Jump to main content or area navigation.

Contact Us

Water: Nonpoint Source Success Stories

Maryland: Constructed Wetlands - Maryland Investigates Dairy Waste Treatment Methods


Constructed Wetlands -
Maryland Investigates Dairy Waste Treatment Methods

Untreated dairy effluent contains high concentrations of nutrients, oxygen- demanding substances, and solids that can adversely affect water quay and the health of aquatic organisms in downstream waters. To address this threat, the State Soil Conservation Service (now the Natural Resources Conservation Service [NRCS]) constructed a waste treatment system consisting of two settling basins, two wetland cells, and a vegetated filter strip at a dairy farm near Frederick, Maryland, in the Monocacy River watershed in July 1993. Other partners in this 319 project are the University of Maryland at College Park, the Maryland Department of Natural Resources, and dairy farmer Clyde Crum.


The project is designed to investigate whether constructed wetlands can provide a cost- effective alternative to conventional technologies, such as lagoons and land application, for controlling animal waste runoff from dairy farms. While constructed wetlands have been found to be effective for treating other waste types (e.g., domestic sewage and industrial waste), few quantitative data are available to document the effectiveness of constructed wetlands for treating dairy waste.

Controlling nutrient releases from dairy farms has some urgency, since Maryland and other states in the Chesapeake Bay watershed have agreed to reduce the input of nutrients to the Bay to 40 percent of 1985 levels by the year 2000. Animal waste, particularly from dairy cows, is a major source of nutrients to the Monocacy River, which flows into the Chesapeake Bay.

Wetland treatment systems

Effluent from the dairy milking parlor flows through one of the settling basins and is then sp to flow into both of the wetland cells in parallel. Runoff from the barnyard flows through the other settling basins and then into one of the wetland cells. Effluent from the wetland cells then flows through the vegetated filter strip and out through a culvert at the downstream end.

The project is designed to investigate whether constructed wetlands can provide a cost-effective alternative to conventional technologies, such as lagoons and land application, for controlling animal waste runoff from dairy farms.

To assess the effectiveness of the wetland treatment system, researchers from the University of Maryland collected surface water samp once each month in each of two settling basins receiving effluent from the milking parlor and barnyard, at the inflow to each wetland cell, and at four to six sites across the length of each cell. Additional samp were taken at the outflow pipes of the cells if discharge was occurring, at a culvert at the downstream end of the vegetated filter strip, and at a groundwater seep area within the strip. The samp were analyzed for several water quay parameters including five-day biochemical oxygen demand (BOD5), total suspended solids (TSS), total Kjeldahl nitrogen (TKN), nitrate-nitrogen (NO3-N), nitrite-nitrogen (NO2-N), ammonia-nitrogen (NH3-N), orthophosphate (PO4-P), and total phosphorus (TP).

Reductions in nutrients

Based on the results of sampling and analysis conducted between May 1995 and January 1997, the treatment system as a whole achieved considerable reductions in all parameters except nitrate and nitrite. The percentage of overall reduction (from the settling basin receiving milking parlor effluent to the outlet of the vegetated filter strip) was 87 percent for BOD, 99.8 percent for TSS, 59 percent for ammonia, 97 percent for total nitrogen, 88 percent for orthophosphate, 94 percent for total phosphorus. Nitrate and nitrite increased from 5.7 to 12.4 mg/L (117 percent), and most of the increase in nitrate and nitrite occurred in the filter strip, suggesting that the strip is a site of nitrification (a precursor to nitrogen removal via denitrification). The settling basins reduced TSS but had tle effect on BOD or nutrients.

Guidelines developed by the NRCS specify as design objectives that constructed wetlands for treating agricultural waste should reduce BOD and TSS to below 30 mg/L and ammonia-nitrogen to below 10 mg/L. Filter strip effluent contained average concentrations of TSS (60 mg/L), BOD (144 mg/L), and ammonia (30 mg/L) that exceeded design objectives. Nonethes, these concentrations are within an order of magnitude of design objectives and represent a tremendous improvement over conditions that existed prior to the treatment system.

Our results suggest that the system could be improved by recirculating effluent through the system or creating another wetland cell downstream of the existing system. The University is continuing to work with the NRCS and Mr. Crum to develop design modifications.

CONTACT: Andrew Baldwin, Ph.D.
Department of Biological Resources Engineering University of Maryland
(301) 405-1198

The Sawmill Creek Project -
Modeling the Watershed Approach

Sawmill Creek -- one of four watersheds selected by the governor's Chesapeake Bay Work Group to develop, demonstrate, and evaluate a coordinated approach to improving water quay and habitat conditions for living resources is using an adaptive management approach to reverse declines in water quay and habitat. Substantial habitat improvements have already been made to a tributary to the creek, and the project may also provide some of the first documented research on the lag times associated with restoration activities.

Profile of the watershed

Sawmill Creek is a second order freshwater stream on Maryland's coastal plain. The watershed drains approximately 8.4 square mi, and the creek flows about 5 mi from its headwaters to its mouth, a tidal estuary near the mouth of the Patapsco River and Baltimore Harbor.

The region was originally known for its productive fruit and vegetable farms. Approximately two-thirds of the watershed has been converted to residential and light industrial land uses over the past 50 years. Development of a major transportation network has had a significant effect on the watershed. The Baltimore Washington International Airport is the center of a web of interconnecting rail lines and interstate highways.

Groundwater withdrawals for municipal drinking water have increased dramatically, and excessive pumping from an unconfined aquifer has reduced the annual base flow in the creek from an average of six cubic feet per second in 1965 to s than one cubic foot per second during more recent dry years.

Project description

A wide spectrum of land owners and land management agencies have pooled their resources to develop the new approach and restore Sawmill Creek: five Anne Arundel county government departments, seven state agencies, three federal agencies, five nongovernmental organizations, several local businesses, and numerous private citizens. Each partner is mandated to use existing programs to achieve the goal. No new funds were allocated for the project, and even section 319 funding was used solely for assessment and monitoring.

The adoption of a watershed perspective (i.e., the coordinated and integrated approach called for by the governor's work group) is intended to be a continual and permanent change in management practice. In this case, the monitoring and implementation teams acted concurrently. The implementation team began to address obvious flaws in historic management practices, while the monitoring team investigated the subtle, cumulative impacts of various land-use practices.

The implementation team drafted a restoration strategy that described the geographic location of each environmental problem, prescribed a general restoration goal, and identified the responsible management agencies for each major problem. The partners then used feedback from the monitoring team's ongoing investigations to revise and improve the details of each restoration project. This interactive process has been described as adaptive management. It continues, but after three years the emphasis has shifted from assessment and planning to implementation and evaluation.

Examples culled from the implementation phase

Table 1. - Habitat scores and fish species on Sawmill Creek.

Substrate and cover 80% 50% 75%
Enbeddedness 65% 25% 60%
Flow 60% 45% 30%
Channel alteration 93% 13% 67%
Pool/riffle/run ration 87% 47% 87%
Bank stability 80% 30% 70%
Bank vegetative stability 90% 40% 90%
Stream side cover 80% 60% 50%
TOTAL SCORE 78% 39% 65%

The implementation phase began in 1994 and actively continues. A wide variety of best management practices have been, and will be, installed as the partners revisit each site using biological health and stream conditions to guide their determination of overall conditions. Thus, for example, the project used a biological survey (EPA's Rapid Bioassessment Protocols) to assess and quantify stormwater problems. Table 1 compares habitat scores and fish populations in a reference stream and a Sawmill Creek tributary before and one year after restoration. The scores shown for each stream parameter are reported as a percentage of a theoretically perfect stream. The last line shows the number of fish species that were found in each stream.

There is documented evidence that six species of fish survive in the restoration area.

The reference stream indicated in Table 1 is covered with second growth forest. It is not pristine but the stream ecosystem is in good condition for a western shore coastal-plain stream. Tributary 9, by comparison, is in an urban portion of the watershed and has approximately 50 percent impervious cover. It drains an area mostly covered by subdivisions built in the late 1940s. Much of the upper stream network was buried in drain pipes under the streets with no stormwater management plan in place to control either the quantity or the quay or runoff.

Habitat improvements on Tributary 9 consisted of reshaping the eroded channel to restore a stable cross-section, gradient, and plane geometry to accommodate the increased stormwater discharge rates. The new channel was stabilized with bioengineering techniques including root-wad revetments, rock weirs, and dense riparian plantings. Table 1 indicates the significant habitat improvements that have evolved from the stream restoration practices installed on Tributary 9. The habitat scores are expected to continue to improve as the riparian plants develop into a mature forest buffer. Experimental stocking of resident nongamefish species has also been accomplished, with help from a local junior high school science club. Thus far, there is documented evidence that six species of fish survive in the restoration area.

Lessons learned

The Sawmill Creek project shows that an ecosystem-based approach can be used to set priorities for watershed management planning. Quantifiable measures of biological health and stream stabiy can be used to guide the integration of a wide variety of best management practices. The approach can be used for both restoration and planning purposes. However, lag time the time that elapses between the installation of a best management practice and the first improved conditions is highly variable depending on the level of action and specific site conditions. As monitoring continues in the watershed, section 319 funding may contribute to research on this aspect of watershed management.

CONTACTS: Larry Lubbers
(410) 260-8701
Watershed Restoration Division

Elysabeth Bonar Bouton
(410) 260-8734
Coastal Zone Management Division
Maryland Department of Natural Resources

Jump to main content.