JEQ Grow Your Career With ASA
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 Related articles in JEQ
Right arrow Similar articles in this journal
Right arrow Similar articles in 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 Web of Science (8)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bomo, A.-M.
Right arrow Articles by Hanssen, J. F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bomo, A.-M.
Right arrow Articles by Hanssen, J. F.
Agricola
Right arrow Articles by Bomo, A.-M.
Right arrow Articles by Hanssen, J. F.
Related Collections
Right arrow Water Management
Right arrow Water Quality
Right arrow Other Environmental Contamination
Right arrow Other Waste Management
Published in J. Environ. Qual. 33:1041-1047 (2004).
© ASA, CSSA, SSSA
677 S. Segoe Rd., Madison, WI 53711 USA

TECHNICAL REPORTS

Waste Management

Bacterial Removal and Protozoan Grazing in Biological Sand Filters

Anne-Marie Bomo*,a, Tor Kristian Stevika, Ine Hovib and Jon Fredrik Hanssenc

a Department of Mathematical Sciences and Technology, Agricultural University of Norway, P.O. Box 5003, 1432 Ås, Norway
b Technical Department of Ski County Council, P.O. Box 3010, 1402 Ski, Norway
c Department of Chemistry, Biotechnology and Food Science, Agricultural University of Norway, P.O. Box 5040, 1432 Ås, Norway

* Corresponding author (anne-marie.bomo{at}imt.nlh.no).

Received for publication September 24, 2002.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
The objective of the study was to investigate the importance of protozoan predation as a biological removal mechanism in sand filters used for purification of bacteria from wastewater. Eleven sand filter columns were seeded with a high dose of wastewater (70 mm d–1) and a high concentration (108 colony forming units [CFU] mL–1) of Aeromonas hydrophila (American Type Culture Collection [ATCC] 14715) for a period of 30 d. Water samples from three filter outlets were analyzed for the concentration of A. hydrophila. In addition, one filter column was sacrificed each sampling day for the quantification of A. hydrophila, culturable bacteria (heterotrophic plate counts, HPC), total bacterial counts, and protozoa in the sand. The mean removal efficiency of A. hydrophila in the sand filter columns was 4.7 log units. The concentration of A. hydrophila in the sand filter effluent, however, had a clearly time-dependent pattern from high (log 6) and unstable concentrations to low and more stable levels (log 2). The removal efficiency of A. hydrophila correlated significantly (P = 0.0005, r2 = 0.6) with numbers of protozoa in the sand filters. Significantly higher (P < 0.05) concentrations of A. hydrophila were observed in sand filter effluents from columns treated with the protozoan inhibitor cycloheximide, compared with nontreated columns. Results from the present study show that protozoan grazing plays an important role as a bacterial removal mechanism in sand infiltration systems.

Abbreviations: CFU, colony forming units • HPC, heterotrophic plate counts


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
WASTEWATER EFFLUENT TREATMENT by soil infiltration and percolation has long been used as a simple, low-cost means of wastewater management throughout the world (Mancl and Peeples, 1991; Tchobanoglous and Burton, 1991; Wotton, 2002). When properly designed, constructed, and operated, soil infiltration systems can provide an effective means of wastewater treatment and disposal with regard to organics, nutrients, and bacteria (Anderson et al., 1994; Duncan et al., 1994; Hagedorn et al., 1981; Sundblad and Wittgren, 1996; Stevik, 1998). However, unless properly designed and operated, infiltration systems may pose a bacterial pollution risk to ground water and freshwater recipients. The majority of the studies of infiltration systems are based on the removal of fecal microorganisms from human waste (Bouma et al., 1972; Hagedorn et al., 1981; Farooq and Al-Yousef, 1993; Stevik et al., 1999a, 1999b; Harrison et al., 2000; Ausland et al., 2002). In addition, studies have shown that low-cost sand infiltration systems are promising as a way of treating fish farm effluents (Vigneswaran et al., 1991; Kristiansen and Cripps, 1996; Palacios and Timmons, 2001) and as an efficient method for removal of fish pathogenic bacteria from fish farm wastewater (Bomo et al., 2003). Bacterial removal mechanisms in infiltration systems are a combination of physical, chemical, and biological factors. The studies of Bomo et al. (2003) indicated that biological wastewater filters are dynamic systems and that biological factors (i.e., predation by the indigenous protozoa) can be of major importance in these systems. According to Acea and Alexander (1988) and Acea et al. (1988), protozoa are the main predators of bacteria. Stevik (1998) found an increase in enteric bacterial removal efficiency in the upper zones of biological filters possibly as a result of protozoan predation. Increasing the knowledge about the dynamics of biological sand filters is highly important as this will aid in optimizing infiltration systems for improved and stable bacterial removal efficiencies. To our knowledge, few studies have worked with the interaction between protozoan predation and bacterial removal from wastewater treated in infiltration systems. An experiment was therefore designed to study the interaction between protozoan grazing and the removal of pathogenic bacteria in biological sand filters.

The present study forms part of a larger project investigating the use of infiltration systems as an alternative method for removal of fish pathogenic bacteria in wastewater from freshwater fish farms. Our experimental system was therefore based on infiltration of wastewater from an inland fish farm and by using the fish pathogenic bacterium Aeromonas hydrophila (ATCC 14715) as our model organism.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Experimental Design
The experiment was performed in 11 plastic cylinder columns with a diameter of 15 cm and a height of 100 cm. The columns were packed with sand (grain size: 1–2 mm, d50 = 1.34) to a height of 80 cm. Each column was loaded with a high dose of wastewater (70 mm d–1 applied in 24 doses of 2.9 mm each) when compared with loading rates in identical systems (Stevik et al., 1999a). The loading was controlled by a system of programmable switches that were connected to magnetic valves and an electric pump. The wastewater was gently dosed into the columns using 4-cm sprinklers that were placed over the filter surfaces. The top of each column was closed with a tight lid. To eliminate the biological effects of temperature variations and avoid replication of A. hydrophila in filter columns, they were placed in a dark room with a temperature of 6 ± 1°C. The column studies were performed over a period of 30 d. Before commencing the experiment, the columns were inoculated with wastewater without addition of A. hydrophila for 8 wk. This was done to establish a microbial community in the filters and to simulate natural conditions. The wastewater was supplied from an inland fish farm every second day (Table 1) and fed into a temperature-regulated tank and stored at 4°C with continuous circulation.


View this table:
[in this window]
[in a new window]
  		Table 1. Composition of fish farm wastewater applied to sand filter columns.{dagger}

 
Bacterial Strains and Growth Conditions
A fresh culture of A. hydrophila ATCC 14715 was grown on tryptic soy agar (TSA) at 22°C for 48 h and then inoculated in Brain Heart Infusion broth (BHIB) and grown at 22°C for 24 h. This culture was washed twice by centrifugation and diluted with wastewater from the inland fish farm to an approximate density of 108 CFU mL–1. This procedure was repeated every second day to maintain a continuous and stable inlet concentration of A. hydrophila during the 30-d experimental run.

Sampling and Bacterial Enumeration in Water (Filter Inlet/Outlet)
Water samples from the inlet and outlets of three filter columns were collected every second day during the 30-d period. One milliliter of the samples was diluted 10-fold in phosphate-buffered saline (PBS) (1.34 g Na2HPO4·H2O L–1, 0.34 g NaH2PO4·2H2O L–1, 8.5 g NaCl L–1) and the colony forming units (CFU mL–1) determined by culturing on selective ampicillin–dextrin agar (AD agar) (Havelaar et al., 1986) with a 0.008% concentration of bromothymol-blue, and incubated for 24 h at 30°C. Yellow colonies with a diameter of 1 to 1.5 mm were counted as A. hydrophila.

A die-off study was performed to investigate the survival of A. hydrophila in fish farm effluent. Flasks (100 mL) containing 25 mL sterile and nonsterile fish farm wastewater were inoculated with a fresh culture of A. hydrophila. Flasks were incubated at 4°C on a shaker machine with gentle stirring. Samples were taken daily from the flasks for a period of 6 d and cultured on AD agar as described above.

Sampling and Protozoan Enumeration in Sand
Sand samples from three different levels, top (0–2 cm), middle (40–42 cm), and bottom (78–80 cm), of the filter columns were collected seven times during the experimental period. One filter column was sacrificed each time. In addition, one column was sacrificed before seeding the system (Day 1) with A. hydrophila. Three sand samples of 30 g were taken from each level and poured into glass bottles with 90 mL of filter-sterilized H2O. Numbers of protozoa in the sand samples were enumerated using the 4,6-diamidino-2-phenylindole (DAPI) direct count (DDC) method (Stevik et al., 1998). Briefly, mechanical treatment on a reciprocal shaking machine (Edmund Bühler VKS; Johanna Otto GmbH, Hechingen, Germany) with a speed of 100 strokes min–1 for 1 min was used to disperse the protozoa from the sand. One milliliter supernatant was then fixed with 40 µL glutaraldehyde (final concentration = 1%) and stained with 0.4 mL DAPI (working concentration: 2 µg mL–1) for 30 min. The fixed and stained sample was filtered on black 0.8-µm polycarbonate filters, 25 mm in diameter (Costar Scientific, Cambridge, MA), without vacuum. The polycarbonate filters were mounted on a glass slide with a drop of paraffin, and numbers of protozoa counted by an epifluorescence microscope (Leica [Wetzlar, Germany] DMRE, Filterblock A; excitation wavelength: 340–380 nm) using the 40x objective. To attain statistical confidence, 40 fields of view per filter were searched.

The criteria used to define those protozoan cells to be counted were (i) cells containing distinctly stained nucleus or nuclei and (ii) a clear cell outline and cytoplasm.

Bacterial Enumeration in Sand
Sand samples from the same three levels, top (0–2 cm), middle (40–42 cm), and bottom (78–80 cm), were collected to quantify bacterial numbers in the filters. Three samples of 10 g were taken from each level and poured into glass bottles with 90 mL PBS. To separate bacteria from sand, the samples were vigorously hand-shaken for 3 min, and thereafter allowed to mix for an additional 10 min on a shaking machine according to Stevik (1998). For determination of the concentration of A. hydrophila in sand, appropriate serial dilutions were plated on AD agar and incubated and colonies counted as described previously. Culturable bacteria were enumerated using heterotrophic plate counts (HPC) (Reasoner and Geldreich, 1985). Appropriate serial dilutions of sand extracts were plated on low-nutrient agar R2A (CM 906; Oxoid Ltd., Hampshire, England) and incubated for 10 d at 20°C in the dark. Total bacterial numbers were quantified using the SYBR Green I method for rapid epifluorescence bacterial counts (Noble and Fuhrman, 1998). One milliliter of an appropriate dilution of sand extract was filtered through a 0.1-µm-pore-size Al2O3 Anodisc 25-mm membrane filter (Whatman, Maidstone, UK) with a slight vacuum (20 kPa). The Anodisc membrane was filtered to dryness and laid sample-side-up on 100 µL of the SYBR Green solution (final dilution = 2.5 x 10–3) and stained for 15 min in the dark. After the staining period, the filter was picked up and any remaining moisture was then carefully wicked away by touching the backside of the membrane with a soft paper. The Anodisc filter was mounted on a glass slide with a drop of a mixture of 50% glycerol and 50% PBS and 0.1% p-phenylenediamine (made fresh daily from frozen 10% aqueous stock). Bacteria were counted with an epifluorescence microscope (Leica DMRE, Filterblock I3; excitation wavelength: 450–490 nm) using the 100x objective. To attain statistical confidence, cells in 30 fields of view per filter were counted.

Predation Studies
Predation effects were also determined using cycloheximide-treated sand filter columns (Marino and Gannon, 1991; Davies et al., 1995) and comparing the removal efficiency of A. hydrophila in these sand filter columns with nontreated sand filter columns. A total of six sand columns were used in this study, with all columns having the same setup and specifications as described previously. After an 8-wk maturation period and before seeding the columns with A. hydrophila, three sand columns were treated with 225 mg L–1 cycloheximide. Determination of the appropriate concentration of cycloheximide was based on the cycloheximide concentration (75 mg L–1) traditionally used in growth media to inhibit growth of eukaryotic organisms (Tronsmo, 2001). This concentration was doubled and tripled and the possible effect of this high concentration on the survival of A. hydrophila was investigated in a microcosm experiment. Flasks containing three levels of cycloheximide (75, 150, and 225 mg L–1) were inoculated with A. hydrophila and incubated at 4°C with continuous shaking. Samples were taken on Days 1, 4, 12, and 20 and diluted and cultured on AD agar as described earlier. No difference in survival of A. hydrophila and increasing concentration of cycloheximide was observed. The highest concentration of cycloheximide (225 mg L–1) was therefore chosen to ensure as high as possible elimination of protozoa. Fish farm wastewater was then sterilized by autoclaving, mixed with cycloheximide to a final concentration of 225 mg L–1, and dosed to the surface of the three filters in the same manner as previously described. By closing the column outlet, columns were slowly saturated with cycloheximide-containing water. Columns were then slowly drained and the treatment was repeated. Three control columns were treated in the exact same way, except that water did not contain cycloheximide. Aeromonas hydrophila was then grown and diluted with sterilized fish farm wastewater to an approximate concentration of 108 CFU mL–1 and continuously added to filter columns in the same way as described earlier. Water samples from filter outlets were collected 12 times during a 25-d period and cultured and analyzed as described previously.

Tracer Measurement
To determine the hydraulic retention time (HRT) through the filter columns, a solution of 100 mg L–1 potassium bromide (KBr) was dosed onto the filter surface of three columns at the same time as an ordinary dose of wastewater. Measurements of HRT were performed after the 8-wk maturation period but before seeding the columns with A. hydrophila. The bromide concentration in the filter effluents was measured with an ion selective electrode (Metrohm [Herisau, Switzerland] 6.0502.100 Br–/0–50°C) to get breakthrough curves for the columns. A first breakthrough of tracer occurred after 4 h in the sand filter. The highest bromide concentration (approximately 10% of initial tracer concentration) was measured after 9 h, after which time the bromide concentration started to decline again.

Statistical Analysis
All bacterial and protozoan counts were log10–transformed before analysis. The data were analyzed using simple linear regression and analysis of variance (ANOVA) with t test and statistically significant differences for P < 0.05. The statistical program used was JMP Version 4 (SAS Institute, 2000).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 REFERENCES
 
The inlet concentration of A. hydrophila was 5.28 x 108 ± 1.59 x 109 CFU mL–1 throughout the experiment, with the sand filter columns having a high average removal efficiency (4.7 log units) of A. hydrophila when assessed over the entire experimental period. The concentration of A. hydrophila in the filter outlet was, however, highly time-dependent with a significant reduction with time (Fig. 1) . Two to four days after seeding the columns with A. hydrophila, the concentration in the filter outlet was relatively high (log 4–6), followed by a rapid (Days 5–7) and then a more steady decrease (Days 7–14). From Day 14, the bacterial concentration in the filter outlet stabilized and stayed at a low level (log 2) during rest of the experimental period. The bacterial removal pattern in the columns can therefore be divided into a start-up period (Days 1–7), an intermediate period (Days 7–14), and a period of steady state (Days 14–30). The mean value for reduction of A. hydrophila in steady state was 5.6 log units.



View larger version (25K):
[in this window]
[in a new window]
  		Fig. 1. Concentration of Aeromonas hydrophila (AH water) in filter effluents (in colony forming units [CFU] mL–1) and protozoan numbers (PR) in top, middle, and bottom layers of the sand columns (protozoa g–1 dry wt. sand). Error bars = standard deviations (n = 3).

 
Low numbers of protozoa (<101 cells g–1 dry wt. sand) were observed in all layers in the sand before seeding the columns with A. hydrophila (Day 1) and during the start-up period with no significant differences between layers (P > 0.05). In the intermediate period (Days 7–14) the number of protozoa in top layers of the columns increased significantly (P < 0.002), followed by stabilization and a relatively constant protozoan number (103 cells g–1 dry wt. sand) throughout the rest of the experimental period (steady state) (Fig. 1). Significant increases in protozoan numbers in the intermediate period were also observed in middle (P = 0.011) and bottom (P = 0.006) layers in the sand columns. In the middle layer, the protozoan number was increasing until Day 21 after which the concentration stabilized. In the bottom layers, no significant difference in protozoan numbers was observed after Day 15.

In intermediate and steady state, the number of protozoa in the top layers was at all times significantly higher (P < 0.0001) than in middle and bottom layers, whereas protozoan numbers in the middle layer was significantly higher (P = 0.0024) than in the bottom layers from Days 21 to 30. Correlating the data for removal efficiency (inlet concentration – outlet concentration) of A. hydrophila with the numbers of protozoa from the top layer in the filter columns gave a significant correlation (P = 0.0005) with r2 = 0.6.

Treating sand columns with cycloheximide to verify the effect of protozoan predation as a bacterial removal mechanism showed that a significantly higher concentration of A. hydrophila was measured in filter outlet from cycloheximide-treated columns compared with nontreated columns (Table 2). A significant difference was observed on each sampling during a 25-d period except on Samplings 7, 8, and 9 when no differences were observed. To verify the possible washout of cycloheximide from the system and consequently an increase in protozoan numbers in the sand columns, the cycloheximide treatment was repeated. A significantly higher concentration of A. hydrophila in cycloheximide-treated sand filter outlets was observed post-treatment (Samplings 10–12) (Table 2).


View this table:
[in this window]
[in a new window]
  		Table 2. Effluent concentrations of Aeromonas hydrophila in cycloheximide-treated and nontreated sand filter columns.{dagger}

 
Possible effect of cycloheximide on A. hydrophila was determined in A. hydrophila cultures with and without cycloheximide at concentrations used in the experiment. Cycloheximide-treated and untreated cultures exhibited equivalent growth curves (results not shown).

No observations of die-off in fish farm effluents were observed for A. hydrophila during a 6-d observation period with the concentration of A. hydrophila in sterile and nonsterile microcosms being relatively stable at log levels of 9.00 ± 0.28 and 8.55 ± 0.24, respectively.

The concentration of A. hydrophila within the sand (CFU g–1 dry wt. sand) declined with increasing column depth, with significant differences (P < 0.0001) between all layers in the columns (top, middle, and bottom) (Fig. 2) with the concentration in the top layer being the highest. This distribution was the same throughout the experimental period. The highest concentration of A. hydrophila in the top layer of the sand columns was measured the day after seeding with the bacterium (Day 1). For the rest of the experimental period the concentration of A. hydrophila in top layer decreased steadily with significant difference (P < 0.0001) for each day except between Days 4 and 7. The steepest decline was observed from Days 11 to 15 where the A. hydrophila concentration decreased with 1 log unit. In middle layer, the concentration of A. hydrophila was 3 log units lower than in the top layer (P < 0.0001) measured on Day 1. Toward Day 11 the concentration of A. hydrophila in middle layers increased whereas from Day 15 and to the conclusion of the experiment, the concentration decreased (P < 0.0001). The approximate same pattern was seen for bottom layers of the sand columns where a significant increase (P < 0.0001) in A. hydrophila concentration was observed from Days 1 to 4; it then stabilized and then declined (P < 0.0001) toward the end of the experiment.



View larger version (16K):
[in this window]
[in a new window]
  		Fig. 2. Concentration of Aeromonas hydrophila in sand in different layers in filter columns (colony forming units [CFU] g–1 dry wt. sand). Results are log10–transformed. Error bars = standard deviations (n = 3).

 
The concentration of culturable bacteria (HPC) in top layers of the sand filters (CFU g–1 dry wt. sand) (Fig. 3) increased significantly (P = 0.0003) the day after seeding the columns (Day 1) with A. hydrophila. Throughout the rest of the experimental period the HPC concentration in top layers remained relatively stable with small changes (P > 0.05) from day to day, except on Day 21 where a significant increase was observed. The concentration of HPC was at all sampling times significantly higher in top layers compared with middle and bottom layers (P = 0.0002), except on Day 11 were no differences (P > 0.05) were observed. The HPC concentration in middle and bottom layers followed nearly identical patterns. A significant increase in HPC concentration was observed on Day 4 with a steady increase toward Day 11 (P < 0.0001). The concentration then declined with insignificant differences (P > 0.05) toward the end of the experiment (Days 21–30).



View larger version (15K):
[in this window]
[in a new window]
  		Fig. 3. Concentration of culturable bacteria (HPC) in sand in different layers in filter columns (colony forming units [CFU] g–1 dry wt. sand). Results are log10–transformed. Error bars = standard deviations (n = 3).

 
Before seeding the columns (Day 1) with A. hydrophila, the number of total bacterial cells (cells g–1 dry wt. sand) was slightly higher in middle layer of the columns (P = 0.03) compared with top and bottom (Fig. 4) , whereas for the rest of the experimental period, differences in total bacterial numbers at all sampling depths were small and not statistically different (P > 0.05). The number of total bacterial cells in top layers of the columns remained stable throughout the experimental period with no significant differences (P > 0.05) between sampling days. The total bacterial numbers in middle and bottom layers were unstable with significant changes from Days 4 to 11 (P = 0.006) and Days 4 to 15 (P = 0.001), respectively. For the rest of the experimental period, the concentration at both levels stabilized.



View larger version (15K):
[in this window]
[in a new window]
  		Fig. 4. Numbers of total bacterial cells in sand in different layers in filter columns (cells g–1 dry wt. sand). Results are log10–transformed. Error bars = standard deviations (n = 3). Bacterial numbers on Day 21 not determined.

 
The numbers of A. hydrophila, HPC, and total bacterial counts in sand were correlated against the numbers of protozoa detected in the sand. A significant correlation (P < 0.0001, r2 = 0.64) was found for the concentration of A. hydrophila and protozoan numbers in the top layers of the columns, whereas for subsurface layers and for HPC and total bacterial counts, no such relationship could be verified statistically (P > 0.05).


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 REFERENCES
 
The significantly higher concentration of A. hydrophila in sand filter effluents from cycloheximide-treated columns compared with nontreated columns revealed that predation by protozoa may play a significant role as a bacterial removal mechanism in sand filter systems. This is consistent with the work of Weber-Shirk and Dick (1997)(1999) where bacterivory was found to be the principal responsible biological removal mechanism for bacteria in slow sand filters and where E. coli concentrations declined more rapidly in effluents from slow sand filters than in protozoan-free control samples. Other predation studies have also shown an increased bacterial growth and survival in sediments treated with a protozoan growth inhibitor compared with nontreated systems (Marino and Gannon, 1991; Davies et al., 1995).

Investigating the removal of A. hydrophila in sand filter columns with time showed that the overall removal efficiency of A. hydrophila was high (4.7 log units) and comparable with removal efficiencies found in pilot-scale infiltration systems (Farooq and Al-Yousef, 1993; Stevik et al., 1999a; Ausland et al., 2002) as well as full-scale buried sand filters (Harrison et al., 2000). However, a significant time-dependent removal pattern was observed, with the bacterial concentrations in filter effluents varying from high (log 4–6) and unstable concentrations, then followed by a decline to low and stable levels (log 2). This is consistent with a previous study in the same experimental system (Bomo et al., 2003). The numbers of protozoa in the sand filter columns were found to correlate significantly (r2 = 0.6, P = 0.0005) with the increased bacterial removal performance of the filters. Similar significant correlations were found in the study of Marino and Gannon (1991) where comparisons of concentrations of bacteria and protozoan numbers in sediments gave an r2 value of 0.6. Previous studies have confirmed the important role protozoa play in consuming bacteria in both sewage treatment and soil ecosystems (Acea and Alexander, 1988; Acea et al., 1988; Ayo et al., 2001; Decamp and Warren, 1998; Eberl et al., 1997; England et al., 1993; Griffiths, 1990), as well as a biological removal mechanism in slow sand filtration (Lloyd, 1996; Weber-Shirk and Dick, 1997, 1999). It is, however, evident that other mechanisms such as adhesion of bacterial cells to filter media surface also play a role in bacterial removal as well as die-off of bacterial cells in the system. In a study comparing attachment versus predation, Lloyd (1996) and Weber-Shirk and Dick (1997) found that protozoan predation was of more significance than attachment of bacterial cells to filter surfaces. Results from our study also showed that protozoan predation was the dominating removal mechanism, verified by the response observed between higher bacterial removal efficiency and increasing protozoan numbers (Fig. 1). Treating sand filter columns with a protozoan inhibitor (cycloheximide) also revealed that a significantly higher effluent concentration of A. hydrophila was observed in treated columns compared with nontreated columns (Table 2). No die-off of A. hydrophila was observed during a 6-d observation study. This, in combination with the relatively short hydraulic retention time in the sand filters (9 h), reveals that die-off as a reason for declining concentrations of A. hydrophila observed both within sand and in sand filter effluents can probably be eliminated.

Before seeding the filter columns with A. hydrophila (Day 1), the numbers of protozoa in the sand were low (<101 cells g–1 dry wt. sand). After seeding and during the start-up period, the concentration of A. hydrophila in the top level of the filter columns was high (approximately log 6 g–1 dry wt. sand), but no immediate effect of increased protozoan populations was observed. This indicates that the filters were subjected to a protozoan lag period (i.e., the time required for protozoa to reach sufficient density to account for an efficient removal of bacteria). In soil, a lagged pulse in protozoan numbers will follow a pulse increase in the level of soil bacteria, normally a few days after the bacterial peak (Killham, 1994). Hunt et al. (1989) noticed a protozoan lag time of 4 d in soil in semiarid ecosystems, whereas Enzinger and Cooper (1976) found a lag period of 2 to 4 d before protozoan predators efficiently reduced the numbers of E. coli in freshwater samples. In our study, the concentration of A. hydrophila in filter effluent started to decrease on Day 5, whereas protozoan numbers started to increase on Day 7. A significant and detectable increase in protozoa was observed from Days 7 to 14, after which time predation by protozoa was great enough to account for a high and stable bacterial removal. This defines our protozoan lag period as being at least 7 d. The length of a lag period will be brief or extended depending on the system that is studied. In our study, the short retention time for the bacteria in the filters may explain the longer protozoan lag, as compared with more stable systems like soils.

The highest numbers of protozoa were found in the top layer of the columns, which is consistent with the observations of Stevik (1998). The concentration of A. hydrophila and HPC was also highest in the surface layers of the filter. This is in agreement with other studies where a higher and more active microbial biomass is observed in surface layers of sand filters (Madoni et al., 2001; Calvo-Bado et al., 2003) and with a reduction of bacteria with increasing depth (Ellis and Aydin, 1995; Pell and Nyberg, 1989). However, the high surface concentration of A. hydrophila was an expected observation given the fact that the sand filters were continuously seeded with this bacterium. Despite the continuous seeding, the surface layer concentration of A. hydrophila steadily declined throughout the 30-d experimental period with the steepest reduction from Days 11 to 15. At the same time, the protozoan numbers in top layers peaked and stabilized. Similar observations were made in all layers in the sand, but not so distinct as in the top layers. Stevik (1998) observed a strong correlation between protozoa and bacteria in surface as well as subsurface layers in infiltration systems, whereas in our study such a correlation between A. hydrophila and protozoa was only verified statistically (P < 0.0001, r2 = 0.64) in the surface layers of the sand filters. Seeding the filter columns with A. hydrophila probably accounted for the significant increase in HPC observed on Day 1 in surface layers and on Day 4 in middle and bottom layers. For the rest of the experimental period, the HPC concentration was relatively stable in all layers in the sand columns and did not show the same declining pattern as the concentration of A. hydrophila and no significant correlation (P = 0.3) was found between protozoan numbers and culturable bacteria (HPC) in the system.

In contrast to the distribution of A. hydrophila and HPC within the sand filters, the distribution of total bacterial cells did not show any significant reduction with filter depth. Stevik (1998) also observed this when counting total bacterial cells in fine weathered sand columns. A likely explanation for our observation was that the relatively high flow rate (70 mm d–1) in combination with a relatively coarse filter medium might have contributed to the low reduction of the total bacterial number with depth (Bahgat et al., 1999; Stevik, 1998). In addition, reduced availability of nutrients with depth in sand columns (Ellis and Aydin, 1995) will reduce the ability of bacterial cells to grow on nutrient agar, whereas quantification of total bacterial cells is based on direct microscopy counts, and will therefore not detect differences in bacterial counts based on nutrient gradients and reduced culturability. The addition of A. hydrophila to the filter columns did not seem to affect the total bacterial numbers in the filters, and as for HPC, no significant correlation (P = 0.5) was observed between total bacterial cells and the increasing numbers of protozoa in the sand. This indicates that the size of the indigenous bacterial community (HPC and total bacterial cells) remained fairly unaffected from the addition of a new bacterium to the system and that the protozoa seemed to fluctuate with the pathogen (A. hydrophila) brought into the system and not with the indigenous bacteria.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
REFERENCES
 
Results from the present study show that protozoan grazing plays an important role as a bacterial removal mechanism in sand infiltration systems. Significantly higher (P < 0.05) concentrations of A. hydrophila were observed in sand filter effluents from columns treated with cycloheximide, compared with nontreated columns. The concentration of A. hydrophila in the sand filter effluent showed a clearly time-dependent pattern from high (log 6) and unstable concentrations to low and stable levels (log 2) where a significant correlation was found between numbers of protozoa in the sand filter and the removal efficiency of A. hydrophila (P = 0.0005).


   ACKNOWLEDGMENTS
 
The Agricultural University of Norway financially supported this work.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 


Related articles in JEQ:

This Issue in Journal of Environmental Quality

JEQ 2004 33: 799-804. [Full Text]  



This article has been cited by other articles:


Home page
 J. Environ. Qual.Home page
J. A. Amador, D. A. Potts, M. C. Savin, P. Tomlinson, J. H. Gorres, and E. L. Nicosia
Mesocosm-Scale Evaluation of Faunal and Microbial Communities of Aerated and Unaerated Leachfield Soil
J. Environ. Qual., May 31, 2006; 35(4): 1160 - 1169.
[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 Related articles in JEQ
Right arrow Similar articles in this journal
Right arrow Similar articles in 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 Web of Science (8)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bomo, A.-M.
Right arrow Articles by Hanssen, J. F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bomo, A.-M.
Right arrow Articles by Hanssen, J. F.
Agricola
Right arrow Articles by Bomo, A.-M.
Right arrow Articles by Hanssen, J. F.
Related Collections
Right arrow Water Management
Right arrow Water Quality
Right arrow Other Environmental Contamination
Right arrow Other Waste Management

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