Though Native Americans lived by, traveled along, and drank from the local waters, much has changed in the past four hundred years. Water quality deteriorated as population increased.  Rivers were used as dump sites for many industrial processes.  Dams and dikes were built which were detrimental to fish migrations. In the 1960s environmental awareness increased, and the resulting Federal Clean Water Act of 1970 began a process to clean up the nation's waters. The Clean Water Act Amendments passed by Congress in 1972 further strengthened the effort to improve water quality by concentrating on the identification of point source pollution (Reimold 1998). Of those waters assessed so far, forty percent were classified as "too polluted for basic uses, such as fishing and swimming (Reimold 1998).  Progress has been made in many rivers by curbing point source pollution, while non-point source pollution still remains nebulous and difficult to remediate.  Education is important to promote understanding of the relevance of lifestyle choices and how these choices can impact the health of rivers.  Watershed management and protection is a vital part of our inheritance.

         The Green Harbor River in Marshfield, Massachusetts is a scenic, shallow river that meanders slowly through marshes on its way to the ocean. It is a second-order stream with cranberry bogs, a golf course, conservation land, and a wildlife sanctuary along the river.  The river flows into the ocean at "the Narrows" near Bluefish Cove in Brant Rock (Appendix 1).   In 1872, a dike and dam were built "to reclaim the marsh lands above for agricultural purposes" (Report of the Joint Board on the Restoration of Green Harbor 1898). The dike restricted the flow of tide into the river, thus lowering the water level as farmers hoped.   However, the land that was primarily salt marsh actually subsided anywhere from a few inches to several feet  (Report of the Joint Board 1898). The dike and its effects on the harbor and river access were highly controversial. One of the major effects was shoaling in the harbor, making boat passage more difficult. Some local residents were adamantly opposed to the dike and a "dike war" ensued between the fishermen and the farmers, the result of which was damage to the structure. The damage was such that the gates were never tight and "at every tide a large quantity of sea-water" entered the river (Report of the Joint Board 1898). In 1898, the Massachusetts legislature voted to remove the dike, but, since there was no money allocated to reimburse the farmers for their losses, the bill was vetoed by the governor.

        According to the Green Harbor River Commission, in 1898 the tide was still entering the river daily, and the river was brackish (Report of the Joint Board 1898). The practice of allowing a tide in the river is assumed to have continued through the early part of the twentieth century, although no specific published references have been found.  The assumption is based on analysis of the 1898 report, early photographs of the river, family histories, and interviews.  In the 1960s, the river was still brackish, with flounder, white perch, eels, striped bass, and many other species living in the river.

        Around 1969, the town did major reconstruction of the dam and sluices.  The old tide gates were replaced with new standard flap gates. Tidal induction was totally eliminated.  Sluice boards were raised up, presumably to keep the river at a predetermined height.  The river became degraded and less brackish.  The Green Harbor River had both anadromous fish runs such as herring and catadromous such as eels.  The fish run was impacted by the flap gates. The herring run used to be enormous according to Harvey.  During the time of Daniel Webster, the menhaden, a species of herring were so abundant that they were caught to fertilize his farm.  He used ten to twelve cartloads per acre (Harvey 1877). It remains to be seen if any herring enter the river past the dike today. Fishermen have reported seeing them in the harbor trying to enter the river. The eels still manage to subsist as the adults can exit on the outgoing tides when ready for their reproductive journey.  The juveniles, returning from the Sargasso Sea are tiny and can enter through small leaks in the tide-gates.  Also, eels can tolerate polluted water.  As an example of their hardiness, by 1955, the increased sewage in England reduced water quality "creating an anaerobic part of the Thames in which the only fish living were eels" (Allan 1995).

        Over time the tide gates leaked again and the river was fairly clean in the early 1990s.  A state river survey in preparation for a 305(b) report showed the river at medium risk of eutrophication, which suggests the river was at least relatively clean at the time of testing (Upton 1997).  Audubon ecologist Elizabeth Colburn found salt water one mile upstream and water level three feet below mean sea level. (Mednick 1993) The Army Corps report (1993) which "assumed "the river to be fresh water" was clearly incorrect.  Local residents and David Clapp of Massachusetts Audubon confirmed that tidal action occurred regularly during the early 1990s.  After the Corps report was published, the town of Marshfield replaced the tide gates, and tidal induction was stopped in 1994.  The estuarine community suffered once again from the lack of salt water and stagnation.

         In the last several years, the gates have been leaking again.  The EPA was aware that one gate was hanging improperly, and, in February 2004, it was noted that one hinge pin was almost falling out.  The town was aware of the problem and contracted Offshore Marine to fix the gate.  On June 24, 2004 the gate was removed and left out for a week.  The tide entered freely and the water level rose, though not anywhere near the predictions in the Corps report which had suggested +1.3 feet NGVD for a high at spring tide (with all sluice boards removed).  According to Rod Procaccino of the DPW, the level never exceeded -0.8 feet (never reached mean sea level).

         A number of groups have been interested in the river this past year.  The site was assessed on February 20, 2004, with representatives from the Environmental Protection Agency (EPA), Massachusetts Coastal Zone Management (CZM), and the U.S. Army Corps of Engineers (USACE).  A meeting with Town of Marshfield Conservation Agent and DPW on October 13 included a presentation of experimental results by Tim Smith of CZM. The EPA, USACE, Mass. Audubon, and National Oceanic and Atmospheric Administration (NOAA) were in attendance.

.

This survey was designed to evaluate a number of points addressed in the U.S. Army Corps report (1993) by determining the water quality, water levels, and biological community. The tide gates were leaking during the study and were intentionally opened for repair and experimentation for three weeks. Water quality monitoring was done approximately once per week.

The hypothesis that the river would be cleaner when tidal flow occurred and degraded during limited tidal action was supported. The water was cleaner during the summer months when tidal flow was regular than in the autumn when flow slowed to a trickle.  The survey of 2003 showed similar results.  In the summer of 2003, there was clean water in July (while the gates leaked heavily) and extremely dirty water in August (when rain, warm temperatures, and lack of inflow contributed to a turbid, coffee colored river).  In the fall of 2003, when  tidal action increased, the water cleared up again (Figure 18). Many variables affect the water quality, some of which are related to tidal induction.  These variables will be discussed later.

Turbidity varied throughout the summer.  After the removal of the tide gate in late June, the secchi disk could be viewed 1.75 meters down one day later.  This was the best rating all summer. One must keep in mind that time of year itself can be a variable in the turbidity.  Early summer seems to have clearer water than late summer. By the following week, when the gate was back in, the turbidity had worsened (0.6 m) because the bottom sediment had been churned up with the substantial tide. An accumulation of brown sediment had collected on the rocks above the dike during the winter, and it was dislodged by the scouring action of the tide. Gradually, the water became clearer as daily tides from leakage entered the river, and the next peak was August 4 at 1.5 m. Afterwards, there was a slight decrease in secchi depth.  In August, the water has historically developed a brown coloration, possibly because of the warmer temperatures and vegetation. This annual change is definitely a variable in turbidity. During the two week experiment in late August, the turbidity increased slightly (1.2 m – 1.1 m). On August 31, two days before the experiment ended, a group met on the dike and commented on the water clarity (no reading taken). 

The secchi depth ranged from 0.6 meters to 1.75 m during summer 2004, compared with 0.2 m to 1.75 m in summer 2003. During August 2003 there was very slight tidal action, and the water looked like weak coffee. Secchi depth was a mere 0.2 m. In comparing autumn of 2004 and 2003, one can see a difference in 2003 when tidal induction was present (1.5 m on 10/11/03) in contrast to the very little daily inflow in 2004 (0.5 m on 10/9/04) (Figure 19).

Total dissolved solids, although not measured quantitatively this year, was visually high after the gate was removed.  There was a great deal of sediment stirred up with the rapid inflow of tide, which caused a rusty deposition on boats, rocks, and shorelines.  These solids eventually settle on the substrate and can impact the river depth and bottom composition.  According to Allan, "sedimentation affects the distribution of fish species which vary widely in their tolerance for silty conditions" (1995).  Sedimentation is assumed to be more pronounced in years without tidal action, since there is more time for sediments to settle in still water than in moving water.  The river depths have changed over the years as was discussed in the survey of the river in 2003.

Water temperature ranged from 64°F to 72°F during July and August (Tables 2 & 3). The mean was 68.5°F. In both locations the temperature was within the acceptable range for Class SB waters and did not exceed 29.4 degrees Celsius (85°F)(MA DEP 1996). The mean surface water temperature in the summer of 2003 was 70°F, with a range of 15°F (60 to 75°F).  July's average was 66.8°F, and August was higher at 74°F.

The surface salinity varied widely, with a range of 30 ppt at Location 1 (L1) and 28 ppt at Location 2 (L2).   The mean surface salinity was 21 ppt at L1 and 17.4 ppt at L2. The proximity to the dike, where the tide enters, affects the salinity.  When the tide comes in, the salt water does not sink immediately because of the momentum. L2, being farther from the tide gates, does not have salinities as high. Both locations had higher salinities at one-meter depths, which was expected since salt water is denser than fresh water. L1 ranged from 21-31 ppt while L2 ranged from 13-34 ppt.  Means were 28 ppt at L1 and 25 ppt at L2. On August 12 the surface salinity was 29 ppt, a questionably high reading that was nonetheless replicated in all parts of the estuary. Between the half drainage situation and high evaporation rate in summer, it may be possible that the salt concentration increases. These high readings, also measured in 2003, led to a hypothesis that low flow estuaries may have higher than expected salinities.  A reference in Cod to salt making being economically viable only in some British estuaries, yet not along the coast, may support this hypothesis (Kurlansky 1997). In a discussion with Pat Harcourt of Waquoit Bay National Estuarine Research Reserve on Cape Cod, Pat confirmed that Waquoit Bay gets similar readings in summer due to high evaporation rates (2005).  She considered the readings as credible.

A Green Harbor River aquarium was set up to observe some of the river organisms.  The salinity increased regularly (to a high of 40 ppt) as the water evaporated.  Only fresh tap water was added to the system, and salt that collected on the edges of the tank was discarded. The increase in salinity occurs when the water level drops. Another possible explanation for the high readings on August 12 is that the SeaTest meter was not very accurate.  However, comparisons with the hydrometer and conversions are generally within 2ppt plus or minus. (The device is less accurate in low salinities than in high salinities.) Future tests will determine whether this anomaly is an actual phenomenon or the result of lack of precision in instruments.

Rainfall impacts the salinity readings. Although in July 2004 the rainfall total was 4.16 inches, there had been little rainfall in August (0.58 inches by August 12).  After 1.45 inches of rain on August 13-14, the surface salinity dropped to 18 ppt (Burtner 2004). An additional 2.1 inches of rain on August 15-16 from the remnants of a hurricane dropped the salinity to 1 ppt. This decline in salinity occurred before the tide gate was chained open on August 18. On a canoe trip upstream on August 17, a salt water wedge was detectable up to Bass Creek (Figure 20). A salinity map was created on August 18 (day the gate was chained open), which showed surface salinities between 0 ppt and 3 ppt and one- meter salinity from 12 ppt to 21 ppt (Figure 21). With a large tidal induction, by August 26 the surface salinity had increased to 28 ppt at L1 and L2. Tim Smith and Jason Burtner took readings upstream during the experiment and found a salt water wedge a distance up the river (Appendix IV).  The gate was closed on September 2, and the surface salinity was 29 ppt on September 6.  On October 9 the level had dropped to 18 ppt and by November 14 to 5.5 ppt.

The dissolved oxygen (DO) levels ranged from 7.3 ppm to 9.0 ppm. (Figure 14).  Tidal action throughout the summer probably contributed to the higher mean levels than the previous summer (7.7 ppm against 6.2 ppm). The summer of 2003 had a greater range (3.2 ppm to 9.0 ppm) which was coincident with tidal action.  During July 2003 there was considerable tidal flow and a mean DO of 7.6 ppm, but in August there was very slight action and warmer temperatures leading to a mean DO of 4.2 ppm. In August 2003 there was a hypoxic event (D.O. 3.2 ppm) in the river, although no fish kills were observed. 

The entire summer of 2004 had sufficient oxygen for aquatic life and met the state's criteria for Class SA waters (not lower than 6.0 mg/l). Dissolved oxygen should not have acted as a limiting factor. The highest DO reading was on June 23, the day after tide gate #3 was removed for repairs.  The excellent mixing associated with a strong tidal inflow increased the DO content.  Another factor could simply be that, earlier in the summer, the mean water temperature is colder, and colder water can hold more oxygen.  The May 25 reading was 8 ppm, and the earliest data from 2003 (on July 10) was 9.1 ppm.  There was good tidal action on July 10, and the measurements were taken on a high tide cycle.  The tidal action has an impact on the temperature, DO, and salinity.  Therefore, all analyses must consider the time of day, the tidal conditions, and location.

Drawbacks of the dissolved oxygen results included the use of the Ward's test kit this year.  The reasons for trying the new protocol were student safety considerations and ease of use.  However, the accuracy and precision are low compared with the LaMotte titration method used in the summer of 2003.  Wards is suitable for approximations, but my confidence in the data is limited. Another weakness was that no DO readings were taken in the fall.  Data would have been helpful to further support the hypothesis.

Mean surface pH was 7.6, the same as in the summer of 2003.  The range was different (2.3 range in 2004 and 1.4 in 2003). One-meter depth means were also similar with 8.0 in 2004 and 7.8 in 2003.  Both summers showed changes related to rainfall events. The mean pH at L1 was slightly more than L2 at surface (7.5) and one-meter depth (7.6). The average values met the Massachusetts standards of 6.5-9.0 (not more than 0.5 units outside background) for Class SC waters.  pH varies according to the amount of dissolved carbon dioxide in the water, which combines to produce carbonic acid, thereby lowering the pH (Allan 1995).  During hours of photosynthesis, more CO₂ is being used, so the pH tends to be higher during the daylight hours (Allan 1995). If data is collected for a 24-hour period, the differences can be seen. The tests were consistently performed in the late morning in order to control for these daily fluctuations.

The Green Harbor River is a shallow and slow moving river. With a gradient of only 0.8 feet per mile from Webster Street to the dike, the flow is insufficient to sustain high water quality without tidal induction.

            The assumption in the Corps Report of 1993 that the river was fresh water at that time is not supported by evidence presented previously.  The concern that tidal inflow would have a negative impact on water quality and the aquatic community is not supported by the research. On the contrary, there appears to be a strong correlation between tidal induction and cleaner water. Also, the aquatic community is not a fresh water community.  All species collected are consistent with an estuarine environment.

In the Corps Report, the predicted water level with one tide gate opened was not supported by either experiment this summer.  When the tide gate was totally removed, the highest level was -0.3 feet (NAVD) (-0.8 feet DPW figures), compared to the Corps model of +1.3 feet (NGVD).  When the gate was chained open, the highest reading was  -0.4 feet (transducer data).    In both cases, the sluice boards could have been much lower, resulting in better drainage and lower water levels at high tide.

In the Green Harbor River, a greater flow could be accomplished by allowing more tide in, and later in the day it would exit with the outgoing tide. As long as daily inflow was equal to daily outflow, there would be no significant difference in flooding potential, while enormous gains could be achieved in habitat improvements and recreation. Not only would the increase in tidal flow keep the water in the river colder, cleaner, and with higher DO, but the exiting tide might help prevent some of the shoaling in the harbor. There are ways to increase flow. One is to lower the sluice boards to get a lower invert. However, unless combined with daily incoming tide, the lowering is only a several-day fix, and the water level remains too low. Flow becomes sluggish when the river is low, and riverbanks are exposed permanently, which has adverse effects on the biological community.  There needs to be sufficient water pressure to open the tide gates on outgoing tides. A permanent restoration of tidal induction would allow for better flow as incoming and outgoing tides were balanced.  The mean water level could remain basically the same as it had been for the past 35 years while clean ocean water could enter daily to protect the ecosystem and slow the growth of Phragmites.

A cleaner river is quickly attainable. The community should strive to restore the river to its former condition: a swimmable and fishable river.  Though careful study is necessary to ensure the protection of all property, the historical and current evidence is clear that tidal flow can and should be allowed into the river during such time as a new management plan for the river can be developed.

1) Water quality monitoring of the river should continue.

2) The biological community should be further documented.

3) Experiments should be conducted with the operation of the tide gates and sluice boards in order to determine a suitable method and configuration that will protect the habitat, water quality, and personal property.

4) The Conservation Commission should change the management policy on the tide gates (if one exists) to permanently allow daily tidal induction.

5) Install a self-regulating tide gate or suitable alternative in order to allow a permanent return of tide to the river.

Addendum :

In December, the Town of Marshfield received a grant to study the feasibility of salt marsh restoration in the Green Harbor River.

The Green Harbor River in Marshfield, MA has been tidally restricted since 1872, when farmers built a dike and dam to reclaim land from the sea.  The ensuing “dike war” caused damage to the dike and the tide gates were never secure afterward.  Tide was still entering regularly according to a comprehensive report completed in 1898.  The river remained brackish until 1969, when the town of Marshfield replaced the tide gates and structure.  The tide was totally eliminated by the resulting changes. Water quality declined and many  species such as crabs, clams, quahogs, horseshoe crabs, striped bass, and flounder disappeared from the river.  Over the years, leakage has occurred, and the depths still appear to have salinity readings consistent with an estuary.  Surface salinities also showed the river to be a brackish waterway.

A survey of the river in the summers of 2003 and 2004 indicated that the water quality was improved when tidal induction occurred and degraded when induction was limited.  Temperatures were higher, dissolved oxygen lower, and turbidity higher during times of low leakage.  Green crabs, shrimp, mummichogs, and sticklebacks were the most captured animals.  No fresh water species were caught.  Clam worms and flounder were also found in the river. 

One of the tide gates was removed in late June,2004  for repair.  Tide freely entered the river for one week.  Even with the large amount of tidal flow, the water level never exceeded mean tide (NAVD). In August one gate was chained open for two weeks, and the effect on the water levels showed no reason for concern.  Some people are under the impression that allowing tide in the river will increase flooding, but that idea was not supported by the experiments or leakage. 

More testing needs to be done, and a large biological sampling would be appropriate to further determine the nature of the river system. Since maintenance on the gates has been completed, the leakage is minimal, and the river is in danger of deteriorating aesthetically and, more importantly, ecologically.  Experimentation must continue to protect the aquatic community living in the river and to assist in restoring a herring run.  The data is clear that tidal flow improves water quality and can do so without raising water levels excessively.  In keeping with the spirit of the Clean Water Act, the town of Marshfield is strongly encouraged to allow regular tidal action once again in the Green Harbor River. 

Allan, JD.  1995.  Stream Ecology:  Structure and Function.  New York:  Chapman and Hall.

Bickford, W & Dymon, U. ed.  1990.  An Atlas of Massachusetts River Systems:  Environmental Designs for the Future.  Amherst, MA:  University of MA Press.

Collette, B. & Klein-MacPhee, G., ed.  2002. Fishes of the Gulf of Maine.  Washington, D.C.:  Smithsonian Institution Press.

Dates, G. & Byrne, J.  1997.  Living Waters:  Using Benthic Macroinvertebrates and Habitat to Assess Your River's Health.  Montpelier, Vermont:  River Network.

Dobson, C. & Beck, GG.  1999. Watersheds:  A Practical Handbook for Healthy Water.  Ontario,Canada: Firefly Books Ltd.

Related sites of interest