| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Physiology, Wayne State University, Detroit, MI
Corresponding author (jeffram{at}med.wayne.edu)
Received for publication October 15, 1999.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
Abbreviations: cfu, colony forming units • MPN, most probable number
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
The present study investigates the use of the zebra mussel as a bacterial uptake mechanism to provide an indicator of prior exposure to bacteria in a freshwater environment. One of the problems with traditional water monitoring methods is the necessity of collecting water samples precisely at the time that bacteria are actually present. In a flowing system, such as a river or stream, a short-term pulse of bacteria flowing past a point may be missed completely if the water sample is not taken at the appropriate time. The present study evaluated whether zebra mussels are able to accumulate and retain E. coli when exposed to high concentrations of bacteria. Retention of bacteria would enable the mussels to be used as an indicator of bacterial contamination even after bacterial concentrations in the water have returned to relatively low concentrations.
The zebra mussel, a bivalve mollusk indigenous to Russia and Europe, was inadvertantly introduced into the Great Lakes region of North America in the mid-1980s (Hebert et al., 1989). It spread rapidly from its original site of introduction throughout much of the eastern United States (National Aquatic Nuisance Species Clearinghouse, 1999; Ram et al., 1992) largely due to its high reproductive capacity and its ability to live in densities as high as 700000 animals per m2 (Kovalak et al., 1993). As a highly efficient filter feeder, the zebra mussel draws water into its inlet siphon, passing it over and through the gills where particle capture is mediated by gill cilia and mucus (Morton, 1969a, b; Silverman et al., 1995; Sprung and Rose, 1988; Ten Winkel and Davids, 1982). Many of the bacteria, algae, and zooplankton captured by this filtration, sorting, and selection process (Lowe et al., 1990; MacIsaac and Sprules, 1990) may still be viable even after passing completely through the gut (MacIsaac and Sprules, 1990; Nichols et al., 1996), indicating that digestion of captured organisms is often incomplete in the zebra mussel. Since bivalves do not normally contain fecal coliform bacteria in their gut, the presence of such bacteria in a zebra mussel is hypothesized to indicate a recent environmental exposure of the mussel to bacteria rather than to reflect their endogenous presence in the mussel. This hypothesis was examined in the present study in both laboratory experiments and field observations, and the temporal parameters of bacterial uptake and depuration by zebra mussels were measured.
The field setting of this study was the Clinton River watershed (Fig. 1) . This watershed is located in southeast Michigan and drains approximately 2000 km2, including portions of Lapeer, Macomb, Oakland, and St. Clair Counties. The basin is highly urbanized with a population of approximately 1.3 million people (Clinton River Watershed Council, personal communication, 1999). The Clinton River receives frequent exposures to fecal bacteria (Ditschman et al., 1995). A new, potentially more efficient method of monitoring the occurrence of high bacterial concentrations was examined in this study.
|
|
|
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
Enumeration Procedures
Two methods for enumerating bacteria in water samples and zebra mussel homogenates were used in these studies: (i) a plate count method and (ii) a commercially available multiwell most probable number (MPN) test (Colilert, IDEXX Laboratories, Westbrook, ME). The plate count method was used in laboratory experiments and preliminary field studies. The MPN method was used for most field studies. Both methods enumerate growth by colony forming units and use specific chromogenic substrates for generally identifying coliform bacteria and specifically identifying E. coli. Development of color in the chromogenic substances is based on the specific presence of glucuronidase in E. coli and galactosidase in common coliform bacteria (Alvarez, 1984; Brenner et al., 1993; Edberg and Edberg, 1988; Kilian and Bulow, 1976; Kilian and Bulow, 1979; Rippey et al., 1987).
The plate count method was modified from procedures for assessing the sanitary quality of shellfish (American Public Health Association, 1984a, b). A sample or a series of dilutions were inoculated into Coliscan medium and poured into a petri dish containing agar. After incubation at 32°C for 24 and 48 h, the number of purple (E. coli) and pink (coliforms) colonies were counted, and their concentrations were calculated.
In the MPN method, a known volume and dilution of the sample to be enumerated was added to 100 mL of the Colilert medium, mixed well, and poured into an IDEXX Quanti-Tray/2000 tray (IDEXX Laboratories, Westbrook, ME). The tray was sealed with the IDEXX Quanti-Tray Sealer, forming 50 large wells and 48 small wells. The entire tray was then incubated for 20 h. Colony forming units were determined by counting the number of yellow compartments in room light and fluorescent compartments with UV illumination. The manufacturer's calibration chart translates this test result into the MPN of coliform and E. coli cfu per 100 mL of starting material.
Bacteria were enumerated in groups of zebra mussels that were removed from experimental tanks or field locations. Mussels to be enumerated were placed in 18°C water for 15 min. Animals that opened during this period were thereby verified as being alive and also could purge bacteria that were in the water of their branchial chambers at the time of collection. Ten animals in each experimental group were shucked, and their soft tissues were weighed. A 10:1 dilution (10 mL medium g-1 of mussel tissue) was made with peptone buffer (Sigma–Aldrich [St. Louis, MO] Peptone Hy-Soy T) and homogenized for 60 s. The homogenate was added to Coliscan medium (Micrology Laboratories, Goshen, IN), shaken, and a portion was assayed for cfu using the plate method. Similarly, a series of duplicates and two to four 10-fold dilutions were performed.
For some laboratory experiments, following the 15 min purge step, mussels were homogenized without shucking. Eight to ten mussels were weighed, shells were cleaned of any adhering debris, and then the mussels were macerated whole in 50 mL of sterile water in a Waring (New Hartford, CT) blender fitted with a stainless steel attachment. A portion of the homogenate (usually 1 mL) was assayed by the plate method or Colilert tray method. Most field data were also obtained using mussels homogenized without shucking, according to these methods.
Zebra mussels used in these experiments ranged considerably in size. Shell lengths varied from 15 to 27 mm. The weight per mussel ranged from 0.4 g to 1.5 g. In order to normalize for any differences in numbers of animals and their sizes, results have been expressed per gram of mussel, taking into account dilutions of the homogenates used in the assays. For comparison, E. coli concentrations in water samples have been expressed as cfu mL-1 of water (i.e., cfu g-1 of water).
Laboratory Experiments
Laboratory experiments were performed to determine the uptake and depuration kinetics of E. coli by zebra mussels and to compare results obtained for shucked mussels with those homogenized whole. In the first uptake kinetics experiment, 150 mussels were placed in a plastic cage in an aquarium containing 38 L of river water known to have a near-zero concentration of E. coli. After acclimating mussels to river water for approximately 1 h, 2 L of raw influent water from the Mt. Clemens sewage treatment plant was added to the aquarium, producing a 20:1 dilution of sewage in river water. Mussels were exposed to this solution for up to 96 h. Water samples and groups of mussels were removed for enumeration at various intervals before and during the sewage treatment.
Two subsequent experiments provided data on both uptake and depuration kinetics. Mussels (238 in one experiment, 538 in the other) were acclimated as above and exposed to a 10:1 solution of river water and sewage. After 12 h of exposure to the E. coli source (uptake phase) the mussels were placed in sterile aquarium water for 80 h (depuration phase). In each experiment, groups of mussels were removed for enumeration at various intervals during the uptake and depuration phases. In the experiment with 538 mussels, enumerations were performed on two groups of mussels. The first group was shucked and enumeration was performed on soft tissues; for the second group, mussels were homogenized whole, with shell.
Field Observations
Zebra mussels were placed in plastic cages (16 x 13 x 12 cm) anchored at strategic sites in the Clinton River watershed (Fig. 1). Permission to place mussels in the Clinton River watershed, which was already colonized by zebra mussels, was obtained from the Michigan Department of Natural Resources. The plastic cages were attached to a nylon rope, anchored a short distance above the bottom of the river by a brick, and tied to an overhanging rope or structure (e.g., bridge). Altogether, mussels have been placed for various periods of time in 16 different sites in the Clinton River and its tributaries (Fig. 1); however, for the present report, which focuses on the utility of the method rather than a complete analysis of the watershed, data obtained from only a few of the sites are reported.
Each cage was initially loaded with approximately 150 to 300 mussels. For sampling, cages were lifted from the water and 15 mussels were removed at each site. Mussels were placed in a plastic bottle in a cooler, transported to the laboratory, and then processed for enumeration of E. coli as described above.
Rain data, reported in several figure captions, were obtained from rain gauges maintained by local cities and the Selfridge Air National Guard base. In each case, rainfall was reported from gauges located within about 4.8 km (3 miles) of the relevant zebra mussel site (see Fig. 1).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
|
|
Depuration
High E. coli concentrations taken up by zebra mussels previously exposed to sewage took several days to decrease to preexposure concentrations. In the experiment illustrated in Fig. 3A, E. coli concentrations in the water were reduced to zero by replacement of E. coli-laden water with sterile water 12 h after the beginning of the experiment. However, the concentration of bacteria in the zebra mussels at 30 h was still approximately 30% of its maximum. Only after 80 h did tissue concentrations of E. coli return to preexposure concentrations. In the second depuration experiment (Fig. 3B), the decrease in E. coli concentration was more rapid for both shucked and whole mussels, but was still approximately 250 cfu g-1 (approximately 15% of the maximum) at 21 h. Significant concentrations of bacteria were still present in mussels at 33 h, and near zero concentrations were observed at 57 h.
Retention of E. coli by zebra mussels was also evident in the first accumulation experiment (Fig. 2). Escherichia coli concentration decreased much more slowly in the mussels than in the surrounding water. Even when E. coli concentrations in water were near zero at 51 h, the mussel concentrations were still approximately 50% of the maximum. Similarly, field data described below reveal many instances in which peaks in E. coli concentrations in mussels took two or more days to return to preexposure concentrations when water E. coli concentrations had returned to low concentrations earlier during the same period.
Escherichia coli Concentrations in Zebra Mussels under Field Conditions
Response to a Brief Pulse of Escherichia coli
The response of zebra mussels in the field to a brief pulse of E. coli is illustrated in Fig. 4A
. Mussels were collected daily from cages placed upstream (Site 12) and downstream (Site 13) of a suspected source of E. coli, the main branch of the Clinton River. Rain occurred during the 24 h preceding the third day of sampling, and E. coli concentrations in the water at the downstream site increased to approximately 180 cfu mL-1. This increase in E. coli concentrations after rain storms is a common observation for the main branch of the Clinton River. The next day, E. coli concentrations in the water returned to a lower level (<60 cfu mL-1), similar to concentrations prior to the rain event. Escherichia coli concentrations remained low for the remainder of the period illustrated. For comparison, Fig. 4B also shows E. coli concentrations measured in the north branch of the Clinton River, upstream from its confluence with the main branch. Escherichia coli concentrations in the north branch remained low throughout this period. Thus, this storm event produced a brief exposure of mussels to E. coli in the water at the downstream site and little change in the E. coli exposure of mussels in the north branch.
Escherichia coli concentrations in the mussels increased in response to high concentrations of E. coli in the water and remained elevated longer than the concentrations in the water. The 1-d pulse of high E. coli concentration in the water at the Clinton River downstream site was accompanied by a rise in E. coli concentrations in the mussels above prior basal concentrations for several days (Fig. 4A). Escherichia coli concentration in the mussels peaked at 2300 cfu g-1 on the first day of the pulse, fell by approximately half during the next 2 d, and attained the prepulse concentration on the third day after exposure. In contrast, the mussels in the north branch were not exposed to high E. coli concentrations and exhibited little elevation of their E. coli content (Fig. 4B). The uptake and depuration responses of the mussels in the downstream location to a pulse of E. coli are similar to responses of mussels in the laboratory.
Responses to Low Concentrations of Escherichia coli
The inherent capability of zebra mussels to concentrate E. coli suggests that they may be used to detect increases in E. coli concentration that might otherwise be missed by conventional monitoring. For example, in Fig. 4A, the small increase in mussel E. coli concentrations on 7 September could have been due to an undetected change in water E. coli concentrations, and the similar concentration on the first day of sampling could have been due to a small rain event that had occurred the day before sampling began. Data from another location (Fig. 5)
illustrate several such small peaks in mussel E. coli concentrations that occurred in the presence of low concentrations of E. coli in surrounding waters.
|
Responses to High Concentrations of Escherichia coli
Field work in Bear Creek and Red Run Drain illustrates some potential applications as well as some problems encountered in using mussels to monitor bacterial pollution concentrations. In our partial survey of the Clinton River watershed, the most consistently high concentrations of E. coli were found in Bear Creek. The median concentration of E. coli in Bear Creek water (Site 7, sampled for 68 d during the summer of 1999) was 96 cfu mL-1 and was greater than 60 cfu mL-1 for 44 d (65% of the 68 d). A subset of these data is illustrated in Fig. 6
. In comparison, E. coli concentrations in the water of Red Run Drain (sampled for 28 d during summer 1999) at the Denton, Red Run, and Metropolitan locations (Sites 3, 8, and 9), exceeded 60 cfu mL-1 only 21, 36, and 29% of the time, respectively. Median E. coli concentrations in the water at Denton, Red Run, and Metropolitan were 17, 36, and 18 cfu mL-1, respectively.
Consistent with the larger concentrations of E. coli in the water in Bear Creek, mussels at this location also showed larger concentrations of E. coli than did mussels in other locations. The median concentrations of E. coli in mussels at the Bear Creek, Denton, Red Run, and Metropolitan monitoring sites were 718, 176, 213, and 185 cfu g-1, respectively. As expected, peaks in water concentrations of E. coli in Bear Creek often coincided with or were followed by rises in mussel concentrations of E. coli. However, we also observed peaks in mussel concentrations of E. coli that did not have any obvious "signature" in the water record. One such event is illustrated in Fig. 6A, during the period of 27 to 29 May.
Data collected at Red Run and Denton, downstream and upstream, respectively, from the confluence of Bear Creek and Red Run Drain, illustrate some relationships at sites of confluence and also some problems encountered in these studies. First, Red Run E. coli concentrations tended to rise whenever Bear Creek E. coli concentrations increased. The relationship was not precise, and the small magnitude of some of the increases in Red Run after large peaks in Bear Creek suggests that Bear Creek contributes only a small proportion of the E. coli in Red Run Drain. Nevertheless, the high bacterial loads in Bear Creek (Site 7) may partially account for the higher median concentrations of bacteria at the Red Run site (Site 8) than at Denton (Site 3). Second, the peak concentrations of E. coli observed in mussels at Red Run were relatively low (always less than 1000 cfu g-1) compared with maximum concentrations seen elsewhere. For example, the relatively high E. coli concentrations in the water at Red Run on 22 and 31 May were followed by only modest increases in E. coli concentrations in the mussels. Third, sometimes the concentration of E. coli measured in mussels was elevated for only a single day and returned the next day to baseline (e.g., 28 May at Denton). This brief peak in E. coli concentration in mussels at Denton seems to be reflected in a much smaller increase downstream at Red Run. Possible causes of such brief "events" will be discussed below. Fourth, the gap from 24 to 26 May in the data from Denton was due to vandalism. At Denton, cages were removed from the water, cages or bricks were broken, or knots were tied in anchoring ropes a number of times. Thus, during the illustrated period, cages vandalized at Denton prior to collection time on 24 May had to be restocked with mussels and allowed to acclimate for several days.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
Although the zebra mussel is most efficient at filtering particles ranging in size from 15 to 40 µm in diameter (Ten Winkel and Davids, 1982), they have also been observed to filter particles as small as 0.7 to 1.0 µm (Silverman et al., 1996; Sprung and Rose, 1988). The size of a typical E. coli is approximately 2 µm long (Brock and Madigan, 1991). While this size is near the lower end of the mussel's filtering ability, these bacteria are often found adsorbed onto particles, such as algae and detritus, which are of the optimum size range for zebra mussel consumption. Lei et al. (1996) demonstrated nearly 100% removal of the 2-µm size class, while Silverman et al. (1995) showed excellent clearance of E. coli. The present study confirmed, characterized, and utilized this capability of zebra mussels at filtering and concentrating E. coli.
Kinetics
An initial uptake rate can be calculated from the initial few hours of exposure to E. coli when purging and digestion is comparatively small. In Fig. 2 and 3, uptake for the first 2 h averaged 950 cfu g-1 h-1. The mussels accumulated viable bacteria up to a concentration of 2000 to 2500 cfu g-1 in the presence of approximately 900 cfu mL-1 in the ambient water.
Depuration kinetics were determined using the three data sets in Fig. 3. Regression analysis of mussel E. coli concentrations vs. time during the depuration phase of each data set according to first order kinetics gave rate constants ranging from -0.043 to -0.062 h-1, with a mean value of -0.053 h-1 and an average R2 of 0.93. Thus, the decrease in E. coli concentration during depuration can be represented by C
= C0 exp
, where C(t) is the E. coli concentration (cfu g-1) in the mussel at time t (h), C0 is the E. coli concentration in mussels at the beginning of the depuration period, and -0.053 h-1 is the average first order rate constant. According to this model, E. coli concentration decreases to 28 and 8% of the peak concentration 24 h and 48 h, respectively, after a bacterial exposure.
Field data on the uptake and depuration of bacteria by zebra mussels is consistent with the kinetics measured in the laboratory. The mussels in the field appear to have concentrated the bacteria relative to the E. coli concentration in the water even more than was seen in the laboratory. However, measurements of E. coli in the water may not have corresponded to the time of maximum bacterial concentrations. Peak E. coli concentrations in mussels of 2000 to 3000 cfu g-1, attained in the illustrated field observations, do not represent a physiological limit to the bacterial concentrations that can occur in zebra mussels, since other studies in the upper reaches of Bear Creek have exhibited peak concentrations of 3000 to 5000 cfu g-1 (data not shown).
Depuration of mussels in the field in the presence of low concentrations of E. coli in the water is shown here following distinct peaks for one event at the Clinton River (Downstream, Fig. 4A), for four events at Spillway (Fig. 5), and one event at Denton (Fig. 6). The event at Denton is unusual and is discussed below. For the other five events, the mussels retained bacteria for at least 1 d past the peak. On average, E. coli concentration the day after a peak decreased to 37 ± 16% (mean ± sd) of the peak value, which is similar to depuration kinetics observed in the laboratory.
Detection of Events Missed by Standard Water Monitoring Methods
Zebra mussels helped detect increases in E. coli concentrations that might have been missed by standard water monitoring methods and schedules. If water monitoring is done infrequently, brief increases in bacteria in water may not be detected but could be retained in mussels. In Fig. 4A, if water had been sampled on 4 or 5 September at the downstream Clinton River site (Site 13), nothing remarkable would have been noted about E. coli concentration. However, if mussels at the downstream site had been sampled on 4 or 5 September, it would be evident that a rise in bacterial concentrations had occurred recently. Using zebra mussels for monitoring can thereby increase the number of days that provide information about increased bacterial concentrations. Thus, sampling mussels every second or third day in a monitoring program could provide information about the occurrence of bacterial "pulses" that would otherwise have required daily water sampling methods to detect.
A second way in which zebra mussels help detect E. coli that standard monitoring techniques miss is that the mussels concentrate the bacteria in the water. For example, at Spillway (Fig. 5), several peaks in zebra mussel bacteria occurred in the presence of low concentrations of E. coli. Three of the four peaks were associated with the occurrence of rain, which might have been expected to increase E. coli concentrations in the water. The fact that the mussels can concentrate the bacteria makes detection of such events possible.
Third, zebra mussels detect increases in E. coli that are not specifically associated with rainfall. At Spillway, the final peak was not accompanied by rainfall. Similarly, the large peak in the concentration of E. coli in zebra mussels in Bear Creek on 27 to 29 May occurred without accompanying rainfall or an evident peak in E. coli concentrations in the water. Brief pulses of E. coli in the water at times other than the sampling times could be the source of the E. coli in the zebra mussels. Such rainfall-independent events might have been caused by a sporadic anthropogenic bacterial source upstream from these sites.
Tracking Bacterial Sources
A potentially important application of zebra mussels for watershed monitoring is to locate sources of bacteria. Mussels placed downstream from a bacterial source would be expected to exhibit larger concentrations of bacteria than mussels placed upstream from the source. The mussels placed in the Clinton River at downstream and north branch locations illustrate this principle and suggest the presence of a bacterial source in the main branch of the Clinton River that is not present in the north branch. In addition, measurements of bacteria in zebra mussels in Bear Creek provide direct evidence of high concentrations of E. coli from this suspected source. High E. coli concentrations in mussels at a downstream location could be used to indicate when upstream mussels should be sampled, as the upstream mussels would still retain bacteria the next day. If the upstream mussels did not show evidence of a peak in E. coli concentration, this would indicate that the source of the bacteria was between the upstream and downstream locations.
Problems in the Use of Zebra Mussels
Zebra mussels can be useful in detecting E. coli in a watershed. However, some practical problems include: (i) anchoring cages for mussels, (ii) collecting zebra mussels, (iii) limiting the method to waters already colonized by zebra mussels, and (iv) vandalism. Problems in data interpretation include: (i) occasional large fluctuations, (ii) variability in environmental bacterial sources, (iii) decreases due to dilution and natural processes of water purification, and (iv) variability in uptake of bacteria.
The problem of finding places to anchor mussels was encountered at the confluence of the main branch and north branch of the Clinton River. It would have been useful to corroborate the suspected bacterial source in the main branch of the Clinton River; however, there was no convenient anchoring site upstream from the confluence. The downstream and north branch sites were relatively easy to monitor as bridges from which mussel cages could be hung cross the rivers at those locations. At another site (Red Run, Site 8), it was necessary to construct a pulley and rope system in order to anchor and retrieve mussels.
Due to the accessibility of waters of Lake St. Clair that are already highly infested with zebra mussels, it was relatively easy to collect thousands of mussels for this study. Study sites further from such highly infested sites would encounter higher costs to collect and transport the mussels. Unless the watershed to be studied is already in a zebra mussel–infested region, zebra mussels should not be used for E. coli monitoring because of their potential for biofouling and disruption of established ecosystems. Moreover, interbasin transfer of mussels between already infested areas should be discouraged due to the possibility of transferring disease pathogens with the mussels.
One problem of data interpretation is that, on occasion, the bacterial concentrations in zebra mussels rose sharply for only a single day, as occurred at Denton on 28 May. The abrupt decrease to basal concentrations on 29 May is much faster than would be expected from laboratory measurements of depuration. The cause of such rare large fluctuations is unknown. Perhaps one or more of the zebra mussels sampled on 28 May may have been anomalously infected with E. coli. Replication of observations is therefore important when significant increases occur.
Another consideration in interpreting data is that E. coli sources change. Although the data for the Clinton River in Fig. 4 for the period of 1 to 8 September clearly indicate that the north branch is not a source of downstream bacteria for this particular event, a subsequent event not shown here exhibited bacteria also in the north branch. Variability in apparent sources also occurred in our sampling in Red Run and Bear Creek. Monitoring of many events under a variety of environmental conditions is necessary to determine the predominant sources and conditions that may contribute to increases in bacteria in a watershed.
The presence of a highly contaminated source upstream from a monitoring site might be expected to produce high concentrations of bacteria downstream, at the monitoring site. However, bacterial concentrations at the monitoring site may be lower than the source due to dilution effects and natural purification processes. Dilution of the highly contaminated Bear Creek waters in the larger volume Red Run probably accounts for the relatively small increases at Red Run (Site 8) compared with Bear Creek (Site 7). Lower nonpeak concentrations of E. coli at Metropolitan (Site 9) than at Red Run (Site 8) may be due to dilution by tributaries and drains that feed into Red Run Drain between the two monitoring sites but may also be due to natural purification processes that remove viable bacteria as water flows downstream (Kunze and Stober, 1981). The fact that peak concentrations at Metropolitan (Site 9) seem to be higher than Red Run (Site 8) when rainfall has occurred could mean that these same tributaries and drains are sources of bacteria during rainfall.
Another problem in interpreting the concentrations of E. coli observed in mussels is that the mussels at different sites may vary in their efficiency at taking up bacteria. Mussels may not filter at the same rate or for the same amount of time at different locations. Temperature, presence of noxious chemicals, and concentrations of particulates in the water are all known to affect filtration rates and could account for differences in bacterial uptake at different sites. Filtration rate generally increases with temperature (Morton, 1971); however, temperatures above 20°C may also cause a decrease in filtration rates (Noordhuis et al., 1992). Although we did not measure water temperatures for most days during this study, measurements at these sites during similar weather conditions have revealed water temperatures between 20 and 30°C. Mussels are known to close up and stop filtering when exposed to noxious conditions (Jenner et al., 1992; Mouabad and Pihan, 1992). Furthermore, the average volume of water being filtered by a zebra mussel is known to decrease when the concentration of particles in the water increases (Lei et al., 1996). Such factors could account for differences in the degree of concentration of E. coli by mussels at Red Run (Site 8) and Metropolitan (Site 9) compared with other sites.
|
CONCLUSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. L. Ram, R. P. Ritchie, J. Fang, F. S. Gonzales, and J. P. Selegean Sequence-Based Source Tracking of Escherichia coli Based on Genetic Diversity of {beta}-Glucuronidase J. Environ. Qual., May 1, 2004; 33(3): 1024 - 1032. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 | |||
