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


     

Published online 9 August 2005
Published in J Environ Qual 34:1591-1599 (2005)
DOI: 10.2134/jeq2004.0293
© 2005 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 Related articles in JEQ
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Rauter, A.
Right arrow Articles by Battin, T. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Rauter, A.
Right arrow Articles by Battin, T. J.
Agricola
Right arrow Articles by Rauter, A.
Right arrow Articles by Battin, T. J.
Related Collections
Right arrow Surface Water Quality
Right arrow Fate
Right arrow Organic Compounds
Right arrow Water Pollution

TECHNICAL REPORTS

Surface Water Quality

Transport and Metabolic Fate of Sewage Particles in a Recipient Stream

Andrea Rautera, Gabriele Weigelhofera, Johann Waringera and Tom J. Battinb,*

a Dep. of Freshwater Ecology, Vienna Ecology Center, Univ. of Vienna, A-1090 Vienna, Austria
b Dep. of Freshwater Ecology, Vienna Ecology Center, Univ. of Vienna, A-1090 Vienna, Austria, and Dep. of Ecology, Univ. of Barcelona, 08082 Barcelona, Spain

* Corresponding author (tomba{at}pflaphy.pph.univie.ac.at)

Received for publication July 30, 2004.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 STUDY SITE
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Although the implementation of wastewater treatment plants (WWTP) has dramatically increased the quality of surface waters in urbanized areas, WWTPs can still discharge noticeable amounts of solutes and particles to recipient streams. Although the fate of WWTP nutrients has received considerable attention, transport and in-stream transformation of sewage-derived particulate organic matter (SDPOM) have not. To investigate the transport and transformation of SDPOM in recipient streams, we experimentally injected fluorescently labeled SDPOM into a headwater stream and tracked its downstream fate at baseflow. Most SDPOM disappeared from the streamwater within a 160-m long reach with an average deposition velocity of 0.14 mm s–1. We further coupled hydrometric measurements of specific water fluxes through the streambed interface with a mixing model to estimate streambed oxygen removal, and found significantly higher oxygen removal in the deposition (0.75 g O2 m–2 d–1) than in the downstream post-deposition (0.36 g O2 m–2 d–1) subreach. Contrary to our expectations, we did not detect any apparent effect of SDPOM deposition on streambed clogging. Our results show the capacity of a recipient stream to retain SDPOM and to reduce its downstream export, and thus contribute to a better understanding of ecosystem services of human-altered streams.

Abbreviations: ANOVA, analysis of variance • DO, dissolved oxygen • DTAF, 5-(4,6-dichlorotriazinyl) amino-fluorescein • EMMA, end-member mixing analysis • FITC, fluoresceine isothiocyanate • FPOM, fine particulate organic matter • SDPOM, sewage-derived particulate organic matter • VHG, vertical hydraulic gradient • WWTP, wastewater treatment plant


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
STUDY SITE
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
WHILE SOCIETY APPRECIATES the ecosystem services of streams and rivers, we must also recognize that human activity has severely diminished the integrity and functionality of these ecosystems (Paul and Meyer, 2001; Meyer and Wallace, 2001). Among the most dramatic anthropogenic impacts are channelization, diversion or streambed sealing, along with increasing point and diffuse sources of nutrients and contaminants. This has dramatic implications. First, headwaters, which are particularly vulnerable to human alterations (Meyer and Wallace, 2001) can affect larger ecosystems such as rivers, estuaries, and even oceans through downstream linkage of local biogeochemical processes (Alexander et al., 2000). Second, stream and ground waters are now recognized as one single resource, and, depending on ecosystem health, streams can be a source of contamination to aquifers, or conversely, contaminated aquifers discharging into streams can induce long-term contamination of surface waters (Alley et al., 2002).

Wastewater treatment plants (WWTP) have substantially reduced emissions to surface waters (Grady et al., 1999), yet during malfunctioning or even regular operation, WWTPs can emit solids to the recipient surface waters. Increased solid emissions due to sludge rising, pin floc formation, or slime bulking (Bergh and Olsson, 1996), for instance, occur more often than usually admitted—a fact that is particularly true for older and low-capacity WWTPs in headwater catchments. Sewage-derived particulate organic matter (SDPOM) consists of microbial extracellular polymeric substances, bacteria, fungi, and protozoa, and exhibits high enzymatic activity (Nielsen, 2002). The SDPOM can serve as vehicles for allochthonous bacteria, pathogens, contaminants, antibiotics, and various pharmaceutical substances (Mugan, 1996; Ternes, 1998; Medema and Schijven, 2001; Schwartz et al., 2003). The SDPOM also represents a food source to riverine organisms (DeBruyn and Rasmussen, 2002) and its preferential consumption can be a pathway of contaminant transfer to higher trophic levels. Despite these attributes of SDPOM, very little is known about its impacts on stream functions.

Stream vulnerability and ecosystem health result from the combined effects of nuisance inputs, in-stream transport and transformation. Both transport and transformation link upstream to downstream reaches—a phenomenon that can formally be described by the spiraling concept (Newbold et al., 1981). Here the streambed exerts major controls on solute and particle transport, and drives biogeochemical processes central to ecosystem functioning (Findlay et al., 1993; Battin et al., 2003a; Packman et al., 2003). Through additions of radiolabeled natural seston (Minshall et al., 2000) or various surrogates such as spores (Wanner and Pusch, 2000), yeast (Paul and Hall, 2002), or pollen (Georgian et al., 2003), it has become clear that there is rapid and continuous deposition of particles onto the streambed. These studies show that once deposited, particles are either retained by mixing into the streambed (Packman et al., 2000), by microbial biofilms (Battin et al., 2003a), or are resuspended within a few days (Cushing et al., 1993). Particles can fuel streambed metabolism (Findlay et al., 1993; Battin et al., 2003b), change the sedimentary flow environment, and induce clogging (Packman and MacKay, 2003). Clogging can inhibit hydrodynamic exchange at the streambed interface, isolate deeper zones of the streambed from streamflow, which, in turn, can degrade the habitat and reduce self-purification.

In this study we investigated SDPOM transport in a recipient stream, its possible impact on streambed hydrodynamic exchange, and its metabolic fate at prolonged baseflow. Specifically, we expected the stream reach below the WWTP to have greater metabolic activity and reduced hydrodynamic exchange with the subsurface water than upstream of the WWTP. We postulate the recipient stream as a functional extension of the WWTP by continuing the treatment process, and consider its retention and transformation capacity an important ecosystem service (Costanza et al., 1997) as it contributes to self-purification.


   STUDY SITE
TOP
 ABSTRACT
 INTRODUCTION
 STUDY SITE
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
The study was conducted in a 660 m long reach of a fifth-order stream (Kleine Erlauf, Gresten; 47°59' N lat; 15°1' E long; Austria) draining a largely calcareous catchment (42 km2) with forest (58%), agriculture (38%), and settlements (6%) as the major land-use categories (Fig. 1) . Average annual discharge is 250 L s–1; we conducted experiments at baseflow that ranged from 70 to 90 L s–1 in summer 2003. Streambed sediments are typically composed of gravel (median grain size: 28.3 ± 7.6 mm) with average benthic chlorophyll a of 81.4 ± 40.6 mg m–2. The stream channel is characterized by sequences of riffles and pools with a series of man-made cascades; channel width averages 5 m and slope 0.003.



View larger version (30K):
[in this window]
[in a new window]
  		Fig. 1. Location of the study site in Austria, wastewater treatment plant (WWTP), SDPOM injection and sampling sites, and installation of the minipiezometers along the reference, deposition, and post-deposition subreaches. Each dot represents minipiezometers nested at 10, 35 cm and in some sites at 50 cm (stream width fivefold).

 
The WWTP (Municipality of Gresten) has a capacity of 2420 population equivalents d–1, and an average effluent of 6 L s–1 with SDPOM that increases background concentrations of natural particulate C by 56% on average. Dissolved nutrient concentrations of the effluent typically average 0.4 ± 0.7 mg L–1 for NH4–N, 13.5 ± 7.3 mg L–1 for NO3–N, and 0.8 ± 0.9 mg L–1 for PO4–P, respectively.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 STUDY SITE
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Particle Labeling and Experimental Additions
The SDPOM was directly collected from the WWTP clarifier and fluorescently stained for experimental, whole-stream injections. In a first injection (9 July 2003), we used 5-(4,6-dichlorotriazinyl) amino-fluorescein (DTAF) (Sigma-Aldrich) to stain total SDPOM biomass and exopolysaccharides. The SDPOM was resuspended in 4 L of 0.05 M NaH2PO4 buffer (pH 9.0) and stained with DTAF at a final concentration of 0.2 g L–1 for 2 h at 60°C (Sherr et al., 1987). Because of reduced availability of DTAF from the manufacturer and its elevated costs, we used fluoresceine isothiocyanate (FITC) (Sigma-Aldrich) for the two following additions (31 July 2003 and 2 Sept. 2003); FITC was also applied to microorganisms and particles (Paul and Hall, 2002). The SDPOM was resuspended in 4 L of FITC solution (4 L carbonate–bicarbonate buffer [625 mL 42.0 g NaHCO3 L–1 deionized water and 375 mL 53 g NaCO3 L–1 deionized water; pH = 9.5], 1 g FITC).

We conducted three short-term releases (duration 30 min) immediately downstream of the WWTP effluent with a peristaltic pump (Masterflex) into a well-mixed stream subreach. Before each release, water travel time and plateau conditions (i.e., steady state conditions) were determined from rhodamine injections. Plateau conditions of rhodamine and tracer particles must not be necessarily identical because of possible particle sedimentation in hydraulic dead zones. We countered this possible source of error by collecting triplicate samples for labeled SDPOM over 30 min during plateau conditions; samples were collected at eight sites downstream of the injection point and distributed along a 480-m long reach. Samples were collected from the middepth water column (average depth 20 cm and well mixed) in the thalweg into wide-mouth polyethylene bottles, preserved with formaldehyde (2% final concentration), and kept refrigerated pending further processing. Concentrations of fluorescent SDPOM-particles were determined from direct counts on irgalan-black-stained filters (0.2 µm, Millipore) with an epifluorescence microscope (Nikon Eclipse E800) on 50 randomly selected fields (40x magnification). Averaged triplicates were used for the computation of SDPOM travel length and deposition velocity.

SDPOM Travel Length and Deposition Velocity
Assuming that turbulence keeps fine particles well mixed throughout the water column of streams, the loss of particles from the water column becomes analogous to the uptake of solute tracer, specifically, the quantity of introduced particles remaining in suspension declines exponentially with distance (Newbold et al., 1981). Therefore, a first-order longitudinal loss rate, kp, of injected SDPOM particles can be derived from the following:

[1]
where F(x) is the quantity remaining in suspension at distance x from the point of introduction and F(0) is the quantity introduced. The average transport distance, Sp, for a particle is expressed by the following:

[2]
The Sp describes SDPOM deposition on the scale of the stream, whereas a related parameter, the deposition velocity, vdep, quantifies deposition as a mass-transfer coefficient across the water–sediment interface. The vdep is obtained from the following:

[3]
where u is mean water velocity and h is mean water column depth.

The fall velocity (vfall) of the SDPOM is routinely measured by the WWTP staff, and averaged 0.66 ± 0.03 mm s–1 (Krösbacher, personal communication, 2004).

Based on the average particle travel length (SP), we partitioned the study reach into a deposition, a downstream post-deposition, and a reference reach immediately upstream of the WWTP.

Hydrodynamic Exchange
We installed a total of 120 nested minipiezometers (10- and 35-cm depth) over 20 sites (triplicates per depth) along the study reach (Fig. 1). In some sites, we also installed individual minipiezometers to a 50-cm depth. Minipiezometers were constructed from PVC tubing that was perforated near the bottom with millimeter-sized holes and covered with 100-µm mesh Nitex fabric (Lee and Cherry, 1978). We attempted to install the minipiezometers evenly across each site, but in many cases this was not possible because of bedrock outcrops. Nevertheless, we managed to install replicate minipiezometers within a distance typically ranging from 0.7 to 1.7 m.

Vertical hydraulic gradients (VHG), which indicate the hydrodynamic exchange between subsurface and surface waters, were measured using a hydraulic potentiomanometer attached to the minipiezometers (Winter et al., 1988). The VHG was calculated as follows:

[4]
where {Delta}h is the hydraulic head between subsurface and surface waters, and {Delta}z is the installation depth of the minipiezometer. The VHG is dimensionless and positive values indicate upwelling (i.e., infiltration of subsurface water) and negative values downwelling (i.e., exfiltration of surface water). Sediment permeability (K) was estimated in each minipiezometer and on each date in situ using the Hvorslev slug test (Hvorslev, 1951). Water fluxes were computed according to the Darcy equation as follows:

[5]
where v is the Darcy velocity equivalent to the water flux (volume per time) per unit area of streambed (Freeze and Cherry, 1979).

We also collected water samples for the measurements of temperature, conductivity (WTW LF196), and dissolved oxygen (DO) (WTW OXI 196) from the same minipiezometers.

End-Member Mixing Analysis
If porewater DO concentration was simply the result of conservative mixing of surface and subsurface waters, its concentration would be between the concentrations of these end-members. Deviations from conservative mixing indicate whether the streambed is a source or a sink for DO. We conducted an end-member mixing analysis (EMMA) to differentiate between sources and sinks of DO within the streambed (Battin et al., 2003b). Conductivity was used as a conservative tracer and we assumed that water at a depth of 10 cm within the streambed was a mixture of downwelling streamwater and upwelling subsurface water passing through the 35-cm layer within the streambed. The relative contributions (x and 1 – x) of both end-members were computed for each sampling date by solving the following:

[6]
where C is conductivity and subscripts refer to the end-members (streamwater and 35-cm sediment layer) or the mixture of end-members (10-cm sediment layer). Based on these relative fractions and on end-member DO concentrations, we then predicted DO concentrations that would result in the 10-cm layer from the conservative mixing of streamwater and the ground water (water from 35 cm). We acknowledge that this approach does not include exchange with atmospheric oxygen or the hyporheic longitudinal flow putatively important in gravel streams. Nevertheless, we assume that these components have similar effects in all three study reaches, which makes their comparison reasonable.

Streambed Oxygen Removal and SDPOM Retention
The DO fluxes through the streambed were estimated from the EMMA residuals (mg O2 L–1) and specific water fluxes (m3 m–2 h–1) through the streambed–water interface (Battin et al., 2003b). The upwelling water flux from the streambed into the streamwater was calculated as net flux of upwelling and downwelling, and the subreach (i.e., reference, deposition, and post-deposition) average was weighted for upwelling, downwelling, and no exchange. Dissolved oxygen was converted to units of C, assuming a respiratory quotient of 1 (i.e., 1 g C = 2.67 g O2).

On each of the three surveys, we collected duplicate samples for natural seston into two 3-L polyethylene containers from the WWTP effluent, immediately upstream of the WWTP and from the lower end of the study reach (i.e., 480 m downstream of the WWTP). Samples were filtered onto precombusted glass-fiber filters (GF/F Whatman), ash-free-dry-mass (450°C, 5 h) of seston determined, and organic C derived assuming 50% weight percentage. Seston retention was calculated from simple mass balance using seston concentration and discharge from the stream and the WWTP effluent.

Statistical Analysis
Differences in DO, conductivity, temperature, and EMMA-derived DO residuals between subreaches were tested with analysis of variance (ANOVA) and Mann–Whitney U-test. Significant differences were determined at an error level of {alpha} = 0.05. All data are presented as the mean ± SD.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 STUDY SITE
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
SDPOM Transportation Behavior
The first injection was conducted with lower input concentrations of SDPOM (DTAF stained) than the two subsequent injections with FITC-stained SDPOM. Yet initial concentrations of labeled SDPOM exponentially decreased along the study reach in all three injections with similar loss rates ranging from 0.0056 to 0.0071 (Fig. 2) . Resulting transport distance (SP) of SDPOM averaged 160 ± 25 m, deposition velocity (vdep) 0.14 ± 0.03 mm, and the vdep/vfall ratio 0.21 ± 0.03 (Table 1).



View larger version (32K):
[in this window]
[in a new window]
  		Fig. 2. Relative loss (%) of fluorescent SDPOM from the streamwater with distance from the injection point. Each point was derived from triplicate samples at plateau conditions (see text). Symbols refer to 9 July 2003 ({blacksquare}), 31 July 2003 ({square}), and 2 Sept. 2003 ({circ}).

 

View this table:
[in this window]
[in a new window]
  		Table 1. SDPOM transport variables determined from three individual injections.

 
Median size of all measured particles (N = 1360) was 71 µm (minimum, 4 µm; maximum, 650 µm). Median particle size did not apparently change along three downstream sites (40 m, 59 µm; 250 m, 80 µm; 430 m, 77 µm), yet the coefficient of variation as a measure for the heterogeneity of particle size decreased along the study reach (40 m, 1.16; 250 m, 0.88; 430 m, 0.78).

Hydrodynamic Exchange
The VHG and streambed permeability did not show any consistent pattern between the three subreaches but exhibited considerable heterogeneity both within and between sites (Table 2, Fig. 3) . Upwelling accounted for 56%, downwelling for 34%, and no exchange (i.e., VHG = 0) for 10% of the piezometric measurements (N = 328). Average subreach VHG at a 10-cm depth generally indicated downwelling (i.e., negative values), whereas upwelling clearly dominated in the 35-cm layer and decreased downstream (Table 2). Temporal variation in VHG was mainly given by a minor storm on 29 July 2003. During prolonged baseflow, reach-averaged 10-cm VHG values were positive (0.19 ± 0.20) on 9 July, whereas post-storm VHG values were negative (31 July 2003, –0.40 ± 0.17); they were more positive (0.04 ± 0.10) again after prolonged baseflow on 2 Sept. 2003 (Fig. 3). Streambed permeability ranged from 0.17 x 10–4 to 3.26 x 10–4 m s–1. Averaged values were slightly elevated in the reference reach, yet because of large variation associated with these numbers, no coherent spatial pattern emerged from our measurements (Table 2). Permeability did not show any clear temporal patterns (Fig. 3).


View this table:
[in this window]
[in a new window]
  		Table 2. Hydrodynamic exchange parameters of the Kleine Erlauf streambed for the reference, deposition and post-deposition subreaches. Data represent averages (± SD) of sites (n refers to number of sites per subreach) for all three sampling dates.

 


View larger version (26K):
[in this window]
[in a new window]
  		Fig. 3. Distribution of vertical hydraulic gradients (VHG, a and b) and permeability (c and d) along the study reach. The shaded bar indicates the deposition subreach, preceded by the reference and followed by the post-deposition reach. Symbols indicate mean ± standard deviation from of two to three replicate minipiezometric measurements, and refer to 9 July 2003 ({blacksquare}), 31 July 2003 ({square}), and 2 Sept. 2003 ({circ}).

 
Streambed VHG and permeability values translate into a reach-scale average Darcy velocity of 0.59 m d–1 (weighted for upwelling, downwelling, and no exchange) across all sites, depths, and dates. Average downwelling Darcy velocity across the 10-cm plane was 0.65 m d–1 and the upwelling Darcy velocity across the 35-cm plane was 1.13 m d–1. The large standard deviations precluded clear spatial patterns emerging from these data (Table 2).

Solute Patterns
Streamwater and porewater temperature closely followed the same spatial pattern with a continuous downstream decrease after the WWTP (Fig. 4) . By contrast, conductivity increased downstream, notably in the post-deposition subreach, and showed similar patterns in the streamwater and porewater. Average temperature and conductivity values did not show any noticeable depth gradients (Table 3). There was no apparent gradient from the 35- to the 50-cm depth minipiezometers. Dissolved oxygen exhibited steep depth gradients with concentrations that were on average 1.8-fold lower in the porewaters than in the streamwater. Both the 10- and 35-cm DO concentrations were significantly lower in the deposition than in the reference subreach, with concentrations as low as 2.6 ± 1.4 mg L–1 in the 10-cm porewaters and 3.2 ± 0.4 mg L–1 in the 35-cm porewaters immediately downstream of the WWTP. Dissolved oxygen concentrations remained relatively low downstream with some increase in the lower part of the post-deposition subreach (Fig. 4).



View larger version (31K):
[in this window]
[in a new window]
  		Fig. 4. Spatial distribution of conductivity (a), temperature (b) and dissolved oxygen (c) along the study reach. Streamwater ({blacksquare}), 10-cm ({square}) and 35-cm porewater ({circ}). Symbols indicate the mean ± standard deviation of two to three replicate minipiezometric measurements.

 

View this table:
[in this window]
[in a new window]
  		Table 3. Summary statistics for streamwater and porewater conductivity, dissolved oxygen, and temperature in the three subreaches. Values indicate mean ± SD and number of individual measurements (n).

 
End-Member Mixing Analysis
Considering the direction of hydrodynamic exchange and the distribution patterns of both conductivity and DO concentration, the 10-cm depth could be reasonably assumed as the mixing product of deeper porewaters (i.e., 35 cm) and the streamwater. Using conductivity measurements and the mixing model equation, we found surface water to contribute 17 ± 16% to the 10-cm porewater in the reference subreach, 22 ± 30% in the deposition and 23 ± 34% in the post-deposition subreach. Plotting the EMMA derived predictions of DO concentrations against the observed DO concentrations in the 10-cm porewaters provides first approximation of DO removal in the streambed with clear overestimates in the deposition subreach (Fig. 5) . Average residuals were significantly (Mann Whitney-U, P < 0.05, n = 27) more negative in the deposition subreach (–1.00 ± 1.22 mg O2 L–1) than in the reference (–0.13 ± 0.79 mg O2 L–1) and post-deposition (–0.35 ± 1.69 mg O2 L–1) subreaches.



View larger version (25K):
[in this window]
[in a new window]
  		Fig. 5. Relationship between observed and predicted dissolved oxygen concentrations as derived from the end-member mixing analysis. The line denotes the 1:1 cases were the model predictions fit the observed concentrations. Values under the line indicate oxygen removal. Symbols refer to the reference ({square}), deposition ({circ}), and post-deposition ({blacksquare}) subreaches.

 
Streambed Oxygen Removal and SDPOM Retention
We estimated average DO removal rates within the superficial streambed from water fluxes as derived from subreach averaged Darcy velocities and DO EMMA residuals. Darcy velocities (Table 2) translate into net upwelling water fluxes of 480, 750, and 1030 L m–2 d–1 in the reference, deposition, and post-deposition reach, respectively. Multiplying this specific water flux with the DO residuals yields average rates of DO removal of 0.06, 0.75, and 0.36 g O2 m–2 d–1 in the reference, deposition, and post-deposition reach, respectively. Assuming that this DO removal is attributable to heterotrophic respiration, this would correspond to 0.02, 0.28, and 0.13 g C m–2 d–1 in the respective subreaches.

The SDPOM load into the stream averaged 0.041 g C s–1, whereas seston background load averaged 0.097 g C s–1 just upstream of the WWTP and 0.106 g C s–1 at the downstream end of the study reach. Thus, using simple mass balance, these numbers correspond to an average areal retention rate of 0.92 g C m–2 d–1 along the deposition and post-deposition subreaches (streambed surface area 3038 m2). Assuming similar travel length (i.e., 178 m) for both natural seston and experimentally injected SDPOM, 63% of the total suspended load would be deposited within this deposition subreach based on the exponential models in Fig. 2. Thus, 0.58 g C d–1 m–2 would be retained within the deposition reach. We have no estimates of benthic organic matter from the three surveys. Yet, the annual average from 2002 shows a clear gradient of ash free dry mass associated with the sandy sediment fraction decreasing from 1.67 ± 0.29 g m–2 shortly after the WWTP to 1.07 ± 0.09 g m–2 approximately 400 m downstream.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 STUDY SITE
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
Transport and transformation dynamics of organic particles are essential processes that link headwater streams to rivers and estuaries (Cushing et al., 1993; Minshall et al., 2000; Caraco and Cole, 2004). To predict this metabolic link, it is important to understand the physical transportation and bioavailability characteristics of the particles under consideration. This link depends in fact on the number of deposition events, the length of time a particle resides on or in the streambed, and the bioreactivity of the deposition site itself (Newbold et al., 1981; Cushing et al., 1993; Caraco and Cole, 2004). To our knowledge this study is the first attempt to investigate the transport behavior of SDPOM in a stream ecosystem. We found labeled SDPOM to deposit within a relatively short distance and to influence streambed oxygen environment. Contrary to our anticipation, SDPOM deposition did not induce any measurable inhibition of the hydrodynamic exchange at the streambed–streamwater interface.

The fall velocity and the vdep/vfall ratios of the SDPOM are closely bracketed by published values from natural fine particulate organic matter (FPOM; Minshall et al., 2000; Thomas et al., 2001) and corn pollen (Georgian et al., 2003). Theory suggests that the deposition of particles from a mixed water column should be proportional to the vfall of the particles (Cushing et al., 1993). While some empirical observations confirmed this, others did not (reviewed by Georgian et al., 2003), and our data show that SDPOM vdep is only 21% of vfall. Georgian et al. (2003) suggest that gravitational settling is one mechanism that can cause large vdep—a relationship that may be particularly important for particles larger than 50 to 100 µm (Thomas et al., 2001). Using a wide spectrum of different particles, Georgian et al. (2003) found in fact a relationship between size and the vdep/vfall ratio, and speculate that particle density and water temperature explain some of the observed scattering. While the size of SDPOM typically corresponds to FPOM (53–106 µm) used by others (Cushing et al., 1993; Minshall et al., 2000; Thomas et al., 2001), its physicochemical properties (e.g., density, adhesion affinity) certainly differ from those of natural FPOM. In fact, microbial extracellular polymeric substances, which constitute the bulk building block of SDPOM (Nielsen, 2002), may enhance deposition through adhesion to streambed biofilms. These latter also produce adhesive polymers, and filamentous biofilm structures (streamers) are thought to filter particles from the streamwater (Battin et al., 2003a). Therefore, we suggest that the surface property of particles is an important factor that influences settling and resuspension.

Our results relate to prolonged baseflow, which may narrow their relevance. However, we argue that baseflow is particularly sensitive to sewage effluent because of reduced dilution capacity of the recipient stream. Furthermore, previous work on stable isotopes clearly showed that most of the SDPOM deposits within a 600-m long reach (i.e., the deposition and post-deposition zones at baseflow) on an annual basis accounting for hydrograph variation (Singer et al., 2005).

While it is recognized that proteinaceous N and C increase bioavailability of sewage particles (Lerch et al., 1992), the effects of SDPOM on the resource pool and metabolism of headwater streams is unknown. Our results present evidence that SDPOM accumulated (supported by the gradient in benthic organic matter) within the major deposition zone significantly affects streambed metabolism at baseflow. In fact, as derived from EMMA and hydrometric estimates of specific water fluxes, we found 12 and 2 times higher respiratory C removal in the deposition subreach than in the reference and post-deposition subreaches, respectively. Assuming that the entire metabolism (i.e., 0.28 g C d–1 m–2) is supported by particulate organic C, approximately 48% of the particles retained (0.58 g C d–1 m–2) are metabolized within the deposition subreach. Whereas a more conservative approach considering that only 60% of the estimated metabolism (i.e., 0.17 g C d–1 m–2) is supported by particulate C (Findlay et al., 1993; Battin et al., 2003b) would yield 30% of the retained particles being metabolized at prolonged baseflow. These data clearly underscore the efficiency of the streambed to retain and metabolize particulate organic C that would otherwise be transported downstream. In fact, as revealed by our experimental additions of SDPOM, the post-deposition subreach receives only approximately 40% of the SDPOM at baseflow, and as suggested by the elevated oxygen removal rates processes a substantial amount of this C.

We are aware that our budgetary estimates are prone to variable sources of error, yet we need to stress that they represent first in situ estimates of SDPOM retention and metabolism in a stream. Uncertainties related to the hydrometric approach include inaccurate hydraulic head measurements and scaling-up from specific discharge measurements at the meter scale to reach average fluxes (Harvey et al., 1996). Similarly, multiple sources of errors may blur the EMMA results, and yet the approach can provide valuable estimates for streambed oxygen removal (Battin et al., 2003b); this should be particularly true for our between-reach comparison assuming the same error in each reach. We solved this possible methodological bottleneck by measuring vertical hydraulic gradients, sediment permeability, and pore-water chemistry in a large number of minipiezometers during each survey. Nevertheless, we found streambed oxygen removal rates reasonably close to those published from the Steina Creek (0.64 g C m–2 d–1; Pusch, 1996) and White Clay Creek (0.38 g C m–2 d–1; Battin et al., 2003b). The lower estimates (0.02 g C m–2 d–1) in the reference subreach are most likely due to a thin topmost layer of the streambed that was nearly impervious and impeded hydrodynamic exchange.

Contrary to our initial expectations, we did not find any apparent effect of SDPOM deposition and accumulation on the hydrodynamic exchange at the streambed interface, yet our results document intriguing heterogeneity of vertical hydraulic gradients and sediment permeability. Laboratory flume experiments showed that kaolinite clay deposition in a sand bed clogged the streambed surface, isolated deeper zones of the bed from stream-flow, and reduced particle deposition (Packman and MacKay, 2003). Similarly, Packman et al. (2003) found sediment clogging to reduce the deposition velocity of natural seston and showed that removal of fines from the streambed increased seston deposition in stream-side flumes.

Considering these experimental findings, we suggest several reasons why we did not find any effect of continuous SDPOM deposition on hydrodynamic exchange. First, stream invertebrates feed on SDPOM (DeBruyn et al., 2003), which we found to almost double invertebrate secondary production (approximately 0.3 g C m–2 d–1) in our study site (G. Singer, unpublished data, 2004). This would imply that invertebrates continuously rework deposited SDPOM through ingesting and resuspension, and thereby change the sedimentary environment (Mermillod-Blondin et al., 2003). Second, due to its aggregation properties, accumulated SDPOM forms a flocculent layer on the streambed, which on the one hand is prone to flood-induced disturbance and, on the other, makes further penetration of particles into streambed interstices and hence clogging difficult. The SDPOM certainly subsidizes microbial heterotrophs, and it could be argued that microbial growth clogs interstitial spaces as it was postulated for a large-river impoundment (Battin and Sengschmidt 1999). Yet, sediment structure and the dynamic nature (i.e., frequent storms) of headwater streams make this explanation rather unlikely. Third, we cannot exclude that the level of resolution of our field approach was simply too low to detect SDPOM-induced variation in hydrodynamic exchange. Yet, we want to stress that findings from laboratory experiments need to be tested under field conditions, and that the hydrometric method as used in this study represents a very valuable step toward this ecosystem approach.

While natural inputs of allochthonous organic matter into headwater streams have been extensively studied (Webster and Meyer, 1997), anthropogenic inputs such as SDPOM have not. Inputs of SDPOM can be considered as spatial subsidies (Polis et al., 2004) that affect not only food webs in the recipient streams (DeBruyn and Rasmussen, 2002; Singer et al., 2005) but may also disrupt the longitudinal continuum in lotic networks. In fact, the river continuum concept (Vannote et al., 1980) postulates metabolic linkage between upstream and downstream ecosystems, which can be disrupted by damming, for instance, as predicted by the serial discontinuity concept (Stanford and Ward, 2001). Our study shows that SDPOM is removed from the streamwater within a relatively short distance, where it supports streambed metabolism. Thus, both deposition and metabolism of SDPOM reduce its downstream export. We consider this as an indication that subsidized SDPOM does not greatly affect the longitudinal linkage—at least not at baseflow—and further suggest this as an ecosystem service of the recipient stream. In conclusion, our study contributes to a better understanding of anthropogenic inputs to streams. This is crucial to value ecosystem services and a prerequisite for the great needs of restoration and the persisting supply of high quality water as aimed by the European Water Framework Directive.


   ACKNOWLEDGMENTS
 
This study was supported by the EU-Project STREAMES (EV1-CT-2000-00081), a Ramon y Cajal and the Hochschuljubiläumsstiftung der Stadt Wien to TJB. We greatly acknowledge the field assistance of the WWTP staff Gresten.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 STUDY SITE
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 


Related articles in JEQ:

This Issue in Journal of Environmental Quality

JEQ 2005 34: v. [Full Text]  




This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Related articles in JEQ
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Rauter, A.
Right arrow Articles by Battin, T. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Rauter, A.
Right arrow Articles by Battin, T. J.
Agricola
Right arrow Articles by Rauter, A.
Right arrow Articles by Battin, T. J.
Related Collections
Right arrow Surface Water Quality
Right arrow Fate
Right arrow Organic Compounds
Right arrow Water Pollution

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