Water table management impacts on phosphorus loads in tile drainage

https://doi.org/10.1016/j.agwat.2006.12.007Get rights and content

Abstract

Increased phosphorus (P) concentrations, consistently exceeding Quebec's surface water quality standard of 0.03 mg L−1 total P, caused increased P loads in tile drainage from controlled drainage/subirrigation (CD/SI) plots compared to free drainage (FD) plots in a field study carried out in 2005 in Coteau-du-Lac (southwestern Quebec, Canada). This happened even though the total outflow volumes from CD/SI plots were reduced by 27% compared to FD plots. Of the total P concentration, around 96% was in the form of dissolved P under both treatments. The results also showed that CD/SI had no effect on the soil P concentration and P saturation.

During the experiment, the SI water added about 0.84 kg P ha−1, which was negligible compared to the fertilizer input on the field, but represented about 8.5 times the average total P loss from CD/SI plots. However, a laboratory soil column experiment, in which the two drainage treatments were simulated and P-free water was used for SI, also showed increased P concentrations under CD/SI. This confirmed that the increased P loads in tile drainage under CD/SI were most likely caused by an increase in P solubility due to the shallow water table inherent to the water table management system rather than by the addition of P from the SI water.

Introduction

Technological advances made in the 20th century have led to more intensive agriculture with the use of large quantities of nitrogen (N) and phosphorus (P) from chemical fertilizers and manure to maintain productivity (AAFCRB, 2000). The over application of P results in critical soil P saturation levels which lead to a higher risk of P loss to the surrounding environment. In a study conducted in southern Quebec, soil Mehlich-III P concentrations exceeded adequate soil P fertility levels for corn and soybeans at 21 of 27 sites (Beauchemin and Simard, 2000).

Artificial subsurface drainage is commonly used to improve poor drainage capacities of agricultural fields. This practice reduces the risk of waterlogging and soil erosion caused by surface runoff, and allows for earlier field operations in the spring. However, it can also enhance nutrient losses since the volume of water exiting the field is increased and the time during which the nutrients could be sorbed by the soil matrix is reduced. Therefore, despite the low mobility of P in soils, the transfer of P to drainage waters may become significant in soils having high P saturation and equipped with drainage systems (Heckrath et al., 1995, Beauchemin et al., 1998, Sims et al., 1998, Hooda et al., 1999). Generally, most of the P lost through subsurface transport is dissolved P (Heckrath et al., 1995, Heathwaite, 1997). In certain conditions, such as with clay soils with large cracks and macropores that enhance preferential flow, particulate P can constitute a major fraction of P losses in tile drainage (Beauchemin et al., 1998).

Since P is a limiting nutrient in the environment, it is an important contributor to eutrophication and toxic algal blooms in aquatic ecosystems (Pitois et al., 2001). Dissolved P is particularly of concern as it is readily available for plant uptake.

Recent studies have highlighted the agronomic and environmental advantages of implementing water table management systems which can provide both controlled drainage (CD) and subirrigation (SI) through already existing tile drains. Because such systems provide crops with the adequate amount of water throughout the growing season, they can increase crop yields (Mejia et al., 2000). Moreover, they can significantly reduce nitrate (NO3) losses in tile drains, mainly by increasing denitrification rates in the soil profile, which in turn decreases NO3 concentrations in drainage waters.

A previous study conducted in 2001 and 2002 in Coteau-du-Lac (southwestern Quebec) showed that significant amounts of total dissolved P (TDP) were lost through tile drainage under CD/SI, compared to conventional free drainage (FD) (Stämpfli and Madramootoo, 2006). In the same study, subsequent analysis of the well water used for SI showed that TDP concentrations exceeded Quebec's surface water quality standard of 0.03 mg P L−1. Therefore, it was judged possible that a significant amount of P had been added by the SI well water in the subirrigated plots and, thus, may have influenced P concentrations in tile drains under the CD/SI treatment. Another hypothesis for the increased P losses was an increase in P solubility related to shallow water tables. Several studies have shown that the prevalence of soluble P usually increases under low reduction-oxidation potentials (reducing conditions) caused by flooded or anoxic conditions (Ponnamperuma, 1984, Sallade and Sims, 1997, Vadas and Sims, 1998).

Given the benefit of using CD/SI systems on water quality with respect to the reduction of NO3 loss, it is essential to better understand why this drainage practice has previously been shown to increase P losses in drainage water. The goal of this research was to evaluate with a field study the effect of CD/SI versus FD on P loads in tile drainage and on soil P concentrations, P saturation and pH. Moreover, a laboratory soil column experiment was conducted to better understand the results from the field study.

Section snippets

Field description and agronomy

The field experiment was conducted in 2005 on a 4.2-ha field located in Coteau-du-Lac, Quebec (74°11′15″W, 45°21′0″N), in the St. Lawrence lowlands. The soil, of sedimentary origin, is a stone-free Soulanges very fine sandy loam (humic gleysol), and extends to a depth of 0.5–0.9 m. It is underlain by a clay parent material (Lajoie and Stobbes, 1950). The limestone bedrock lies at an estimated depth of 21 m below soil surface (Broughton, 1972). Surface topography is flat, with an average slope of

Climatic data

The 2005 climatic data were compared to the long-term averages between 1971 and 2000 computed by Environment Canada for the Coteau-du-Lac weather station. The station is located less than 500 m from the experimental field. The 2005 growing season was warm, with an average increase of 1.6 °C compared to the long-term averages. Except May, which was on average 1.6 °C colder than the long-term average, all months had mean temperatures above normal. From May to October there was 29% more precipitation

Conclusions

The results from the field study showed that CD/SI had no significant effect on the soil P concentration and P saturation whereas the soil pH was significantly greater in CD/SI plots. Overall, tile drainage was significantly reduced under CD/SI (−27%). However, TP, TDP and DRP concentrations in drainage water from CD/SI plots were on average increased by 131%, 136% and 178%, respectively, compared to FD plots. As a consequence, overall P loads in tile drainage were increased in CD/SI plots.

Acknowledgements

This research was supported by Natural Science and Engineering Research Council of Canada (NSERC). We gratefully acknowledge Kenton Ollivierre for his help on the field and Mr. Guy Vincent, the host farmer, for allowing us to conduct the study on his field.

References (33)

  • A.L. Heathwaite

    Sources and pathways of P loss

  • G. Heckrath et al.

    Phosphorus leaching from soils containing different phosphorus concentrations in the Broadbalk experiment

    J. Environ. Qual.

    (1995)
  • W.H. Hendershot et al.

    Soil reaction and exchangeable acidity

  • P.S. Hooda et al.

    Phosphorus loss in drainflow from intensively managed grassland soils

    J. Environ. Qual.

    (1999)
  • J.W. Kaluli et al.

    Subirrigation systems to minimize nitrate leaching

    J. Irrigat. Drain. Eng.

    (1999)
  • P. Lajoie et al.

    Soil Survey of Soulanges and Vaudreuil Counties in the Province of Quebec. King's Printer and Controller of Stationery

    (1950)
  • Cited by (50)

    • Effect of controlled drainage on nitrogen losses from controlled irrigation paddy fields through subsurface drainage and ammonia volatilization after fertilization

      2019, Agricultural Water Management
      Citation Excerpt :

      The proportions of groundwater levels below the WTC were 12.5%, 0.0%, and 0.0% in 2012 and 25.0%, 50.0%, and 37.5% in 2013 for CWT1, CWT2, and CWT3, respectively. Previous studies observed lower groundwater levels than the targeted levels in CD due to the deep seepage below the plastic curtain (Sanchez Valero et al., 2007). In the current experiment, the groundwater levels in CI paddy fields were mostly higher than the WTC during the first week after fertilization due to the presence of more irrigation water than that in dry farmlands.

    • Dominant glacial landforms of the lower Great Lakes region exhibit different soil phosphorus chemistry and potential risk for phosphorus loss

      2018, Journal of Great Lakes Research
      Citation Excerpt :

      This finding has important implications for designating appropriate water management strategies for these fields. Indeed, fields under controlled drainage and sub-irrigation management systems have reported increased SRP concentrations in surface runoff and subsurface drainage compared with free drainage fields (Tan and Zhang, 2011; Valero et al., 2007). Even in the absence of controlled drainage systems, saturated field conditions often occur across the lower Great Lakes region during the non-growing season (e.g., Lam et al., 2016; Macrae et al., 2010; Van Esbroeck et al., 2017) and may pose a risk of anaerobic FeOx mineral dissolution with associated Pi release during these periods.

    • Nitrate and phosphorus transport through subsurface drains under free and controlled drainage

      2018, Water Research
      Citation Excerpt :

      The non-significant difference in nitrate concentrations was similar to what the majority of other studies have reported (Fausey, 2005; Adeuya et al., 2012; Williams et al., 2015a), although some have observed lower nitrate concentration in CD fields. The increase in SRP was comparable to observations from Valero et al. (2007), Feser et al. (2010), and Tan and Zhang (2011). Nutrient concentrations in the soil are affected by different factors such as water table level, soil organic matter, and temperature.

    View all citing articles on Scopus
    View full text