward or downward trend during the years sampled, nor did they identify any change after FCS installation. Comparison of the measured concentrations with sediment-quality guidelines that are used to assess the ecological quality of substrate environments indicated that the wetland was moderately to severely impaired—an assessment consistent with the benthic-macroinvertebrate biotic indices.

During the precontrol period (1990–96), the wetland was a sink for particulate constituents (removal efficiencies for total phosphorus and total suspended solids were 28 and 47 percent, respectively), but had little effect on conservative constituents (chloride and sulfate). The wetland was a source of orthophosphate and ammonia (removal efficiencies were -38 and -84 percent, respectively).

During the postcontrol period (1997–2001), the wetland continued to be a sink for particulate constituents (removal efficiencies for total phosphorus and total suspended solids were 45 and 52 percent, respectively); the exportation of orthophosphate by the wetland decreased (by 7 percent), whereas that of ammonia increased (by about 70 percent). The outflow loads of orthophosphate and ammonia represented about 15 and 2.3 percent of total phosphorus and total nitrogen loads, respectively. Changes in the loads of conservative constituents were negligible, and the overall removal efficiencies for other constituents during the precontrol period differed from those of the postcontrol period by no more than 5.4 percent.

Statistical analyses of monthly inflow and outflow loads indicated significant differences between inflow and outflow loads of most constituents during the pre- and postcontrol periods. Load data were adjusted to remove the effects of dissimilar hydrologic conditions that prevailed during the pre- and postcontrol periods, and to isolate the water-quality-improvement effect that could be attributed solely to the FCS. Results indicated that the FCS contributed significantly to the decrease in total phosphorus loads, and slightly to a decrease in ammonia-plus-organic nitrogen loads, but had little or no significant effect on loads of other constituents.