JEQ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 1 November 2007
Published in J Environ Qual 36:1725-1734 (2007)
DOI: 10.2134/jeq2006.0517
© 2007 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 Haggard, B. E.
Right arrow Articles by Brye, K. R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Haggard, B. E.
Right arrow Articles by Brye, K. R.
Agricola
Right arrow Articles by Haggard, B. E.
Right arrow Articles by Brye, K. R.
Related Collections
Right arrow Wetlands and Aquatic Processes
Right arrow Surface Water Quality
Right arrow Sorption/Exchange

TECHNICAL REPORTS

Wetlands and Aquatic Processes

Variations in Stream Water and Sediment Phosphorus among Select Ozark Catchments

Brian E. Haggarda,*, Douglas R. Smithb and Kristofor R. Bryec

a Biological and Agricultural Engineering Dep., Univ. of Arkansas, 203 Engineering Hall, Fayetteville, AR 72701
b USDA–ARS National Soil Erosion Research Lab., 275 South Russell Street, Purdue Univ., West Lafayette, IN 47907
c Crop, Soil, and Environmental Sciences Dep., Univ. of Arkansas, 115 Plant Sciences, Fayetteville, AR 72701

* Corresponding author (haggard{at}uark.edu).

Received for publication November 28, 2006.

   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 Materials and Methods
 Results and Discussion
 Conclusions
 REFERENCES
 
Stream sediments play a large role in the transport and fate of soluble reactive phosphorus (SRP) in stream ecosystems, and equilibrium P concentrations (EPC0) of benthic sediments at which P is neither adsorbed nor desorbed are often related to stream water SRP concentrations. This study evaluated (i) the variation among water chemistry and sediment-P interactions among streams draining catchments that varied in the land use; (ii) the relations between SRP concentration, sediment EPC0, and other measured abiotic factors (e.g., particle size distribution, slope of linear sorption isotherms, etc.) in the stream sediments; and (iii) the use of the traditional Mehlich-3 (M3) soil extraction on stream sediments to elucidate other abiotic factors (e.g, M3P, P saturation ratio, etc.) related to SRP concentration in stream sediments. Stream water and sediments were sampled at 22 selected Ozark streams in northwest Arkansas during fall 2003 and spring 2004. Nitrate-N concentrations in the water column (r = 0.69) and modified P saturation ratios (PSRmod) of the benthic sediments (r = 0.79) at the selected streams increased with an increase in percent pasture in the catchments, whereas SRP concentration (r = –0.56) and Mehlich-3–extractable P (M3P) content (r = –0.47) decreased with an increase in the percent forested area. Soluble reactive P concentrations in the stream water were positively correlated to sediment EPC0 (r = 0.51), although sediment EPC0 was generally greater than SRP. The M3 soil extraction was useful in identifying abiotic factors related to SRP concentrations in the selected streams, in particular SRP concentrations were positively correlated to M3P contents (r = 0.50) and PSRmod (r = 0.71) of the benthic sediments. Thus, M3P and EPC0 estimates from stream sediments may be valuable yet simple indicators of whether benthic sediments act as sinks or sources of P in fluvial systems, as well as estimating changes in stream SRP concentrations.

Abbreviations: Abbreviations: ANOVA, analysis of variance • Chl-a, chlorophyll-a • DO, dissolved oxygen • EPC0, equilibrium P concentration • LSD, least significant difference • M3, Mehlich-3 extraction • M3P, Mehlich-3 P • Max, maximum • Med, median • Min, minimum • NH4–N, ammonium-N • NO3–N, nitrate N • NO2–N, nitrite N • PSR, P saturation ratio • Q, stream discharge • SRP, soluble reactive P • SD, standard deviation • TN, total N • TP, total P • u, water velocity • WWTP, wastewater treatment plant


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
Materials and Methods
 Results and Discussion
 Conclusions
 REFERENCES
 
SEVERAL studies have evaluated factors controlling soluble reactive P (SRP) concentrations in streams, particularly the ability of stream sediments to provide a dissolved P buffering mechanism. Stream sediments have the ability to adsorb dissolved P from the water column or release P to overlying waters depending on the equilibrium exchange concentration, or sediment equilibrium P concentration (EPC0; Froelich, 1988). The sediment EPC0 is the concentration in the aqueous phase (e.g., water column in streams) at which net adsorption or desorption of dissolved P from the benthic sediment does not occur (Taylor and Kunishi, 1971; Klotz, 1988), i.e., the aqueous and solid phases are in a dynamic equilibrium with respect to dissolved P. The sediment EPC0 concept is often used to determine if benthic sediments are a potential source or sink of dissolved P to the overlying waters in streams.

A myriad of factors influence the dissolved P equilibrium concentration between the water column and benthic sediments in streams, including biotic processes and various abiotic characteristics. Important abiotic characteristics revolve around the ability of stream sediments to adsorb dissolved P from the water column, such as benthic sediment particle size distribution (Klotz, 1988; Haggard et al., 1999), the content of water-extractable or other P forms in the sediments (McDowell et al., 2002; Ekka et al., 2006), divalent cations in aqueous solution (Klotz, 1991) and exchangeable in sediments, and the strength with which dissolved P is adsorbed to sediments, among other potential factors. The influence of sediments on stream water P concentrations is similar to that observed in soil–water systems, where dissolved P concentrations in runoff water from small plots are positively correlated to water-extractable P and Mehlich-3 P (M3P) content of soils (Pote et al., 1999, Vadas et al., 2005).

Dissolved P equilibrium concentrations between stream sediments and water can also be affected by external factors, such as municipal wastewater treatment plant (WWTP) effluent discharges high in dissolved P (Popova et al., 2006; Ekka et al., 2006). Biotic processes in benthic sediments are extremely important in the uptake and release of dissolved P (Gächter and Meyer, 1993) and often account for a large fraction (more than 38%) of dissolved P adsorbed from aqueous solutions (e.g, see Haggard et al., 1999; Khoshmanesh et al., 1999). The importance of biotic processes is further supported by interactions among sediment EPC0, benthic organic matter content (Smith et al., 2005), and alkaline phosphatase activity in the benthos (Klotz, 1985).

It is clear that uptake and release of P by stream sediments are important components in the fate and transport of dissolved P through fluvial systems and in the modification of P export from catchments (House and Warwick, 1998; McDowell et al., 2003). Many studies have evaluated sediment and dissolved P interactions in individual streams (e.g., Klotz, 1988, 1991; House and Warwick, 1998; McDowell et al., 2002) or on a small number of streams in a particular region and similar land use types (e.g., Haggard et al., 1999; Ekka et al., 2006). However, few studies have evaluated sediment EPC0 and potential abiotic regulating factors over a large number of streams (greater than 20) draining catchments with differing proportions of forest, pasture, and urban development. Thus, the main of objective of our study was to evaluate the interactions of dissolved P between stream water and sediments at 22 streams in the Ozark Highlands of northwest Arkansas, USA. Specifically, we focused on variations in water chemistry and sediment P parameters with catchment land uses (Objective 1), the relation between water column dissolved P concentrations and sediment EPC0 at the selected streams (Objective 2), and the use of the traditional Mehlich-3 (M3; Mehlich, 1984) soil extraction to elucidate factors related to dissolved P concentrations in streams and benthic sediment EPC0 (Objective 3).


   Materials and Methods
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
Results and Discussion
 Conclusions
 REFERENCES
 
Study Site Description
We selected one site on 22 different streams in the Ozark Highlands of Benton, Washington, and Madison Counties in Northwest Arkansas, USA (Table 1 , Fig. 1 ). These streams drained catchments of various sizes (122 to 11200 ha) and multiple land uses, including forest, pasture, and urban development. ArcGIS software (ESRI GIS and Mapping Software, St. Charles, MO) was used to determine catchment area and land use classifications for each of the water quality and sediment sampling sites; land use was determined using 1999 land use and land cover data layers from the University of Arkansas Center for Advanced Spatial Technology (CAST, 2006). The land use and land cover information was classified into four categories: forest, pasture, urban, and transitional areas or barren lands. The selected sites drained catchments that varied from ~90% forest to ~86% pasture to ~61% urban, and barren and transitional areas were always the smallest fraction (less than 5%) of designated land use in each of the catchments. The 22 selected streams were separated into land use categories (Table 1), which were designated as: Forest, Mixed, Pasture, and Urban. The Forest catchments were those that had more than 60% forested land use and less than 10% urban land use; Pasture catchments were those that had more than 60% pasture land use and less than 10% urban; and Urban catchments were those that had approximately 10% or more urban development. The Mixed catchments did not fit into any of the other land use categories, and these catchments generally had a close to equal mix between forested and pasture land use. We acknowledge that the separation of the selected catchments into these designated land use categories is somewhat arbitrary, but this was done to facilitate comparisons between selected streams. The 22 selected sites were sampled in late summer and fall 2003 (late August through early November) and in spring 2004 (mid May through mid June), and sampling dates were selected when the stream was in seasonal base flow conditions. Seasonal base flow discharge shows natural temporal and spatial variation in the Ozarks, and base flow conditions were preferred because benthic sediment control or maintenance of stream water SRP concentrations would occur during these low flow periods (McDowell et al., 2002).


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

  		Table 1. Catchment area, percent of forested, pasture and urban land use and land use classification of the selected Ozark streams, Northwest Arkansas, USA.

 

Figure 1
View larger version (117K):
[in this window]
[in a new window]

  		Fig. 1. Map delineating catchment area and land use distribution of the selected study sites on 22 different streams in the Ozark Highlands of Northwest Arkansas, USA. Note: Site numbers correspond to the numbers listed in parentheses in Table 1.

 
Field Methods
Water and sediment samples were collected at three points along the stream continuum, where water samples were collected in the thalweg of the defined stream channel, and sediment samples were collected from the top 5 to 10 cm along a transect perpendicular to stream flow until approximately 2 L in volume was collected. The gravel and cobble stream bottoms in typical Ozark streams have relatively large transient storage areas (see Haggard et al., 2001), where surface water interacts with bottom sediments often much deeper than 5 to 10 cm. Three individual water samples and three different composite sediment samples were collected at the selected sites along a stream reach that varied from 50 to over 100 m in distance. Water samples collected for chemical analyses included an unfiltered sample (~250 mL), a filtered acidified sample (~20 mL, 0.45 µm membrane, syringe filtration, pH < 2), and a filtered unacidified sample (~20 mL, 0.45 µm membrane, syringe filtration); two other unfiltered water samples (~1 L each) were collected at each point for sestonic chlorophyll-a (Chl-a) analysis and also the sediment extractions. We measured pH (pH Testr 3, Oakton Instruments, West Cladwell, NJ), dissolved oxygen (DO) (YSI Model 85, Yellow Springs, OH), and conductivity and temperature (Orion Meter 115A plus, Beverly, MA) at the middle site. Water velocity (u) was measured at the middle site along a transect divided into equally spaced intervals (Flo-Mate 2000, Marsh-McBirney, Inc., Frederick, MD), and depth was measured at the mid-point of each interval. Discharge (Q) was estimated as the sum of u multiplied by the cross-sectional area (A2) of each interval along each transect.

Laboratory Methods
After sample processing in the field, all water samples for chemical analyses were frozen until specific analyses were completed. Filtered, acidified samples were analyzed for SRP using the automated ascorbic acid method (APHA, 1998) and ammonium–nitrogen (NH4–N) using the sodium nitroprusside and salicylate method (APHA, 1998) on a Skalar San Plus Wet Chemistry Autoanalyzer (Skalar, the Netherlands). Nitrite N (NO2–N), nitrate N (NO3–N), and Cl– were measured on filtered, unacidified samples using colorimetric determination of NO2–N (APHA, 1998), Cd–Cu reduction of NO3–N and subsequent colorimetric analysis of NO2–N (APHA, 1998), and the mercuric thiocyanate reaction for Cl– (Skalar Method, the Netherlands) on the autoanalyzer. A known volume of water (1 L) was filtered through a Whatman GF/F glassfiber filter, and the glassfiber filter was frozen until Chl-a analysis could be performed. The frozen filter was placed in a glass vial with 5 mL of aqueous acetone saturated with MgCO3 and then shredded. The glass vial was centrifuged and then the supernatant analyzed for Chl–a using the trichromatic method (APHA, 1998).

After return to the laboratory, sediments were immediately sieved through a 4.5-mm sieve, and particles less than 4.5 mm were used in the following extraction procedure. Stream water (~1 L) was filtered through a 0.45-µm membrane and used in the subsequent extractions as defined. Approximately 25 g of fresh, wet sediment were placed into a 125 mL Erlenmeyer flask, and then 100 mL of different solutions were added to each flask. Our solutions were comprised of filtered (0.45 µm) stream water spiked with additional amounts of PO4–P at rates of 0.00, 0.10, 0.25, 0.50, 1.00, and 2.50 mg L–1. For example, if the stream water had an SRP concentration of 0.01 mg L–1, then our series of solutions would have initial SRP concentrations of 0.01, 0.11, 0.26, 0.51, 1.01, and 2.51 mg L–1. The sediment slurry was shaken in a reciprocating type shaker for 1 h, and then allowed to settle for approximately 30 min. A 15- to 20-mL aliquot was filtered through a 0.45-µm membrane, and then analyzed for SRP as previously described. The remaining sediment slurry was transferred into Al pans and dried for ~48 h at 80°C to determine sediment dry mass. Simple linear regression of P sorbed (mg Psorbed kg–1 dry sediment) against final SRP concentration (SRPfinal, mg L–1) in the solution was used to estimate sediment EPC0, where the x intercept represents the point of negligible P adsorption or release from sediments to the aqueous solution (Fig. 2 ). The slope (Kslope) of this line was used as a measure of the ability of stream sediments to adsorb P, where greater Kslope values would indicate a stronger ability to adsorb P from the aqueous solution (Froelich, 1988).


Figure 2
View larger version (11K):
[in this window]
[in a new window]

  		Fig. 2. Graphical display of the amount of phosphorus adsorbed by the sediment (P sorbed) as a function of initial and final soluble reactive phosphorus (SRP) concentration in the aqueous solution. Note: The linear regression of P sorbed and final SRP concentration was used to estimate the slope of the relation (Kslope) and sediment equilibrium P concentration (EPC0) in this study, and the data point from the highest level of P enrichment, i.e., ambient SRP concentrations plus an additional 2.5 mg PO4–P L–1, was not used in the linear regression used to determine Kslope and sediment EPC0 in this example.

 
In this study, we used fresh, wet sediments in our sorption studies because previous studies have shown that sediment drying affects P sorption characteristics of sediment (e.g., see Baldwin, 1996; Watts, 2000) and the estimation of sediment EPC0. We also used filtered (0.45 µm) stream water as the basis for our series of P solutions, because several studies (e.g., Klotz, 1988; House and Denison, 2000) have shown that dissolved cations (i.e., Ca, Mg, Fe, and Mn) can influence sediment EPC0 measurement. In fact, Popova et al. (2006) showed that sediment EPC0 was significantly less when determined using a CaCl2 solution of similar conductivity to the natural stream water than when using filtered stream water, as in this study.

In spring 2004, sediment samples from the 22 selected streams were also extracted using the M3 (Mehlich, 1984) method of soil analysis. The M3 extracts were analyzed using inductively coupled plasma–optical emission spectrometry (ICP–OES) to determine M3-extractable Al, As, B, Ca, Cd, Co, Cr, Cu, Fe, K, Mg, Mn, Mo, Na, Ni, P, Pb, S, Se, Ti, and Zn. The M3 elemental contents are reported on a mg element kg–1 dry sediment basis. We used the M3 data to determine the P saturation ratio (PSR) of these stream sediments (see Sims et al., 2002):

Formula 1[1]

However, we also calculated a modified PSR (PSRmod) based on the results obtained in this study that showed M3Fe, M3Mn, and M3Mg were important factors related to M3P in the sediment extracts (see Results and Discussion):

Formula 2[2]
Phosphorus saturation ratios were calculated on a molar basis using mmol kg–1 dry sediment.

Data Analyses and Comparisons
The intent of our study was to focus on broad comparisons and correlations across the 22 selected Ozark streams. Therefore, we used the geometric mean of the data when collected in replicates, where the replicates represent different water or sediment samples collected at the separate transects. The geometric mean was used to minimize the influence of extreme values, and it is common to use log transformations in water quality data analyses and comparisons (Hirsch et al., 1991). We used a paired t-test of the data from fall 2003 and spring 2004 to ascertain any seasonal differences in parameters across the 22 selected streams, and we used simple linear or stepwise linear regression analyses to evaluate relations between two or more parameters. The strength of associations discussed in the text (see Results and Discussion) are quantified by correlation coefficients (r) due to uncertain causality, whereas coefficients of determination (R2) are reported in figures as part of the presentation of the regression models. We used analysis of variance (ANOVA) with mean separation (least significant difference, LSD) to determine if any parameter varied among the four defined land use categories. All statistical comparisons were made using an a priori significance level of 0.05 and Statistix 8.0 software (Analytical Software, Tallahassee, FL).


   Results and Discussion
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results and Discussion
Conclusions
 REFERENCES
 
Stream Flow and Water Chemistry
Stream Q was highly variable among the 22 selected sites on these Ozark streams (Table 2 and 3 ), ranging from 3 L s–1 at Moores Creek to ~890 L s–1 at Little Osage Creek in fall 2003 and from 2 L s–1 at Mud Creek Tributary to ~1900 L s–1 at Little Osage Creek in spring 2004. Stream Q increased significantly with an increase in catchment area (r = 0.63, P < 0.01), and Q was also positively correlated with u (r = 0.64, P < 0.01), which varied from a combined fall 2003 and spring 2004 average of 0.04 m s–1 at Moore's Creek to 0.68 m s–1 at Flint Creek. Stream Q and u were significantly greater (Paired t-test, P < 0.001) in spring 2004 compared to fall 2003 across the selected streams, likely because seasonal base flow discharge was greater in the spring season due to increased precipitation and replenishing of alluvial ground waters over the winter and spring seasons which contribute to seasonal base flow.


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

  		Table 2. Data summary of water chemistry, sediment–phosphorus sorption, and benthic particle size distribution (limited to particles < 4.5 mm in diameter) across the 22 selected sampling sites at Ozark streams in fall 2003 (late August through early November).

 

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

  		Table 3. Data summary of water chemistry, sediment–phosphorus sorption, and benthic particle size distribution (limited to particles <4.5 mm in diameter) across the 22 selected sampling sites at Ozark streams in spring 2004 (May through June).

 
Water chemistry and temperature varied between the seasons sampled where pH, conductivity, and DO were greater during base flow conditions in fall 2003, but water temperature and NH4–N, NO2–N, and NO3–N concentrations were greater in spring 2004 (Table 2 and 3). However, conductivity and sestonic Chl-a, SRP, TP, and Cl concentrations did not vary between the seasons across the selected sampling sites. Overall, the supply of N (i.e., NH4–N and NO3–N) at the selected streams was much greater than that of P (i.e, SRP), because molar N/P supply ratios ranged from 10 to 368. Studies in the southwestern Ozark Highlands have shown that periphytic algal growth in Ozark streams is often limited by the supply of P relative to N or is co-limited by the supply of N and P, when nutrients are the primary limiting factor for periphytic algal growth (Matlock et al., 1999; Popova et al., 2006). However, sestonic Chl-a concentrations were not correlated to the supply of either N and or P at the selected Ozark streams and were generally very low (<1 µg L–1), except at the slowly flowing Moore's Creek (1.2–21.2 µg L–1). In contrast, Lohman and Jones (1999) showed that sestonic Chl-a concentrations were positively correlated to the nutrient supply in northern Ozark streams, although these northern Ozark streams were generally larger than those sampled in our study (i.e, catchment area was much greater).

Average NO3–N concentrations increased with an increase in stream Q across the selected streams (data not shown, r = 0.57, P < 0.01), but it appears that catchment land use has the greatest influence on stream nutrient concentration and conductivity. Average NO3–N concentrations increased with an increase in the percent pasture land use within the catchment (r = 0.69, P < 0.001) (Fig. 3B ), and NO3–N concentrations were greatest (ANOVA LSD, P < 0.001) in the streams draining catchments in the Pasture and Mixed land use categories (Fig. 4B ). On the other hand, average SRP and TP concentrations decreased with an increase in the percent forested land use within the catchment (SRP: r = –0.56, P < 0.01; TP: r = –0.53, P < 0.05) (Fig. 3A), and SRP concentrations were least in the streams draining catchments in the Forest category (ANOVA LSD, P < 0.05) (Fig. 4A). Several studies have shown that agricultural land use (i.e., pasture) can influence water chemistry in ground and surface waters in the Ozark Highlands (Petersen et al., 1999; Haggard et al., 2003) and in other basins (e.g., see McFarland and Hauck, 1999; Buck et al., 2004). However, conductivity and Cl concentrations in the selected streams were positively correlated to the percent urban land use in the catchments (data not shown; Cond: r = 0.60, P < 0.01; Cl: r = 0.65, P < 0.01). The footprint of catchment land use on stream water chemistry, particularly nutrient concentrations, is rather consistent among studies, whether water quality monitoring was simply seasonal (as in this study) or more long term (multiple years as in the other citations).


Figure 3
View larger version (13K):
[in this window]
[in a new window]

  		Fig. 3. The relation between (A) soluble reactive phosphorus (SRP) concentrations (mg L–1) in the water column at the selected stream and the percent (%) forested land use in the catchments; (B) nitrate-nitrogen (NO3–N) concentrations (mg L–1) in the water column at the selected stream and the percent (%) pasture land use in the catchments; (C) Mehlich-3 P (M3P) content (mg kg–1 dry sediment) in the stream sediments and the percent (%) forested land in the catchments; and (D) the modified P saturation ratio (PSRmod) of the stream sediments and the percent (%) pasture land use in the catchments.

 

Figure 4
View larger version (14K):
[in this window]
[in a new window]

  		Fig. 4. Mean (plus standard deviation) of (A) soluble reactive phosphorus (SRP) concentrations (mg L–1), (B) nitrate-nitrogen (NO3–N) concentrations (mg L–1), (C) sediment equilibrium P concentrations (EPC0, mg L–1), and (D) the modified P saturation ratio (PSRmod) as a function of the various defined land use categories. Note: Different letters above bars denote significant differences between land use categories using ANOVA LSD P < 0.05.

 
Sediment Equilibrium Phosphorus Concentrations
Benthic sediments displayed a linear P sorption isotherm (all regressions: R2 ≥ 0.75, P ≤ 0.05), especially at initial aqueous SRP concentrations less than 1.1 mg L–1 (e.g, see Fig. 2). The linear portion of the sorption isotherm that was used to estimate sediment EPC0 often did not include the data from the highest level of P enrichment, i.e., 2.5 mg L–1 PO4–P plus ambient SRP solution. This would suggest the possibility of a two-phase sorption process, and the isotherms from fall 2003 had a greater number of occurrences of this type than did isotherms from spring 2004. The slopes of the linear isotherms (Kslope) were greater in fall 2003 compared to spring 2004 (paired t-test, P < 0.05), which coincided with a greater proportion of silt particles (paired t-test, P < 0.001) (Table 2 and 3). The slope of the linear isotherm (Kslope) was positively correlated to the proportion of silt particles (0.002–0.05 mm) in the benthic sediments (r = 0.57, P < 0.05), suggesting that the P buffering capacity of these sediments increases with an increase in fine sediments. Overall, these benthic sediments were largely made up of particles sand size (0.05–2.0 mm) or greater (Table 2 and 3). Benthic sediment EPC0 also exponentially decreased (R2 = 0.38, P < 0.01) with an increase in Kslope, which represents the ability of the sediments to adsorb P (Fig. 5D ).


Figure 5
View larger version (15K):
[in this window]
[in a new window]

  		Fig. 5. The relation between (A) soluble reactive phosphorus (SRP) concentrations (mg L–1) in the water column and benthic sediment equilibrium P concentrations (EPC0, mg L–1) at the selected streams; (B) SRP concentrations (mg L–1) in the stream water and benthic sediment Mehlich-3 P (M3P) contents (mg kg–1 dry sediment); (C) SRP concentrations (mg L–1) in the stream water and the modified P saturation ratio (PSRmod) of the benthic sediments; and (D) benthic sediment EPC0 (mg L–1) and the slope (Kslope, L kg–1) of the linear regression used to estimate benthic sediment EPC0.

 
Sediment EPC0 varied by more than an order of magnitude (Table 2 and 3), ranging from less than 0.001 mg L–1 at Beaty Creek and Lollars Creek to 0.329 mg L–1 at Ballard Creek. The range in sediment EPC0 was much larger than that observed with average SRP concentration in the water column of the streams (0.003–0.072 mg L–1). Sediment EPC0 at the selected streams was greater (paired t-test, P < 0.001) than SRP concentrations in the water column, although sediment EPC0 and SRP concentrations were significantly correlated (r = 0.51, P < 0.05) across the selected streams (Fig. 5A). This observation suggests that benthic sediments in Ozark streams play a role in the regulation and/or maintenance of SRP concentration in the water column, as observed in several other studies nationwide (e.g., see Klotz, 1988; McDowell et al., 2001; Novak et al., 2004). However, average sediment EPC0 explained only a small portion (R2 = 0.26) of the variability observed in average SRP concentrations, indicating that many other factors contribute to the regulation and maintenance of dissolved P in Ozark streams.

Benthic sediments might be a source of dissolved P to the overlying water in Ozark streams, because sediment EPC0 was greater than SRP and desorption of P often occurred in our lab studies at low initial SRP concentrations, i.e., generally 0.1 mg PO4–P L–1 plus ambient SRP concentration or less. Sediment EPC0 was generally greater (ANOVA LSD, P < 0.05) in the selected streams draining catchments in the Pasture or Mixed land use category (Fig. 4C). Benthic sediments in Pasture streams, or those draining catchments with a Mixed land use, may assist in maintaining elevated SRP concentrations. Other studies have also shown that benthic sediments adjacent to agricultural lands have a greater P content than those adjacent to forested lands (McDowell et al., 2002). However, municipal effluent discharges in Ozark streams have a much greater impact on sediment EPC0 than catchment land use, where sediment EPC0 downstream from effluent discharges has ranged from ~0.1 to 7 mg L–1 (Popova et al., 2006; Ekka et al., 2006). The selected streams in this study, including the highly urbanized streams, were not influenced from municipal effluent discharges, only nonpoint source impacts.

Sediment Mehlich-3 Extractions
Contents of M3–extractable elements were highly variable between elements and within individual elements (Table 4 ). Contents of M3–extractable Al, Ca, Fe, K, Mg, Mn, and Na exceeded 1 mg kg–1 dry sediment at one or more of the selected sites, although M3–extractable elements were much less than the contents typically observed in soils (e.g., see Sims et al., 2002; Zhang et al., 2005; Brye, 2006). In particular, M3P contents in stream sediments are low (0.27–1.94 mg kg–1 dry sediments) compared to levels considered optimal for productive agricultural soils (~50 mg kg–1) in Arkansas and other regions (e.g., Maguire and Sims, 2002). Contents of M3P in the sediments of these selected streams are much less than typical pasture soils in Arkansas, which range from less than 10 mg kg–1 in a soil never amended with poultry litter to more than 400 mg kg–1 in a soil that has received long-term applications of poultry litter (Daniels et al., 2001).


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

  		Table 4. Phosphorus saturation ratios (PSR and PSRmod) and contents of Mehlich 3 (M3)–extractable elements measured using inductively coupled plasma–optical emission spectrometry (ICP–OES) across the 22 selected sampling sites at Ozark streams in spring 2004 (May through June).

 
Several studies have evaluated the relation between soil M3P content and SRP concentrations in runoff water (Pote et al., 1999; Vadas et al., 2005), and our study showed that sediment M3P was correlated (r = 0.50, P < 0.05) to SRP concentrations in the stream water (Fig. 5B). The slope of linear relation between SRP concentration and M3P content of the benthic sediments (0.022; Fig. 5B) is an order of magnitude greater than that observed by Vadas et al. (2005) between runoff SRP concentrations and M3P content of a wide range of soils (slope = 0.002). Even so, it suggests similar controls may exist between both solid and water phases in determining particulate-bound P release, and that sediment with low M3P contents (i.e., 0.2–2 mg M3P kg–1 dry sediment) might release proportionately more dissolved P into aqueous solution compared to soils with greater M3P contents. Furthermore, sediment EPC0 and M3P content using step-wise linear regression explained about half of the variability observed in SRP concentrations at the selected streams (R2 = 0.50, SRP = 0.002 + 0.105 EPC0 + 0.021 M3P, P < 0.01). Similar to SRP, M3P content in the sediments decreased (r = –0.47, P < 0.05) with an increase in the percent forested lands in the catchments (Fig. 3C).

Mehlich-3 extractions can be effectively used to evaluate P saturation in soils (Sims et al., 2002; Kleinman and Sharpley, 2002; Zhang et al., 2005), and PSRs are good indicators of the potential for P loss through various hydrologic pathways (Maguire and Sims, 2002; Allen et al., 2006). Mehlich-3 PSRs in soils generally range from less than 0.01 to as much as 0.59 (Sims et al., 2002; Zhang et al., 2005). However, M3 PSRs in the sediments of the selected Ozark streams were between 0.02 and 0.11 (Table 4) and on the lower end of the range typically observed in soils. McDowell and Sharpley (2001) observed that M3 PSRs (~0.03, calculated from mean M3Al, M3Fe, and M3P) in bed sediments of an agricultural catchment were within the range observed in our study. Mehlich-3 PSRs of benthic sediments were not significantly correlated to sediment EPC0 or SRP concentration in the water column at the selected streams, as has been observed between M3 PSRs in soils and dissolved P concentrations in runoff water (Sims et al., 2002; Allen et al., 2006) or leachate (Maguire and Sims, 2002).

Mehlich-3 PSR and water-extractable P in calcareous soils have not shown a significant correlation (Ige et al., 2005), and other parameter combinations that included M3Ca and M3Mg provided a significant correlation. We used step-wise linear regression to determine what parameters were related to M3P and found that M3Fe, M3Mg, and Kslope of the linear sorption isotherm were significant factors (R2 = 0.67, M3P = 0.581 + 0.084 M3Fe – 0.071 M3Mg – 0.977 Kslope, P < 0.05); the contents of M3Fe and M3Mn were also highly correlated (r = 0.87, P < 0.001). The removal of M3Fe from the PSR denominator showed that M3P/M3Al in the sediments was not significantly correlated to SRP concentration in the stream water, whereas M3P/M3Fe was positively correlated to SRP concentration at these streams (r = 0.53, P < 0.05). However, PSRmod was correlated (r = 0.71, P < 0.001) with SRP concentration in the water column at the selected streams (Fig. 5C), and substitution of M3Ca in place of M3Mg did not produce a significant correlation.

Some catchment-scale effect on the contents of M3-extractable elements and the degree of P saturation existed, similar to the effects observed on water chemistry and sediment EPC0. The degree of P saturation (i.e., PSRmod) in the sediments generally increased with an increase in the proportion of pasture land use in the catchments (r = 0.80, P < 0.001) (Fig. 3D). Sediments from the selected streams in the Pasture and Mixed land use catchments had greater PSRmod compared to the other land use categories (ANOVA LSD, P < 0.01) (Fig. 4D).


   Conclusions
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results and Discussion
 Conclusions
REFERENCES
 
This study evaluated the variation among water chemistry and sediment P interactions among streams draining catchments that varied in the percent forested, pasture, and urban land use (Objective 1). Soluble reactive P concentrations in the water column and M3P contents of the benthic sediments at the selected Ozark streams decreased with an increase in percent forested area in the catchments, whereas NO3–N concentrations and PSRmod increased with an increase in percent pasture land use within the catchments. Thus, catchment land use affects water chemistry and sediment P parameters measured in this study.

This study also evaluated the relations between SRP concentration, sediment EPC0, and other measured abiotic factors in the stream sediments (Objective 2). Soluble reactive P concentrations in the water column of these selected streams were positively correlated to sediment EPC0, although sediment EPC0 was generally greater than SRP at individual streams. However, sediment EPC0 explained only 26% of the variation in SRP concentrations, indicating that many other abiotic and or biotic factors influence SRP.

This study evaluated the use of the traditional M3 extraction on stream sediments to elucidate other abiotic factors related to SRP concentration in stream sediments (Objective 3). Mehlich-3 P contents in benthic sediments were related to SRP concentration measured in the stream water, and the combination of M3P and sediment EPC0 explained over 50% of the variability in SRP concentrations. Mehlich-3 extractions were useful in estimating PSRs, and a modified PSR (PSRmod) of the benthic sediments was strongly correlated to SRP concentrations in the water column of the selected Ozark streams. Thus, M3P and EPC0 estimates from stream sediments may be valuable yet simple indicators of whether benthic sediments act as sinks or sources of P in fluvial systems, as well as estimating changes in stream SRP concentrations.


   ACKNOWLEDGMENTS
 
This study was funded by the UA Division of Agriculture and the USDA Agricultural Research Service. This project would have not been completed without the field and laboratory assistance from S. Williamson, R. Avery, and A. Erickson, and GIS support from J. Giovannetti. This manuscript benefited from comments provided by A. Sharpley, R. Stoner, and other anonymous technical reviewers. Mention of a trade name, proprietary product, or specific equipment does not constitute a guarantee or warranty by the contributing agencies and does not imply its approval to the exclusion of other products that may be suitable.


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




This article has been cited by other articles:


Home page
 J. Environ. Qual.Home page
R. O. Maguire, G. H. Rubaek, B. E. Haggard, and B. H. Foy
Critical Evaluation of the Implementation of Mitigation Options for Phosphorus from Field to Catchment Scales
J. Environ. Qual., August 24, 2009; 38(5): 1989 - 1997.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
M. D. McDaniel, M. B. David, and T. V. Royer
Relationships between Benthic Sediments and Water Column Phosphorus in Illinois Streams
J. Environ. Qual., February 6, 2009; 38(2): 607 - 617.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
B. E. Haggard and L. D. Bartsch
Net Changes in Antibiotic Concentrations Downstream from an Effluent Discharge
J. Environ. Qual., January 13, 2009; 38(1): 343 - 352.
[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 Haggard, B. E.
Right arrow Articles by Brye, K. R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Haggard, B. E.
Right arrow Articles by Brye, K. R.
Agricola
Right arrow Articles by Haggard, B. E.
Right arrow Articles by Brye, K. R.
Related Collections
Right arrow Wetlands and Aquatic Processes
Right arrow Surface Water Quality
Right arrow Sorption/Exchange

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