| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Hampshire College, 895 West St., Amherst, MA 01002-5001
b Inst. of Ecosystem Studies, 65 Sharon Turnpike, P.O. Box AB, Millbrook, NY 12545
* Corresponding author (rhale01{at}gmail.com)
Received for publication April 25, 2006.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: DEA, denitrification enzyme activity • LSD, least significant difference • LWD, large woody debris
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
An important change resulting from urbanization is an increase in impervious surface cover which affects the hydrology of urban and suburban watersheds. Impervious surfaces increase surface runoff and flood peak discharges and decrease infiltration and groundwater recharge (Booth and Jackson, 1997; Paul and Meyer, 2001). The modification of streams with engineered channels and storm drainage systems compounds the effects of impervious surface cover by directing surface runoff directly into streams (Paul and Meyer, 2001; Walsh et al., 2005).
Increased flows associated with impervious surface have dramatic effects on stream channel geomorphology, widening and deepening channels and eroding stream banks (Paul and Meyer, 2001). High flows can also lead to stream scouring, reducing the presence of large woody debris (LWD) in urban streams. Finkebine et al. (2000) found that streams in areas with greater than 20% impervious surface generally have little LWD. Large woody debris is important to the formation and maintenance of organic debris dams which obstruct stream flow (Bilby, 1981). Debris dams can eventually become water-tight and form small pools upstream (Bilby and Likens, 1980; Valett et al., 2002). Because of their high organic matter content, debris dams are "hotspots" for nutrient cycling in small streams (Valett et al., 2002; Bernot and Dodds, 2005). Bernhardt et al. (2003) found that large inputs of woody debris to a stream during an ice storm significantly increased in-stream N processing by increasing microbial immobilization of N and stimulating denitrification (conversion of NO3– to N gas). Groffman et al. (2005) found that denitrification rates were higher in debris dams than in gravel bars, pools, and riffles.
In addition to affecting stream flow, urbanization leads to changes in stream chemistry. Concentrations of inorganic N are typically elevated in urban and suburban streams due to fertilizer use, sanitary sewer leaks and overflows, and septic tank discharges (Makepeace et al., 1995; Paul and Meyer, 2001; Walsh et al., 2005). In addition to N, Cl– concentrations are often very high in urban streams, especially during the winter due to deicing activities (Kaushal et al., 2005). Several studies have demonstrated that Cl– inhibits nitrification and microbial activity in soils (Hahn et al., 1942; Roseberg et al., 1986; Groffman et al., 1995), but there have been no studies on its effects in suburban streams.
In this study, we quantified the frequency, dimensions, and type of debris dams in large and small forested and suburban streams and then investigated the effects of elevated Cl– concentrations on microbial N cycle processes in debris dam material from a subset of these streams. To carry out the evaluation of microbial processes, we collected samples from dams in one forested and two suburban streams with different concentrations of Cl– and NO3– and incubated these samples in laboratory microcosms with stream water and different amendments for 30 d, measuring nitrification, net nitrogen mineralization, denitrification, and microbial respiration. Our objectives were to (1) quantify the effect of urbanization on the extent and frequency of organic debris dams in large and small streams, (2) compare rates of N cycle processes in debris dams in streams with different NO3– and Cl– concentrations, (3) evaluate the relative importance of processes that produce (mineralization, nitrification) and consume (denitrification, immobilization) inorganic N, and how this relative importance is linked to C dynamics (respiration), and (4) determine the effects of NO3– and Cl– additions on N cycle processes in streams with different histories of NO3– and Cl– exposure. The work was conducted in streams that are monitored by the National Science Foundation-funded urban long-term ecological research (LTER) project, the Baltimore Ecosystem Study (BES), and builds on previous BES water quality (Groffman et al., 2004; Kaushal et al., 2005) and in-stream nitrogen processing research (Groffman et al., 2005).
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
20% impervious surface) with high concentrations of both Cl– and NO3– (Table 1). Dead Run is a more urbanized watershed (40% impervious surface).
Debris Dam Survey
Of the five streams surveyed, three were small (1.2- to 1.7-m width) and two were large (5.4- to 7.8-m width). The three small streams included a forested reference (Pond Branch) and two suburban streams (Glyndon and Dead Run tributary at Westfield). The two small suburban streams differed in that one (Dead Run tributary at Westfield) exhibited the channel degradation and incision common to suburban streams, whereas the other (Glyndon) did not. The two large streams included forested reference (Baisman Run) and suburban streams (Dead Run) (Table 1).
Stream segments of 100 to 200 m were surveyed for debris dams in December 2004. Debris dams were defined as accumulated fresh and decomposed organic matter in the stream channel after Groffman et al. (2005). Current stream width and bank full width were measured at each debris dam and 1 m above and below the dam. The supporting structure of the debris dam was noted (i.e., cobbles, LWD, other, or an absence of supporting structure). In the case of LWD, the length and diameter of each piece was measured, as well as whether or not it spanned the stream channel and its position (i.e., perpendicular, at angle, or parallel to stream flow).
The dimensions of each debris dam were measured and used to calculate volume. The quality of organic matter was noted (i.e., coarse leaves and twigs or well decomposed debris) along with the presence or absence of any urban debris. Debris dam height both in and above the water was measured. Pool formation was noted and its mean depth recorded. Frequency of debris dams was calculated per 100 m.
Laboratory Microcosm Studies
We collected samples from organic debris dams in Pond Branch, Baisman Run, and Glyndon on 3 June 2004. Three distinct dams, separated by at least 10 m were sampled in each stream. Samples were taken from submerged sections of the debris dams and consisted of a mixture of leaves and trapped sediments. Large sticks and insects were removed from samples which were then homogenized in a blender with small quantities of ambient stream water. Samples were homogenized to provide a uniform material for assessment of potential microbial activity in materials that had been exposed to different stream chemistries and for evaluation of the effects of amendments on this activity.
Samples were stored at 4°C between sampling and analysis. The moisture content of the debris slurries was determined by drying at 60°C for 24 h (McInnes et al., 1994) and organic matter content by loss on ignition at 450°C for 4 h (Nelson and Sommers, 1996). Ambient stream Cl– and NO3– concentrations were determined by ion chromatography.
Laboratory microcosms consisted of 150 g of debris dam material incubated in sealed mason jars with 40 mL of stream water. Microcosm lids were fitted with septa to allow for gas sampling to determine CO2 production. Samples from each of the three replicate debris dams from each stream were incubated with native, ambient water as well as with native water with different amendments. Samples from Pond Branch (forested reference stream) were incubated with native stream water amended with 2 mg NO3– L–1 and/or 80 mg Cl– L–1 to determine the effects of NO3– and Cl– concentrations common in suburban streams on N cycle processes in a forested stream. Pond Branch samples were also incubated with 2500 mg Cl– L–1 to assess the effects of the high Cl– concentrations found during winter in urban study streams. Pond Branch samples were also incubated with stream water from Glyndon (suburban stream) to separate the effects of NO3– and Cl– from any other characteristics of this water. Samples from Baisman Run were incubated with native water, with native water amended with 80 or 2500 mg Cl– L–1, or with Glyndon water. Samples from Glyndon were incubated with native water or with native water amended with 2500 mg Cl– L–1.
Microcosms were incubated at room temperature (22°C) for 30 d, were not shaken, and therefore likely became anaerobic between sampling events. Headspace samples were taken by syringe at 0, 5, 10, 20, and 30 d. After gas sampling, jars were opened and samples of the debris slurry (sediment plus water) were removed from the jar for inorganic N (NO3– and NH4+) extraction with 2 M KCl (5 g) and DEA analysis (20 g, on Days 0, 15, and 30).
Gas samples were analyzed for CO2 by thermal conductivity gas chromatography, and microbial respiration was quantified as the amount of CO2 evolved over each incubation period. Inorganic N (NH4+ and NO3–) was measured by colorimetric analysis with a Perstorp 3000 flow injection analyzer. Potential net N mineralization was calculated as the amount of NO3– + NH4+ accumulated over the 30 d incubation and net nitrification was calculated from the accumulation of NO3– alone. Rates are expressed on a per g of dry sediment basis, accounting for soil removed for analysis over the course of the incubation.
Denitrification enzyme activity was measured using the method described by Groffman et al. (1999). Sediment samples were amended with NO3– (100 mg N kg–1), dextrose (40 mg C kg–1), chloramphenicol (10 mg kg–1), and acetylene (10% headspace), and incubated (shaken) under anaerobic conditions for 90 min in 120-mL Erlenmeyer flasks. Gas samples were taken at 30 and 90 min and analyzed for N2O by electron capture gas chromatography.
Statistical Analysis
For the debris dam survey, nonnormal data were log transformed for all statistical analysis. Differences among streams and stream geomorphic structures were evaluated using one-way analysis of variance with Tukey's Studentized Range (HSD) Test used to determine differences between specific sites. For the laboratory study, sites and treatments were compared with a repeated measures analysis of variance, with least significant difference (LSD) multiple comparison tests used to determine differences between specific sites and treatments. Relationships between variables were explored with Pearson Product Moment correlations. The Statistical Analysis System (SAS Institute, 1988, Release 6.03) was used for all analyses.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
|
|
|
Microbial respiration showed a complex response to amendments (Table 3). In the forested reference stream, additions of either 2 mg N L–1 NO3– plus 80 mg L–1 Cl– or 2500 mg L–1 Cl– inhibited (p < 0.10) respiration relative to the control. However, additions of NO3– alone (2 mg N L–1), Cl– alone (80 mg L–1), or Glyndon water had no effect on respiration.
In the less developed suburban stream (Baisman Run) microbial respiration was stimulated (p < 0.10) by addition of 80 and reduced by addition of 2500 mg Cl– L–1. Addition of 2500 mg Cl– L–1 to debris dam material from the more developed suburban (Glyndon) stream, which had a history of exposure to high levels of Cl–, had no effect on respiration.
In many cases, potential net N mineralization was negative, indicating a loss of inorganic N over the course of the 30-d incubation to immobilization or denitrification (Fig. 2, bottom, Table 3). Levels of NO3– were very low in all materials making it impossible to calculate rates of net nitrification. In material from the forested reference stream net mineralization was negative in control incubations, but was positive in all amended samples, indicating that the amendments either increased the production or reduced the consumption of inorganic N. Net mineralization was highest (p < 0.10) in samples amended with 2500 mg Cl– L–1 and lowest (p < 0.10) in the control (Table 3). Amendments had no effects on net mineralization in material from the streams with a history of exposure to NO3– and Cl–.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Debris Dam Occurrence and Characteristics
The frequency of debris dams in our forested streams was comparable to densities reported in previous studies (Bilby and Likens, 1980; Potts and Anderson, 1990; Kraft et al., 2002). Our densities of 11.5 to 18 dams per 100 m for small streams are similar to those (14 per 100 m) reported by Bilby and Likens (1980) for the Hubbard Brook watershed in New Hampshire and by Kraft et al. (2002) (0 to 14 dams per 100 m) for first-order streams in the northeastern USA. Our densities of 1.25 to 3.5 dams per 100 m for large streams are comparable to those reported by Kraft et al. (2002) (1.3 to 4.6 dams per 100 m) for third-order streams in the northeastern USA. Debris dam density is naturally highly variable both within and between watersheds and landscapes (Gregory et al., 1993).
The effects of urbanization on the nature and extent of debris dams was more complex than expected. We expected that high flows associated with impervious surfaces would greatly reduce the development and maintenance of debris dams in suburban streams. However, our results show that debris dams can be abundant in these streams. The small suburban stream (Glyndon) that showed little evidence of the incision common to urban streams was similar to the small forested stream in many measures, including volume of woody debris and debris dams and pool formation. While we cannot definitely determine what prevents stream incision in this watershed, factors such as culvert size and storm water controls appear to be important in allowing the accumulation and maintenance of debris dams in this stream and should be considered in future studies.
The results from the Glyndon stream suggest that there is potential to design and manage suburban streams to facilitate the accumulation of organic debris. Hilderbrand et al. (1997) reported that adding LWD both systematically and randomly to urban streams increased pool formation but did not significantly affect total benthic macroinvertebrate abundances. Larson et al. (2001) evaluated the effectiveness of LWD additions in six urban in-stream rehabilitation projects by measuring LWD and pool frequency and benthic macroinvertebrate abundances. They also found that LWD additions increased pool frequency; however, pool spacing was lower in rehabilitated urban streams than in forested streams with the same quantity of LWD. Pool formation is of critical importance as it affects retention time and hyporheic dynamics of water and N. These studies suggest that restoration of functionally significant organic debris dams in urban streams will be complex and will likely require consideration of watershed flow regimes as well as organic debris supply. The Glyndon stream illustrates this connection between hydrological regimes and debris dam occurrence and demonstrates that it is possible to manage urban streams to maintain a level of organic matter comparable to forested streams of similar size.
Microbial Processes
As found in previous studies (Groffman et al., 2005), debris dams in suburban streams had higher denitrification potential than those in forested streams. The increase in denitrification potential in debris dam samples from the forested reference stream that were amended with 2 mg NO3––N L–1 confirms that this "urban effect" on denitrification is due to the high NO3– concentrations in urban and suburban streams. Many studies have found that denitrification in organic-rich, anaerobic sites such as organic debris dams is NO3––limited and therefore increases with increasing NO3– concentrations (Holmes et al., 1996; Martin et al., 2001; Kemp and Dodds, 2002; Kellman, 2004; Richardson et al., 2004; Wall et al., 2005; Tuerk and Aelion, 2005). These results suggest that high denitrification potential in debris dams rich in organic matter may play an important role in maintaining water quality by reducing NO3– loading to receiving waters.
Our results suggest that the maintenance of debris dams could be an important strategy for increasing N retention in urban watersheds. Ultimately, the importance of debris dams in controlling stream NO3– concentrations depends on the amount of water that passes through debris dams with sufficient residence time to allow for denitrification (Bernot and Dodds, 2005). Clearly the presence of debris dams in some of our suburban streams is not sufficient to consume all anthropogenic NO3– in these streams. However, without the heterogeneity in flow and C and N dynamics caused by the presence of these dams, NO3– concentrations in these streams would likely be higher (Fisher et al., 2004; Gücker and Boëchat, 2004; Kellman, 2004). Whole-stream manipulations of debris dams are needed to truly quantify their importance as regulators of stream NO3– levels.
Increases in stream Cl– concentrations are a common effect of urbanization (Herlihy et al., 1998). In our streams, Cl– concentrations ranged from <10 mg L–1 in our forested reference stream to over 5000 mg L–1 in our most developed suburban stream during winter (Kaushal et al., 2005). Our data here suggest that increased Cl– has significant and complex effects on N processing in these streams.
Additions of Cl– clearly had significant effects on N processing in debris dam material from the forested reference stream. Denitrification potential (DEA) was decreased by high Cl– concentrations (2500 mg L–1), and lower concentrations (80 mg L–1) inhibited increases in DEA in response to NO3– inputs. Net N mineralization in debris dam material from the forested reference stream was stimulated by Cl– additions due either to stimulation of mineralization or inhibition of coupled nitrification/denitrification (Hahn et al., 1942; Roseberg et al., 1986; Groffman et al., 1995). These additions changed the sign of net mineralization from negative (consumption of inorganic N) to positive (net production of inorganic N), suggesting that Cl– additions have the potential to convert debris dams from sinks to sources of inorganic N in streams.
In contrast to results from the forested reference stream, data from our suburban sites, which have a history of exposure to high Cl– levels, suggest that microbial communities have some potential to adapt to chemical changes induced by urbanization. The suburban streams had inherently higher levels of DEA than the forested reference stream, and additions of Cl– to debris dam materials from these streams had no effect on DEA or net N mineralization over our 30 d laboratory incubation. Respiration was also unaffected by Cl– additions in the stream with the highest previous exposure to Cl– (Glyndon). Respiration in the stream affected by moderate Cl– exposure (Baisman Run) was more complex, with stimulation by addition of 80 mg Cl– L–1 and inhibition by addition of 2500 mg Cl– L–1. These results suggest that the response of microbial communities in stream debris dams to stream chemistry are complex and functionally significant and worthy of further study.
It is impossible to tell from our data if high denitrification in the presence of high Cl– in our suburban streams was fostered by adaptation of microbial communities to high Cl– levels (Elshahed et al., 2004), by the development of new microbial communities (Zumft, 1999), or by altered coupling between C and N cycle processes. For example, levels of microbial respiration were significantly lower in the suburban streams than in the forested stream, likely due to the lower organic content of the debris dam material in these streams. Carbon status and microbial growth rate strongly influence microbial immobilization and mineralization of N (Bengtson and Bengtsson, 2005) and the high denitrification in the presence of high Cl– levels in our suburban streams could be a product of lower growth rates and/or immobilization activity. Interactions between C status and Cl– effects could be particularly important in fall and winter in streams where litterfall is a dominant C source, or during summer in streams where floating mats of algae and macrophytes are important C sources (Schaller et al., 2004). The high NO3– levels in these streams could also facilitate high denitrification in the presence of high Cl– as NO3– influences redox status, which strongly influences the balance of processes that produce and consume NO3– (Pett-Ridge and Firestone, 2005). There is a clear need for more detailed analysis of the effects of changes in stream chemistry induced by urbanization on microbial communities and processes.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  E. S. Bernhardt, L. E. Band, C. J. Walsh, and P. E. Berke Understanding, Managing, and Minimizing Urban Impacts on Surface Water Nitrogen Loading Ann. N.Y. Acad. Sci., June 1, 2008; 1134(1): 61 - 96. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M.L. Cadenasso, S.T.A. Pickett, P.M. Groffman, L.E. Band, G.S. Brush, M.F. Galvin, J.M. Grove, G. Hagar, V. Marshall, B.P. McGrath, et al. Exchanges across Land-Water-Scape Boundaries in Urban Systems: Strategies for Reducing Nitrate Pollution Ann. N.Y. Acad. Sci., June 1, 2008; 1134(1): 213 - 232. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. H. Duff, A. J. Tesoriero, W. B. Richardson, E. A. Strauss, and M. D. Munn Whole-Stream Response to Nitrate Loading in Three Streams Draining Agricultural Landscapes J. Environ. Qual., May 1, 2008; 37(3): 1133 - 1144. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
