Phase 2 – Previous Research

Stream Stability, Nutrient Dynamics,
and Bacteria Dynamics

Introduction

Figure 1. Aerial photograph of the Phase II marsh area taken in June 2009

During the second phase of the project, (completed in 2007 with final tree planting in 2008) 22 ha of non-riparian hardwood wetland, 9 ha of riparian freshwater wetlands, and 14 ha of tidal marsh have been constructed. This phase required much more extensive earthwork than Phase I, so much insight into design and construction techniques was gained. The use of specialized equipment suited for extreme wet conditions, well planned construction sequencing and timing, and topsoil replacement are examples of ways to efficiently ensure restoration success while minimizing cost.

The tidal marsh portion created in this second phase has provided a unique demonstration in terms of location (prior converted farmland), scale (14 ha), and species planted (included infrequently used but native Juncus roemerianus). Research on the marsh has focused mainly on the stability of the tidal stream, downstream water quality (since agricultural drainage water also has been diverted through the system), and vegetation survival. Primary research goals for this effort included:

1. Evaluate the design and construction techniques utilized in restoring a large-scale coastal marsh.
2. Assess the ideal target elevations for the marsh vegetation used in the planting plan
3. Determine the post-construction stability of the tidal stream portion of the marsh restoration.
4. Establish whether the tidal marsh provides water quality improvement to drainage water diverted into the site, in terms of N, P, and fecal bacteria
5. Provide NCEEP with recommendations based on our research experience at the site that will guide future tidal marsh restorations

Figure 2. Final restoration design developed and implemented by the NCSU team. Research area on the tidal marsh portion of the restoration indicated in the red box.

Post Construction Tidal Stream and Marsh Stability

Stream stability monitoring consisted of periodic cross section (twice yearly) and longitudinal (yearly) surveys. A total of 27, 12.2 m (40 ft) wide permanent cross sections (10 pools+17 runs) were established along the stream corridor and surveyed seven times from August 2006 to June 2009. Additionally, three longitudinal surveys (completed 1/07, 3/08, and 3/09) of the thalweg, top and bottom of the stream bank were conducted using a total station to monitor changes in stream pattern and profile.

Results

In the three years of post construction monitoring, very little change has occurred in the dimension and pattern of the tidal stream. Each cross section has maintained fairly consistent width, depth, bank slope and bankfull cross sectional areas. There is no evidence of any significant bank failures at any point along the stream.

Figure 5. Example cross sectional survey results

The yearly total station surveys of the plan form of the stream show that there has been little to no lateral migration of the stream channel. Both the cross sectional and longitudinal profile surveys show evidence of changes in bed slope (or profile). The entire reach shows a pattern of pool and run features located in the expected plan-form locations.

Figure 6. Thalweg profile of the restored tidal stream

Conclusions

Based on the three years of monitoring data, the tidal stream and marsh appear to be stable. The stream appears to be maintaining its dimension and pattern, and there is evidence of pool/run development throughout the stream reach. However, some aggradation in the channel bed has been noted of late, and has most notably contributed to deceased depth of a few pools formed previously in the meander bends. The changes in stream profile are not significantly affecting channel dimensions, and the stream appears to still be able to convey design flows and maintain good floodplain connection.

Nutrient and Bacteria Dynamics within the Tidal Stream

Water quality monitoring of the restored tidal stream (Broome’s Branch) was conducted for two years post construction (2007-09) to determine if the restored stream and marsh provided any water quality benefits to influent agricultural drainage water.

Water quality monitoring consisted of a combination of grab and time weighted composite sampling. Automated ISCO samplers were installed at three locations (BBup, BBmid, and BBdown) along Broome’s Branch (BB). Monthly grab samples were collected at each sampler location, from the main drainage canal from Open Grounds Farm (MD01) and from the reference tidal stream (P2Ref). The samples were analyzed for total kjeldahl nitrogen (TKN), ammonium nitrogen (NH4-N), nitrate nitrogen (NO3-N), and total phosphorus (TP). Additional monthly grab samples were collected on rising and falling tides from each location and analyzed for bacteria (Fecal Coliform and E. coli) by Dr. Bill Kirby-Smith at the Duke Marine Lab.

Figures 7 & 8. Example ISCO Sampler Setup

Figure 9. Sampling locations

Results

Nutrients

Analysis of monthly grab samples show all sites to have higher (0.7 to 1.1 mg/L) median concentrations of TKN compared to the median concentrations of NH4-N, NO3-N, and TP (all <0.2 mg/L). Nutrient concentrations were lower at reference stream (P2Ref) than in the restored tidal stream (BB) and the agricultural drainage canal (MD01). TKN and TP concentrations decreased from upstream to downstream in the restored tidal stream. NO3-N concentrations were similar at all three sites in the tidal stream, but less than MD01 and greater than P2Ref.

Figure 10. Median nutrient concentrations Bacteria

Bacteria

The results of the bacteria sampling are highly variable from site to site and between rising and falling tides. Counts frequently exceeded testing limits (1600 MPN) and no immediate trends have been detected in the restored stream or between rising and falling tides. However, it is evident is that the bacteria concentrations observed in the restored tidal stream are within the range of what was observed in the reference stream.

Figure 11. Median bacteria counts

Dye Tracer Study

In order to compare hydraulic retention times of the tidal stream versus the bypassing agricultural drainage canal, a dye tracer study was conducted. The travel path from the dye injection location to the sampling location was about 40 m (5%) longer in Broome’s Branch. The travel times were an average of two falling tides, recorded on consecutive days in July 2008. Analysis of the dye samples collected from the lower ends of Broome’s Branch and the drainage canal show that the travel time through the constructed tidal stream is approximately one hour (60%) longer than down the canal. Reduced velocities in the tidal stream are attributed to increased sinuosity, decreased width, and increased frictional resistance from vegetated banks.

Conclusions

The two years of post construction monitoring show the restored tidal stream to be providing some water quality treatment, particularly retaining NO3-N and TP. This indicates that the tidal stream and marsh are providing an environment for denitrification, nutrient uptake by marsh vegetation, and/or settling of sediments and associated bound nutrients. The results of the dye tracer study show that there is increased travel time in the restored stream and thus greater opportunity for these processes to occur. Periodic flooding of the marsh also contributes to the treatment potential of the system.

The results of the bacteria monitoring are not as conclusive. The restored stream and marsh could potentially be a source of bacterial contamination due to the increasing number of wildlife species present. The system may not be providing the appropriate environment for bacteria die-off (exposure to UV light, salinity, high temperatures) or ample time for treatment. It must be noted that similar concentrations of bacteria were observed in the reference stream area. Further investigation into the sources of bacteria at both locations would be beneficial.