Use of Constructed Wetlands to Improve Water Quality in Finfish Production

The general objective of this study was to evaluate the use of constructed wetlands in improving pond water quality and assess the associated benefits and costs in intensive finfish pond production. The experiments were conducted at the Mississippi State University-Coastal Aquaculture Unit for three catfish growing seasons. Six of the ponds were used in a replicated fashion to evaluate the effectiveness of constructed marshes toward improving water quality while the remaining ponds were used in a range testing fashion to determine optimal wetland design criteria and operation parameters.


Water quality parameters were monitored on both daily and weekly schedules and included factors assessed as part of standard pond management practice (e.g., dissolved oxygen, temperature, pH, total ammonia, nitrite) as well as additional factors (e.g., nitrate and total phosphorous) and parameters (e.g., total suspended solids, chlorophyll a concentration) that were needed for a more detailed assessment of nutrient removal. Dissolved oxygen and temperature were measured twice per day (early morning, mid-afternoon). Salinity, pH, total ammonia, nitrate, nitrite, total phosphorus, total suspended solids, and photosynthetic pigments (chlorophyll a and phaeophytin) were measured on a weekly basis.

The condition of the plant population in each wetland was assessed after the final fish harvest. The number of stems and biomass of plants were measured in each of three 0.25 m2 plots haphazardly located at both the influent and effluent ends of each wetland, for a total of six samples per wetland. Plants were oven-dried to constant weight at 105 C. Each plant variable was compared for the inlet and outlet portions of wetlands between treatments directly as well as expressed as percentage change.

Fish growth was measured by taking the absolute increase in weight from stocking to harvesting time. Food conversion was assessed at harvest time by comparing the quantity of feed used with fish production. Delta Pride Catfish, Inc. conducted the organoleptic testing for off-flavor using standard industry protocol. Individual weekly readings of electric consumption of each pond were recorded. Water flow rates were monitored weekly at the influent end of the marsh cells. Checking and cleaning of the inlet and outlet pipes were conducted on a weekly basis.

Analysis of variance was performed using General Linear Model (GLM) procedure in order to determine the effects of wetland size and water flow rate on water quality variables, critical biological and economic variables and plant biomass. The means of these variables were compared by wetland size and retention time by using the Student-Newman-Keuls (SNK) multiple range test.

The benefits and costs associated with the use of constructed wetlands in intensive and recirculating catfish pond production were estimated by applying the results of the experiment conducted at MSU-Coastal Aquaculture Unit to a multi-enterprise commercial farm in the Mississippi Black Belt area. The point of entry of constructed wetlands in a multi-enterprise farm located in the Mississippi Black Belt was evaluated by using a mixed integer linear programing model.

Mature constructed wetlands were shown to significantly reduce levels of most parameters monitored at the wetland inlets and outlets at the following rates: total ammonia nitrogen (2-63%), nitrite-nitrogen (29-97%), nitrate-nitrogen (28-80%), total phosphorus(52-95%), total suspended solids (2-76%) and pH (0.5-10%). In general, wetlands of standard size and flow rate showed the highest levels of reduction for most parameters monitored at the inlet and outlet ends.

The vegetative characteristics of constructed wetlands varied between treatments and between the inlet and outlet areas of each system. Plant biomass at the inlet portion of wetlands was greater compared to outlet areas, suggesting that plants near the inlets remove and utilize a larger proportion of available nutrients as these compounds flow through these wetlands. For wetlands of variable sizes and fixed retention times, plant biomass was significantly greater for the inlets of standard size wetlands and for the outlets of small size wetlands compared to the same sample points in other configurations. For wetlands of fixed size and variable flow rates, plant biomass was not significantly different across flow rates for either inlet or outlet portions of wetlands.

Results of pond monitoring showed that ponds with wetlands had dampened daily fluctuations in dissolved oxygen and reduced levels of photosynthetic pigments compared to control ponds. These ponds also required less aeration than control ponds. Levels of nutrients within ponds with wetlands were highly variable among treatments and were not less than those in control ponds, but were within tolerable ranges for fish production. These patterns may, in part be related to the large observed variation in rates of fish survival and production across ponds that in turn affect water quality.

The observed improvements in pond water quality as the pond-wetland systems were pushed to their limits by higher fish stocking densities did not produce the expected increase in fish survival, growth, yield and off-flavor and feed conversion. Overall catfish survival during the three growing seasons averaged 77, 68 and 57%, respectively. There were no significant variations observed in fish growth between control and treatment ponds. Observed fish mortality rates were lower during the final growing season when ponds were stocked most intensively and wetland plants had reached maturity. The overall average observed fish mortality rates were 2.46% in 1996 and 1.08 % in 1997. Since the overall average observed fish mortality rates were low, the low fish survival rates during the second and third growing seasons can be largely attributed to predation, primarily by double-crested cormorants. Fish harvest did not vary significantly between control ponds and ponds with standard wetlands, among ponds with variable wetland sizes and among ponds with variable retention times during each of the experimental periods. Except for the ponds with small wetlands, the fish in control and treatment ponds had some degree of off-flavor. The overall net feed conversion ratios were 1.0 to 2.6, 1.8 and 2.4 during each of the three growing seasons.

Ponds with standard wetlands used less electricity to recirculate water from the wetlands to the ponds than those with smaller and larger wetlands. Ponds with wetlands having longer retention times consumed less electricity than those with shorter retention times. Overall aeration time was lower in ponds with standard wetlands than control ponds during each of the three growing seasons. Increased aeration was required when ponds were stocked with larger and more fingerlings during the second and third growing seasons. The costs of chemicals (lime, herbicides, salt) used in experimental ponds did not show any significant variation.

The initial investment requirements for six 8-water-surface-acre catfish ponds on a multi-enterprise farm in the Mississippi Black Belt area added up to $169,312 or $3,527/acre. With six 1.2-water-surface-acre wetlands built adjacent to the catfish ponds, total initial investment requirements rose by $127,498 or $2,656/acre. To build six 2.0-water-surface-acre wetlands adjoining the catfish ponds, an additional initial investment amounting to $199,703 or $4,160/acre would be required. Total initial investment would rise by $271,908 or $5,665/acre when six 2.8-water-surface-acre wetlands are built adjacent to the catfish ponds.

Total specified operating and ownership costs of the 48-acre Mississippi Black Belt catfish enterprise without constructed wetlands amounted to $184,074/year, averaging $3,835/acre or $0.57/lb. When six 1.2-acre constructed wetlands were added to the catfish enterprise, total costs rose by $16,776 annually, $350/acre or $0.052/lb. The use of six 2.0-acre constructed wetlands in the 48-acre catfish enterprise led to an increase in total costs by $19,225 annually, $401/acre or $0.059/lb. Building and operating six 2.8-acre constructed wetlands adjacent to the six 8-acre catfish ponds would increase total costs by $25,412 annually, $529/ acre or $0.079/lb.

Catfish production in a multi-enterprise farm in the Mississippi Black Belt is a profitable enterprise. Assuming present technology, the use of constructed wetlands in intensive recirculating catfish pond production, however, is not economically viable. The use of constructed wetlands in a 48-water acre catfish farm in a multi-enterprise Mississippi Black Belt area would be acceptable to farmers if the annual yield of marketable catfish would increase by at least 5% above that of the current yield of ponds operated without wetlands.

The annual cost estimates for catfish production are lower than those reported recently for catfish production in the Mississippi Delta. The lower cost of production is attributable to the nature of the catfish production systems in the Mississippi Black Belt area. Some farm-wide assets (management, building, tractor) are jointly used in several enterprises (catfish, crops) thereby reducing fixed cost. There were no pond renovation costs included in the variable cost of producing catfish in the Mississippi Black Belt area. The average costs of pond renovation, management and water supply for a 160-acre catfish farm in the Mississippi Delta were $0.0103, $0.0071, and $0.0300/lb, respectively.

The investment requirements for constructed wetlands were estimated by using one-foot plants in creating the vegetative cover within a short period of time. This method of creating vegetative coverage resulted in a very large cost outlay for the purchase and planting of the desired vegetation. Other cheaper and equally effective methods of creating vegetative coverage have not been explored at this stage of developing this technology. Likewise, other plant species need to be tested for their effectiveness in improving water quality and fish production.

The results of experimental trials conducted at MSU-Coastal Aquaculture Unit showed that at 8,000 fingerlings/water acre stocking density, the expected revenues were not sufficient to justify the costs associated with constructed wetlands. Further testing of this technology needs to be undertaken in order to ascertain the effects on the yields of marketable catfish or other fish species under intensive commercial scale operations.