JEQ Grow Your Career With ASA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Related articles in JEQ
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (12)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by McDowell, R. W.
Right arrow Articles by Wilcock, R. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by McDowell, R. W.
Right arrow Articles by Wilcock, R. J.
Agricola
Right arrow Articles by McDowell, R. W.
Right arrow Articles by Wilcock, R. J.
Related Collections
Right arrow Surface Water Quality
Right arrow Water Quality
Right arrow Watershed and Landscape Processes
Right arrow Wetlands and Aquatic Processes
Right arrow Phosphorus
Published in J. Environ. Qual. 33:2111-2121 (2004).
© ASA, CSSA, SSSA
677 S. Segoe Rd., Madison, WI 53711 USA

TECHNICAL REPORTS

Landscape and Watershed Processes

Particulate Phosphorus Transport within Stream Flow of an Agricultural Catchment

R. W. McDowella,* and R. J. Wilcockb

a AgResearch Ltd, Invermay Agricultural Centre, Private Bag 50034, Mosgiel, New Zealand
b National Institute of Water and Atmospheric Research, P.O. Box 11 115, Hamilton, New Zealand

* Corresponding author (richard.mcdowell{at}agresearch.co.nz)

Received for publication April 14, 2004.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
There is interest in quantifying phosphorus (P) loss from intensively grazed dairy landscapes to identify key pathways and target remediation methods. The Bog Burn drains a dairying catchment in Southland, New Zealand, and has been monitored at fortnightly intervals over a 12-mo period at four sites for suspended sediment (SS), dissolved reactive phosphorus (DRP), and total phosphorus (TP). Time-integrated samplers, deployed at 0.6 median water depth at each site (calculated from previous year's flow data), collected sediment samples, which were analyzed for SS, bioavailable phosphorus (BAP), and TP. Mean concentrations of DRP and TP in stream flow and BAP and TP in sediment were generally highest in summer or autumn (0.043 mg DRP L–1, 0.160 mg TP L–1, 173 mg BAP kg–1, 2228 mg TP kg–1) and lowest in winter or spring (0.012 mg DRP L–1, 0.034 mg TP L–1, 6 mg BAP kg–1, 711 mg TP kg–1), while loads were highest in winter. Analysis of 137Cs concentrations in trapped sediment, topsoil, subsoil, and stream bed and bank sediment indicated that trapped sediment was derived from topsoil and entered the stream either through tile drainage or, to a lesser extent, overland flow. Because concentrations of DRP and TP in stream flow are in excess of recommended limits for good water quality (>0.01 mg DRP L–1, 0.033 mg TP L–1), management should focus on the topsoil and specifically on decreasing P loss via tile drainage. This is best achieved by decreasing soil Olsen P concentrations, especially because, on average, Olsen P concentrations in the catchment were above the agronomic optimum.

Abbreviations: BAP, bioavailable phosphorus • DRP, dissolved reactive phosphorus • PP, particulate phosphorus • SS, suspended sediment • TP, total phosphorus


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
THE TRANSPORT AND DYNAMICS of P in fluvial systems are central to the P-limited eutrophication of lowland streams. Much of this transport can occur as P bound to particulate material in suspended sediment (Brunet and Brian-Astin, 1998). The measurement of total phosphorus (TP) is a poor predictor of P available to periphyton, because algae, the dominant form of periphyton, take up only phosphate and P desorbable from particulate phosphorus (PP) (Ekholm, 1998; Reynolds and Davies, 2001). The proportion of PP bioavailable to algae in aerobic systems can be estimated by anion exchange membranes or Fe-oxide strips (Sharpley et al., 1995; Uusitalo and Ekholm, 2003). A number of authors have shown that bioavailable PP varies with surrounding land use and management, and seasonally (Pionke and Kunishi, 1992; Sharpley et al., 1995; Steegen et al., 2001). Studies have also shown the proportion of bioavailable PP in suspended sediment (SS) to vary from 7 to 50% (Huettl et al., 1979; Uusitalo et al., 2001). However, such data are derived from catchments either in North America or Europe where management practices such as grazing during winter differ dramatically from those used in New Zealand or Australia, where winter temperatures are commonly milder. Differences in management are compounded by different flow regimes and P loss mechanisms (e.g., flood irrigation). Although P is still lost via overland or subsurface routes, the concentrations and quantities differ from those in North America or Europe, especially between seasons (Gillingham and Thorrold, 2000). For instance, little P loss can occur from frozen fields in much of North America and parts of Europe in winter (Xue et al., 1998), whereas this is when most P loss occurs in New Zealand catchments (Gillingham and Thorrold, 2000).

Coupled to seasonality and flow mechanisms, once in the stream the impact and fractionation of P is determined largely by hydrodynamics: the bioavailability of P for periphyton and subsequent growth is influenced by flow rates and bed materials (Nikora et al., 2003). While hydrodynamics can influence when sources such as stream bank or bed sediment or topsoil come into flow and when they are deposited, the origin of the sediment can be the principal factor in determining how much bioavailable P is held by SS. For instance, sediment derived from topsoil that receives regular P inputs is likely to contain more bioavailable P than sediment derived from the eroded subsoil. One method of determining sediment origin is to use sediment-specific characteristics, which can then be correlated with the sediment source. Such a characteristic is 137Cs concentration. This anthropogenic isotope is the result of fallout from atmospheric nuclear testing and is still detectable in soil. Strong fixation to soil means that 137Cs is restricted to topsoil, and that the concentration decreases exponentially with depth. Consequently, 137Cs has been used as a measure of erosion rates (e.g., Walling and He, 1997) and sediment origin in drain flow and overland flow (Laubel et al., 1999; Uusitalo et al., 2001).

With recent advances in methodology it has become possible to obtain a representative time-integrated sample of SS in stream flow (e.g., Russell et al., 2000; Phillips et al., 2000). Consequently, the main objective of this paper was to determine and relate the bioavailability and concentration of P in SS sampled at four locations along the primary reach of a stream draining an agricultural catchment to overall P loss dynamics. A secondary objective was to suggest the most effective form of managing this loss.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Study Area
The study area was the 8760-ha Bog Burn catchment in Southland, New Zealand (Fig. 1) . Land use within the catchment is predominantly pastoral (largely dairy farming) with some exotic production forestry (Pinus radiata spp.) at the headwaters. Elevation ranges from more than 500 m above sea level in the headwaters to 50 m above sea level near the catchment outflow with the majority of pastoral land between 100 and 50 m above sea level, which is artificially drained by a network of mole-tile drains at a 70- to 100-cm depth. Slope ranges from 0.5 to 20%, with only areas near the headwaters >5%. Mean annual rainfall in the catchment is about 900 mm while the mean annual temperature is 10.2°C. Soils within the catchment are Pukemutu silt loam soils (New Zealand Soil Classification: Argillic-mottled Fragic Pallic; USDA Taxonomy: Typic Fragiochrept) with a mean Olsen bicarbonate-extractable P concentration (within dairy farms) of 42 mg P kg–1 soil. Dairy shed effluent (about 20 kg P ha–1 yr–1) is spread throughout the year onto defined blocks of land within each dairy farm (about 10%), except in winter months when cows are not milked (dried-off) and grazed off-farm. Superphosphate application rates range from 40 to 100 kg P ha–1 yr–1 with an average rate of 61 kg P ha–1 yr–1. Superphosphate is generally applied in late spring and over summer. The mean stocking rate is 2.9 cows ha–1, which return in dung between 10 and 23 g P d–1 cow–1 to the soil surface (Haynes and Williams, 1993).



View larger version (33K):
[in this window]
[in a new window]
  		Fig. 1. Land use (left), elevation (right), and sampling sites (1–4) along the Bog Burn catchment, Southland, New Zealand.

 
Sampling sites are indicated in Fig. 1, with samples of stream flow and time-integrated SS taken fortnightly for one year beginning at the start of January 2002. Hourly flow records were kept at Site 3 via a permanent gauging station. However, expense prohibited the continuous measurement of flow (Q) at other sites. A staff gauge has been maintained at Site 2 and a relationship established with the continuous recorder data at Site 3, such that Q2 = 2.244Q0.78173. The catchment areas above Sites 2 and 3 are 1610 and 2506 ha, respectively. The four sites, numbered 1 to 4, represent areas influenced by forestry and significant slope (Site 1) and increasingly downstream (Sites 2–4) by inputs from flat land (slope < 2%) used for dairy farming and connected to the stream via tile drains.

Stream Water Collection and Analysis
Each fortnight, water samples (1 L) were collected at each site and analyzed for dissolved reactive phosphorus (DRP, that is, P measured following filtration through a 0.45-µm filter) and TP, after a persulfate digestion, by flow injection analysis and colorimetry (American Public Health Association, 1998). A measure of SS at the time of sampling was also gained by filtration of a 2-L sample through a Whatman (Maidstone, UK) GF/C filter paper and weighing the oven-dried residue (American Public Health Association, 1998).

Monthly P load was calculated using an interpolation procedure (Walling and Webb, 1982):

[1]
where Ci is the instantaneous individual DRP concentration (mg L–1), Qi is the instantaneous discharge at time of sampling (L s–1), Qr is the mean discharge for the time of record (L s–1), and K is the conversion factor to take account of period of record.

This method has been employed as an estimate of nutrient load (e.g., McDowell and Trudgill, 2000). It is not the best method identified by Walling and Webb (1982) of calculating loads. However, given the poor nature of rating curves between P and flow (r2 generally < 0.10) and the frequency of sampling, this method was seen as most appropriate.

Suspended Sediment and Soil Sampling and Analysis
Time-integrated sediment samplers, based on the design of Phillips et al. (2000), were installed at 0.6 median water depth and attached using rubber ties to steel uprights. Briefly, samplers operate in situ and each one consists of a narrow (4-mm diameter) inlet and outlet tube, with a larger cavity (98-mm diameter) in between. This cavity (7.5-L volume) has a cross-sectional area approximately 600 times greater than the inlet and outlet tubes and drastically decreases water flow velocity relative to ambient flow and enhancing sedimentation. Samplers commonly capture more than 70% of sediment that passes through them, while obtaining a sample representative of SS over a defined time period for small streams (Phillips et al., 2000; Russell et al., 2000).

Each fortnight (±1–2 d), sediment samplers were removed, sediment was retained within the cavity collected, and the sampler was flushed with stream water before being tied back in place. Due to flooding in June (Fig. 2) samplers were swept away and a sampling period was missed. Sediment was isolated from the collected samples via filtration through a 0.7-µm borosilicate microfiber filter (GF75-MFS; Advantec, Pleasanton, CA) and air-dried, and the residue was weighed. Dried sediment (no more than 150 mg) was placed in a 40-mL centrifuge tube along with 30 mL of deionized water and a Fe-oxide strip designed to sequester bioavailable P (Method 4; Bramley and Roe, 1993). This suspension was shaken overnight and the strip removed. Any particles were washed off back into the centrifuge tube and P was desorbed from the strip by shaking it with 30 mL of 0.1 M H2SO4 for 1 h. The P released is termed bioavailable phosphorus (BAP) and determined via colorimetry of a neutralized aliquot (American Public Health Association, 1998). Phosphorus remaining in sediment after removal of BAP was isolated by centrifuging the suspension at 2600 x g for 15 min, decanting the supernatant, oven-drying the residue (105°C), and determining TP following an aqua regia digest (Crosland et al., 1995).



View larger version (21K):
[in this window]
[in a new window]
  		Fig. 2. Fortnightly suspended sediment (SS), dissolved reactive phosphorus (DRP), and total phosphorus (TP) concentrations in stream flow at each site and daily mean flow at Site 3.

 
For particle size analysis of dispersed and nondispersed trapped sediment, three subsamples for each season and sampler were taken and split in half (500 L). One half was dispersed using a high speed homogenizer (25000 rpm, Kinematica; Polytron-Aggregate, Luzern, Switzerland) for 5 min. Both dispersed and nondispersed samples were analyzed by a laser particle sizer (Mastersizer; Malvern Instruments, Malvern, UK).

Topsoil (plow layer, 0- to 20-cm depth) samples were taken at each site in each of six paddocks that were >100 m away from the stream channel. At the same time, subsoil samples were taken from the 20- to 100-cm depth to encompass mole-tile drains and four samples of stream bed and bank sediment (surface 1 cm) taken at each site. Soil and sediment were air-dried, ground, and sieved to <1 mm before analysis of BAP, TP, and particle size (only on dispersed samples) as above. Samples (500 g) were also analyzed for 137Cs concentration via gamma spectrophotometry for 16 h (Model BE5030; Canberra Industries, Meriden, CT). Due to low 137Cs concentrations, samples of trapped sediment for each site were bulked together within a season to get >60 g of material, and counted via gamma spectrophotometry for 7 d.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Stream Flow
A summary of water quality data is given in Table 1 and Fig. 2 for the four sites along Bog Burn. Continuous stream flow data were only recorded at Site 3, which varied from 44 to 5186 L s–1 with a mean and median flow rate of 389 and 163 L s–1, respectively. The highest flows were recorded in autumn and winter months from May to July, with a flood at the beginning of June sweeping away sediment samplers at Sites 3 and 4. Samplers were not reinstated until three weeks later, while regular fortnightly sampling was unaffected.


View this table:
[in this window]
[in a new window]
  		Table 1. Mean suspended sediment and dissolved reactive phosphorus (DRP) and total phosphorus (TP) concentrations in stream flow at each site for each season.

 
Suspended sediment concentrations were generally highest at Site 1, which received runoff from steeply sloping forest land (150–500 m above sea level), and lowest at Site 4 near the catchment outflow, although differences between sites were generally small. Mean SS was not significantly different between Sites 2, 3, and 4. During the year, the highest mean concentration of SS occurred in spring (Table 1). In general, SS concentrations rarely exceeded 10 mg L–1, reflecting low sediment losses expected from a spring-fed stream in a catchment dominated by slopes of <2% and pastoral land use. This was further emphasized by the low annual (2002 calendar year) sediment yield of 98 kg ha–1, calculated using Eq. [1] and continuous flow data for Site 3 (approximate catchment area of 2506 ha).

In contrast to SS, the estimated annual yield for DRP at Site 3 was high at 0.34 kg ha–1 compared with other mixed grazing pasture catchments in New Zealand (range = 0.04–0.3 kg ha–1; Wilcock et al., 1999). On the other hand, the annual estimated yield for TP of 0.71 kg ha–1 was in the middle to lower end of the range established by Wilcock et al. (1999) of 0.3 to 1.7 kg ha–1 for New Zealand catchments. The high concentration of DRP, and the high DRP to TP ratio (46%), reflect both intensive dairying within the catchment (positive P balance estimated as 34 kg P ha–1 yr–1; R. Monaghan, personal communication, 2002), and the widespread use of tile drains (100% below Site 1) that discharge less particulate-associated P compared with catchments dominated by overland flow (McDowell et al., 2001). Monaghan et al. (2002) showed that about 60% of TP loss from tile-drained Pukemutu soils was via subsurface flow with the majority (50–60%) of that as DRP.

At each of the four sites, the highest mean concentration of DRP and TP was recorded during summer (except for Site 3 when winter > summer), while the lowest mean concentration was recorded during winter or spring (Table 1). The relationship between SS and TP was weak (r2 < 0.4, P < 0.05) and probably caused by dissolved P dominating TP in flow.

Using continuous flow data for Site 3, loads for DRP and TP were calculated for each season. For DRP, the lowest load occurred during summer (54 kg) as low flow conditions dominated. Conversely, the opposite was true of DRP loads in winter when high flows flushed much P from the system (513 kg lost). The seasonal trend for TP loss was slightly different; the lowest TP loss occurred during autumn (257 kg) while as with DRP, most TP loss occurred in winter (878 kg).

Sediment Samplers
The sediment samplers collect more sediment as flow rate and the quantity of sediment transported in flow increases (Phillips et al., 2000). As such, trapped sediment was related to cumulative flow, Q{Delta}T m3 [e.g., at Site 1; sediment (mg) = 0.4 exp(0.0005 x cumulative flow); r2 = 0.5, P < 0.05]. To estimate the overall efficiency of the samplers to trap sediment, the quotient of sampler cross-sectional area and the cross-sectional area of the stream channel at Site 3 (assuming depth equivalent to the annual median flow rate) was multiplied by annual SS load (248 Mg), which would have yielded 0.092 kg of SS if 100% efficient. The actual SS load was 0.039 kg (assuming SS entrapment for the missed sample was equivalent to the mean of all winter samples) giving an estimated efficiency of 42%. This value is probably an overestimate because most sediment will be transport during high flow when cross-sectional area will be greater.

The mass of sediment trapped by each sampler tended to mirror the concentration of SS detected in the fortnightly stream sampling; that is, most sediment was trapped at Site 1, while the least sediment was, in general, trapped at Sites 3 and 4 (Fig. 3) . In contrast to trapped sediment, concentrations of BAP and TP were highest at Sites 3 and 4 and lowest at Site 1 (Table 2). This may be due to agricultural intensification downstream and particle size sorting in overland or stream flow, which favors P-rich fine materials over heavy coarse-sized particles (Sharpley and Smith, 1990). Data in Fig. 4 show that when mechanically dispersed, the proportion of fines in sediment collected at Sites 1 and 4 increased from 23 at Site 1 to 37% at Site 4. Overall, the increase in fines from Sites 1 to 4 was 12% (n = 12), and did not significantly differ between seasons (P > 0.05).



View larger version (22K):
[in this window]
[in a new window]
  		Fig. 3. Sediment and bioavailable phosphorus (BAP) and total phosphorus (TP) concentrations trapped in samplers located at each site. Note the different scales on the P concentration axes.

 

View this table:
[in this window]
[in a new window]
  		Table 2. Mean sediment load and bioavailable phosphorus (BAP) and total phosphorus (TP) concentrations trapped by samplers at each site for each season.

 


View larger version (23K):
[in this window]
[in a new window]
  		Fig. 4. The relationship between bioavailable P (solid symbols) and total P (open symbols) and mass of sediment trapped at each site. The dashed vertical line represents the point of inflexion for the split-line model (not shown for clarity).

 
Data in Table 2 show that, like DRP and TP in stream flow, BAP and TP in trapped sediment were greater in summer and autumn (except in Site 1) than during winter or spring. In contrast, trapped sediment load paralleled flow and was greater in winter and spring than summer or autumn. Overall, there was a significant negative exponential relationship (P < 0.05) between BAP or TP concentration and the mass of sediment trapped, with small quantities of sediment containing a greater P concentration than large quantities (Fig. 5 ; Sharpley and Smith, 1990). This P-poor material was prevalent in the sediment trapped at Site 1, and could reflect either the P status of the forested soil or entrapment of coarse P-poor material from steep slopes. A split-line model (McDowell and Sharpley, 2001) showed that BAP and TP were significantly enriched with P when <0.5 g of sediment was trapped. This occurred once for Site 1, and with increasing frequencies downstream (4, 8, and 8 times for Sites 2, 3, and 4, respectively; Fig. 5). Using flow data for Site 3, a 0.5-g sediment load equated to elevated BAP and TP concentrations when flow was <185 L s–1.



View larger version (23K):
[in this window]
[in a new window]
  		Fig. 5. Laser particle size distribution of dispersed and nondispersed sediment samples from Sites 1 and 4 on 21 Jan. 2003.

 
Table 3 shows that BAP and TP concentrations of topsoil and subsoil were less enriched with P than in trapped sediment at the same site. In addition, concentrations of BAP and TP were greater downstream in topsoil and at the Site 4 stream bank, but not different between subsoils or stream bed sediments at each site (Table 3). While selective sorting of particles in flow may account for P-rich fines downstream, Fig. 6 shows that the 137Cs concentration of trapped sediment was not significantly different between sites. Furthermore, 137Cs concentration of trapped sediment was similar to that in topsoil, but not subsoil or stream bed or bank sediment. Because 137Cs is more concentrated in the clay fraction of topsoils (Uusitalo et al., 2001), size sorting will enrich 137Cs concentrations in flow compared with the source. Using the specific surface area ratio and the equations of He and Walling (1996), enrichment ratios for 137Cs concentration indicated that trapped sediment was only similar to topsoil (mean enrichment of 1.65). No significant difference was noted (P > 0.05) between seasons for 137Cs concentrations in trapped sediment at each site.


View this table:
[in this window]
[in a new window]
  		Table 3. Mean concentrations of bioavailable phosphorus (BAP) and total phosphorus (TP) in topsoil, subsoil, and stream bed and bank sediments from each site.

 


View larger version (29K):
[in this window]
[in a new window]
  		Fig. 6. Cesium-137 concentration in stream bank and bed sediments, topsoil and subsoil, and bulked suspended sediment (SS) samples for each sampling site. Error bars are 95% confidence intervals.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 REFERENCES
 
Variation of Phosphorus Export
Fluctuations in P export along a stream vary according to P inputs and discharge. Sources of sediment and P into Bog Burn upstream of Site 1 are probably overland flow from steep slopes in the headwaters (Fig. 1), while subsurface tile drainage and to a lesser extent overland flow from intensive agriculture (largely dairying) are the most likely causes of P downstream. Downstream of Site 1, P loads may also be enhanced by inputs from stock, stream bank destabilization, and effluent and fertilizer applications (Wilcock et al., 1999).

Seasonal variation in mean P concentrations at each site is clear from Table 1, and typical of most catchments in temperate maritime climates. More unique to Bog Burn are the dominance of pastoral land use and the prevalence of 100% tile drainage in flat (<2% slope) pastoral land downstream of Site 1. Both characteristics decrease the influence of overland flow and particulate P input into stream flow. This is evident by seasonal fluctuations where most P is flushed from soils by drainage during autumn and winter, and by the greater ratio of DRP to TP at Site 4 (dominated by inputs from tile drainage) compared with Site 1 (more likely to be dominated by overland flow).

The variation of P concentration with season led Brunet and Brian-Astin (1998) to conclude that most retention of P within a catchment occurs during spring and summer while most loss occurs during autumn and winter. Similarly, May et al. (2001) concluded for the upper River Cherwell in Oxfordshire, UK, that most retention occurred in spring to early autumn, and the release of stored P occurred during winter. In Bog Burn, with little erosion and low concentrations of particulate P, seasonal variation is dominated by dissolved P loss in winter. Assisting dissolved P loss in winter are wet conditions (soils near saturation; R. Monaghan, personal communication, 2002) and soil damage via compaction. Both factors cause anaerobic conditions in soil microsites and subsequent release of P from ferric hydroxides (Jensen et al., 1998). Losses in summer are episodic (i.e., little continuous drainage occurs), but have a high P concentration reflecting dung inputs via direct stock access or increases in soil Olsen P in summer from warmer temperatures or P inputs via fertilizer or effluent spreading (e.g., Saunders and Metson, 1971; McDowell and Trudgill, 2000). However, with most flow occurring in winter, flushing of P built up in topsoils in summer means that on a loading basis, most P loss occurs at this time.

Losses of P from tile-drained plots (four replications, each 0.18 ha) under dairying with the same soil type (Pukemutu) were 1 kg P ha–1 yr–1, with the majority lost via tile drainage (50–60%; Monaghan et al., 2002). These losses are about 30% greater than those calculated to pass Site 3, but not unexpected for a number of reasons. Perhaps the most pertinent is the removal of dissolved P from drainage waters by sediment in tile drains, ditches, and culverts because the catchment has only been used for intensive dairying for 8 yr. Sediments sampled from drainage ditches in dairy catchments in the North Island of New Zealand have shown a large capacity to remove dissolved P from drainage waters (Nguyen and Sukias, 2002).

Particulate Associated Phosphorus Loss
Sediment transported in streams and rivers comes from topsoil via overland flow (Dorich et al., 1984) or subsurface flow including tile drainage (Ulén and Mattsson, 2003), subsoil (Wallbrink et al., 2003), or stream banks and beds (Laubel et al., 2003). Depending on the system, the relative contribution of each source will vary (Walling et al., 1997). For instance, Laubel et al. (2003) showed that for 14 Danish agricultural catchments stream bank material was of a similar total P concentration to topsoil and that erosion of stream bank material accounted for about 50% of catchment P loss. In contrast, Wallbrink et al. (2003) showed that for a large (8700 ha) eastern Australian catchment dominated by steep (>20%) forested slopes, 62% of sediment P loss was from subsoil erosion of gullies. However, Chapman et al. (2001), Laubel et al. (1999), and Uusitalo et al. (2001) showed that for gently sloping (3.7 and 2% slope, respectively) tile-drained land, sediment P loss via drainage or overland flow was topsoil derived. Data presented here for 137Cs concentrations link topsoil to SS in stream flow (Fig. 6). Topsoil may have moved via overland flow or drainage. Consequently, subsoil and stream bank erosion and resuspension of stream bed sediment are less important. This is not surprising because these processes often require much erosive power or large and highly variable flow regimes; neither occur in Bog Burn.

Upstream of Site 1, all land is used for forestry where slopes are much steeper than below Site 1. Steep slopes and a lack of artificial drainage mean that overland flow is more likely than in other areas of the catchment. With overland flow comes the likelihood of losing coarser-sized sediment, compared with fine-sized sediment lost in drainage (Quinton et al., 2001). Data showed that trapped sediment was coarser at Site 1 than at Site 4 (Fig. 4). This suggests that size sorting may have occurred. Uusitalo et al. (2001) showed that concentrations of 137Cs and P are enriched in fine particles compared with coarse particles. Site 1 contained the highest 137Cs concentration in topsoil, yet concentrations were lowest in trapped sediment. Concentrations of 137Cs had increased by about 10% from Sites 1 to 4, in line with the increase in fines noted by particle size analysis (on average, 12%). This suggests that coarse materials poor in 137Cs and P (Fig. 5) eroded in flow above Site 1, and settled out downstream.

While sediment size sorting in stream flow can account for some of the increase in P downstream, if we apply the same logic and assume that particles subject to sorting are sorted in the same manner for P (Sharpley and Smith, 1990) as for 137Cs, then P should have increased by 12% on average from Sites 1 to 4. Tables 1 and 2 clearly show that the increase was an order of magnitude greater than that. Consequently, P concentrations and inputs to the Bog Burn via topsoil must have changed (see Table 3). Clearly, P losses in this tile-drained catchment are a direct function of topsoil management and conditions.

Both stream flow P concentrations and concentrations of BAP and TP in trapped sediment indicated a seasonal influence. However, no seasonal influence was noted among sites for 137Cs concentrations. This indicates that while the source of trapped sediment did probably not change, topsoil P concentrations may have. As noted earlier, seasonal changes in soil P concentrations have been widely noted for soil test P (e.g., Olsen P) largely due to changing P inputs in fertilizer or dung. However, in a study of BAP in topsoil, Sharpley et al. (1995) also noted that BAP increased in warmer months due to increased biological activity, mineralization, and fertilizer application. The link between topsoil and trapped sediment suggests that topsoil moves either via subsurface drainage (e.g., macropore flow) or overland flow into the stream channel and that seasonality in topsoil P and management defines BAP and TP in SS and dissolved P and TP in stream flow.

Management Implications
With the presence of tile drains throughout the lower three-quarters of the catchment, a direct conduit exists for the transfer of P-enriched water and topsoil from intensively farmed soils. At present, mean concentrations of DRP and TP loss exceed the Australian and New Zealand Environment and Conservation Council limits of 0.01 and 0.033 mg P L–1 for lowland streams in New Zealand (Table 1; Australian and New Zealand Environment and Conservation Council, 2000). Consequently, P loss can be best decreased by focusing on topsoil management.

Management practices to prevent the loss of P from tile drains are limited, but worthwhile because several studies have noted that artificial drainage decreases overall P loss by minimizing overland flow (Haygarth et al., 1998). Management directed at mitigating P loss from tile drains has involved subsoiling, effectively breaking apart preferential flow pathways and forcing dissolved P to flow through soil where it is sequestered en route to drains (Djodjic et al., 2002). Clearly, this restricts drainage and would probably lead to a significant loss in on-farm production while simply transferring the pathway of P loss to overland flow. Another practice is to use a wetland (natural or constructed) at the end of tile drains to remove P from drainage water. The problem with this is that many tile drain outlets are either obstructed or very close to stream level, making wetlands either difficult to position or obtrusive to stream flow (causing siltation). Perhaps the best management practice is to minimize the potential for P loss on-farm. This means not overloading topsoil with P and increasing the potential for P loss. Dairy shed effluent spreading occurs year-round except when cows are not milked in June and July; if effluent is spread on wet soil there is risk of direct transfer to drainage water. Monaghan et al. (2002) recommended that effluent should not be spread when soil moisture deficit is <20 mm.

Currently, fertilizer P application rates within the catchment are in the order of 49 to 86 kg P ha–1 yr–1 with a mean application rate of 68 kg P ha yr–1, together with a P surplus within the catchment of 34 kg P ha–1 yr–1; this has increased soil P. The catchment's Pukemutu soil is poor in Al and Fe oxides, which limits P retention and enhances the potential for P loss. Loss of P is significantly (P < 0.05) greater beyond a soil Olsen P concentration of approximately 35 mg kg–1 (McDowell et al., 2003). As such, 45% of the soils within the catchment have been classified as moderate or high risk (i.e., ≥35 mg Olsen P kg–1 soil; Drewry et al., 2003). This also represents an Olsen P concentration beyond the optimum level (20–35 mg kg–1) for pasture production (Roberts and Morton, 1999). Clearly, problems in the Bog Burn relate to elevated soil P. Additional management strategies are less likely to be effective in decreasing P losses but include minimizing stream bank erosion from stock damage, excluding direct access by stock to the stream, and minimizing the contribution of P loss from laneways and surface drains by lessening their connectivity with the stream (e.g., with intercepting riparian plantings).


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
REFERENCES
 
Concentrations of DRP and TP in stream flow and BAP and TP in trapped sediment followed a seasonal and geographical pattern, generally increasing downstream, highest in summer, and lowest in winter or autumn. Loads were highest in winter. A high proportion of P in stream flow was in dissolved form, which entered the stream largely via tile drainage in the lower three-quarters of the catchment. Changes in concentration were linked to management of topsoil. Similarly, analysis of 137Cs indicated that topsoil entering via tile drainage, and to a lesser extent, overland flow was the controlling factor for sediment-associated P in the stream. Because concentrations of DRP and TP in stream flow are in excess of current limits for good freshwater quality in New Zealand, it is suggested that management be focused on topsoil and specifically minimizing the potential for P loss via tile drainage. At present, this is probably best achieved by decreasing fertilizer and effluent applications and hence soil Olsen P status, especially because, on average, soil Olsen P within the catchment is beyond the agronomic optimum.


   ACKNOWLEDGMENTS
 
Funding for this work was provided by the New Zealand Foundation for Research, Science and Technology under contracts AGRX002, C10X0320, and C01X0215. We are also grateful for funding from the New Zealand dairy industry and the Ministry of Agriculture and Forestry Sustainable Farming Fund. The assistance of staff at Environment Southland (K. Hamill, M. White, and S. Levington) and NIWA Dunedin (I. Maze) is also gratefully acknowledged.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 


Related articles in JEQ:

This Issue in Journal of Environmental Quality

JEQ 2004 33: 1947-1953. [Full Text]  



This article has been cited by other articles:


Home page
 J. Environ. Qual.Home page
A. N. Sharpley, P. J. A. Kleinman, A. L. Heathwaite, W. J. Gburek, G. J. Folmar, and J. P. Schmidt
Phosphorus Loss from an Agricultural Watershed as a Function of Storm Size
J. Environ. Qual., March 1, 2008; 37(2): 362 - 368.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
R. W. McDowell and R. J. Wilcock
Sources of Sediment and Phosphorus in Stream Flow of a Highly Productive Dairy Farmed Catchment
J. Environ. Qual., March 1, 2007; 36(2): 540 - 548.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
L. E. Gentry, M. B. David, T. V. Royer, C. A. Mitchell, and K. M. Starks
Phosphorus Transport Pathways to Streams in Tile-Drained Agricultural Watersheds
J. Environ. Qual., January 25, 2007; 36(2): 408 - 415.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
L. B. Owens and M. J. Shipitalo
Surface and Subsurface Phosphorus Losses from Fertilized Pasture Systems in Ohio
J. Environ. Qual., May 31, 2006; 35(4): 1101 - 1109.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Related articles in JEQ
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (12)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by McDowell, R. W.
Right arrow Articles by Wilcock, R. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by McDowell, R. W.
Right arrow Articles by Wilcock, R. J.
Agricola
Right arrow Articles by McDowell, R. W.
Right arrow Articles by Wilcock, R. J.
Related Collections
Right arrow Surface Water Quality
Right arrow Water Quality
Right arrow Watershed and Landscape Processes
Right arrow Wetlands and Aquatic Processes
Right arrow Phosphorus

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Vadose Zone Journal Journal of Plant Registrations
Journal of Natural Resources
and Life Sciences Education		 Soil Science Society of America Journal