- A few studies indicate seepage wetlands remove between 75 and 98 percent of nitrate from water.
- Most N is removed via denitrification and uptake by wetland plants. Denitrification is reliant on shallow horizontal seepage (mixing between surface and upper wetland soils) and represents permanent N loss from water, whereas uptake by plants represents temporary storage until the plants die.
- Smooth, low-cover vegetation works best to promote filtration and prevent flow channels from forming. Channels reduce contact time between water and soil, reducing the effectiveness of denitrification.
- Livestock should be excluded from shallow (not just deep) wetlands, as they are more likely to enter these areas and cause damage to the soil. This may reduce wetland effectiveness (especially through soil compaction). Light grazing when the wetland soil is dry promotes smooth vegetative cover without degrading the soil.
Forms of nitrogen
N occurs in water in several different forms, including nitrate (NO3), nitrite (NO2) and ammonium (NH4) ions; dissolved inorganic N (DIN = NO3 + NO2 + NH4); dissolved organic N (DON); particulate organic N (PON); and total N (TN = DIN + DON + PON). Some forms of N are bioavailable (notably DIN) and can induce excessive growths of algal slime and weeds in streams, and algal blooms in lakes. DON and PON are less bioavailable but can be broken down by bacteria, fungi and sunlight into bioavailable forms.
What are seepage wetlands?
Occurring naturally along stream banks or at the heads of streams, seepage wetlands are characterised by water-tolerant plants; saturated, organically-enriched, anaerobic soils; and standing water. They are generally disliked by farmers because of the risk of livestock and vehicles getting stuck and because they're unproductive. However, these areas are useful sinks for nitrogen (N), phosphorus (P), sediment and pathogens washed off paddocks.
Seepage wetlands are mainly fed by subsurface water flow from springs that emerge from a single point, or by seepage emerging from the ground along a line or surface without a distinct origin. Their degree of saturation ranges from temporary dryness to permanent saturation with standing water.
Seepage wetlands typically have three layers: a dense mat of plant roots (generally native grasses, rushes, sedges and raupo1) at the top, sitting over a porous, anaerobic, saturated organic soil, which lies on top of a less permeable soil layer, such as clay. They are located at the change of slope where particulate solids, including mineral sediments and organic matter, accumulate. Being small (10 to 5000 square metres), seepage wetlands are rarely identified in regional wetland inventories or managed any differently from surrounding pasture. However, one study found seepage wetlands that covered only five percent of a catchment area2 intercepted more than 20 percent of runoff3.
Constructed (man-made) wetlands attempt to mimic seepage wetlands and optimise contaminant trapping and removal by forcing water to pass through shallow flooded beds of emergent aquatic plants such as raupo.
How do wetlands work?
Wetland soils are typically saturated, have a high organic content and are anaerobic. Such conditions favour denitrification – the reduction of nitrate (NO3) to gaseous forms – which permanently removes nitrogen (N) from runoff. Several studies have measured high potential denitrification rates (quantified as the denitrification enzyme activity: DEA) in wetland soils. Two studies have measured high actual (in situ) denitrification rates from wetlands4. However, denitrification does not explain all NO3 removal, implying that some is transformed to ammonium (NH4), dissolved organic nitrogen (DON) and/or particulate organic nitrogen (PON).
Water travels through wetlands in a number of ways (Figure 1). Generally, more water travels across the top of a wetland than seeps through the microbially-active soils. High NO3 removal (25 percent of added NO3 removed over 1.5 metres) has been measured from surface flow during dry weather, although removal was less effective during rainfall5.
Surface water can also mix vertically into the top 5 to 10cm of the porous wetland soils, bringing NO3 into contact with the denitrifying bacteria, which is where most NO3 removal takes place (Figure 1). While soils 15 to 20 cm deep have high DEA (i.e. the potential to remove NO3), porosity decreases with depth and reduces the vertical mixing. Consequently, NO3 concentrations and removal rates decrease in soil depth below about 15 to 20cm5.
NO3 concentrations and denitrification rates also decrease as water moves from the top to the bottom of a wetland. NO3 concentrations are high where water first enters and encounters soils with high DEA, but low near the outlet stream, reflecting that most NO3 has been removed as the water passes through4. Regardless of where in the wetland most denitrification occurs, maximising the contact between inflowing water and wetland soils increases NO3 removal rates through denitrification.
How effective are they at removing nitrogen?
So far, only a handful of reliable wetland studies have been undertaken in New Zealand. That is because seepage wetlands are challenging to study. Their water inputs are spread out, their soils are unconsolidated, and they are home to complex biogeochemical activities. Furthermore, key processes such as denitrification are difficult to measure.
However, a recent review of New Zealand seepage wetland studies found that, in comparisons of in-flow versus out-flow, all studies reduced NO3 by 75 to 98 percent4. This was true regardless of the methods used and whether concentrations or loads were compared.
Studies of constructed wetlands are simpler to undertake because they have more defined in-flow and out-flow paths. The studies have quantified the uptake of nutrients (including N) by plants for growth, and the generation of organic carbon (which promote denitrification) when they deteriorate. In-flowing N removal from constructed wetlands in New Zealand has been found to range from 65 to 92 percent, and N removal increases linearly with plant biomass6. Plants almost certainly remove N from natural seepage wetlands, although rates attributable to plant uptake have not been quantified.
Wetlands are highly effective at removing N over the course of a year (i.e. on average). However, they perform best during low summer flows and are usually net sinks of all forms of nitrogen during this time. Conveniently, this occurs when nitrogen poses the highest risk, because it causes excessive plant and algal growth in receiving streams, rivers, lakes and estuaries. During other times of the year, wetlands may vary in performance. One study found wetlands can be net sources (putting out more than goes in) of some forms of N (NH4, DON and/or PON) at times of higher flows4. But most other studies have found wetlands were sinks for total N (the sum of all organic and inorganic forms of N found in a water sample) even during times high flows4,7.
How does a farmer assess wetlands’ effectiveness?
The most common method for farmers to assess a wetland’s effectiveness is by using Overseer, although this nutrientmodelling software does come with some limitations.
Overseer assumes an average rate of 250 milligrams per square metre per day (mgm-2d-1) (at 20°C), which is adjusted by wetland condition and temperature. Compared with four studies with measured removal rates4,5, Overseer predicted only 36 to 67 percent of the measured NO3 removal rates, indicating Overseer predictions were underestimated. It is not clear, however, whether Overseer is conservative for estimating the removal of all forms of bioavailable N. This is partly because we have an incomplete understanding of the bioavailability of organic nitrogen exports from wetlands. In addition, it is not clear what proportion of DON and PON losses from farmland is included in Overseer losses.
User inputs also affect the Overseer estimation. Users must enter data into Overseer, including inflows and the condition of wetland soils and vegetation. Although look-up tables are provided, users report difficulties providing input data objectively. NIWA is working with Overseer to make the process simpler and, consequently, more widely-accepted in nutrient budgets.
Even so, Overseer allows farmers to assess the potential of seepage wetlands to reduce N loss from farms, and to see how removal varies with wetland/catchment area ratio, flow channelisation, vegetation and stock damage.
How are seepage wetlands best managed?
When cattle enter wetlands, they can degrade the water quality directly through faecal and urine inputs, and soil disturbance; and indirectly by altering soil physical properties (e.g. compaction) and damaging vegetation. High total and organic N exports8 have been measured from small, shallow seepage wetlands after livestock incursions.
Currently, farmers use fencing to prevent cattle becoming trapped in deep wetland soils. However, cattle tend to avoid deeper wetlands7, so shallow wetlands (standing water less than one metre deep) are likely to benefit more from livestock exclusion. Unlimited cattle access to unfenced wetlands deeper than two metres caused no observable impact on water quality. This was mainly due to the lack of wetland ingress by the cattle and the ability of the dense wetland grasses to capture and remove (attenuate) contaminants from the water. This deep wetland was also very effective at attenuating particulate matter and associated nutrients from the steep adjacent hillslopes7.
Portable electric fences offer a flexible, effective way to exclude cattle from wetlands that could be grazed lightly during periods when the wetland soils have dried out. Light grazing is beneficial as it maintains smooth low cover vegetation, preventing channels from forming. We do not recommend bulldozing benches around wetlands for permanent fencing, because bare earth is vulnerable to erosion and benches may divert runoff away from wetlands.
Planting seepage wetlands with large vegetation (e.g. flax, shrubs, trees) is not advised as these are not as effective as smaller wetland plants, although some large plants may help protect wetlands from ‘washout’ during storms. Plants provide organic matter, promote denitrification and trapping solids, but larger plants encourage flow channels to form. Channels, like drains, reduce the contact time between water and the soils where denitrification and plant uptake occur.
- Clarkson, B. R., P. D. Champion, B. D. Rance, P. N. Johnson, K. A. Bodmin, L. Forester, P. Gerbeaux, and P. N. Reeves. 2013. Wetland indicator status ratings for New Zealand species. www.landcareresearch.co.nz/__data/assets/ pdf_file/0014/64400/wetland_rating_species_December_2013.pdf (accessed June 2018).
- McKergow, L. A., J. C. Gallant, and T. I. Dowling. 2007. Modelling wetland extent using terrain indices, Lake Taupo, NZ. Proceedings of MODSIM 2007 International Congress on Modelling and Simulation, Christchurch, New Zealand, pp. 74-80.
- Rutherford, J. C., D. Schroer, and G. Timpany. 2009. How much runoff do riparian wetlands affect? New Zealand Journal of Marine and Freshwater Research 43: 1079-1094.
- Rutherford, J. C. 2017. Review of Nitrogen Attenuation in New Zealand Seepage Wetlands. NIWA Client Report 2017241HN.
- Rutherford, J. C., and M. L. Nguyen. 2004. Nitrate removal in riparian wetlands: Interactions between surface flow and soils. Journal of Environmental Quality 33(3): 1133-1143.
- Tanner, C. C. 1996. Plants for constructed wetland treatment systems — A comparison of the growth and nutrient uptake of eight emergent species. Ecological Engineering 7(1): 59-83.
- Hughes, A. O., C. C. Tanner, L. A. McKergow, and J. P. S. Sukias. 2016. Unrestricted dairy cattle grazing of a pastoral headwater wetland and its effect on water quality. Agricultural Water Management 165: 72-81.
- McKergow, L. A., J. C. Rutherford, and G. C. Timpany. 2012. Livestock- Generated Nitrogen Exports from a Pastoral Wetland. Journal of Environmental Quality 41(5): 1681-1689.
This article was originally published in Technical Series September 2018