Characterising phosphorus loss in surface and subsurface hydrological pathways

https://doi.org/10.1016/S0048-9697(00)00393-4Get rights and content

Abstract

The magnitude and composition of the phosphorus (P) load transported in surface and subsurface hydrological pathways from a grassland catchment depends on the discharge capacity of the flow route and the frequency with which the pathway operates. Surface runoff is an important pathway for P loss, but this pathway is spatially limited and temporarily confined to high magnitude, high intensity rainfall events. High P concentrations (mean: 1.1 mg TP l−1) were recorded, with most P transported in the dissolved fraction. Preferential flow pathways, particularly soil macropores and field drains, are important contributors to the overall P load; most P is transported in the particulate fraction and associated with organic or colloidal P forms. High P concentrations (mean: 1.2 mg TP l−1) were recorded in macropore flow in the upper 0–15 cm of a grassland soil, and generally declined with increasing soil depth. On average, P concentrations in drainflow were over six times greater in stormflow compared to baseflow. Stormflow P losses in drainflow were predominantly in the particulate fraction; significant correlation (P<0.01) was recorded with suspended sediment concentrations in drainflow. Phosphorus concentrations in groundwater were low (<0.2 mg TP l−1 at 150 cm), although this pathway may contribute to stream flow for the majority of the year.

Introduction

With stringent water quality standards being set for many large point source nutrient discharges in response to European legislation, the relative contribution of non-point sources (NPS) to deteriorating water quality is increasing (Foy and Withers, 1995). Nutrients derived from agricultural NPS are currently viewed as the main cause of eutrophication in many European rivers and lakes (Heathwaite et al., 1996). Non-point source pollution from agricultural land can be particularly detrimental to water quality because discharges are untreated, often contain high nutrient and organic matter loads, and losses occur sporadically during intense rainfall events from sources that are difficult to identify, quantify and control.

Surface waters are highly sensitive to P loss from agriculture because critical concentrations for eutrophication control (10–20 μg P l−1) are an order of magnitude lower than soil P concentrations required for crop growth (200–300 μg P l−1). Thus P loss from agricultural land may accelerate the eutrophication of P-sensitive waters (OECD, 1982). Symptomatic changes include the proliferation of potentially toxic algal blooms, the death of invertebrates and fish due to de-oxygenation and the long-term loss of biodiversity.

The national balance sheet for P in UK agriculture shows a net P surplus of 10 kg P ha−1 year−1 (Withers, 1996). Whilst this surplus cannot be accounted for directly by increased P inputs to land in fertilisers and manures because these have altered little over the past 25 years, changes in land management have enhanced the potential for P transport. Particularly important are a greater proportion of arable land sown to winter cereals, enhancing P loss linked to soil erosion; more land drainage, enhancing subsurface P losses and increasing the likelihood of winter grazing; and increased livestock densities which result in the enrichment of surface soils and disposal problems for P-rich organic waste (Heathwaite et al., 2000).

Given these changes in agricultural land management, two key factors govern the amount and forms of P lost from agricultural land to receiving waters. First, soil biochemical processes, which control the forms of soil P available for transport, and second, hillslope hydrology, which defines the mechanisms and pathways of loss. Heathwaite (1997) showed the risk of P loss from agricultural land is greater than previously thought, with both surface and subsurface (especially preferential flow) hillslope hydrological pathways being important in P transport. However, one of the problems in understanding hydrological pathways of nutrient loss from land to stream is their dynamic nature, both in time and space. A hillslope may generate only subsurface flow during a gentle rainstorm; infiltration-excess surface runoff during a deluge; or subsurface flow alone during a short rainstorm and saturation-excess runoff during a long one (Dunne, 1983). Fortunately, sophisticated field monitoring has improved our understanding of the hydrological pathways of water movement from land to stream (e.g. Brammer and McDonnell, 1996). However, such advances have yet to be applied to an understanding of nutrient transport pathways from agricultural land. Thus for P, the assumption that most loss from agricultural land occurs via surface runoff in association with soil particles is still widely held (Boardman, 1990, Sharpley et al., 1994). Whilst this may be the case for cultivated land with soils that are susceptible to erosion (e.g. Gburek et al., 1996, Fraser et al., 1999), a growing body of evidence suggests that other pathways of P loss may make an important contribution to the stream P load (e.g. Dils and Heathwaite, 1996, Heathwaite et al., 2000).

Currently, little is known about P loss in individual flow pathways despite these routes forming the critical link between P sources and P outputs measured in streamflow. Thus, whilst recent P research has significantly improved our knowledge of edge-of-field P losses, there are few attempts to translate such losses into actual inputs to receiving waters (Heathwaite and Sharpley, 1999). The magnitude and composition of the P load transported in individual flow pathways will depend on: (i) land management practices (e.g. grazing intensity and fertiliser application rate), which influence the total amount, spatial heterogeneity and form of soil P; (ii) rainfall characteristics and hillslope hydrology, which define the mechanisms and pathways of water flow; and (iii) soil type and antecedent moisture status, which determines the contact time between ‘new’ water and the soil, and the frequency with which the pathway operates.

We propose that different hydrological pathways of P loss are characterised by distinct P ‘signatures’ (forms and concentrations). As P fractions vary in their bioavailability, the relative contribution from discrete surface and subsurface hydrological pathways will influence the overall P composition in agricultural runoff and its impact on receiving waters. We focus our research on managed grasslands which, despite covering 70% of the UK land area (Waters, 1994) have received little attention in terms of P loss relative to arable land. This may be because inorganic fertiliser inputs to grassland are low (10–40 kg P ha−1 year−1) and erosion risk is minimal (McGrath and Loveland, 1992). However, recent work by Haygarth and Jarvis (1997) recorded high TP loss (1.77 kg TP ha−1 year−1 for surface and 0.38 kg TP ha−1 year−1 for subsurface pathways) even where the soil P status was low (Olsen P approx. 5 mg l−1).

Section snippets

Site

The 120-ha Pistern Hill catchment (NGR: SK 352 197) is situated near Ashby-de-la-Zouch, Leicestershire, UK (Fig. 1). Catchment geology is dominated by glacial till with Triassic marl and Triassic sandstone. At depth, there are naturally occurring phosphatic nodules embedded in marine clay bands (T. Harrod, personal communication). Catchment soils are a complex patchwork of 10 different soil series dominated by the Hodnet and Salop series (slightly stony sandy loam or clay loam over moderately

Field site details

The field set-up used enabled an integrated approach to the measurement of P loss in the key hydrological pathways: surface, subsurface (matrix and macropore) and drainflow. We installed nests of field instruments in two parallel hillslope transects, both 150 m in length and 10 m apart. The nests were positioned at top-, mid- and base-slope; the layout is shown in Fig. 1. Each sampling nest consisted of a surface runoff trough (5.5 m2 collection area), three zero-tension macropore samplers

Surface pathways of P transport

Table 1 summarises the P fractionation recorded in surface runoff for the period September 1994 to February 1996. Approximately 62% TP loss in surface runoff is recorded in dissolved form, primarily as DIP (70%). The large concentration ranges and standard errors associated with the surface runoff data indicate a high degree of variability in surface runoff losses, both in time and space. This is primarily a reflection of the effect of rainfall duration and intensity on the magnitude of surface

Phosphorus export in surface runoff from agricultural land

The mean TP concentration in surface runoff (1140 μg P l−1, Table 1) for our study is considerably higher than that reported for grassland catchments by other researchers (e.g. McColl et al., 1977, Sharpley and Syers, 1979, Haygarth and Jarvis, 1996). A number of factors may account for this difference, and include variation in soil total P, hydrologically effective rainfall, erosion potential, farm type and land use, including fertiliser and manure amendments, the presence or absence of

Conclusions

The trends recorded at the plot- and hillslope-scale reported in this paper do not reveal the extent of spatial or temporal variations at the field or catchment scales, and must, therefore, be interpreted with caution, especially in relation to land management advice. Future research is developing approaches for integrating different scales of research from plots through to catchments.

Of equal importance, in terms of understanding and managing P transport, is an evaluation of the thresholds of

Acknowledgements

This work was funded by MAFF studentship AE8750 to RD and undertaken at the Department of Geography, University of Sheffield. We acknowledge ADAS for weir installation at the field site and Mr John Smith, the host farmer for allowing us access to his land.

References (52)

  • A. Breeuwsma et al.

    Phosphate-saturated soils: a ‘new’ environmental issue

  • O. Broberg et al.

    Particulate and dissolved phosphorus forms in freshwater: composition and analysis

    Hydrobiologia

    (1988)
  • Chambers, B.J., Smith, K., 1998. Soil fertility benefits from farm manures. In: Organic amendments to soils....
  • J.L.B. Culley et al.

    Suspended soils and phosphorus loads from a clay soil: I. Plot studies

    J. Environ. Qual.

    (1983)
  • Dils, R.M., 1997. Phosphorus fractionation in hillslope hydrological pathways contributing to agricultural runoff....
  • R.M. Dils et al.

    Phosphorus fractionation in hillslope hydrological pathways contributing to agricultural runoff

  • R.M. Dils et al.

    Tracing phosphorus movement through agricultural soils

  • R.H. Foy et al.

    The contribution of agricultural phosphorus to eutrophication

    Proc. Fert. Soc.

    (1995)
  • A.I. Fraser et al.

    The effect of rainfall on soil erosion and particulate phosphorus transfer from arable soils

    Water Sci. Technol.

    (1999)
  • W.J. Gburek et al.

    Identification of critical source areas for phosphorus export from agricultural catchments

  • Harris, R., Heathwaite, A.L., Haygarth, P.M., 1995. High temporal resolution sampling of P exported from grassland soil...
  • P.M Haygarth et al.

    Pathways and forms of phosphorus losses from grazed grassland hillslopes

  • P.M. Haygarth et al.

    Phosphorus cycling through grassland systems

    Soil Use Manage.

    (1996)
  • P.M. Haygarth et al.

    Size distribution of colloidal molybdate reactive phosphorus in river waters and soil solution

    Water Res.

    (1997)
  • P.M. Haygarth et al.

    Form of phosphorus transfer and hydrological pathways from soil under grazed grassland

    Eur. J. Soil Sci.

    (1998)
  • P.M. Haygarth et al.

    Hydrological factors for phosphorus transfer from agricultural soils

    Adv. Agron.

    (2000)
  • Cited by (335)

    • The Phosphorus Cycle

      2023, Wetzel's Limnology: Lake and River Ecosystems, Fourth Edition
    • Analysis of reactive phosphorus treatment by filter materials at the edge of tile-drained agricultural catchments: A global view of the current status and challenges

      2022, Journal of Environmental Management
      Citation Excerpt :

      Agriculture has long been considered the main source of diffuse P losses to surface waters due to continued application of fertilizers and manure, which commonly exceeds crop requirements and results in P accumulation prone to leaching (Penn et al., 2016). Agricultural P losses vary largely in time and space (25–75%) and occur through surface and subsurface runoff (Withers and Haygarth, 2007; Sharpley et al., 2011; Heathwaite and Dils, 2000). The losses in surface runoff can be minimized through management practices that prevent soil erosion (Sharpley et al., 2001).

    View all citing articles on Scopus
    View full text