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 (21)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Marti, E.
Right arrow Articles by Sabater, F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Marti, E.
Right arrow Articles by Sabater, F.
Agricola
Right arrow Articles by Marti, E.
Right arrow Articles by Sabater, F.
Related Collections
Right arrow Surface Water Quality
Right arrow Nutrient Cycling
Right arrow Nutrient Management
Right arrow Nitrogen
Right arrow Phosphorus
Published in J. Environ. Qual. 33:285-293 (2004).
© ASA, CSSA, SSSA
677 S. Segoe Rd., Madison, WI 53711 USA

TECHNICAL REPORTS

Surface Water Quality

Nutrient Retention Efficiency in Streams Receiving Inputs from Wastewater Treatment Plants

Eugènia Marti*,a,d, Jordi Aumatella, Lluís Godéb, Manel Pocha and Francesc Sabaterc

a Laboratori d'Enginyeria Química i Ambiental, Facultat de Ciències, Universitat de Girona, Campus Montilivi, s/n, 17071 Girona, Spain
b Agència Catalana de l'Aigua, Departament de Medi Ambient, Generalitat de Catalunya, Provença 204-208, 08036 Barcelona, Spain
c Departament d'Ecologia, Universitat de Barcelona, Diagonal 645, 08028 Barcelona, Spain
d Centre d'Estudis Avançats de Blanes (CSIC), Accés a la Cala St. Francesc 14, 17300 Blanes, Girona, Spain

* Corresponding author (eugenia{at}ceab.csic.es).

Received for publication October 2, 2002.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
We tested the effect of nutrient inputs from wastewater treatment plants (WWTPs) on stream nutrient retention efficiency by examining the longitudinal patterns of ammonium, nitrate, and phosphate concentrations downstream of WWTP effluents in 15 streams throughout Catalonia (Spain). We hypothesized that large nutrient loadings would saturate stream communities, lowering nutrient retention efficiency (i.e., nutrient retention relative to nutrient flux) relative to less polluted streams. Longitudinal variation in ambient nutrient concentration reflected the net result of physical, chemical, or biological uptake and release processes. Therefore, gradual increases in nutrient concentration indicate that the stream acts as a net source of nutrients to downstream environments, whereas gradual declines indicate that the stream acts as a net sink. In those streams where gradual declines in nutrient concentration were observed, we calculated the nutrient uptake length as an indicator of the stream nutrient retention efficiency. No significant decline was found in dilution-corrected concentrations of dissolved inorganic nitrogen (DIN) and phosphate in 40 and 45% of streams, respectively. In the remaining streams, uptake length (estimated based on the decline of nutrient concentrations at ambient levels) ranged from 0.14 to 29 km (DIN), and from 0.14 to 14 km (phosphate). Overall, these values are longer (lower retention efficiency) than those from nonpolluted streams of similar size, supporting our hypothesis, and suggest that high nutrient loads affect fluvial ecosystem function. This study demonstrates that the efficiency of stream ecosystems to remove nutrients has limitations because it can be significantly altered by the quantity and quality of the receiving water.

Abbreviations: ACA, Catalan Water Treatment Agency • DIN, dissolved inorganic nitrogen • DOC, dissolved organic carbon • WWTP, wastewater treatment plant


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
A TIGHT LINKAGE between terrestrial and aquatic ecosystems is indicated by the influence of watershed processes on stream water nutrient concentrations (Dillon and Kirchner, 1975; Webster et al., 1983; Likens and Bormann, 1995). Direct anthropogenic inputs greatly affect terrestrial solute exports from either point sources (e.g., wastewater treatment plants), or through diffuse pathways (Casey et al., 1993; Jordan and Weller, 1996), which increase stream loads. Anthropogenic nutrient sources are a major cause of large increases in stream nutrient concentrations measured worldwide, often with documented declines in water quality (Howarth et al., 1996). Degradation of water quality is a universal issue because it directly affects human health and the ecological integrity of freshwater ecosystems.

Human well-being and economic development are highly dependent on freshwater quantity and quality. There is an obvious need to implement management strategies that minimize degradation of freshwaters. During the last 50 yr, massive effort has been directed on developing wastewater treatments to reduce direct sewage input to streams (Tchobanoglous and Burton, 1991). In this sense, wastewater treatment plants (WWTPs) have certainly improved water quality. However, some aspects of the water quality problem remain unsolved because (i) nonpoint sources are difficult to control by infrastructure, (ii) the removal efficiency of WWTP nutrient loads is technologically limited, and (iii) world population and associated activity continue to increase. Under these conditions there is a need to identify and implement alternative solutions to enhance water quality.

Physical, chemical, and biological processes within streams influence the transport, transformation, and retention of nutrients during downstream transport (Stream Solute Workshop, 1990). Several studies have reported high efficiency of first to third order streams in nutrient retention (Mulholland et al., 1985; Triska et al., 1989; Munn and Meyer, 1990; Marti and Sabater, 1996; Valett et al., 1996; Marti et al., 1997; Peterson et al., 1997; Mulholland et al., 2000; Peterson et al., 2001). This intrinsic property, the so-called "self-purifying capacity," of stream ecosystems could partially ameliorate water quality problems by reducing the nutrient loads within relatively short distances (Elosegui et al., 1995) if this capacity is not overwhelmed by excessive nutrient loading. Most existing data on nutrient retention efficiency has been obtained from nearly pristine streams, and there is limited study on retention under "polluted conditions" (Haggard et al., 2001). Based on the "Subsidy-Stress Hypothesis" (Odum et al., 1979), we hypothesized that large, long-term nutrient loads (such as those from WWTP effluents) would stress the stream communities and lower the nutrient retention efficiency relative to less polluted streams. The objective of this study was to examine the effect of large nutrient inputs, caused by point sources, on stream nutrient retention efficiency (N and P) by (i) evaluating the effects of nutrient inputs from WWTPs on nutrient loads and measuring stream nutrient retention efficiency downstream of the WWTPs inputs; and (ii) comparing our results to those reported from nonpolluted streams. This study represents one of the first attempts to characterize stream nutrient retention efficiency under polluted or nonpristine conditions.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Study Site
This study was conducted in Catalonia, a region of 31895 km2 located in NE Spain (Fig. 1) . Population density in this region is 193 people km–2. Over centuries, the landscape has been highly modified by agriculture and, more recently, by industry and tourism. Throughout this region we selected 15 streams to examine the effect of nutrient inputs from WWTP on stream nutrient retention efficiency.



View larger version (24K):
[in this window]
[in a new window]
  		Fig. 1. Location of study sites within the Catalonian region (40°30'–42°30' N and 0°15'–3° 15' E).

 
The selected streams covered a wide range of environmental settings characteristic of Catalonia. These streams drain a variety of geomorphologic and subclimatic settings (Table 1); however, most were located in calcareous and lowland watersheds characteristic of the region. In terms of stream hydrology at the watershed scale, half of the streams were permanent and half of them were intermittent (Table 1). This region is dominated by a Mediterranean climate with subclimate differences dependent on altitude and distance from the coast (Table 1). The study streams have some commonalties. They all are second- to third-order tributaries of major rivers draining the Catalonian region, and are located in watersheds highly influenced by agricultural and urban–industrial activity. In all sites, discharge can vary orders of magnitude within and between years. Low flow is typically summer and winter; peak flows, if they occur, are usually spring and fall. From an environmental policy perspective, these streams drain areas considered sensitive—according to the 91/271/EEC European Directive on urban sewage—by the Catalan Water Treatment Agency (ACA). This government agency is responsible for directing water quality improvement within the Catalonian domain. A primary objective of the ACA is to control the quality of the effluents from WWTPs to minimally impact the receiving fluvial ecosystems, especially those classified as sensitive.


View this table:
[in this window]
[in a new window]
  		Table 1. Physical characteristics of the study sites at the watershed and the reach scales. Subclimate settings are within the Mediterranean climate. Lithology indicates the dominant geology within the watershed. Hydrology emphasizes whether stream flow at watershed scale is continuous over time (permanent) or not (intermittent). Reach characteristics data refer to the selected study reach within each stream. Channel width is the average of several measures done along the reach.

 
In each stream we selected a reach located downstream of a WWTP outfall. All WWTP were of medium size, between 10000 and 100000 inhab.-eq; where 1 inhab.-eq is the biodegradable organic matter load equivalent to a BOD5 (biochemical oxygen demand) of 60 g O2 d–1. These WWTPs had either activated sludge (67%) or lagoon (33%) biological treatments, but none chlorinated the outflow of sewage water or had the technology to actively remove N or P.

The length of the selected reaches ranged from 200 m to 2 km depending on discharge. Location of the reaches cover a wide altitudinal range (Table 1). The study reaches also cover a variety of channel morphologies and dominant substrata types (Table 1). No tributaries joined the streams along any study reach.

Methods
Nutrient retention efficiency for each stream was determined from the longitudinal patterns in stream nutrient concentration along the study reach. We sampled water at several sites (at least four) along each reach. Sample collection was completed within <2 h to ensure minimal variation in effluent volume and element composition during sample collection. We also collected water samples from the WWTP outflows (hereafter referred to as effluents) and from the streams above the effluent input. All water samples were analyzed for ammonium-N (NH4–N), nitrate-N (NO3–N), nitrite-N (NO2–N), phosphate-P (PO4–P), dissolved organic carbon (DOC), and chloride (Cl–). Water collections were made on two dates, May and August 1998. During this period water discharge is lowest; therefore, WWTP impact on water quality is presumably greatest (Gasith and Resh, 1999). On each date, we measured the stream discharge upstream of the effluent input, at the WWTP outflow, and at some locations downstream of the effluent input. Discharge was estimated from width, depth, and water velocity measurements. Water velocity was the average of several measures done in a cross-section of the stream using a current meter (NEURTEK Instruments S.A.).

Water samples were kept on ice and transported to the laboratory where they were filtered on arrival through pre-washed fiberglass filters (Whatman GF/F). Filtered samples were kept in the refrigerator until analyzed. All chemical analyses were completed within a week after sample collection. Ammonium-N concentration was analyzed using a Continuous Flow Analysis system (Ammonia Analyzer mod. 255, European Analytical & Scientific Instruments, Technologies S.A.). Chloride, NO3–N, NO2–N, and PO4–P were analyzed using the capillary electrophoresis technique [Waters, CIA-Quanta 5000; Romano and Krol (1993)]. The DOC was analyzed using a Skalar 12SK TOC analyzer with UV/promoted persulfate oxidation.

We estimated the relative contribution of the WWTP to the overall stream Cl– and dissolved inorganic nutrient loads by comparing the amount of Cl– and nutrients transported by the stream after the effluent input to the amount delivered in the effluent. Chloride and dissolved inorganic nutrient loads (kg d–1) in the stream, above the WWTP, and in the effluent were calculated by multiplying concentration (mg L–1) times discharge (L s–1).

Longitudinal variation in Cl– (i.e., a hydrologic tracer) concentration was used to estimate dilution along the reach. Because there were no tributaries along the reaches and the effluent constituted a major source of Cl– and nutrients to the stream, decreases in Cl– along the reach were assumed to be caused by lateral or vertical subsurface-to-surface water inflow through diffuse sources that were more depleted in Cl– concentration relative to water surface. Percentage of water dilution (D) at each sampling site along the reach was calculated using the following equation:

where Clx and Cl1 are Cl– concentrations at the sampling Site x and at the head of the reach (where the effluent has completely mixed with the stream water), respectively.

The longitudinal pattern in stream Cl– relative to nutrient concentrations downstream of the WWTP was used to estimate nutrient retention efficiency. Patterns of N were examined for NH4–N, NO3–N, and total dissolved inorganic N (DIN = NH4–N + NO2–N + NO3–N). Downstream changes in ambient nutrient concentration below the WWTP reflected the net result of physical, chemical, or biological uptake and release processes, whereas changes in Cl– were caused only by physical dilution. Therefore, correction of longitudinal nutrient (P and N) patterns by the longitudinal pattern of Cl– concentration excluded those changes in nutrient concentrations that were due to dilution. Variation in the relative rate of chemical and biological uptake and release processes determines whether the stream acts as a net source (i.e., gradual increase along the reach) or a net sink (i.e., gradual decline along the reach) of nutrients to downstream environments.

In cases where chloride-corrected nutrient concentrations gradually declined along the study reach (i.e., the stream was acting as a net sink of nutrients), we calculated the nutrient uptake length as an index of the stream nutrient retention efficiency (Newbold et al., 1981). That is, the amount of nutrient being removed relative to the nutrient flux being transported. In contrast to retention rates calculated using the nutrient budget approach, the estimation of nutrient uptake length is not dependent of the size of the selected reach. Therefore, the use of uptake lengths allowed comparison of results from reaches of different length. Shorter uptake lengths indicate higher retention efficiencies than longer distances. In the past, nutrient uptake length has been estimated based on short-term additions of low nutrient levels together with a conservative tracer (Mulholland et al., 1985; Munn and Meyer, 1990; D'Angelo and Webster, 1991; Marti and Sabater, 1996). In this study, we used the long-term addition of chloride and high nutrient levels from the WWTP into the stream to estimate nutrient uptake length. Nutrient uptake length was calculated as proposed by Marti et al. (1997) for those cases in which uptake lengths were estimated based on the net decline of nutrient concentrations at ambient levels. Briefly, if any net uptake occurred along the reach (i.e., longitudinal decline in ambient nutrient concentration), the variation of nutrient concentration along the reach can be described as follows:

where N is nutrient concentration and Cl is chloride concentration at Station 1 (first sampling point below the WWTP input) and the downstream points (x distance from Station 1), and the inverse of the slope (–1/b) is the nutrient uptake length from nutrient concentration decline at ambient levels.

Nutrient concentrations and nutrient uptake lengths obtained on the two sampling dates were compared using a Student's paired t-test to examine if there were significant differences on these parameters between dates. Student's paired t-test was also used to compare nutrient concentrations between the stream, upstream of the WWTP input, and the WWTP effluent. Relationships among concentration of the different nutrients considered were examined using Pearson correlation analysis. We also examined the relationship between nutrient uptake length (as dependent variable) and stream discharge, NH4–N, NO3–N, DIN, PO4–P, and DOC concentrations (as independent variables) using linear regression analyses. The effect of the parameters that characterize the study streams at watershed and reach scale (see Table 1) on nutrient uptake length was examined using one-way ANOVA. Variables were log-transformed to meet normality requirements. Significance level used for all tests was p < 0.05. The SYSTAT computer program (Wilkinson, 1989) was used for all statistical analyses.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Comparison of Wastewater Treatment Plant Effluent and Stream Chloride and Nutrient Loads
The WWTP outflows ranged from 4 to 219 L s–1 and the stream discharges (measured upstream of the effluent input) ranged from 0 (dry) to 110 L s–1 (Table 2). Outflows and stream discharges were slightly lower in August than in May (Table 2). The WWTP outflow exceeded stream discharge in 77 and 73% of the study cases in May and August, respectively (Table 2).


View this table:
[in this window]
[in a new window]
  		Table 2. Comparison between stream discharge upstream of the effluent input (Qstr, L s–1) and the wastewater treatment plant outflow (Qeff, L s–1). Study sites are grouped in three categories according to the stream/effluent discharge ratio. The percentage of stream water dilution (due to lateral diffuse inputs) at the end of the study reach is also shown. Data are from the two sampling dates. Site code definitions are provided in Table 1.

 
Nutrient and Cl– concentrations did not significantly differ between sampling dates. Therefore, data from the two dates were combined to characterize the streams as well as effluents Cl– and nutrient concentrations. The range of Cl– concentration was greater in the effluent than upstream of the WWTP (Table 3). Combining the sites where the stream flowed upstream of the WWTP, Cl– concentrations downstream of the effluent increased on average (±1 SEM) 83.5 ± 46 mg L–1 from background concentrations. Chloride loads from the effluent averaged 3.6 times greater than above the WWTP, representing a contribution of 59.3 ± 8.7% of Cl– transported downstream of the WWTPs (Table 4).


View this table:
[in this window]
[in a new window]
  		Table 3. Range and mean (±1 SEM) concentrations of chloride, dissolved inorganic N forms (NH4–N, NO3–N, NO2–N, and DIN = NH4 + NO3 + NO2), PO4–P, and dissolved organic C (DOC) for all the study sites on the two sampling dates (May and August). Table compares data from the streams, upstream of the wastewater treatment plants (WWTP), to those from the WWTP effluents.{dagger}

 

View this table:
[in this window]
[in a new window]
  		Table 4. Percent contribution of the wastewater treatment plant (WWTP) effluent to the chloride and nutrient load transported by the stream below the WWTP input. Data are only for streams with discharge above the WWTP.{dagger}

 
As with Cl–, the ranges of NH4–N, NO2–N, NO3–N, PO4–P, and DOC concentrations above the WWTPs were narrower than in the effluents (Table 3). Mean concentrations of NH4–N, NO2–N, PO4–P, and DOC in the effluents were significantly higher (20.7, 8.5, 12.2, and 2.4 times, respectively) than in the streams, but not NO3–N. In some streams nutrient concentrations upstream of the WWTP were relatively high. Streams and effluents differed in the nutrient quantity, and, in the case of N, in nutrient composition. The DIN was dominated by NO3–N in the streams (mean ± 1 SEM = 71 ± 8%) and by NH4–N in the effluents (mean ± 1 SEM = 54 ± 9%). On average, NO2–N represented 13.7 and 12.6% of the DIN in the effluents and the streams, respectively. Concentrations of NH4–N and NO3–N in the effluents were inversely related (p = 0.01, r = –0.53, n = 28). Above WWTPs, there was a positive relationship between NH4–N and PO4–P (p = 0.05, r = 0.51, n = 15), and between NH4–N and DOC (p = 0.03, r = 0.59, n = 13).

Where streams flowed upstream of the WWTP (54% of the cases), the effluent contributed a large proportion of the stream nutrient load (Table 4). On average, the WWTP contribution for PO4–P was greater than for DIN or DOC, and the fraction of stream NH4–N load that came from the effluent was greater than the fraction of NO3–N (Table 4).

Longitudinal Patterns in Solute Concentrations Downstream of the Wastewater Treatment Plant
The Cl– concentration along the reaches either decreased or remained relatively constant. Based on the longitudinal patterns of Cl– concentration, the percentage of water dilution at the bottom of the study reaches ranged between 0 (no dilution) to 57.5% (Table 2). Combining all sites and dates together, dilution was <10% in 54% of the cases.

After correcting nutrient concentrations for dilution, we observed a consistent increase along the reaches in 27% of the cases for NH4–N, 35% for NO3–N, 31% for DIN, and 29% for PO4–P (Table 5). The longitudinal pattern was unclear in 11% of the cases for NH4–N, 22% for NO3–N, 8% for DIN, and 4% for PO4–P. In the remainder, concentrations gradually declined along the reach. Uptake lengths for NH4–N, NO3–N, DIN, and PO4–P varied quite broadly among sites and between dates (Table 5), but averages were in the order of km (mean ± 1 SEM, 3.2 ± 1.6 km for NH4–N, 6.5 ± 3.1 km for NO3–N, 5.3 ± 1.8 km for DIN, and 3.7 ± 1.0 km for PO4–P). A broader range of nutrient uptake lengths was observed in May than in August (Table 5), although paired t-tests indicated no significant differences between the two dates. We did not find significant differences between uptake lengths of the different nutrients. There were no significant relationships between nutrient uptake length and stream discharge or concentration of NH4–N, NO3–N, DIN, PO4–P, and DOC. We also conducted one-way ANOVAs with the uptake length data using watershed lithology and hydrology, and reach morphology and substrata type (see the parameters reported in Table 1) as factors to examine the influence of these physical factors on the nutrient retention efficiency. Results from these analyses indicated that watershed lithology had a significant effect on NO3–N uptake length (F = 8.053, p = 0.02, n = 10) and watershed hydrology had a significant effect on PO4–P uptake length (F = 5.014, p = 0.04, n = 16). The NO3–N uptake length was shorter in streams located in calcareous (2.2 ± 0.8 km) than in siliceous (23.4 ± 8.5 km) watersheds, and PO4–P uptake length was shorter in intermittent streams (2.1 ± 0.6 km) than permanent streams (6.3 ± 2.1 km). No significant effects were found for the rest of nutrients or for the rest of factors cited above.


View this table:
[in this window]
[in a new window]
  		Table 5. Longitudinal patterns in nutrient concentrations along the study reaches on the two dates. Study sites are listed following the same order as in Table 1.{dagger}

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 REFERENCES
 
Effects of Wastewater Treatment Plants on Stream Nutrient Loads
Dissolved inorganic nutrient loads from the WWTP were a large proportion of total stream load. Water quality of the studied streams was significantly depleted, according to ACA accepted standards, by this source of nutrients and organic matter. Water quality shifted from being acceptable for a wide range of uses to only acceptable for minimal uses. Our results support previous findings that point sources continue to be a major threat to local water quality (House and Denison, 1997; Robson and Neal, 1997; Koning and Roos, 1999). Water inputs from WWTP increased nutrient availability and, in the case of N, shifted the dominant form in transport. This shift from NO3–N to NH4–N has negative implications for the stream biota, especially fish communities (Miltner and Rankin, 1998; Benito and Puig, 1999).

The effect of effluent on nutrient loads was magnified by low stream discharge. In fact, discharge downstream of the WWTP was dominated by WWTP effluent, representing almost 100% of flow, in more than half the streams. Under these conditions, the quantity and quality of water released from WWTPs are crucial to stream ecosystem health. Although the scarcity of surface flow is an intrinsic feature of arid streams, drying may become widespread in more mesic regions in the future due to the increasing human demand for water (Petts, 1994; Gasith and Resh, 1999). Therefore, nutrient point sources will further degrade stream water quality because reduced stream flow will result in reduced point source dilution. This fact can be especially relevant in those cases where water is a source of domestic water supply or irrigation.

European legislation regulates the quality of the WWTP effluent based on solute concentrations. According to the European Community Directive 91/271/EEC, concentrations of total N and P in effluent should not surpass 15 mg N L–1 and 2 mg P L–1, respectively, if the WWTP (i) operates in urban concentrations of 10000 to 100000 inhab.-eq., (ii) has the mechanisms to eliminate nutrients, and (iii) discharges in a "sensitive area." Applying these established thresholds to arid and semiarid regions indicates that stream water cannot be used for any human activity at the high end of permitted concentration. In addition, these levels far exceed average natural concentrations measured in fluvial ecosystems (Meybeck, 1982). This conclusion raises the issue of how to identify the most appropriate strategy to minimize the impact of nutrient point sources on the stream ecosystems. Results from our study suggest that established effluent threshold concentrations are not the most suitable strategy to reduce the impact of WWTP on stream water quality, especially when selected thresholds are so high. This conclusion is particularly applicable in arid and semiarid regions, which account for two-thirds of the world's continental area.

Nutrient Retention Efficiency in Streams with High Nutrient Loads
Based on the longitudinal pattern of nutrient concentrations observed along the reaches, nutrients were being transported downstream without any significant net retention in one-third to one-half of the streams studied. The consistent increase in nutrient concentration corrected for dilution found in some streams indicates that sources in addition to the WWTP exceed assimilation capacity (i.e., streams acted as a net source of nutrients). These additional sources could be (i) lateral and vertical subsurface inputs along the reach (Dent and Grimm, 1999), or (ii) a dominance of in-stream release processes (e.g., precipitation–dissolution dynamics, mineralization of organic matter, and nitrification). Dilution in some of the streams suggests some diffuse water sources along these streams. However, there are no available data on nutrient concentrations from these additional groundwater sources to support hypothesis (i). In 40% of all cases, declines in NH4–N concentration correlated with increases in NO3–N concentration, suggesting partial nitrification of effluent NH4–N during transport, supporting hypothesis (ii) for NO3–N. In those cases, NH4–N was being transformed rather than retained. Alternatively, a lack of a longitudinal pattern in nutrient concentration could be caused by the inhibition of in-stream processes due to some substance in the effluent. Unfortunately, the scope of our data was insufficient to examine this hypothesis.

More than half of the streams had a consistent longitudinal decline in nutrient concentrations corrected for dilution (i.e., streams acted as a net sink of nutrients). The nutrient retention efficiency (expressed as uptake length) varied widely among streams, between dates, and among nutrients. Nevertheless, average uptake lengths for NH4–N, NO3–N, and PO4–P were in the order of km, suggesting low efficiency of nutrient retention in these polluted streams. In other words, these streams retained a small fraction of the dissolved inorganic nutrients that were transported. Uptake lengths from this study were similar to those reported by Reddy et al. (1996) for a stream receiving large P inputs from the watershed, and those from Haggard et al. (2001) for a stream receiving water from a WWTP effluent. Combined, these studies indicate that the footprint of local urban and industrial activity has implications for fluvial ecology both locally (i.e., large effect of WWTP on stream nutrient loading) and regionally (i.e., large nutrient loads are transported over long distances affecting water quality further downstream).

Among all the physical parameters considered in this study (i.e., watershed lithology and hydrology, or reach morphology and substrata type, Table 1), differences in uptake lengths among sites were affected only by watershed lithology in the case of NO3–N and watershed hydrology in the case of PO4–P. These results partly agree with findings from pristine streams by Munn and Meyer (1990) and Marti and Sabater (1996) in which they showed that watershed geology as well as reach morphology and substrata type affect nutrient retention efficiency. These similar findings between pristine and polluted streams suggest that the physical template, at least at watershed scale, for NO3–N and PO4–P has a strong effect on stream nutrient retention efficiency.

Comparison of Nutrient Retention Efficiency between Pristine and Polluted Streams
Studies from the last 20 yr have revealed that streams have a high capacity to transform and retain part of the nutrients in transport (Meyer et al., 1988; Stream Solute Workshop, 1990; Peterson et al., 2001). In contrast to streams in this study, most previous work has been conducted in nearly pristine streams with low background nutrient concentrations. This difference allows comparison of nutrient retention efficiency in WWTP-dominated streams with those in pristine environments. This comparison was done under the consideration that stream discharge influences nutrient uptake length (Stream Solute Workshop, 1990). The PO4–P and NH4–N uptake lengths from a broad variety of nonpolluted streams have been shown to be positively related to stream discharge (Butturini and Sabater, 1998; Peterson et al., 2001; Fig. 2a and 2b) . As discharge increases, the stream becomes less efficient in retaining nutrients (i.e., longer uptake lengths). Because most biological activity in small to mid-size streams is associated with the stream bottom, contact between available nutrients and "active" sediments diminishes at high discharges, increasing nutrient uptake length. In contrast to previous findings, PO4–P and NH4–N uptake lengths from our study do not show any significant relationship with stream discharge (Fig. 2a and 2b). Additionally, we examined the uptake length–stream discharge relationship for NO3–N in nonpolluted streams and in polluted streams and found no significant relation in either case (Fig. 2c). For nonpolluted streams this contrasted with results for PO4–P and NH4–N and suggests that other factors may obscure the relationship. One possible explanation is the fact that the range of discharge for NO3–N uptake length available data is much narrower than that for PO4–P and NH4–N (Fig. 2c). Uptake lengths for NH4–N, NO3–N, and PO4–P from polluted (high nutrient concentrations) streams were on average an order of magnitude longer than those from nearly pristine (low nutrient concentrations) streams with similar discharge (Fig. 2). One possible cause of this difference can be the fact that uptake lengths from this study reflect the net decline (uptake minus regeneration) in ambient nutrient concentration, whereas uptake lengths from short-term nutrient additions conducted in pristine streams primarily reflect gross nutrient removal (Marti et al., 1997). Nevertheless, differences between NO3–N uptake lengths from either natural decline or short-term nutrient additions for a nonpolluted stream (Marti et al., 1997) were smaller than those found in this study between polluted and nonpolluted streams. All together, these results suggest that stream nutrient retention efficiency is reduced by high nutrient loads.



View larger version (21K):
[in this window]
[in a new window]
  		Fig. 2. Relationship between stream discharge and (a) phosphate–phosphorus uptake length, (b) ammonium-N uptake length, and (c) nitrate-N uptake length in streams located in nonpolluted areas (open circles) and in polluted streams (solid circles). Figures (a) and (b) are modified from Butturini and Sabater (1998) for nonpolluted streams, and expanded with data from this study for polluted streams and data from Davis and Minshall (1999) for nonpolluted streams. Data for polluted streams in figure (c) are from this study, and data for nonpolluted streams are from the following studies: Munn and Meyer (1990), Webster et al. (1991), Wallace et al. (1995), Marti and Sabater (1996), Valett et al. (1996), Marti et al. (1997), and Davis and Minshall (1999).

 
Ecological Implications of Nutrient Point Sources
Besides their effect on stream water quality, several studies illustrate the ecological implications of WWTP nutrient inputs on stream biota. High nutrient inputs impoverish biological diversity (Prenda and Gallardo-Mayenco, 1996), modify the structure of algal, macroinvertebrate, and fish communities, and enhance dominance by tolerant species (Ogbogu and Hassan, 1996; Miltner and Rankin, 1998; Koning and Roos, 1999). These inputs also affect stream metabolism by favoring chemoheterotropic activity of biofilms over photoautotropic activity (Masseret et al., 1998). Our study shows that stream nutrient retention efficiency can also be altered. In their "Subsidy-Stress Hypothesis," Odum et al. (1979) proposed that anthropogenic perturbations resulting in inputs of usable substances may enhance (i.e., subsidy effect) or depress (i.e., stress effect) the ecosystem response depending on the amount of substance. Within this theoretical context, our results suggest that nutrient loads entering most of the study streams far exceeded the subsidy thresholds. Under such conditions, retention efficiency was lower than expected compared with less polluted streams (i.e., nutrient loads have a stress effect), supporting our hypothesis. Therefore, the potential influence of the in-stream processes on stream nutrient concentrations is decreased. This fact reduces the in-stream potential to ameliorate water quality.

Constanza et al. (1997) ranked lakes and streams as the biome type that was third in importance (after wetlands and marshes) in the estimation of the average global value of the annual ecosystem waste treatment service (sensu Constanza et al., 1997). This ecological property is critical to the integral management of freshwater ecosystems. In fact, some management strategies to ameliorate water quality have been proposed based on the idea that this ecosystem property could complement the removal of nutrients from wastewater treatment plants (Craggs et al., 1996), assuming that it is not overwhelmed by large nutrient inputs. This study demonstrates that the efficiency of stream ecosystems to remove nutrients has limitations because it can be significantly altered by the quantity and quality of the receiving water.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
REFERENCES
 
Water inputs from WWTP increased nutrient availability in all study streams. These increases represented a depletion of water quality, according to ACA accepted standards. The effect of WWTP inputs on stream nutrient loads was magnified by the low stream discharge characteristic of streams located in semiarid areas such as the Catalonian region (NE Spain). Longitudinal patterns in nutrient concentration downstream of the WWTP inputs allowed examination of the influence of in-stream nutrient processes on nutrients in transport in streams affected by high nutrient loads. Longitudinal variation of nutrient concentrations at ambient levels are the net result of uptake and release processes occurring along the study reaches. More than one-third of the streams exhibited no clear pattern or a consistent increase of nutrient concentration along the reach, indicating that nutrients were being transported downstream without any significant removal. In the rest of the cases, nutrient concentrations consistently decline along the reach, indicating that the stream was acting as a net sink for nutrients. In these streams, uptake lengths were in the order of km. These values are at least an order of magnitude longer than those measured in streams draining relatively pristine watersheds. Together, these results suggest that stream nutrient retention efficiency decreases under high nutrient load conditions. This study complements existing knowledge about the effect of WWTP inputs on ecological attributes of stream ecosystems, in particular on nutrient dynamics. The high nutrient loads and relatively low nutrient retention efficiency measured downstream of the WWTP inputs provide evidence that the effect of these nutrient-point sources have ecological implications not only locally, but also regionally (i.e., large nutrient loads are transported over long distances affecting water quality farther downstream).


   ACKNOWLEDGMENTS
 
The authors thank N. Vidal, E. Llorens, A. Alsina, and S. Centelles for field and laboratory assistance. The Serveis Cientifico Tècnics of the Universitat de Barcelona and the Water Laboratory of the Departament EQATA (Universitat de Girona) provided their facilities and technical help in chemical analyses. We also thank J.L. Meyer and three anonymous reviewers for helpful comments on an earlier version of this manuscript. This research was supported by the funding of the Agència Catalana de l'Aigua (ACA), Departament de Medi Ambient, Generalitat de Catalunya, and the STREAMES project (U.E. EVKI-CT-2000-00081).


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 




This article has been cited by other articles:


Home page
 J. Environ. Qual.Home page
K. W. King, J. C. Balogh, K. L. Hughes, and R. D. Harmel
Nutrient Load Generated by Storm Event Runoff from a Golf Course Watershed
J. Environ. Qual., May 25, 2007; 36(4): 1021 - 1030.
[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 (21)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Marti, E.
Right arrow Articles by Sabater, F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Marti, E.
Right arrow Articles by Sabater, F.
Agricola
Right arrow Articles by Marti, E.
Right arrow Articles by Sabater, F.
Related Collections
Right arrow Surface Water Quality
Right arrow Nutrient Cycling
Right arrow Nutrient Management
Right arrow Nitrogen
Right arrow Phosphorus

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