JEQ Journal of Natural Resources and Life Sciences Education
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 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 (4)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Grey, M.
Right arrow Articles by Henry, C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Grey, M.
Right arrow Articles by Henry, C.
Agricola
Right arrow Articles by Grey, M.
Right arrow Articles by Henry, C.
Related Collections
Right arrow Watershed and Landscape Processes
Right arrow Surface Water Quality
Right arrow Nitrogen
Right arrow Phosphorus
Right arrow Municipal Waste
Journal of Environmental Quality 31:926-936 (2002)
© 2002 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America

TECHNICAL REPORTS
Surface Water Quality

Phosphorus and Nitrogen Runoff from a Forested Watershed Fertilized with Biosolids

Mark Grey*,a and Chuck Henryb

a Synagro West, Inc., Box 7027, Corona, CA 92878-7027
b College of Forest Resources, Box 352100, Univ. of Washington, Seattle, WA 98195

* Corresponding author (mgrey{at}synagro.com)

Received for publication February 14, 2001.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Municipal biosolids are typically not used on the steepest of forested slopes in the U.S. Pacific Northwest. The primary concern in using biosolids on steep slopes is movement of biosolids particles and soluble nutrients to surface waters during runoff events. We examined the pattern and extent of P and N runoff from a perennial stream draining a small, forested 21.4-ha watershed in western Washington before and after biosolids application. In this study, we applied biosolids at a rate of 13.5 Mg ha-1 (700 kg N ha-1 and 500 kg P ha-1) to 40% of the watershed following nearly 1.5 years of pre-application water sampling and 1.5 years thereafter. There was no evidence of direct runoff of P or N from biosolids into surface water. Elevated surface water discharge did not change the concentration of PO4–P, biologically available phosphorus (BAP), bioavailable particulate phosphorus (BPP), or total P nor did it affect the concentration–discharge relationship. Some instances of total P concentrations exceeding the USEPA surface water standard of 0.1 mg L-1 were observed following biosolids application. However, total P in 27 Creek was predominately in particulate form and not labile, suggesting that detritus moving into the main creek channel and ephemeral drainage courses may be the principal P source. Ammonium N concentrations in runoff water were consistent before and after biosolids application, ranging from below detection limits (0.01 mg L-1) to 0.1 mg L-1; no concentration–discharge relationship existed. Biosolids application changed the 27 Creek concentration–discharge relationship for NO-3–N. Before application, no relationship existed. Beginning nine months after biosolids application, increases in discharge were positively related to increases in NO-3–N concentrations. Nitrate concentrations in runoff following biosolids application were approximately 10 times less than the USEPA drinking water standard of 10 mg L-1.

Abbreviations: BAP, biologically available phosphorus • BPP, bioavailable particulate phosphorus • PNW, Pacific Northwest


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
APPLYING BIOSOLIDS TO FORESTED sites in the U.S. Pacific Northwest (PNW) increases the growth of Douglas-fir [Pseudotsuga menziesii (Mirb.) Franco], especially on low site index stands (Henry et al., 1993, 1994). The reason for positive growth response is the fertilizing effect of biosolids and its ability to supply nitrogen (N) and other plant essential macro- and micronutrients. Because of this evidence, there is interest in using biosolids to fertilize forests and reclaim disturbed lands located in diverse terrain in the PNW, including steeply sloped areas. Pure biosolids without composting are typically not applied to steep slopes due to concerns about runoff of biosolids into receiving waters. Direct particle movement and nutrient losses are primary considerations. Phosphorus losses from treated watersheds could lead to surface water eutrophication (Correll, 1998) and N inputs to streams may contribute to eutrophication as well as increased periphyton production (Bisson et al., 1992). In forested ecosystems in the PNW, nutrient and sediment export can be accelerated by heavy rainfall (Fredriksen et al., 1975). Heavy and prolonged rainfall induces transport of ions through the soil and mobilizes particulate matter in surface water runoff. Biosolids add high concentrations of nutrient elements to the forest floor in both soluble and particulate forms. The introduction of potentially mobile constituents may, therefore, change the amount and pattern of export for many elements including P and N.

The University of Washington monitored water quality in two small 1.2-ha treated and control watersheds at Pack Forest, Washington following application of liquid biosolids at a rate of 45 Mg ha-1 (University of Washington, 1986). Investigators found no statistically significant difference in organic or mineral forms of N and P in runoff waters comparing treated and untreated watersheds. A 15-m buffer between the stream and biosolids was used and the monitoring period was approximately one year following application. Kimmins et al. (1991) applied liquid biosolids at a rate of 11 Mg ha-1 to a 3.6-ha area within the East Creek watershed in British Columbia, Canada. Concentrations of total N, NO-3–N, NH+4–N, and total and PO4–P in East Creek and ephemeral drainages were almost equal to or below pre-biosolids application levels for one year following application. The study of runoff from biosolids application on a plot scale has been done in agriculture and arid range settings (Aguilar and Loftin, 1992; Bruggeman and Mostaghimi, 1993; Dunigan and Dick, 1980; Harris-Pierce et al., 1995; Kelling et al., 1977; Kladviko and Nelson, 1979). However, reports in forestry literature are lacking. An inherent problem in comparing data on runoff water quality from plot scale or range studies is the vastly different soils examined, differences in rainfall amounts and intensities used, and the contact distance between biosolids and collection devices before collecting runoff waters. For the study of runoff patterns from forests, it is difficult to locate replicate watersheds.

The intent of this research was to examine the mobility of P following biosolids application to a young Douglas-fir forest within a steep, forested watershed. Phosphorus mobility in the soil is thought to be strongly limited in acidic, PNW forest soils by sorption onto amorphous oxides of iron and aluminum (Johnson et al., 1986). The mobile nature of N following biosolids application to PNW forests has been well documented (Henry et al., 1999). For this reason, N was also included in the watershed study for contrast with P. The objectives of the study were to (i) establish if surface water runoff is a pathway of P and N movement from a biosolids application site located on steep terrain and (ii) determine if runoff events induce movement of P and N in biosolids via surface runoff. The experiment described herein uses a before and after approach and tests the assumption that if constituents in biosolids are to move from a site, movement would occur during runoff events that are induced by heavy rainfall and high soil moisture conditions. Under heavy rainfall conditions, stream discharge increases and the source area for runoff water expands within a watershed. This situation could be encountered anytime between October and June in watersheds of all sizes throughout western Washington.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Study Site Description
The 27 Creek watershed is located within the University of Washington's 1720-ha Charles Lathrop Pack Demonstration Forest (46°51' N, 122°18' W; Fig. 1) . Pack Forest is at the base of the Cascade foothills about 110 km south of Seattle. Dominant vegetation on the forest is second-growth Douglas-fir with some western hemlock [Tsuga heterophylla (Raf.) Sarg.], western red cedar (Thuja plicata Donn ex D. Don), and red alder (Alnus rubra Bong). The climate is typical maritime, with relatively dry summers and wet winters, and moderate temperatures throughout the year. Annual precipitation is about 1200 mm, with approximately 50% falling between October through January (Pack Forest precipitation records). Rainfall during the period of July through August is usually less than 120 mm, often resulting in drought-like conditions on well-drained soils. During periods of normal temperature and precipitation, evapotranspiration is estimated to be between 380 to 560 mm annually (Henry, 1989).



View larger version (65K):
[in this window]
[in a new window]
  		Fig. 1. The 27 Creek watershed, Pack Forest, Washington.

 
The study site is a 21.4-ha headwater watershed and is drained by a first-order perennial stream (27 Creek). The watershed ranges between 420 and 600 m in elevation with a northerly aspect (Fig. 1). Topography is steep, with some slopes exceeding 60%. Nearly all of the watershed was clear-cut in 1982 and replanted the following year, and consists of 18-yr-old second-growth Douglas-fir intermixed with naturally regenerated western hemlock and red alder. Most of the watershed was pre-commercially thinned in March and April 1996; felled trees were left in place.

The soil within the watershed is mapped as the Wilkeson soil series (fine-loamy, isotic, mesic Vitrandic Haploxeralf). Forest soils within Pack Forest fall into two general types: soils formed in relatively unweathered glacial outwash and more weathered soils formed in residuum (andesite and basalt). Soils formed in glacial outwash are predominate in elevations below 300 m, while those formed in residuum (andesite and basalt) are located at elevations greater than 300 m. Forest soils in the PNW are generally nitrogen limited (Walker and Gessel, 1991) and, at Pack Forest, soils formed in both outwash and residuum have a large capacity to sorb P (Grey, 1999; Johnson et al., 1986).

Biosolids Characteristics and Application
Anaerobically digested, dewatered (approximately 20% solids) biosolids from King County's South Treatment Plant were used. The biosolids contained 320 g C kg-1, 53 g total N kg-1, 12 g NH+4–N kg-1, 36 g total P kg-1, 6 g water-soluble P kg-1, and 4 g bicarbonate P kg-1 on a dry weight basis. The pH at the time of application was 8.5. Biosolids samples were analyzed after air-drying for one week and light grinding to pass through a 2-mm sieve. Biosolids were applied to 8.4 of 21.4 hectares of the watershed during the third week of January 1997 (95% of total application area) and the first week of May 1997 (5% of total application area) to the hatched areas shown in Fig. 1 at a rate of 13.5 Mg ha-1. Wet conditions and unstable roads in January 1997 prevented application from being completed at one time due to safety concerns. The application rate was based on N demand for a 15-yr-old Douglas-fir stand on soils never fertilized with biosolids. The approximate loading rates are 700 and 500 kg ha-1 for N and P, respectively. A total buffer distance around 27 Creek and ephemeral drainages was maintained at 20 m.

27 Creek Watershed Hydrology and Runoff Event and Baseflow Separation
Creek discharge was measured where 27 Creek exited the watershed. An ISCO (Lincoln, NE) Model 4120 submerged probe flow logger was installed inside a 61-cm-diameter galvanized culvert. Discharge was determined using the Manning equation fitted for closed conduits. A roughness coefficient of 0.24 was used based on corrugated surface characteristics of the culvert exiting the watershed. Calibration was done using a ruler to measure water level over the top of the submerged probe. Frequent recalibration was performed and observed readings were compared with measured discharge using a 121-L plastic container and stop watch. The flow logger was integrated with an ISCO 3700 portable water sampler and Model 510 rain gauge located in a clearing near the culvert. Daily average discharge and daily rainfall for water years 1996–1997 and 1997–1998 are shown in Fig. 2 and 3 . Several runoff events were selected for analysis, with an attempt made to include at least one runoff event per season exclusive of mid-July–September rainstorms (during this time, relatively little runoff is generated within the 27 Creek watershed). Runoff events chosen for analysis were separated from baseflow using the technique of Hewlett and Hibbert (1967). From the point of initial hydrograph rise, a line sloping upward at a rate of (1.42 L s-1) x 2.59(km2) is plotted and extended until it intercepts the hydrograph. This method is acceptable for separating runoff hydrographs in watersheds less than 52 km2 (Hewlett and Hibbert, 1967).



View larger version (23K):
[in this window]
[in a new window]
  		Fig. 2. Discharge and rainfall in 27 Creek for the water year 1996–1997.

 


View larger version (24K):
[in this window]
[in a new window]
  		Fig. 3. Discharge and rainfall in 27 Creek for the water year 1997–1998.

 
Water Quality Sampling and Analytical Methods
From November 1995 until November 1998, water samples were removed at least once a month by taking a grab sample at a single point in 27 Creek, followed by analysis for total P and total N (total N = organic N plus NH+4–N and NO-3–N). Beginning in October 1996 numerous rainfall events causing a substantial rise in the stream hydrograph initiated automatic water sampling. After sampler initiation, water samples remained in the automatic sampler on ice until removal and transport. Water samples were also removed periodically during baseflow conditions after October 1996. After October 1996, water samples were analyzed for total P, PO4–P, BAP, total N, NH+4–N, and NO-3–N. Water analyses were performed by the King County Water and Land Resources Division Environmental Laboratory in Seattle, Washington.

Total N and P on unfiltered water samples as well as PO+4–P, NH4–N, and NO-3–N were determined using standard methods for the examination of water (Clesceri et al., 1994). Biologically available phosphorus (BAP) was determined by filtering a known volume of 27 Creek water (0.45-µm filter) and analyzing the extract for orthophosphate (Clesceri et al., 1994). The particulate material remaining on the filter paper is extracted in a dilute sodium hydroxide–sodium chloride solution overnight, neutralized, filtered, and analyzed for orthophosphate. The sum of both determinations is BAP.

Total biosolids N and C were determined by dry combustion (PerkinElmer [Wellesley, MA] CHN Analyzer Model 2400). Total biosolids P was determined by digesting samples in HNO3–H2O2–HCl and determining P colorimetrically (Murphy and Riley, 1962) after neutralization of sample aliquots. Water-soluble and bicarbonate P were determined sequentially by equilibrating a 0.5-g sample of biosolids in deionized water for 16 h, centrifuging at 15000 rpm, decanting the supernatant, and repeating the procedure using 0.5 M NaHCO3. Labile P was determined by analyzing both extracts colorimetrically (Murphy and Riley, 1962). A PerkinElmer Model 55E spectrophotometer was used for all colorimetric analyses using a wavelength of 880 nm.

Data Transformation and Statistical Analysis
Phosphorus and nitrogen concentrations in 27 Creek water (mg L-1) were log transformed in order to meet normality assumptions necessary in exploratory statistical analyses. For examination of concentration and discharge relationships, both linear and nonlinear regression analyses were performed on untransformed and log-transformed data. Flow-weighted P and N parameters were calculated using the following formula:

where Ci = concentration (mg L-1) at time 1 (t1), t2...ti; Qi = discharge at t1, t2...ti; and {sum}Qi = sum of discharge over ti.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
Concentration Range of Phosphorus in 27 Creek Before and After Biosolids Application
The frequency distributions of dissolved PO4–P, BAP, and total P concentrations in 27 Creek before and after biosolids application are shown in Fig. 4 . The data are untransformed and data points represent either a single grab or automatic sample taken from 27 Creek from October 1996 to November 1998 for PO4–P and BAP, and from October 1995 to November 1998 for total P. The median concentration of PO4–P and BAP decreased after biosolids application (Fig. 4A). The generally narrow range of labile P concentrations in 27 Creek suggest that biosolids have had no effect on 27 Creek labile P concentrations during the study period. It is important to recognize that for all parameters analyzed in this study, more post-application samples than pre-application samples were taken. Because of the lack of background data prior to biosolids application, it is possible that natural variation in 27 Creek P concentrations was not accounted for during the study. Thus, the normal range of P concentrations could be lower or higher than detected during this study and this must be accounted for in drawing conclusions regarding the effect of biosolids application.



View larger version (14K):
[in this window]
[in a new window]
  		Fig. 4. (A) Frequency distribution of PO4–P and biologically available phosphorus (BAP) and (B) total P concentrations in 27 Creek before and after biosolids application. Horizontal lines above and below each box represent the 10th and 90th percentiles of each parameter, while the upper, lower, and middle line within each box represent the 75th, 25th, and 50th percentiles, respectively. The arithmetic mean of each sample group is represented by the black square.

 
The median and mean total P concentrations in 27 Creek are higher after application than before application (Fig. 4B), while the range of concentrations increased. The variation may be due in part to sample collection differences with respect to 27 Creek discharge at the time of sampling. When the average monthly flow-weighted total P concentration is plotted over time (Fig. 5) , the increase in total P is less apparent, but a seasonal trend develops in elevated total P concentrations during spring and autumn months. Interpreting the trend of increases in 27 Creek total P concentration after biosolids application is difficult. Runoff events prior to October 1996 were not monitored, and total P was not determined for some of the runoff events during early autumn 1996. An analysis of post-application data shows that elevated total P concentrations were noted primarily during spring months and occasionally during periods of increased discharge from the watershed. The highest total P concentration noted before application was approximately 0.1 mg L-1 in early November 1996, while the highest concentration after application was fourfold greater (0.4 mg L-1) in May 1997. Natural variation in P movement and export from the watershed is expected and it appears that certain periods of increased discharge are important in total P export. Thus, the limited number of storm events monitored prior to application and exclusion of spring events may not adequately represent the pattern of total P movement from the watershed.



View larger version (23K):
[in this window]
[in a new window]
  		Fig. 5. Monthly average total P and N concentrations in 27 Creek, November 1995–October 1998.

 
Concentration Range of Nitrogen in 27 Creek Before and After Biosolids Application
The frequency distributions of NH+4–N and NO-3–N concentrations before and after biosolids application are shown in Fig. 6 . Ammonium N has remained relatively constant, ranging from below the detection limit (0.01 mg L-1) to 0.1 mg L-1. The distribution of NH+4–N includes zero values in Fig. 4A to reflect the frequency of concentrations falling below the detection limit. Despite fertilization of the watershed with nitrogen and ammonium-rich biosolids, no change in 27 Creek NH+4–N concentrations was observed at any time after application. No seasonal trends for NH+4–N were observed. Consistent NH+4–N concentrations in surface water runoff water may be due to moderate volatilization losses of ammonia N in the days and weeks following application as well as NH+4–N adsorption by the soil. After a winter application of biosolids, uptake of NH+4–N by plants and immobilization by microorganisms following application may be minor compared with sorption reactions; over time the importance of these processes reverses and uptake and microbial immobilization limit NH+4–N mobility. At the time of application, ammonia N represented approximately 15% of the NH+4–N contained in biosolids. Watershed scale studies evaluating water quality effects of urea N fertilization have shown much more dramatic increases in stream water NH+4–N concentrations. For example, Hetherington (1985) reported peak NH+4–N concentrations of up to 1.5 mg L-1 twenty-four hours after fertilization, with concentrations returning to pre-fertilization conditions within 14 d. Elevated levels of NH+4–N in streamwater have been reported to persist following urea fertilization of watersheds dominated by Douglas-fir for several months (Bisson et al., 1992). Ammonia N concentrations in surface water (temperature and pH dependent) greater than 0.1 mg L-1 generally indicate polluted water.



View larger version (15K):
[in this window]
[in a new window]
  		Fig. 6. Frequency distribution of (A) NH+4–N and (B) NO-3–N concentrations in 27 Creek before and after biosolids application. Horizontal lines above and below each box represent the 10th and 90th percentiles of each parameter, while the upper, lower, and middle line within each box represent the 75th, 25th, and 50th percentiles, respectively. The arithmetic mean of each sample group is represented by the black square.

 
Nitrate N concentrations in 27 Creek increased after application of biosolids (Fig. 6B). Pre-application concentrations ranged from 0.05 to 0.5 mg L-1, while post-application concentrations were as high as 1.5 mg L-1. The effect of biosolids on increasing 27 Creek NO-3–N concentrations was observed approximately nine months after the January 1997 application, and elevated concentrations (approximately 1.0 mg L-1) generally persisted during the study. The increase in NO-3–N concentrations is reflected in a marked increase in the average monthly concentration of total N in 27 Creek after biosolids application (Fig. 5). Elevated NO-3–N levels in streams draining watersheds fertilized with urea N have been reported to persist for a year or more (Bisson et al., 1992); in many instances buffers around drainages were not used. While specific nitrogen transformations were not measured following biosolids application, an increase in 27 Creek NO-3–N concentration appears to be a result of biosolids decomposition, mineralization of organic N, and nitrification of NH+4–N, which has elevated the concentration of NO-3–N supported in the soil solution.

Runoff Event Phosphorus and Nitrogen Concentrations Before and After Biosolids Application
Phosphorus
Several runoff events were monitored before and after biosolids application, with at least one runoff event during each season selected for analysis except during low-flow, July–October conditions. More events were monitored after biosolids application than before it. Flow-weighted concentrations for PO4–P, BAP, and total P during selected runoff events are shown in Table 1. Flow-weighted PO4–P and BAP concentrations during runoff events were consistent throughout the study. Phosphate P concentrations peaked in early October 1996 at 0.07 mg L-1 during a small autumn runoff event (data not shown) but dropped at this point and remained relatively constant between 0.01 to 0.05 mg L-1. The BAP concentrations were consistent throughout the study and over several seasons; rarely did flow-weighted concentrations exceed 0.03 mg L-1.


View this table:
[in this window]
[in a new window]
  		Table 1. Flow-weighted phosphorus and nitrogen concentrations in 27 Creek during runoff event conditions, 1996–1998.

 
Both dissolved PO4–P and BAP are useful in assessing the potential effect of biosolids application on 27 Creek water quality during runoff events. The average concentration of water-soluble PO4–P in biosolids used in this study is 6 g kg-1, while that of labile P (analogous to BAP in natural waters) is approximately 10 g kg-1 on a dry weight basis (labile P as defined here is equal to the sum of sequential H2O- and 0.5 M NaHCO-3–extractable P fractions). If direct runoff of P from biosolids occurred at any time following application, even slight PO4–P increases in runoff water should be measurable given: (i) the analytical sensitivity for detecting PO4–P (detection limit = 0.005 mg L-1) and (ii) the study design's emphasis on sampling 27 Creek water over a wide range of discharge rates. Likewise, BAP concentrations during runoff events did not change following biosolids application, suggesting that labile, P-rich particles in biosolids are not mobile during periods of heavy rainfall and subsequent runoff.

Flow-weighted total P concentrations range from 0.05 to 0.3 mg L-1 (Table 1). Concentrations during runoff events were generally comparable, but tended to be higher after biosolids application. Increases in total P were noted during a late spring 1997 runoff event five months after biosolids application, and again to a lesser extent during two events in 1998. As labile P concentrations varied little, an increase in total P during certain runoff events may be within the normal concentration range for this watershed. An examination of P fractions provides some insight. Total phosphorus (TP), soluble phosphorus (SP), and BAP were measured directly. Through subtraction, the concentration of particulate phosphorus (PP) and BPP can be calculated:

The relative amount of organic versus inorganic particulate P was not determined, nor was the concentration of dissolved organic P. The concentration of the PP fraction has increased following biosolids application due to increases in total P. However, the concentration of labile particulate phosphorus (BPP) has not changed. Therefore, episodes where total P and PP levels are elevated in 27 Creek runoff water while those of PO4–P, BAP, and BPP remain constant suggest that P in 27 Creek water is primarily in particulate form and not labile. Sources of particulate P within the 27 Creek watershed include the heavily vegetated and densely covered 27 Creek channel and ephemeral drainage courses. Another source of particulate P may be from downed young trees (<18 yr) located throughout the 27 Creek watershed, which was precommercially thinned approximately six months before the study began (trees were left in place).

Most of the P in anaerobically digested, dewatered biosolids is in an inorganic solid phase, and of this fraction, most is found as Al, Ca, or Fe phosphates of varying stability and reactivity (Frossard et al., 1994; Hinedi et al., 1989). Organic P in anaerobically digested, dewatered biosolids is a minor fraction on a dry weight basis and occurs primarily as mono- and di-esters and polyphosphates (Hinedi et al., 1989) and these compounds are readily hydrolyzed upon application and reaction with the soil and theoretically persist for short periods of time. Organic P compounds added in biosolids presumably decompose and release phosphate soon after application, minimizing the potential for movement of P rich particles. Thus, transfer of particles rich in inorganic or organic P from biosolids to surface waters is unlikely in this case and the data support this assumption.

Nitrogen
Flow-weighted NH+4–N concentrations, when detected, were low, averaging between 0.01 and 0.07 mg L-1. In approximately 30% of water samples taken for this study, NH+4–N was below detection limits (0.01 mg L-1). Ammonium N concentrations in biosolids are relatively high compared with other nutrients, ranging from 0.5 to 1.5% on a dry weight basis (1997 King County biosolids monitoring data). Consequently, runoff waters directly contacting and moving through biosolids would be susceptible to high NH+4–N concentrations immediately after biosolids application. Conversely, NO-3–N concentrations in anaerobically digested biosolids are often below 50 mg kg-1. Therefore, in assessing effects of biosolids application on forest stream water quality, NH+4–N is a useful early indicator, while NO-3–N is useful over the long term as biosolids decompose and organic N mineralizes and nitrifies.

Flow-weighted NO-3–N concentrations during runoff events were relatively constant before biosolids application, ranging between 0.23 and 0.35 mg L-1 (Table 1). The pre-biosolids application NO-3–N concentration in 27 Creek water generally mirrored that of precipitation. The five-year (1992–1996) average annual volume-weighted mean NO-3–N concentration in precipitation at Pack Forest is 0.34 mg L-1 (National Atmospheric Deposition Program, 1998). Beginning in November 1997, the average flow-weighted NO-3–N concentration noticeably increased to approximately 1 mg L-1 (Table 1). In addition, total N concentrations increased after biosolids application (Fig. 5). The increase in NO-3–N concentration in 27 Creek is clearly an effect of biosolids in increasing the soil water NO-3–N concentration due to biosolids decomposition and N mineralization and nitrification processes. However, nitrate N concentrations in 27 Creek after biosolids application are approximately 10% of the USEPA drinking water standard of 10 mg L-1.

Effect of Discharge on Phosphorus and Nitrogen Concentrations in 27 Creek
Phosphorus
The concentration of PO4–P and BAP in 27 Creek did not change with increasing discharge before and after biosolids application (Fig. 7 and 8) . For separate runoff events, changes in PO4–P and BAP concentrations were not related to changes in discharge at any time (r2 < 0.1). This is not unexpected, as the lack of a relationship between dissolved phosphate concentration and discharge has been reported for a number of forest types (Hobbie and Likens, 1973; Meyer and Likens, 1979; Schreiber et al., 1976). Biologically available P is not commonly used in studies of runoff from forested watersheds. It is useful in this case because of the high concentration (approximately 1%) of labile P in municipal biosolids and the potential for transfer of fine particulates from steep slopes via lateral flow networks (Dunne, 1978).



View larger version (19K):
[in this window]
[in a new window]
  		Fig. 7. Phosphate P versus discharge in 27 Creek before and after biosolids application.

 


View larger version (18K):
[in this window]
[in a new window]
  		Fig. 8. Biologically available phosphorus (BAP) versus discharge in 27 Creek before and after biosolids application.

 
Dilution effects were not observed for PO4–P and BAP. The lack of any dilution effect suggests that soil water within 27 Creek's source area and runoff water PO4–P and BAP concentrations may be in equilibrium. Thus, a reasonable assumption is that mass export of labile P would increase linearly with increasing discharge and variation in amounts exported during runoff events is controlled by changes in discharge. Furthermore, for PO4–P and BAP a reasonable assumption is that annual variation in discharge is probably more important than variation in runoff event discharge in controlling labile P export from this watershed. The absence of any concentration discharge relationship is important relative to the mobility of P following biosolids application. In general, low soil solution phosphate concentrations are maintained by the soil's ability to buffer phosphate additions, primarily through P sorption reactions with Al and Fe oxides in low-pH soil environments (Berkheiser et al., 1980). Soil water and subsurface runoff water must maintain a low equilibrium phosphorus concentration and this concentration is nearly equal to that of streamwater under the discharge range within the 27 Creek watershed. Biosolids, while adding approximately 500 kg total P ha-1, did not appear to exceed the soil's retention capacity for added P and any phosphate that was released into solution from biosolids was rapidly sorbed by the soil near the application zone.

There is little relationship (r2 < 0.1) between total P concentrations and discharge from the 27 Creek watershed (Fig. 9) . In some cases elevated total P concentrations have been measured seasonally and during periods of increased discharge following biosolids application, but there has been no consistent pattern during the runoff events selected for analysis. A runoff event during late May 1997 produced the highest total P concentrations in runoff water measured during the study, but total discharge was moderate compared with other runoff events. A positive relationship between total P concentration and discharge has been reported for streams draining some forested areas (Meyer and Likens, 1979; Meyer et al., 1988) and the increase in total P concentration has usually been attributed to mobilization of organic detritus from source areas in stream networks. Mobilization of particulate P is commonly part of the natural cycle of P export from most forested watersheds and mass export is strongly dependent upon episodic runoff events.



View larger version (16K):
[in this window]
[in a new window]
  		Fig. 9. Total P versus discharge in 27 Creek before and after biosolids application.

 
An apparent trend from Fig. 9 is the elevation of total P concentrations after biosolids application over most of 27 Creek's discharge range. Elevated total P concentrations have been measured during both spring seasons after biosolids application, as well as during runoff events in November 1997 and January 1998. This may be an effect of biosolids increasing the equilibrium P concentration supported in the soil solution for a P fraction other than PO4–P or BAP (such as soluble organic P that was not hydrolyzed during determination of soluble P using the Murphy and Riley [1962] procedure). Another possible explanation may be P release from 27 Creek sediments or organic matter mineralization in response to warm temperatures (spring events) and abundant detrital sources near the stream corridor (autumn and spring events). Research has shown that stream sediments can sorb and release phosphate, with sorption positively correlated with organic matter, extractable aluminum, and decreasing sediment particle size (Meyer, 1979).

The variation in the total P concentration–discharge relationship after biosolids application and the consistency in BPP concentrations suggests that runoff events do not induce widespread movement of P from application areas nor would these events probably change the P export pattern. Certainly it would be expected that an increasing mass of total P is exported from the watershed during high discharge events and the relationship between mass export of P on a per hectare basis and discharge would be linear. Changes in observed total P concentrations in 27 Creek following biosolids application may have more to do with seasonal and event-related source area dynamics than with biosolids application.

Nitrogen
There is no relationship between NH+4–N concentrations and discharge (r2 < 0.1) from the 27 Creek watershed (Fig. 10) . The absence of a concentration–discharge relationship prior to biosolids application is consistent with other studies examining runoff water quality in forested watersheds (Feller and Kimmins, 1979; Meyer et al., 1988). It is clear that biosolids have not altered the NH+4–N concentration–discharge relationship in any way nor has the application of biosolids had any effect on NH+4–N concentrations in 27 Creek. This situation is similar to that of labile P and suggests that the NH+4–N concentration in 27 Creek is independent of discharge. The lack of any change in NH+4–N concentration following biosolids application is an important finding in that biosolids contain high concentrations of NH+4–N and because elevated levels of NH+4–N can be toxic to fish populations. It is also important in determining whether or not biosolids applied to steep forested slopes results in direct runoff of nutrient-rich water into streams draining biosolids-treated areas. Clearly, this has not occurred for NH+4–N, both in the wet months immediately after biosolids application and over several seasons.



View larger version (17K):
[in this window]
[in a new window]
  		Fig. 10. Ammonium N versus discharge in 27 Creek before and after biosolids application. Values below the detection limit (0.01 mg L-1) were set to zero.

 
The relationship between NO-3–N concentrations in 27 Creek and discharge has been altered by biosolids application (Fig. 11) . Before application, the NO-3–N concentration in 27 Creek was not related to changes in discharge as shown by the horizontal scatter of pre-application data over a wide discharge range; NO-3–N concentrations in 27 Creek rarely exceeded 0.5 mg L-1. Nine months after biosolids application, a noticeable increase in NO-3–N concentrations occurred, and in general, increases in discharge result in increases in NO-3–N concentrations in 27 Creek (r2 = 0.31, p < 0.0001). The elevation of NO-3–N concentrations was observed beginning in October 1997 and it continued for the duration of the study. The data cluster beginning at approximately 0.75 mg L-1 and extending upward coincides with the onset of autumn 1997 runoff events.



View larger version (18K):
[in this window]
[in a new window]
  		Fig. 11. Nitrate N versus discharge in 27 Creek before and after biosolids application.

 
The lag effect of increasing NO-3–N soil water and runoff water concentrations nine months after application is consistent with the current understanding of biosolids nitrogen dynamics and the release of organic N through the mineralization process. Organic N mineralization from anaerobically digested biosolids occurs slowly and, within one year of application for western Washington Douglas-fir forests, approximately 40% of the organic N is mineralized (Henry et al., 1999). Therefore, organic N mineralization provides a large pool of NH+4–N from which NO-3–N is ultimately produced through microbially mediated transformations. The mobility of NO-3–N in the soil following biosolids application in the Pacific Northwest has been demonstrated repeatedly and for several different soil types and stand ages (Henry et al., 2000).

Plant uptake, microbial immobilization, and soil storage probably removed a portion of the organic N released through mineralization and nitrification between January and October 1997. No changes in 27 Creek NO-3–N concentrations were noted during this time (Fig. 11). The data are scattered horizontally across the full range of 27 Creek discharge rates and concentrations are similar comparing before and after conditions. Autumn rain storms beginning in late October 1997 mobilized NO-3–N from soil water in response to the input of low NO-3–N concentration rain water, elevating concentrations in subsurface runoff water and noticeably in 27 Creek runoff. The replenishment of soil water NO-3–N has continued, due in part to the moderate rate of biosolids organic N mineralization and nitrification of NH+4–N, followed by its release as NO-3–N during runoff events. Elevated NO-3–N concentrations in 27 Creek and the development of a positive concentration–discharge relationship should be placed into context, however, with the mass of N applied to the watershed versus mass exported. Approximately 5900 kg of total N was added to the watershed with biosolids fertilization. From 1 Oct. 1997 to 30 June 1998, 90 kg of NO-3–N was exported via surface water runoff assuming a constant NO-3–N concentration of 1 mg L-1 and total discharge of approximately 9.0 x 107 L. Export of 90 kg NO-3–N is relatively small and includes nitrate produced from biosolids and natural sources. Given background nitrate concentrations below 0.4 mg L-1, nitrate losses from biosolids represent less than 1% of the original mass of total N applied.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 

  1. Phosphate P, BAP, BPP, and NH+4–N were not mobile following biosolids application to the 27 Creek watershed. Direct runoff of these compounds into 27 Creek during runoff events was not observed.
  2. Total P appears to be immobile following biosolids application. Seasonal episodes of elevated total P concentrations were noted during some periods of increased discharge. However, particulate P entering 27 Creek during runoff events or release of P from sediments cannot be ruled out in elevating total P concentrations.
  3. The concentrations of PO4–P, BAP, BPP, and NH+4–N in 27 Creek are not related to discharge; for most events, discharge had no effect on total P concentrations as well.
  4. Biosolids changed the 27 Creek concentration–discharge relationship for NO-3–N. Before application, no relationship existed. Beginning nine months after biosolids application, increases in discharge were positively related to increases in NO-3–N concentrations. The longevity of this effect was not determined.


   ACKNOWLEDGMENTS
 
This research was supported by funding from the King County Department of Natural Resources, Biosolids Management Program, Seattle, Washington. The authors acknowledge the contributions of Neil Cowley and Dan Bennett, University of Washington Pack Forest staff, and King County Biosolids Management Program and Analytical Laboratory staff.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 




This article has been cited by other articles:


Home page
 J. Environ. Qual.Home page
A. L. Friend, S. D. Roberts, S. H. Schoenholtz, J. A. Mobley, and P. D. Gerard
Poultry Litter Application to Loblolly Pine Forests: Growth and Nutrient Containment
J. Environ. Qual., April 3, 2006; 35(3): 837 - 848.
[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 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 (4)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Grey, M.
Right arrow Articles by Henry, C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Grey, M.
Right arrow Articles by Henry, C.
Agricola
Right arrow Articles by Grey, M.
Right arrow Articles by Henry, C.
Related Collections
Right arrow Watershed and Landscape Processes
Right arrow Surface Water Quality
Right arrow Nitrogen
Right arrow Phosphorus
Right arrow Municipal Waste

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