HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Watson, C. J. Articles by Matthews, D. I. Search for Related Content PubMed PubMed Citation Articles by Watson, C. J. Articles by Matthews, D. I. Agricola Articles by Watson, C. J. Articles by Matthews, D. I. Related Collections Surface Water Quality Phosphorus Field-Scale Studies

ENVIRONMENTAL ISSUES

Increase in Phosphorus Losses from Grassland in Response to Olsen-P Accumulation

^a Agriculture, Food and Environmental Science Division
^b Biometrics Branch, Agri-Food and Biosciences Institute, Newforge Lane, Belfast BT9 5PX, United Kingdom

^* Corresponding author (Catherine.Watson{at}afbini.gov.uk).

Received for publication May 26, 2006.
ABSTRACT

The Olsen-P status of grazed grassland (Lolium perenne L.) swards in Northern Ireland was increased over a 5-yr period (March 2000 to February 2005) by applying different rates of P fertilizer (0, 10, 20, 40, or 80 kg P ha^–1 yr^–1) to assess the relationship between soil P status and P losses in land drainage water and overland flow. Plots (0.2 ha) were hydrologically isolated and artificially drained to v-notch weirs, with flow proportional monitoring of drainage water and overland flow. Annually, the collectors for overland flow intercepted between 11 and 35% of the surplus rainfall. Single flow events accounted for up to 52% of the annual dissolved reactive phosphorus (DRP) load. The Olsen-P status of the soil influenced DRP and total phosphorus (TP) concentrations in land drainage water and overland flow. Annual TP loss was highly variable and ranged from 0.19 to 1.55 kg P ha^–1 yr^–1 for the plot receiving no P fertilizer and from 0.35 to 2.94 kg P ha^–1 yr^–1 for the plot receiving 80 kg P ha^–1 yr^–1. Despite the Olsen-P status in the soils ranging from 22 to 99 mg P kg^–1, after 5 yr of fertilizer P applications it was difficult to identify a clear Olsen-P concentration at which P losses increased. Any relationship was confounded by annual variability of hydrologic events and flows and by hydrologic differences between plots. Withholding P fertilizer for over 5 yr was not long enough to lower P losses or to have an adverse effect on herbage P concentrations.

Abbreviations: AFWM, annual flow-weighted mean • AE, actual evaporation • DRP, dissolved reactive phosphorus • DUP, dissolved unreactive phosphorus • PP, particulate phosphorus • TDP, total dissolved phosphorus • TP, total phosphorus

DIFFUSE losses of phosphorus from agricultural land to surface water have been increasing in recent years, leading to increased eutrophication of many rivers, lakes, and impounding reservoirs worldwide (Foy and Bailey-Watts, 1998; Sharpley and Rekolainen, 1997). Evidence suggests that the increased loss of agricultural P is due to farms operating an annual P surplus of approximately 15 to 20 kg P ha^–1 yr^–1, and, as a result, their soils have become enriched with P (Edwards and Withers, 1998). In Northern Ireland, Foy et al. (1995) estimated that the rate of accumulation in soils in the Lough Neagh catchment over the last 50 yr was 10 kg total P ha^–1 yr^–1. Monitoring the six major rivers in this predominantly grassland catchment showed that the rates at which the dissolved reactive phosphorus (DRP) loadings were increasing were in the range of 13.0 to 14.5 g P ha^–1 yr^–1 over a 17-yr period (Foy et al., 1995). This increase is much greater than could be accounted for by changes in rural human population densities or industrial sources and was sufficient to counterbalance the benefits of a P removal program operating at sewage treatment works. Similar increases in P loads were found over a 28-yr period in Scandinavia by Krug (1993) and over a 15-yr period in Finland by Rekolainen (1989). These increases were postulated to be due to increases in surface and subsurface losses of P resulting from P accumulation in soils originating from an imbalance between agricultural inputs and outputs. It is estimated that the agricultural contribution to total P losses to inland waters in Northern Ireland is 58% and that the annual economic costs of eutrophication are in the range of 36 to £61 million (Smith et al., 2005). The target of the European Union's Water Framework Directive (European Commission, 2000) is to restore all surface water to good ecologic and chemical status by 2015.

Evidence that increases in DRP concentrations in land drainage water were a direct response to soil P accumulation has been shown by Smith et al. (2003) in a temporal study on a 9-ha grassland catchment at Greenmount Agricultural and Horticultural College, County Antrim, Northern Ireland, that had no domestic or industrial sources of P and no farm buildings. They showed that for the period 1989 to 1997, annual median DRP concentrations in land drainage water increased by 2.4 µg P L^–1 yr^–1 in response to a mean P surplus of 23.4 kg P ha^–1 yr^–1.

The question arises as to what soil P status is optimum to minimize environmental losses of P to surface waters while maintaining grass yields. Several dose-response curves have been produced where agricultural P losses are plotted against their soil P status. A breakpoint for DRP losses to drainage water of around 60 mg Olsen-P kg^–1 was shown by Heckrath et al. (1995) from soils of the Broadbalk continuous wheat experiment at Rothamsted. However, Hesketh and Brookes (2000) showed that the breakpoint measured after laboratory extraction of different soils with CaCl₂ ranged from 10 to 119 mg Olsen-P kg^–1. They were unable to find any causal relationships between differences in soil chemical and physical properties or management and the different breakpoints. Knowledge of the relationship between P losses to water and soil P status, together with quantifying the contribution of overland flow to total P loss, is fundamental to developing a sustainable P management policy for agriculture. The aim of the current study was to increase the Olsen-P status of grazed grassland swards over a 5-yr period by applying P fertilizer and to assess the relationship between soil P status and P losses in overland flow and land drainage water. Particular emphasis is given to the DRP fraction because of its high bioavailability (Watson and Foy, 2001).

Materials and Methods

Study Site
The grassland experimental site was located at the Agricultural Research Institute of Northern Ireland at Hillsborough, County Down on a slightly gleyed, sandy clay-loam soil (48% sand, 31% silt, 21% clay, and 12% organic matter) overlying Silurian shale. Five plots, each 143 m x 14 m (0.2 ha), were laid side by side on the northerly aspect of a sloping drumlin. The plots were hydrologically isolated and artificially drained to separate v-notch weirs, which monitored flows in the range of 0.01 to 7.0 L s^–1 to an accuracy of ±4%. Details of the installation of the drains have been described elsewhere (Watson et al., 2000). To prevent cross-plot flows of surface water, residual soil-spoil was placed on both sides of the drain trench. Interplot fences were erected on these spoil lines. The drains were installed during the spring and summer of 1987, and the pasture was plowed and reseeded to perennial ryegrass in August 1987. From March 1989 to February 1999, the plots received different rates of N fertilizer (0–500 kg N ha^–1 yr^–1) as ammonium nitrate–calcium carbonate (Watson et al., 2000) and were continuously grazed by beef steers from April to October of each year. The same, single annual application of P (as single superphosphate) and K (as muriate of potash) was applied to all plots, according to soil analysis and standard recommendations for grazing management (Ministry of Agriculture, Fisheries, and Food, 1994).

Ground limestone (54% neutralizing value) was surface applied to all plots in April 1999 at a rate of 5 t ha^–1 to increase soil pH from 5.8 to 6.2. All plots received 250 kg N ha^–1, 19 kg K ha^–1, and 18 kg S ha^–1 during 1999 and were under the same grazing management as in previous years. The N fertilizer was applied in six equal applications from March to August, the K fertilizer was applied as a single dressing in June, and the S fertilizer was applied with the N source on three occasions. No P fertilizer was applied during 1999. Because all plots received the same fertilizer applications during 1999, this acted as a break year from the previous 10-yr study, using different N fertilizer rates. There was no evidence that previous N application rates had any effect on P losses to drainage water (Watson, unpublished). The experimental P treatments were imposed from March 2000 to February 2005. The plots received 0, 10, 20, 40, or 80 kg P ha^–1 as triple superphosphate (46% P₂O₅) in six equal applications between March and September each year. All plots received the same inputs of N (250 kg N ha^–1 yr^–1 applied with the P fertilizer), K (20 kg K ha^–1 yr^–1), and S (18 kg S ha^–1 yr^–1) and were continuously grazed by beef steers (7–12 mo old) from April to October to maintain a constant sward height of 7 cm. Output was estimated from live-weight gain of animals (measured every 20 d) allocated to each plot randomly. For most of the grazing season, the same two cattle grazed each plot, but the stocking rate was varied depending on growing conditions. The average number of grazing days for the 5-yr study period was 1718 d ha^–1 yr^–1 (or 344 d yr^–1 per plot), and the average animal live-weight gain was 1226 kg ha^–1 yr^–1.

Sampling and Analysis
Drainage flow through each weir was monitored continuously and logged at 15-min intervals using Warren Jones WJ 460 FM flow monitors (Warren Jones Engineering, Bicester, Oxon, UK) fitted with ultrasonic probes capable of measuring water depths in the weirs to ±1 mm. Warren Jones Spartan-2 Magnum refrigerated autosamplers collected a daily composite water sample from each plot, comprising four 100-mL samples taken at 6-h intervals. The daily water samples were bulked to give a weekly composite on which various P fractions were analyzed.

A collector for overland flow was installed in a shallow trench across the width of each plot at the lowest end and consisted of a 185-mm diameter solid PVC pipe cut in half with a 45-mm flange attached to the top edge of one side. This was pushed horizontally into the soil to an approximate depth of 50 mm. The collector was connected to a mini v-notch weir (152 mm) and portable 6700 series ISCO water sampler (ISCO, Inc., Lincoln, NE) with a bubbler module. Flow data were recorded using ISCO's Flowlink software and downloaded using a 581 Rapid Transfer Device. Each collector for overland flow was fitted with an easily removable gray PVC lid (200 mm wide), which, when closed, rested on top of the grass stubble on the collection side. This prevented contamination of the pipe with bird droppings or debris but allowed overland flow (including shallow interflow to 50 mm) to enter the pipe. When overland flow occurred, 24 x 200 mL water samples were taken at 20-min intervals, and a composite sample was produced for the event. Phosphorus loads were calculated from recorded flows and measured P concentrations for each event. The overland flow collectors were protected from trampling by cattle with an electrified fence placed across the width of the plot approximately 0.3 m up from the collector pipe. The vegetation immediately around the collectors and water samplers was cut on a regular basis to a height of approximately 50 mm and removed.

Dissolved reactive P (DRP) was determined on a 0.45-µm, membrane-filtered (Millipore, UK) water samples by the acidic molybdate-ascorbic acid method of Murphy and Riley (1962). Total dissolved P (TDP) and total P (TP) were determined on filtered and unfiltered samples, respectively, by digestion with potassium persulphate and sulfuric acid, followed by analysis of the digest as for DRP (Eisenreich et al., 1975). Dissolved unreactive phosphorus (DUP = TDP – DRP) and particulate phosphorus (PP = TP – TDP) fractions were calculated. Annual flow-weighted mean (AFWM) concentrations were calculated for each P species by Eq. [1]:

[1]

where w is week number (1–52), f_w is the weekly flow through the weir, and c_w is the concentration of the determinant in the weekly composite sample collected from each weir. Each experimental year ran from 1 March to 28 or 29 February of the following year.

Daily rainfall and actual evaporation (AE) data were obtained from the Central Climate Unit, Meteorological office, Exeter, Devon, UK using the Meteorological Office Rainfall and Evaporation Calculation System (MORECS) for Hillsborough. Daily rainfall was manually recorded at the site using a tipping bucket collector and was not significantly different from the MORECS data. Daily rainfall and AE values were summed to give monthly and annual total values. The efficiency with which the drains intercepted flow at the start of the 10-yr experimental period varied considerably (range, 72–103%) (Watson et al., 2000). For this reason, annual loads of P lost to drainage water were determined using the theoretical drainage volume calculated from meteorologic data (i.e., AFWM concentration was multiplied by annual rainfall minus AE). It was assumed that the volume intercepted on each plot was a representative sample of soil drainage water.

For soil sampling, each plot was divided into three sections (top, middle, and bottom), running down the slope. On a weekly basis, from January 2000 to February 2005, 30 cores (15 mm diameter x 75 mm depth) were randomly taken from each section and bulked. Soil samples were air dried at 30°C, ground to pass a 2-mm sieve, and analyzed for bicarbonate-extractable inorganic P (Olsen-P) according to the method of Olsen et al. (1954) using a soil to Olsen reagent ratio of 1:20. A linear model (Eq. [2]) was fitted to the weekly Olsen-P data over the 5 yr of the study:

[2]

where Y is the estimated P status in mg kg^–1, P_fert is the rate of P applied as kg P ha^–1 yr^–1, and Date is the date of the observation in Julian format. To allow for random variation in the observed values, the annual increase was derived by calculating the estimated Olsen-P value at the start and the end of the fertilizer year and subtracting the two values.

Grass clippings (>5 cm) were taken at approximately 2- to 3-wk intervals, depending on growing conditions, from March/April to September/October each year. The grass samples were dried at 102°C for 18 h and milled to pass through a 1-mm mesh. Samples were analyzed for N by dry combustion using a Carlo Erba NA1500 elemental analyzer and for P, K, Ca, Mg, and S by inductively coupled plasma spectroscopy after digestion in a (4:1) mixture of nitric/perchloric acid (Ministry of Agriculture, Fisheries, and Food, 1986).

Annual P balances (inputs minus outputs) were calculated for each plot over the 5-yr study. Total P inputs were the sum of P applied in fertilizer and wet deposition, which averaged 0.31 kg P ha^–1 yr^–1. Total P outputs were calculated as the sum of P lost in drain-flow and overland flow and removed in animal live-weight gain. Phosphorus removed in animal product was calculated as 8 g kg^–1 empty body weight, which was determined by dividing the animal live-weight gain by 1.09 (Agricultural Research Council, 1980).

Data Analysis
Logarithmic transformations were performed on the weekly concentrations of DRP, DUP, and PP in drainage water before analysis against weekly drainage flow because their distribution was found to be non-normal using the Kolmogorov-Smirnov test (Mood et al., 1974). The average monthly P concentrations in drainage water were analyzed by ANOVA to test for differences between summer months (May–September) and winter months (October–April), having adjusted for yearly variation. A linear contrast was estimated to test whether there was an increase in P concentrations with the rate of fertilizer P applied. The Seasonal Kendall test (Hirsch et al., 1982) was used to determine whether there was an overall trend in average monthly P concentrations in drainage water and flows for each plot, from March 2000 to February 2005, having adjusted for seasonal variation.

Linear Mixed Model of Phosphorus Concentrations in Overland Flow
The P concentrations in overland flow were logarithmically transformed to normalize the distribution. A linear mixed model (McCulloch and Searle, 2001) was used to determine if differences occurred between plots:

[3]

where LP_rp is the log of P concentration for the overland flow event r on plot p; a_r is a random intercept for each overland flow event (identified by the date of occurrence), which is distributed as N(a_r,a²); x_p is the estimated fixed effects for each plot; y_c is the estimated fixed effects of the years; and is random variation which is normally distributed N(0,²).

A residual maximum likelihood (Patterson and Thompson, 1971) method was used to obtain estimates for the marginal means and SEs. Paired comparisons of the marginal means for the different plots were estimated, and the significance values were adjusted for multiple comparisons using the Bonferroni method (Bonferroni, 1928).

Results

Over the 5-yr study, the average annual rainfall (March–February) and actual evaporation were 895 mm and 520 mm, respectively. The hydrologically effective rainfall (surplus rainfall) was calculated by difference and ranged from 625.7 mm in 2002/2003 to 160.1 mm in 2003/2004 (Table 1). The Olsen-P status of the plots receiving 0, 10, 20, 40, or 80 kg P ha^–1 yr^–1 was 23.5, 19.3, 21.9, 22.6, and 20.7 mg kg^–1 (equivalent to 16.0, 13.1, 14.9, 15.4, and 14.1 mg P L^–1), respectively, in February 2000. In February 2005, this had increased to 35.3, 41.2, 55.9, and 98.5 mg kg^–1 for the plots receiving 10, 20, 40, and 80 kg P ha^–1 yr^–1, respectively (Fig. 1 ). The change in weekly Olsen-P status in the plots receiving zero P fertilizer and 80 kg P ha^–1 yr^–1 is shown in Fig. 2 from March 2000 to February 2005.

View larger version (31K): [in this window] [in a new window]   Fig. 1. Olsen-P (mg kg–1) status in February each year for plots receiving 0, 10, 20, 40, or 80 kg P ha–1 yr–1.

View larger version (33K): [in this window] [in a new window]   Fig. 2. Change in weekly Olsen-P (mg kg–1) status in the plots receiving zero P and 80 kg P ha–1 yr–1.

Land Drainage Water
Generally there was a negative correlation between weekly DRP, DUP, and PP concentrations and weekly drainage flow (results not shown), probably reflecting a dilution effect. Figure 3 shows the weekly DRP concentrations in drainage water, for each calendar month, for the plots receiving zero P and 80 kg P ha^–1 yr^–1 for 5 yr. An ANOVA showed that there was significant (P < 0.001) seasonal variation in DRP concentrations for all plots, with higher concentrations occurring between May and September compared with the rest of the year (Fig. 3). This increase may be related to lower flows during the May to September period compared with the winter months. The seasonal difference in DRP concentrations was significantly greater for the plot receiving 80 kg P ha^–1 yr^–1 compared with the other plots. A linear contrast showed that there was a significant increase (P < 0.05) in DRP concentrations with the rate of P applied. There was also significant (P < 0.001) seasonal variation in DUP and PP concentrations in the drainage water of all plots with higher concentrations occurring during the summer months (May to September) compared with the other months. However, in contrast to the DRP concentrations, there was no significant difference between plots.

View larger version (18K): [in this window] [in a new window]   Fig. 3. Weekly dissolved reactive phosphorus (DRP) concentrations in land drainage water for each calendar month for the plots receiving (a) zero P and (b) 80 kg P ha–1 yr–1 for 5 yr (March 2000 to February 2005). The line is drawn through the mean values.

View larger version (35K): [in this window] [in a new window]   Fig. 4. Histograms showing the mean weekly differences in DRP concentrations for the period March 2002 to February 2005 compared with the period March 2000 to February 2002 in land drainage water from plots receiving 0, 10, 20, 40, or 80 kg P ha–1 yr–1.

The CV in DRP and TP concentrations in overland flow showed a marked increase with increasing fertilizer P application (Table 6). The CV in DUP and PP concentrations were higher in overland flow than in land drainage and averaged 0.42 and 0.50, respectively. The annual flow-weighted mean DRP and TP concentrations in overland flow were considerably higher than in land drainage water and increased with the rate of P fertilizer applied (Table 7). Maximum DRP and TP concentrations were generally recorded when overland flow occurred soon after fertilizer P application. For example, a maximum DRP concentration of 21500 µg P L^–1 was recorded in overland flow from the plot receiving 80 kg P ha^–1 yr^–1 on 4 May 2003. Fertilizer P had been applied 2 d earlier. This single event accounted for 52% of the annual DRP load in overland flow for this plot.

View this table: [in this window] [in a new window]   Table 8. Multiple comparisons table showing the significance of pairwise differences between the plots for the log10 of mean dissolved reactive phosphorus (DRP) concentrations in overland flow (averaged over all years).

Total Phosphorus Loads in Land Drainage and Overland Flow
The annual average TP load (kg P ha^–1 yr^–1) in overland flow and drainage water for each plot for the 4-yr period (March 2001 to February 2005) is shown in Fig. 5 . Year 2000 data could not be included in Fig. 5 because the volume of overland flow was not measured during 2000. Total P lost (land drainage plus overland flow) was highly variable from year to year. On the zero P plot, TP lost ranged from 0.19 kg P ha^–1 in 2003/04 to 1.55 kg P ha^–1 in 2002/03. On the plot receiving 80 kg P ha^–1 yr^–1, the range was from 0.35 kg P ha^–1 in 2001/02 to 2.94 kg P ha^–1 in 2002/03. Consistent trends in TP loss with soil P status were not obvious. However, when all years were averaged, TP losses were higher from the plots receiving the highest fertilizer P applications. There was a problem in recording the overland flows on the plot receiving 40 kg P ha^–1 yr^–1 during the summer of 2002 due to instrumentation failure. This resulted in anomalously low P losses in that year and is the reason why the average TP load in overland flow from this plot was lower than the plot receiving 20 kg P ha^–1 yr^–1 (Fig. 5).

View larger version (13K): [in this window] [in a new window]   Fig. 5. Average annual total phosphorus (TP) load (kg P ha–1) in overland flow and land drainage water (March 2001 to February 2005) for plots receiving 0, 10, 20, 40, or 80 kg P ha–1 yr–1. P loads in overland flow from the plot receiving 40 kg P ha–1 yr–1 were anomalously low due to instrumentation failure during the summer of 2002. The average weekly Olsen-P concentration (mg P kg–1), from March 2001 to February 2005, is shown above each plot.

Phosphorus balances (inputs of fertilizer P and wet deposition minus outputs in drain-flow, overland flow, and removal in animal live-weight gain) for the plots receiving 0, 10, 20, 40, and 80 kg P ha^–1 yr^–1 averaged –8.4, 1.2, 10.8, 29.5, and 71.4 kg P ha^–1 yr^–1, respectively, over the 5-yr study period. Regression analysis showed a good linear relationship (R² 0.999) between annual P balance and the observed annual increase in Olsen-P status in the 0- to 75-mm soil depth as shown in Eq. [4]:

[4]

where P_y is the yearly change in Olsen-P (mg P kg^–1) and P_b is the annual P balance.

Discussion

In the current study, the Olsen-P status of grazed grassland swards was increased over a 5-yr period by surface applying different rates of P fertilizer. Regression analysis predicted that every 10 kg P ha^–1 yr^–1 surplus in the P budget would increase Olsen-extractable P in the top 75 mm of soil by 2.5 ± 0.17 mg P kg^–1 yr^–1. Assuming a soil bulk density in the field of 1 g cm^–3 (Cruickshank, 1997), this indicates that 18.9% of the P surplus in soil to a depth of 75 mm is Olsen extractable. This increase in Olsen-P influenced P concentrations in land drainage water and overland flow.

There was significant seasonal variation in all P fractions in land drainage water, with the highest concentrations occurring between May and September each year. This may be related to low flows occurring during these months compared with other times of the year. Because the seasonal effect in DRP concentrations was greatest on the plot receiving 80 kg P ha^–1 yr^–1, compared with the other plots, it suggested that part of the seasonal increase may be due to fertilizer P applications. On the plot receiving 20 kg P ha^–1 yr^–1, daily TP concentrations in land drainage water were measured for part of the study period. A maximum concentration of 2674 µg P L^–1 occurred on 10 Sept. 2004, 3 d after fertilizer P application and heavy rainfall (18 mm d^–1) even though the soil was at a moisture deficit of 24.8 mm. This provides evidence of preferential flow pathways facilitating the loss of fertilizer P from the soil surface to drainage water at 1 m deep. Preferential flow pathways have been reported in many soil types, especially clay soils that are susceptible to cracking (Simard et al., 2000; Turner and Haygarth, 2000; Stamm et al., 2002).

Total phosphorus lost in land drainage water ranged from 0.12 to 1.49 kg P ha^–1 yr^–1 and was agronomically small relative to the amount of P being applied (<3.7%). These small losses are important factors in the eutrophication process leading to AFWM concentrations of TP of up to 237 µg P L^–1. These concentrations are well above 100 µg P L^–1, which would be classified as hypertrophic if they occurred in a lake environment (Organisation for Economic Cooperation and Development, 1982), but below the guideline of 300 µg P L^–1 set by the European Union for water abstraction for drinking water purposes (Newman, 1988).

Dissolved reactive phosphorus and TP concentrations in overland flow were considerably higher than in land drainage water and increased with P application rate. Concentrations in overland flow were highly dependent on the timing of the event relative to fertilizer application. A maximum DRP concentration of 21500 µg P L^–1 was recorded from the plot receiving 80 kg P ha^–1 yr^–1 when a large flow event occurred 2 d after fertilizer P was applied. Single overland flow events can account for a large percentage (up to 52%) of the annual DRP load. Overland flow is a major loss pathway for P from grazed grassland swards, and its contribution to total P loss can be well above that from land drainage water. There is substantial evidence linking high levels of soil P to increased losses of P in overland flow from other studies (Sharpley, 1995; Pote et al., 1999; McDowell et al., 2002), but few have quantified the relative contribution of overland flow to total P loss. Management practices should focus on controlling dissolved P losses in overland flow by considering not only the rate and timing of fertilizer applications but also on strategies to improve soil structure and hence water infiltration.

In contrast to other studies (Heckrath et al., 1995; Hesketh and Brookes, 2000), it was difficult to identify a clear Olsen-P concentration above which TP losses increased, despite the Olsen-P status in the soils ranging from 22 to 99 mg P kg^–1, after 5 yr of fertilizer P applications. Any relationship was confounded by annual variation in hydrologic events and flows as well as hydrologic differences between plots. This emphasizes the importance of medium- to long-term studies. Phosphorus concentrations in land drainage water were significantly increased with fertilizer P application rates of 40 kg P ha^–1 yr^–1 or above. However, P concentrations in overland flow were increased with P applications above 10 kg P ha^–1 yr^–1.

There was little change in the Olsen-P status of the plot receiving zero P fertilizer over the 5 yr of the study (decrease of 1.10 mg P kg^–1 yr^–1 or 0.75 mg P L^–1 yr^–1) and no evidence of a decrease in DRP or TP concentrations in overland flow or land drainage water with time. AFWM concentrations of TP in drainage water from the zero P plot ranged from 50 to 111 µg P L^–1 and were well above the 35 µg P L^–1 believed to trigger eutrophic effects in lakes (Organisation for Economic Cooperation and Development, 1982). AFWM concentrations of TP in overland flow from this plot ranged from 400 to 1021 µg P L^–1. Withholding P fertilizer for 6 yr (5 yr of study plus the equilibration year of 1999/2000) from a grazed grassland sward at the optimum Olsen-P status (24–37 mg kg^–1 or 16–25 mg P L^–1) was not long enough to lower soil P losses to drainage water or overland flow or to have an adverse effect on herbage P concentrations.

The average annual P balance (inputs minus outputs) for the plot receiving zero P was a deficit of 8.4 kg P ha^–1. A nutrient management scheme was introduced for 3 yr (1998–2000) in the Greenmount Agricultural and Horticultural College catchment, which reduced the P balance (inputs minus outputs) from a surplus of 23.5 kg P ha^–1 yr^–1 to a deficit of 15.0 ha^–1 yr^–1 (Smith et al., 2003). This led to a significant reduction in concentrations of DRP in land drainage in 2000 compared with 1997. The major difference between the two studies was the removal of P in cut silage (18 kg P ha^–1 yr^–1) in the Greenmount catchment. In the current grazed study, total P output (loss to water plus P removal in animal product) averaged 8.8 kg P ha^–1 yr^–1.

Having built up a range of Olsen-P concentrations in the soils (from 22 to 99 mg P kg^–1) over a 5-yr period, an interesting question is how quickly the Olsen-P status of the soils declines when no fertilizer P is applied and how loss rates of P to water are affected. Recent soil samples from intensive grassland farms in Northern Ireland showed that 51% of silage fields were above the agronomic optimum for Olsen-P (>37 mg kg^–1 or >25 mg P L^–1) (Jordan, personal communication, 2006). For Northern Ireland to comply with a range of European Union Directives (European Community, 1991; European Commission, 2000) on the aquatic environment, it has declared a "total territory" approach to control the agricultural contribution to nutrient losses. Part of the proposed action plan includes withholding fertilizer P application on soils of high Olsen-P status. Curtailing fertilizer P applications at the current site should demonstrate how rapidly environmental improvements can be achieved under grazing management and provide information to support policy decisions. Results suggest that improving water quality with respect to P may be a slow process unless P is redistributed within the soil profile by, for example, ploughing for the reseeding of pastures.

ACKNOWLEDGMENTS

We thank the staff in the Agriculture, Food, and Environmental Science Division, Agri-Food and Biosciences Institute, particularly Phil Dinsmore's team in the freshwater chemistry laboratory for their excellent technical support with sample collection and analysis. We also thank the staff at the Agricultural Research Institute for Northern Ireland at Hillsborough for managing the experimental site and Raymond Stewart and Briege McCarney for the collection of soil and herbage samples.

NOTES

All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher.

REFERENCES

• Agricultural Research Council. 1980. Composition of the ruminant's body and its products. p. 1–58. In K. Blaxter (ed.) The nutrient requirements of ruminant livestock. Commonwealth Agricultural Bureaux, Slough, UK.
• Bailey, J.S., J.A.M. Beattie, and D.J. Kilpatrick. 1997a. The diagnosis and recommendation integrated system (DRIS) for diagnosing the nutrient status of grassland swards: I. Model establishment. Plant Soil 197:127–135.[CrossRef][Web of Science]
• Bailey, J.S., A. Cushnahan, and J.A.M. Beattie. 1997b. The diagnosis and recommendation integrated system (DRIS) for diagnosing the nutrient status of grassland swards: II. Model calibration and validation. Plant Soil 197:137–147.[CrossRef][Web of Science]
• Bonferroni, C.E. 1928. On the measure of variation of a collectivity. J. Inst. Actuar. 59:62–65.
• Cruickshank, J.G. 1997. Soil and environment: Northern Ireland. Agricultural and Environmental Science Division, DANI and The Agricultural and Environmental Science Dep., The Queen's Univ. of Belfast, Newforge Lane, Belfast, UK.
• Edwards, A.C., and P.J.A. Withers. 1998. Soil phosphorus management and water quality: A UK perspective. Soil Use Manage. 14:124–130.[CrossRef]
• Eisenreich, S.J., R.T. Bannerman, and D.E. Armstrong. 1975. A simplified phosphorus analysis technique. Environ. Lett. 9:43–53.[Web of Science][Medline]
• European Commission. 2000. Directive 2000/60/EC of the European Parliament and of the Council of the European Union of 23 October 2000 establishing a framework for Community action in the field of water policy. Off. J. Eur. Commun. L327:1–72.
• European Community. 1991. Council directive concerning the protection of water against pollution caused by nitrates from agricultural sources. EC 91/676. Off. J. Eur. Commun. L375:1–8.
• Foy, R.H., and A.E. Bailey-Watts. 1998. Observations on the spatial and temporal variation in the phosphorus status of lakes in the British Isles. Soil Use Manage. 14:131–138.[CrossRef]
• Foy, R.H., R.V. Smith, C. Jordan, and S.D. Lennox. 1995. Upward trend in soluble phosphorus loadings to Lough Neagh despite phosphorus reduction at sewage treatment works. Water Res. 29:1051–1063.
• Heckrath, G., P.C. Brookes, P.R. Poulton, and K.W.T. Goulding. 1995. Phosphorus leaching from soils containing different phosphorus concentrations in the Broadbalk experiment. J. Environ. Qual. 24:904–910.[Web of Science]
• Hesketh, N., and P.C. Brookes. 2000. Development of an indicator for risk of phosphorus leaching. J. Environ. Qual. 29:105–110.[Abstract/Free Full Text]
• Hirsch, R.M., J.R. Slack, and R.A. Smith. 1982. Techniques of trend analysis for monthly water quality data. Water Resour. Res. 18:107–121.
• Krug, A. 1993. Drainage history and land use pattern of a Swedish river system: Their importance for understanding nitrogen and phosphorus load. Hydrobiologia 251:285–296.[CrossRef][Web of Science]
• McCulloch, C.E., and S.R. Searle. 2001. Generalized, linear, and mixed models. John Wiley & Sons, New York.
• McDowell, R.W., A.N. Sharpley, P.J.A. Kleinman, and W.J. Gburek. 2002. Hydrological source management of pollutants at the soil profile scale. p. 197–223. In P.M. Haygarth, and S.C. Jarvis (ed.) Agriculture, hydrology and water quality. CAB Int., Wallingford, UK.
• Ministry of Agriculture, Fisheries, and Food. 1986. The analysis of agricultural materials. 3rd ed. Reference Book 427. HMSO, London.
• Ministry of Agriculture, Fisheries, and Food. 1994. Fertilizer recommendations for agricultural and horticultural crops. 6th ed. RB209. HMSO, London.
• Mood, A.M., F.A. Graybill, and D.C. Boes. 1974. Nonparametric methods. p. 504–526. In Introduction to the theory of statistics. McGraw-Hill, Inc., Tokyo, Japan.
• Murphy, J., and J.P. Riley. 1962. A modified single solution method for the determination of phosphate in natural waters. Anal. Chim. Acta 27:31–36.[Medline]
• Newman, P.J. 1988. Classification of surface water quality: Review of schemes in EC member states. Heinemann Professional Publ., Oxford.
• Olsen, S.R., R.V. Cole, F.S. Watanabe, and L.A. Dean. 1954. Estimation of available phosphorus in soils by extraction with sodium bicarbonate. p. 1–19. Circular No. 939. U.S. Dep. of Agriculture, Washington, DC.
• Organisation for Economic Cooperation and Development. 1982. Eutrophication of waters. Monitoring, assessment and control. OECD, Paris.
• Patterson, H.D., and R. Thompson. 1971. Recovery of inter-block information when block sizes are unequal. Biometrika 58:545–554.[Abstract/Free Full Text]
• Pote, D.H., T.C. Daniel, D.J. Nichols, A.N. Sharpley, P.A. Moore, Jr., D.M. Miller, and D.R. Edwards. 1999. Relationship between phosphorus levels in three ultisols and phosphorus concentrations in runoff. J. Environ. Qual. 28:170–175.[Web of Science]
• Rekolainen, S. 1989. Phosphorus and nitrogen load from forest and agricultural areas in Finland. Aqua Fenn. 19:95–107.
• Sharpley, A.N. 1995. Dependence of runoff phosphorus on extractable soil phosphorus. J. Environ. Qual. 24:920–926.[Abstract/Free Full Text]
• Sharpley, A.N., and S. Rekolainen. 1997. Phosphorus in agriculture and its environmental implications. p. 1–53. In H. Tunney et al. (ed.) Phosphorus loss from soil to water. CAB Int., Wallingford, Oxford.
• Simard, R.R., S. Beauchemin, and P.M. Haygarth. 2000. Potential for preferential pathways of phosphorus transport. J. Environ. Qual. 29:97–105.[Abstract/Free Full Text]
• Smith, R.V., C. Jordan, and J.A. Annett. 2005. A phosphorus budget for Northern Ireland: Inputs to inland and coastal waters. J. Hydrol. 304:193–202.[CrossRef]
• Smith, R.V., S.D. Lennox, and J.S. Bailey. 2003. Halting the upward trend in soluble phosphorus transported from a grassland catchment. J. Environ. Qual. 32:2334–2340.[Web of Science]
• Stamm, C., R. Sermet, J. Leuenberger, H. Wunderli, H. Wydler, H. Fluhler, and M. Gehre. 2002. Multiple tracing of fast solute transport in a drained grassland soil. Geoderma 109:245–268.[CrossRef][Web of Science]
• Turner, B.L., and P.M. Haygarth. 2000. Phosphorus forms and concentrations in leachate under four grassland soil types. Soil Sci. Soc. Am. J. 64:1090–1099.[Abstract/Free Full Text]
• Watson, C.J., and R.H. Foy. 2001. Environmental impacts of nitrogen and phosphorus cycling in grassland systems. Outlook Agric. 30:117–127.[Web of Science]
• Watson, C.J., C. Jordan, S.D. Lennox, R.V. Smith, and R.W.J. Steen. 2000. Inorganic nitrogen in drainage water from grazed grassland in Northern Ireland. J. Environ. Qual. 29:225–232.[Abstract/Free Full Text]

This article has been cited by other articles:

				C. van der Salm, W. J. Chardon, G. F. Koopmans, J. C. van Middelkoop, and P. A.I. Ehlert Phytoextraction of Phosphorus-Enriched Grassland Soils J. Environ. Qual., February 25, 2009; 38(2): 751 - 761. [Abstract] [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Watson, C. J. Articles by Matthews, D. I. Search for Related Content PubMed PubMed Citation Articles by Watson, C. J. Articles by Matthews, D. I. Agricola Articles by Watson, C. J. Articles by Matthews, D. I. Related Collections Surface Water Quality Phosphorus Field-Scale Studies