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


     

Published online 4 January 2008
Published in J Environ Qual 37:69-78 (2008)
DOI: 10.2134/jeq2006.0422
© 2008 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
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 Similar articles in this journal
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 Google Scholar
Google Scholar
Right arrow Articles by Young, E. O.
Right arrow Articles by Briggs, R. D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Young, E. O.
Right arrow Articles by Briggs, R. D.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Young, E. O.
Right arrow Articles by Briggs, R. D.
Related Collections
Right arrow Biogeochemical Processes
Right arrow Nutrients
Right arrow Nutrient Management
Right arrow Ground Water Quality

TECHNICAL REPORTS

Landscape and Watershed Processes

Phosphorus Concentrations in Soil and Subsurface Water: A Field Study among Cropland and Riparian Buffers

Eric O. Younga,* and Russell D. Briggsb

a Dep. of Plant and Soil Science, Univ. of Vermont, Hills Agricultural Science Building, Burlington, VT 05405
b Dep. of Forest and Natural Resources Management, SUNY-ESF, 1 Forestry Drive, Syracuse, NY 13210

* Corresponding author (eoyoung{at}uvm.edu).

Received for publication October 2, 2006.

   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 Materials and Methods
 Results and Discussion
 Conclusion
 REFERENCES
 
Riparian buffers can be effective at removing phosphorus (P) in overland flow, but their influence on subsurface P loading is not well known. Phosphorus concentrations in the soil, soil solution, and shallow ground water of 16 paired cropland-buffer plots were characterized during 2004 and 2005. The sites were located at two private dairy farms in Central New York on silt and gravelly silt loams (Aeric Endoaqualfs, Fluvaquentic Endoaquepts, Fluvaquentic Eutrudepts, Glossaquic Hapludalfs, and Glossic Hapludalfs). It was hypothesized that P availability (sodium acetate extractable-P) and soil-landscape variability would affect P release to the soil solution and shallow ground water. Results showed that P availability tended to be greater in crop fields relative to paired buffer plots. Soil P was a good indicator of soil solution dissolved (<0.45 µm) molybdate-reactive P (DRP) concentrations among plots, but was not independently effective at predicting ground water DRP concentrations. Mean ground water DRP in corn fields ranged from ≤20 to 80 µg L–1, with lower concentrations in hay and buffer plots. More imperfectly drained crop fields and buffers tended to have greater average DRP, particulate (≥0.45 µm) reactive P (PRP), and dissolved unreactive P (DUP) concentrations in ground water. Soil organic matter and 50-cm depth soil solution DRP in buffers jointly explained 75% of the average buffer ground water DRP variability. Results suggest that buffers were relatively effective at reducing soil solution and shallow ground water DRP concentrations, but their impact on particulate and organic P in ground water was less clear.

Abbreviations: DRP, dissolved reactive phosphorus • PRP, particulate reactive phosphorus • TDP, total dissolved phosphorus • DUP, dissolved unreactive phosphorus • WTD, depth to the water table


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
Materials and Methods
 Results and Discussion
 Conclusion
 REFERENCES
 
PHOSPHORUS (P) management for profitable crop production and water quality protection is a concern in the USA and other countries. The mobility of P in the landscape poses water quality concerns because of its contribution to cultural eutrophication (Carpenter et al., 1998). In the USA, agricultural nonpoint P is considered a leading source of impairment to rivers and lakes (USEPA, 2000). Although annual P losses from agricultural land are generally low (<1 to 2 kg P ha–1yr–1), water quality impairments can still occur (Sims and Pierzynski, 2005).

In general, the potential for P loss from the landscape is a function of P availability to water flow and the flow mechanics that control transport to surface waters (McDowell et al., 2004; Gburek et al., 2005). The majority of P losses from cropland are generally attributed to transport in overland flow, but subsurface flows can be important P transport pathways in some settings (Beauchemin et al., 1998; Sims et al., 1998; Simard et al., 2000; Turner and Haygarth, 2000; Gentry et al., 2007). Maguire and Sims (2002) suggested that quantitative prediction of P leaching potential is critical for developing soil P tests that can be used with site hydrology for identifying areas with high leaching potential. Some P leaching studies have reported significant relationships between extractable P and subsurface drainage water P concentrations (Heckrath et al., 1995; Hesketh and Brookes, 2000; Hooda et al., 2000; Maguire and Sims, 2002; McDowell et al., 2002), whereas others have shown no correlation (Haygarth et al., 1997; Turner and Haygarth, 2000; Brye et al., 2001; Kleinman et al., 2003; Akhtar et al., 2003). The high spatial and temporal variability of leaching dynamics confounds attempts to predict P leaching based on P availability and soil water movement, but in general, subsurface P loss is more likely where soil P availability is high and/or manure is applied, and where preferential flow paths exist.

Dissolved and particulate P forms (inorganic and organic) have been reported in subsurface drainage waters at concentrations that could cause potential water quality concerns (Beauchemin et al., 1998; Sims et al., 1998; Heathwaite and Dils, 2000; Simard et al., 2000; Turner and Haygarth, 2000; Uusitalo et al., 2001; van Es et al., 2004; Heathwaite et al., 2005). Dissolved molybdate-reactive P (DRP) is generally assumed to be mostly orthophosphate and highly bioavailable. Subsurface drainage water with elevated DRP could therefore pose water quality concerns if discharged to streams via tile drains and/or discharging ground water. In studying P leaching from grassland soils in the UK, Turner and Haygarth (2000) reported that DRP losses tended to dominate, though particulate P (21 to 46%) losses were also important. The proportion of dissolved and particulate P forms in subsurface water varies widely and is affected by crop management, soil texture, and weather patterns. Simard et al. (2000) suggested that both dissolved and particulate P forms can be important in subsurface drainage water, and that the risk of P transfer to subsurface water could be greater when rainfall is preceded by drier periods. Though several studies have demonstrated P leaching at the field scale, few studies have simultaneously measured ground water and soil pore water P concentrations, and therefore, have provided little information on the potential for P transfer to shallow ground water.

Riparian buffers can reduce P in overland flow emanating from crop fields, but their impact on subsurface P transport is not well known (Uusi-Kamppa et al., 2001). Studies have reported mixed results on the P removal capacity of buffers with respect to shallow ground water, and some studies have reported little or no P removal (Osborne and Kovacic, 1993; Jordan et al., 1993; Clausen et al., 2000; Spruill, 2000; Carlyle and Hill, 2001). Given the widespread use of riparian buffers as a best management practice for reducing P movement from crop fields, information on subsurface P dynamics among cropland-riparian buffer areas is needed for an improved understanding of their role in nonpoint source water quality.

We characterized P concentrations in soil, soil solution, and shallow ground water for 16 paired cropland-buffer plots in central New York during 2004 and 2005. The objectives of the study were to determine if P concentrations in soil and subsurface water differed significantly among cropland (hay and corn) and riparian buffers (recently established grass and willow, and forest buffers), and to examine relationships among P availability, ground water P concentrations, and soil-landscape factors (e.g., drainage class, organic matter content, and redox status).


   Materials and Methods
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
Results and Discussion
 Conclusion
 REFERENCES
 
Study Areas
The study sites were located at two farms in adjacent northward draining valleys in Onondaga County, New York. Paired cropland-buffer plots were established along portions of Spafford Creek (42°82' N; 76°23' W) and Onondaga Creek (42°85' N; 76°13' W) in 2002. Shallow, unconfined aquifers of variable thickness occur in both valleys and contribute to stream discharge (Winkley, 1989). Crop fields are managed in corn silage, alfalfa-grass, and/or permanent reed canary grass hay in various stages of rotation. Corn fields receive surface applied liquid dairy manure in the spring at rates of ~70 to 100 Mg ha–1 (wet basis). Additional information on the soils, cropping, and riparian buffer characteristics at the sites is presented elsewhere (Young and Briggs, 2007).

Riparian Buffers
A cool season grass riparian buffer (timothy and orchard grass) was planted along the east and west sides of Spafford Creek in August 2002. Ten buffer strip plots (9 x 50 m) were then established along both sides of the stream, with plots spaced 100 to 125 m apart on the same stream side. In May 2003, existing grass buffers (predominantly reed canary grass) at the Onondaga Creek site were widened to 9 m, and six buffer strip plots (9 x 50 m) were then established. All recently established grass buffers were then randomly assigned to one of two vegetation treatments consisting of grass alone (n = 6) or in combination with willow (n = 6). Willow cuttings (Salix discolor; variety S365) from the SUNY-ESF Willow Biomass Program were planted in June 2003 following moldboard plowing, disking, and treatment with residual herbicide. Individual willow cuttings were planted in two double rows (0.6 x 0.6 m within double rows, and 0.80 m between double rows) along the stream bank edge for a distance of 50 m. Two sampling plots in existing riparian forest buffers at each site were also established. These buffers were approximately 10 m in width, located on very poorly drained hydric soils, and occupied by black willow, green ash, and red maple.

Soil and Water Sampling
Ground water monitoring wells and tension lysimeters (25- and 50-cm depth) were installed in each plot to sample shallow ground water and soil solution for two seasons (2004 and 2005). Water table depths and ground water dissolved oxygen concentrations were measured periodically both years. Details on soil solution and ground water sampling methods at the sites are described elsewhere (Young and Briggs, 2007). Soil solution and ground water samples were collected in polyethylene bottles and immediately placed on ice following collection. Samples were taken back to the lab after each collection and stored at 5°C until P analysis. Eight randomly selected soil samples were collected with augers (0–30 cm) from each plot (9 x 50 m) in April 2004 (before manure or fertilizer application) and composited. Samples were sieved (2 mm), air-dried, pulverized, and stored in paper bags until chemical analysis.

Soil and Water Chemical Characterization
Air-dried soils were extracted with sodium acetate solution (0.72 N NaOAc+0.52 N CH3COOH; Morgan, 1941) to determine available P (abbreviated as NaOAc-P) and cations following the procedure of Wolf and Beegle (1995). Extracts were analyzed on a PerkinElmer Optima 3300DV ICP–OES (PerkinElmer Corporation, Norwalk, CT) using standard techniques. Soil organic matter was determined by loss on ignition after Bickelhaupt and White (1982). Acid ammonium oxalate-extractable P, Fe, and Al were used to estimate the degree of P saturation (Psat), calculated as the molar ratio of P/(Fe +Al). Some studies have used 0.50[Fe +Al] in the Psat calculation under the assumption that 0.5 mmol P is the maximum amount of P sorbed per mmol of Fe + Al. However, the use of 0.50 is arbitrary and depends on equilibration time and initial P concentrations, and should not be applied unless predetermined (Hooda et al., 2000). Therefore, we calculated Psat based on other studies that determined Psat independent of an {alpha} value (Beauchemin and Simard, 2000; Hooda et al., 2000; Vadas et al., 2005). All soil extractions were performed in duplicate.

Dissolved reactive P (DRP) was measured following filtration (<0.45 µm) and reaction with molybdate (generally within 48 h) using the stannous chloride molybdate method (APHA, 1989). A 10-cm path length was utilized with the spectrophotometer to increase sensitivity. Concentrations of DRP were measured on all soil solution and ground water samples in 2004 and 2005. Due to time constraints, other P forms could not be measured as frequently as DRP. Since particulate and organic P in subsurface waters can be important P forms, particulate (≥0.45 µm) reactive P (PRP) and total dissolved P (TDP) were measured for a subset of samplings (PRP was measured on 3 Sept. 2004, and 5 May, 29 July, 2 Sept., 20 Oct., and 21 Nov. 2005; TDP was measured on 20 June, 29 July, and 20 Oct. 2005). The concentration of PRP was calculated as the difference between filtered and unfiltered molybdate-reactive P (Turner and Haygarth, 2000). Concentrations of TDP were determined on filtered samples using the ammonium persulfate method (Pote and Daniel, 2000). The dissolved unreactive P (DUP) concentration was calculated as the difference between TDP and DRP, which represents organic P (Haygarth and Sharpley, 2000).

Statistical Analysis
A General Linear Modeling (GLM) approach was used for hypothesis testing and variance explanation (SAS Institute, 1999). A partially balanced complete block (by site) Analysis of Variance (ANOVA) was used to test the null hypothesis of equivalent soil solution and shallow ground water P concentrations among corn (n = 8), hay (n = 8), grass buffers (n = 6), grass-willow buffers (n = 6), and riparian forest buffers (n = 4). Average ground water P concentrations among soil drainage classes within cropland and buffers were also analyzed by ANOVA. Least square means were separated by a priori linear contrasts. Linear associations among variables were assessed with Pearson correlation coefficients or the Spearman rank correlation procedure for highly skewed variables. Linear and stepwise multiple linear regression were applied to examine relationships among soil-landscape factors and ground water P concentrations.


   Results and Discussion
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results and Discussion
Conclusion
 REFERENCES
 
General Soil Properties
Soils at the sites were characterized by a range of drainage classes and fertility. Riparian soils tended to be more imperfectly drained with lower NaOAc-P concentrations relative to cropland plots (Table 1 ). Hydric soils (Wayland series) occupied the forest buffers (SB9, SB10, OB5, OB6) and two grass-willow plots (OB1, OB3). The more poorly drained fields and buffers tended to have shallower depths to the water table. The average (from 2004 to 2005) depth to the water table (WTD) was greatest in corn field plots (mean WTD = 1.25 m; SEM = 0.12), and slightly greater than the average depths among hay, grass, and grass-willow plots (range = 0.93–1.17 m). The average WTD in forest riparian buffers (mean WTD = 0.53; SEM = 0.18) was lower (0.0029 ≤ p ≤ 0.097) than all other plots.


View this table:
[in this window]
[in a new window]

  		Table 1. Soil series names, drainage class, and select chemical properties for paired cropland-buffer plots at the Spafford and Onondaga Creek sites.

 
With respect to New York field crop guidelines (Ketterings et al., 2003), 57 and 16% of the samples were in the high (4.5 ≤ mg P kg–1 ≤ 19.5) and very high P categories (≥20 mg P kg–1), respectively (Table 1). Many of the fields at the sites have received long-term annual dairy manure applications, which probably contributed to relatively high P fertility. The OC2 location was characterized by continuous (>10 yr) corn silage, the greatest NaOAc-P concentration (60 mg kg–1), and anomalously high organic matter (113 g kg–1) relative to other plots.

Phosphorus Availability and Soil Solution Dissolved Reactive Phosphorus Concentrations
Available P concentrations (NaOAc-P), soil P sorption saturation (Psat), and soil solution DRP concentrations tended to be greater in corn field plots (Table 1). Buffer plots tended to have lower NaOAc-P, Psat, and soil solution DRP relative to corn field plots (Table 1, 2 ). Much of the variation in soil P availability among crop fields and buffers was likely related to differences in long-term P applications and transport processes occurring between fields and the near-stream areas. This is supported by the fact that corn fields received greater historical P applications (e.g., dairy manure) than hay fields and had correspondingly higher P availability. In addition, there were significant correlations between paired cropland and buffer NaOAc-P concentrations (r = 0.62, p = 0.01) and between cropland and buffer Psat (r = 0.77, p = 0.005), suggesting that the spatial variability of P among crop fields affected the P status of adjoining buffers.


View this table:
[in this window]
[in a new window]

  		Table 2. Mean soil solution dissolved reactive P (DRP) concentrations for 25- and 50-cm depth lysimeters in cropland and buffer plots during 2004 and 2005.

 
Mean soil solution DRP concentrations in corn field plots were lower at the 50-cm depth, but were only significantly lower than 25-cm depths for June (p = 0.05) and September 2005 (p = 0.02) (Table 2). Concentrations of DRP at 25-cm depth were highly correlated (0.74 ≤ r ≤ 0.98, p < 0.0001) with 50-cm depth DRP for individual samplings among plots. There was an overall trend for lower DRP at 50-cm depth, with most of the differences occurring where P availability was greater. Based on linear regression (R2 = 0.94, p < 0.0001) between average 25- and 50-cm depth DRP (2004 to 2005), 50-cm depth concentrations were about 30% lower. Soil Psat and NaOAc-P were significantly correlated with DRP concentrations (0.55 ≤ r ≤ 0.93, p ≤ 0.001) at both depths for all samplings, and the concentration of NaOAc-P provided reasonable curvilinear estimates of the average soil solution DRP and Psat (Fig. 1 ).


Figure 1
View larger version (12K):
[in this window]
[in a new window]

  		Fig. 1. Relationships between sodium acetate-extractable P (NaOAc-P) and average 25- and 50-cm depth soil solution dissolved reactive P (DRP) concentration (a), and soil P sorption saturation percentage (Psat) and NaOAc-P concentration for all plots (b).

 
The concentration of NaOAc-P is used as an index of plant availability for New York agricultural soils (Ketterings et al., 2002). The significant curvilinear relationship between NaOAc-P and soil solution DRP suggests its utility as an index of plant availability and potential release to the soil solution. Kleinman et al. (2000) reported a significant relationship between the concentration of NaOAc-P and the quantity of P extracted by 0.01 M calcium chloride, which has been shown to relate well with P release to water extracts and simulated runoff (Magdoff et al., 1999; Maguire et al., 2005). In our study, the NaOAc-P concentration also provided a reasonable estimate of the Psat, though Psat did not relate as parsimoniously to soil solution DRP as compared to NaOAc-P (data not shown). Due to the high spatial and temporal variability in soil solution DRP observed in this study, the relatively small data set, and the heterogeneity of soil chemical processes, relationships among NaOAc-P, Psat, and soil solution DRP should only be interpreted as potentials of P release to matrix water flow in these soils.

The lack of significant differences between 25- and 50-cm depth soil solution DRP concentrations in cropland and buffers suggests that the average P availability was broadly similar at both depths. Similarly, Kleinman et al. (2003) reported no significant differences in drainage water DRP between 20- and 50-cm depth lysimeters (15-cm diameter cores) before, or following manure application. These authors also reported no correlation between profile P concentrations and drainage water DRP, and concluded that P leaching was likely dominated by preferential flow. In contrast, Maguire and Sims (2002) found high correlations between surface soil P concentrations and drainage water DRP using lysimeters similar to Kleinman et al. (2003). The porous cup tension lysimeters used in our study should have captured mostly matrix flow, and presumably reflected soil water that had greater equilibration time with solid phases relative to water transported preferentially (e.g., water flow in macropores that bypasses the soil matrix). This could have contributed to the relatively high correlation between NaOAc-P and soil solution DRP observed in this study. Additionally, agricultural activity at the sites has homogenized upper horizons. The combination of these factors probably contributed to the similar DRP concentrations between depths, and to the correlation between NaOAc-P and soil solution DRP.

Considerable seasonal variation in soil solution DRP concentrations was observed in 2004 and 2005, with higher average concentrations in 2005. Average 25-cm depth DRP concentrations in corn and hay fields were approximately 1.5- and 2.8-fold greater in 2005, respectively. Since cropping practices were similar both years of the study, it is possible that some of this variation was related to differences in weather patterns (e.g., rainfall and temperature), which affects both P availability and leaching. There was about 25 cm more rainfall during the 2004 early growing season (May 1 to July 31) compared to 2005. Drying and rewetting cycles can influence P solubility by altering inorganic and organic P, and rewetting of dry soils can cause aggregate breakdown, which can expose additional P within aggregates (Nevo and Hagin, 1966; Bartlett and James, 1980; Amezketa, 1999). Turner and Haygarth (2003) reported that drying soils increased inorganic and organic bicarbonate-extractable P in pasture soils up to 165 and 137%, respectively, and attributed the P increases to enhanced solubility of organic matter and P upon drying. It is possible that the greater soil solution DRP concentrations observed in 2005 at our study sites may have been related to increased P availability and/or lower P losses from the profile caused by the drier conditions.

Phosphorus Forms and Concentrations in Shallow Ground Water
Mean concentrations of DRP in shallow ground water were low and generally <20 µg L–1 during 2004, with no significant differences among cropland and buffer plots (Fig. 2 ). The exception was the 8 July 2004 sampling when DRP in hay fields was >50 µg L–1 following about 5.0 cm of rainfall on July 7. Mean DRP concentrations in corn fields showed up to eightfold increases in 2005, and were significantly higher than other plots for five out of the six samplings (Fig. 2). Due to the relatively high variation in DRP, PRP, and DUP concentrations, average concentrations over the study were used to compare differences among plots and soil drainage classes (Fig. 3 , 4 ). Average DRP concentrations were highest (0.005 ≤ p ≤ 0.07) in corn field plots, while DUP and PRP tended to be greater (0.009 ≤ p ≤ 0.33) in forest buffers (Fig. 3). Average P concentrations in cropland plots were about two to five times greater in somewhat poorly drained (SPD) soils compared to well (WD) and moderately well drained (MWD) fields, while poorly drained (PD) buffers had from 1.3- to 9.5-fold greater average concentrations of P in ground water compared to the SPD and WD buffers (Fig. 4).


Figure 2
View larger version (27K):
[in this window]
[in a new window]

  		Fig. 2. Mean ground water dissolved reactive P (DRP) concentrations among cropland and buffer plots in 2004 (a) and 2005 (b). Means within sampling date among cropland and buffers with different letters are different at p ≤ 0.05. Error bars are estimated standard errors. Note that the DRP concentration was below the limit of detection in grass buffers for the 27 Oct. 2005 sampling.

 

Figure 3
View larger version (14K):
[in this window]
[in a new window]

  		Fig. 3. Mean dissolved reactive phosphorus (DRP), particulate reactive P (PRP), and dissolved unreactive P (DUP) in the shallow ground water of cropland and buffer plots. Means for each plot were calculated from study average DRP (n = 13), PRP (n = 5), and DUP (n = 3). Means within each P form with different lowercase letters are significantly different at p ≤ 0.07. The number of plots for corn, hay, grass, grass-willow, and forest buffers is 8, 8, 6, 6, and 4, respectively.

 

Figure 4
View larger version (19K):
[in this window]
[in a new window]

  		Fig. 4. Average dissolved reactive phosphorus (DRP), particulate reactive P (PRP), and dissolved unreactive P (DUP) concentrations by soil drainage class in cropland (a) and buffer plots (b). Means for each drainage class were calculated from study average DRP (n = 13), PRP (n = 5), and DUP (n = 3). Means within each P form with different lowercase letters are significantly different for crop fields and buffers at p ≤ = 0.09 and p ≤ 0.03, respectively. The number of plots for the well, moderately well, and somewhat poorly drained crop fields is 6, 3, and 7, respectively; for the well/moderately well, somewhat poorly, and poorly drained buffers the number of plots is 3, 7, and 6, respectively. Error bars are the standard error of the mean.

 
Linear correlations among soil P measures (Psat, NaOAc-P, soil solution DRP) and average ground water DRP concentrations for cropland plots were weak. Stepwise multiple linear regression selected 50-cm depth DRP as the only significant (r2 = 0.41, p = 0.02) predictor of average cropland ground water DRP. However, the distribution of model residuals was suspect and nonparametric correlation was weak (Spearman r = 0.37, p = 0.16). There were significant correlations between buffer ground water DRP and buffer 25-cm depth DRP (r = 0.68, p = 0.004) and 50-cm depth DRP (r = 0.67, p = 0.005). Soil organic matter (OM) content in buffer soils was also significantly correlated with average (2004 to 2005) buffer ground water DRP (r = 0.68, p = 0.003) and PRP concentration (r = 0.57, p = 0.02). Soil OM and average 50-cm depth DRP jointly explained 75% of the average buffer ground water DRP variability by multiple linear regression (Fig. 5 ).


Figure 5
View larger version (16K):
[in this window]
[in a new window]

  		Fig. 5. Relationship between observed average buffer ground water dissolved reactive P (DRP) and predicted buffer ground water DRP by multiple linear regression with buffer soil organic matter (OM) and 50-cm depth DRP as independent variables. Average buffer ground water DRP = 0.630 + 0.091 (50-cm depth DRP) + 1.21 (soil OM), R2 = 0.75, n = 16, p = 0.0002, F = 18.1.

 
Ground Water Dissolved Reactive Phosphorus Trends in Cropland
Average concentrations of ground water DRP tended to remain low in hay field and buffer plots for both years of the study, but DRP in corn fields showed large increases in 2005. Since cropping and fertilization practices were similar both years, other factors may have contributed to the differences. Rainfall and soil moisture during 2004 and 2005 varied substantially and could have altered soil water fluxes to shallow ground water. The soils also had relatively high proportions of silt and clay (mean silt + clay = 70%, SD = 10) and were subject to cracking, as was evident on the soil surface and along profile walls of small pits observed during drier conditions. Some research has reported that finer-textured soils can display elevated leaching losses due to greater preferential water flow through macropores (e.g., cracks, fissures, biopores) (Turtola and Jaakkola, 1995; Beauchemin et al., 1998; Stamm et al., 1998; van Es et al., 2004). The corn fields at the Spafford Creek site are also tile drained, which can exacerbate preferential flow in the vicinity of the tiles (Lennartz et al., 1999).

The two highest ground water DRP concentrations in 2005 occurred following drier periods. Rainfall in May (–7.0 cm) and June (–4.4 cm) were below the long-term average (recorded at Hancock International Airport, Syracuse, NY), and most of the rainfall for June occurred the week before ground water sampling on 23 June. Similarly, the 27 October sampling was preceded by below average rainfall conditions (–6 cm in September) and had sustained rainfall before sampling. Research with fine-textured cropland soils in Canada showed that dissolved and particulate P concentrations in subsurface drainage water can increase when rainfall followed dry periods (Beauchemin et al., 1998; Simard et al., 2000). The authors hypothesized that preferential flow through cracks and fissures formed during drier conditions increased preferential transport of P to tile drains. At our sites, it is possible that the drier conditions in 2005 caused greater overall solute movement to ground water due to a greater tendency for preferential flow in some of the corn fields. In addition, soil solution DRP concentrations also tended to be higher in 2005, which would have increased the amount of P available for leaching to ground water. The fact that hay plots did not show elevated DRP in 2005 could be partially related to the lower soil P levels, and because hay fields did not receive dairy manure in the spring, which would have also decreased the amount of P available for leaching relative to corn field plots.

The lack of correlation between soil P and ground water DRP in corn fields supports the idea that preferential flow paths may have contributed to the elevated ground water DRP in 2005. However, more imperfectly drained crop fields tended to have greater ground water DRP concentrations. Since the depth to the water table reflected soil drainage classes for these sites, ground water in more imperfectly drained fields interacted to a greater degree with upper soil horizons, where soil P availability would tend to be greater. Therefore, in addition to possible preferential flow pathways, there could have been the potential for some direct P mobilization by ground water in these fields during elevated water table conditions.

Ground Water Dissolved Reactive Phosphorus Trends in Buffers
Average riparian ground water DRP was largely explained by the average 50-cm depth soil solution DRP and OM content of buffer soils, suggesting that soil-landscape variables such as soil series, organic carbon, and available P levels were important influences on ground water DRP concentrations. More imperfectly drained riparian soils tended to have greater concentrations of OM and oxalate-extractable Fe and Al, and there were significant correlations between soil OM and oxalate Fe (r = 0.80, p = 0.0002) and Al (r = 0.62, p = 0.004). Young and Briggs (2007) showed that the more poorly drained riparian soils (as measured by the depth to water table) at these sites had significantly lower dissolved oxygen and nitrate concentrations, and significantly greater soil OM, suggesting that ground water in more imperfectly drained buffers was more reduced, with presumably lower redox potentials. Low redox conditions in soil porewaters and aquifer sediments can increase the likelihood of P release from Fe-P compounds (Carlyle and Hill, 2001; Young and Ross, 2001; Chardon and Schoumans, 2002; Meissner et al., 2002; Pierzynski et al., 2005; Sims and Pierzynski, 2005). For the buffers in this study, there was also a weak negative correlation between WTD and DRP (r = –0.47, p = 0.06). While it is possible that some DRP was released to ground water upon reduction of Fe-P compounds in the more poorly drained buffers, P release from other compounds (e.g., Ca-P, Al-P, organic P complexes) in soils and aquifer sediments would also be important. Since the poorly drained buffers had the shallowest average depth to the water table, ground water interacted to a greater degree with upper soil horizons relative to more well drained buffers. Upper soil horizons tend to be enriched with organic carbon and P relative to subsurface horizons, and therefore, ground water in these buffers could have been exposed to greater total P concentrations, thereby increasing the relative likelihood of some P release to shallow ground water. The combination of these factors probably contributed to the greater average DRP concentrations observed in the more poorly drained buffers.

Dissolved molybdate-reactive P is thought to be largely orthophosphate and considered readily bioavailable. Whether or not shallow ground water DRP contributes P to stream discharge is dependent on the ground water hydrology, the DRP concentration in the volume of discharging ground water, the DRP concentration in overlying stream water, and the biogeochemical conditions along flow paths and the hyporheic zone (Spruill, 2000). Carlyle and Hill (2001) concluded that riparian ground water would function as a net source of DRP to streams because riparian ground water DRP was higher than stream water DRP. Study average stream and riparian ground water DRP at the Spafford Creek site was 7.0 (SD = 5.0) and 4.8 µg P L–1 (SD = 1.6), respectively. At the Onondaga Creek site, the average stream water (for the main tributary) and riparian ground water DRP concentrations were 22.0 (SD = 13.0) and 8.4 µg P L–1 (SD = 3.1), respectively. Based on these averages and the fact that agricultural activity upstream from the sites contributed runoff to stream water, discharging ground water was probably not a major source of DRP to the streams.

Ground Water Particulate Reactive Phosphorus, Dissolved Unreactive Phosphorus and Total Phosphorus Trends
Similar to the DRP trends, more imperfectly drained soils tended to have greater concentrations of PRP and DUP in ground water. Somewhat poorly drained crop fields had greater average PRP (46 µg P L–1) and DUP (72 µg P L–1) concentrations compared to PRP (6 and 16 µg P L–1) and DUP (34 and 39 µg P L–1) in moderately well and well drained fields, respectively. Poorly drained buffers also had greater average PRP and DUP concentrations relative to moderately well and somewhat poorly drained buffers (Fig. 4). More imperfectly drained buffers tended to have greater OM and amorphous Al and Fe, which readily form complexes with P. There was a weak correlation (r = 0.57, p = 0.02) between soil OM and average PRP in buffers, and a negative correlation (r = –0.63, p = 0.01) between average water table depth and PRP. It is possible that the shallower depth to the water table in more imperfectly drained buffers increased the likelihood of particulate and/or organic P release to shallow ground water. In addition, P bound to particles, colloids, and/or organic complexes is thought to be more mobile in the soil solution and ground water compared to orthophosphates due to decreased affinity for sorption sites (Gschwend and Reynolds, 1987; Hens and Merckx, 2001; Heathwaite et al., 2005; Condron et al., 2005). Other studies have reported that a significant fraction of the total P in subsurface drainage can be comprised of organic and/or particulate P (Heathwaite and Dils, 2000; Uusitalo et al., 2003). The bioavailability of particulate and organic P forms is not well known, but research has suggested that some fraction of particulate P could be bioavailable (Uusitalo et al., 2003). The particulate P measured in our study was molybdate-reactive, and it is generally assumed that much of the P that reacts with molybdate is inorganic, thus some of the PRP may have been bioavailable.

Since only filtered ground water samples were digested in this study, there was not a true measure of total P concentration (e.g., unfiltered samples are digested in the standard method for total P determination). Therefore, we used the sum of the average concentrations of DRP, DUP, and PRP as an estimate of average total P concentration.

Average total P in forest buffers (205 µg P L–1) was two- to threefold greater than concentrations in corn (97 µg P L–1), hay fields (97 µg P L–1), grass (69 µg P L–1), and grass-willow buffers (62 µg P L–1). Though total P concentrations in grass buffers and grass-willow buffers were not significantly different than concentrations in corn and hay field plots, these data suggest that the recently established grass and grass-willow buffers tended to reduce P in shallow ground water flow from cropland, and that they were more effective than the forest buffers sampled in this study. Other studies have reported little or no P removal from shallow ground water flow in riparian buffers (Jordan et al., 1993; Clausen et al., 2000; Dosskey, 2001).

The USEPA guideline for ground water discharging to lakes and streams is 100 µg L–1 of total P. Since the total P estimates used in this study are probably low, some of the total P concentrations could be considered a potential water quality concern. At the Spafford and Onondaga Creek sites, average total P concentrations in riparian ground water and stream water were 124 µg L–1 and 72 µg L–1, and 59 µg L–1 and 165 µg L–1, respectively. Both sites had agricultural activities upstream from the study areas that contributed runoff to the streams, which likely contributed largely to the P loads. Though these P concentrations were variable and reflect considerable uncertainty, they do provide a relative indication of potential P inputs to stream water from discharging ground water. Based on these averages, discharging ground water could be a possible source of P to Spafford Creek, whereas ground water at the Onondaga Creek site would be considered a smaller source of P given the much larger total P concentration in the stream water.


   Conclusion
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results and Discussion
 Conclusion
REFERENCES
 
Concentrations of P were characterized in soil, soil solution, and shallow ground water for 16 paired cropland-buffer plots over two seasons. Results showed that soil P availability was a reasonable indicator of soil solution DRP concentrations among plots, but was not a good predictor of shallow ground water DRP concentrations. Buffers tended to maintain lower soil solution and ground water DRP concentrations relative to paired cropland plots. Average shallow ground water DUP and PRP concentrations were not significantly different among cropland and buffers. Buffers in this study appeared to be more effective at reducing DRP in shallow ground water flow as compared to particulate and organic P. Greater total P concentrations in ground water were associated with more imperfectly drained buffers, suggesting that soil-landscape factors that enhance nitrate removal from shallow ground water (e.g., poor drainage, organic carbon accumulation, low redox potential) may not be as effective for P removal.


   ACKNOWLEDGMENTS
 
We thank the USEPA for funding, the Snavlin and CoVale dairy farms, Parratt-Wolff Drilling Inc., Drs. Tim Volk and Larry Abrahamson, Ken Burns, Don Bickelhaupt, USDA-NRCS, and the Onondaga County Soil and Water Conservation District.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results and Discussion
 Conclusion
 REFERENCES
 
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
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results and Discussion
 Conclusion
 REFERENCES
 




This article has been cited by other articles:


Home page
 Agron. J.Home page
J. L. Kovar and N. Claassen
Growth and Phosphorus Uptake of Three Riparian Grass Species
Agron. J., September 1, 2009; 101(5): 1060 - 1067.
[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 Similar articles in this journal
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 Google Scholar
Google Scholar
Right arrow Articles by Young, E. O.
Right arrow Articles by Briggs, R. D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Young, E. O.
Right arrow Articles by Briggs, R. D.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Young, E. O.
Right arrow Articles by Briggs, R. D.
Related Collections
Right arrow Biogeochemical Processes
Right arrow Nutrients
Right arrow Nutrient Management
Right arrow Ground Water Quality

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