Nutrient removal by a constructed wetland treating subsurface drainage from grazed dairy pasture

https://doi.org/10.1016/j.agee.2004.05.008Get rights and content

Abstract

Nitrogen and phosphorus budgets over two annual periods are presented for an establishing surface-flow constructed wetland treating subsurface drainage from rain-fed, dairy cattle grazed pasture in the North Island of New Zealand. Drainage flows to the wetland (occupying ∼1% of the catchment area) were highly pulsed, associated with rainfall and soil water status, and differed between years (305 and 197 mm drainage). Flow-proportional sampling of inflow and outflow concentrations were combined with continuous flow records to calculate mass balances for the wetlands. Influent nitrate concentrations were high (median 11 g m−3) in both years, but transient loads of organic N were also an important form of N in the first year. Mass removal of total nitrogen (TN) and its main constituent forms nitrate/nitrite and organic N was recorded for all seasons over both annual periods studied. TN mass removal efficiency of 79% (841 g m−2 per year) in the first year, declined to 21% (40 g m−2 per year) in the second year, associated with changes in the magnitude, speciation and seasonal pattern of N export from the catchment. Ammoniacal N (NH4-N), which comprised <0.5% of TN loadings to the wetland, was generated in small amounts during passage through the wetland in both years. Total phosphorus (TP) in the drainage waters occurred at median concentrations of 0.1–0.2 g m−3, mainly in dissolved reactive forms (DRP 92% by mass). TP export rose by 101% (5.0 g m−2 per year) after passage through the wetland in the first year, but decreased by 12% (0.2 g m−2 per year) in the second year. The results show that constructed wetlands comprising ∼1% of catchment area can markedly reduce N export via pastoral drainage, but may be net sources of NH4-N, DRP and TP during establishment. Performance of the wetland appeared to be affected by both establishment/maturation factors and year-to-year climatic variations. Longer-term studies, supplemented by process-based laboratory and mesocosm investigations, are required to evaluate sustainable nutrient removal rates over a range of climatic conditions, and identify the key factors regulating performance.

Introduction

To reduce off-site impacts of agriculture on the quality and biodiversity of water resources, farmers and regulatory agencies require information on practical, cost-effective tools to manage diffuse pollution of waterways. Interception and filtering of surface and subsurface farm runoff with vegetated buffer strips, riparian zones and wetlands can complement good grazing and cropping, fertilisation, irrigation, drainage and waste management practices to reduce rates of nutrient loses from agricultural lands (Dinnes et al., 2002, Mitsch et al., 2001, Osborne and Kovaic, 1993, Petersen et al., 1992). In intensively grazed pastures with low gradients and slowly permeable soils, much of the runoff and associated contaminants are routed through subsurface and surface drains (Gilliam et al., 1999, Ritter and Shirmohammadi, 2001), which effectively short-circuit shallow groundwater and riparian zone pathways. Alternative design and management of these drainage systems is required to integrate environmental goals with traditional agricultural water and soil management goals. Constructed wetlands offer a potentially cost-effective, practical option for treatment of these numerous small, highly-pulsed sources of diffuse pollutants before they reach streams, rivers, lakes and estuaries.

There is now a substantial database on the performance of wastewater treatment wetlands with relatively constant inflows (IWA, 2000), but this is hard to apply directly to systems receiving less organic-enriched drainage discharges with highly variable flow characteristics. Information on pulse-loaded wetlands receiving urban stormwater flows is developing rapidly, but in these situations most contaminants are strongly associated with particulates (Strecker et al., 1992). Sediments are also the most important form of nutrient export in surface run-off and surface drainage from agricultural lands, particularly cultivated croplands (Gilliam et al., 1999, Ritter and Shirmohammadi, 2001). Most of the information on constructed wetland treatment of agricultural runoff relates to such surface-derived runoff, where the main removal mechanism is settling and retention of sediments and associated nutrients (Baskerud, 2001, Higgins et al., 1993, Jansson et al., 1994, Jordan et al., 1999).

In contrast, subsurface drainage generally decreases export of sediments and associated contaminants by reducing the potential for surface run-off (Armstrong and Garwood, 1991), but increases export of mobile constituents from the soil (Gilliam et al., 1999, Johnes and Burt, 1993, Ritter and Shirmohammadi, 2001, Sims et al., 1998). Nutrient export via subsurface-drainage therefore occurs largely in dissolved forms such as nitrate and orthophosphate, although particulate forms can also be important, particularly in soils that are sandy or exhibit rapid macropore flow (Laubel et al., 1999, Turner and Haygarth, 2000).

Information on wetland treatment of subsurface drainage is limited. Kovaic et al. (2000) undertook a study of three wetlands (each ∼3% of contributing catchment area) treating cropland tile drainage in the upper Embarras River watershed in Illinois (annual rainfall ∼790–1000 mm with 23–35% drainage yield). Overall removal efficiencies were 37% total nitrogen (TN; predominantly NO3) and 22% dissolved reactive phosphorus (DRP), but only 2% for total phosphorus (TP) ranging from −64 to +80% annual TP removal for the individual wetlands. Hunt et al. (1999) reported 37% annual TN removal (0.3 g m−2 per day; predominantly as NO3) for an in-stream wetland occupying ∼0.8% of an agricultural watershed in the Coastal Plains of North Carolina, USA.

Summarising data on wetland NO3 removal from river waters for multi-year studies carried out in six off-stream wetlands at two sites in mid-western USA, Mitsch et al. (2001) reported NO3-N removal rising as a power function from ∼12 to 45 g m−2 per year as loading increased from 20 to 200 g m−2 per year. Using data from 65 surface-flow wetlands, including all those noted above, Kadlec (in press) derived a mean first-order areal removal rate constant (k) of 34 ± 3 m per year for NO3 removal, and a mean Arrhenius temperature coefficient of 1.09. However, the dataset, which included wetlands treating a wide range of NO3 concentrations and loadings, water types (wastewaters, stormwaters, agricultural drainage and river water) and flow regimes, showed a wide range of mean k values (<10 to >60 m per year for wetlands not receiving carbon supplements).

Mitsch et al. (1995) studied P retention in four constructed marshes at Des Plaines, IL, receiving pumped river waters draining agricultural lands that contained 0.11–0.18 g m−3 TP (36–50% dissolved reactive P). Removal of 0.5–3 g TP m−2 per year (53–99% annual removals) was attributed mainly to sediment retention. Both Mitsch et al. (1995) and Moustafa (2000) found wetland P removal performance was well described by a simple empirical Vollenweider-type model in which removal (described by a single rate constant) is proportional to TP concentration in the water column. In further studies at the Des Plaines wetlands involving discrete hydraulic loading events (equivalent to ∼1 nominal detention time) with waters containing ∼0.24 g m−3 TP, Kadlec (1999) used analogous first order k-C* models coupled to dynamic hydraulic models. TP removal averaged 58% over all events, with significantly higher rate constants and background concentrations than found for previous steady-state operation. Initial displacement of antecedent treated water, in addition to removal during the event, significantly influenced wetland performance. Based on data from 83 wetlands in the USEPA North American Database for treatment wetlands, a mean first order plug-flow TP removal rate constant of k = 12 ± 6 m per year was calculated for non-forested surface-flow wetlands (Reddy et al., 1999).

The primary objective of the present study was to quantify mass removal of key forms of nitrogen and phosphorus from grazed pastoral subsurface drainage in a surface-flow constructed wetland system. Wetland treatment performance is likely to change during establishment and maturation, and to vary depending on flow characteristics (hydraulic residence and inter-event times), concentrations and forms of nutrient inputs, vegetation and soil types, and climatic conditions (IWA, 2000). This study evaluates wetland treatment performance, after 1 year’s establishment, for two annual periods. It was carried out on an operational dairy farm within a national dairy industry monitoring catchment in a typical prime dairying region in the North Island of New Zealand. Long-term monitoring of this and other sites with differing environmental and farm management characteristics will be used to assess performance, develop practical design guidelines, and demonstrate the concept to farmers.

Section snippets

Study site and constructed wetland

Constructed wetland treatment of subsurface drainage from grazed (2.9 cows ha−1), fertilised (split application of urea at 150–175 kg N ha−1 per year and superphosphate at 9.2–18.4 kg P ha−1 per year), non-irrigated ryegrass (Lolium perenne) – white clover (Trifolium repens) dairy pastures was studied over two annual periods in the Toenepi catchment at Kiwitahi near Hamilton, Waikato in the North Island of New Zealand (37°44′S, 175°35′E). Installation of the flexible perforated plastic drainage

Soil characteristics

Properties measured in the soils collected from the first and second stages of the wetland are summarised in Table 1. Levels of NO3 in the soil were low compared to NH4+. Soil DEA values were similar to those found in other wetlands that exhibit significant NO3 removal via denitrification (61–90% of added NO3; Burns and Nguyen, 2002, Matheson et al., 2002, Matheson et al., 2003). Under flooded anoxic conditions, and with suitable contact times, soils in the Toenepi wetland are therefore

Conclusions

  • Annual mass removal of TN from subsurface drainage in a surface-flow wetland comprising ∼1% of a grazed pastoral catchment varied from 21 to 79% (52–840 g m−2 per year) for the first 2 years of operation. The wetland generally released more TP and DRP than it received during this period.

  • NO3-N and Org-N were the predominant forms of N exported from the drainage system. Passage through the wetland resulted in reductions of 11–49% in seasonal NO3-N loads to receiving waters. Large pulses of Org-N

Acknowledgements

This study was funded by the New Zealand Dairy Industry through the NZ Dairy Research Institute and Fonterra Research, and by the NZ Foundation for Research Science and Technology (contract C01X0215). Don Tindale undertook field sampling and maintained the site, Margaret Bellingham assisted with data compilation, and the NIWA Inorganic Chemistry Laboratory performed water analyses. We are grateful to Prof. R.H. Kadlec (Wetland Management Services, Chelsea, MI, USA) for providing useful comments

References (70)

  • P.A.M. Bachand et al.

    Denitrification in constructed free-water surface wetlands. II. Effects of vegetation and temperature

    Ecol. Eng

    (2000)
  • B.C. Baskerud

    The influence of vegetation on sedimentation and resuspension of soil particles in small constructed wetlands

    J. Environ. Qual

    (2001)
  • Biggs, B.J.F., 2000. New Zealand periphyton guideline: detecting, monitoring and managing enrichment of streams....
  • Brakensiek, D.L., Osborn, H.B., Rawls, W.R. (Eds.), 1979. Field Manual for Research in Agricultural Hydrology,...
  • Bruce, J.D., 1978. Soil map of part of Raglan county, South Auckland, New Zealand. Bulletin 41, New Zealand Soil...
  • D.A. Burns et al.

    Nitrate movement and removal along a shallow groundwater flow path in a riparian wetland within a sheep-grazed pastoral catchment: results of a tracer study

    N. Z. J. Mar. Freshwater Res

    (2002)
  • M.B. David et al.

    Carbon, plant and temperature control of nitrate removal from wetland mesocosms

    Trans. IL State Acad. Sci

    (1997)
  • D.L. Dinnes et al.

    Nitrogen management strategies to reduce nitrate leaching in tile-drained midwestern soils

    Agron. J

    (2002)
  • Environment Waikato, 1998. Waikato State of the Environment Report 1998. Environment Waikato/Waikato Regional Council,...
  • M.J. Fisherman et al.

    Evaluation of methods for preservation of water samples for nutrient analysis

    Int. J. Environ. Stud

    (1986)
  • Gilliam, J.W., Baker, J.L., Reddy, K.R., 1999. Water quality effects of drainage in humid regions. In: Skaggs, R.W.,...
  • C.W. Hickey et al.

    Toxicity of ammonia to nine native New Zealand freshwater invertebrate species

    Archiv. Environ. Contam. Toxicol

    (1994)
  • Higgins, M.J., Rock, C.A., Bouchard, R., Wengrezynek, B., 1993. Controlling agricultural runoff by use of constructed...
  • P.S. Hooda et al.

    Phosphorus loss in drainflow from intensively managed grassland soils

    J. Environ. Qual

    (1999)
  • P.G. Hunt et al.

    In-stream mitigation of nitrogen contamination in a USA coastal plain stream

    J. Environ. Qual

    (1999)
  • IWA, 2000. Constructed Wetlands for Pollution Ccontrol: Processes, Performance, Design and Operation. International...
  • M.R.A. Jansson et al.

    Wetlands and lakes as nitrogen traps

    Ambio

    (1994)
  • Johnes, P.J., Burt, T.P., 1993. Nitrate in surface waters. In: Burt, T.P., Heathwaite, A.L., Trudgill, S.T. (Eds.),...
  • Jordan, T.E., Whigham, D.F., Hofmockel, K., Gerber, N., 1999. Restored wetlands in crop fields control nutrient runoff....
  • Kadlec, R.H., 1999. Phosphorus dynamics in event driven wetlands. In: Vymazal, J. (Ed.), Proceedings of the Third...
  • Kadlec, R.H. Nitrogen farming for pollution control. J. Environ. Sci. Health, in...
  • KaleidaGraph, 2000. KaleidaGraph, version 5.51. Synergy Software, Reading,...
  • D.A. Kovaic et al.

    Effectiveness of constructed wetlands in reducing nitrogen and phosphorus export from agricultural tile drainage

    J. Environ. Qual

    (2000)
  • A. Laubel et al.

    Subsurface drainage loss of particles and phosphorus from field plot experiments and a tile-drained catchment

    J. Environ. Qual

    (1999)
  • S.F. Ledgard

    Nitrogen inputs and losses from clover/grass pastures grazed by dairy cows, as affected by nitrogen fertilizer

    J. Agric. Sci

    (1999)
  • Cited by (135)

    • Economic impact of predatory piscivorous birds on small-scale aquaculture farms in Kenya

      2019, Aquaculture Reports
      Citation Excerpt :

      Finally, by hosting birds and other wildlife, rural aquaculture farms provide expanded habitat areas for biodiversity, besides possibly serving as refugia for some wetland-specialist species displaced by the loss of their natural wetland habitats (Barrett et al., 2018). Furthermore, as constructed wetlands, many aquaculture farms may absorb excessive loads of nutrients flowing on ground surface; as a result, they may offer vital ecological regulatory services in the landscapes in which they are located (Tanner et al., 2005). For these two reasons, aquaculture farmers may have a justifiable case for deserving economic incentives, from the wildlife and local government authorities, in the form of payments or compensation for important ecosystem services rendered to the environment by aquaculture infrastructure across the landscape (Hishamunda et al., 2014), particularly in light of the economic damage and losses incurred through the depredations of wildlife on pond fish.

    View all citing articles on Scopus
    View full text