| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Environmental Science Graduate Program, The Ohio State Univ., 590 Woody Hayes Dr., Columbus, OH, 43210
b Dep. Veterinary Preventative Medicine, The Ohio State Univ., 1920 Coffey Rd., Columbus, OH, 43210
c Dep. Food, Agricultural and Biological Engineering, The Ohio State Univ., 590 Woody Hayes Dr., Columbus, OH, 43210
* Corresponding author (morgan.559{at}osu.edu).
Received for publication March 6, 2007.
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMaterials and MethodsResultsDiscussionConclusionsREFERENCES |
|---|
Abbreviations: TC, total coliform • AEES, Advanced Ecologically Engineered System • WETS, Waterman Ecological Treatment System • BOD, biological oxygen demand • PC, principle components • HRT, hydraulic retention time
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and MethodsResultsDiscussionConclusionsREFERENCES |
|---|
Ecological treatment systems rely on renewable resources and consist of a combination of anaerobic reactors, vegetated reactors, and wetlands. These systems have successfully treated municipal and industrial effluents with reduced costs compared to conventional methods (Todd and Josephson, 1996; Austin, 2000), and could meet the need for technically sound and economically feasible agricultural wastewater treatment. Nutrients, solids, and biochemical oxygen demand (BOD) also have been effectively removed from dairy wastewater using ecological treatment systems (Lansing and Martin, 2006). To date, most of the research on ecological treatment systems has focused on the ability of these systems to remove excess nutrients, turbidity, and BOD from wastewater (Peterson and Teal, 1996; Austin, 2000). However, another factor key to making these systems successful is ensuring their ability to reduce coliform concentrations in wastewater.
Wastewater is the primary source of fecal contamination in aquatic ecosystems (Griffin et al., 2001; George et al., 2002), containing total and fecal coliforms on the order of 108–1010 and 107–109 colony forming units (CFU) L–1, respectively (George et al., 2002). The use of fecal contaminated water can lead to gastrointestinal and respiratory illness as well as eye, ear, and skin infections (Cheung et al., 1990). Total and fecal coliforms, and E. coli are commonly used as pathogen indicator organisms because they are present in human and animal intestines and excreted in feces. These indicator organisms are not pathogens themselves, but their presence indicates the existence of pathogenic organisms.
While many studies have documented large reductions in coliform concentrations between influent and effluent of wastewater treatment systems, few studies have attempted to explain the reduction mechanisms. Mechanisms of coliform reduction in wastewater treatment systems are speculated to include: adsorption onto suspended organic matter followed by primary settling (An et al., 2002; George et al., 2002), physical filtration, predation by heterotrophs, exudation of allelopathogens from hydrophytic vegetation (Coombes and Collett, 1995), and exposure to ultraviolet light (George et al., 2002). Long hydraulic retention times increase the efficiency of these removal mechanisms (Toet et al., 2005). Temperature and substrate type have also been found to correlate with fecal coliform removal (Rivera et al., 1995; An et al., 2002). E. coli concentrations in wastewater, and the incidence of infection in cattle, have been shown to reach maximum levels in warmer months (Hancock et al., 1994; Wang et al., 1996).
Ecological treatment systems are a relatively new technology and have not been used commercially in the U.S. to treat agricultural wastewater. Each component of the treatment system is contained within a separate tank, allowing the specific locations, and therefore mechanisms, of nutrient and coliform removal to be identified and enhanced. To date, a single study evaluated the ability of an ecological treatment system to remove bacterial coliforms from municipal wastewater (Austin, 2000). At the Vermont Advanced Ecologically Engineered System (AEES), fecal coliform removal ranged from 99.97 to 99.99% (Austin, 2000). The primary fecal coliform removal mechanisms in the AEES were identified as sedimentation in the clarifier, filtration after the clarifier, and predation by zooplankton.
Reduced economic and energy inputs required by ecological treatment systems (Austin, 2000) and successful water quality improvements (Lansing and Martin, 2006) may result in their increased use on dairy farms. However, to reuse the wastewater on the farm, or release it to surface water, monthly geometric averages for E. coli concentrations cannot exceed 1260 CFU L–1 (USEPA, 2004). Consequently, the ability of ecological treatment systems to remove coliforms from wastewater is a critical issue necessitating further research.
Although ecological treatment systems have successfully treated municipal and industrial wastewater, their capacity to treat agricultural wastewater is unknown. To be effective for agricultural wastewater treatment the systems must be able to effectively treat a high strength of wastewater. The objectives of this study were to determine whether the efficiency of total coliform and E. coli reduction within an ecological treatment system decreased as the strength of dairy wastewater increased from low to high, and to gain insight regarding the primary mechanisms responsible for these reductions. By sampling at multiple locations within the ecological treatment system, the location of coliform reduction can be identified, and correlations between water quality variables and coliform reduction can be calculated. Based on the success of past ecological treatment systems, we hypothesized that total coliforms and E. coli concentrations would be significantly reduced between the influent and effluent of the ecological treatment system during the low strength dosing, and that reductions of coliforms would decrease during high strength dosing.
|
Materials and Methods |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and MethodsResultsDiscussionConclusionsREFERENCES |
|---|
Each treatment line is designed in the following manner: one anaerobic reactor (R1) (0.587 m3), one anoxic reactor (R2) (0.416 m3), one closed aerobic reactor (R3) (0.416 m3), one vegetated aerobic reactor (R4) (0.416 m3), one clarifier (C1) (0.586 m3), one subsurface flow gravel wetland (W1) (1.2 x 0.6 x 1.2 m) (length x width x depth), two vegetated aerobic reactors (R5 & R6) (0.34 m3), one clarifier (C2) (0.34 m3), and two subsurface flow gravel wetlands (W2) (1.2 x 0.6 x 0.3 m) (length x width x depth) (Fig. 1 ).
|
Plastic racks affixed to the top of the aerobic reactors support vegetation on the water surface and allow the roots to be suspended in the reactors, where they provide surface area for attachment of nitrifying and denitrifying bacteria. The aerobic reactors are dominated by Cyperus papyrus L. and Colocasia esculentus Schott, and have lower abundances of Iris pseudocorus L., Hibiscus moscheutos L., Canna spp., Saururus cernuus L., Salix nigra Marshall, and Lemna spp. The wetland mesocosms are vegetated with a mix of Schoenoplectus tabernaemontani Vahl, Chloris spp., Polygonum persicaria L., Juncus spp., and Salix nigra. Stem density averaged 44 stems 0.5 m–2 in R4, 65 stems 0.5 m–2 in R5, and 124 stems 0.5 m–2 in R6. Stem densities were not recorded for the wetlands as the stems were so dense that a hand could not be inserted between the stems for counting; they were 100% vegetated.
Clarifiers were not aerated creating a calm environment for the settling and digestion of suspended solids. Calcium carbonate substrate in the wetlands provides exchange sites for phosphorus adsorption and further filtering of suspended solids. Additionally, the anaerobic environment of the wetlands promotes denitrification.
A two phase dosing experiment was conducted during the summer and early fall of 2005, to assess the capacity of the WETS to reduce coliform concentrations in dairy wastewater. During July, a small volume of wastewater was pumped into the dosing tank. The wastewater was diluted with well water at a ratio of one part wastewater to three parts well water for 4 wk, creating ‘low strength’ wastewater. Following a 2-wk break without dosing, a large volume of wastewater was pumped into the dosing tank for 6 wk. The wastewater to well water ratio was 2:1, creating a ‘high strength’ wastewater. The dilution ratios were selected based on the ease with which they could be created by manually operating the waste pump.
Microbiological Methods
Samples collected from the WETS were analyzed for total coliform and E. coli concentrations to assess the coliform reduction efficiency of the system. Sample locations were selected to provide information about coliform reduction mechanisms. System influent was collected from a sampling port installed between the dosing tank and first anaerobic reactor. Additionally, effluent of the anoxic reactor (R2), first clarifier (C1), second clarifier (C2), and last wetland (effluent) were collected in 50 mL sterile tubes from each sampling location on each treatment line (Fig. 1). Samples were collected during the second and fourth weeks of the low volume dosing, and second, fourth, and sixth weeks of the high volume dosing. Following collection, samples were immediately transported to the lab and processed.
Upon arrival at the lab, samples were immediately serially diluted in phosphate buffer solution to an end dilution of 10–5 (APHA Method 9050C, 1998). Total coliform (TC) and E. coli dilutions were cultured following a modification of method 9215 A in Standard Methods for the Examination of Water and Wastewater (APHA, 1998). The dilutions were spread on plates containing MacConkey agar for TC identification and incubated for 24 h at 37°C (Sanderson et al., 2005). For E. coli determination, samples were spread on plates containing Fluorocult MacConkey agar, containing 4-methylumbelliferyl-beta-D-glucuronide (MUG), and incubated 24 h at 37°C. E. coli colonies cleave MUG using enzyme β-D-glucuronidase releasing fluorogen, allowing them to be identified based on their ability to fluoresce under a UV lamp (APHA Method 9221F, 1998). Duplicate plates were inoculated for both TC and E. coli at each dilution. Although stressed coliforms often fail to grow on rich and selective media, we were only concerned with healthy, viable coliforms that might persist outside the host organism.
Plate counts were performed after 24 h of incubation. Coliform concentrations were calculated as number of colonies per volume of sample plated and reported as CFU per L. Standard counting procedure (APHA, 1998) dictates that only plates containing between 30 and 300 colonies are counted. However, because this was an experimental study to identify reduction mechanisms and efficiency, all colonies on all plates containing one to 300 colonies were counted. Furthermore, the USEPA discharge standards are based on a method that counts all colonies on all plates containing one to 200 colonies (USEPA, 2002). Percent reduction in coliform concentrations between sample locations was calculated using ((C1-C2)/C1)*100, where C1 is the concentration in the influent and C2 is the concentration in the effluent.
Water Quality Parameters
Several water quality parameters identified in the literature as relating to coliform reduction were measured. Wastewater samples were collected from the influent and effluent of the WETS twice a week in acid-washed nalgene bottles and transported to the lab on ice. For quality control, one field blank and one duplicate from a randomly chosen treatment line were collected once a week. Once during each dosing regime, water samples were collected from the anoxic reactor, the first clarifier, and the second clarifier in addition to influent and effluent. Upon arrival at the lab, 75 mL of the effluent samples, and 20 mL of the influent sample, were vacuum filtered through a Whatman glass fiber filter to collect solids for total suspended solids (TSS) analysis and dried to constant weight at 60°C (APHA, 1998). The filtrate was further pressure filtered through a 0.45 µm nylon membrane filter for NO3–N analysis on a Lachat Automated Flow Injection Analyzer (QC8000; Hach Chemical Company, Loveland, CO). All filtered samples were preserved with 1:5 H2SO4, stored at 4°C, and analyzed within 28 d of collection (APHA, 1998). Dissolved oxygen and temperature were measured in each tank once a week using a hand held YSI multiprobe (MPS 556; YSI, Yellow Springs, OH). Five-day carbonaceous biochemical oxygen demand (CBOD5) was assessed on influent and effluent samples once a week following APHA (1998) standard methods.
Statistical Analyses
Duplicate plates from the coliform counts were averaged to obtain a mean coliform count for each collected sample. The geometric means of the coliform concentrations were calculated for the influent and effluent samples and used to calculate percent reductions. Geometric means are a better reflection of central tendencies than arithmetic means, represent median microbial levels (Rose et al., 1996), and are used in regulations set by the USEPA. Arithmetic means are influenced by the frequency of very large sample values and may be a better representation of mass loading; therefore arithmetic means of coliform concentration and water quality variables were calculated for each sample location.
Normality of the water quality parameters was assessed with the Anderson-Darling test and homogeneity of variance was assessed with Levene's test for any continuous distribution. Based on the results of these tests, a natural log transformation was performed on TSS and temperature, and a natural log transformation plus one (i.e., ln(Y+1)) was performed on DO, NO3, TC, and E. coli. All tests of significance were evaluated at the alpha = 0.05 level.
Tests for Significant Differences
Two-way t tests and Mann–Whitney tests performed with Minitab software (Minitab, 2003) were used to analyze differences among influent and effluent coliform concentrations, during the different wastewater strengths.
Nested ANOVAs were used to differentiate the effect of the two categorical variables (wastewater strength and sample location) on water quality using SYSTAT software (Systat Software, Inc., 2004). Two types of nested ANOVAs were run. The first ANOVA nested wastewater strength within sample location to determine whether measured variables at a particular location differed between low and high wastewater strengths. The second ANOVA nested location within wastewater strength, to determine whether measured variables differed among the sample locations during a single wastewater strength.
Analysis of Coliform and Water Quality Variable Relationships
Principle components analysis (PCA) was used as an exploratory method to identify the water quality parameters of most importance in the wastewater. Correlation coefficients were calculated for each wastewater strength to determine whether significant relationships existed between coliform concentration and the other water quality variables.
|
Results |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and MethodsResultsDiscussionConclusionsREFERENCES |
|---|
|
Overall, TC concentration of the influent increased 67% from the low to high wastewater dosing, while E. coli concentrations increased 11%. Regardless of wastewater strength, TC and E. coli concentrations were consistently reduced by at least 99% from influent to effluent of the WETS (Table 1). The reduction was statistically significant for TC and E. coli during the high wastewater strength (two-way t test (TC) and Mann Whitney (EC) P < 0.001); as well as for TC during the low wastewater strength (P = 0.051). No significant differences were found in effluent concentrations of TC and E. coli when the high and low wastewater strengths were compared (two-way t test P = 0.296 and 0.284, respectively). The lowest effluent concentrations of both TC and E. coli occurred during the low wastewater strength (Table 1).
Coliform Relationships to Other Water Quality Variables
The first principle component (PC 1) of the PCA attributed 55% of the variation in water quality to TC, E. coli, and NO3– concentrations (Fig. 2a
). Total coliform and E. coli were highly correlated but negatively associated with NO3–. Principle component two was highly influenced by water temperature, but water temperature was unrelated to all other water quality variables. The PCA also illustrated that the water quality of influent and anoxic reactor sample locations was considerably different from all other sample locations (Fig. 2b). Nitrate, DO, temperature, and TSS all had significant loadings on the first two PCs and were therefore included in the nested ANOVAs.
|
|
0.292).
Examining differences in sample location during a single wastewater strength (i.e., nesting sample location within wastewater strength) revealed that temperature did not differ among sample locations within a given wastewater strength (nested ANOVA P = 0.975). As expected, TC, E. coli, DO, NO3–, and TSS concentrations did significantly differ among sample locations for a given wastewater strength (P 
0.001). Dissolved oxygen declined from the influent to anoxic reactor, and then increased throughout the system (Table 3
). Nitrate concentration increased from the anoxic reactor through the first clarifier, reflecting the effect of increased DO on nitrification. During the low wastewater strength, effluent from the first clarifier contained lower TSS than effluent from the anoxic reactor; however, during the high wastewater strength effluent from the first clarifier had higher TSS than the anoxic reactor effluent.
|
|
Discussion |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and MethodsResultsDiscussionConclusionsREFERENCES |
|---|
Regardless of wastewater strength, over 75% of the coliform concentration was reduced between entering the WETS and leaving the anoxic reactor (Table 3, Fig. 2b). The substantial reduction in coliforms in the anaerobic and anoxic reactors is likely the result of a combination of factors. Enteric coliforms are facultative anaerobic bacteria that can use nitrate as a terminal electron acceptor. Consequently, the large reduction in DO that occurs as the waste enters the first anaerobic digester will have a detrimental effect on the bacteria only if NO3– concentration is also low. Nitrate concentration in animal wastewater is generally low (Newman et al., 2000) and increases only as waste is oxygenated and nitrification occurs. During both the low and high wastewater strengths, NO3– and DO concentrations of the influent wastewater were very low, and coliform concentrations were dramatically reduced in the anaerobic and anoxic reactors.
The reduction of organic matter, or carbon sources, is generally one of the primary goals of wastewater treatment and is measured with the BOD5 technique. Carbon sources are required to sustain coliform populations. Consequently, the reduction of BOD can result in the simultaneous reduction of coliform populations (van der Steen et al., 2000). Although not measured in the current study, BOD was found to decline 79% between the influent and third aerobic reactor of the ecological treatment system in 2004 (Lansing and Martin, 2006). Williams et al. (1995) measured a stronger correlation of fecal coliform with BOD than with suspended solids and suggested that adsorption of fecal coliforms may be more important than sedimentation. Thus, the low oxygen and NO3– concentrations and high BOD reduction in the first two reactors likely account for the majority of the coliform reduction within the WETS.
The continued decline in coliform concentration throughout the system was likely influenced by increasing DO concentrations, as a negative correlation was measured between DO and coliform concentration (Table 2). Previous studies have also found decreases in coliform concentration at elevated DO concentrations (Curtis et al., 1992; van der Steen et al., 2000) and negative regression coefficients between coliform and DO concentrations (Júnior Athayde et al., 2000).
Coliform concentrations consistently declined throughout the WETS during the low wastewater strength. However, under the high wastewater strength, TCs and TSSs increased between the anoxic reactor and first clarifier (Table 3). During the high wastewater strength the system was overloaded with solids, causing solids that should have settled out in the clarifier to overflow to the wetlands. Total coliform and TSS concentrations were positively correlated (Table 2) and had a linear regression R2 of 66.5%. About 25% of coliforms in wastewater treatment system effluent are hypothesized to be attached to suspended matter >3 to 5 µm (George et al., 2002). Thus the larger mass of suspended solids leaving the clarifier compared to the anoxic reactor would result in a higher TC count. Settling of organic matter with adsorbed coliforms has been proposed as a possible mechanism of coliform removal (Coombes and Collett, 1995). A high level of correlation between TSS and TC was also found in a study by Gearheart (1999), indicating that sedimentation of suspended matter is an important coliform removal mechanism. Furthermore, reduction of coliforms in the clarifier of the WETS corresponds with findings of the AEES (2000), and confirms that sedimentation in clarifiers is an important mechanism of coliform removal in ecological treatment systems.
As expected, temperature of the wastewater differed between the two different wastewater strength periods as summer progressed into fall and days became cooler, but within a single wastewater strength temperature did not significantly differ among the sample locations. Although coliform removal is hypothesized to correlate with temperature (Khatiwada and Polprasert, 1999; An et al., 2002), studies have shown that temperatures greater than 37°C must be maintained for 15 d to kill coliforms (Kudva et al., 1998; Larney et al., 2003). Temperatures within the WETS never exceeded 26°C and were not correlated with coliform concentration, except on one occasion.
Although significant reductions between influent and effluent were accomplished in the WETS, 88% of the effluent samples had E. coli concentrations that exceeded those allowable for discharge, following our conservative plate counting methods. Further treatment of the wastewater may be necessary to reach discharge requirements and could be achieved through exposure to UV light. The effectiveness of UV light exposure as a bactericide has been extensively studied (Baron and Bourbigot, 1996; Davies-Colley et al., 1997; George et al., 2002). Exposure to 4 to 6 s of UV light has been found to reduce total coliform concentrations an additional 89% (Gearheart, 1999). Fecal coliforms are most susceptible to UVB light (290–320 nm), which induces the formation of pyrimidine dimers in DNA (Davies-Colley et al., 1997). In the current design, wastewater in the WETS is not exposed to UV light at a duration that would lead to bacterial cell death. Reactors in the WETS are either covered or vegetated, limiting UV light exposure.
|
Conclusions |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and MethodsResultsDiscussionConclusionsREFERENCES |
|---|
Increasing the wastewater strength did not affect treatment mechanisms operating in the WETS. Effluent concentrations of E. coli and total coliform did not significantly increase with increased wastewater strength. Based on these results we conclude that the high wastewater strength is most efficient as it handles the most waste with no significant reduction in treatment level. However, additional testing of these and higher strengths is needed to definitively determine the optimal loading rates based on coliform reductions.
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and MethodsResultsDiscussionConclusionsREFERENCES |
|---|
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and MethodsResultsDiscussionConclusionsREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 | |||
