top of page

Extracting phosphorus through constructed wetlands

There is no scientific consensus regarding the growing appearance of blue-green algae (cyanobacteria) in Lake Kinneret and other water bodies. Some scientists ascribe this to periodic natural cycles, though they appear to be growing in intensity. Others refer to the increasing phosphorus concentrations coupled with declining nitrogen as a condition which favors cyanobacteria, capable of N-fixation. Hence increased soluble phosphorus which is not consumed by algae is proof that this is the source of eutrophication.

Either way, global data connects excess phosphorus (even at fairly low levels of <1 mg /L ) to the deterioration of lakes, and a significant worldwide effort is made to prevent this phenomenon. The excess phosphorus that comes from agricultural leachate and the trout farms in Lake Kinneret headwaters is not high in absolute values, however, it accumulates to a high level of phosphorus that reaches the Kinneret.

During the recent years, there have been many studies on the ability of different types of constructed wetland systems to act as effective long term solutions in this matter.  Many lab studies have shown high ability to reduce phosphorus with the help of soil-integrated plant systems. Nevertheless, the ability of these studies to predict results is unclear. There is justified criticism against the notion of trying to prove the claim of the efficiency of phosphorus levels reduction, for two main reasons:

1.            Laboratory studies cannot simulate complete ecological systems, and therefore it is impossible to conclude from its results how the levels will be reduced in real conditions.

2.            These are often very short-term studies, while the mechanisms for reducing phosphorus split into long and short-term and, naturally, when planning a plant-based system for 30-50 years, we are interested in learning about the long-term processes.

Dierberg et al. (2002) studied the effect of submerged as opposed to emergent macrophytes in P-accumulation and removal. The team discusses the ability to draw conclusions from study results carried out on a small area regarding bigger systems, with all the systems installed in true conditions. They tested facilities of about 3 sqm, a fifth of hectare and a 146 hectares system, and found reasonable correlation between the results. Nonetheless, in this article we choose to refer to large-scale systems in real conditions, and for a long term.

Australian researchers Lund, Lavery and Froend (2001) attempted to distinguish between long and short term processes for extracting phosphorus and between soil and plant absorption processes. Their model was found to predict well in a three-year period. They found that the short-term extraction was lower compared to longer processes. In practice, it was found that the level of extraction increases over time and does not decline, as expected in short-term absorption processes.

In a conference that took place in November 2000 in Orlando, Florida, many studies were presented that examined the reduction level of phosphorus in actual systems over time. Thus, even if we ignore the prediction ability of lab studies, there are many studies that show the ability of plant and soil integrated purifying systems in regard to phosphorus.

Many of the studies were carried out in the Everglades area, a huge, million-square kilometer open laboratory with many different research studies conducted in it, costing a mere 650 million dollars. In their study, DeBusk, Dierber and Reddy (2001) review 25 years of research and the results of working facilities to reduce phosphorus in Florida’s everglades. In 2000, there were 20 such systems and the researchers presented the results of the oldest, 35-40-year water hyacinth systems with good results, even in low loads and concentrations of phosphorus. Phosphorus removal reached nearly 52 g P/m2/yr in the systems at Loxahatchee and Melbourne, which also had the highest P loading rates (108 and 354 g/m2/yr, respectively).

Other studies conducted in these sites found that the absorption of the phosphorus was almost exclusively carried out by the soil and the sub-soil plant parts. In the long term, the upper plant parts do not have much importance.

The researchers reviewed two additional large projects: "sampling" facilities of 1540 hectares for treating leachates of agricultural areas in the Everglades, and a purifying system by circulation through plants of Lake Apopka (200 hectares).  The sampling at the Everglades demonstrates the efficiency of the plant's purification in alkaline, high carbonate water with reduction from 1.0 mg/L to 0.01 mg/L phosphorus. This study, carried out for 15 years, showed significant differences between different types of plants, with the numbers varying from 5.5 grams P a year per sqm for the weakest plant, to 6.19 gram P a year per sqm for the most efficient plant. In Lake Apopka, that suffers from similar problems as the Kinneret in Israel, the water is circulated through the wetland with an unintentional variety of plants and the permanent result is 4 grams per sqm per year.

The Norwegian researcher, Braskerud (2002), examined 4 facilities built to reduce phosphorus over a period of 3-7 years. Significant differences were found between the different systems, planted with different species, but in all of the cases there was a significant reduction every year, ranging from 20-44% removal, including during the cold Norwegian winter. It is difficult to compare the systems, due to the significant difference between the phosphorus values at the start, and the rate of flow (i.e. P loading rates). However, it was generally found that the higher the level of the dissolved phosphorus, the more efficient the cost of removal. Importantly, modeling results of three systems indicated that every system should be adapted to its unique conditions.

The study of Comeau et al (2001) focuses on water passing through a fish farming pond in Quebec, Canada. The process is very similar to that in the trout farms on the Dan River in the North of Israel: streaming high-quality water from the stream to the fishpond and returning the water to the stream. The return water has low phosphorus levels, but accumulation of the P-loads can reach quantities that affect water in downstream ponds. The daily flow rate to the fish farm in Quebec producing 100 tons of fish a year is 10,000 cubic meters. In Israel, the flow rate is even higher.

In Canada (as in Israel), a mechanical filter is used at the exit from the farm. In this case, a 60-micron filter that filters 60% of the particulate phosphorus but has hardly any effect on the soluble reactive phosphorus (SRP), which is the main eutrophic contributor. The purification pools were intended to purify the sludge that accumulates in the filter and to "polish" the water extracted from the filter.  The pools, in the size of 200 sqm, reduced 95% of the particulate phosphorus and 80% of the dissolved phosphorus (a total of 92% of the phosphorus) while reaching the objective of 0.03 mg/L.

In Lake Stokkavannet in Norway, tests performed in the late 90s demonstrated low water quality. Since this was the drinking water reservoir of the nearby town, the municipality decided to use a constructed wetland system. The article by Braskerud et al (2005) focuses on the system (2000 sqm) that started operating in 1994, with measurements continuing through 1999-2000. The average detention time was 20 hours. The system operated well even at low temperatures (<5 C) removing 70-85% of P loads.

Returning to the Everglades, Gu et al (2001), examined an area of 2610 hectares receiving water that already passed through a constructed wetland and the phosphorus level in the water dropped to 0.44 mg/L. However, as this level of P can be eutrophic, there is need for another polishing procedure before releasing the water into the nature reserve. The objective of the study was to characterize the optimal depth and delay time and to examine the effect of harvesting the plants.

The findings demonstrated that there was no significance to an extensive detention time in reducing phosphorus (as was found in additional studies). It also demonstrated that there was no significance to great depth, in spite of the increase in the detention time as a result. An interesting finding was that harvest of the plants could be inefficient and even harmful.

Measurements were carried out during 1994-1999. In the first years the reduction was minor, however, it has stabilized with time, increasing from 26% to 70% over the years.

Nungesser and Chimmney (2001) are Florida Water Authority officials, and they examined the results of the  1,545 ha Everglades Nutrient Removal Project (ENRP) in the period 1994-1999. The objective of reducing phosphorus level beneath 0.05 mg/L and reaching a 75% reduction in the entry values was achieved in full, with detention times varying between 5 and 15 days. The researchers attempt to measure the performance of the different areas according to the plants inside them, by analyzing aerial photos. Indeed, area 4, which had submerged plants, as compared to emergent and floating plants in the other cells, was found to demonstrate outstanding results, even though it is located along the flow line and receives low phosphorus values.

Lynch and Newman also focused on the optimal detention time, and found that dissolved phosphorus is not affected at all by the delay time (reduction by 80% in all times) but the particulate phosphorus is affected by detention times.

Dierberg et al (2002) focus on a question that troubled many researchers: the release ability of the phosphorus absorbed to the soil, when conditions change, e.g from reductive to oxidative. The researchers took soil from different areas in the Everglades and caused extreme changes to them (PH, redox, different chemicals, etc.). Only 11-12% of the phosphorus was released.

Therefore, it seems that CW systems can successfully reduce levels of phosphorus, even at high flow rates, with a short detention time, even when used as “polishing” and at low values, which would escape other methods such as flocculation.  It is clear that the specific soil, plants and type of flow should be adapted to this goal.

Even if we do not have full confidence in the causal relation between the amount of phosphorus and cyanobacteria/eutrophic condition of Lake Kinneret, it does not release us from the need to reduce these amounts and to enable the Kinneret,  that has been stable for many years, to continue to function over time. This can be carried out using nature based systems, such as constructed free water wetlands and we must consider this option seriously.


Braskerud, B.C. (2002) Factors affecting phosphorus retention in small constructed wetlands treating agricultural non-point source pollution. Ecological Engineering 19: 41–61.

Braskerud, B.C., Tonderski, K.S., Wedding, B., Bakke, R., Blankenberg, A.-G.B., Ulén, B., and Koskiaho, J. (2005) Can Constructed Wetlands Reduce the Diffuse Phosphorus Loads to Eutrophic Water in Cold Temperate Regions? Journal of Environmental Quality 34: 2145–2155.

Comeau, Y., Brisson, J., Réville, J.-P., Forget, C., and Drizo, A. (2001) Phosphorus removal from trout farm effluents by constructed wetlands. Water Science and Technology 44: 55–60.

DeBusk, T.A., Dierber, F.E., and Reddy, K.R. (2001) The use of macrophyte-based systems for phosphorus removal: an overview of 25 years of research and operational results in Florida. Water Sci Technol 44: 39–46.

Dierberg, F.E., DeBusk, T.A., Jackson, S.D., Chimney, M.J., and Pietro, K. (2002) Submerged aquatic vegetation-based treatment wetlands for removing phosphorus from agricultural runoff: response to hydraulic and nutrient loading. Water Research 36: 1409–1422.

Gu, B., DeBusk, T.A., Dierberg, F.E., Chimney, M.J., Pietro, K.C., and Aziz, T. (2001) Phosphorus removal from Everglades agricultural area runoff by submerged aquatic vegetation/limerock treatment technology: an overview of research. Water Sci Technol 44: 101–108.

Lund, M.A., Lavery, P.S., and Froend, R.F. (2001) Removing filterable reactive phosphorus from highly coloured stormwater using constructed wetlands. Water Sci Technol 44: 85–92.

Newman, J.M. and Lynch, T. (2001) The Everglades Nutrient Removal Project test cells: STA optimization--status of the research at the north site. Water Sci Technol 44: 117–122.

Nungesser, M.K. and Chimney, M.J. (2001) Evaluation of phosphorus retention in a South Florida treatment wetland. Water Sci Technol 44: 109–115.

bottom of page