Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Mechanisms and factors regulating organic carbon removal in surface flow constructed wetlands receiving… Tao, Wendong 2006

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2006-130407.pdf [ 16.15MB ]
Metadata
JSON: 831-1.0063297.json
JSON-LD: 831-1.0063297-ld.json
RDF/XML (Pretty): 831-1.0063297-rdf.xml
RDF/JSON: 831-1.0063297-rdf.json
Turtle: 831-1.0063297-turtle.txt
N-Triples: 831-1.0063297-rdf-ntriples.txt
Original Record: 831-1.0063297-source.json
Full Text
831-1.0063297-fulltext.txt
Citation
831-1.0063297.ris

Full Text

MECHANISMS AND FACTORS REGULATING ORGANIC CARBON REMOVAL IN SURFACE FLOW CONSTRUCTED WETLANDS RECEIVING WOOD WASTE L E A C H A T E  by WENDONG TAO M.Sc, Beijing Normal University, 1990  A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Civil Engineering)  THE UNIVERSITY OF BRITISH COLUMBIA November 2005 ©Wendong Tao, 2005  ABSTRACT A n amber leachate was generated by rainfall on an uncovered woodwaste pile beside Fraser River in Mission, B C , Canada. The "young" leachate in the pile's placement period was acidic (pH 3.4-3.7) and nutrient-poor, with high concentrations of chemical oxygen demand (COD 12559-14254 mg L" ), tannin and lignin ( T & L 3066-5150 mg L" ) and volatile fatty acids 1  1  (VFAs 1564-2132 mg L" ). The leachate at 1.5 years into the closure period became less 1  acidic, ammonia-rich, more refractory, and lower in organic strength. More than 98% of leachate constituents were soluble. Laboratory tests demonstrated insignificant volatilization, limited effect of sediment adsorption, and the major role of biological degradation for organic carbon removal from woodwaste leachate in surface flow constructed wetlands (CWs). The activity of the microbial community was investigated by measuring  H-leucine  incorporation into bacterial protein, C-glucose turnover, and bacterial assimilation and 14  mineralization of C-acetate in pilot-scale CWs receiving the young leachate. The effects of 14  hydraulic retention time (HRT), influent strength and mass loading rate on treatment performance, microbial biomass and heterotrophic activities were examined in mesocosm wetlands receiving the "aging" leachate. There were insignificant longitudinal variations in heterotrophic activities and insignificant vertical variation in biomass. It could take up to 6 weeks for microbial maturation. A H R T value of 5 d was enough to avoid adverse effects on bacterial communities. When a VFAs-rich influent was fed, C O D and T & L reduction efficiencies increased significantly with HRT. When a refractory influent of aged leachate was fed, reduction efficiencies increased slightly with HRT. Heterotrophic activities were regulated by the availability of organic substrates, electron acceptors and inorganic nutrients, which varied with influent strength and HRT. Nanoflagellate grazing controlled a proportion of active bacteria. The influents with up to 40% of the organic strength of young leachate did not inhibit the acclimatized microorganisms. Vegetation made little or no difference in performance. Reduction rates and rate constants were correlated to the heterotrophic activities of planktonic, epiphytic and sedimentary bacteria. Bacterioplankton and sedimentary bacteria usually contributed 77-99% to the total heterotrophic activities. The relative importance of sedimentary bacteria increased under substrate-limiting conditions, while bacterioplankton was favored by V F As-rich influents.  ii  TABLE OF CONTENTS ABSTRACT  ii  T A B L E OF CONTENTS  .iii  LIST OF T A B L E S  xi  LIST OF FIGURES  xiii  LIST OF A B B R E V I A T I O N S  xvi  ACKNOWLEDGMENTS  xvii  CO-AUTHORSHIP  xviii  1  L I T E R A T U R E R E V I E W A N D R E S E A R C H SCOPE  1.1  Constructed Wetlands for Water Quality Improvement  1 1  1.1.1  Definition and properties of natural wetlands  1  1.1.2 .  Properties of constructed wetlands  2  1.1.3  Diverse applications of wetlands for wastewater treatment  4  1.1.4  Treatment mechanisms of constructed wetlands  5  1.2  Design Considerations for Surface Flow Constructed Wetlands  8  1.2.1  Wetland layout and structures  8  1.2.2  Water depth and vegetation  .9  1.2.3  Hydraulic retention time and loading rate  10  1.2.4  Mass loading rate and oxygen transport  11  1.3  Characterization of Microorganisms in Aquatic Systems  12  1.3.1  Microbial diversity of aquatic systems  13  1.3.2  Spatial distribution of microorganisms in wetlands and biofilm collection  16  1.3.3  Determination of microbial biomass and enumeration  18  1.3.4  Determination of heterotrophic bacterial production  23  1.3.5  Determination of heterotrophic bacterial activity on organic matter  26  1.4 1.4.1  Background and Research Scope  29  Development of technology for treatment of wood leachate iii  29  1.4.2  Current situation of treatment wetland design  30  1.4.3  Research objectives and methodology  33  1.5 2  References  38  C H A R A C T E R I Z A T I O N OF W O O D W A S T E L E A C H A T E  56  2.1  Introduction..  56  2.2  Materials and Methods  57  2.2.1  Woodwaste pile and leachate pool  57  2.2.2  Field measurement and laboratory analysis  59  2.2.3  Statistical analysis  60  2.3  Results  '.  61  2.3.1  Physical and toxicological properties  61  2.3.2  Change over age of woodwaste pile  61  2.3.3  Composition of oxygen demand  63  2.3.4  Nutrient levels  65  2.3.5  Seasonal variation  66  2.4  Discussion  66  2.4.1  Characteristics of woodwaste leachate  66  2.4.2  Temporal variation of chemical properties  68  2.4.3  Implications for selection of treatment processes  69  2.5  Conclusions  70  2.6  References  70  3  L A B O R A T O R Y STUDY O N POTENTIAL M E C H A N I S M S F O R T R E A T M E N T OF W O O D W A S T E L E A C H A T E IN W E T L A N D S  73  3.1  Introduction  73  3.2  Materials and Methods  75  3.2.1  Woodwaste leachate and chemical analysis  75  3.2.2  Jar test for volatilization  75  3.2.3  Batch soil adsorption tests  75 iv  3.2.4  Setup and operation of bench-scale bioreactors  77  3.2.5  Statistical analysis  79  3.3  Results  79  3.3.1  Volatilization  79  3.3.2  Soil adsorption  79  3.3.3  Biological degradation  82  3.4  Discussion  86  3.4.1  Volatility  86  3.4.2  Soil adsorption capacity  86  3.4.3  Biological treatability  87  3.5  Conclusions  89  3.6  References  89  4  HETEROTROPHIC B A C T E R I A L ACTIVITIES A N D T R E A T M E N T P E R F O R M A N C E OF PILOT-SCALE W E T L A N D S  93  4.1  Introduction  93  4.2  Methodology  95  4.2.1  Operation of pilot-scale constructed wetlands  95  4.2.2  Field measurement and laboratory analysis  96  4.2.3  Sample preparation for radioisotope incubation  97  C-glucose uptake assays  98  4.2.4  l4  4.2.5  3  4.2.6 4.3  H-leucine incorporation assays  101  Statistical analysis  102  Results  102  4.3.1  Operating conditions and treatment performance  102  4.3.2  Dynamics of bacterioplanktonic activities  104  4.3.3  Effects of nutrient amendment and vegetation on heterotrophic activities  106  4.4  Discussion  106  4.4.1  Implications of heterotrophic dynamics  106  4.4.2  Effects of nutrient and electron acceptor amendment  108  v  4.4.3  Effect of vegetation  109  4.4.4  Relative importance of water, sediment and epiphyton  110  4.5  Conclusions  Ill  4.6  References  112  5  HETEROTROPHIC B A C T E R I A L UTILIZATION OF A C E T A T E IN PILOT-SCALE CONSTRUCTED WETLANDS  1.16  5.1  Introduction  116  5.2  Methods and Materials  117  5.2.1  Constructed wetlands  117  5.2.2  Radioisotope incubation and assays  117  5.2.3  Statistical analysis  119  5.3  Results  119  5.3.1  Microbial colonization  119  5.3.2  Evaluation of heterotrophic bacterial activity  119  5.3.3  Effects of fertilization and vegetation  121  5.4  Discussion  121  5.4.1  Development of microbial colonization  121  5.4.2  Comparison to similar treatment systems  122  5.4.3  Temporal and longitudinal variations  123  5.4.4  Relative contributions of water, sediment and epiphyton  123  5.4.5  Effects of fertilization and vegetation  124  5.5  Conclusions  125  5.6  References  126  6  P E R F O R M A N C E E V A L U A T I O N A N D EFFECTS OF H Y D R A U L I C RETENTION TIME A N D M A S S L O A D I N G R A T E O N T R E A T M E N T OF W O O D W A S T E L E A C H A T E IN M E S O C O S M W E T L A N D S  129  6.1  Introduction  129  6.2  Materials and Methods  131 vi  6.2.1  Setup of mesocosm wetlands  13  6.2.2  Field measurement and chemical analysis  13:  6.2.3  Determination of microbial biomass  13!  6.2.4  Calculations of operating parameters and statistical analysis  13;  6.3  Results  13'  6.3.1  Operating conditions  13'  6.3.2  Temporal and vertical variations of biomass  14i  6.3.3  Variations of treatment performance and kinetics  14;  6.4  Discussion  14  6.4.1  Time for microbial maturation  14  6.4.2  Disturbance of weather to performance  14  6.4.3  Implications of vertical variation  14i  6.4.4  Effect of hydraulic retention time  14i  6.4.5  Effect of mass loading rate  14  6.5  Conclusions  141  6.6  References  14'  7  EFFECTS OF H Y D R A U L I C RETENTION TIME A N D M A S S L O A D I N G R A T E O N MICROORGANISMS IN M E S O C O S M W E T L A N D S  1  7.1  Introduction  1  7.2  Materials and Methods  1  7.2.1  Sample preparation  1  7.2.2  Determination of cellular ATP  1  7.2.3  Determination of heterotrophic production  1  7.2.4  Determination of heterotrophic assimilation and mineralization  1  7.2.5  Statistical analysis  1  7.3  Results  1  7.3.1  Variation of microbial biomass  1  7.3.2  Variation of heterotrophic production  1  7.3.3  Variations of heterotrophic acetate assimilation and mineralization  1  vii  7.3.4 7.4  Correlation of treatment performance with heterotrophic activities Discussion  166 168  7.4.1  Establishment of microbial communities and effect of hydraulic retention time....168  7.4.2  Effects of mass loading rate and influent strength  168  7.4.3  Microbial diversity  170  7.4.4  Microbial bridges of pollutant loading and wetland performance  171  7.5  Conclusions  172  7.6  References  173  8  T R E A T M E N T OF W O O D W A S T E L E A C H A T E I N M E S O C O S M W E T L A N D S : EFFECTS OF INFLUENT STRENGTH A N D V E G E T A T I O N  177  8.1  Introduction  177  8.2  Materials and Methods  178  8.2.1  Woodwaste leachate and mesocosm wetlands  178  8.2.2  Field measurement, sampling and laboratory analysis  179  8.2.3  Data processing and statistical analysis  181  8.3  Results and Discussion  181  8.3.1  Operating conditions  181  8.3.2  Effect of influent strength on treatment performance  184  8.3.3  Effect of vegetation on treatment performance  187  8.3.4  Treatability of woodwaste leachate in wetlands  187  8.4  Conclusions  189  8.5  References  189  9  M I C R O B I A L C O M M U N I T Y STRUCTURE A N D HETEROTROPHIC ACTIVITIES OF M E S O C O S M W E T L A N D S : EFFECT OF INFLUENT S T R E N G T H  192  9.1  Introduction  192  9.2  Materials and Methods  194  9.2.1  Sample preparation for microbiological examinations  194  9.2.2  Enumeration of protozoa  194 viii  9.2.3  Enumeration of respiring bacteria  194  9.2.4  Determination of cellular ATP  195  9.2.5  Determination of heterotrophic production  196  9.2.6  Determination of heterotrophic assimilation and mineralization  196  9.2.7  Statistical analysis  196  9.3  Results  196  9.3.1  Operating conditions of mesocosm wetlands  196  9.3.2  Variation of heterotrophic activities  197  9.3.3  Variation of microbial community structure  198  9.4  Discussion  202  9.4.1  Microbial community structure  202  9.4.2  Factors regulating heterotrophic activities  204  9.4.3  Importance of heterotrophic bacteria in organic carbon removal  205  9.5  Conclusions  205  9.6  References  206  10  G E N E R A L DISCUSSION A N D CONCLUSIONS  :  211  Appendix 1. Parameters of woodwaste leachate occasionally monitored  215  Appendix 2. Quality assurance and quality controls for chemical analysis  216  Appendix 3. Insignificance of volatilization of woodwaste leachate  219  Appendix 4. Kinetics of adsorption on silt loam in woodwaste leachate  220  Appendix 5. Extraction and analysis of chlorophyll-a of algae  222  Appendix 6. Procedure of C-glucose uptake experiments  223  14  Appendix 7. Temporal variation of operating conditions of pilot-scale wetland #6 in 2001..225 Appendix 8. Temporal variation of operating conditions of pilot-scale wetlands in 2002....226 Appendix 9. Variation of heterotrophic acetate utilization in pilot-scale wetland #6 in 2001  228  Appendix 10. Extraction and assay of microbial ATP ix  229  Appendix 11. Temporal variations of influent and effluent of mesocosm wetlands in 2003  232  Appendix 12. Depth profiles of operating conditions in the mesocosm wetlands in 2003....235 Appendix 13. Properties of sediment and vegetation in the mesocosm wetlands in 2003  237  Appendix 14. Procedure of H-leucine incorporation experiments  238  3  Appendix 15. Adaptation of H-leucine incorporation technique to mesocosm samples  240  Appendix 16. Procedure of C-acetate uptake experiments  244  Appendix 17. Adaptation of C-acetate uptake technique to mesocosm samples  246  3  14  14  Appendix 18. Microbial biomass and heterotrophic activities in mesocosm wetlands in 2003  248  Appendix 19. Properties of sediment and vegetation in the mesocosm wetlands in 2004......250 Appendix 20. Temporal variations of operating conditions of mesocosm wetlands in 2004..251 Appendix 21. Dilution-live counting of protozoa by phase contrast microscopy  255  Appendix 22. Counting CTC-reducing bacteria by epifluorescence microscopy  257  Appendix 23. Microbial biomass, abundances and heterotrophic activities in mesocosm wetlands in 2004  260  LIST OF TABLES Table 1-1. Examples of acidophiles in natural environments  13  Table 2-1. Variation of woodwaste leachate quality oyer years  62  Table 3-1. Adsorption characteristics of silt loam in woodwaste leachate at 23°C and reduction efficiency at equilibrium  81  Table 3-2. Summary of operating parameters of the bench-scale reactors treating woodwaste leachate after 19 d of acclimatization  83  Table 3-3. Summary of performance of bioreactors for treatment of woodwaste leachate after 19 d of acclimatization  85  Table 4-1. Longitudinal variations of ammonia and nitrate in pilot-scale constructed wetlands  103  Table 4-2. Operating conditions and treatment performance of pilot constructed wetlands... 104 Table 4-3. Effects of ammonium nitrate addition and vegetation on heterotrophic bacterial activities in constructed wetlands treating woodwaste leachate  107  Table 4-4. Comparison of relative contributions of water, epiphyton, and sediment to the total heterotrophic activities by water and water equivalents in three types of constructed wetlands treating woodwaste leachate  110  Table 5-1. Bacterial activity in a pilot-scale surface flow constructed wetland receiving woodwaste leachate  120  Table 5-2. Effects of fertilization and vegetation on heterotrophic activity in surface flow constructed wetlands treating woodwaste leachate  122  Table 5-3. Effects of fertilization and vegetation on heterotrophic acetate utilization in 2002  125  Table 6-1. Operating conditions of the mesocosm wetlands fed with a weak influent during the first operating period  136  Table 6-2. Operating conditions of the mesocosm wetlands fed with a strong influent during the second operating period  137  xi  Table 6-3. Performance of woodwaste leachate treatment in the mesocosm wetlands fed with a weak influent during the first operating period  138  Table 6-4. Performance of woodwaste leachate treatment in the mesocosm wetlands fed with a strong influent during the second operating period  139  Table 7-1. Relative contributions of water, epiphyton and sediment to the total microbial biomass and heterotrophic activities of mesocosm wetlands treating woodwaste leachate  161  Table 7-2. Two-way analysis of variance of heterotrophic acetate utilization in mesocosm wetlands with different HRTs during two operating periods at different influent strengths  163  Table 7-3. Coefficients of Spearman's rank correlation of acetate utilization with mass loading rates to the mesocosm wetlands  165  Table 7-4. Nutrient balance of mesocosm wetlands treating woodwaste leachate in 2003... 171 Table 8-1. Operating conditions of mesocosm wetlands receiving different strengths of raw and dilute woodwaste leachate during the first operating period  182  Table 8-2. Operating conditions of mesocosm wetlands receiving refractory woodwaste leachate during the second operating period  183  Table 8-3. Nutrient balance of mesocosm wetlands treating woodwaste leachate in 2004... 184 Table 9-1. Average percentage contributions of water:epiphyton:sediment toward microbial biomass and heterotrophic activities of mesocosm wetlands treating woodwaste leachate  198  Table 9-2. Cellular A T P concentrations in water, epiphyton and sediment of mesocosm wetlands treating woodwaste leachate  xii  201  LIST OF FIGURES Figure 1-1. Appearance of natural wetlands  1  Figure 1-2. Illustration of surface flow constructed wetland and subsurface flow constructed wetland  3  Figure 1-3. Pathways in the degradation of organic matter in wetlands  14  Figure 1-4. Carbon flow in a surface flow constructed wetland treating high-organic-strength wastewater  32  Figure 1-5. Summary of research hypotheses tested and pertinent setup of pilot-scale vegetated and open constructed wetlands  34  Figure 1-6. Summary of research hypotheses tested and pertinent setup of vegetated and open mesocosm wetlands fed at different flow rates, with different strengths of influent, or operated in series  36  Figure 2-1. Woodwaste pile in October 2004 and leachate pool in March 2004  58  Figure 2-2. Variation of monthly air temperature and precipitation at Abbotsford A meteorological station  58  Figure 2-3. Temporal variation of woodwaste leachate quality  63  Figure 2-4. Temporal variation in composition of chemical oxygen demand of the leachate from a woodwaste pile  64  Figure 2-5. Yearly variation of the proportion of individual to total volatile fatty acids  65  Figure 2-6. Correlation of pH to monthly concentrations of oxygen-demanding substances of woodwaste leachate  65  Figure 3-1. Sketch of bench-scale aerated and anaerobic complete-mix reactors  78  Figure 3-2. Illustration of adsorption kinetics curves of silt loam in woodwaste leachate at 23°C  80  Figure 3-3. Correlation of reduction efficiency via silt loam adsorption at 23°C to initial aqueous concentration  81  xiii  Figure 3-4. Adsorption isotherm and capacity estimation with Langmuir equation at 23°C...82 Figure 3-5. Treatment performance of the bench-scale complete-mix bioreactors  84  Figure 4-1. Site plan of the pilot-scale constructed wetland treatment system and the structure of a vegetated wetland cell  96  Figure 4-2. Epiphyton sampler, syringes for incubation of water and epiphyton, and a flask for incubation of sediment  98  Figure 4-3. Leucine incorporation and glucose turnover by planktonic bacteria in a vegetated constructed wetland in 2001  105  Figure 4-4. Correlation of glucose turnover by planktonic bacteria to leucine incorporation and glucose mineralization percentage in a vegetated constructed wetland in 2001  106  Figure 5-1. Development of epiphyton in a surface flow constructed wetland treating woodwaste leachate from August 1 to September 25, 2001  119  Figure 5-2. Correlation between bacterial uptake rate and mineralization percentage in a surface flow constructed wetland treating woodwaste leachate in 2001  121  Figure 6-1. Layout of the mesocosm wetland treatment system and established mesocosm vegetation  131  Figure 6-2. Weekly precipitation and evapotranspiration in comparison to the hydraulic loading rate to mesocosm wetlands #1 and #4 in 2003  140  Figure 6-3. Depth profiles of water temperature, DO and redox potential in the mesocosm wetlands with different HRTs during the first operating period with a weak influent and during the second period with a strong influent.  141  Figure 6-4. Variation of cellular A T P in mesocosm wetlands with different HRT after switching influent from sewage to diluted woodwaste leachate  142  Figure 7-1. Variation of cellular A T P concentration over mesocosm wetlands with different HRTs during two operating periods at different influent strengths.... 160 Figure 7-2. Variation of heterotrophic leucine incorporation rate over mesocosm wetlands with different HRTs during two operating periods at different influent xiv  strengths  162  Figure 7-3. Variation of heterotrophic acetate uptake rates over mesocosm wetlands with different HRTs during two operating periods at different influent strengths  164  Figure 7-4. Conceptual model of wetlands for removal of COD, T & L , and V F A s from woodwaste leachate  167  Figure 8-1. Treatment of woodwaste leachate in vegetated and open mesocosm wetlands fed with different strengths of influent during two operating periods in 2004... 180 Figure 8-2. Variations of treatment performance and kinetics of mesocosm wetlands with influent strength  186  Figure 8-3. Comparison of treatment performance and kinetics of mesocosm wetlands in the vegetated mesocosm #1 - vegetated mesocosm #2 series and the vegetated mesocosm #1 - open mesocosm #3 series during the second operating period  188  Figure 9-1. Variation of heterotrophic production with influent strength of mesocosm wetlands treating woodwaste leachate  ....199  Figure 9-2. Variations of heterotrophic acetate assimilation and gross uptake with influent strength of mesocosm wetlands  200  Figure 9-3. Comparison of microbial structure between mesocosm wetland #1 fed with a strong influent and mesocosm wetland #4 fed with a weak influent  xv  202  LIST OF ABBREVIATIONS ATP  adenosine triphosphate  BOD  biochemical oxygen demand  COD  chemical oxygen demand  CTC  5-cyano-2,3-ditolyl tetrazolium chloride  CW  constructed wetland  DO  dissolved oxygen  DPM  radioactivity count in disintegration per minute  HRT  hydraulic retention time  LSD  least significant difference  MLVSS  mixed liquor volatile suspended solids  SD  standard deviation  TCA  trichloroacetic acid  ThOD  theoretical oxygen demand  TKN  total Kjeldahl nitrogen  T&L  tannin and lignin  TP  total phosphorus  TSS  total suspended solids  VFA  volatile fatty acids  xvi  ACKNOWLEDGEMENTS This research was supported by a Natural Science and Engineering Research Council grant to Ken Hall and Network of Centres of Excellence - Sustainable Forest Management funding to Sheldon Duff. The first thing I should mention is that this thesis could not have addressed so diverse mechanisms and factors without Dr. Ken Hall's intent of interest and continuous support. He input many valuable days to work with us in the odorous pilot-scale wetlands, made a wise decision to set up mesocosm wetlands on U B C campus at the right moment, and rarely missed a trip to Mission to collect leachate with me. I would also like to thank Dr. Ken Hall for his excellent academic guidance, laboratory assistance, constructive comments and suggestions on the draft manuscripts, and financial support. Drs. Susan Baldwin, Eric Hall, Don Mavinic and William Ramey are acknowledged for their invaluable supervision, advice, and comments on the draft thesis. Special thanks go to Eric Hall and William Ramey for their inspiration, discussion and assistance in some experiments. Anonymous reviewers assigned by several journal editors offered constructive comments, which improved the presentation of this thesis. The technical assistance of Susan Harper and Paula Parkinson in the U B C Environmental Engineering Laboratory is greatly appreciated. The chemical analyses made by Paula Parkinson are acknowledged. It is this laboratory that cultivated my interest in experimentbased research. I am grateful to the Radiation Protection Office of U B C for allowing me to use their liquid scintillation counter. Thanks to Dr. Tom Beatty for his permission and kind instructions to use his epifluorescence microscope. I would also like to express my sincere gratitude to Frederic Koch for his support in allowing me to set up mesocosm wetlands in the U B C Pilot Wastewater Treatment Plant as well as his advice and assistance in maintaining the mesocosm wetlands. B i l l Leung is recognized for building reliable mesocosm wetlands. Arash Masbough's cooperation and kindness in operating the pilot-scale constructed wetlands made our field trips fun. Many thanks also go to Jody Addah, Wayne Lo, Jordan Zhang, and the other kind people who helped me in field. Finally, I would like to say special thanks to my family for their patience and support in my years' pursuit of a doctoral degree, especially in those long lab days and field trips.  xvii  CO-AUTHORSHIP The following chapters are based on co-written manuscripts: •  Chapter 2: Wendong Tao conducted three years (2002-2004) of field measurement and laboratory analysis, did data analysis, drafted the manuscript, and made revisions. K . Hall proposed and supervised the field measurement  and laboratory analyses,  participated in field measurement, and made comments on the draft manuscript. K . Frankowski initiated the monitoring in 1999. A . Masbough conducted the field measurement and laboratory analysis in 2000 and 2001. S. Duff provided partial financial support to maintain the study site. •  Chapter 3: Wendong Tao proposed the hypotheses, designed the test programs, conducted the tests and laboratory analysis, did data analysis, and drafted the manuscript. K . Hall helped in experimental setup, and provided comments on the draft manuscript. E. Hall provided advice on experimental setup, discussed on the interim results, and commented on data analysis and the draft manuscript.  •  Chapter 4: Wendong Tao designed the study program, conducted the  field  measurement and laboratory analysis, did data analysis, drafted the manuscript, and made revisions. K . Hall proposed the intent, instructed the thesis author with radioisotopic techniques, participated in field measurement and sample preparation, and made comments on the draft manuscript. S. Duff provided partial financial support to maintain the study site. •  Chapter 5: Wendong Tao designed the study program, conducted the field sampling and laboratory bioassays, did data analysis, drafted the manuscript, and made revisions. K . Hall proposed the intent, instructed the thesis author with radioisotopic techniques, participated in sample preparation, and made comments on the draft manuscript.  •  Chapter 6: Wendong Tao proposed the hypotheses, designed the research program, conducted the field measurement and laboratory analysis, did data analysis, drafted the manuscript, and made revisions. K . Hall participated in setup and operation of the mesocosm wetlands, and provided comments on the draft manuscript. S. Duff provided partial financial support for leachate collection. xviii  Chapter 7: Wendong Tao proposed the hypotheses, designed the research program, conducted sample collection and laboratory analysis, did data analysis, and drafted the manuscript. K . Hall participated in setup and operation of the mesocosm wetlands, and provided comments on the draft manuscript. S. Duff provided partial financial support for leachate collection. Chapter 8: Wendong Tao proposed the hypotheses, designed the research program, conducted the field measurement and laboratory analysis, did data analysis, and drafted the manuscript. K . Hall participated in operation of the mesocosm wetlands, and provided comments on the draft manuscript. S. Duff provided partial financial support for leachate collection. Chapter 9: Wendong Tao proposed the hypotheses, designed the research program, conducted sample preparation and laboratory analysis, did data analysis, and drafted the manuscript. K . Hall provided comments on the draft manuscript. W. Ramey provided instructions on using microscopes and advice on microorganism enumeration, and commented on the draft manuscript.  xix  1 1.1  LITERATURE REVIEW AND RESEARCH SCOPE Constructed Wetlands for Water Quality Improvement  Wetlands used mainly for water quality improvement are known as treatment wetlands. Treatment wetlands could be natural wetlands or constructed wetlands (CWs). However, there was an increasing concern in the 1990s that unrestricted use of natural wetlands as receptacles for point and non-point sources of pollution would have an adverse effect on wetland integrity (USEPA, 1996). Since natural wetlands are protected more stringently from receiving wastewater in many parts of the world, CWs have to be established for wastewater treatment.  1.1.1  Definition and properties of natural wetlands  "Wetlands" are generally defined as those areas that are inundated or saturated by surface or ground water at a frequency and duration sufficient to support a prevalence of vegetation typically adapted for life under saturated soil conditions. Wetlands encompass a broad range of wet environments, such as swamps, marshes, and bogs (Figure 1-1). Since wetlands are transitional areas between land and water, the boundaries between wetlands and uplands or deep water are not always distinct. Nevertheless, wetlands feature specific hydroperiods, hydric soil, and hydrophytes (Mitsch and Gosselink, 2000; U S D A, 1995).  Figure 1-1. Appearance of natural wetlands: marsh (left), bog (middle), and swamp (right).  The hydrology of wetlands is generally a slow flow through shallow waters or saturated substrates. Water depth fluctuates frequently, although the substrate is saturated long enough during the growing season to create oxygen-poor conditions. Hydric soils are formed under conditions of repeated periods of saturation or inundation during the growing season. After 1  formation of hydric soils, the morphological features (e.g., mottling, and organic layers) tend to persist in the soil during both wet and dry periods, making them particularly useful in wetland identification. The plant communities of wetlands are dominated by hydrophytes, which are adapted for long or frequently recurrent periods of inundation or saturation. The hydrological regime largely determines the vegetation development. The vegetation in turn modifies the hydrological conditions by trapping sediment, interrupting water flows, and building peat deposits. Wetlands provide a wide array of functions that are directly related to their physical, chemical, and biological integrity (Mitsch and Gosselink, 2000). In addition to water quality improvement, the other common functions of wetlands include the provision of wildlife habitats, flood storage and peak flow alteration, groundwater recharge, sediment stabilization, carbon storage in the form of peat, and recreation in and on the water. Modification of existing wetlands to improve treatment capability is often very disruptive to the integrity of wetland ecosystems (USEPA, 1996). Natural wetlands are often afforded the same level of protection as other surface waters with regard to discharge standards. Greater interests have been shifted to CWs for wastewater treatment (Cole, 1998; USEPA, 2000).  1.1.2  Properties of constructed wetlands  CWs are always regulated as wastewater treatment facilities rather than "wetlands". CWs are generally designed to mimic the natural wetlands in overall structure while fostering those processes that are thought to contribute the most to the improvement of water quality. CWs consist of shallow ponds or channels, open or vegetated. They are typically constructed with uniform depths and regular shapes near the source of the wastewater and often in upland areas where no wetlands have historically existed (USEPA, 2000). CWs have impervious clay or synthetic liners and engineered structures to control the flow direction (Davis 1995; USEPA, 2000). Although a number df modifications and hybrids have been developed, CWs are classified by the literature and practitioners into two basic types (Kadlec and Knight, 1996; USEPA, 2000) as illustrated in Figure 1-2. Surface flow CWs (or free water surface wetlands) resemble natural wetlands in appearance, which typically consist of parallel cells or channels with relatively impermeable bottom soil, emergent vegetation, and shallow water depths of 0.1 to 0.6 m. Treatment occurs as water flows slowly above ground through the leaves and stems of emergent plants. Alternatively, 2  surface flow CWs may have floating and submergent plants. Subsurface flow CWs (also known as vegetated submerged beds, root zones, or rock-reed filters) contain channels or trenches of media (such as small stones, gravel, sand or soil) that may have been planted with aquatic plants. Studies have indicated that subsurface flow CWs without plants perform better than or as well as those with plants in many cases (e.g., Baptista et al., 2003; Liehr, 2000). Wastewater flows beneath the surface of the media. The porous media allow long-term subsurface flow without clogging.  Inlet Settling Zone  Zone 1 Fully Vegetated  Floating and Emergent Plants  Submerged Growth Plants  Zone 2 Open-Water Surface  Floating and Emergent Plants  Zone 3 FuBy Vegetated  Pretreated (Settled) Influent  1  •<ic£Jo° Liner  -Bottom Slope (£1%)  Figure 1-2. Illustration of surface flow constructed wetland (upper) and subsurface flow constructed wetland (bottom). From U S E P A (2000).  CWs typically rely on plants, soils and microorganisms to remove contaminants from wastewater by means of various physical, chemical and biological processes. Pretreated 3  wastewater is normally applied continuously to such systems. Surface flow CWs are normally used after ponds or lagoons because of their ability to handle the excess solids generated in the ponds. A surface flow C W may also be designed with consideration to create new wildlife habitats or enhance nearby existing natural wetlands, while offering its main objective of treatment. Subsurface flow CWs are more appropriately set behind a process designed to minimize suspended and settleable solids (USEPA, 2000).  1.1.3  Diverse applications of wetlands for wastewater treatment  Utilization of natural wetlands for wastewater disposal dated back to 1912 (Kadlec and Knight, 1996). Studies on treatment wetlands began in Europe in the 1950's, and in the US in the late 1960's (USEPA, 2000). It is only within the last 20 years that CWs have been utilized worldwide (Cole, 1998; Cooper and Findlater, 1990; Cooper et al., 1999; Haberl et a l , 1997; Hammer, 1989; Kadlec and Brix, 1995; Kadlec and Knight, 1996; Mbwette et al., 2003; Moshiri, 1993; Reddy and Kadlec, 2001). Cole (1998) estimated that there were over 500 wetland treatment systems in use in Europe and more than 600 in the U S . No new natural wetland treatment systems have begun since 1990. At least one-third of the natural wetland treatment systems included in the "North American Wetlands Treatment Systems Database" are no longer operating (USEPA, 2000). The Version 2 of the Database includes 40 natural wetlands and 205 CWs, of which 159 CWs are used to treat municipal wastewater. These wetlands are located in 34 US states and 6 Canadian provinces. The number of wetlands per state is probably more a function of having an advocate for treatment wetlands than climate or other favorable conditions (USEPA, 2000). Successful operation of CWs has been reported from the tropics (e.g., Braungart et al., 1997; Polprasert et al., 1996) to the Arctic (e.g., Jokela and Pinks, 1998; Pries, 1994). The surface flow wetland system in Listowel, Ontario is operated year-round with temperature as low as 3°C. The feasibility of operation through the winter depends on the water temperature in a C W and the treatment objectives. CWs have been used to treat stormwater runoff, municipal wastewater, industrial wastewater, agricultural wastewater, acid mine drainage, and landfill leachate. The contaminants being effectively removed in treatment wetlands range from conventional pollutants such as total suspended solids (TSS), chemical oxygen demand (COD), biochemical oxygen demand (BOD), nitrogen, phosphorus, and pathogens to such priority pollutants as metals, chlorinated organic compounds, and hydrocarbons. Recent reviews on treatment wetlands have been made 4  on treatment of stormwater runoff (Carleton et al., 2001), municipal wastewater (Knight, 2004; USEPA, 1999, 2000), and industrial wastewater (Knight, 2004). These reviews intended to derive kinetic models and statistical relationships for C W design. Although large cities, such as Phoenix, Arizona and Orange County, Florida, use CWs for wastewater treatment, wetlands have been utilized mainly by small- to medium-sized communities ranging from 5,000 to 50,000 in population (USEPA, 2000). Compared to wastewater treatment plants, CWs are low in cost, require only periodic on-site labor for maintenance, and provide a natural appearance (Kadlec and Knight, 1996; USEPA, 2000). Wetland treatment systems are especially suitable for small communities where inexpensive land is available and skilled operators are hard to find and keep. Along with ponds and vegetative biofilters, CWs are structural best management practices for non-point pollution control (USEPA, 2002). CWs may be used alone or in series with other treatment processes, depending on the treatment goals.  1.1.4  Treatment mechanisms of constructed wetlands  The processes occurring in C W systems encompass many of those used in wastewater treatment plants, such as flocculation and sedimentation, biological degradation, filtration, gas transfer and stripping, adsorption, ion exchange, chemical precipitation, chemical oxidation and reduction, and acid-base reactions (Kadlec and Knight, 1996; USEPA, 2000). There are also other processes  unique to natural treatment systems, such as photosynthesis,  photooxidation, and plant uptake. In natural treatment systems, the processes occur at 'natural' rates as opposed to wastewater treatment plants in which processes occur at accelerated rates with energy input (Metcalf & Eddy, 1991). The functional mechanisms of CWs are dependent on the wastewater constituents that are to be removed. The relatively low velocity and large surface area in subsurface flow CWs provide opportunities for TSS separation by sedimentation, filtration, and adsorption on biofilms. Subsurface flow CWs must be designed to minimize loss of infiltrative capacity due to clogging. In surface flow CWs, sedimentation is enhanced by very low flow velocity and shallow depth. Typically, sedimentation processes will remove material larger than about 50 p  with specific gravity of about 1.2 (USEPA, 2000). Some particles may be intercepted by  angular emergent plant tissue as would occur in settling basins equipped with plate or tube settlers (USEPA, 2000). The biofilm on plant surfaces and surface sediment in surface flow 5  CWs can adsorb colloidal and supracolloidal particles. However, there has been little work done on the evaluation of natural flocculation and interception-adhesion in surface flow CWs. Gearheart and Finney (1996) presented the only known application of the particle-size theory to show that colloidal fractions were flocculated in surface flow C W s M n addition; TSS are produced by C W processes, such as through death of invertebrates, fragmentation of detritus from plants, production of microorganisms, and formation of chemical precipitates. Resuspension of solids may occur primarily due to bioturbation, wind-induced turbulence, and gas-lift. The TSS of a wetland effluent is rarely from nonremovable TSS in the influent, and is often dictated by the wetland processes that generate TSS in the wetland (USEPA, 2000). Separation of particulate organic matter would occur by the same mechanisms as those for TSS. The settled and entrapped particulate organic matter is finally converted to soluble organic carbon by hydrolysis. Soluble organic polymers are hydrolyzed to organic monomers through exoenzymatic activity, and fermented further to labile organic compounds under anaerobic conditions. Soluble organic matter may be removed by adhesion to plant surfaces and sorption on sediment or bed medium. Many of the wetland solid surfaces are also renewed by continuous turnover of biomass that makes up the major component of the sorbent. The degree and rate of sorption and desorption are dependent on the characteristics of both the organic adsorbate and the adsorbent. Biodegradable organic matter is removed under either aerobic or anaerobic conditions (Metcalf & Eddy, 1991; USEPA, 2000; Westermann, 1993). Dissolved oxygen (DO) is an important factor controlling the rate of biochemical reactions. The water column of open CWs may be aerobic due to surface aeration. Due to gas transfer through the vascular tissues and diffusion of oxygen out of the root tissues, a thin layer of aerobic sediment forms around the rhizosphere (Bezbaruah and Zhang, 2005; Brix, 1997; Hammer, 1997). Sediment and water column of the CWs populated with dense emergent macrophytes are mainly anoxic. The predominant biological mechanism in subsurface flow CWs is likely to be facultative or anaerobic (USEPA, 2000). Volatilization may also account for loss of certain organics. Generally, organic matter entering a wetland following pretreatment will not contain significant quantities of volatile compounds (USEPA, 2000). However, some volatile organic compounds may be produced by biological processes. Some organic compounds in wastewater, such as tannin, lignin, humic substances, and various aromatic compounds, strongly absorb U V radiation ( A P H A et al., 1999). Recalcitrant organic  6  polymers may be photolyzed to labile compounds by solar radiation in several wavebands (Cole, 1999; Engelhaupt et al., 2003; Wetzel et al., 1995). Particle-associated organic nitrogen may be transported to the sediment with suspended solids. Recalcitrant organic nitrogen will accumulate and eventually become a part of soil humus. Some organic nitrogen is hydrolyzed to soluble amino acids that may undergo further breakdown to release ammonia. Soluble ammonia can be removed by volatilization. Volatilization of ammonia is relatively minor, <10% in CWs (Metcalf & Eddy, 1991), especially at a pH less than 7. Most of the converted ammonia in a natural system is adsorbed temporarily through ion exchange reactions on soil particles and charged organic particles. Adsorbed and free ammonia are available for uptake by plants and microorganisms and for biological nitrification under aerobic conditions. Nitrate is also a nutrient, which is used after ammonia. Plant uptake of nitrogen only occurs near the root zone during the growing season and then released in the fall and early spring. Plant uptake is a function of plant species and yield (Burgoon et al., 1991). Harvesting removes <20% of the influent nitrogen (Reed et al., 1995) at conventional loading rates. When a wetland treatment system has both aerobic and anaerobic zones, nitrification of ammonia to nitrate and denitrification of nitrate to nitrogen gas and nitrous oxide should occur in sequence. The denitrification rate in CWs is affected by nitrate availability, the concentration of organic substrates, temperature and pH (Gale et al., 1993; Kozub and Liehr, 1999; Toet et a l , 2003; Xue et al., 1999). Meanwhile, nitrogen gas may be fixed to organic nitrogen in water, sediment, rhizosphere of plants, and on the leaf and stem surfaces of plants (Reddy and Graetz, 1988). Phosphorus occurs in natural waters and wastewater primarily as phosphates. Particulate phosphate may be removed with TSS. The major processes for removal of soluble phosphate in natural treatment systems are chemical precipitation and adsorption onto the biofilms of plant surfaces and clay minerals of sediment, although plants do take up some phosphorus (Burgoon et al., 1991; Metcalf & Eddy, 1991; USEPA, 2000). Newly placed soils or media will have a greater phosphorus sorption capacity than a mature system that has most sorption sites already saturated. New plants growing in a freshly planted wetland will take up more phosphorus than a mature wetland, which will have phosphorus leaching from dying plants. Adsorbed phosphate could be released from the metal complexes, depending on the redox potential and pH of the sediment. Soluble organic phosphate, insoluble inorganic phosphate 7  and particulate organic phosphate are not usually available to plants until transformed to a soluble inorganic form by microorganisms. There is no important gaseous component in the biogeochemical cycle of phosphorus. The recalcitrant phosphate separated from the water column and accumulated as accreted sediment represents the total net removal.  1.2 1.2.1  Design Considerations for Surface Flow Constructed Wetlands Wetland layout and structures  Wetlands are likely to form where landform directs water to shallow basins. Wetland layout is dictated by the topography, geology, and land availability. The regional landscape should be integrated into the site selection process (Campbell and Ogden, 1999). On level C W sites, cells can be created with dikes. On sloping sites, cells can be terraced. C W cells have been configured in a number of shapes, including rectangles, polygons, ovals, kidney shapes, and crescent shapes. No data support one shape being superior to another shape in terms of effluent quality (USEPA, 1999). A site-sensitive design that incorporates existing features of the site reduces the amount of earthmoving required and increases the visual attractiveness of the site. Generally, C W cells are designed and built with an aspect ratio of 3:1 to 5:1 (USEPA, 2000). A higher aspect ratio is intended to minimize short-circuiting, and approach plug-flow hydraulics. Finger dikes are often used to create serpentine flow paths and mitigate shortcircuiting (Davis 1995; USEPA, 2000). Use of long, narrow channels, however, may lead to overloaded conditions at the inlet. Inlet is an open-end pipe, channel, or gated pipe that releases water into the wetland. Multiple inlets must be fully and independently adjustable to ensure an even distribution of flow. A n inlet settling zone is usually designed to retain setfieable particulates i f lacking pretreatment. Outlet structure can be a weir, spillway, or adjustable riser pipe. Water levels are controlled by the outlet. C W structures should be sized to handle maximum design flows. Multiple cells with appropriate piping between them offer greater operational flexibility. Each cell should be completely drainable for repairs. The number of cells depends on topography, hydrology, wastewater quality, and treatment targets (USEPA, 1999). A "sequential model" has been developed by Gearheart and Finney (1996). The sequential model recognizes that the treatment objectives beyond secondary require a minimum of three general wetland "compartments", an anaerobic vegetated zone for flocculation and sedimentation, an aerobic 8  open zone for soluble BOD reduction and nitrification, and an anaerobic vegetated zone for denitrification and further reduction in TSS and associated constituents.  1.2.2  Water depth and vegetation  The operating water depths for surface flow CWs in the North American Wetlands Database have ranged from 0.1 to over 2.0 m with a typical depth of 0.15-0.60 m. USEPA (2000) recommends water depths of 0.6-0.9 m in fully vegetated zones and 1.2-1.5 m in open-water zones in surface flow CWs treating municipal wastewater. In cold climates, the operating depth in the winter is normally increased to allow ice formation on the surface and to provide a longer hydraulic retention time (HRT) value (Metcalf & Eddy, 1991). The design water depth depends on the optimum depth for the selected vegetation. The water column with a depth greater than 1.2-1.5 m and planted with submergent plants will not be rapidly encroached by emergent plants. The water column of 0.5-1.0 m deep that is planted with a species of emergent plants will prevail over submergent plants (USEPA, 2000). The emergent plants commonly used in CWs include cattails, common reeds, rushes, bulrushes, and sedges. A l l of these plants are ubiquitous and tolerate freezing conditions. Cattails tend to dominate in water depths over 0.15 m; bulrushes grow well at depths of 0.05 to 0.25 m; reeds grow well in water up to 1.5 m, but are poor competitors in shallow waters; and sedges normally occur in waters shallower than that for bulrushes (Metcalf & Eddy, 1991). Cattail rhizomes and roots extend to a substrate depth of approximately 0.3 m, reeds extend to more than 0.6 m, and bulrushes to more than 0.8 m (Metcalf & Eddy, 1991). As research and application of CWs have expanded, documentation of actual performance differences between emergent plant species has become increasingly less valuable to C W designers (USEPA, 2000). The substrate for wetland vegetation should be agronomic in nature (e.g., loam), well loosened, and at least 15 cm deep (USEPA, 2000). Native species should be encouraged in planting, seedling, and other types of vegetative establishment for CWs. Plant communities in CWs undergo significant changes following initial planting. Several growing seasons may be needed to obtain a mature vegetative density (Mitsch, 1992; USEPA, 2000). In general, species diversity in CWs increases as ecosystems mature. Vegetation may play a variety of roles in surface flow CWs: >  provides surfaces for microbial attachment;  9  >  enhances flocculation and sedimentation;  >  acts as a supplementary carbon source for the microbial communities;  >  transfers oxygen into the vicinity of plant roots;  >  controls algal growth by shading;  >  reduces the heat-loss effects of wind; and  >  traps snow to insulate water surface from cold air for winter operation.  Vegetation may also have adverse impacts on water quality. Emergent and floating plants block atmospheric aeration. Several floating plants, especially duckweed and water hyacinth, have very high rates of primary production, which add large quantities of organic matter. The emergent plants with high lignin content remain standing as dead biomass that decays slowly after the growing season.  1.2.3  Hydraulic retention time and loading rate  A slow flow prolongs the exposure time for contaminants to be attacked. USEPA (2000) recommends a minimum H R T of 2 d for fully vegetated zones and a maximum HRT of 2-3 d for open water zones to avoid unwanted algal blooms. Precipitation and evapotranspiration have significant effects on C W water balance (Kadlec and Knight, 1996; Reed et al., 1995; USEPA, 2000). Tracer studies (Rash and Liehr, 1999) suggest that surface flow wetlands are not subject to short-circuiting, and the nominal H R T is usually a good estimate of the actual HRT. The nominal H R T of a surface flow C W can be expressed as: H R T = 2*Ve/(Qi+Qe)  0-1)  Ve = As*d*£  (1-2)  where As = surface area; Ve — effective water volume; Qi = inflow rate; Qe = outflow rate; d = water depth; and e = void fraction. 10  The vegetation and litter occupy a portion of the water column of a CW. The volume of the living plants ranges from 0.02 m m" to 0.10 m m" (USEPA, 2000). Kadlec and Knight 3  3  3  3  (1996) report that the average void fraction of treatment wetlands is usually greater than 0.95. A void fraction of 0.65 to 0.75 for fully vegetated zones is usually used in design calculations (Crites and Tchobanoglous, 1998; Reed et al., 1995; USEPA, 2000). The use of conservative values provides a factor of safety. In practice, H R T fluctuates due to rainfall events and temporal variation of evapotranspiration. The HRT required to achieve a specified level of efficiency is usually estimated by a kinetic or empirical model incorporating HRT (Carleton et al., 2001; Crites and Tchobanoglous, 1998; Kadlec and Knight, 1996; USEPA, 2000). Some wetland treatment systems show a more consistent correlation with area and hydraulic loading rate than with H R T (Davis 1995). This is reasonable for a shallow water system that receives energy inputs (sun light and gases) on an areal basis (Metcalf & Eddy, 1991). Hydraulic loading rate represents the depth of water distributed to the wetland surface over a specified time interval. Hydraulic loading rates used in practice range from 15 to 50 mm d" (USEPA, 1988). The hydraulic loading rate can be 1  used to estimate the land area requirement. The area requirement can also be estimated with the design HRT. A factor of 1.2-1.4 is usually used to convert the C W cell area to the total site area (USEPA, 2000).  1.2.4  Mass loading rate and oxygen transport  A n aerobic environment is highly desirable in wastewater treatment systems in which the target is effective removal of BOD. The design ultimate B O D loading rate should not exceed one-half the oxygen transfer rate (USEPA, 1988). U S E P A (2000) recommends a maximum B O D loading of 6 g m" d" to a surface flow C W receiving municipal wastewaters to 2  1  consistently attain an effluent B O D of <30 mg L" . U S E P A (1999) suggests a maximum B O D 1  loading rate of 5 g m" d" . A n upper B O D limit of 11 g m ' d" is recommended by WPCF 2  1  2  1  (1990). Care must be taken in using areal loading criteria because the actual load is not applied uniformly but concentrated near the inlets, whereas oxygen is usually supplied uniformly over the surface. For systems treating wastewaters with a significant fraction of settleable organic solids, the loading must be even lower, or distributed by step feeding to avoid overloading at the head of C W cells. In addition, an internal B O D load may be significant for polishing CWs. The internal loading occurs from the plant exudates and intermediates of anaerobic digestion 11  of the previously settled organic solids. The internal loading begins in the spring as water temperature rises, and continues until the backlog of settleable organics and plant detritus accumulated over the winter are exhausted (USEPA, 2000). The D O concentration of treatment wetlands is dependent on the rates of oxygen transfer and uptake. Oxygen transfer into surface flow CWs may be accomplished by vegetation transport, photosynthesis and surface aeration. The principal consumers of oxygen in wetlands include microorganisms that consume oxygen for respiration and plants that carry out respiration when sunlight energy is unavailable. The open water zones consistently exhibit higher DO with diurnal changes up to 10 mg L" (USEPA, 2000). In the vegetated zones, however, the 1  emergent stands along with floating plants and litter provide a heavy canopy that obscures surface aeration and photosynthesis. Increased oxygen transfer on a system-wide basis can likely be achieved by using alternating vegetated and open-water cells. Rates of oxygen transport through the leaves and stems to the rhizomes and roots of emergent plants range from 0 to 3 g m" d" (Brix, 1990; U S E P A , 2000). Brix and Schierup (1990) and Brix (1990) 2  1  found that the oxygen transported by Phragmites almost exactly balanced the respiratory demand of the belowground plant tissues, leaving only 0.02 g m" d" to be released to the 2  1  surrounding soil. In addition, any oxygen "leak" to the rhizosphere will likely be consumed by the large benthic oxygen demand that normally exists in wetlands (Metcalf & Eddy, 1991). The role of emergent plants on oxygen transfer to the wetland is a subject of controversy.  1.3  Characterization of Microorganisms in Aquatic Systems  Microorganisms within an aquatic system may include bacteria, algae, and protozoa. Heterotrophic bacteria decompose particulate and soluble organic matter and regenerate nutrients. Algae fix inorganic carbon into aquatic system. Protozoa in aquatic systems include heterotrophic flagellates and ciliates that feed on bacteria, algae, and detritus. Bacteria, algae, and protozoa form microbial food webs, along which carbon is mineralized or transformed. Distribution among the wetland components (water, epiphyton and sediment) and community dynamics of microbial biomass, abundance, production rate, and mineralization rate unveil the biological treatment mechanism in CWs. Methods for microbiological examination have been continuously introduced to aquatic systems, covering microscopy as well as radioisotopic, fluorogenic, and molecular techniques. 12  A l l of the methods, however, possess their respective applicability and limitations for environmental samples. Cautions have to be taken to choose appropriate methods for specific CWs. Many microbiological methods are originally established for laboratory cultures, and can be easily adapted for water samples from aquatic system. Special treatment is usually required for sediment and epiphyton samples. More than one method could be used to cross check, but change of methods might cause problems in interpreting the data consistently.  1.3.1  Microbial diversity of aquatic systems  The microorganisms in a wetland are composed of variable, mixed species established during long residence time under variable substrate and ambient conditions. Many acidic environments like most wetlands are well populated by bacteria, algae, fungi and protozoa (Kushner, 1993; Westermann, 1993). Some acidophiles are limited to acid conditions, while others can grow over wide p H ranges. Table 1-1 gives the pH range for growth of representative acidophiles. The internal p H of such bacteria is near neutrality (Kushner, 1993). The biochemical adaptations necessary for such abilities usually involve the cell envelope, especially the cytoplasmic membrane that controls ion transport.  Table 1-1. Examples of acidophiles in natural environments Organisms Bacteria  Fungi  Algae  Protozoa  pH for growth Thiobacillus thioooxidans  0.9-4.5  Thiobacillus ferrooxidans  1.5-4.0  Thermoplasma acidophilum  1.0-4.0  Bacillus acidocaldarius  2.0-6.0  Sulfolobus acidocaldarius  0.9-5.8  Aspergillus/Penicillium/Fusarium spp.  2-10  Acontium velatium/Cephalosporium  2.0-7.0  Chlorella pyrenoidosum  2-10  Cyanidium caldarium  <2-5 1.4-9.6  Polytomella caeca  Source: Kushner, 1993.  13  The vertical heterogeneity in the water column caused by physical and biological processes, such as solar radiation, surface aeration, and algal growth, may increase the diversity of heterotrophic bacteria in wetlands (Figure 1-3). The complete oxidation of organic carbon to C O 2 can be carried out by aerobic bacteria in the oxic surface layer. Under anoxic conditions, organic particles are hydrolyzed to soluble organic matter by exoenzymatic activity of fermentative bacteria. A major fraction of soluble organic matter consists of compounds with a high molecular weight, such as proteins, polysaccharides, nucleic acids, and humic substances, which are not directly available to microheterotrophs (Chrost, 1989). The high-molecularweight organic matter has to be further hydrolyzed enzymatically into low-molecular-weight compounds, such as glucose and amino acids. Fermentative bacteria are facultative or strict anaerobes that decompose polymers, monomers, and oligomers.  H 0  O, Particulate organic carbon, soluble organic polymers  Hydrolysis Aerobic bacteria arid fungi  Monomers  2  Aerobic bacteria  NO,  *C0  2  N,  Hydrolysis Exoenzymes  ->co  2  Denitrifying bacteria  Mn'4+ Oligomers, monomers  Mn 2+  Oxidation  •»co  2  -»co  2  *co  2  -»co  2  Manganese reducing bacteria  Fe,3+ Fermentative Fermentation bacteria  Fe,2+  Iron reducing bacteria  so 2  Fatty acids, alcohols, H , C 0 , etc. 2  2  4  Oxidation  Sulfate reducing bacteria  CO, Methanogens  Figure 1-3. Pathways in the degradation of organic matter in wetlands. Adapted from Westermann (1993).  14  The metabolic activity yielding the highest amount of energy from a common electron donor will dominate i f the electron acceptor is present. The thermodynamic criteria result in a sequential reduction of O2, N O 3 , M n , Fe , SO 4 and H C O 3 as the redox potential decreases + 4  +3  (Zehnder and Stumm, 1988). Different bacterial groups are responsible for anaerobic oxidation of soluble organic compounds (Pedros-Alio et al., 1993), depending on redox potential and the presence of terminal electron acceptors. Spatial overlap of respiration processes using different electron acceptors occurs in microbial mats (Canfield and Des Marais, 1991). Nevertheless, few studies (Flood et al., 1999) have been carried out on the phylogenetic identification of bacteria in surface flow CWs. Respiration with nitrate as electron acceptor is generally initiated when DO is <0.3 mg L"  1  (Westermann, 1993). The bacteria responsible for denitrification are facultative anaerobes, such as Pseudomonas species. Manganese reduction from M n  3 +  and M n  4 +  oxides linked to  oxidation of organic substrates occurs under aerobic and anaerobic conditions (Lovley, 1991; Lovley and Phillips, 1988). Respiration with M n as electron acceptor is carried out by a wide range of aerobic and facultatively anaerobic bacteria and microfungi (Ghiorse, 1988). Below the zone of denitrification and manganese reduction, iron reduction might take place, utilizing ferric iron as an electron acceptor. Many different microorganisms from aerobic, facultatively anaerobic and obligately anaerobic bacterial groups are able to perform ferric reduction. Studies on waterlogged soil (Komatsu et al., 1978) and eutrophic lake water (Jones et al., 1983) have suggested that denitrifying bacteria are mainly responsible for observed iron reduction. Sulfate reduction occurs in the strata below or microniches within the more oxidized zones at redox potentials below -120 mV (Connell and Patrick, 1968; Widdel, 1988). Sulfate-reducing bacteria are strict anaerobes that use organic acids and alcohols as electron donors and sulfate as terminal electron acceptor. Sulfate-reducing bacteria grow slowly because H S , the respiration product, decreases the growth rate. In most CWs, the 2  concentration of inorganic electron acceptors relative to available carbon is low, resulting mainly in methanogenesis (Westermann, 1993). Methanogenesis is performed by a specialized group of obligately anaerobic archaebacteria. Heterotrophic methanogens split acetate and formate into CO2 and CH4. Humus is composed of darkly colored amorphous high-molecular-weight organic substances resulting from the biotransformation and repolymerization of phenolic and aromatic 15  components in litter, such as lignin, tannins and secondary metabolites. Humus and quinone analogues have a distinct role as electron acceptors for anoxic respiration of a variety of substrates (Field et al., 2000). Acetate respiration and fermentation of glucose and lactate have been shown to include humic acid and quinone or anthraquinone disulfonate as electron acceptors. After heterotrophic bacteria, heterotrophic flagellates are frequently the most abundant group of organisms observed in biofilms (Hunt and Parry, 1998). The abundance of protozoa (2-200 um in size) is generally greater in sediment, water-sediment interfaces, submerged surfaces, and organically enriched sites, than in water column (Baldock and Sleigh, 1988; Bott and Kaplan, 1989; Gasol, 1993). The factors that affect the number of protozoa in freshwater sediment include sediment porosity, texture, organic content and C:N:P ratios, redox potential and temperature (Gasol, 1993). Bacterial abundance may significantly decrease at plentiful food supply due to grazing (del Giorgio et al., 1996). Grazing on bacteria may speed recycling of nutrients, thus promoting organic matter mineralization (Gasol, 1993). Nanoflagellates selectively crop the larger and actively growing cells leaving the slow growing or dormant cells behind (del Giorgio et al., 1996), causing a large proportion of inactive population. Selective protistan grazing has been shown to influence mean per-cell growth of the active bacteria (Pernthaler et al., 1996). Algae are ubiquitous in wet habitats. Algae in open zones can form a canopy that blocks sunlight from penetrating the water column to submerged vegetation, resulting in reduced DO levels. The presence of open water near the outlet of a surface flow C W typically promotes seasonal blooms of algal species, which result in elevated concentrations of TSS and particulate nutrient forms in the effluent (USEPA, 2000).  1.3.2  Spatial distribution of microorganisms in wetlands and biofilm collection  Bacteria in surface flow CWs may inhabit the water column, submerged surfaces of plants and litter, and surface sediment. Attention has been paid to the role of bacterioplankton, epiphyton, or sedimentary bacteria in the decomposition of organic matter in aquatic environments over the last three decades. Few studies have evaluated the relative importance of the water column, sediment and plant surfaces within a given wetland. Moran and Hodson (1992) found that sediment supported the bulk of bacterial secondary production, while the remainder was contributed approximately equally by bacteria in the water column and on detritus in a 16  freshwater marsh. In macrophyte beds receiving piggery effluent, the major fraction of bacterial uptake for acetate and glucose was associated with the bacterioplankton (Toerien and Toerien, 1985). In surface flow CWs receiving a secondary sewage effluent, the denitrification rate was considerably higher in epiphyton than in water and sediment (Toet et al., 2003). There is a dynamic balance among planktonic, epiphytic, and sedimentary microorganisms (Caldwell, 1987). Microbial attachment is favored in a bulk fluid that has a chemical composition of sufficiently dilute nutrients such that suspended growth is not possible (White et al., 1999). A n attached form of growth is beneficial to the organisms, especially under low nutrient conditions (Mueller, 1996). Nutrient levels may be higher in the surfaces than the bulk fluid due to the surface-associated organic material (Davis and McFeters, 1988; LeChevallier and McFeters, 1990). The movement of water along solid-liquid interfaces provides microorganisms with continuous supply of substrates and removes the inhibitory metabolic wastes as they are produced (Caldwell, 1987). Attached growth can also enhance cell survival with respect to biocides or other chemical stressors (Chen et al., 1993; LeChevallier and McFeters, 1990; van der Wende et al., 1989). Investigations into the structure of a bedrock epilithic community using scanning electron microscopy (Stock and Ward, 1989) showed the existence of a rather loosely attached assemblage of microorganisms overlying a more firmly attached layer. In addition, the firmly attached component was not homogeneous but had a complex and varied structure. Heterogeneity in organism distribution and metabolic activities is common within the biofilms of attached communities (White et al., 1999). This property of biofilm structure requires a serious consideration of sampling techniques. Biofilms can be collected by removing parts of the system carrying the attached assemblage (Fischer and Pusch, 1999; Flood et al., 1999; Pollard et al., 1995) or scratching the assemblage from the attached surfaces. Thin layers can be seen i f a white tissue is used for wiping (Schaule et al., 2000). Utilization of a removable, smooth synthetic substratum ("coupon") facilitates reproducible collection of a homogeneous biofilm over time and at different systems. The size of the coupons generally ranges from a few square centimeters to about twenty square centimeters. Except for the rectangular coupons, some substrata are used in the form of discs (Hunt and Parry, 1998; Hunt et al., 1999; Jones and Lock, 1991). The coupons are immersed in a given microbial habitat, left for a period of microbial colonization, and then retrieved for examination. 17  Microbial colonization on synthetic substrata in aquatic systems is dependent on surface roughness (Harmsworth and Sleigh, 1993; Hunt and Parry, 1998) and material type of substrata (Fletcher and Loeb, 1979; Harmsworth and Sleigh, 1993; Hunt and Parry, 1998). Several kinds of synthetic substrata have been used in biofilm studies. The commonly used substrata are glass slides, high-density polyethylene, polyvinyl chloride, and stainless steel or copper. Biofilm samples have also been collected on sandpaper (Hunt and Parry, 1998; Hunt et al., 1999), glass beads ( A G H T M , 1999), rock (Muller et al., 2000), and membranes (Jones and Lock, 1991). Glass beads do not readily facilitate bacterial attachment ( A G H T M , 1999), but they are inexpensive, and easy to disinfect and clean, and have a very smooth surface. The concentration of biofilm A T P was similar on copper and stainless steel and 3 times higher on cross-linked polyethylene during recirculation of tap water at 25-35°C (van der Kooij et al., 2005). In contrast to copper and stainless steel, plastic (hardened polyethylene and polyvinyl) coupons were colonized by bacteria in higher population densities in water distribution systems (Schwartz et al., 2000). Sandpaper with a given roughness (e.g., 800 grit) maximizes reproducibility in surface roughness in comparison to hand-roughened surfaces (Hunt and Parry, 1998). In addition, the roughness of sandpaper  can be classified using the  manufacturer's classification system. Long-term exposure experiments will determine i f the differences in material-related biofilm growth are only a time-dependent phenomenon in biofilm formation, rather than for mature biofilm (Hunt and Parry, 1998; Schwartz et al., 2000). Studies (Hunt and Parry, 1998; McCormick et al., 1988; Oemke and Burton, 1986; Shamsudin and Sleigh, 1994; Stewart et al., 1985; Stock and Ward, 1989) show that after 1014 d exposure of a synthetic substratum in aquatic environments, initial colonization often reaches a mature biofilm community. The properties of a mature biofilm can fluctuate with sloughing and protozoan grazing (Hunt and Parry, 1998).  1.3.3  Determination of microbial biomass and enumeration  Microbial biomass has largely been estimated by various techniques of determining numbers, cell volume, and biomarker concentrations. Many methods have been developed for water ( A P H A et al., 1999; Kemp et al., 1993; Metcalf & Eddy, 2003), biofilm (Melo et al., 1992) and sediment samples (Kirka et al., 2004). Sediment has the advantage over water in that bacterial populations are more numerous. The disadvantage of sediment is the humic materials and soil particles in and on which the bacteria may inhabit, restricting the applicability of 18  many methods to sediment samples. Moreover, soil is heterogeneous, and contains diverse microhabitats (Kirka et al., 2004), resulting in a high variability in microbial diversity and population size. Classical microbiological identification methods are based on dispersion and cultivation of the organisms in agar or suspension media. These methods essentially give the relative abundance of different physiological types of heterotrophic microorganisms, but often underestimate microbial populations in environmental samples (Hobbie and Ford, 1993; Schaule et al., 2000; Kirka et al., 2004) because all media are selective to a lesser or greater extent. Furthermore, the bacteria that grow in culture are not necessarily the ones that are active in the environment, and bio volume is not directly available with these methods. Flow cytometry is ideal for rapid characterization of populations and cell size distribution of suspended bacteria by direct light scatter, staining, or fluorescent labeling. However, the data obtained from flow cytometers are not often verifiable by other methods. It faces the interference of detritus. Comparing to the epifluorescence microscopy, flow cytometry has limited resolution. Most flow cytometers analyze very small volumes (<0.5 mL) so that cells occurring at less than 1000 mL" cannot be conveniently analyzed (Button and Robertson, 1  1993). Microscopic techniques have long been practiced to enumerate microorganisms. Image analysis systems can be employed in conjunction with microscopy to reduce the tedium and subjectivity for enumeration. Cell biovolume can be determined with a phase contrast microscope based on dimensions of different morphological types. Biomass can be estimated from the number and biovolume. The factors used to convert cell numbers and biovolume to biomass may vary with the morphology and physiology of microbial cells (Fry, 1990; Lee and Fuhrman, 1987; Norland, 1993). Direct enumeration with traditional light microscopy has the inherent difficulty of discriminating viable cells from detritus and dead cells as well as bacteria from microalgae and nanoflagellates. It may not accurately count cells within tight microbial clusters. Epifluorescence microscopy is considered the most reliable method for the evaluation of bacterial community dynamics. Fixed bacteria are stained with a fluorescent dye and collected on a membrane filter. The fluorescence distinguishes living bacteria from dead bacteria, organic detritus, protozoa, and autotrophs under an epifluorescence microscope set at certain 19  excitation and cutoff wavelengths. Numerous fluorescing dyes are available, but acridine orange and 4,6-diamidino-2-phenyl indole (DAPI) are the two commonly used dyes. A l l of the fluorochromes differ somewhat in their ability and specificity to stain. The most appropriate dye for a particular sample is to some degree a matter of trial and error. DAPI is preferred for most environmental samples with a high particulate loading because of lower nonspecific background staining. Fluorescently stained bacteria in water samples can be easily counted by epifluorescence microscopy (Porter and Feig, 1980). Epiphyton and sediment samples can be treated by a combination of chemical and physical procedures to disperse bacteria from their attached sites or aggregated forms (Fischer and Pusch, 1999; Kaplan and Bott, 1989; Velji and Albright,  1993), followed by the  standard  epifluorescence  microscopic method for  enumeration of dispersed bacteria. Fluorescent staining, however, does not distinguish metabolically active bacteria from dormant and senescent bacteria. Some metabolic indicators are nonfluorescent, while being readily reduced via electron transport activity to insoluble fluorescent formazans. When a metabolic indicator, such as 5cyano-2,3-ditolyl tetrazolium chloride (CTC) in an oxidized.form (Rodriguez et al., 1992; Schaule et al., 1993) and 2-(p-iodophenyl)-3-(p-nitrophenyl)-5-phenyl tetrazolium chloride (Bott and Kaplan, 1985; Johnson and Ward, 1993), is added to fresh samples and incubated in the dark for a certain period, the fluorescent formazans are deposited intracellularly. The formazans can be visualized in wet-mount preparations with black polycarbonate membrane filters (Rodriguez et al., 1992) or in biofilms associated with optically opaque surfaces (Fischer and Push, 1999) under an epifluorescence microscope. The fluorescence of the formazans provides an indication to distinguish active (respiring) from inactive (non-respiring) bacteria. CTC is more suitable for samples with elevated levels of non-biological material or in thick biofilm (Schaule et al., 2000). A n incubation of >2 h with C T C at a final concentration of 2-6 m M at 28°C resulted in optimal formazan deposition for P. putida (Rodriguez et al., 1992). A n incubation of water and biofilm samples in drinking water with 0.5 m M C T C for one hour at 23°C was sufficient to obtain intracellular reduction of C T C (Schaule et al., 1993). Shorter reduction time would reduce the likelihood of cell division and decrease the overall assay time. The optimum C T C concentration and incubation time should be verified and adjusted according to the microbial community structure and aquatic environments. Controls should be prepared to correct abiotic reduction of the indicators for low-redox samples. In strongly reduced samples with redox potential as low as -200 mV, 2-(p-iodophenyl)-3-(p20  nitrophenyl)-5-phenyl tetrazolium chloride is reduced abiotically to formazan. CTC is reduced abiotically from -220 mV (Schaule et al., 2000). Protozoa could be enumerated  by the epifluorescence microscopic method at high  concentrations of fluorescent dyes (Sherr et al., 1993). However, some major problems have been realized regarding preservation and filtration of protozoan samples (Gasol, 1993; SimeNgando et al., 1990; Sherr et al., 1993), such as lysis and loss of naked flagellate and ciliate cells that are sensitive to filtration, decrease in cell biovolume and distortion of cell shape upon fixation. Quantification of protozoa is more easily accomplished by direct live counting with a phase contrast microscope (Gasol, 1993; Massana and Gude, 1991; Sime-Ngando et al., 1990). Flagellates and ciliates can be easily detected by their movement and size. Molecular methods are being introduced to environmental samples for phylogenetic identification, enumeration, and biomass estimation. In the nucleic acid probe method (Madigan et al., 1997), a nucleic acid probe is synthesized that is complementary to a D N A or R N A sequence unique to the targeted organisms. The probe is then labeled with a radioisotope ( S or P ) , a fluorescent dye, or an enzyme. The labeled probe is hybridized to the nucleic 35  32  acid extracted from bacterial cells. The hybridized probes can be detected on a sheet of X-ray film sensitive to the type of label for dot blot hybridization, under a fluorescence microscope (DeLong et al., 1989) or with flow cytometry (Amann et al., 1990) for fluorescent in situ hybridization, or by autoradiography with a scintillation counter or photographic film for radio-labeled hybridization (Kemp et al., 1993). Highly specific probes capable of differentiating organisms within a phylogenetic domain are increasing. One limitation of the in situ hybridization method for environmental samples is the lack of sensitivity (Kirka et al., 2004). Polymerase chain reaction is a rapidly improving molecular technique. It involves amplification of the D N A of the microorganisms being tested by using oligonucleotide primers complementary to sequences in the genes of interest (Madigan et a l , 1997). The primer triggers a reaction that results in the production of billions of copies of the target D N A . The resulting products can be separated in different ways (Dorigo et al, 2005; Kirka et al., 2004). It requires that the nucleotide sequence of a portion of the desired gene be known. Real-time P C R makes the quantitative determination of D N A and R N A much more accurate (Dorigo et al., 2005). Nevertheless, there are problems associated with the application of this 21  technique to environmental samples, including the small sample volumes that can be assayed, variable lysis efficiency, variation among D N A extraction and cleanup procedures, bias due to differential amplification, inhibition by interfering constituents, and inability to differentiate between viable and inactivated microorganisms (Kirka et al., 2004; Metcalf & Eddy, 2003). Depending on the molecular technique used, the information gained can be quite different (Dorigo et al, 2005). Application of immunological methods to microbial ecology (Madigan et al., 1997) involves labeling an antibody with a fluorescent dye, such as fluorescein and rhodamine. Once the fluorescent antibody has found the organism in question, it becomes attached to a target protein. The sample can be examined optically to detect the target organism against a background of non-target organisms. The utility of immunological techniques in aquatic microbiology was demonstrated by the early work of Ward (1982) on marine nitrifying bacteria. In combination with epifluorescence microscopy or flow cytometry, immunological methods enable classification to major taxonomic division and perhaps species. Antibodies to specific enzymes have been used to enumerate microorganisms containing target enzymes, e.g., nitrogenase (Currin et al., 1990). The heterogeneous distribution of cell sizes and morphologies and the variable cell biovolume-to-biomass conversion factors complicate methods using cell number and biovolume to estimate biomass. Adenosine triphosphate (ATP) is ubiquitous in all living cells, has a relatively short half-life following cell death and autolysis, and is present at a constant intracellular concentration regardless of nutritional mode. Cellular A T P can be rapidly and efficiently extracted from living cells and stabilized in boiling buffers, cold mineral or organic acids, or a variety of organic solvents (Karl, 1993), varying in their efficiency and simplicity. Realizing the thermal gradients around sediment particles in a hot extracting agent and immediate inactivation of transphosphorylating enzymes, cold extraction may be superior to hot extraction for sediment samples. The ability of sediment to adsorb and retain large amounts of inorganic phosphate is controlled i f the adsorbing sites are saturated with added phosphate (Bulleid, 1978). A very small amount of A T P can even be measured by bioluminescent reactions with reduced luciferin, luciferase, M g , and oxygen in excess. The 2 +  light emission is directly proportional to the concentration of A T P in solution. The initial rise (in 0-2 s), the peak height (in 0-5 s), or a predetermined integrated portion (e.g., 15-75 s) of 22  the light emission decay curve can be used to relate A T P concentrations in standards to those in the unknown sample extracts. Phospholipid fatty acids concentration in environmental samples is proportional to microbial biomass, and signature fatty acids can differentiate major taxonomic groups within a community (Steenbergen et a l , 1993; Kirka et al., 2004). Phospholipid fatty acids can be extracted and quantified with excellent sensitivity to determine both composition and biomass of microbial populations in water, soil and sediment (Fredrickson et al., 1986; Zelles and Bai, 1993). However, interpretation of phospholipid fatty acids profiles is still in progress, although data on functional groups (e.g., sulfate-reducing bacteria and denitrifiers) are rapidly becoming available. The method is suitable for relatively high concentrations of biomass as encountered in sediment and biological reactors (Tunlid and White, 1990; Werker and Hall, 2000). With the limited knowledge to date, it would not be possible to ensure which microbial groups are changing in population. Different protocols of microbial fatty acids extraction could derive distinct levels of the absolute amounts of microbial fatty acids (Werker and Hall, 2000). Algal biomass can be distinguished from the total biomass derived from ATP measurements. Ash-free dry mass, biovolume measurement and pigment analysis are three of the commonly used approaches to measure algal biomass. The gravimetric method is simple, but leads to underestimation i f there is a large fraction of non-algal organic material. Enumeration and volume measurement of algal cells provides a direct, but labor-intensive method. Both ashfree dry mass and biovolume measurement do not distinguish between live and dead cells. The concentration of photosynthetic pigments is used extensively to estimate phytoplankton and phytoperiphyton biomass. Chlorophyll-a occurs in greatest abundance in aquatic environments (Wetzel and Likens, 2000). It can be extracted in acetone solution, and determined by standard methods with spectrophotometric, fluorometric, or high-performance liquid chromatographic techniques ( A P H A et al., 1999). Fluorometry is more sensitive than spectrophotometry, and requires less sample volume.  1.3.4  Determination of heterotrophic bacterial production  Unlike traditional wastewater treatment processes, aquatic systems usually have low concentrations of heterotrophic bacteria or volatile suspended solids in water. The rate of bacterial production in aquatic systems is usually estimated by: 23  >  determining the increase of bacterial number, biomass, or turbidity during an incubation of isolated samples or in continuous culture systems at the in situ temperature (Kemp et al., 1993; Madigan et al., 1997); or  >  tracking bacterial incorporation of radio-labeled substrates when a certain volume of a sample is incubated in isolated samples for a definite time in the dark at the in situ temperature (Kemp et al., 1993; Ward and Johnson, 1996; Wetzel and Likens, 2000).  Enumeration and biomass determination are usually time-consuming. Biovolume and generation time are likely to change during a long course of incubation. Turbidity determination may be subject to interferences for colored samples. Radio-labeled substrate incorporation techniques have been commonly used for determining heterotrophic bacterial production in natural environments that have less than usual bacterial abundance. Radioactive tracers are sensitive enough to follow heterotrophic production of the microorganisms in natural environments. Radio-labeled substrates used for determination of bacterial production include H-leucine, thymidine, adenine, etc. 3  When a labeled substrate is incubated with fresh samples under in situ conditions, the microorganisms retain their original reproduction rate for a limited period (Kaplan et al., 1992). Bacteria incorporate most of the labeled leucine into protein, and thymidine into D N A . The leucine content of bacterial protein remains constant, and D N A labeling represents a nearly constant proportion of the labeled macromolecules. Cyanobacteria, algae, and fungi generally lack transport mechanism needed to assimilate thymidine (Wetzel and Likens, 2000). This character restricts thymidine uptake in a short incubation to D N A synthesis to the heterotrophic bacteria. The bacteria in natural aquatic systems apparently maximize their utilization of extracellular leucine for protein synthesis, rather than synthesizing leucine de novo (Kirchman et al., 1985). Most of the assimilated leucine (78-88%) is incorporated into bacterial proteins (Jorgensen, 1992). As a whole, thymidine and leucine incorporation rates reflect bacterial biomass production in aquatic systems. After incubation, epiphyton samples are usually sonicated for minutes to detach bacteria (Fischer and Pusch, 1999; Hunt and Parry, 1998; Schaule et al., 2000; Thomaz and Wetzel, 1995). The labeled protein or D N A is extracted, collected on membrane filters, and assayed for the radioactivity incorporated. Killed controls are incubated along with live subsamples to correct abiotic adsorption of radioisotopes on membrane filters. The rates of nucleotide ( H 3  24  thymidine) incorporation into bacterial D N A and amino acids ( H-leucine or H-amino acids) incorporation into bacterial proteins are determined to estimate heterotrophic bacterial production. Heterotrophic productivity is derived by incubating replicates at a series of increasing substrate concentrations (Simon and Azam, 1989; Wetzel and Likens, 2000). A single incubation at the saturation concentration is also recommended (Jorgensen, 1992; Ward and Johnson, 1996) to derive productivity. A high extracellular concentration of leucine inhibits de novo synthesis of leucine (Kirchman et al., 1985), minimizing intracellular isotope dilution due to leucine biosynthesis. A high concentration of added leucine also reduces the risk of external isotope dilution. A number of procedures have been used to extract D N A labeling from H-thymidine 3  incorporation (Torreton and Bouvy, 1991; Kaplan et al., 1992). Usually, thymidine-labeled D N A is extracted with 5-20% (final concentration) ice-cold trichloroacetic acid (TCA) solution (Fuhrman and Azam, 1982; Jorgensen, 1992; Simon and Azam, 1989). Protein is extracted with 5% T C A at 85-100°C for 30 min followed by 5% cold T C A washing (e.g., Fischer and Pusch, 1999; Moran and Hodson, 1992; Simon and Azam, 1989; Thomaz and Wetzel, 1995) or 5% cold T C A for 15 min followed by 5% T C A washing (e.g., Kirchman, 1992; Wicks and Robarts, 1988). TCA-insoluble extract may contain peptidoglycan, lipid, and some complex polysaccharides in addition to protein (Hollibaugh and Wong, 1992; Kirchman et a l , 1985; Wicks and Robarts, 1988). Cold ethanol (80%) washing may follow T C A washing to remove non-protein extracts. Some studies (e.g., Chin-Leo and Kirchman, 1988; Jorgensen, 1992; Riemann and Azam, 1992) found no significant difference between hot and cold T C A extractions. The extracted protein can be collected on membrane filters with a pore size of 0.2 um or 0.45 p , Hollibaugh and Wong (1992) found large differences in the amount of H retained on the controls between Millipore cellulose acetate filters as used by Jorgensen 3  (1992) and Simon and Azam (1989) and Nuclepore polycarbonate filters as used by Thomaz and Wetzel (1995). Labeling of D N A by H-thymidine may not be tenable in some ecosystems (Hollibaugh, 1988; 3  Riemann et al., 1982; Robarts et al., 1986; Torreton and Bouvy, 1991). Some bacteria, such as acetate-utilizing and sulfate-reducing bacteria, are unable to incorporate thymidine (Gilmour et al., 1990; Pollard and Moriarty, 1984). The amount of thymidine incorporated is converted to cell number increase by a factor of approximately 2 x l 0 25  18  cells mol" thymidine (Wetzel and 1  Likens, 2000). To estimate carbon production, however, one needs to know the per-cell carbon content that varies with cell size. H-leucine is often chosen over other amino acids and 3  thymidine (Simon and Azam, 1989). Leucine incorporation is one order of magnitude more sensitive than thymidine, because over time bacterial cells incorporate about 10 times more leucine than thymidine. The leucine content of bacterial protein remains more constant than other amino acids.  Bacterial protein production could be converted to bacterial carbon  production based on a reasonably predictable relationship between cell carbon and protein contents (Simon and Azam, 1989). Nevertheless, some anaerobic bacteria do not take up appreciable amounts of soluble leucine (Ward and Johnson, 1996). More than 50% of the bacterial population takes up leucine in natural aquatic environments (Kirchman et al., 1985). A study (Tabor and Neihof, 1982) has demonstrated that 45-76% of the total microscopic direct counts assimilated H-amino acids in the Chesapeake Bay water. A fraction of leucine 3  taken up is respired (Jorgensen, 1992), or transformed to other amino acids (Kirchman et al., 1985; Monticello and Costilow, 1982; Simon and Azam, 1989) or H 0 (Jorgensen, 1992). 3  2  1.3.5  Determination of heterotrophic bacterial activity on organic matter  The bacteria in aquatic systems include those groups that hydrolytically decompose organic polymers and aerobically or anaerobically mineralize labile organic compounds. Heterotrophic activity is measured as the rate of a specific metabolic reaction, such as oxygen consumption, utilization of an organic substrate or electron acceptor, heterotrophic C O 2 assimilation, formation of gases and intermediate degradation products, and exoenzymatic activity (Hesselsoe et al., 2005; Kemp et al., 1993; Sorokin and Kadota, 1972; Wetzel and Likens, 2000). Cautions have to be taken to select a reaction representing the naturally occurring process. The determination of oxygen consumption and gas producing rates usually requires a long closed incubation, during which the microbial community structure and the availability of substrates may change. Measuring the heterotrophic assimilation and mineralization of a radio-labeled organic substrate has been the most common approach to  determine  heterotrophic bacterial activity of the bacterial communities in aquatic environments. The assimilation and mineralization rates may vary with the kinds of substrates. The overall heterotrophic activity for organic carbon removal is still difficult to measure directly. 14  C-glucose and C-acetate are the most frequently used substrate tracers. A radio-labeled 14  organic substrate is added to a fresh sample in a closed system. After a period of dark 26  incubation at the in situ temperatures, the radio-labeled bacterial cells and respiration products ( CC»2) are collected separately, and assayed for the radioactivities assimilated and respired. 14  Killed controls are incubated to correct abiotic adsorption of radioisotopes on membrane filters. It is important to keep incubation time short and demonstrate the linearity of bacterial response over time. After incubation, epiphyton samples are usually sonicated for minutes to detach bacteria as for determination of production rate. Heterotrophic bacterial activity is expressed as substrate assimilation and mineralization rates that are derived from the radioactivities of bacterial particulates and CG"2 respectively. Heterotrophic uptake of C 14  1 4  substrates is diluted by the extracellular pools of substrates present in natural water (Chrost, 1989; King and Berman, 1984; King and Klug, 1982). When a sample with an unknown natural substrate concentration is incubated in a series of replicates with  different  concentrations of the labeled substrate, maximum substrate uptake rates are evaluated with the Michaelis-Menten enzyme kinetic model or one of its modifications (Wetzel and Likens, 2000; Wright and Hobbie, 1966). If the natural substrate concentration can be measured independently, in situ uptake rates can be determined by incubation with a given amount of the labeled substrate (Cavari et al., 1978; Chidthaisong et al., 1999). The radioisotopic technique has been extensively used to measure the potential heterotrophic activity in water and sediment of freshwater systems. Limitations of this technique include intracellular isotope dilution (King and Berman, 1984) and assumption of complete mineralization. The anaerobic degradation products may include CIL; in addition to CC»2. 14  l4  The relative yields of CH4 and CC»2 from incubation vary with the tracer and incubation 14  14  medium (Federle and Schwab, 1992; King and Klug, 1982; Lovley and Klug, 1982). Some intermediate decomposition compounds may be released in a soluble form other than CO2 (Chidthaisong et a l , 1999; King and Klug, 1982; Lovley and Klug, 1982; Sawyer and King, 1993; Zinder, 1986). A recent study (Kong et al., 2005) found that two morphotypes of polyphosphate-accumulating organisms did not take up short-chain fatty acids (e.g., acetate) and glucose. Extracellular enzyme activity indicates the capacity of the prevailing bacterial community for hydrolysis of C, N and P polymeric compounds (Marx et al., 2001; Hakulinen et al., 2005). Fluorometric assay is commonly used to determine exoenzymatic activity in aquatic environments. When a fluorogenic model substrate is added to a fresh sample and incubated 27  under in situ conditions, the nonfluorescent substrate is hydrolyzed enzymatically to fluorescent product 4-methylumbelliferone (MUF) or 7-amino-4-methylcoumarin (AMC). A model substrate is usually composed of a fluorescent tracer molecule and an organic molecule that are linked by a specific binding. Fluorescence is observed after enzymatic splitting of the non-fluorescent MUF-substrate to fluorescent M U F . The hydrolytic activity of a specific extracellular enzyme is derived from the fluorescence increase. Incubation time may vary as long as fluorescence increases are linear, depending on the nature of the samples, temperature, substrate type and sensitivity of the fluorometer. Any fluorometer may be used for fluorescence measurement at an excitation of 355-390 nm and an emission of 440-460 nm (Chrost, 1989; Hakulinen et al., 2005; Hoppe, 1993; Radl et al., 2005). A microplate fluorometer is used for microplate fluorometric assays (Hakulinen et al., 2005; Marx et al., 2001; Radl et al., 2005). In order to derive enzyme kinetics based on the Michaelis-Menten equation, final model substrate concentrations should cover a sufficient range (Chrost, 1991; Hoppe, 1993; Marx et al., 2001). Application of a single, high concentration (Hakulinen et al., 2005; Radl et al., 2005) is recommended, only if measurements of relative activity are desired. Hydrolysis of model substrates is competitively inhibited by a variety of natural compounds with the same structural characteristics, so only the maximum velocity of hydrolysis can be measured by adding high or increasing quantities of model substrate to samples (Hoppe, 1993). The model substrate to use may be selected according to the targeted type of enzymes. Model substrates have been applied to aquatic systems to determine the activities of phosphatase (Chrost and Overbeck, 1987; Marxsen and Schmidt, 1993), phosphomonoesterase (Hakulinen et al., 2005; Radl et al., 2005), sulfatase (Chappell and Goulder, 1992; Hakulinen et al., 2005), P-D-galacotosidase (Chrost and Krambeck, 1986), aminopeptidase (Hakulinen et al., 2005; Hoppe et al., 1988; Radl et al., 2005), acetate-esterase (Hakulinen et al., 2005), butyrateesterase (Hakulinen et al., 2005), P-cellobiosidase (Hakulinen et al., 2005), p-glucosidase (Hakulinen et al., 2005; Radl et al., 2005), p-D-glucosidase (Chrost, 1989; Somville, 1984), ctglucosidase (Hakulinen et al., 2005), and P-xylosidase (Hakulinen et al., 2005). In order to cover the diverse exoenzymes in aquatic systems, up to 11 enzymatic activities have been determined simultaneously in lake sediment (Hakulinen et al., 2005).  28  1.4 1.4.1  Background and Research Scope Development of technology for treatment of wood leachate  In regions where forestry is one of the major industrial sectors, collection and treatment of wood leachate are a challenging task. Wood leachate may be generated when rainfall, runoff or sprinkling water percolates through woodpiles, log yards, wood product storage sites, and woodwaste disposal sites. Usually, wood leachate is dark, acidic, of very high oxygen demand, and toxic to aquatic organisms (Bailey et al., 1999b; Field et al., 1988; Frankowski, 2000; Hunter et al., 1993; Masbough, 2002; Peters et al., 1976; Taylor and Carmichael, 2003; Taylor et al., 1996; Woodhouse and Duff, 2004; Zenaitis et al., 2002). The organic carbon is a mixture of volatile fatty acids (VFAs), tannin and lignin (T&L), and other soluble organic matter. The causes of toxicity have not been fully investigated, but the toxic effects were usually attributed to tannins, lignins, lignans, terpenes, tropolones, resin acids, metals, oxygen depletion, or low pH. The acute toxicity (96-h LC50 down to <1% leachate) and sub-acute toxic effect to fish (Atwater, 1980; Bailey et al., 1999a, b; Buchanan et al., 1976; Field et a l , 1988; Frankowski, 2000; Pease, 1974; Peters, 1974; Peters et al., 1976; Taylor et al., 1996; Temmink et al., 1989), invertebrates (Buchanan et al., 1976; Hensel, 1990; Peters et al., 1976; Taylor et al., 1996) and algae (Taylor et al., 1996), as well as the high oxygen demand (up to 14 g L" COD) justify the necessity for removal of both biodegradable organic matter and 1  recalcitrant, toxic compounds from wood leachate. Meanwhile, the low pH (down to 4), recalcitrance and toxicity to bacteria (Field et al., 1988; Taylor and Carmichael, 2003; Taylor et al., 1996; Woodhouse and Duff, 2004; Zenaitis et al., 2002) pose an adverse environment for biological treatment. Developing a treatment technology suitable for wood leachate has received very little attention. Cedar extracts are metabolized or bound by contact with soil during overland flows (Atwater, 1980). Logyard runoff has been efficiently treated by batch aerobic reactors (Zenaitis et al., 2002) and a trickling filter (Woodhouse and Duff, 2004). Ozonation significantly reduces toxicity and T & L from logyard runoff (Zenaitis and Duff, 2002). Surface flow vegetated CWs and open CWs (or ponds) have become cost-effective alternatives for onsite treatment of wood leachate (Frankowski, 2000; Hunter et a l , 1993; Masbough, 2002). CWs are able to accept varying quantities and concentrations of wastewater, are inexpensive to construct, and are easy to maintain with low energy and labor requirements. The feasibility 29  for treatment of wood leachate in CWs was first examined by Hunter et al. (1993) in a smallscale open channel and an adjoining vegetated channel with a H R T of 10 d and an influent C O D of 1024 mg L " . With aerobic microcosm wetlands, Frankowski (2000) proved the 1  applicability of CWs for reduction of COD, T & L and toxicity from woodwaste leachate. In two years of operation of pilot-scale CWs, Masbough (2002) confirmed the effectiveness of CWs in reduction of C O D , B O D , T & L , and V F A s from woodwaste leachate. A batch test with aspen leachate in a vat without aeration (Taylor et al., 1996) achieved 4.1% COD and 89% V F A s removal in 65 d. Taylor and Carmichael (2003) found a B O D decline over 10-12 d of storing aspen leachate in a leachate catch basin. Further studies are needed to derive the design parameters for CWs to treat wood leachate.  1.4.2  Current situation of treatment wetland design  CWs have been applied for water quality improvement for decades all over the world. Efficient treatment of a variety of wastewaters has been reported anecdotally. Despite the diverse applications and mechanisms conceivable for water quality improvement in the literature, the optimal design of CWs has not yet been determined. Furthermore, U S E P A (2000) indicated four misconceptions commonly existing in the application, design and performance assessment of CWs: >  Misconception #1—Wetland design has been well-characterized by published design equations;  >  Misconception #2—CWs have aerobic as well as anaerobic treatment zones;  >  Misconception #3—CWs can remove significant amounts of nitrogen;  >  Misconception #4—CWs can remove significant amounts of phosphorus.  Misconception #2 is attributed to extrapolation of treatment mechanisms that function in many larger polishing wetlands and natural wetlands, which account for most of the earlier applications of wetlands for water quality improvement. The oxygen demand from wastewater in CWs for secondary treatment is usually far more than the oxygen "leaks" (Brix, 1997; Brix and Schierup, 1990; Hammer 1997; Otte et al. 1995) from the roots of emergent plants to the surrounding soils. Field experience and recent studies with microelectrodes (e.g., Bezbaruah and Zhang, 2005) have shown that the small amount of oxygen leaked from plant roots is insignificant compared to the oxygen demand. Misconception #3 is related to misconception 30  #2 since co-existence of aerobic and anaerobic zones in a C W would make nitrification and denitrification significant mechanisms for nitrogen removal. Misconception #4 mainly stems from earlier studies with monitoring over a short period of initial operation instead of a long period covering initial phosphorus adsorption on soil, uptake of plants during a growing season, and release during senescence of plants. For the polishing wetlands that receive wastewater with a low phosphorus concentration, a small amount of mass removal could result in large removal efficiency. Misconception #1 actually refers to the earlier design guidelines and texts (Kadlec and Knight, 1996; Metcalf & Eddy, 1991; Reed et al. 1995; T VA , 1993; U S D A , 1995; USEPA, 1988; WPCF, 1990) that established models and suggested design criteria based mainly on aerobic treatment of wastewater or derived from short periods of performance data. Treatment efficacy could be attributed to several physical, chemical and biological processes (Section 1.1.4) occurring simultaneously. Design values are mainly derived in an empirical approach, considering a wetland cell as a black box to extrapolate the influent-effluent relationships of existing CWs or aggregating all the treatment mechanisms to one first-order reaction. However, performance has varied among the systems that have been monitored and it is usually difficult to derive design parameters from the performance data of existing CWs since the effects of diverse factors are always mingled (Carleton et al., 2001; USEPA, 2000). Moreover, most of the CWs are used for treating municipal wastewater and purifying agricultural runoff (Cole, 1998; USEPA, 2000). Woodwaste leachate is different from municipal wastewater and agricultural runoff, especially in organic strength and composition. The performance could not be extrapolated to CWs for treatment of wood leachate. Different mechanisms may act on different types of contaminants as described in Section 1.1.4. Organic contaminants are removed or transformed by a combination of physical, chemical, and biological mechanisms in CWs. Figure 1-4 illustrates the potential mechanisms for organic carbon removal through surface flow CWs receiving an influent of high organic strength like wood leachate. Heterotrophic bacteria play the most important role in the ultimate removal of organic matter in CWs. Heterotrophic bacteria may either suspend in the water column or attach on sediment, litter and plant surfaces. Bacteria and algae may be grazed by flagellates and ciliates, and the organic carbon dissipates along the aquatic food webs (Kisand and Noges, 2004; Steenbergen et al., 1993; Wetzel, 2001). To improve C W 31  design and operation for treatment of a specific wastewater, the functional mechanisms and the regulating factors should be addressed. A great kinetic-based alternative of C W design was to extend one overall reaction rate to both suspended and attached growth for biological removal of organic carbon (Polprasert and Agrawalla, 1994; Polprasert and Khatiwada, 1998; Polprasert et a l , 1998).  Influent  Water  Sediment  Figure 1-4. Carbon flow in a surface flow constructed wetland treating high-organic-strength wastewater. Dotted arrow = particulate organic carbon (POC); open arrow = soluble organic carbon (DOC).  The predominant mechanisms are dependent on the external input parameters, internal interactions, and the wetland characteristics (USEPA, 2000). The external input parameters of concern include influent quality, mass loading rate, and hydraulic loading rate or HRT. The rates of the internal processes shown in Figure 1-4 and their spatial distribution are affected by the environmental conditions. Performance of surface flow CWs is usually reported through short-term or long-term monitoring of the influent and effluent. Fewer studies have investigated the internal treatment mechanisms for organic carbon removal. Taylor et al. (1996) found that aeration facilitated degradation over 65 d of aspen leachate stored in vats. Flood et al. (1999) and Pollard et al. (1995) assessed bacterial abundance, growth rate, respiratory activity and phylogenetic composition of the biofilm communities of surface flow wetlands receiving secondary sewage effluent. Toerien and Toerien (1985) investigated the relative contributions of water, macrophyte surfaces, settled sludge, and sediment to heterotrophic activity of the macrophyte beds receiving piggery effluent. Polprasert et al. 32  (1998) compared the organic carbon removal rates by bacteria in water and sediment of surface flow CWs. Further quantification of the mechanisms and their regulating factors under different operating conditions will contribute to kinetic-based C W modeling.  1.4.3  Research objectives and methodology  With regard to the limited knowledge in the technology for wood leachate treatment and the kinetics of surface flow CWs, this study took an adaptive approach to examine the potential mechanisms and their regulating factors for treatment of woodwaste leachate in surface flow CWs. Because of the difficulty in separating the effects of individual mechanisms on-site, laboratory tests on volatilization, adsorption and biological treatability were first done to evaluate the major mechanisms for organic carbon removal from woodwaste leachate through surface flow CWs. This study was then focused on the major mechanism, biological degradation, for treatment of woodwaste leachate in surface flow CWs. Although wetlands are typically taken as black boxes while investigating the treatment effectiveness of CWs, assumptions on internal processes have been made to interpret their performance. The overall objective of the research with CWs was to promote kinetic-based C W design by bridging design parameters and treatment performance with microbial community structure and heterotrophic activities inside wetland cells. The initial objectives during studies with both the pilot-scale and mesocosm wetlands were to assess the spatial and temporal variations of the internal microbial processes so that the following experimental programs were simplified. Meanwhile, attention was paid to the integrity of CWs, which consist of three components, i.e., water column, epiphyton and sediment. The pilot-scale CWs provided a site to evaluate the overall performance under actual operating conditions, e.g., variable influent quality, temperature and rainfall events. Figure 1-5 summarizes the setup of the pilot-scale surface flow CWs to test the following research hypotheses: >  Hypothesis 1—Microbial colonization develops over time;  >  Hypothesis 2—Microbial communities in CWs are dynamic in heterotrophic activity;  >  Hypothesis 3—There are longitudinal variations in heterotrophic activities because the hydraulics of CWs lie somewhere between the plug-flow and complete-mix conditions;  33  Hypothesis 4—The insignificant effect of vegetation on treatment of woodwaste leachate in CWs is due to the limited contribution of the bacteria attached on plant surfaces to heterotrophic activity; and Hypothesis  5—Amendment  of nutrients  and electron acceptors  can  stimulate  heterotrophic activity, and hence improve treatment performance of CWs.  Pilot wetland #6: HRT>14 d  T  2001 Summer and Fall •  Microbial colonization progress in acetate uptake (hypothesis 1);  •  Long-term dynamics in leucine incorporation, acetate uptake and glucose turnover (hypothesis 2);  •  Longitudinal variations in leucine incorporation, acetate uptake and glucose turnover (hypothesis 3).  Pilot wetland #4: HRT= 12d  Y  2002 Summer and Fall •  Effects of vegetation and fertilization on leucine incorporation, acetate uptake, and  Pilot wetland #5: HRT= 13 d  glucose turnover (hypotheses 4 & 5); —CS]—T>  NH NO 4  / J  Pilot wetland #6: HRT==14 d  T  •  Effects of vegetation and fertilization on treatment performance (hypotheses 4 & 5).  Figure 1-5. Summary of research hypotheses tested and pertinent setup of pilot-scale vegetated (T) and open constructed wetlands.  The mesocosm wetlands were more controllable than the pilot-scale CWs so that effects of individual design parameters could be examined. Figure 1-6 summarizes the setup of the surface flow mesocosm wetlands to test the following hypotheses: >  Hypothesis 6 — C W microbial communities maturate over time;  >  Hypothesis 7—There is a vertical gradient in microbial biomass due to vertical variations of water temperature, D O and redox potential in CWs; 34  >  Hypothesis 8—HRT affects treatment performance and development of microbial communities in CWs;  >  Hypothesis 9—Woodwaste leachate inhibits heterotrophic activities and biological treatment is less efficient when wetlands are fed with a strong influent at high mass loading rates;  >  Hypothesis 10—Protozoan grazing controls the abundance of active bacteria;  >  Hypothesis 11—Vegetation may play a greater role for treatment of VFAs-poor woodwaste leachate than for treatment of V F As-rich leachate; and  >  Hypothesis 12—Recalcitrant contaminants can be removed gradually by bacteria established in CWs at long HRTs.  Based on the effects of the above design parameters and natural factors tested in the CWs, this research identified the factors that regulate organic carbon removal from woodwaste leachate through microbial degradation in surface flow CWs. The relative contributions of the wetland components were evaluated under different operating conditions. Simultaneous employment of both geochemical and biochemical approaches made it possible to bridge the external input parameters to effluent quality with microbial metabolisms inside the CWs. Determination of cellular A T P provided a rapid estimation of the total biomass of all living microorganisms. Chlorophyll-a was measured to distinguish algal biomass from the total microbial biomass. Epifluorescence microscopy was used in the later stage of this study to enumerate respiring bacteria. Phase-contrast microscopy was used in the later stage of this study to count protozoa since protozoan grazing seemed to pose a pressure on bacterial biomass and activity. The community structure in biomass of algae, bacteria and protozoa demonstrated the microbial food webs in surface flow CWs. Incorporation of H-leucine into 3  bacterial protein was determined to assess bacterial production rate. Assimilation and mineralization of  14  C-acetate were tracked to represent the heterotrophic activity on  degradation of labile organic compounds. Turnover of C-glucose was determined in the 14  pilot-scale wetlands to represent the overall heterotrophic activity on decomposition of soluble organic compounds. Utilization of these techniques was expected to limit the bias inherent in using any one technique, and to provide a complete profile of the microbial communities in the treatment wetlands. 35  Flow 1  Y  Mesocosm* 1: HRT==5(5)d  •  Flow 2  [~J  2003 Summer and Fall  »l Mesocosm#2: HRT=9(9)d  (hypothesis 6);  Flow 3  Y  Mesocosm#3: HRT=16(14)d Flow 4  Initial microbial maturation in cellular ATP  •  Vertical variations in ATP, temperature, DO, and redox potential (hypothesis 7);  Mesocosm#4: HRT=25(19) d  Y  •  Effects of HRT on treatment performance, A T P , leucine incorporation, and acetate uptake (hypothesis 8); and  •  Effects of mass loading rate on treatment performance, ATP, leucine incorporation, and acetate uptake (hypothesis 9).  Mesocosm#l: HRT= =13 d  Mesocosm#2: HRT= 13d  Y Y  2004 Summer and Fall •  Effects of influent strength on treatment performance, ATP, leucine incorporation,  Mesocosm#3: HRT= =i3 d  Y  Mesocosm#4: HRT= o d  Y  . Strength 2 Mesocosm#l:HRT= -12 d  Y  and acetate uptake (hypothesis 9).  2004 Fall Effects of influent strength on abundances  Mesocosm#2: HRT=29 d  l-O-H  Y  of protozoa and active bacteria (hypothesis 10);  Mesocosm#3: HRT=29 d  Effects of vegetation on treatment of old -Strength' Mesocosm#4: HRT= 12 d  woodwaste leachate (hypothesis 11); and  Y  •  Treatment performance of wetland trains (hypothesis 12).  Figure 1-6. Summary of research hypotheses tested and pertinent setup of vegetated (Y) and open-water mesocosm wetlands fed at different flow rates, with different strengths of influent, or operated in series.  36  The following chapters of this thesis are presented in a manuscript format: >  Chapter 2 characterizes the woodwaste leachate with six years of monitoring data.  >  Chapter 3 presents the laboratory tests on volatilization, adsorption and biological treatability and discusses the potential mechanisms for treatment of woodwaste leachate in CWs.  >  Chapters 4 and 5 summarize the operating conditions and treatment performance of the pilot-scale CWs in 2001 and 2002, evaluate the spatial and temporal variations of heterotrophic activities, assess the effects of vegetation and fertilization on treatment performance and heterotrophic activities, and discuss the relative importance of the three wetland components.  >  Chapter 6 summarizes the operating conditions of the mesocosm wetlands in 2003, examines the maturation of microbial communities, discusses the  appropriate  parameters for performance evaluation, and evaluates the variations of treatment performance with H R T and mass loading rate. >  Chapter 7 examines the responses of heterotrophic activities to H R T and mass loading rate of the mesocosm wetlands in 2003, and correlates heterotrophic activities with treatment performance.  >  Chapter 8 summarizes the operating conditions of the mesocosm wetlands in 2004, and evaluates the variations of treatment performance with influent strength and vegetation.  >  Chapter 9 examines the effects of influent strength on heterotrophic activities in the mesocosm wetlands in 2004, and discusses the impacts of protozoa grazing on heterotrophic activities and respiratory activity of bacteria.  >  Chapter 10 concludes by summarizing the treatment performance and microbial activities over years, extracting general findings on the major mechanisms and regulating factors, and suggesting design considerations for practical application of CWs to treatment of woodwaste leachate. Major issues worthy of further studies for treatment of woodwaste leachate in CWs are recommended.  37  1.5 References A G H T M . 1999. Standard method to evaluate aquatic biofilms. In: C.W. Keevil, A . Godfree, D. Holt, and C. Dow (eds.), Biofilms in the Aquatic Environment. The Royal Society of Chemistry: Cambridge, U K . pp210-219. Amann, R.I., B . J . Binder, R.J. Olson, S.W. Chisholm, R. Devereux, and D.A. Stahl. 1990. Combination of 16S rRNA-targeted oligonucleotide probes with flow cytometry for analyzing mixed microbial populations. Applied and Environmental Microbiology, 56: 1919-1925. A P H A et al. 1999. Standard Methods for the Examination of Water and Wastewater, 20th ed. American Public Health Association, American Water Works Association, and Water Environment Federation. Atwater, J.W. 1980. Fraser River Estuary Study, Water Quality: Impact of Landfills. Government of Canada and Province of British Columbia, Vancouver, B C , Canada. Bailey, H.C., J.R. Elphick, A . Potter, and B . Zak. 1999a. Zinc toxicity in stormwater runoff from sawmills in British Columbia. Water Research, 33: 2721-2725. Bailey, H.C., J.R. Elphick, A . Potter, E. Chao, D. Konasewich, and J.B. Zak. 1999b. Causes of toxicity in stormwater runoff from sawmills. Environmental Toxicology and Chemistry, 18: 1485-1491. Baldock, B . M . , and M . A . Sleigh. 1988. The ecology of benthic protozoa in rivers: seasonal variation in numerical abundance in fine sediments. Archiv fur Hydrobiologie, 111: 409-421. Baptista, J.D.C., T. Donnelly, D. Rayne, and R.J. Davenport. 2003. Microbial mechanisms of carbon removal in subsurface flow wetlands. Water Science and Technology, 48: 127-134. Bezbaruah, A . N . , and T. C. Zhang. 2005. Quantification of oxygen release by bulrush (Scirpus validus) roots in a constructed treatment wetland. Biotechnology and Bioengineering, 89: 308318. Bott, T.L., and L . A . Kaplan. 1985. Bacterial biomass, metabolic state, and activity in stream sediments: relation to environmental variables and multiple assay comparison. Applied and Environmental Microbiology, 50: 508-522.  38  Bott, T.L., L . A . Kaplan. 1989. Densities of benthic protozoa and nematodes in a Piedmont stream. Journal of the North American Benthological Society, 8: 187-196. Braungart, M . , K . Hansen, and D . Mulhall. 1997. Biomass nutrient recycling: an affordable closed-cycle process with added benefits. Water Environment and Technology, 9: 41-45. Brix, H . 1990. Gas exchange through the soil-atmosphere interface and through dead culms of Phragmites Australis in a constructed reed bed receiving domestic sewage. Water Research, 24: 259-266. Brix, H . 1997. Do macrophytes play a role in constructed treatment wetlands? Water Science and Technology, 35: 11-17. Brix, H . , and H . Schierup. 1990. Soil oxygenation in constructed reed beds: the role of macrophyte and soil-atmosphere interface oxygen transport. In: P.F. Cooper and B . C . Findlater (eds.), Constructed Wetlands in Water Pollution Control. Pergamon Press: New York, N Y . pp53-65. Buchanan, D.V., P.S. Tate, and J.R. Moring. 1976. Acute toxicities of spruce and hemlock bark extracts to some estuarine organisms in southeastern Alaska. Journal of the Fisheries Research Board of Canada, 33: 1188-1192. Bulleid, N.C.. 1978. A n improved method for the extraction of adenosine triphosphate from marine sediment and seawater. Limnology & Oceanography, 23: 174-178. Burgoon, P.S., K.R. Reddy, T.A. DeBusk, and B. Koopman. 1991. Vegetated submerged beds with artificial substrates. II: N and P removal. Journal ASCE-EED,  117: 408-424.  Button, D.K., and B.R. Robertson. 1993. Use of high-resolution flow cytometry to determine the activity and distribution of aquatic bacteria. In: P.F. Kemp, B.F. Sherr, E.B. Sherr, and J.J. Cole (eds.), Handbook of Methods in Aquatic Microbial Ecology. Lewis Publishers: Boca Raton, FL.ppl63-174. Caldwell, D.E. 1987. Microbial colonization of solid-liquid interfaces. Annals of the New York Academy of Sciences, 506: 275-280. Campbell, C.S., and M . H . Ogden. 1999. Constructed Wetlands in the Sustainable Landscape. John Wiley & Sons: New York, N Y .  39  Canfield, D.E., and D.J. Des Marais. 1991. Aerobic sulfate reduction in microbial mats. Science, 251: 1471-1473. Carleton, J.N., T.J. Grizzard, A . N . Godrej, and H.E. Post, 2001. Factors affecting the performance of stormwater treatment wetlands. Water Research, 35: 1552-1562. Cavari, B.Z., G. Phelps, and O. Hadas. 1978. Glucose concentrations and heterotrophic activity in Lake Kinneret. Verhandlungen der Internationalen Vereinigungfur Limnologie, 20: 2249-2254. Chappell, K.R., and R. Goulder. 1992. Epilithic extracellular enzyme activity in acid and calcareous headstreams. Archivfur Hydrobiologie, 125: 129-141. Chen, C.-L, T. Griebe, R. Srinivasan, and P.S. Stewart. 1993. Effects of various metal substrata on accumulation of Pseudomonas aeruginosa biofilms and the efficacy of monochloramine as a biocide. Biofouling, 7: 241-251. Chidthaisong, A., B . Rosenstock, and R. Conrad. 1999. Measurement of monosaccharides and conversion of glucose to acetate in anoxic rice field soil. Applied and Environmental Microbiology, 65: 2350-2355. Chin-Leo, G., and D.L. Kirchman. 1988. Estimating bacterial production in marine waters from the simultaneous incorporation of thymidine and leucine. Applied and Environmental Microbiology, 54: 1934-1939. Chrost, R.J. 1989. Characterization and significance of (3-glucosidase activity in lake water. Limnology and Oceanography, 34: 660-672. Chrost, J. R. 1991. Environmental control of the synthesis and activity of aquatic microbial exoenzymes. In: R.J. Chrost (ed.), Microbial Enzymes in Aquatic Environments. SpringerVerlag: New York, N Y . pp29-59. Chrost, R.J., and H.J. Krambeck. 1986. Fluorescence correction for measurements of enzyme activity in natural waters using methylumbelliferyl-substrates. Archiv fur Hydrobiologie, 106: 79-89. Chrost, R.J., and J. Overbeck. 1987. Kinetics of alkaline phosphatase and phosphorus availability for phytoplankton and bacterioplankton in lake PluPsee (north German eutrophic lake). Microbial Ecology, 13: 229-248. 40  Cole, S. 1998. The emergence of treatment wetlands. Environmental Science and Technology, 32:218-223. Cole, J.J. 1999. Aquatic microbiology for ecosystem scientists: new and recycled paradigms in ecological microbiology. Ecosystems, 2: 215-225. Connell, W.E., and W.H. Jr Patrick. 1968. Sulfate reduction in soil: effects of redox potential and pH. Science, 159: 86-87. Cooper, P.F., and B . C . Findlater (eds.). 1990. Constructed Wetlands in Water Pollution Control. Pergamon Press: New York, N Y . Cooper, P., and Members of the Scientific Committee (eds.). 1999. Proceedings of International Conference on Wetland Systems for Water Pollution Control, held in Eguasde Sao Pedro, Brazil, 27 September - 2 October 1998. Crites, R.W., and G. Tchobanoglous. 1998. Small and Decentralized Wastewater Management Systems. WCB~McGraw-Hill: New York, N Y . Currin, C.A., H.W. Paerl, G.K. Suba, and R.S. Alberte. 1990. Immunofluorescence detection and characterization of N2-fixing microorganisms from aquatic environments. Limnology and Oceanography, 35: 59-71. Davis, L . 1995. A Handbook of Constructed Wetlands Volume I - General Considerations. U S D A - N R C S and USEPA-Region III: Washington, D C . Davis, D.G., and A . McFeters. 1988. Growth and comparative physiology of Klebsiella oxytoca attached to granular carbon particles and in liquid media. Microbial Ecology, 15: 165175. del Giorgio, P. A., J.M. Gasol, D. Vaque, S. Agusti, and C M . Duarte. 1996. Bacterioplankton community structure: protists control net production and the proportion of active bacteria in a coastal marine community. Limnology and Oceanography, 41: 1169-1179. DeLong, E.F., G.S. Wickham, and N.R. Pace. 1989. Phylogenetic stains: ribosomal R N A based probes for the identification of single cells. Science, 243: 1360-1363. Dorigo, U . , L . Volatier, J.-F. Humbert. 2005. Molecular approaches to the assessment of biodiversity in aquatic microbial communities. Water Research, 39: 2207-2218.  41  Engelhaupt,  E.,  T.S. Bianchi,  R.G. Wetzel,  and  M . A . Tarr.  2003.  Photochemical  transformations and bacterial utilization of high-molecular-weight dissolved organic carbon in a southern Louisiana tidal stream (Bayou Trepagnier). Biogeochemistry, 62: 39-58. Federle, T.W., and B.S. Schwab. 1992. Mineralization of surfactants in anaerobic sediments of a Laundromat wastewater pond. Water Research, 26: 123-127. Field, J.A., M.J.H. Leyendeckers, R.S. Alvarez, G . Lettinga, and L . H . A . Habets. 1988. The methanogenic toxicity of bark tannins and the anaerobic biodegradability of water soluble bark matter. Water Science and Technology, 20: 219-240. Field, J.A., F.J. Cervantes, F.P. van der Zee, and G . Lettinga. 2000. Role of quinines in the biodegradation of priority pollutants: a review. Water Science and Technology, 42: 215-222. Fischer, FL, and M . Pusch. 1999. Use of the [ C]leucine incorporation technique to measure 14  bacterial production in river sediments and the epiphyton. Applied and Environmental Microbiology, 65: 4411-4418. Fletcher, M . , and G.I. Loeb. 1979. Influence of substratum characteristics on the attachment of a marine pseudomonad to solid surfaces. Applied and Environmental Microbiology, 42: 886896. Flood, J.A., N.J. Ashbolt, and P.C. Pollard. 1999. Complementary independent molecular, radioisotopic and fluorogenic techniques to assess biofilm communities in two wastewater wetlands. Water Science and Technology, 39: 65-70. Frankowski, K . A . 2000. The Treatment of Wood Leachate Using Constructed Wetlands. M.A.Sc. thesis, The University of British Columbia, Vancouver, Canada. Fredrickson, H . L . , T.E. Cappenberg, and J.W. De Leeuw. 1986. Polar lipid fatty acid compositions of Lake Vechten seston, an ecological application of lipid analysis. FEMS Microbiology Ecology, 38: 381-396. Fry, J.C. 1990. Direct methods and biomass estimation. Methods in Microbiology, 22: 41-85. Gale, P.M., I. Devai, K.R. Reddy, and D.A. Graetz. 1993. Denitrification potential of soils from constructed and natural wetlands. Ecological Engineering, 2: 119-130. Gasol, J.M. 1993. Benthic flagellates and ciliates in fine freshwater sediments: calibration of a live counting procedure and estimation of their abundances. Microbial Ecology, 25: 247-262. 42  Gearheart, R . A . , and B . A . Finney. 1996. Criteria for design of free surface constructed wetlands based upon a coupled ecological and water quality model. In: R. Haberl, R. Perfler, J. Laber, and P. Cooper (eds.), Proceedings of the 5th International Conference on Wetland Systems for Water Pollution Control, at Vienna, Austria. Ghiorse, W.C. 1988. Microbial reduction of manganese and iron. In: A.J.B. Zehnder (ed.), Biology of Anaerobic Microorganisms. John Wiley and Sons: New York, N Y . pp305-332. Gilmour, C.G., M . E . Leavitt, and M . P . Shiaris. 1990. Evidence against incorporation of exogenous thymidine by sulfate-reducing bacteria. Limnology and Oceanography, 35: 14011409. Haberl, R., R. Perfler, J. Laber, and P. Cooper (eds.). 1997. Proceedings of the 5th International Conference on Wetland Systems for Water Pollution Control, at Vienna, Austria. Hakulinen, R., M . A . Kahkonen, M . Salkinoja-Salonen. 2005. Vertical distribution of sediment enzyme activities involved in the cycling of carbon, nitrogen, phosphorus, and sulfur in three boreal rural lakes. Water Research, 39: 2319-2326. Hammer, D . A . (ed.). 1989. Constructed Wetlands for Wastewater Treatment, Municipal, Industrial, and Agricultural. Lewis Publishers: Chelsea, MI. Hammer, D.A. 1997. Creating Freshwater Wetlands, 2  nd  ed. Lewis Publishers: New York.  Harmsworth, G.C., and M . A . Sleigh. 1993. Colonization of non-living surfaces in streams by Peritrich ciliates. European Journal of Protistology, 29: 294-301. Hensel, L . 1990. Assessment of Chronic Impacts from Woodwaste Leachate. Report No.90-01, B.C. Ministry of Environment, Surrey, B C , Canada. Hesselsoe, M . , J.L. Nielsen, P. Rosier, and P.H. Nielsen. 2005. Isotope labeling and microautoradiography of active heterotrophic bacteria on the basis of assimilation of C02. 14  Applied and Environmnetal Microbiology, 71: 646-655. Hobbie, J.E., and T.E. Ford. 1993. A perspective on the ecology of aquatic microbes. In: T.E. Ford (ed.), Aquatic Microbiology: An Ecological Approach. Blackwell Scientific Publications: Boston, M A . pp 1 -14. Hollibaugh, J.T. 1988. Limitations of the ( H) thymidine method for estimating bacterial 3  productivity due to thymidine metabolism. Marine Ecology Progress Series, 43: 19-30. 43  Hollibaugh, J.T., and P.S. Wong. 1992. Ethanol-extractable substrate pools and the incorporation of thymidine, L-leucine, and other substrates by bacterioplankton. Canadian Journal of Microbiology, 38: 605-613. Hoppe, H.-G. 1993. Use of fluorogenic model substrates for extracellular enzyme activity (EEA) measurement of bacteria. In: P.F. Kemp, B.F. Sherr, E.B. Sherr, and J.J. Cole (eds.), Handbook of Methods in Aquatic Microbial Ecology. Lewis Publishers: Boca Raton, F L . pp421-431. Hoppe, H.-G., S.J. K i m , and K . Gocke. 1988. Microbial decomposition in aquatic environments: combined process of extracellular enzyme activity and substrate uptake. Applied and Environmental Microbiology, 54: 784-790. Hunt, A.P., and J.D. Parry. 1998. The effect of substratum roughness and river flow rate on the development of a freshwater biofilm community. Biofouling, 12: 287-303. Hunt, A.P., J.D. Parry, and J. Hamilton-Taylor. 1999. Algal-bacterial relationships in streams with varying D O C concentration and source. In: C W . Keevil, A . Godfree, D. Holt, and C. Dow (eds.), Biofilms in the Aquatic Environment. The Royal Society of Chemistry: Cambridge, U K . pp25-32. Hunter, R., A . E . Birkbeck, and G. Coombs. 1993. Innovative marsh treatment systems for control of leachate and fish hatchery wastewaters. In: G . A . Moshiri (ed.), Constructed Wetlands for Water Quality Improvements. Lewis Publishers: Baca Raton, F L . pp477-484. Johnson, M . D . , and A . K . Ward. 1993. A comparison of iNT-formazan methods for determining bacterial activity in stream ecosystems. Journal of the North American Benthological Society, 12: 168-173. Jokela, J.B., and C. Pinks. 1998. Constructed wetlands for stormwater treatment in Alaska. In: F. Hayes (ed.), Proceedings of the ASCE Wetlands Engineering and River Restoration Conference. American Society of Civil Engineers: Reston, V A . Jones, S.E., and M . A . Lock. 1991. Peptidase activity in river biofilms by product analysis. In: R. J. Chrost (ed.), Microbial Enzymes in Aquatic Environments. Springer-Verlag: New York. Jones, J.G., S. Gardener, and B . M . Simon. 1983. Bacterial reduction of ferric iron in a stratified eutrophic lake. Journal of General Microbiology, 129: 131-139. 44  Jorgensen, N.O.G. 1992. Incorporation of H-leucine and H-valine into protein of freshwater bacteria: uptake kinetics and intracellular isotope dilution. Applied and Environmental Microbiology, 58: 3638-3646. th  Kadlec, R.H., and H . Brix (eds.). 1995. Proceedings of the 4 International Conference on Wetland Systems for Water Pollution Control, held in Guangzhou, China, 6-10 November 1994. Kadlec, R.H., and R.L. Knight. 1996. Treatment Wetlands. Lewis Publishers: Boca Raton, F L . Kaplan, L . A . , and T.L. Bott. 1989. Diel fluctuations in bacterial activity on streambed substrata during algal blooms: effects of temperature, water chemistry, and habitat. Limnology and Oceanography, 34: 718-733. Kaplan, L . A . , T.L. Bott, and J.K. Bielicki. 1992. Assessment of [ H]thymidine incorporation 3  into D N A as a method to determine bacterial productivity in stream bed sediments. Applied and Environmental Microbiology, 58:3614-3621. Karl, D . M . 1993. Total Microbial Biomass Estimation Derived from the Measurement of Particulate Adenosine-5'-Triphosphate. In: P.F. Kemp, B.F. Sherr, E.B. Sherr, and J.J. Cole (eds.), Handbook of Methods in Aquatic Microbial Ecology. Lewis Publishers: Boca Raton, F L . pp359-372. Kemp, P.F., B.F. Sherr, E . B . Sherr, and J.J. Cole. 1993. Handbook of Methods in Aquatic Microbial Ecology. Lewis Publishers: Boca Raton, F L . King, G . M . , and T. Berman. 1984. Potential effects of isotopic dilution on apparent respiration in C heterotrophy experiments. Marine Ecology Progress Series, 19: 175-180. 1 4  King, G . M . , and M.J. Klug. 1982. Glucose metabolism in sediments of a eutrophic lake: tracer analysis of uptake and product formation. Applied and Environmental Microbiology, 44: 1308-1317. Kirchman, D. L . 1992. Incorporation of thymidine and leucine in the subarctic Pacific: application to estimating bacterial production. Marine Ecology Progress Series, 82: 301-309. Kirchman, D.L., E. K'Nees, and R. Hudson. 1985. Leucine incorporation and its potential as a measure of protein synthesis by bacteria in natural aquatic Environmental Microbiology, 49: 599-607. 45  systems. Applied and  Kirka, J.L., L . A . Beaudettea, M . Hartb, P. Moutoglisc, J.N. Klironomosb, H . Lee, and J.T. Trevors. 2004. Methods of studying soil microbial diversity. Journal of Microbiological Methods, 58: 169-188. Kisand, V . , and T. Noges. 2004. Abiotic and biotic factors regulating dynamics of bacterioplankton in a large shallow lake. FEMS Microbiology Ecology, 50: 51-62. Knight, R.L. 2004. N C A S I Technical Bulletin, 876. Komatsu, Y . , M . Takagi, and M . Yamaguchi. 1978. Participation of iron in denitrification in waterlogged soil. Soil Biology and Biochemistry, 10: 21-26. Kong, Y . , J.L. Nielsen, and P.H. Nielsen. 2005. Identity and ecophysiology of uncultured actinobacterial polyphosphate-accumulating organisms in full-scale enhanced biological phosphorus removal plants. Applied and Environmental Microbiology, 71: 4076-4085. Kozub, D.D. and S.K. Liehr. 1999. Assessing denitrification rate limiting factors in a constructed wetland receiving landfill leachate. Water Science and Technology, 40: 75-82. Kushner, D.J. 1993. Microbial life in extreme environments. In: T.E. Ford (ed.), Aquatic Microbiology: An Ecological Approach. Blackwell Scientific Publications: Boston, M A . pp383-407. Lechevallier, M.W., and G.A. McFeters. 1990. Microbiology of activated carbon. In: G . A . McFeters (ed.), Drinking Water Microbiology. Springer Verlag: New York, N Y . ppl04-l 19. Lee, S., and J.A. Fuhrman. 1987. Relationship between biovolume and biomass of natural derived marine bacterioplankton. Applied and Environmental Microbiology, 53: 1298-1303. Liehr, S.K. 2000. Constructed Wetlands Treatment of High Nitrogen Landfill Leachate. W E R F report No. 94-IRM-U. Lovley, D.R. 1991. Dissimilatory Fe(III) and Mn(IV) reduction. Microbiology Review, 55: 259-287. Lovley, D.R., and M . J . Klug. 1982. Intermediary metabolism of organic matter in the sediments of a eutrophic lake. Applied and Environmental Microbiology, 43: 552-560.  46  Lovley, D.R., and E.J. Phillips. 1988. Novel mode of microbial energy metabolism: organic carbon oxidation coupled to dissimilatory reduction of iron and manganese. Applied and Environmental Microbiology, 54: 1472-1480. Madigan, M.T., J.M. Martinko, and J. Parker. 1997. Brock Biology of Microorganisms, 8 ed. th  Prentice Hall: Upper Saddle River, NJ. Marx, M . - C , M . Wood, and S.C. Jarvis. 2001. A microplate fluorimetric assay for the study of enzyme diversity in soils. Soil Biology and Biochemistry, 33: 1633-1640. Marxsen, J., and H.-H. Schmidt. 1993. Extracellular phosphatase activity in sediments of the Breitenbach, a Central European mountain stream. Hydrobiologia, 253: 207-216. Masbough, A . 2002. The Effectiveness of Constructed Wetlands for Treatment of Wood Leachate. M.A.Sc. thesis, The University of British Columbia, Vancouver, Canada. Massana, R., and H . Gude. 1991. Comparison between three methods for determining flagellate abundance in natural waters. OPHELIA, 33: 197-203. Mbwette, T.S.A., J.H.Y. Katima, S.E. Jorgensen, and S Kayombo (eds.). 2003. Selected Proceedings of the 8? IWA International Conference for Wetland Systems in Water Pollution h  Control VIII, held at Arusha, Tanzania, 16-19 September 2002. McCormick, P.V., J.R. Pratt, D.G. Jenkins, and J.J. Cairns. 1988. A comparison of protozoan, algal, and metazoan colonization of artificial substrates of differing size. Transactions of the American Microscopical Society, 107: 259-268. Melo, L.F., T.R. Bott, M . Fletcher, and B . Capdeville. 1992. Biofilms: Science and Technology. Applied Sciences, N A T O ASI Series E, 223. Metcalf & Eddy. 1991. Wastewater Engineering - Treatment, Disposal and Reuse, 3rd ed. McGraw-Hill: New York, N Y . Metcalf & Eddy. 2003. Wastewater Engineering - Treatment and Reuse, 4th ed. McGraw-Hill: New York, N Y . Mitsch, W.J. 1992. Landscape design and the role of created, restored and natural riparian wetlands in controlling nonpoint source pollution. Ecological Engineering, 1:27-47. Mitsch, W.J. and J.G. Gosselink. 2000. Wetlands, 3rd ed. Jogn Wiley: New York, N Y .  47  Monticello, D.J., and R . N . Costilow. 1982. Interconversion of valine and leucine by Clostridium sporogenes. Journal of Bacteriology, 152: 946-949. Moran, M . A . , and R.E. Hodson. 1992. Contributions of three subsystems of a freshwater marsh to total bacterial secondary productivity. Microbial Ecology, 24: 161-170. Moshiri, L . (ed.). 1993. Constructed Wetlands for  Water Quality Improvement. Lewis  Publishers, Inc.: Chelsea, MI. Mueller, R.F. 1996. Bacterial transport and colonization in low nutrient environments. Water Research, 30:2681-2690. Muller, E., B . Abmus, A . Hartmann, and K.-P. Seiler. 2000. The in situ detection of a microbial biofilm community on karst rock coupon in a groundwater habitat. In: H.-C. Flemming, U . Szewzyk, and T. Griebe (eds.), Biofilms: Investigative Methods & Applications. Technomic Publishing Company, Inc.: Lancaster, PA. ppl55-164. Norland, S. 1993. The relationship between biomass and volume of bacteria. In: P.F. Kemp, B.F. Sherr, E . B . Sherr, and J.J. Cole (eds.), Handbook of Methods in Aquatic Microbial Ecology. Lewis Publishers: Boca Raton, NJ. pp303-307. Oemke, M.P., and T . M . Burton. 1986. Diatom colonization dynamics in a lotic system. Hydrobiologia, 139: 153-166. Otte, M . L . , C C . Kearns, and M.O. Doyle. 1995. Accumulation of zinc in the rhizosphere of wetland plants. Bulletin of Environmental Contamination and Toxicology, 55: 154-161. Pease, B . C . 1974. Effects of Log Dumping and Rafting on the Marine Environment of Southeast Alaska. U S D A technical report PNW-22. Pedros-Alio, C, J.G. Cantizano, and J.L. Calderon. 1993. Bacterial production in anaerobic water columns. In: P.F. Kemp, B.F. Sherr, E . B . Sherr, and J.J. Cole (eds.), Handbook of Methods in Aquatic Microbial Ecology. Lewis Publishers: Boca Raton, F L . pp519-530. Pernthaler, J., B . Sattler, K . Simek, A . Schwarzenbacher, and R. Psenner. 1996. Top-down effects on the size-biomass distribution of a freshwater bacterioplankton community. Aquatic Microbial Ecology, 10: 255-263.  48  Peters, G.B., H.J. Dawson, B.F. Hrutfiord, and R.R. Whitney. 1976. Aqueous leachate from western red cedar: effects on some aquatic organisms. Journal of the Fisheries Research Board of Canada, 33: 2703-2709. Pollard, P.C., and D.J.W. Moriarty. 1984. Validity of the tritiated thymidine method for estimating bacterial growth rates: measurement of isotope dilution during D N A synthesis. Applied and Environmental Microbiology, 48: 1076-1083. Pollard, P.C., J.A. Flood, and N.J. Ashbolt. 1995. The direct measurement of bacterial growth in biofilms on emergent plants of an artificial wetland. Water Science and Technology, 32: 251-256. Polprasert, C , and B . K . Agrawalla. 1994. A facultative pond model incorporating biofilm activity. Water Environment Research, 66: 725-732. Polprasert, C , and N.R. Khatiwada. 1998. A n integrated kinetic model for water hyacinth ponds used for wastewater treatment. Water Research, 32: 179-185. Polprasert, C , N.P. Dan, and N . Thayalakumaran. 1996. Application of constructed wetlands to treat some toxic wastewaters under tropical conditions. Water Science and Technology, 34: 165-171. Polprasert, C , N.R. Khatiwada, and J. Bhurtel. 1998. A model for organic matter removal in free water surface constructed wetlands. Water Science and Technology, 38: 369-377. Porter, K . G., and Y.S. Feig. 1980. The use of DAPI for identifying and counting aquatic microflora. Limnology and Oceanography, 25: 943-948. Pries, J. 1994. Wastewater and stormwater applications of wetlands in Canada. Sustaining Wetlands Issues, Paper No. 1994-1, North American Wetlands Conservation Council, Ottawa, Ontario, Canada. Radl, V., K . Pritsch, J.C. Munch, and M . Schloter. 2005. Structural and functional diversity of microbial communities from a lake sediment contaminated with trenbolone, an endocrinedisrupting chemical. Environmental Pollution, 137: 345-353. Rash, J.K., and S.K. Liehr. 1999. Flow pattern analysis of constructed wetlands treating landfill leachate. Water Science and Technology, 40: 305-309.  49  Reddy, K.R., and D.A. Graetz. 1988. Carbon and nitrogen dynamics in wetland soils. In: D.D. Hook et al. (eds.), The Ecology and Management of Wetlands, Vol. I. Timber Press: Portland, OR. Reddy, K . R . , and R . H . Kadlec (eds.). 2001. Selected Proceedings of the 7th International Conference on Wetland Systems for Water Pollution Control 2000, held in Lake Buena Vista, Florida, 11-16 November 2000. Reed, S.C., R.W. Crites, and E J . Middlebrooks. 1995. Natural Systems for Waste Management and Treatment, 2nd ed. McGraw-Hill: New York, N Y . Riemann, B., and F. Azam. 1992. Measurements of bacterial protein synthesis in aquatic environments by means of leucine incorporation. Marine Microbial Food Webs, 6: 91-105. Riemann, B., J.A. Fuhrman, and F. Azam. 1982. Bacterial secondary production in freshwater measured by H thymidine incorporation method. Microbial Ecology, 8: 101-114. 3  Robarts, R.D., R J . Wicks, and L . M . Sephton. 1986. Spatial and temporal variations in bacterial macromolecule labeling with (methyl- H) thymidine in a hypertrophic lake. Applied 3  and Environmental Microbiology, 52: 1368-1373. Rodriguez, G. G., D. Phipps, K . Ishiguro, and H.F. Ridgway. 1992. Use of a fluorescent redox probe for direct visualization of actively respiring bacteria. Applied and Environmental Microbiology, 58: 1801-1808. Sawyer, T.E., and G . M . King. 1993. Glucose uptake and end product formation in an intertidal marine sediment. Applied and Environmental Microbiology, 59: 120-128. Schaule, G., H.-C. Flemming, and H.F. Ridgway. 1993. Use of 5-cyano-2,3-ditolyl tetrazolium chloride for quantifying planktonic and sessile respiring bacteria in drinking water. Applied and Environmental Microbiology, 59: 3850-3857. Schaule, G., T. Griebe, and H.-C. Flemming. 2000. Steps in biofilm sampling and characterization in biofouling cases. In: H.-C. Flemming, U . Szewzyk, and T. Griebe (eds.), Biofilms: Investigative Methods & Applications. Technomic Publishing Company, Inc.: Lancaster, PA. ppl-22. Schwartz, T., S. Hoffmann, and U . Obst. 2000. Formation and bacterial composition of young, natural biofilms obtained from public bank-filtered drinking water systems. In: H.-C. 50  Flemming, U . Szewzyk, and T. Griebe (eds.), Biofilms: Investigative Methods & Applications. Technomic Publishing Company, Inc.: Lancaster, PA. ppl23-142. Shamsudin, L . , and M . A . Sleigh. 1994. Seasonal changes in composition and biomass of epilithic algal floras of a chalk stream and a soft water stream with estimates of production. Hydrobiologia, 273: 131-146. Sherr, E.B., D . A . Caron, and B.F. Sherr. 1993. Staining of heterotrophic protests for visualization via epifluorescent microscopy. In: P.F. Kemp, B.F. Sherr, E.B. Sherr, and J.J. Cole (eds.), Handbook of Methods in Aquatic Microbial Ecology. Lewis Publishers: Boca Raton, FL. pp213-227. Sime-Ngando, T., H.J. Hartmann, and C.A. Groliere. 1990. Rapid quantification of planktonic ciliates: comparison of improved live counting with other  methods. Applied and  Environmental Microbiology, 56: 2234-2242. Simon, M . , and F. Azam. 1989. Protein content and protein synthesis rates of planktonic marine bacteria. Marine Ecology Progress Series, 51:201-213. Somville, M . 1984. Measurement and study of substrate specificity of exoglucosidase activity in eutrophic water. Applied and Environmental Microbiology, 48: 1181-1185. Sorokin, Y.I., and H . Kadota. 1972. Techniques for the Assessment of Microbial Production and Decomposition  in Fresh  Waters. IBP Handbook No. 23, Blackwell  Scientific  Publications: Oxford, U K . Steenbergen, C . L . M . , R . A . J.-P. R . A . Sweerts, and T.E. Cappenberg. 1993. Microbial biogeochemical activities in lakes: stratification and eutrophication. In: T.E. Ford (ed.), Aquatic Microbiology: An Ecological Approach. Blackwell Scientific Publications: Boston, M A . pp69-99. Stewart, P.M., J.R. Pratt, J.J. Cairns, and R.L. Lowe. 1985. Diatom and protozoan species accrual on artificial substrates in lentic habitats. Transactions of the American Microscopical Society, 104: 369-377. Stock, M.S., and A . K . Ward. 1989. Establishment of a bedrock epilithic community in a small stream: microbial (algal and bacterial) metabolism and physical structure. Canadian Journal of the Fishiries Aquatic Science, 46: 1874-1883. 51  Tabor, P.S., and R.A. Neihof. 1982. Improved microautoradiographic method to determine individual microorganisms active in substrate uptake in natural waters. Applied and Environmental Microbiology, 44: 945-953. Taylor, B.R., and N.B. Carmichael. 2003. Toxicity and chemistry of aspen wood leachate to aquatic life: field study. Environmental Toxicology and Chemistry, 22: 2048-2056. Taylor, B.R., J.S. Goudey, and N . B . Carmichael. 1996. Toxicity of aspen wood leachate to aquatic life: Laboratory studies. Environmental Toxicology and Chemistry, 15: 150-159. Temmink, J.H.M., J.A. Field, J.C. van Haastrecht, and R . C . M . Merkelbach. 1989. Acute and sub-acute toxicity of bark tannins in carp (Cyprinus carpio L.). Water Research, 23: 341-344. Thomaz, S.M., and R.G. Wetzel. 1995. [ H]Leucine incorporation methodology to estimate 3  epiphytic bacterial biomass production. Microbial Ecology, 29: 63-70. Toerien, D.F., and M . C . Toerien. 1985. Microbial heterotrophy in an effluent treatment system using macrophytes. Agricultural Wastes, 12: 287-312. Toet,  S., L.H.F.A.  Huibers, R.S.P. Van Logtestijn, and J.T.A. Verhoeven. 2003.  Denitrification in the periphyton associated with plant shoots and in the sediment of a wetland system supplied with sewage treatment plant effluent. Hydrobiologia, 501: 29-44. Torreton, J.P., and M . Bouvy. 1991. Estimating bacterial D N A synthesis from [ H]thymidine 3  incorporation: discrepancies among macromolecular extraction procedures. Limnology and Oceanography, 36: 299-306. Tunlid, A . , and D.C. White. 1990. Use of lipid biomarkers in environmental samples. In: A . Fox, S.L. Morgan, L . Larsson and G . Odham (eds.), Analytical Microbiology Methods. Plenum Press: New York, N Y . pp259-274. T V A . 1993. General Design, Construction and Operation Guidelines: Constructed Wetlands Wastewater Treatment Systems for Small Users Including Individual Residences. T V A / W M 93/10, Water Management Resource Group, Tennessee Valley Authority, Chattanooga, T N . USDA.  1995. Handbook of Constructed  Wetlands. U.S. Department of Agriculture,  Washington, D.C.  52  USEPA. 1988. Design Manual for Constructed Wetlands and Aquatic Plant Systems for Municipal Wastewater Treatment. E P A 625/1-88-022, U.S. Environmental Protection Agency, Cincinnati, OH, September. USEPA. 1996. Protecting Natural Wetlands: A Guide to Stormwater Best Management Practices. EPA-843-B-96-001, U.S. Environmental Protection Agency, Washington, D.C. USEPA. 1999. Free Water Surface Wetlands for Wastewater Treatment: A Technology Assessment. EPA/832/R-99/002, U.S. Environmental Protection Agency, Washington, D.C. USEPA. 2000. Constructed Wetlands Treatment of Municipal Wastewaters. Manual of EPA/625/R-99/010, U.S. Environmental Protection Agency, Cincinnati, OH. USEPA. 2002. Considerations in the Design of Treatment Best Management Practices (BMPs) to Improve Water Quality. EPA/600/R-03/103, U.S. Environmental Protection Agency, Cincinnati, OH, September. van der Kooij, D., H.R. Veenendaal, and W.J.H. Scheffer. 2005. Biofilm formation and multiplication of Legionella in a model warm water system with pipes of copper, stainless steel and cross-linked polyethylene. Water Research, 39: 2789-2798. van der Wende E., W.G. Characklis, and D.B. Smith. 1989. Biofilms and bacterial drinking water quality. Water Research, 23: 1313-1322. Velji, M.I., and L . J . Albright. 1993. Improved sample preparation for enumeration of aggregated aquatic substrate bacteria. In: P.F. Kemp, B.F. Sherr, E.B. Sherr, and J.J. Cole (eds.), Handbook of Methods in Aquatic Microbial Ecology. Lewis Publishers: Boca Raton, FL.ppl39-142. Ward, B . B . 1982. Oceanic distribution of ammonium-oxidizing bacteria determined by immunofluorescence assay. Journal of Marine Research, 40: 1155-1172. Ward, A . K . , and M . D . Johnson. 1996. Heterotrophic microorganisms. In: F.R. Hauer and G.A. Lamberti (eds.), Methods in Stream Ecology. Academic Press: New York, N Y . pp254-260. Werker, A . G . , and E.R. Hall. 2000. Assessment of microbial fatty acid analysis for quantifying mixed culture biomass during batch growth. Water Environment Research, 72: 484-489.  53  Westermann, P. 1993. Wetland and swamp microbiology. In: T.E. Ford (ed.), Aquatic Microbiology: An Ecological Approach. Blackwell Scientific Publications: Boston, M A . pp215-239. Wetzel, R . G . 2001. Limnology: Lake and River Ecosystems, 3rd ed. Academic Press: San Diego, C A . Wetzel, R.G., and G.E. Likens. 2000. Limnological Analyses, 3rd ed. Springer-Verlag: New York, N Y . Wetzel, R., P. Hatcher, and T.S. Bianchi. 1995. Natural photolysis by ultraviolet irradiance of recalcitrant dissolved organic matter to simple substrates for rapid bacterial metabolism. Limnology and Oceanography, 40: 1369-1380. White, D.C. et al. 1999. The biofilm ecology of microbial biofouling, biocide resistance and corrosion. In: C W . Keevil, A . Godfree, D. Holt and C. Dow (eds.), Biofilms in the Aquatic Environment. The Royal Society of Chemistry: Cambridge, U K . pp 120-130. Wicks, R.J., and R.D. Robarts. 1988. Ethanol extraction requirement for purification of protein labeled with [ H]leucine in aquatic bacterial production studies. Applied and Environmental 3  Microbiology, 54:191-193. Widdel, F. 1988. Microbiology and ecology of sulfate- and sulfur-reducing bacteria. In: A.J.B. Zehnder (ed.), Biology of Anaerobic Microorganisms. John Wiley and Sons: New York, N Y . pp469-585. Woodhouse, C , and S.J.B. Duff. 2004. Treatment of log yard runoff in an aerobic trickling filter. Water Quality Research Journal of Canada, 39: 232-238. WPCF. 1990. Natural Systems for Wastewater Treatment, Manual of Practice FD-16, Water Pollution Control Federation, Alexandria, V A . Wright, R.T., and J.E. Hobbie. 1966. Use of glucose and acetate by bacteria and algae in aquatic ecosystems. Ecology, 47: 447-464. Xue, Y . , D.A. Kovacic, M . B . David, L . E . Gentry, R.L. Mulvaney, and C W . Lindau. 1999. In situ measurements of denitrification in constructed wetlands. Journal of Environmental Quality, 28: 263-269.  54  Zehnder, A.J.B., and W. Stumm. 1988. Geochemistry and biogeochemistry of anaerobic habitats. In: A.J.B. Zehnder (ed.), Biology of Anaerobic Microorganisms. John Wiley and Sons: New York, N Y . ppl-39. Zelles, L., and Q.Y. Bai. 1993. Fatty acid patterns of phospholipid and lipopolysaccharides in environmental samples. Chemosphere, 28: 391-411. Zenaitis, M . G . , and S.J.B. Duff. 2002. Ozone for removal of acute toxicity from logyard runoff. Ozone: Science and Engineering, 24: 83-90. Zenaitis, M . G . , H . Sandhu, and S.J.B. Duff. 2002. Combined biological and ozone treatment of log yard runoff. Water Research, 36: 2053-2061. Zinder, S.H. 1986. Patterns of carbon flow from glucose to methane in a thermophilic anaerobic bioreactor. FEMS Microbiology Ecology, 38: 243-250.  55  2  CHARACTERIZATION OF WOODWASTE L E A C H A T E *  2.1 Introduction Leachate is generated when water percolates through a large mass of solid material as contaminated liquid. Wood leachate may be generated when rainfall, runoff, or sprinkling water percolates through woodpiles, log yards, wood product storage areas, and woodwaste disposal sites of various wood processing mills and barge loading facilities. Naturally occurring wood  constituents,  preservatives, metals  from  the  facilities,  and  wood  decomposition by-products may be present in wood leachate. Wood leachate extracted by water in laboratory study (Field et al., 1988; Peters et al., 1976; Taylor et al., 1996; Temmink et al., 1989) or leached in open fields (Bailey et al., 1999; Hunter et al., 1993; Peters et al., 1976; Taylor and Carmichael, 2003; Woodhouse and Duff, 2004; Zenaitis et al., 2002) has been characterized as acidic, of very high oxygen demand, and toxic to aquatic organisms. The reported chemical composition of wood leachate varies with wood type, wood component (bark, heartwood, sapwood, or foliage), waste compaction, leaching time, and characteristics of leaching water. In regions such as British Columbia, Canada, where forestry is one of the major industrial sectors, collection and disposal of woodwaste (process trimmings, off-specification wood chips, bark, and sawdust) is an immense task. Woodwaste, especially when it is not covered, poses a serious threat to surface and ground water due to leachate generation. Leachate flow rate and contaminant concentrations vary significantly over time as the landfill develops, through closure and after closure (Farquhar, 1989; Forgie, 1988). Operation flexibility should be considered in treatment plant design with regard to the time-variable nature of the leachate from woodwaste piles. The types, amounts, and production rates of contaminants appearing in the leachate at a landfill are influenced by several factors that are difficult to quantify reliably in a landfill (El-Fadel et al., 1997; Farquhar, 1989). It is, therefore, necessary to rely on data  A version of this chapter has been accepted for publication. Tao, W., K . J . Hall, A . Masbough, K . Frankowski, and S.J.B. Duff. 2005. Characterization of leachate from a woodwaste pile. Water Quality Research Journal of Canada, 40: in press. 56  and experience from other landfill investigations. However, long-term monitoring and characterization of woodwaste leachate have not been reported to date. Refuse age and corresponding landfill decomposition stage are usually the major determinants of leachate composition (El-Fadel et al., 1997). The present study had been monitoring the leachate generated by precipitation on a cedar woodwaste pile for 6 consecutive years, including both the placement and closure periods. Based on continual and occasional monitoring data of leachate quality, this study assessed the physical-chemical properties and toxicity of the woodwaste leachate. Temporal variations in leachate quality were examined with regard to the development sequence of the woodwaste pile and local weather conditions. Further, it discussed the implications of leachate characteristics to treatment of woodwaste leachate.  2.2  Materials and Methods  2.2.1 Woodwaste pile and leachate pool The woodwaste pile of concern was a large open storage of sawdust, shredded bark and roots, off-specification wood chips, and process trimmings. It was located adjacent to Fraser River in Mission, British Columbia, Canada. Several wood processing facilities, mostly cedar shake and shingle mills, stacked their cedar woodwaste together in an effort to gain a sufficient quantity for either selling it as fuel to electrical cogeneration facilities or using it as a raw material for some other industrial processes (e.g., charcoal production). Since 1992, this pile had been receiving new woodwaste and growing up from an area of approximately 150 m to 200 m in diameter and to 20 m high in 1999, adjoining the working site of a wood chips company at a higher elevation. There were no intentional compaction, placement sequence, bottom lining, and cover while placing the waste. The stored woodwaste was delivered to a pulp mill as a fuel supplement to its cogeneration boiler on an as-requested basis in 1999 and 2000. Consequently, the pile was reduced to about 12 m high in 2000. The pile size increased again to 20 m high in early 2001. In the summer of 2001, the 8-m-high cap of the pile was removed to become a working site (Figure 2-1), and new piles formed in the adjacent area. In general, the pile of concern underwent a placement period before the summer 2001, followed by closure.  57  Figure 2-1. Woodwaste pile in October 2004 (left) and leachate pool in March 2004 (right).  This area is characterized by a coastal climate with colder, wet winters and warm, drier summers. The climate information during the study period at the nearby "Abbotsford A " meteorological station (elevation of 57.9 m) indicated that the mean monthly temperature ranged from 20.3°C in July 2004 to 2.9°C in January 2004 (Figure 2-2). The annual precipitation was 1255-1797 mm between 1999 and 2004, with 22.5-44.4% occurring during the spring-summer period (April-September).  Year and month (January-December) Figure 2-2. Variation of monthly air temperature and precipitation at Abbotsford A meteorological station. Data from Environment Canada (2005).  58  When precipitation fell directly on the pile and percolated through the woodwaste, it produced an amber leachate. A pool of leachate (Figure 2-1), approximately 20 m by 70 m, had formed in a natural depression on one side of the woodwaste pile since its establishment. A very strong pungent, woody smell emanated from the leachate pool. In the dry summers from July to September, the pool was shallow, but it could collect one meter of leachate through the wetter months. The quantity of leachate generated is site-specific. It depends on weather conditions, landfill cover and lining, and moisture content and field capacity of the solid waste. No runoff was observed on this woodwaste pile. Based on the annual precipitation and evaporation norms between 1971 and 2000 of nearby meteorological stations (Environment Canada, 2005), an annual leachate generation of 24,680 m was estimated with a simplified water balance model. 3  Directly underlying the pile was an impermeable stratum consisting of silt and clay, ranging from 0.9 to 4.1 m thick. Beneath this stratum was an aquifer. The leachate pool could only hold about 1000 m leachate, while the majority might flow to the other sides of the pile or 3  penetrate weak spots of the impermeable stratum beneath the pile.  2.2.2 Field measurement and laboratory analysis The woodwaste leachate had been collected from the leachate pool in 1-L plastic bottles at weekly intervals mostly during the warmer months (March to October) between 1999 and 2004. The leachate was analyzed to facilitate other studies and not conducted on a consistent frequency. C O D , T & L , V F A s , ammonia, nitrate, nitrite, orthophosphate, and p H of the woodwaste leachate had been measured continually through the placement and closure periods of the woodwaste pile. TSS and 5-day B O D were measured in 1999 and 2000 and checked occasionally in the closure period. The 96-h median lethal concentration to rainbow trout was determined in 1999. Total Kjeldahl nitrogen (TKN) and total phosphorus (TP) were analyzed twice and three times respectively. Some other parameters (organic carbon, color, metals, etc.) analyzed only a few times are given in Appendix 1. Field measurements were made while sampling. Water temperature and DO were measured in the leachate pool from 1999 to 2002, using a portable Y S I Model 54 D O meter (Yellow Springs Instrument, Yellow Springs, OH). Redox potential (Eh) was determined from millivolt readings from 2002 to 2004, using a Horiba D-13 US model p H meter (Horiba, Ltd., Kyoto,  59  Japan) equipped with a Pt-Ag/AgCl combination electrode (Broadley James Corporation, Irvine, CA). Standard methods (APHA et al., 1999) were followed for analysis of pH (4500-H  +  electrometric) with the Horiba pH meter, solids (2540), C O D (5220D closed reflux, colorimetric) with a Hach DR/2000 spectrophotometer (Hach Co., Loveland, CO), T & L (5550B Folin phenol colorimetric) with the Hach spectrophotometer, B O D (521 OB 5-day incubation; seeded with sediment of the leachate pool), ammonia (4500-NH3 H flow injection), nitrate plus nitrite (45OO-NO3" I cadmium reduction flow injection), T K N (4500Norg D block digestion and flow injection), TP (4500-P H manual digestion and flow injection), and orthophosphate (4500-P G flow injection). Flow injection analyses were done on a Lachat Quick-Chem 8000 automatic flow-injection ion analyzer (Lachat Instruments, Milwaukee, WI) with absorbance detectors. T & L were analyzed with tannic acid  (C76H52O46  of Fisher Scientific, Fair Lawn, NJ) as standards and reported as equivalent tannic acid. The analyses of C O D and T & L were not affected by the color of water samples because the samples were diluted 25 and 250-1000 times respectively with Millipore-Q ultra-pure water to bring them within method ranges, 20-900 mg L" C O D and <8 mg L" T & L . V F A s were 1  1  analyzed by gas chromatography (HPGC 5880A; Supelco, Inc G C Bulletin 751G). The G C method quantified acetic acid ( C H 3 C O O H ) , propionic acid (C2H5COOH), iso-butyric acid and butyric acid (C3H7COOH), valeric acid (C4H9COOH), and hexanoic acid (C HnCOOH). 5  Acute toxicity tests using rainbow trout followed the procedure described by Environment Canada (1990), with sufficient oxygen and pH adjusted to 4.5-5.0.  2.2.3  Statistical analysis  The monthly averages of leachate quality parameters, instead of values of individual samples, were used for statistical analysis because the leachate was collected at variable frequencies across months. The value of n stands for the number of months monitored. The correlations of leachate quality to air temperature and precipitation were analyzed by multiple linear regressions, which output the coefficient of determination, R . The temporal variation of the 2  ratio of V F A s theoretical oxygen demand (ThOD) to C O D and the correlations between parameters were checked by Spearman's rank correlation analysis. A correlation was considered significant at a probability value of P <0.05.  60  2.3 Results 2.3.1  Physical and toxicological properties  The leachate pool acted as a settling pond, with the majority of solids and organic matter present in a dissolved form (through microfiber glass filters with pore size of 0.7 urn). The woodwaste leachate had <0.5 mg L" settleable solids, and 19-43 mg L" TSS. Over 99% of the 1  1  solids present in the leachate were in dissolved form. More than 98% COD was soluble. The woodwaste leachate exhibited an amber color at p H around 4, and became darker with increasing pH, suggesting that it is largely due to some forms of colored weak acids, similar to tannic acid, that undergo a shift in their protonation as p H changes and subsequently their spectral properties. Anion exchange resin Duolite S3 7 columns (10 cm long x2) completely removed the visual color of woodwaste leachate. Less than 5% T & L was found left in the column filtrate. Leachate temperatures (Table 2-1) were higher than ambient air temperatures (Figure 2-2), with an average difference of 7.1°C between 1999 and 2002. The higher leachate temperature could result from absorption of solar energy by the dark woodwaste leachate in the leachate pool and from anaerobic decomposition in the woodwaste pile. The woodwaste leachate exerted acute toxicity to rainbow trout at a 96-h median lethal concentration of 0.74% (v/v) full-strength leachate in 1999. The toxicity may be attributed to tannins, lignins, zinc, and low pH, but not confirmed yet.  2.3.2  Change over age of woodwaste pile  Figure 2-3 shows variation of the woodwaste leachate quality at monthly intervals over six years. The yearly averages and deviations are presented in Table 2-1. There was not an obvious change in leachate quality in the placement period from 1999 to 2001. The "young" leachate in the placement period contained very high concentrations of COD, T & L , and V F A s . The concentrations of COD and V F A s were similar to the young leachate of municipal solid waste landfills (El-Fadel et al., 1997; Farquhar, 1989). The concentrations of COD, T & L , and V F A s began to decrease in approximate 1, 1.5, and 1.5 years respectively after stopping placement of new woodwaste in the summer of 2001. There was a rapid decrease in the concentrations of COD, T & L and V F A s in 2004, 2.5 years into the closure period. C O D  61  concentration of the old woodwaste leachate decreased to values similar to that of municipal landfill leachate at a refuse age of 10-20 years (Farquhar, 1989). pH varied between 3.3 and 3.9 from 1999 to 2002, and began to increase starting from 2003. Because of very high oxygen demand, D O was always below 0.6 mg L" . Redox potential did not show a clear trend of 1  variation from 2002 to 2004.  Table 2-1. Variation of woodwaste leachate quality over years Year  1999  2000  2001  2002  2003  2004  Monitoring  May-Dec.  May-Sep.  Jun.-Oct.  Jun.-Nov.  Feb.-Oct.  Mar.-Oct.  period  (n = 3-5)  (» = 4-5)  (« = 5)  (n = 3-5)  (n = 4-5)  (n = 7-8)  Temperature, °C  12.6±1.8  28.9+2.2  19.7+4.1  22.3+1.2  ND  ND  DO, mg L"  0.3+0.1  0.2+0.1  0.4+0.1  0.5+0.1  ND  ND  ND  ND  ND  258+8  329+48  219+41  3.6±0.1  3.4+0.0  3.7+0.1  3.7+0.1  4.3+0.3  5.8+0.7  1564±463  2132+303  2063+60  2211+426  994+326  191+221  12.66±4.37 14.25+3.05 12.56+1.30 10.21+1.07 6.07+1.45  3.91+1.63  3070+622 5150+1210 3580+437  1  Redox potential (Eh), m V pH V F A s , mg L" COD, g L"  1  1  T & L , mg L"  1  POj-.mgPL-  1  NH , mgNL-  1  3  NO" +NO",  3370+228  1840+491  1100+271  3.5+0.9  4.3+1.0  3.3+0.2  2.1+0.5  1.5+1.1  2.9+1.8  1.1+0.9  3.1+0.9  1.8+1.0  1.8±0.9  3.2+4:3  11.3+10.5  0.27+0.33  0.15+0.10  0.14+0.12  0.08+0.05  0.50+0.14  0.26+0.15  mgNL"' a  Mean ± standard error of monthly average monitoring data; n = number of months  monitored. N D : no data.  62  18000 o Volatile fatty acids  16000  A Tannin & lignin • Chemical oxygen demand  14000 12000 10000 8000 6000  AA  4000  AA  A  A  2000  o  04  A  O oo o  . 1 1 . 1 1 1 1  o o o  OS OS ON  AAA  AA^  ooooo IIIIII  I  I  I  I  o  I I  © o  CN  I  IIII I  CN  8 o  o  IIIII I  o  o. o  o o  <N  *0 A  A  , 0u ~  o o  f  l  "AA  AOpOr-H  o o  tN  CN  Year and month (January-December)  30 • Ammonia nitrogen 25  opH  20 15 10  OOO 0\  OS OS  o o o  CN  oooo  QSQO  o o  ••••  o  - ••• 0 0 o  I I I I I I I PI I I I I I I I I I I I <— © © CN  CN © © CN  III  o o  <N  o o8  I  II  1 p>l  oooo  0  o oo i  I  II I  o o  fN  Year and month (January-December)  Figure 2-3. Temporal variation of woodwaste leachate quality.  2.3.3  Composition of oxygen demand  The young woodwaste leachate had a BODsiCOD ratio of 0.33 on average from June through December 1999, and older leachate had a B O D : C O D ratio of 0.14 on average in July and 5  August 2004. Tannins and lignins are noted for their resistance to biodegradation. On an 63  annual basis, the ratio of T & L ThOD:COD varied between 0.33 and 0.45. VFAs are easily biodegradable carbon sources. The ratio of annual V F A s ThOD:COD apparently increased (r$ = 0.78, P O . 0 1 ) from 0.16 in 1999 to 0.34 in 2002, and then significantly decreased (r = s  0.93, P O . 0 0 ) to 0.25 in 2003 and 0.06 in 2004 (Figure 2-4). T & L and VFAs together accounted for 43-73% of COD on an annual basis.  0.8 -i  o VFAs ThOD: COD  0.7 -  • T&LThOD:COD  0.6 Q O U Q O  -C  H  0.5 0.4 -  •eS  0.3 -  o  0.2 -  O  0.1 0.0 ON OS  ° o °  •  °°°°  o  • o  • • •  o  o  o o o rs  o •  I I I I I I I I  rs o o  o o  CM  r«-> © o  o o o  o o  Monitoring month (January-December) and year  Figure 2-4. Temporal variation in composition of chemical oxygen demand of the leachate from a woodwaste pile.  The proportions of lower molecular weight V F A s (acetic and propionic) decreased gradually from 1999 to 2002 and then increased (Figure 2-5). In general, acetic acid accounted for 3645% of the total V F A s . The concentrations of COD, T & L , and V F A s were significantly, negatively correlated to pH (Figure 2-6), implying that COD was largely influenced by organic acids, such as V F A s and the tannic acid like complexes.  64  • Acetic B Propionic E! Butyric • Valeric DHexanoic 100%  -j  1999  2000  2001  2002  2003  2004  Year  Figure 2-5. Yearly variations of the proportion of individual to total volatile fatty acids.  • Chemical oxygen demand  A Volatile fatty acids  • Tannin & lignin  •  \\  •*  3  2  y =-2b-\2x+ 8EO&C -O.OOLc +8.43 .K = 0.86 2  * y =-4E-10tc + 2E-06* - 0.004c +6.43 3  2  .R =0.96 2  0  2000  4000  6000  8000  10000  Concentration, mg L  12000  14000  16000  18000  1  Figure 2-6. Correlation of pH to monthly concentrations of oxygen-demanding substances of woodwaste leachate.  2.3.4  Nutrient levels  Ammonia, nitrate, nitrite, and orthophosphate concentrations in the woodwaste leachate were very low either relative to the concentrations of C O D , T & L and V F A s (Table 2-1) or in comparison to the nitrogen and phosphorus concentrations in municipal landfill leachate (El-  65  Fadel et al., 1997; Farquhar, 1989), although there was a sharp increase in ammonia nitrogen in late 2004 (Figure 2-3). The concentration of TP in woodwaste leachate was 4.0 mg L" in 1  1999 and 6.6 mg L" in 2004. The concentration of T K N in woodwaste leachate was 47.3 mg 1  L" in 2004. The concentrations of T K N and TP in woodwaste leachate fell into the category 1  of leachate from >20-years-old landfill refuse (Farquhar, 1989).  2.3.5  Seasonal variation  The number of months monitored each year was inadequate to show a complete seasonal variation in woodwaste leachate quality. The monthly air temperature and precipitation were not correlated with monthly average concentrations of COD (R = 0.03, P = 0.76) and V F A s 2  (R = 0.07, P = 0.63), while T & L concentration was significantly correlated to air temperature 2  and precipitation (R = 0.37, P = 0.03) during 1999-2002. The correlations of COD, T & L , and 2  V F A s concentrations to air temperature and precipitation were not improved by replacing precipitation with the cumulative precipitation of the most recent two months.  2.4 2.4.1  Discussion Characteristics of woodwaste leachate  It can take several months for the leachate to arrive at the base of the landfill, depending on the refuse type, compaction, and depth (Farquhar, 1989). A significant correlation of T & L concentration with air temperature and precipitation suggested immediate generation of leachate from the woodwaste pile, probably because of lack of covers, little compaction, and low field capacity of the woodwaste. In general, the "young" woodwaste leachate produced in the pile's placement period was amber, acidic, nutrient-poor, of very high oxygen demand, and very toxic to aquatic life; the "older" leachate in the late closure period had lower oxygen demand and higher ammonia, and became less acidic and darker. This open woodwaste pile contained sawdust and shredded bark and roots, providing greater exposure to physical, chemical and biochemical reactions than woodpile and compacted woodwaste. The leachate from this woodwaste pile, except in the late closure period, therefore, had lower p H and much higher oxygen demand in terms of COD, T & L , or B O D 5 than the leachate from a 2-year-old cedar waste landfill (Peters et al., 1976), leachate from highway entrances and off-ramps filled with woodwaste (Hunter et al., 1993), leachate from woodpiles (Taylor and Carmichael, 2003), bark wastewater (Field et al., 66  1988), stormwater runoff from sawmills (Bailey et al., 1999), and log yard runoff (Woodhouse and Duff, 2004; Zenaitis et al., 2002). The organic carbon of woodwaste leachate was a mixture of lignins, tannins, fatty acids, and other unknown soluble organic compounds. Characterization of wood extractives (Gabrielii et al., 2000; Sun et al., 2001) suggests that hemicellulose was probably among the unknown organic carbon components of wood leachate. The significant correlation of COD, T & L , and V F A s to pH suggested that the woodwaste leachate contained such organic acids as humic and fulvic acids that are similar to tannic acid, in addition to V F A s . The young woodwaste leachate had a smaller BODs:COD ratio (0.33) and a larger T & L ThOD:COD ratio (0.33-0.45), indicating resistance to biodegradation. The woodwaste leachate had a BODs:COD ratio similar to and a T & L ThOD:COD ratio higher than log yard runoff (Woodhouse and Duff, 2004; Zenaitis et al., 2002). The contribution of easily biodegradable V F A s to COD in the late closure period decreased rapidly over time, resulting in a more recalcitrant leachate. V F A s declined faster than T & L and C O D declined in the closure period, suggesting that microbial decomposition within the woodwaste pile reduced the availability of labile organic compounds for percolation. Acetic acid, the main component of V F A s in woodwaste leachate, is a catabolic product of anaerobic fermentation. Other V F A s with even numbers of carbon atoms are formed from acetic acid through condensation reactions; the V F A s with odd numbers of carbon atoms are formed through various combinations of microorganism-induced cleavage and condensation reactions (Forgie, 1988). The woodwaste leachate color responded to the change of pH that was significantly correlated to concentrations of COD, T & L and V F A s . V F A s C2-C6 species are colorless, except that hexanoic acid may be slightly yellow. T & L and humic substances are highly colored compounds that appear to be the major source of woodwaste leachate color. The toxicity of wood leachate is mostly attributed to phenolic compounds (including tannins), lignins, tropolones, terpenes, zinc, lignans, low pH, and other unidentified constituents (Bailey et al., 1999; Borga et al., 1996; Field et al., 1988; Peters et al., 1976; Taylor et al., 1996; Taylor and Carmichael, 2003; Temmink et al., 1989). With reference to the toxicity bioassays  67  by these previous studies, T & L and low pH were most likely responsible for toxicity of the woodwaste leachate.  2.4.2  Temporal variation of chemical properties  The temporal variation of woodwaste leachate quality showed a pattern (Figure 2-3) similar to, but shorter than, that of municipal landfill leachate (Farquhar, 1989). It may have reached peak concentrations in the earlier months of leaching (Farquhar, 1989). In the placement period, the leachate came from both fresh and leached woodwaste, and its strength varied at high concentrations, likely with the weather conditions and the amount of new waste placed. The woodwaste pile in this study had no cover and the waste was placed without compaction, undergoing a faster leaching process in comparison to aspen woodpiles (Taylor and Carmichael, 2003) and municipal landfills (Farquhar, 1989). Obvious decreases in the easily biodegradable V F A s and recalcitrant T & L occurred 1.5 years after placement of new woodwaste stopped. Microbial decomposition contributes to the characteristics of landfill leachate (Farquhar, 1989; Forgie, 1988). When fresh woodwaste was continually piled up, hydrolysis and fermentation solubilized waste components and produced organic acids and alcohols. As rainfall percolated through the waste, contaminants were mobilized into the liquid phase through dissociation and suspension from the stationary phase, thus producing a concentrated leachate. As the low DO and redox potential of the woodwaste leachate in the leachate pool suggested, anaerobic decomposition might have dominated the degradation process in the woodwaste pile in the closure period. Anaerobic reactions decomposed the biodegradable organic matter to methane, CO2 and metabolic intermediates, and left higher molecular weight organics, including lignins and tannins. With development of anaerobic decomposition, the ratios of BOD5 to COD and V F A s ThOD to C O D decreased. Like in "older" municipal landfills (Carley and Mavinic, 1991; Henry, 1985), the anaerobic reducing environment within the woodwaste pile caused an increase in the concentration of ammonia. With reference to the leachate temperature, the woodwaste pile was likely far from optimum for the anaerobic processes (Metcalf & Eddy, 2003). Higher temperature in the summers (Figure 2-2) would improve microbial metabolism in the woodwaste pile to produce concentrated leachate. In the wet winters, frequent rainfall (Figure 2-2) would raise moisture content of the woodwaste and likely enhance microbial activity and leaching of soluble 68  compounds. T & L is more resistant to biodegradation than V F A s and the aggregate parameter COD. The concentration of T & L in the leachate pool should be close to the fresh woodwaste leachate. The significant correlation of T & L concentration to air temperature and precipitation confirmed the effect of environmental factors on the woodwaste leachate quality. Taylor and Carmichael (2003) reported a significant BOD decline due to 10-12 d of storing wood leachate in a catch basin under anoxic conditions. V F A s and COD in the leachate pool were probably reduced by biodegradation during the warm summer periods compared to the colder winter periods. Subsequently, correlations of V F A s and C O D to air temperature and precipitation diminished. The seasonal variation in monitoring results was a mixed effect of precipitation and air temperature on the generation of woodwaste leachate in the woodwaste pile as well as dilution or concentration and degradation in the leachate pool. Consequently, no simple pattern of seasonal variation was identified in leachate quality.  2.4.3 Implications for selection of treatment processes Treatment options for leachate include aerobic and anaerobic biological processes, and various types of physical-chemical treatment processes (Forgie, 1988). Very low concentrations of suspended solids suggest that sedimentation and flocculation are not practical options for treatment of woodwaste leachate. High concentrations of natural polymers, such as tannins and lignins, imply removal potential by adsorption. High concentrations of V F A s in the woodwaste leachate could be efficiently removed by biological processes. In the cases Forgie (1988) reviewed for leachate treatment, aerobic biological systems operated better with a BODs:N:P ratio of about 100:5:1. The "young" woodwaste leachate had a BOD :(inorganic N):(orthophosphate P) ratio of 1195:0.4:1 on average in 1999, suggesting 5  the need for nutrient supplementation for aerobic treatment of woodwaste leachate. Biological treatment processes rely on the establishment and activity of a mixture of microorganisms, which may be inhibited by toxic components of the waste stream. Inhibition on bacterial metabolism by aspen leachate has been found at wood leachate concentrations below 0.3% (Taylor et al., 1996). Median effective inhibition concentrations derived from bacterial luminescence assays were <10% of full-strength aspen wood leachate and more often approached 1% (Taylor and Carmichael, 2003), and 1.9-91% (v/v) of contaminated log yard runoff (Woodhouse and Duff, 2004; Zenaitis et al., 2002). The 50% inhibitory concentration of bark tannins to methanogens averaged approximately 600 mg L" C O D or 350 mg L ' 1  69  1  tannin solids (Field et al., 1988). Low pH and high concentration of T & L are probably the major inhibitory factors for biological treatment of woodwaste leachate. Conventional biological leachate treatment is most appropriate when the BODsiCOD ratio is high (>0.4) and the molecular weight of the majority of the organics is less than 500 (Forgie, 1988). The woodwaste leachate had a lower B O D : C O D ratio (0.14-0.33), and T & L that 5  accounted for 33-47% of COD has a molecular weight probably higher than or similar to 1701 for tannic acid. With regard to low pH, low B O D : C O D ratio, high BOD :N:P ratio, and 5  5  potential toxicity to microorganisms, anaerobic treatment systems with a long HRT, such as CWs (Hunter et al., 1993; Tao and Hall, 2004), are preferred for treatment of woodwaste leachate.  2.5  Conclusions  A woodwaste pile without cover and compaction generated a leachate with quality changing faster over the years than that of municipal landfills. The mixed effect of air temperature and precipitation resulted in a woodwaste leachate quality without a simple pattern of seasonal change. Overall, the woodwaste leachate was highly colored, acidic, of very high oxygen demand, nutrient-poor, and very toxic to aquatic life. The concentrations of COD, T & L , and V F A s changed over years as the woodwaste pile developed, especially after closure. The treatment processes should consider flexibility in influent quality with regard to the yearly variable nature of woodwaste leachate. The ratio of V F A s ThOD:COD increased as the woodwaste pile developed, and decreased as it stopped receiving new waste. T & L accounted for more than one-third (33-45%) of COD on an annual basis. The leachate characters of recalcitrance, lack of nutrients, and possible inhibition on microbial activity suggest a preference for CWs to treat woodwaste leachate at a long HRT.  2.6 References A P H A et al. 1999. Standard Methods for the Examination of Water and Wastewater, 20th edition. A P H A , A W W A , and WEF.  70  Bailey, H.C., J.R. Elphick, A . Potter, E. Chao, D. Konasewich, and J.B. Zak. 1999. Causes of toxicity in stormwater runoff from sawmills. Environmental Toxicology and Chemistry, 18: 1485-1491. Borga, P., T. Elowson, and K . Liukko. 1996. Environmental loads from water-sprinkled softwood timber: 2. influence of tree species and water characteristics on wastewater discharges. Environmental Toxicology and Chemistry, 15: 1445-1454. Carley, B.N., and D.S. Mavinic. 1991. The effects of external carbon loading on nitrification and denitrification of a high ammonia landfill leachate. Water Environment Research, 63: 5159. El-Fadel, M . , A . N . Findikakis, and J.O. Leckie. 1997. Modeling leachate generation and transport in solid waste landfills. Environmental Technology, 18: 669-686. Environment Canada. 1990. Biological Test Method: Acute Lethality Test Using Rainbow Trout. Environmental Protection Series EPA1/RM/9, Ottawa, O N , Canada. Environment Canada. 2005. National Climate Data and Information Archive, last reviewed on January 26, 2004. http://www.climate.weatheroffice.ec.gc.caAVelcome_e.html. Last access on March 6, 2005. Farquhar, G.J. 1989. Leachate: production and characterization. Canadian Journal of Civil Engineering, 16: 317-325. Field, J.A., M.J.H. Leyendeckers, R.S. Alvarez, G . Lettinga, and L . H . A . Habets. 1988. The methanogenic toxicity of bark tannins and the anaerobic biodegradability of water soluble bark matter. Water Science and Technology, 20: 219-240. Forgie, D.J.L. 1988. Selection of the most appropriate leachate treatment methods Part 1: a review of potential biological leachate treatment methods. Water Pollution Research Journal of Canada, 23: 308-328. Gabrielii, I., P. Gatenholm, W . G . Glasser, R . K . Jain, and L . Kenne. 2000. Separation, characterization and hydrogel-formation of hemicellulose from aspen wood. Carbohydrate Polymers, 43: 367-374. Henry, G.J. 1985. New developments in landfill leachate treatment. Water Pollution Research Journal of Canada, 20: 1-9. 71  Hunter, R., A . E . Birkbeck, and G. Coombs. 1993. Innovative marsh treatment systems for control of leachate and fish hatchery wastewaters. In: G . A . Moshiri (ed.), Constructed Wetlands for Water Quality Improvements. Lewis Publishers: Baca Raton, FL. pp477-484. Metcalf & Eddy. 2003. Wastewater Engineering - Treatment and Reuse, 4th edition. McGraw-Hill: New York, N Y . p984-987, 1506. Peters, G.B., H.J. Dawson, B.F. Hrutfiord, and R.R. Whitney. 1976. Aqueous leachate from western red cedar: effects on some aquatic organisms. Journal of the Fisheries Research Board of Canada, 33: 2703-2709. Sun, R.C., J . M . Fang, J. Tomkinson, Z.C. Geng, and J.C. L i u . 2001. Fractional isolation, physico-chemical characterization and homogeneous esterification of hemicelluloses from fast-growing poplar wood. Carbohydrate Polymers, 44: 29-39. Tao, W., K . J . Hall. 2004. Dynamics and influencing factors of heterotrophic bacterial utilization of acetate in constructed wetlands treating woodwaste leachate. Water Research, 38: 3442-3448. Taylor, B.R., J.S. Goudey, N . B . Carmichael. 1996. Toxicity of aspen wood leachate to aquatic life: laboratory studies. Environmental Toxicology and Chemistry, 15: 150-159. Taylor, B.R., and N . B . Carmichael. 2003. Toxicity and chemistry of aspen wood leachate to aquatic life: field study. Environmental Toxicology and Chemistry, 22: 2048-2056. Temmink, J.H.M., J.A. Field, J.C. van Haastrecht, and R . C . M . Merkelbach. 1989. Acute and sub-acute toxicity of bark tannins in carp (Cyprinus carpio L.). Water Research, 23: 341-344. Woodhouse, C , and S.J.B. Duff. 2004. Treatment of log yard runoff in an aerobic trickling filter. Water Quality Research Journal of Canada, 39: 232-238. Zenaitis, M . G . , H . Sandhu, and S.J.B. Duff. 2002. Combined biological and ozone treatment of log yard runoff. Water Research, 36: 2053-2061.  72  3  LABORATORY STUDY ON POTENTIAL MECHANISMS FOR TREATMENT OF WOODWASTE L E A C H A T E IN WETLANDS*  3.1  Introduction  Leachate may be induced by precipitation, runoff or sprinkling water in woodpiles, log yards, wood product piles, and woodwaste disposal sites. Wood leachate has been characterized as dark, acidic, of very high oxygen demand, and toxic to aquatic organisms (Bailey et al., 1999; Field et al., 1988; Hunter et a l , 1993; Peters et al., 1976; Taylor et al., 1996; Taylor and Carmichael, 2003; Temmink et al., 1989; Woodhouse and Duff, 2004; Zenaitis et al., 2002). Weak organic acids, such as fatty acids, are mainly responsible for the low pH. Acute toxicity is usually attributed to tannin, lignin, tropolone, terpene, lignan, and low pH. Removal of organic carbon, therefore, is the focus of wood leachate treatment. The organic carbon of wood leachate is a mixture of T & L , VFAs, and other soluble organic compounds (Field et al., 1988; Taylor et a l , 1996; Taylor and Carmichael, 2003; Woodhouse and Duff, 2004; Zenaitis et al., 2002). Field et al. (1988) found that anaerobic digestion helped degrade soluble COD from bark. Log yard runoff has been effectively treated by a bench-scale aerobic batch bioreactor (Zenaitis et al., 2002), a laboratory-scale aerobic trickling filter (Woodhouse and Duff, 2004), and ozonation (Zenaitis and Duff, 2002). Because of high capital cost, the intermittent nature of leachate quality and quantity, and specialized operating skills required by conventional treatment plants, vegetated CWs and non-vegetated ponds (Frankowski, 2000; Hunter et al., 1993; Masbough, 2002; Taylor and Carmichael, 2003) become attractive alternatives for onsite wood leachate treatment. CWs can take advantage of the cheaper land near the sites of logging, wood processing and woodwaste disposal, do not need sophisticated management and high energy input, and accept the variability of leachate in quantity and quality. Frankowski (2000), Hunter et al. (1993), and Masbough (2002) have proved the effectiveness of CWs in  * A version of this chapter has been submitted for publication. Tao, W., K . Hall, and E. Hall. Laboratory study on potential mechanisms for treatment of woodwaste leachate in constructed wetlands. Journal of Environmental Engineering and Science, in review. 73  reduction of COD, B O D , T & L , V F A s , and acute toxicity of wood leachate. Taylor and 5  Carmichael (2003) reported a decline of 5-day B O D in aspen leachate over 10-12 d of storing in a leachate catch basin. CWs usually have a large air-water interface, soil at the bottom, and emergent or submerged plants. Treatment in CWs may rely upon a combination of biological, physical, and chemical processes that occur at natural rates (USEPA, 2000). However, few studies have investigated individual processes that may happen in CWs for wood leachate treatment. Tao and Hall (2004) found that bacterial activity in the CWs treating wood leachate was influenced by the concentrations of organic substrates and inorganic nutrients, and that sedimentary bacteria contributed to the majority of total heterotrophic acetate uptake. Taylor et al. (1996) found that aeration facilitated degradation over 65 d of aspen leachate storage in vats. Performance assessment and optimal design of CWs have to be based on further understanding of the physical, chemical and biological processes occurring therein. The woodwaste leachate of concern in this study had a very high concentration of V F A s , which are easily biodegradable. Moreover, CWs may cost-effectively provide a retention time long enough for acclimatization and establishment of acidophiles to decompose the recalcitrant and easily biodegradable organic matter. Biological degradation, hence, may play a key role in woodwaste leachate treatment in CWs. However, ecological maturation of CWs may take a few months or years (Mitsch, 1992; Werker et al., 2002). Initially, when the aquatic ecological community has not fully developed, soil adsorption and some other mechanisms may dominate organic carbon removal. The woodwaste leachate in the present study emanated a strong odor of organic acids, so volatilization should contribute to removal of organic carbon from the leachate. Typical natural polymers, such as lignin and tannin, may be adsorbed readily on solid surfaces of sediment and aquatic plants. Each polymeric molecule can have many functional groups (hydroxyl, carboxyl, etc.) that potentially facilitate adsorption (Stumm and Morgan, 1996). With the high T & L concentration, soil adsorption should be one of the main mechanisms of woodwaste leachate treatment in CWs. Less than 43 mg L" suspended solids and >98% soluble COD meant that little treatment would be achieved 1  through flocculation and sedimentation. The dark color of woodwaste leachate curtailed penetration of solar U V and accordingly, minimized the possibility of photodegradation.  74  Laboratory verification of the promising treatment mechanisms would provide a sound basis for design and management of treatment wetlands. Solely based on monitoring data of overall treatment performance, it is impossible to distinguish the contributions of individual mechanisms. Laboratory tests with woodwaste leachate were then conducted on volatilization, soil adsorption, and biological treatment to examine their treatment potential, kinetics and regulating factors. This chapter also discusses the implications of laboratory test results to wood leachate treatment in CWs.  3.2 3.2.1  Materials and Methods Woodwaste leachate and chemical analysis  The concentrations of COD, T & L and V F A s varied over time in the leachate pool (Figure 23). The woodwaste leachate used for volatilization, adsorption and biodegradation studies was collected at different times, and the quality is given separately in the relevant sections. Solids, pH, temperature,  DO, C O D , T & L , V F A s , ammonia, nitrate plus nitrite, and  orthophosphate were analyzed with the same methods and instruments as described in Section 2.2.2. Method detection limits and errors of field sampling and subsanipling were checked to ensure data quality (Appendix 2).  3.2.2  Jar test for volatilization  Woodwaste leachate and its dilutions were held in four 1-L wide-mouth open jars at 23°C in a fumehood. The raw leachate (prefiltered through 50-urn mesh) had 13.1 g L" COD, 2.8 g L" 1  1  V F A s , and p H 3.7. Dilution was made with filter-sterilized domestic sewage that had 102 mg L" C O D and 6 mg L" V F A s . To enhance volatilization, N2-sparging with a submerged fine1  1  bubble diffuser was applied to the jars that contained 40 mL leachate plus 760 mL sewage, 160 mL leachate plus 640 mL sewage, and 800 mL leachate. Nitrogen gas flow was adjusted to show significant surface agitation. Another jar that contained 800 mL leachate was kept stagnant. The jars were sampled initially and after 0.5 h, 1 h, 2 h, 4 h, 8 h, and 12 h for analysis of COD and V F A s . pH was measured while sampling.  3.2.3  Batch soil adsorption tests  Batch tests were conducted to evaluate the adsorption kinetics, isotherm, and temperature effect. Filter-sterilized woodwaste leachate was diluted with sterile water to prepare different 75  concentrations of adsorbates. Adsorbent was the silt loam collected with corers from a pilotscale C W receiving woodwaste leachate. Three sediment cores (8 cm long; without the soft surface layer) were freeze-dried, sifted through 2-mm mesh, and homogenized. The dried silt loam contained 18% sand (0.05-2 mm), 0.3% moisture and 5.1% organics (loss on ignition at 550°C for 6 h). Dilute leachate had been treated in the pilot-scale CWs at water depths around 25 cm. Therefore, 40 g adsorbent (dry weight) per L leachate was set for batch adsorption tests, based on 25 cm water overlying 2 cm effective sediment in the CWs. Three combinations of adsorbate and adsorbent were tested first with freeze-dried silt loam at initial T & L concentrations of 179 mg L" , 712 mg L" , and 3540 mg L" . The freeze-dried silt 1  1  1  loam was further autoclaved in a thin layer at 125°C for 50 min. Three combinations with initial T & L concentrations of 79 mg L" , 276 mg L" , and 1494 mg L ' were tested with the 1  1  1  autoclaved silt loam, plus one at initial T & L concentration of 291 mg L" with the non1  autoclaved freeze-dried silt loam. No significant difference between the two combinations that had initial T & L concentrations of 276 mg L" and 291 mg L" was found in the remaining 1  1  aqueous T & L concentration (P = 0.93) and C O D concentration (P = 0.86), verifying a negligible contribution of sedimentary microorganisms to the loss of adsorbate from the aqueous phase. The results of these combinations were then pooled and processed together. The batch adsorption tests were conducted in closed glass jars in a horizontal Innova 4230 incubator/shaker (New Brunswick Scientific, Edison, NJ) in the dark at 150 rpm and 23°C. The initial pH was between 3.9 and 4.8, varying with dilution ratio. Initially and after shaking (for up to 96 h, depending on the equilibration trend), 5 mL mixed sample was taken out of each jar and centrifuged at 5600xg for 5 min. The supernatant was analyzed for remaining T & L and COD to determine adsorption kinetics and to derive adsorption isotherms. The monthly air temperature at the leachate generation site varied between 3-20°C over the study period (Figure 2-2). Batch adsorption was further tested at a temperature of 15°C with 0 g, 1 g, 2 g, 3 g, 5 g, 7 g, and 10 g silt loam added to 100 mL filter-sterilized leachate, which contained 1.35 g L" T & L . The initial pH was 4.2. After shaking for 72 h in the incubator, the 1  samples were centrifuged at 5600xg for 5 min and the supernatant was analyzed for T & L . The double-log plots of aqueous concentration versus shaking time were fitted with a straight line (R = 0.80-0.98), which was used to estimate the equilibration time and aqueous  76  concentration at equilibrium. Equilibrium was assumed to have been reached when a ^ % concentration decrease was found over 6 h from the fitting curve. The solid phase concentration (mass of adsorbate adsorbed on the adsorbent) was calculated from the ratio of silt loam dry weight to liquid volume and the difference between the initial and remaining aqueous concentrations. Adsorption capacity was determined as the theoretically maximum solid phase concentration using the linear modification of Langmuir adsorption isotherm equation (Stumm and Morgan, 1996). Reduction efficiency was calculated as the percentage of concentration reduction at equilibrium over initial aqueous concentration.  3.2.4  Setup and operation of bench-scale bioreactors  Previous studies (Frankowski, 2000; Masbough, 2002) demonstrated that aerobic microcosm wetlands performed better than anaerobic pilot-scale CWs for treatment of the woodwaste leachate. Bench-scale aerated and anaerobic reactors were set up in parallel (Figure 3-1). The reactors were operated as complete-mix, flow-through systems at room temperature (23±1°C). Each cylindrical Plexiglas reactor had a working volume of 3.4 L (15.5 cm in inner diameter), including a 0.8-L internal cylindrical clarifier attached on the reactor wall, from which the effluent overflowed. The clarifier had small holes arranged around its wall to keep a uniform hydraulic exchange with the mixed liquor and quiescent conditions inside for clarification. Two air diffusers were placed close to the bottom of the aerated reactor for aeration and mixing. A mechanical mixer was used to provide gentle mixing (60 rpm) for the anaerobic reactor. The two reactors were fed continuously with the same dilute woodwaste leachate at the same flow rate. The raw leachate had 4.16-11.38 g L ' COD, 1.48-3.64 g L" T & L , and 0.82-2.74 g 1  1  L" V F A s . Each batch of leachate was diluted with 1-4 times prefiltered (through 50-u.m mesh) 1  domestic sewage and stored at 4°C to keep a relatively constant influent quality. The prefiltered sewage had a pH of 6.6 and 110 mg L" COD. 1  At the beginning of operation, the mixed liquor pH was adjusted to 7.5 by adding 15 N NaOH. In the cases reviewed (Forgie, 1988) for leachate treatment, aerobic biological systems operated best with a BODs:N:P ratio of about 100:5:1. In the present study, the woodwaste leachate had a COD:(inorganic N):(orthophosphate P) ratio of 4743:1.1:1. Therefore, NH4CI and KH2PO4 were added to the influent and the initial mixed liquor to make nutrient supply in  77  excess relative to the BOD :N:P ratio of 100:5:1, based on the average B O D : C O D ratio of 5  5  0.33 in the woodwaste leachate. The leachate contained 83 mg Ca L" , 44 mg M g L" , and 75 1  1  mg Fe L" (Appendix 1). The sewage used for dilution should also have added some nutrients. 1  2-head pump  Effluent  Aerated reactor  IS Effluent  Anaerobic reactor  Influent  Figure 3-1. Sketch of bench-scale aerated (left) and anaerobic (right) complete-mix reactors.  For the systems to establish microorganisms acclimatized to the woodwaste leachate, the mixed liquor was seeded initially with 2.5 mL of a sediment slurry from the pilot-scale CWs treating woodwaste leachate and 5 mL more slurry on the 19th day, three hours after which the mixed liquor volatile suspended solids (MLVSS) were 80 mg L" in the aerated reactor and 70 1  mg L" in the anaerobic reactor. The mixed liquor was monitored for temperature, D O , 1  M L V S S , and pH, and the influent and effluent were analyzed for COD, T & L , and V F A s every 4-7 d. The operation was controlled at a HRT of 6.7 d from 28 November 2002 to 22 January 2003. In the second run (29 January to 14 March 2003), the mixed liquor was initially made of 2.4 L influent and 1 L activated sludge from a chem-thermal mechanical pulp mill effluent treatment plant. The initial M L V S S was 2360 mg L" in the aerated reactor and 2320 mg L" in 1  1  the anaerobic reactor. The mixed liquor, influent, and effluent were monitored after 19 d of acclimatization at a frequency of once every 7-10 d. The HRT was around 6.3 d in the second run. Efficiency was calculated as concentration reduction percentage. The first-order reaction rate constant was calculated using the mass-balance model for a "complete-mix reactor with reaction" under steady state (Metcalf & Eddy, 2003). 78  3.2.5  Statistical analysis  The difference between means was assessed by a paired /-test. Regression analysis gave the coefficient of determination, i? . The value of n stands for the number of samples taken at each 2  time point. Correlations and differences were considered significant at a probability value of P <0.05.  3.3 Results 3.3.1  Volatilization  Volatilization in terms of reduction of COD and V F A s concentrations was not measured over 12 h of monitoring in raw or dilute woodwaste leachate under stagnant and N2-sparging conditions at 23°C (Appendix 3). The coefficient of variation was 1-7% for COD and V F A s concentrations. No significant p H change was noticed at >20% leachate, which indicated that the p H was probably controlled by organic acids such as V F A s . The loss of CO2 due to N2sparging of the 5% leachate raised p H by nearly 2 units after 4 h.  3.3.2  Soil adsorption  Figure 3-2 demonstrates the kinetics of T & L and COD adsorption on silt loam (see Appendix 4 for the others). Significant adsorption occurred in the first 12 h. The equilibration time of T & L adsorption, 35-78 h, decreased linearly (R = 0.90, P = 0.00) with increasing initial 2  aqueous concentration. The equilibration time of COD adsorption, 12-43 h, decreased linearly (R = 0.72, P = 0.03) with increasing initial aqueous concentration. 2  Table 3-1 summarizes the results of batch adsorption tests at 23°C. Reduction efficiency, 2242% T & L and 6-27% COD, decreased linearly with increasing initial aqueous concentration (Figure 3-3). C O D and T & L adsorption on the silt loam followed the Langmuir isotherm (Figure 3-4). The Langmuir adsorption capacity was estimated at 36 mg T & L g' silt loam. 1  The adsorption capacity of COD could not be confidently derived because of the insignificant P value for the intercept (Figure 3-4), which is the reciprocal of adsorption capacity. The aqueous T & L concentrations at equilibrium in the batch test at 15°C ended up in a small range (1.04-1.24 g L" ) with reference to the previous adsorption isotherm (Figure 3-4), and 1  exhibited a linear relationship with the solid phase concentration (R = 0.81, P = 0.01). 2  Reduction efficiency, 8-23%, increased linearly with increasing amount of adsorbent at the 79  same initial T & L concentration (R - 0.88, P = 0.00). At the same adsorbent:adsorbate ratio of 2  (40 g):(l L x 1.35 g T & L L" ), T & L removal efficiency at 15°C (15%) was 0.44 times that at 1  23°C (34%), and solid phase concentration at 15°C (5.4 mg g" ) was 0.44 times that at 23°C 1  (12.2 m g g ) . 1  14000 6 c o  60  13000  e co  12000 o  e o o  11000  CO  10000  3 O u 3  4000  Chemical oxygen demand  60  e  /c =0.97 2  u o C o o  CO  3 O u 3  Raw leachate  9000 48  72  3200 H — •  2400  log(y )=-0.0231og(;c +1>+4.132  24  Tannin & lignin  7? =0.86 2  1600 800  Raw leachate  0  96  0  24  Shaking time, h  250 60  B  100  Chemical oxygen demand  I?  200  a o  o G  o u CO  3 O o 3  150  log(y )=-0.0741og£c+l>+2.316  100  e u  R =O.S7 2  o C o  CJ CO  50  3 o 3  Most diluted leachate  0 0  12  24  36  48  48  72  96  Shaking time, h  C  o a a>  •  log(y )=-0.0631og(x +l)+3.540  60  72  Tannin & lignin  80 60 40  o"~^-©_ log(y)=-0.1241og(x+l)+1.894 R =0.9% 2  20  Most diluted leachate  0 12  Shaking time, h  24  36  48  60  72  Shaking time, h  Figure 3-2. Illustration of adsorption kinetics curves of silt loam in woodwaste leachate at 23°C.  80  Table 3-1. Adsorption characteristics of silt loam in woodwaste leachate at 23°C and reduction efficiency at equilibrium Leachate and its dilutions  Raw  Dl  D2  D3  D4  D5  3540  1490  712  284  179  79  35  60  58  74  78  72  2780  994  423  173  105  46  19.1  12.5  7.2  2.8  1.9  0.8  22  33  41  39  41  42  13570  4190  2730  874  762  216  12  23  29  27  43  42  12790  3640  2270  745  584  157  19.7  13.8  11.4  3.2  4.5  1.5  6  13  17  15  23  27  1000  2000  3000  4000  Tannin and lignin Initial aqueous concentration, mg L"  1  Equilibration time, h Equilibrium aqueous concentration, mg L"  1  Solid phase concentration, mg g"  1  Reduction efficiency, % Chemical oxygen demand Initial aqueous concentration, mg L"  1  Equilibration time, h Equilibrium aqueous concentration, mg L"  1  Solid phase concentration, mg g"  1  Reduction efficiency, %  0  3000  6000  9000  12000 15000  initial concentration (Co), mg L"  0  Initial concentration (Co), mg L  1  1  Figure 3-3. Correlation of reduction efficiency via silt loam adsorption at 23°C to initial aqueous concentration. 81  60 60  6 c o  25 Tannin & lignin 20 15 10  JS  o on  0  5 0  1  0  3000 6000 9000 12000 15000 Aqueous concentration (Ce), mg L  1  1  1  1  500 1000 1500 2000 2500 3000 Aqueous concentration (Ce), mg L  1  0.8  H  1.5  Chemical oxygen demand  1.2  0.6  1  Tannin & lignin  H  '60  S so 0.4 -5"  _  l/^=O.10/Ce-K).O7  Mq =0.05/Ce +0.03  fl =0.95, P slope =0.00, P intercept =0-08 2  0.2 0.0 -04  4 VCe,Lg  R =0.99, 2  P slope =0-00, P intercept =0-02  6  10  l  1  20  25  l/Ce,Lg  Figure 3-4. Adsorption isotherm (upper) and capacity estimation with Langmuir equation (lower) at 23°C.  3.3.3 Biological degradation The influent characteristics for the first and second biodegradation runs were respectively pH 4.7 ± 0.3 and 4.3 ± 0 . 3 , COD 2480 ± 363 and 2080 ± 122 mg L" , T & L 749 ± 74 and 639 ± 53 1  mg L" , and V F A s 418 ± 79 and 285 ± 6 1 mg L ' . The operating parameters are summarized in 1  1  Table 3-2. The M L V S S in the first run was low, and effluent suspended solids were less than 40 mg L" . Higher M L V S S was established in the second run, and effluent suspended solids 1  were less than 80 mg L" . Based on the average BODs:COD ratio of the "young" woodwaste 1  82  leachate and M L V S S measurements, the food-to-microorganism ratio was estimated at 4 g B O D g" M L V S S d" in the first run and 0.4 g BOD g" M L V S S d" in the second run. 1  1  1  1  Table 3-2. Summary of operating parameters of the bench-scale reactors treating woodwaste leachate after 19 d of acclimatization Sampling period  Seed source Type of reactor  17 December 2002-22  17 February - 14 March  January 2003 (n = 5-6)  2003(n = 4)  Wetland sediment  Activated sludge  Aerated  Anaerobic  Aerated  Anaerobic  6.7  6.7  6.3  6.3  23.8+0.6  24.4+0.6  23.5+0.5  24.0+0.7  6.2+1.1  <0.2  8.0+0.3  <0.2  pH  6.7+0.5  5.9+0.3  6.3+0.7  6.3+0.4  Mixed liquor volatile suspended  52+21  52+18  617+240  583+208  Hydraulic retention time, d Temperature, °C Dissolved oxygen, mg L"  solids, mg L"  1  1  Mean ± SD.  The treatment performance is presented for C O D , T & L and V F A s in Figure 3-5. The performance assessment is summarized in Table 3-3. The aerated reactor was much more efficient than the anaerobic reactor in removal of COD and T & L . Up to nearly 500 mg L"  1  V F A s in the influent could be completely removed through the aerated reactor, raising p H by 2 units. The anaerobic reactor could reduce about 300 mg L" V F A s from the influent. The 1  aerated reactor exhibited a reaction rate constant and reduction efficiency for COD that were similar to those for T & L , and much lower than those for V F A s . The anaerobic reactor exhibited a reaction rate constant and reduction efficiency for C O D higher than those for T & L , and much lower than those for VFAs.  83  Influent o Aerated effluent A Anaerobic effluent  Influent o Aerated effluent A Anaerobic effluent  3000  11  3000  Seeded with activated sludge  oo  f  a  2400  •  8 1800 g S 1200 x  •  o  o  °  600 J2  U  14 21  28  35  42  0  Ii MI I.  Influent o Aerated effluent A Anaerobic effluent  5 .S  o  o  in .1 In  42  49  Influent o Aerated effluent A Anaerobic effluent 1000  1000  t  A  14 21 28 35 Elapsed days  49 56  Elapsed days  t B  1,  A UA  A  800  Seeded with activated sludge  800  J  00  E 6 0 0  .S  ^  J>  400 o °  200 0  °  o  o  .5  o  M  Seeded with sediment I  IMMMI  14 21  600 400 o  200  T | VTT-n I | l I I I I I I I I I I I I [ I I 1 I I I j I I 1 I I 1 I I  i . . . . . . 11  28  35  o  42  0  49 56  7  14  21  28  35  42  49  Elapsed days  Elapsed days  Influent o Aerated effluent A Anaerobic effluent  • Influent o Aerated effluent A Anaerobic effluent  500 op  1°  1 "3 >  O >  14  21  28  35  42  7  49 56  14  21  28  Elapsed days  Elapsed days  Figure 3-5. Treatment performance of the bench-scale complete-mix bioreactors.  84  35  42  49  Small amount of sediment seed did not develop to a high M L V S S concentration even at a high food-to-microorganism ratio in the first run; and a large amount of activated sludge seed did not sustain a high M L V S S concentration at a lower food-to-microorganism ratio in the second run. Reduction efficiency did not differ much due to much higher food-to-microorganism ratio in the first run from that in the second run. The mass reduction rates in the first run with a low M L V S S were even larger than those in the second run with a much higher M L V S S , except for T & L reduction in the anaerobic reactor.  Table 3-3. Summary of performance of bioreactors for treatment of woodwaste leachate after 19 d of acclimatization Sampling period  Type of reactor  17 December 2002-22  17 Feburary-14 March  January 2003 (« = 5-7)  2003 (n = 4)  Aerated  Anaerobic  Aerated  Anaerobic  COD loading rate, g m" d"  370±54  370±54  298±17  298±17  T & L loading rate, g m" d"  112+11  112+11  91±8  91±8  62±12  62+12  41±9  41±9  COD reduction rate, g m" d"  249±51  106±54  205±22  95±32  T & L reduction rate, g m" d"  73+13  16+8  68±6  20±13  V F A s reduction rate, g m" d"  62+12  49±7  44±9  43±11  COD reduction efficiency, %  67±6  28±12  62±5  29±9  T & L reduction efficiency, %  65±6  13+5  67+1  19+12  V F A s reduction efficiency, %  100±0  80±7  98±1  94±5  Reaction rate constant of COD, d"  0.32±0.09  0.06±0.03  0.26±0.06  0.07+0.03  Reaction rate constant of T & L , d"  0.29+0.08  0.03±0.01  0.32+0.01  0.04±0.03  19±3  0.68±0.35  7.0±1.4  2.3±1.4  3  1  3  1  V F A s loading rate, g m" d" 3  1  3  1  3  1  3  1  1  1  Reaction rate constant of V F A s , d"  1  Mean ± SD. 85  3.4  Discussion  3.4.1  Volatility  The volatilization rate depends upon the mass transfer coefficient and the difference between the liquid concentration and the saturation concentration (Metcalf & Eddy, 2003). Quiescence and N2-sparging were supposed to reflect the effect of mixing intensity. Different dilution factors or initial V F A s concentrations imposed different driving forces for volatilization. In comparison to volatile and semi-volatile organic compounds (Metcalf & Eddy, 2003), V F A s C2-C6 have low air saturation concentration and very high water solubility (CCOHS, 2003). These V F A s have Henry's constants far beyond 100, representing air-phase controlled transfer of relatively involatile compounds (Stumm and Morgan, 1996). For example, acetic acid has a Henry's constant of 1.84xl0 at 20°C. These large Henry's constants explain the 4  insignificance of V F A s volatilization from the woodwaste leachate. Because no loss of COD was observed in the volatilization test, the strong odor that emanated from the woodwaste leachate must have been due to organic compounds that have very low odor thresholds. The petroleum-like odor of wood-residue leachate is primarily due to the presence of phenolic compounds and volatile oils (McNeeley et al., 1979). Different wood species can confer different, characteristic odors.  3.4.2  Soil adsorption capacity  Theoretically, 1 g tannic acid equivalent of T & L has an oxygen demand of 1.24 g. Comparison of the solid phase concentrations of T & L and COD (Table 3-1) reveals that T & L was responsible for the majority of C O D sorbed on silt loam from woodwaste leachate, especially at higher solid phase concentrations. Equilibration of T & L adsorption on silt loam occurred in less than 4 d, depending on the initial concentration. The short equilibrium time implies that adsorption has an immediate effect on treatment performance relative to the generally long HRT in CWs, and that sediment adsorption affects treatment mainly during initial and transitional operating periods. The adsorption capacity of powdered activated carbon was approximately 13.8 mg tannin g"  1  and 4.4 mg lignin g" after 5-h contact (Seo et al., 1997). The adsorption capacity of activated 1  charcoal was 0.44 mg tannin g" and 0.42 mg lignin g" after 1-h contact (Mohan and 1  1  86  Karthikeyan, 1997). Zirconium pillared clay had a tannin adsorption capacity of 78 mg g' (Vinod and Anirudhan, 2002). Dentel et al. (1995) achieved a tannin adsorption capacity of 40-250 mg g" with organoclays. Soils as used in the present study are the most common 1  media in CWs, and they represent low cost adsorbents. Compared to the adsorption capacity of activated carbon, activated charcoal, and clays, the silt loam in the present study had a higher adsorption capacity. Navia et al. (2003) observed a reduction efficiency of 67-74% COD and 71-87% T & L by adsorption in allophanic soil columns (17.7 cm long). In  practice,  the  reduction  efficiency  of adsorption  in CWs is affected  by  the  adsorbentadsorbate ratio. CWs are rarely designed to dredge sediments and replace fresh soil or other solid materials. However, water depth in a C W may vary frequently due to a variety of factors. Accordingly, reduction efficiency through adsorption would vary temporally. The actual reduction efficiency would differ from the results in this laboratory study, which assumed 25 cm water depth in CWs. Moreover, it is still not known to what extent desorption could happen when the operating conditions become unfavorable to adsorption, for example, a pH change, leachate aging, and lower temperatures. Polymers like T & L may be adsorbed onto soil and clay by hydrophobic expulsion and electrostatic force (Dentel et al., 1995; Mohan and Karthikeyan, 1997; Stumm and Morgan, 1996). Adsorption capacity is influenced by size distribution of adsorbents and partition coefficient of contaminants. The partition coefficient for the sorption of a nonpolar organic solute on a solid phase is a function of the fraction of organic carbon of the solid material and the octanol-water partition coefficient (Stumm and Morgan, 1996). In the long term, the organic carbon content and texture of the solid material in CWs may change in the process of ecological maturation and seasonally i f the system is vegetated. Due to the phenolic and carboxyl functional groups on T & L , their adsorption capacity changes with pH (Das and . Patnaik, 2000; Mohan and Karthikeyan, 1997; Sakurai et al., 2001; Vinod and Anirudhan, 2002). The adsorption capacity of silt loam for T & L was also influenced by temperature. Temperature and pH may vary diurnally and seasonally in CWs. Therefore, the reduction efficiency through adsorption may fluctuate temporally in CWs.  3.4.3 Biological treatability A batch test for decomposition of wood leachate (Taylor et al., 1996) achieved 4.1% C O D reduction and 89% fatty acids reduction in a vat without aeration, and 66% C O D reduction 87  and 96% fatty acids reduction in a vat with aeration after 65 d. The present study achieved similar or better performance in a shorter retention time for woodwaste leachate treatment with pH adjustment and nutrient enrichment. Practically, combined treatment of woodwaste leachate with nutrient-rich wastewaters would enhance the potential of biological treatment. The treatability may also vary with temperature. Zenaitis et al. (2002) and Woodhouse and Duff (2004) achieved better performance for treatment of log yard runoff through batch aerobic reactors at 35°C and 34°C respectively than for the present study at 24 °C. V F A s are catabolic products of anaerobic fermentation. Field et al. (1988) reported that 4161% of bark water soluble COD could be acidified to V F A and methane, and 7-36% could be converted by methanogenesis to methane after 8-9 d of anaerobic digestion. For the woodwaste leachate in this study, much higher reduction efficiency for V F A s than for COD and T & L through the anaerobic reactor suggests that fermentation is the limiting stage for biological treatment in anaerobic CWs. In comparison to V F A s , T & L and the other organic compounds associated with COD of the woodwaste leachate exhibited a poor biodegradability, but could be removed to some extent. T & L removal may be attributed to mixed liquor solids adsorption, biochemical oxidation, and chemical oxidation (Helmreich et al., 2001; Larrea et al., 1989a and 1989b). Higher T & L reduction efficiency and reaction rate constant at higher M L V S S in the second run than those at much lower M L V S S in the first run was likely due to adsorption onto sludge. The lignin that was adsorbed onto an activated sludge consisted of up to 30% of the mixed liquor suspended solids in aerated sequencing batch reactors treating paper mill wastewater at a solids retention time of 10-40 d (Helmreich et al., 2001). Larrea et al. (1989a) attributed lignin reduction in activated sludge treatment of Kraft effluent to polycondensationadsorption. Much more C O D and T & L were removed through the aerated reactor than the anaerobic reactor. However, CWs cannot maintain aerobic conditions at the high oxygen demand that the woodwaste leachate poses. Oxygen transfer into wetlands may be accomplished through surface reaeration, plant transport, and photosynthesis. Photosynthesis should be very limited due to the dark color of woodwaste leachate. Microbial respiration and plant respiration when sunlight energy is unavailable consume oxygen. Rates of oxygen transport by macrophytes through the exposed leaves and stems to the rhizomes and roots range from 0 to 3 g O2 m" d" 2  88  1  (Bezbaruah and Zhang, 2005; USEPA, 2000); and any oxygen transported to the root zone will likely be consumed by the large benthic oxygen demand that normally exists in wetlands. One field study by Brix and Schierup (1990) found that the oxygen transported by Phragmites almost exactly balanced respiration by the plants. As a whole, the bulk system of CWs treating woodwaste leachate is anaerobic (Tao and Hall, 2004), though local aerobic conditions may exist in a thin water surface layer and around the rhizomes. Therefore, the actual performance in CWs treating the woodwaste leachate would probably fall in between the treatability of the aerated and anaerobic reactors.  3.5  Conclusions  Volatilization is insignificant for removal of V F A s and C O D from woodwaste leachate. In long-term operation, biological degradation would be the major mechanism for removing organic carbon from woodwaste leachate in CWs, which, however, may require a supplement of nutrients to improve performance. Increasing oxygen supply instead of microbial biomass would improve performance in CWs treating high oxygen-demanding woodwaste leachate, especially for removal of COD and T & L . Soil adsorption has an immediate effect on performance in T & L removal. Adsorption may offer a T & L removal potential that is higher than that of anaerobic biological degradability. However, the rapid initial adsorption implies that this mechanism affects the treatment performance mainly during initial and transitional operating periods of CWs. Adsorption also contributes to the variable performance due to temporal variations in temperature, pH, actual influent quality, water depth, and sediment organic content and texture.  3.6  References  Bailey, H.C., J.R. Elphick, A . Potter, E. Chao, D. Konasewich, and J.B. Zak. 1999. Causes of toxicity in stormwater runoff from sawmills. Environmental Toxicology and Chemistry, 18: 1485-1491. Bezbaruah, A . N . , and T. C. Zhang. 2005. Quantification of oxygen release by bulrush (Scirpus validus) roots in a constructed treatment wetland. Biotechnology and Bioengineering, 89: 308318.  89  Brix, H . , and H . Schierup. 1990. Soil oxygenation in constructed reed beds: the role of macrophyte and soil-atmosphere interface oxygen transport. In: P.F. Cooper and B . C . Findlater (eds.), Advances in Water Pollution Control: Constructed Wetlands in Water Pollution Control. Pergamon Press: New York, N Y . pp53-65. CCOHS. 2003. Search CHEMINFO  on CCINFOweb.,  Accessed on May 24, 2003 at  http://www.ccohs.ca/products/databases/cheminfo.htm. Das, C. P., and L . N . Patnaik. 2000. Removal of lignin by industrial solid wastes. Practice Periodical of Hazardous, Toxic, and Radioactive Waste Management, 4: 156-161. Dentel, S.K., J.Y. Bottero, K . Khatib, H . Demougeot, J.P. Duguet, and C. Anselme. 1995. Sorption of tannic acid, phenol and 2,4,5-trichlorophenol on organoclays. Water Research, 29: 1273-1280. Field, J.A., M.J.H. Leyendeckers, R.S. Alvarez, G . Lettinga, and L . H . A . Habets. 1988. The methanogenic toxicity of bark tannins and the anaerobic biodegradability of water soluble bark matter. Water Science and Technology, 20: 219-240. Forgie, D.J.L. 1988. Selection of the most appropriate leachate treatment methods Part 1: a review of potential biological leachate treatment methods. Water Pollution Research Journal of Canada, 23: 308-328. Frankowski, K . A . 2000. The Treatment of Wood Leachate Using Constructed Wetlands. M . A . Sc. thesis, The University of British Columbia, Vancouver, Canada. Helmreich, B . , C. Schiegl, and P.A. Wilderer. 2001. Fate of lignin in the process of aerobic biological treatment of paper mill wastewater. Acta Hydrochimica et Hydrobiologica, 29: 296300. Hunter, R., A . E . Birkbeck, and G. Coombs. 1993. Innovative marsh treatment systems for control of leachate and fish hatchery wastewaters. In: G . A . Moshiri (ed.), Constructed Wetlands for Water Quality Improvements. Lewis Publishers: Baca Raton, FL. pp477-484. Larrea, L., C.F. Forster, and D. Mele. 1989a. Changes in lignin during diffused air activated sludge treatment of Kraft effluents. Water Research, 23: 1073-1080. Larrea, L . , C.F. Forster, and D. Mele. 1989b. Kraft lignin behaviour in diffused aeration of Kraft effluents. Water Science and Technology, 21: 241-253. 90  Masbough, A . 2002. The Effectiveness of Constructed Wetlands for Treatment of Wood Leachate. M.A.Sc. thesis, The University of British Columbia, Vancouver, Canada. McNeeley, R. N . , V.P. Neimanis, and L . Dwyer. 1979. Water quality sourcebook: a guide to water quality parameters. Inland Waters Directorate, Water Quality Branch, Environment Canada, Ottawa, Ontario. Metcalf & Eddy, Inc. 2003. Wastewater Engineering - Treatment and Reuse, 4th edition. McGraw-Hill: New York, N Y . Mitsch, W.J. 1992. Landscape design and the role of created, restored and natural riparian wetlands in controlling nonpoint source pollution. Ecological Engineering, 1: 27-47. Mohan, S.V., and J. Karthikeyan. 1997. Removal of lignin and tannin colour from aqueous solution by adsorption onto activated charcoal. Environmental Pollution, 97: 183-187. Navia, R., L . Levet, M . L . Mora, G . Vidal, and M . C . Diez. 2003. Allophanic soil adsorption system as a bleached kraft mill aerobic effluent post-treatment. Water, Air, and Soil Pollution, 148:323-333. Peters, G.B., H.J. Dawson, B.F. Hrutfiord, and R.R. Whitney. 1976. Aqueous leachate from western red cedar: effects on some aquatic organisms. Journal of the Fisheries Research Board of Canada, 33: 2703-2709. Sakurai, A . , T. Yamamoto, A . Makabe, S. Kinoshita, and M . Sakakibara. 2001. Removal of lignin in a liquid system by an isolated fungus. Journal of Chemical Technology and Biotechnology, 77: 9-14. Seo, G.T., S. Ohgaki, and Y . Suzuki. 1997. Sorption characteristics of biological powdered activated carbon in B P A C - M F (biological powdered activated carbon-microfiltration) system for refractory organic removal. Water Science and Technology, 35: 163-170. Stumm, W., and J.J. Morgan. 1996. Aquatic Chemistry, 3rd edition. John Wiley & Sons, Inc.: New York, N Y . Tao, W., and K . J . Hall. 2004. Dynamics and influencing factors of heterotrophic bacterial utilization of acetate in constructed wetlands treating woodwaste leachate. Water Research, 38: 3442-3448.  91  Taylor, B.R., J.S. Goudey, and N . B . Carmichael. 1996. Toxicity of aspen wood leachate to aquatic life: Laboratory studies. Environmental Toxicology and Chemistry, 15: 150-159. Taylor, B.R., and N . B . Carmichael. 2003. Toxicity and chemistry of aspen wood leachate to aquatic life: field study. Environmental Toxicology and Chemistry, 22: 2048-2056. Temmink, J.H.M., J.A. Field, J.C. van Haastrecht, and R . C . M . Merkelbach. 1989. Acute and sub-acute toxicity of bark tannins in carp (Cyprinus carpio L.). Water Research, 23: 341-344. USEPA. 2000. Constructed Wetlands Treatment of Municipal Wastewaters. Manual of EPA/625/R-99/010, U.S. Environmental Protection Agency, Cincinnati, OH. 13, 36-37. Vinod, V.P., and T.S. Anirudhan. 2002. Sorption of tannic acid on zirconium pillared clay. Journal of Chemical Technology and Biotechnology, 11: 92-101. Werker, A . G . , J.M. Dougherty, J.L. McHenry, W.A. van Loon. 2002. Treatment variability for wetland wastewater treatment design in cold climates. Ecological Engineering, 19:1-11. Woodhouse, C , and S.J.B. Duff. 2004. Treatment of log yard runoff in an aerobic trickling filter. Water Quality Research Journal of Canada, 39: 232-238. Zenaitis, M . G . , and S.J.B. Duff. 2002. Ozone for removal of acute toxicity from logyard runoff. Ozone: Science and Engineering, 24: 83-90. Zenaitis, M . G . , H . Sandhu, and S.J.B. Duff. 2002. Combined biological and ozone treatment of log yard runoff. Water Research, 36: 2053-2061.  92  4  HETEROTROPHIC BACTERIAL ACTIVITIES AND TREATMENT PERFORMANCE OF PILOT-SCALE WETLANDS*  4.1  Introduction  Surface flow CWs are natural treatment systems. Like other natural aquatic systems, CWs depend on microorganisms to decompose organic matter (USEPA, 2000). To analyze the rates of reactions involved with in situ decomposition and mineralization of organic matter, both biochemical and geochemical approaches have been taken (Wetzel and Likens, 2000). Geochemical methods are the most commonly used, simple approach, usually taking a C W as a black box to assess the difference of inflow and outflow in chemical composition. Biochemical methods address the carbon flux through the microbial loop in a CW, providing insights to the biological treatment mechanism. Heterotrophic bacteria mineralize organic matter and convert organic compounds into biomass. Tao and Hall (2004) evaluated the dynamics of heterotrophic bacterial utilization of acetate in CWs treating woodwaste leachate. Flood et al. (1999) and Pollard et al. (1995) assessed bacterial growth rate of the biofilm communities near the inlets and outlets of surface flow wetlands receiving secondary sewage effluent. Toerien and Toerien (1985) investigated the relative contributions of water, epiphyton, settled sludge and sediment to heterotrophic activity of the macrophyte  beds receiving piggery effluent.  Bacterial responses to  environmental conditions may be different in terms of growth rate, respiration rate, biomass, and phylogenetic composition (Flood et al., 1999). Understanding the dynamics and factors influencing heterotrophic bacterial production and mineralization in CWs would undoubtedly elucidate performance variation and improve design and operation of treatment wetlands. Heterotrophic bacteria in surface flow CWs may either be suspended in water or attached to sediment and plants. Attached growth is thought to play a major role in improving the water  * A version of this chapter has been accepted for publication. Tao, W., K.J. Hall, and S.J.B. Duff. 2006. Heterotrophic bacterial activities and treatment performance of surface flow constructed wetlands receiving woodwaste leachate. Water Environment Research, 78: in press. 93  quality of CWs (Brix, 1994; USEPA, 2000). Masbough (2002), however, did not find large differences in treatment performance between the vegetated and open wetland cells receiving woodwaste leachate. Tao and Hall (2004) reported a negligible contribution of epiphyton to heterotrophic acetate uptake in the CWs treating woodwaste leachate. Establishment and maintenance of vegetation usually account for an important part of the total cost for CWs. More information on both treatment performance and microbial activities is needed to assess the role of vegetation in decomposition of organic matter in CWs. Bacterial activities and treatment performance in natural treatment systems are dependent on environmental factors. Kozub and Liehr (1999) found that denitrification was limited by the availability of easily degradable organic carbon in a C W receiving landfill leachate. The woodwaste leachate is acidic, nutrient-poor, and quite toxic (Frankowski, 2000), suggesting an unusual environment for microbial metabolism. Combined treatment provides a possibility to increase influent p H and nutrient content, and reduce toxicity to heterotrophic bacteria. In order to look for an appropriate source of wastewater for combined treatment with woodwaste leachate, it is necessary to examine the effects of nutrient amendment on heterotrophic activities and removal efficiency of organic matter in CWs. Bacterial incorporation of radio-labeled compounds has been adopted by many researchers as a standard technique to measure heterotrophic production. Kirchman et al. (1985) first introduced the  H-leucine incorporation method to measure growth of natural bacterial  assemblages. This technique has been applied to water, sediment and epiphyton of freshwater systems, including marshes (Fischer and Pusch, 1999; Moran and Hodson, 1992; Thomaz and Wetzel, 1995). C-glucose has been used as a sensitive tracer of bacterial substrate utilization 14  (Wetzel and Likens, 2000). Natural aquatic systems usually have a low glucose concentration, 1-11 p;g L" in freshwater (Hicks and Carey, 1968; Jones and Simon, 1975) and 17-272 u.g L" 1  1  in sediment pore water (King and Klug, 1982; Wood and Chua, 1973). Wood and Chua (1973) found that glucose turnover time was not related to the ambient glucose concentrations up to 61 p,g L" in water and 272 u-g L" in sediment pore water. Glucose turnover time reflects the 1  1  rate at which organic substrates are utilized by bacteria for synthesis and respiration. Glucose mineralization percentage expresses the respiratory activity of heterotrophic bacteria. In addition to monitoring the influent and effluent quality to determine treatment performance, this study employed bacterial incorporation of H-leucine and uptake of C-glucose to 3  94  14  examine the longitudinal and temporal variations of bacterioplanktonic activities, to assess the effects of ammonium nitrate amendment and vegetation on heterotrophic activities and treatment performance, and to evaluate the relative importance of water, epiphyton, and sediment in the surface flow CWs treating woodwaste leachate. Utilization of the three independent and complementary methods enabled assessment of overall treatment performance and shed light on the biological treatment mechanism in CWs. 4.2 4.2.1  Methodology Operation of pilot-scale constructed wetlands  This study was conducted on six pilot-scale surface flow CWs receiving woodwaste leachate. The pilot wetlands were located on the north bank of Fraser River near Mission, BC, Canada. Figure 4-1 illustrates the layout of the pilot-scale CW system. Since October 1999, these CWs have been used to treat the woodwaste leachate diluted with nearby slough water (2x) in 2001 and well water (9x) in 2002. A 9.3-m tank was used for dilution and mixing. The tank was 3  situated to feed the CW cells by gravity. The cells were operated in parallel. The cells had a trapezoidal cross section with a side slope of 35°. Each cell was 17 m long and 5 m wide at full berm. The bottom and side wall were covered with 0.5-mm PVC liner. Silt loam was laid evenly on the bottom in a 30-cm layer to support plants. The surface sediment (top 2 cm) had 5.1% organic content and 0.5 g dry weight per mL fresh sediment on average. The influent was evenly distributed through a lateral spreader. Each CW cell had a forebay (1.8 m long), serving as a settling basin. Four cells had broad-leaved cattails (Typha latifolid) in a density of 10-30 plants m" in years 2001 and 2002. Two open cells did not have plants. The effluent was 2  discharged through a lateral perforated PVC pipe buried underneath gravel, and pumped back to the woodwaste pile. This study was undertaken in two periods during 2001 and 2002. From August 30 to October 12 in 2001, it examined the longitudinal and temporal variations of heterotrophic activities of the bacterioplankton in one vegetated cell (#6). Nitrate is one of the strongest biologically useful oxidants in anaerobic environments. Due to the anaerobic condition of the wetland water and the very low concentration of nitrogen in the influent, ammonium nitrate  (NH4NO3)  fertilizer solution was pumped continuously from 22 August to 4 November 2002 to the influent spreader of one vegetated CW cell. Fertilization was intended to increase nitrate as the 95  electron acceptor for denitrification and inorganic nitrogen as a supplemental nutrient source. Heterotrophic activities were examined in one open C W cell (#5), one vegetated C W cell (#4), and one fertilized, vegetated C W cell (#6) to assess the effects of vegetation and fertilization.  Figure 4-1. Site plan of the pilot-scale constructed wetland treatment system (left) and the structure of a vegetated wetland cell (right). Site plan is not in exact dimensions.  4.2.2 Field measurement and laboratory analysis Field measurement was usually conducted weekly. Water temperature, D O and redox potential were measured at the depth of 10 cm in the C W cells using the same methods and instruments as described in Section 2.2.2 for woodwaste leachate. Measurements were usually made in the center of the C W cells, and were made at several points along the fertilized wetland cell on 29 August 2002. Water depth was recorded from a water level gauge in the center of each C W cell. The inflow to each cell was regulated by a valve. The inflow rate decreased gradually over a week due to fouling, although it was adjusted to a given rate on every field trip. A volumetric method was employed to measure flow rates of the influent and effluent. Water volume was a function of water depth, which was established based on the geometry and cross section of each C W cell. The nominal H R T was calculated from water volume divided by average flow rate. Influent and effluent samples were collected with 1-L high-density polyethylene bottles on the same day as field measurement, and stored at 4°C until analysis. A Beckman <J>44 p H meter  (Beckman Instruments, Irvine, C A ) was used to measure p H on stirred samples in the lab. COD, T & L , V F A s , ammonium and nitrate were analyzed with the same methods and instruments as described in Section 2.2.2 for woodwaste leachate. Chlorophyll-a was extracted with aqueous acetone and determined by the fluorometric method (APHA et al., 1999) with a 10-AU fluorometer (Turner Designs, Sunnyvale, CA) (Appendix 5). The numbers of plants and submerged leaves in a one-meter transverse strip were counted at the beginning and whenever a change in density or size was observed. The perimeter of submerged stems and width of submerged leaves were selectively measured and the averages were estimated. Submerged plant surface area was estimated with the measured water depth, average sizes of submerged plant stems and leaves, and plant number. Intensity of photosynthetically active radiation (400-700 nm) was measured with a Li-COR light meter (Li-COR Biosciences, Lincoln, NE) equipped with an underwater quantum sensor at noon on a sunny day in August. Light penetration was assessed by comparing the light intensity in the air just above water surface and at different depths of a C W receiving a mixture of 5 woodwaste leachate:95 tap water.  4.2.3  Sample preparation for radioisotope incubation  Water, epiphyton and sediment samples were collected from the center of the C W cells on 23 September 2002 for H-leucine incubation and 4 November 2002 for C-glucose incubation. 3  14  Water samples were collected near the inlet and outlet of the vegetated C W cell, every 1-2 weeks from August 30 to October 12, 2001. Water samples for radioisotope assays were carefully drawn into 20-mL disposable plastic syringes. The epiphyton was allowed to form on 600-grit silicon carbide sandpaper discs (0.95 cm each), which were adhered to duct tape, wrapped around Plexiglas plates (Figure 4-2), and hung in the wetlands at a water depth of 10-15 cm for 5 weeks. The sandpaper discs provide a substratum that is more reproducible than plant stems and leaves, while mimicking the surface roughness of cattail stems and leaves. Once the disc samples were retrieved, they were placed in 20-mL incubation syringes. Nine mL of wetland water was drawn into each syringe through a membrane filter (0.45 urn pore size). Sediment was sampled with Plexiglas corers (5.1 cm in internal diameter). About 8 cm of sediment was sealed with overlying water in each corer. The corers were stored in the dark at 4°C. Only the top 2-cm sediment where the bacteria are usually concentrated (Lovley and Klug, 1982; Marxsen, 1996; Moriarty et al., 1991) was 97  separated and mixed by a spatula just before radioisotope incubation. One mL of the sediment slurry was transferred in duplicate to measure oven-dry weight (>24 h at 105°C). Sediment suspension was made by diluting the mixed sediment slurry with 20 times (w/w) filtersterilized wetland water.  Figure 4-2. Epiphyton sampler (left), syringes for incubation of water and epiphyton (right), and a flask for incubation of sediment (right).  Syringes were chosen to eliminate headspace and to minimize the contact of water and epiphyton samples with molecular oxygen. Thomaz and Wetzel (1995) found that disruption of the epiphytic bacterial matrices and dispersion of the cells before incubation with radioisotopes might not mimic the field growth. The epiphytic samples were hence maintained as attached growth through sample preparation and radioisotope incubation. Although solute uptake rate is not greatly affected by the presence or absence of oxygen in anoxic sediments (Hall et al., 1972; Toerien and Cavari, 1982), sediment samples were quickly prepared to minimize exposure to the air.  4.2.4  C-glucose uptake assays  14  This study employed a modification of the procedures described by Wetzel and Likens (2000) and Cavari et al. (1978) to determine glucose turnover and mineralization by heterotrophic bacteria in water, sediment and epiphyton (Appendix 6). Water (9 mL) and epiphyton (1 sandpaper disc in 9 mL filter-sterilized wetland water) were prepared in 20-mL plastic syringes. Sediment (1 mL sediment suspension and 8 mL sterilized wetland water) was prepared in 25-mL flasks (Figure 4-2). Duplicate sediment suspension was transferred to measure oven-dry weight (24 h at 105°C). For each sample, one killed control and two live  98  subsamples were prepared. Formaldehyde (final concentration 3.7% w/w) was added to the controls 30 min before adding C-glucose solution. 14  One mL of D-[ C(U)]-glucose (DuPont Nen products, Boston, M A ; specific activity of 287 14  Ci mol" ) solution was added to each syringe or flask, giving a final concentration of 5-12 p.g 1  L" , or 0.1-0.2 uCi per syringe or flask. The actual radioactivity of each batch of C-glucose 1  I4  solution, disintegrations per minute (DPMs) of 0.1 mL solution, was counted with a Beckman LS6500 multi-purpose scintillation counter (Beckman Instruments, Inc., Fullerton, C A ) . Toerien and Cavari (1982) found that respiration rate was not influenced by addition of up to 0.1 mg L" glucose. The amount of labeled glucose added was selected to yield a significant 1  radioactivity for bacterial particles and respired CO2, and to prevent stimulation of bacterial activity. Incubation with C-glucose was conducted in the syringes and flasks in the dark at the in situ 14  water temperature (7-17°C) for 2 h. The syringes were inverted and flasks swirled several times every 30 min. Prior to the end of incubation, epiphyton samples were sonicated for 3 min in an ultrasonic cleaner (Aquasonic Model 50HT, 45 Watts; V W B Scientific, West Chester, PA) to detach bacteria. The sonication time was chosen to achieve effective detachment and minimize lysis (Schaule et al., 2000; Thomaz and Wetzel, 1995; Zips et al., 1990). At the end of incubation, formaldehyde (final concentration 3.7% w/w) was added to the live subsamples to terminate incubation. The incubation solution in the syringes was transferred to 25-mL flasks. Each flask was sealed with a serum cap, through which the rod of a small plastic cup was inserted before replacing the cap. The cup was used to hold a folded glass filter (24 mm in diameter) and was set in the flask's headspace. Through the serum cap, each flask was injected with 0.2 mL of 5N  H2SO4  to release respired  CO2. Phenethylamine (0.2 mL) was injected into the folded glass filter in the small cup to absorb CO2 for 45 min, during which the flasks were swirled several times. Bacterial particles were collected by filtration through a nitrocellulose membrane filter (0.45 urn pore size). Each particle-retaining or C02-trapping filter was placed in 5 mL Ecolite scintillation cocktail (ICN, Costa Mesa, C A ) . DPMs of the C-glucose assimilated (on particle-retaining filter) and 14  respired (in C02-trapping filter) by bacteria were counted with the scintillation counter. Glucose turnover times and mineralization percentage were calculated as follows:  99  Tw=Tc  (4-1)  0.009Z m Te = Tex , x-^—x 0.95cm 1000L r  2 X  n  _ °M l_  T  c  t o  =  x  wl p  7 b < p x  d  I0 cm , xa = 7fcxgx0.009x 10/0.95 m 4  w X  2  2  <»!i!  (4-2)  (4-3)  dxw  ( P P M added) (Total D P M of particle retaining + C 0 trapping filters) / AO  (  4  4  )  2  ,• o/ Mineralization % =  (DPM of C O 2 trapping filter) x l 00 (Total D P M of particle retaining filter + C O 2 trapping filter) b  (4-5)  where Tw = glucose turnover time by bacterioplankton (d); Te = glucose turnover time by epiphytic bacteria (d); Ts = glucose turnover time by sedimentary bacteria (d); 0.009 = the volume of wetland water incubated (L); 0.95 = the area of each epiphyton disc (cm ); 2  w = dry weight of the sediment applied to each flask (mg); a = specific plant surface area (m submerged plant surface area per m wetland water); 2 = the active sediment depth (assuming top 2-cm as above mentioned); d - water depth (cm); p = the dry weight (g) per m sediment slurry; and 3  At = incubation time (d). Tc denotes the time to overturn the glucose in 9 mL wetland water incubated in a syringe or flask by the heterotrophic bacteria in wetland water, on one disc of epiphyton or in one mL sediment suspension. Te and Ts convert Tc to the time required to overturn the glucose in the C W water column by the epiphyton submerged in the water column or sedimentary bacteria underlying the water column. The relative contributions of water, epiphyton and sediment to the overall glucose turnover by bacterioplankton, epiphyton, and sedimentary bacteria together in a given wetland are: 100x(l/7V)/(l/7ViH-l/7'e+l/7i) by water, 100x(l/re)/(l/rw+l/re+l/rs) by epiphyton, and  100  \OOx(l/Ts)/(\/Tw+\/Te+VTs)  4.2.5  by sediment.  H-Ieucine incorporation assays  3  Leucine incorporation has been proved applicable to anaerobic water (Bastviken and Tranvik, 2001; Cole and Pace, 1995; McDonough et al., 1986). The procedure for H-leucine assays 3  followed Jorgensen (1992) and Ward and Johnson (1996). Water, epiphyton, or sediment in the same amount as in glucose uptake assays was incubated with H-leucine in 20-mL plastic syringes or 25-mL flasks in the dark at the in situ water temperature for one hour. One killed control and two live subsamples were prepared for each sample. One mL 37% w/w formaldehyde, was used to fix controls and terminate incubation. One mL of dilute L-[4,53  H]leucine solution (Amersham Bioscience, Buckinghamshire, England; specific activity of  155 uCi nmoi" ) was added to each incubation syringe or flask. The final concentration of H 1  3  leucine was in the range of 0.47-0.92 p.g L" . As in C-glucose assays, sonication was 1  14  introduced to detach bacteria from sandpaper after incubation with leucine. At the end of incubation, bacterial protein was extracted for 30 min with hot T C A (final concentration of 5 %) and filtered through a polycarbonate filter (0.2 um pore size). Each filter with retained proteins was transferred to 5 mL Ecolite scintillation cocktail. DPMs of the H-leucine 3  incorporated into bacterial proteins were counted with the Beckman scintillation counter. Leucine incorporation rate was calculated as follows: JDPM -DPM  vl  sample  )xl42  control  2.2xl0  6  xRsxvxAt  where v = amount of water (L), epiphyton (m ), or sediment (g dry weight) used; VI = leucine incorporation rate of water (ng L" h" ), epiphyton (ng m" h" ), or sediment (ng 1  1  2  1  g" dry weight h" ); 1  1  DPM= radioactive disintegration per minute of live subsamples or controls; 2.2xl0 = DPMs in 1 uCi; 6  142 = molecular weight of H-leucine (g); 3  Rs = specific activity of H-leucine solution (uCi nmol" ); and 3  1  At = incubation time (h). 101  The measured H-leucine incorporation rate was corrected by an intracellular leucine dilution factor of 2 (Jorgensen, 1992; Simon and Azam, 1989). When the measured leucine incorporation rate for epiphyton (ng h" m" ) and the rate for sediment (ng h" g" dry weight), 1  2  1  1  were converted as follows to unified rate, U in ng h" m" (water equivalent), the relative 1  3  importance of water, epiphyton and sediment in a C W could be compared: U = 1 OOOx VI, for water  (4-7)  U = Vl*a, for epiphyton  (4-8)  U = VI* p x 2/d, for sediment  4.2.6  (4-9)  Statistical analysis  Statistical analysis followed Townend (2002) and was performed using Microsoft Excel 'data analysis' command and statistical functions. The value of n stands for the number of individual field samples. The difference between means was assessed by single-factor A N O V A . The least significant difference (LSD) was calculated to identify which two means were significantly different. The Bonferroni method was used to correct the significance level for multiple comparisons. Spearman's rank correlation analysis was used to test the significance of a monotonic relationship between two variables. Correlations and differences were considered significant at P <0.05.  4.3 Results 4.3.1  Operating conditions and treatment performance  Water depth varied between 13 and 26 cm in 2001 (Appendix 7), and was usually at 34-38 cm in 2002 (Appendix 8). The bulk water in all of the CWs was anaerobic (DO<0.3 mg L" ). Due 1  to the extremely dark color of the woodwaste leachate, light penetration was limited to a couple of centimeters. There was little algal development in the CWs. Chlorophyll-a was measured at 1.8-7.4 p.g L" in surface water and 2.0-3.6 mg m" on plant surfaces in 2001. This 1  2  created a very large difference between these CWs and a natural wetland where the relation between primary production and heterotrophic production is often significant. The influent had very low concentrations of ammonia-, nitrite- and nitrate-nitrogen, 0.07-1.19 mg L" in 2001 (Masbough, 2002) and 0.02-1.23 mg L" in 2002 (Appendix 8). The continuous 1  1  addition of 1.0 kg NH4NO3-N per week to one vegetated C W cell reduced the influent 102  C O D ^ N H ^ - N + N 0 -N) ratio from 3614 to 50 and increased the N 0 - N concentration from 3  3  0.04 mg L" to 47 mg L" on average. One week after starting NH4NO3 addition, depletion of 4 1  1  mg L" nitrate-N occurred within 5 meters in the fertilized, vegetated C W cell; and the 1  ammonium concentration decreased gradually along the cell (Table 4-1).  Table 4-1. Longitudinal variations of ammonia and nitrate in pilot-scale constructed wetlands Sampling point  Non-fertilized wetlands  Fertilized wetland  NH;  NO-  N H ;  NO"  0.00  0.03  36.7  32.5  Forebay  19.9  4.13  5 m from forebay  16.6  0.01  10 m from forebay  11.2  0.01  5.8  0.02  Influent  Effluent (14 m from forebay)  0.01-0.02  0.00  Concentrations (mg N L" ) of one grab sample from each point collected on the 8th day of 1  continuous NH4NO3 addition to the influent in year 2002.  Table 4-2 summarizes the operating conditions. and treatment performance (see Appendix 7 and Appendix 8 for original data). In 2002, the three C W cells had similar HRTs. There was no significant difference in COD removal (P = 0.66, LSD =17) and T & L removal (P = 0.33, LSD = 27) between any two of the three C W cells. The vegetated, fertilized wetland cell had a significantly higher V F A s reduction efficiency than the other two cells (P = 0.00, LSD = 20). The three C W cells had significantly different pH increases (P = 0.00, L S D = 0.3). In 2001, the actual inflow rate to the non-fertilized, vegetated C W cell was not known. There was no effluent discharge in most of the operating days due to evapotranspiration (Masbough, 2002). The HRT was likely longer than that in 2002 (Table 4-2). The influent applied to this C W cell in 2001 was stronger than in 2002 in terms of COD, T & L , and VFAs. It achieved much higher removal efficiencies for COD and T & L in 2001 than in 2002, probably due to the longer HRT. There was a lower V F A s reduction efficiency in 2001 than in 2002, which was expected since more V F A s would be generated through fermentation at higher COD concentrations. 103  Table 4-2. Operating conditions and treatment performance of pilot constructed wetlands Sampling period  30 Aug. - 12 Oct.  29 Aug. - 4 Nov. 2002 (n = 7)  2001 (n = 6) b  Type of wetland  Vegetated,  Vegetated,  Open,  Vegetated,  non-fertilized  non-fertilized  non-fertilized  fertilized  Not available  12±2  13±2  14±2  13±4  12±5  12±4  11±4  Not available  213±86  119±92  142±110  Influent pH  4.2±0.2  4.5±0.3  4.5±0.3  4.5±0.3  Effluent pH  5.0±0.4  5.4±0.7  5.9±0.3  6.8±0.2  pH increase  0.8±0.4  1.0±0.4  1.4±0.2  2.3±0.2  4892±1681  1208±202  1208±202  1208±202  2667±1339  921±169  853±194  847±156  45±18  23±13  29±18  29±19  1432±695  331±42  331±42  331±42  691±258  293±48  270±32  276±42  49±12  12±8  18±5  16±6  842±440  216±50  216±50  216±50  419±240  106±71  87±25  20±23  49±21  54±28  59±11  92±9  Hydraulic retention time, d Water temperature, °C Redox potential, mV  Influent COD, mg L"  1  Effluent COD, mg L"  1  COD reduction efficiency, % Influent T & L , mg L"  1  Effluent T & L , mg L"  1  T & L reduction efficiency, % Influent V F A s , mg L"  1  Effluent V F A s , mg L"  1  V F A reduction efficiency, % a  Mean ± SD.  b  Derived from Masbough (2002).  4.3.2  Dynamics of bacterioplanktonic activities  Water samples were collected simultaneously near the inlet and outlet of a C W to examine the longitudinal variation of heterotrophic activities of planktonic bacteria in 2001. The ratios of 104  heterotrophic activity between the inlet and outlet varied over time, with a mean±SD of 0.9±0.4 for the leucine incorporation rate, 1.7+1.7 for the glucose turnover time, and 1.9±1.6 for the glucose  mineralization percentage. However, the  longitudinal variation was  insignificant in leucine incorporation rate (P = 0.84), glucose turnover time (P = 0.93) and glucose mineralization percentage (P =0.28). The heterotrophic activities of planktonic bacteria varied widely and irregularly over seven weeks in 2001 (Figure 4-3). The mean leucine incorporation rate was 58.8 ng L" h" , with a 1  1  coefficient of variation of 0.89. The mean glucose turnover time was 7.9 d, with a coefficient of variation of 1.16. Glucose mineralization percentage was 51% on average, with a coefficient of variation of 0.37. Leucine incorporation rate, glucose turnover time, and glucose mineralization percentage were not correlated to water temperature or the concentrations of COD, T & L , and V F A s in C W water (P>0.05, -0.68<r <0.47). 5  Glucose turnover time was positively correlated with leucine incorporation rate and negatively correlated with glucose mineralization percentage (Figure 4-4). However, there was not a significant relationship between glucose mineralization percentage and leucine incorporation rate (r, = 0.60, P>0.05).  I Near inlet  • Outlet  I Near inlet  • Outlet  Figure 4-3. Leucine incorporation and glucose turnover by planktonic bacteria in a vegetated constructed wetland in 2001. No glucose data on September 7. Each bar represents the result of radioisotope bioassays with one control and two live subsamples.  105  6  6  > O 3  > o B  O o 3  O o  5  5  E  3  3  o  40  80  120  20  160 l , -U  Leucine incorporation rate (ng L h )  40  60  80  Glucose mineralization %  Figure 4-4. Correlations of glucose turnover by planktonic bacteria to leucine incorporation and glucose mineralization percentage in a vegetated constructed wetland in 2001.  4.3.3 Effects of nutrient amendment and vegetation on heterotrophic activities Water, epiphyton, and sediment responded similarly in heterotrophic activities to ammonium nitrate addition and vegetation. The vegetated, fertilized C W cell had the highest leucine incorporation rate and shortest glucose turnover time, followed by the open cell, then the nonfertilized vegetated cell (Table 4-3). Glucose mineralization percentage was highest in the non-fertilized vegetated cell, followed by the open cell, and lowest in the fertilized, vegetated cell (Table 4-3).  4.4  Discussion  4.4.1 Implications of heterotrophic dynamics Insignificant longitudinal changes in leucine incorporation rate and glucose turnover time by bacterioplankton suggested no inhibition or sufficient mixing that brought about a relatively even distribution of bacterioplankton. The insignificant variation in heterotrophic production of planktonic bacteria is consistent with the subtle longitudinal variation of biofilm community growth rate within two surface flow wetlands receiving secondary sewage effluent (Flood et al., 1999). The insignificant longitudinal variation of glucose turnover by bacterioplankton is consistent with the insignificant longitudinal variations of acetate uptake by planktonic and sedimentary bacteria in the same C W (Tao and Hall, 2004).  106  Table 4-3. Effects of ammonium nitrate addition and vegetation on heterotrophic bacterial activities in constructed wetlands treating woodwaste leachate  a  Wetland cell  Vegetated,  Open,  Vegetated,  non-fertilized  non-fertilized  fertilized  68.1 ngm" h"'  120.2 ugm- h"'  N/A  8.2 ug m ' h"  Leucine Incorporation  Water  37.6 ug m" h" 3  Epiphyton  rate  b  2.1 L i g m ^ h 3  b  l.lngg'h-  time  Glucose Mineralization  1  3  2  1  (32.7 ug i n h' )  1  3  12.4 n g g V  3.3 n g g h" 1  1  1  (31.3 ugm- h-')  (86.5 ug m" h" )  (365 ug i n h' )  Water  7.6 d  1.1 d  0.2 d  Epiphyton  284 d  N/A  2.1 d  Sediment  2.4 d  0.1 d  0.05 d  Water  49%  31%  10%  Epiphyton  64%  N/A  10%  Sediment  58%  20%  20%  3  Turnover  - 1  3  (2.4 ugm" !!- ) Sediment  Glucose  1  3  1  c  3  1  percentage  a  Results of radioisotope bioassays with one control and two live subsamples.  b  Brackets give the rate of epiphyton and sediment as water equivalent.  0  Not applicable.  Heterotrophic activity is related to the trophic level of habitats (Stanley and Staley, 1977) and organic substrates (Wetzel, 1993). Toerien and Toerien (1985) reported a glucose turnover time of 0.43-0.46 h by water in macrophyte beds receiving diluted piggery effluent. This effluent had a COD concentration similar to the woodwaste leachate influent and a very high concentration of total nitrogen (200 mg L" ). Like the macrophyte beds, the CWs treating 1  woodwaste leachate were anaerobic. Glucose turnover time on average in the C W in 2001 was 426-fold longer than that in the macrophyte beds. Glucose turnover time in the C W was likely  107  retarded by the extremely low nutrients as revealed by the reduced turnover time in the C W amended with ammonium nitrate (Table 4-3). Moreover, the more recalcitrant character of the woodwaste leachate (i.e., the high concentration of T & L relative to COD) would also affect the structure of the microbial community and its heterotrophic potential. Significant variations in bacterial production rate have been noted from one system to another (Cole et al., 1988). Even under stable conditions, the characteristics of the microbial community can vary (Stanley and Staley, 1977). Leucine incorporation rate, glucose mineralization percentage and glucose turnover time were not significantly correlated to water temperature or the concentration of COD, T & L , or V F A s in the C W treating woodwaste leachate. Leucine incorporation rate, glucose mineralization percentage and glucose turnover time correlated differently from each other in the vegetated C W treating woodwaste leachate. When glucose was turned over slowly in the C W cell in 2001, less glucose taken up by bacterioplankton was respired and there was a greater leucine incorporation rate (Figure 4-4). However, a higher leucine incorporation rate was accompanied by a shorter glucose turnover time when comparing among the open, vegetated, and fertilized C W cells in 2002 (Table 4-3). This suggests a possibility of different structures of the microbial community in the three C W cells. Similarly, biomass, growth rate, and phylogenetic composition of epiphytic biofilm differed independently of each other between four sites in a wetland system treating secondary sewage effluent (Flood et al., 1999). Long-term monitoring of operating conditions, treatment performance, and heterotrophic activities of water, epiphyton and sediment is needed to provide complementary information for optimization of CWs.  4.4.2  Effects of nutrient and electron acceptor amendment  Studies (Kuparinen and Heinanen, 1993; Mueller, 1996) have emphasized that not only the amount of available soluble organic compounds, but also the proper ratio of inorganic nutrients to organic carbon in the water might be key elements in regulating bacterial production. Biofilm growth rates were found to be associated with a nutrient gradient along a surface flow C W (Pollard et al., 1995). Uptake rates of simple carbohydrates generally increase with increasing concentrations of total inorganic nitrogen and TP (Wetzel, 1993). The addition of ammonium nitrate to the C W treating woodwaste leachate demonstrated a significant increase of leucine incorporation rate and more rapid glucose turnover (Table 4-3). 108  Nitrogen and nitrous oxide are the major products of denitrification at positive redox potentials, while the relative significance of nitrate reduction to ammonia increases at negative redox potentials (Jones and Simon, 1981; Sorensen, 1978). The positive redox potential in the water column (161-242 mV) of the fertilized C W cell indicated conversion of nitrate to nitrogen and nitrous oxide rather than ammonia. Therefore, a gradual decrease of ammonium was observed because of slow utilization of ammonium nitrogen as a nutrient. The different longitudinal gradients of ammonium and nitrate implied that electron acceptors like nitrate might be more limiting than nutrients for heterotrophic production and mineralization in the CWs treating woodwaste leachate. Similarly, Flood et al. (1999) did not find a significant difference in bacterial growth rate by epiphyton among four sites with different levels of ammonium and total phosphate in two wetlands receiving secondary sewage effluent. The addition of ammonium nitrate yielded significant improvement in V F A s removal and only a slight increase in removal efficiency for COD and T & L . This implies that V F A s were the priority substrates for denitrifying bacteria. pH increased significantly due to improved V F A s removal and denitrification. A significant increase in acetate uptake rate in the same C W cell has been reported by Tao and Hall (2004). In a C W receiving landfill leachate, Kozub and Liehr (1999) found that denitrification rates were limited by availability of easily degradable organic carbon, such as acetate. 4.4.3  Effect of vegetation  Removal of suspended solids and biochemical oxygen demand in surface flow CWs was reported to be better in wetland cells with plants than in adjoining cells without plants (Gearheart et al., 1989; Thut, 1989) because of sedimentation, interception and adhesion enhanced by the standing plants. In the present study, there was not a significant effect of vegetation on woodwaste leachate treatment (Table 4-2). The contributions of epiphyton to total heterotrophic activities (Table 4-4) were so low that vegetation played a negligible role in woodwaste leachate treatment. The higher leucine incorporation rate and shorter glucose turnover time of water in the open C W cell than in the vegetated C W cell might result from more DO supplied by surface aeration due to wind-driven turbulence. The lower leucine incorporation rate and longer glucose turnover time of sediment in the vegetated C W cell may result from competition for nutrients between plants and sedimentary bacteria. However, glucose mineralization 109  percentage was apparently higher in the vegetated C W cell. The different responses of leucine incorporation rate and glucose turnover time from glucose mineralization percentage suggest a difference in species composition of the heterotrophic community in CWs due to vegetation.  Table 4-4. Comparison of relative contributions of water, epiphyton, and sediment to the total heterotrophic activities by water and water equivalents in three types of constructed wetlands treating woodwaste leachate Contribution to  Contribution to  overall glucose turnover time  total leucine incorporation rate  Vegetated,  Open,  Vegetated,  non-fertilized non-fertilized  fertilized  Vegetated,  Open,  Vegetated,  non-fertilized non-fertilized fertilized  39%  8%  21%  53%  44%  23%  Epiphyton  1%  N/A  2%  3%  N/A  7%  Sediment  60%  92%  77%  44%  56%  70%  Water  4.4.4  Relative importance of water, sediment and epiphyton  By applying H-leucine incorporation technique to a freshwater marsh with average soluble 3  organic carbon concentration of 42 mg L" , Moran and Hodson (1992) found that 46-88% total 1  bacterial productivity on an areal basis was attributed to sediments, and that the remainder was contributed approximately equally by bacteria in water column and on plant detritus. The present study found a similar contribution of sediment and much less contribution of epiphyton to the total production rate in the CWs treating woodwaste leachate (Table 4-4). With regard to glucose uptake rate, most bacterial activity in macrophyte beds receiving piggery effluents was associated with the bacterioplankton and much less with epiphyton (Toerien and Toerien, 1985). The present study found a negligible contribution of epiphyton to glucose turnover, and 8-39% contribution by bacterioplankton (Table 4-4). The two kinds of treatment systems had similar water depths. The difference in the relative importance of water and sediment may be caused by different soil textures that influence the surface area available  110  for bacterial attachment. The importance of planktonic and sedimentary bacteria in glucose turnover (Table 4-4) is similar to that in acetate uptake (Tao and Hall, 2004). The addition of ammonium nitrate stimulated denitrification and depleted VFAs. Starvation of bacteria from such good substrates as V F A s increased the relative importance of epiphyton and sediment, and decreased the relative importance of planktonic bacteria (Table 4-4). In low nutrient environments, attachment might be used as a strategic survival mechanism of organisms (Mueller, 1996). The relative importance of sedimentary bacteria to the total heterotrophic activities in the CWs receiving woodwaste leachate implies that higher treatment performance may be achieved at shallow water depths. Nevertheless, the epiphytic community in addition to bacterioplankton should be enhanced when the land requirement is minimized by increasing water depth for a given treatment capacity.  4.5 Conclusions Bacterial production, mineralization, and their distribution among water, epiphyton and sediment in surface flow CWs were influenced by the supply of inorganic nutrients and electron acceptors, and the availability of organic substrates. Ammonium nitrate addition had an insignificant effect on the removal of COD and T & L , and a significant effect on p H and V F A s removal. Vegetation only increased pH. However, both ammonium nitrate  addition and vegetation resulted in apparently different leucine  incorporation rates, glucose turnover times, and glucose mineralization percentages by bacteria suspended in water and attached on plant surfaces or sediment. To better interpret the effects of such factors as vegetation and nutrient amendment on treatment performance, it is suggested to conduct simultaneous long-term monitoring of treatment performance and examination of biomass and phylogenetic composition in addition to determination of bacterial production rate and substrate uptake rate. Improved performance in woodwaste leachate treatment may be achieved with a shallow water depth. Nevertheless, bacterioplankton would be preferred when deeper water is designed to minimize land requirement and maximize treatment capacity. Utilization of vegetation for bacterial attachment is not important in woodwaste leachate treatment.  Ill  4.6  References  A P H A et al. 1999. Standard Methods for the Examination of Water and Wastewater, 20th ed. American Public Health Association, American Water Works Association, and Water Environment Federation. Bastviken, D., and L . Tranvik. 2001. The leucine incorporation method estimates bacterial growth equally well in both oxic and anoxic lake waters. Applied and Environmental Microbiology, 67: 2916-2921. Brix, H . 1994. Use of constructed wetlands in water pollution control - historical development, present status, and future perspectives. Water Science & Technology, 30: 209223. Cavari, B.Z., G. Phelps, and O. Hadas. 1978. Glucose concentrations and heterotrophic activity in Lake Kinneret. Verhandlungen der Internationalen Vereinigungfur Limnologie, 20: 2249-2254. Cole, J.J., S. Findlay, and M . L . Pace. 1988. Bacterial production in fresh and salt-water ecosystems: a cross-system overview. Marine Ecology Progress Series, 43: 1-10. Fischer, H., and M . Pusch. 1999. Use of the [ C]leucine incorporation technique to measure 14  bacterial production in river sediments and the epiphyton. Applied and Environmental Microbiology, 65: 4411-4418. Flood, J.A., N.J. Ashbolt, and P.C. Pollard. 1999. Complementary independent molecular, radioisotopic and fluorogenic techniques to assess biofilm communities in two wastewater wetlands. Water Science and Technology, 39: 65-70. Frankowski, K . A . 2000. The Treatment of Wood Leachate Using Constructed Wetlands. M.A.Sc. thesis, the University of British Columbia, Vancouver, Canada. Gearheart, R.A., F. Kloop, G. Allen. 1989. Constructed free surface wetlands to treat and receive wastewater: pilot project to full scale. In: D.A. Hammer (ed.) Constructed Wetlands for  Wastewater Treatment, Municipal, Industrial, and Agricultural.  Chelsea, M L ppl21-137.  112  Lewis Publishers:  Hall, K.J., P . M . Kleiber, and I. Yesaki. 1972. Heterotrophic uptake of organic solutes by microorganisms in the sediments. Proceedings of the IBP-UNESCO Symposium on Detritus  and Its Role in Aquatic Ecosystems, Pallanza, Italy, May 23-27, 1972. pp441-471. Hicks, S.E., and F.G. Carey. 1968. Glucose determination in natural waters. Limnology and Oceanography, 13: 361-363. Jones, J.G., and B . M . Simon. 1975. Some observations on the fluorometric determination of glucose in freshwater. Limnology and Oceanography, 20: 882-887. Jones, J.G., and B . M . Simon. 1981. Differences in microbial decomposition processes in profundal and littoral lake sediments with particular reference to the nitrogen cycle. Journal of General Microbiology, 123: 297-312.  Jorgensen, N.O.G. 1992. Incorporation of H-leucine and H-valine into protein of freshwater 3  3  bacteria: uptake kinetics and intracellular isotope dilution. Applied and Environmental Microbiology, 58: 3638-3646. King, G . M . , and M.J. Klug. 1982. Glucose metabolism in sediments of a eutrophic lake: tracer analysis of uptake and product formation. Applied and Environmental Microbiology, AA: 1308-1317. Kirchman, D.L., E. K'Nees, and R. Hudson. 1985. Leucine incorporation and its potential as a measure of protein synthesis by bacteria in natural aquatic systems. Applied and Environmental Microbiology, 49: 599-607.  Kozub, D.D., and S.K. Liehr. 1999. Assessing denitrification rate limiting factors in a constructed wetland receiving landfill leachate. Water Science and Technology, 40: 75-82. Kuparinen,  J, and A . Heinanen.  1993. Inorganic  nutrient  and carbon controlled  bacterioplankton growth in the Baltic Sea. Estuary and Coastal Shelf Science, 37: 271-286. Lovley, D.R., and M . J . Klug. 1982. Intermediary metabolism of organic matter in the sediments of a eutrophic lake. Applied and Environmental Microbiology, 43: 552-560. Marxsen, J. 1996. Measurement of bacterial production in stream-bed sediments via leucine incorporation. FEMS Microbiology Ecology, 21:313-325. Masbough, A . 2002. Effectiveness of Constructed Wetlands in Wood Leachate Treatment.  M.A.Sc. thesis, the University of British Columbia, Vancouver, Canada. 113  Moran, M . A . , and R.E. Hodson. 1992. Contributions of three subsystems of a freshwater marsh to total bacterial secondary productivity. Microbial Ecology, 24: 161-170. Moriarty, D.J.W., G.W. Skyring, G.W. O'Brien, and D.T. Heggie. 1991. Heterotrophic bacterial activity and growth rates in sediments of the continental margin of eastern Australia. Deep Sea Research, 38: 693-712. Mueller, R.F. 1996. Bacterial transport and colonization in low nutrient environments. Water Research, 30: 2681-2690. Pollard, P.C., J.A. Flood, and N.J. Ashbolt. 1995. The direct measurement of bacterial growth in biofilms of emergent plants (Schoenoplectus) of an artificial wetland. Water Science and Technology, 32: 251-256. Schaule, G., T. Griebe, and H.-C. Flemming. 2000. Steps in biofilm sampling and characterization in biofouling cases. In: H.-C. Flemming, U . Szewzyk, and T. Griebe (eds.), Biofilms: Investigative Methods & Applications. Technomic Publishing Company, Inc.: Lancaster, PA. pp 1-21. Simon, M . , and F. Azam. 1989. Protein content and protein synthesis rates of planktonic marine bacteria. Marine Ecology Progress Series, 51: 201-213. Sorensen, J. 1978. Capacity for denitrification and reduction of nitrate to ammonia in a coastal marine sediment. Applied and Environmental Microbiology, 35: 301-305. Stanley, P . M . , and J.T. Staley. 1977. Acetate uptake by aquatic bacterial communities measured by autoradiography and filterable radioactivity. Limnology and Oceanography, 22: 26-37. Thomaz, S.M., and R.G. Wetzel. 1995. [ H]leucine incorporation methodology to estimate 3  epiphytic bacterial biomass production. Microbial Ecology, 29: 63-70. Thut, R.N. 1989. Utilization of artificial marshes for treatment of pulp mill effluents. In: D.A. Hammer (ed.) Constructed Wetlands for Wastewater Treatment: Municipal, Industrial, and Agricultural. Lewis Publishers: Chelsea, MI, pp239-244. Tao, W., and K.J. Hall. 2004. Dynamics and influencing factors of heterotrophic bacterial utilization of acetate in constructed wetlands treating woodwaste leachate. Water Research, 38:3442-3448. 114  Toerien, D.F., and B . Cavari. 1982. The effect of temperature on heterotrophic glucose uptake, mineralization,  and  turnover  rates in lake sediments.  Applied  and  Environmental  Microbiology, 43: 1-5. Toerien, D.F., and M . C . Toerien. 1985. Microbial heterotrophy in an effluent treatment system using macrophytes. Agricultural Wastes, 12: 287-312. Townend, J. 2002. Practical Statistics for Environmental and Biological Scientists. John Wiley & Sons: West Sussex, England. USEPA. 2000. Manual:  Constructed Wetlands Treatment of Municipal Wastewaters;  EPA/625/R-99/010, U.S. Environmental Protection Agency, Cincinnati, OH. Ward, A . K . , and M . D . Johnson. 1996. Heterotrophic microorganisms. In: F.R. Hauer, and G.A. Lamberti (eds.), Methods in Stream Ecology. Academic Press: San Diego, C A . pp 254260. Wetzel, R.G. 1993. Limnology, 2nd ed. CBS College Publishing: New York, N Y . Wetzel, R.G., and G.E. Likens. 2000. Limnological Analyses, 3rd ed. Springer-Verlag: New York, N Y . Wood, L.W., and K . E . Chua. 1973. Glucose flux at the sediment-water interface of Toronto Harbour, Lake Ontario with reference to pollution stress. Canadian Journal of Microbiology, 19:413-420. Zips, A., G. Schaule, and H.-C. Flemming. 1990. Ultrasound as a means of detaching biofilms. Biofouling, 2: 323-333.  115  5  HETEROTROPHIC BACTERIAL UTILIZATION OF ACETATE IN PILOT-SCALE CONSTRUCTED WETLANDS*  5.1  Introduction  Surface flow CWs are natural treatment systems. Organic carbon removal is mainly achieved by heterotrophic microorganisms (USEPA, 2000). The  organic  carbon  flux  through  microorganisms in aquatic systems is usually so low that only radioactive tracers are sensitive enough to follow it (Hall et al., 1972). Parsons and Strickland (1962) first discovered that assimilation of C-labeled soluble organic compounds by natural microorganisms follows I4  Michaelis-Menten kinetics. Hobbie and Crawford (1969) proposed a respiration correction for bacterial uptake of soluble organic compounds. Since then, this technique has been extensively used to measure the heterotrophic activity of microorganisms in water, epiphyton, and sediment of natural aquatic systems. Substrate uptake rate and mineralization percentage give information about the size and function of a bacterial community (Wright and Hobbie, 1966). Acetate is a principal intermediate in a number of intracellular reactions of biosynthesis and energy metabolism. Along with C-glucose, C-acetate is one of the frequently used 14  14  radiolabels to model soluble organic compounds. Probably all bacterial respiration processes are capable of consuming acetate in wetlands (Hines et al., 2001). Occurrence of either aerobic or anaerobic reactions and their rates in treatment systems are dependent on wastewater characteristics and environmental variables such as temperature, DO, pH, and concentrations of inorganic nutrients and organic substrates. Wood leachate is acidic, nutrient-poor, and toxic to microorganisms, but has a very high organic strength. These features suggested a possibly adverse environment for microorganisms in CWs treating woodwaste leachate. Understanding the dynamics and influencing factors of heterotrophic bacterial activity would undoubtedly facilitate design and operation of treatment wetlands  * A version of this chapter has been published. Tao, W., and K . J. Hall. 2004. Dynamics and influencing factors of heterotrophic bacterial utilization of acetate in constructed wetlands treating woodwaste leachate. Water Research, 38: 3442-3448. 116  receiving woodwaste leachate. However, bacterial activity has rarely been investigated quantitatively in CWs. CWs usually have three components, i.e. water column, sediment and emergent plants. Heterotrophic bacteria may either be suspended in water or attached on sediment and plants. Attached growth is thought to play a major role in improving the water quality of CWs (Brix, 1994; USEPA, 2000). Nevertheless, no quantitative assessments have been reported to support the importance of attached bacteria within a particular C W . The relative contributions of water, sediment and epiphyton would provide a critical basis for optimal design of treatment wetlands. This study used heterotrophic bacterial uptake of C-acetate to follow the development of 14  microbial colonization, assess the effects of fertilization and vegetation, and evaluate the temporal and longitudinal variations as well as the relative importance of water, sediment and epiphyton in the CWs treating woodwaste leachate.  5.2  Methods and Materials  5.2.1  Constructed wetlands  This study investigated bacterial activity in the pilot-scale surface flow CWs treating woodwaste leachate in Mission, British Columbia, Canada. This study involved one wetland cell (#6) during August 1-October 12, 2001, and compared a fertilized cell (#6), a vegetated cell (#4), and an open cell (#5) on November 4, 2002 (Figure 4-1). The influent in 2001 had pH 4.0-4.6, C O D 2480-6560 mg U , V F A s 109-1616 mg L " , T & L 489-2510 mg L " , l  1  1  ammonia and nitrate 0.07-1.19 mg N L" , and orthophosphate 0.23-3.15 mg P L" (Masbough, 1  1  2002). The influent in 2002 (Table 5-2) was more dilute than in 2001. The loading rate was 80-214 g C O D m" d" . The operating conditions and treatment performance have been presented in Section 4.3.1 (details in Appendix 7 and Appendix 8).  5.2.2  Radioisotope incubation and assays  Epiphyton was collected on sandpaper discs at a depth of 10-15 cm for 3 weeks in 2001 and 5 weeks in 2002. Sandpaper discs were also retrieved after one and two weeks in 2001 to follow the progress of bacterial colonization. Water, epiphyton, and sediment samples were prepared for C-acetate incubation in the same way as for C-glucose incubation (Section 4.2.3). One 14  14  117  rnL of dilute [l- C]acetate solution (specific activity of 56 C i mol" ), or 0.1-0.3 uCi, was 14  1  added to each syringe or flask. Samples were incubated and bioassayed with C-acetate in the 14  same procedure as with C-glucose in Section 4.2.4. 14  The natural concentration of acetic acid in water samples was determined as a component of V F A s (Section 2.2.2). Michaelis-Menten kinetics holds for acetate concentrations from below 0.5 mg L" (Wright and Hobbie, 1966) for natural systems and up to 30 mg L" in waste 1  1  treatment systems (Wetzel, 1993). Because of the high acetic acid concentration of wetland water (6-235 mg L" ), the addition of 12.4-29.8 ug L" C-acetate would not change the 1  1 14  natural pool size and subsequently provided a good method for measuring maximum bacterial uptake rate. Acetate mineralization percentage was calculated in equation 4-5 as for glucose mineralization. The (gross) uptake rate of acetate was calculated as follows: (DPM  v  -DPM )xMW — 2.2x\0 xRsxvx At sample  Va =  control  6  9xSn + Ra— Ra  (5-1)  where S„ = concentration of acetate in mesocosm water, iig mL" ; 1  Ra = C-acetate added, ug; 14  DPM= radioactivity assimilated in bacterial particulates plus radioactivity respired to C02 14  of fixed controls or live subsamples, disintegration per minute; MW= molecular weight of acetate, ug mol" ; 1  2.2x 10 = conversion factor of uCi to disintegration per minute; 6  Rs = specific radioactivity of C-acetate, uCi mol" ; 14  1  v = amount of sample used, L water, m epiphyton, or g (d.w.) sediment slurry; 2  At = incubation time, h; and Va - acetate (gross) uptake rate, ug L" h" water, ug m" h" epiphyton, or ug g" h" 1  1  2  1  1  1  sediment. Similar to leucine incorporation rates (equations 4-8 and 4-9), the areal acetate uptake rate of epiphyton and the gravimetric rate of sediment were converted to the rate of water equivalent. Based on the same unit, the relative importance of water, epiphyton and sediment could be compared in a CW. 118  5.2.3  Statistical analysis  Statistical analysis was performed using Microsoft Excel data analysis. The value of n stands for the number of field samples. Correlation analysis generated the Pearson's product moment correlation coefficient, r. The differences between means were assessed by paired /-tests. When normal distribution and equal variance assumptions were not met, log or arcsine transformation was applied to normalize the data. Correlations and differences were considered significant at P<0.05.  5.3 Results 5.3.1  Microbial colonization  This study followed development of the epiphytic bacterial community using C-acetate as a 14  substrate tracer. Figure 5-1 shows a significant increase in acetate uptake rate over three weeks of microbial colonization. A considerable increase in mineralization percentage was also observed, from 16-20% on average in the first two weeks to 64% in the third week.  co >>  .C  B<  'S. u  o  '6  60  E  M  o. 3  .  a] -*-»  3 week 1  week 2  week 3  Colonization time  Figure 5-1. Development of epiphyton in a surface flow constructed wetland treating woodwaste leachate from August 1 to September 25, 2001. n = 4. Error bar = SD.  5.3.2  Evaluation of heterotrophic bacterial activity  The long-term heterotrophic activities of bacterioplankton, epiphyton after three weeks of colonization, and sedimentary bacteria in the vegetated C W treating woodwaste leachate are 119  presented in Table 5-1. The large coefficients of variation of bacterial activity for water (0.510.63), epiphyton (0.42-0.62) and sediment (0.96-1.42) samples may result from heterogeneity of sediment and oscillation of microbial community in nature.  Table 5-1. Bacterial activity in a pilot-scale surface flow constructed wetland receiving woodwaste leachate Wetland component Water  Measured acetate  Uptake rate as of  Contribution to  Mineralization  uptake rate  water equivalent  total uptake rate  percentage  20%  55.3+28.3%  8%  63.8±26.8%  72%  16.1+22.8%  408±258 ug L" If 1  1  408±258 mg m" h'  1  1  3  Epiphyton  67.7±41.7mgm" h"'  167±115mgm" h-  Sediment  26.7±25.6 ug g" h'  1471±1640 mg m ' h'  2  1  1  3  3  1  Mean or mean ± SD of water (n = 8), epiphyton (n = 4), and sediment (n = 14) samples collected between August 1 and October 12, 2001.  The temporal variations of acetate uptake rate and mineralization percentage in the C W in 2001 (Appendix 9) did not show a significant, positive correlation with water temperature or acetic acid concentration. However, there was a significant, positive correlation between acetate uptake rate and mineralization percentage (Figure 5-2), indicating that the more acetate the bacteria take up, the more active in respiration the bacteria are. This is similar to the above-indicated development of the epiphytic community. This study compared the acetate uptake rate and mineralization percentage between the "inlet" (the front of the planted zone) and "outlet" (the end of the planted zone) of a wetland. The longitudinal variations for both water and sediment varied irregularly in wide ranges over two months. Although bacteria were a little more active near the outlet on average, no statistically significant difference in acetate uptake rate between the inlet and outlet was found for water (n = 4,P = 0.34) and sediment (n = l,P = 0.12). Likewise, no statistically significant difference in mineralization percentage was found for water (n = 4, P = 0.85) and sediment {n-l,P 0.96). 120  =  100 -j  100  Bacterioplankton  80 a o  1  .1  80 §  60 -  Sedimentary bacteria  60  '•P  40 -  S S3 40 .3 20  r =0.77 P =0.02  20 -  r =0.80 P =0.00  s  0-  0 200  400  600  800  20  1000  Acetate uptake rate, ug L"' h"  40  60  80  Acetate uptake rate, ug g h  1  1  100  1  Figure 5-2. Correlation between bacterial uptake rate and mineralization percentage in a surface flow constructed wetland treating woodwaste leachate in 2001.  5.3.3  Effects of fertilization and vegetation  Table 5-2 presents acetate uptake and mineralization in the three types of wetlands in 2002. Compared to the non-fertilized vegetated wetland, the fertilized wetland exhibited higher acetate uptake rates in sediment and water, being consistent with the responses of leucine incorporation and glucose turnover to fertilization. Unlike leucine incorporation and glucose turnover (Table 4-3), acetate uptake rate of epiphyton was not improved by fertilization. Acetate mineralization percentages of water and epiphyton were higher than glucose mineralization percentages (Table 4-3), and did not respond significantly to fertilization except in sediment. Acetate mineralization percentage was similar in the vegetated and open wetlands. Acetate uptake rates of water and sediment were much lower in the open wetland than those in the vegetated wetland.  5.4 5.4.1  Discussion Development of microbial colonization  Bacterial uptake rates of simple carbohydrates generally increase with greater bacterial densities and biomass (Wetzel, 1993; Wright and Hobbie, 1966). Similarly, the increasing trend of both acetate uptake rate and mineralization percentage over the colonization period in the C W indicated possible dependence of uptake rate on both microbial biomass and growth yield in the development stage of epiphyton. 121  Table 5-2. Effects of fertilization and vegetation on heterotrophic activity in surface flow constructed wetlands treating woodwaste leachate  Acetate uptake rate  Mineralization percentage  Fertilized, vegetated  Non-fertilized,  Non-fertilized, open  wetland  vegetated wetland  wetland  360 u g L " ! ^  92 ug L" h-  Water  797 ug L ' h'  Epiphyton  17.4mgm" h''  23.3 mg m" h"  N/A  Sediment  60.7 jagg h"  15.9 ug g-'h"  2.2 u g g h"  1  1  2  1  1  1  2  1  1  1  1  Water  79%  86%  91%  Epiphyton  76%  73%  N/A  Sediment  18%  0%  0%  1  1  Sampling on November 4, 2002 after adding ammonium nitrate for 9 consecutive weeks.  Microbial colonization and biofilm formation vary not only with the properties of microbial cells in the bulk water and surface roughness of substrata (Hunt and Parry, 1998; Mueller, 1996; Vanhaeke et al., 1990), but also with environmental parameters such as temperature, pH, nutrients, and ionic strength (Abbott et al., 1983; Stanley, 1983). The time for epiphytic maturation or acclimatization in CWs would depend upon the specific conditions.  5.4.2  Comparison to similar treatment systems  Heterotrophic potential is related to the trophic level of habitats (Stanley and Staley, 1977) and organic substrates (Wetzel, 1993). Nevertheless, polluted waters had a much higher substrate uptake rate than natural eutrophic lake water (Wright and Hobbie, 1966). Except for low pH and nutrient limitation, the CWs had much higher contents of excellent substrates such as V F A s (49-738 mg L" ). Reasonable comparison could only be made for acetate uptake to other 1  treatment systems. Toerien and Toerien (1985) reported maximum acetate uptake rates of 12.3-15.0 mg L" h" by 1  1  water, 0.52-1.36 mg m" h" by macrophyte leaves, and 1.0 ug cm" h" by surface sediment 2  1  2  1  (top 2-4 mm) in macrophyte beds receiving piggery effluent. Both the CWs treating 122  woodwaste leachate in this study and the macrophyte beds receiving piggery effluent were anaerobic, with high concentrations of acetic acid and C O D . Acetate uptake rate of bacterioplankton in the macrophyte beds was 34-fold greater than that in the C W (Table 5-1), probably because of the extremely high nitrogen (total N 2000 mg L" ) in the piggery effluent 1  that provided a good nutrient source and electron acceptor for denitrification. The C W had a bacterioplanktonic acetate uptake rate (Table 5-1) similar to that of a pulp mill waste aerated lagoon (194-726 fig L" h" ) that had acetate concentration less than 1 mg L" and excess levels 1  1  1  of nutrients. The C W in this study had 72 times higher acetate uptake rate of epiphyton, and 27 times higher uptake rate of sedimentary bacteria (Table 5-1) when compared to the macrophyte beds. The low acetate uptake rate in macrophyte beds was thought by Toerien and Toerien (1985) to result from grazing of epiphyton by large numbers of rat-tailed maggots. The higher acetate uptake rate by epiphyton and sedimentary bacteria in the C W might be attributed to the extremely low inorganic nutrients, which favor attached growth (Mueller 1996).  5.4.3  Temporal and longitudinal variations  Seasonal variation of bacterial activity is generally attributed to changes of temperature and substrate concentration (Sawyer and King, 1993; Toerien and Cavari, 1982; Wetzel, 1993). Like acetate uptake by the epiphytic microorganisms in macrophyte beds (Toerien and Toerien, 1985) and acetate mineralization in lake sediments (Hall et al., 1972), heterotrophic activity in the C W , however, was not stimulated by increased temperature. There are two possible reasons: a mixed population of microorganisms that had different optimum temperatures was using acetate; or the magnitudes of the changes of water temperature (717°C) and substrate concentration (such as 70-235 mg L" acetic acid) were not critical. 1  The vegetated zone of surface flow CWs usually follows plug-flow conditions (USEPA, 2000). The insignificant longitudinal variation in the vegetated C W in the present study might be attributed to wind-driven mixing because this C W did not have dense vegetation. In addition, acetate might always be saturated for bacterial uptake.  5.4.4  Relative contributions of water, sediment and epiphyton  Surfaces can stimulate bacterial activity, especially in substrate-limiting and low nutrient environments (Kjelleberg et al., 1982; Mueller, 1996; White et al., 1999). Attached bacteria 123  are more dense and bigger than free-living bacteria in natural aquatic environments (Kirchman and Ducklow, 1987). The attached cells have to assimilate more organic matter than freeliving bacteria, and take up substrate more rapidly than suspended cells (Fletcher, 1986). Coincidently, removal of suspended solids and C O D in surface flow CWs is reported to be better in wetland cells with plants than in adjoining cells without plants (Gearheart et al., 1989; Thut, 1989). Masbough (2002), however, did not find a remarkable difference in treatment performance between the vegetated and open wetland cells receiving woodwaste leachate through parallel operation. This phenomenon can be explained by the negligible contribution of epiphyton to the total heterotrophic uptake rate as Table 5-1 shows. A n earlier review (van Loosdrecht et al., 1990) has doubted the previous studies supporting the dogma of surface-enhanced growth. Sedimentary bacteria accounted for 72% of the total acetate uptake in the C W (Table 5-1). However, Toerien and Toerien (1985) found that major bacterial acetate uptake in macrophyte beds receiving organic effluents was associated with the bacterioplankton, and the contributions of sediment and epiphyton on the macrophytes were negligible. The difference in the relative importance of sediment could be explained by the different nutrient levels in these two treatment systems.  5.4.5  Effects of fertilization and vegetation  Nutrient conditions in the bulk water influenced the growth of suspended and attached microorganisms, as well as the cellular transport to the solid-water interface and the rate of attachment onto the substratum (Mueller, 1996). Acetate uptake rate by bacterioplankton was increased by fertilization to more than double that without fertilization. Sedimentary bacteria were stimulated substantially for both acetate uptake and mineralization by fertilization. The enhanced activity of microorganisms attached to sediment particles may be attributed to depletion of the excellent electron donor, V F A s , in the fertilized wetland due to promoted denitrification. Kozub and Liehr (1999) found that denitrification in the water of a wetland receiving landfill leachate was limited by the availability of easily degradable sources of organic carbon. In comparison to the activity of microorganisms attached on sediment, the epiphytic activity decreased somewhat due to fertilization, probably because the inert synthetic substratum could not take advantage of the plant surface for supply of electron donors and nutrients. A higher 124  specific plant surface area of the fertilized wetland (Table 5-2) converted its slightly lower measured acetate uptake rate of epiphyton to a higher rate as of water equivalent (Table 5-3). A n attached form of growth is beneficial to bacteria under low nutrient conditions, where starvation may occur (Kjelleberg et al., 1982; Mueller, 1996; White et al., 1999). Fertilization improved acetate uptake rate, especially of sedimentary bacteria, but the relative contribution of epiphyton stayed negligible (Table 5-3). Heterotrophic acetate utilization might be limited by lack of inorganic nutrients in water of the open wetland, while additional nutrients were likely released from the senescent emergent plants in the vegetated wetlands. Long-term analysis of heterotrophic acetate uptake is needed to clarify the effect of vegetation.  Table 5-3. Effects of fertilization and vegetation on heterotrophic acetate utilization in 2002 Type of wetland cell  Acetate uptake rate, mg m" h" 3  Water  Epiphyton Sediment 3  Contribution to total uptake, %  1  3  Water  Epiphyton Sediment  Fertilized, vegetated  797  69  1784  30  3  67  Non-fertilized, vegetated  360  27  455  43  3  54  Non-fertilized, open  92  N/A  58  61  N/A  3  As of water equivalent.  b  Not applicable.  b  39  5.5 Conclusions The total heterotrophic uptake rate and its distribution among water column, epiphyton and sediment were influenced by the levels of organic substrates and inorganic nutrients. Sedimentary bacteria contributed to the majority of the total acetate uptake in the C W while using only 16% for respiration; epiphytic bacteria accounted for 8% acetate uptake; and bacterioplankton used 55% acetate uptake for respiration. The CWs treating woodwaste leachate should be designed with deeper water to take advantage of the higher mineralization 125  percentage of suspended bacteria. This kind of design would reduce land requirement for wetland construction. Fertilization improved heterotrophic uptake of acetate. Combined treatment of woodwaste leachate with wastewater that has high concentrations of nitrate and nutrients, and a low concentration of organic carbon would probably improve overall microbial degradation. Wastewater from greenhouses is a possible source of nutrient amendment for woodwaste leachate treatment in CWs. An appropriate stabilization period should be considered in investigating and evaluating newly established treatment wetlands, during which treatment performance may be retarded by lower acetate uptake rate and mineralization percentage.  5.6 References Abbott, A., P.R. Rutter, and R.C.W. Berkeley. 1983. The influence of ionic strength, pH, and a protein layer on the interaction between Streptococcus mutans and glass surfaces. Journal of General Microbiology, 129: 439-445. Brix, H . 1994. Use of constructed wetlands in water pollution control -  historical  development, present status, and future perspectives. Water Science and Technology, 30: 209223. Fletcher, M . 1986. Measurement of glucose utilization by Pseudomonas fluorescens that are free-living and that are attached to surfaces. Applied and Environmental Microbiology, 52: 672-676. Gearheart, R.A., F. Kloop, and G. Allen. 1989. Constructed free surface wetlands to treat and receive wastewater: pilot project to full scale. In: D.A. Hammer (ed.), Constructed Wetlands for  Wastewater Treatment—Municipal, Industrial, and Agricultural. Lewis Publishers:  Chelsea, MI. ppl21-137. Hall, K.J., P . M . Kleiber, and I. Yesaki. 1972. Heterotrophic uptake of organic solutes by microorganisms in the sediments. Proceedings of the IBP-UNESCO  Symposium on Detritus  and Its Role in Aquatic Ecosystems, Pallanza, Italy. pp23-27. Hines, M . E . , K . N . Duddleston, and R.P. Kiene. 2001. Carbon flow to acetate and CI compounds in northern wetlands. Geophysical Research Letters, 28: 4251-4254. 126  Hobbie, J.E., and C C . Crawford. 1969. Respiration corrections for bacterial uptake of dissolved organic compounds in natural waters. Limnology and Oceanography, 14: 528-532. Hunt, A.P., and J.D. Parry. 1998. The effect of substratum roughness and river flow rate on the development of a freshwater biofilm community. Biofouling, 12: 287-303. Kirchman, D.L., and H.W. Ducklow. 1987. Trophic dynamics of particle-bound bacteria in pelagic ecosystems: a review. In: D.J.W. Moriarty and R.S.V. Pullin (eds.), Detritus and Microbial  Ecology in Aquaculture. International Center for Living Aquatic Resource  Management, Manila. pp54-82. Kjelleberg, S., B . A . Humphrey, and K . C Marshall. 1982. Effect of interfaces on small, starved marine bacteria. Applied and Environmental Microbiology, 43: 1166-1172. Kozub, D.D., and S.K. Liehr. 1999. Assessing denitrification rate limiting factors in a constructed wetland receiving landfill leachate. Water Science and Technology, 40: 75-82. Masbough, A . 2002. Effectiveness of Constructed Wetlands in Wood Leachate Treatment. M.A.Sc. thesis, the University of British Columbia, Vancouver, Canada. Moriarty, D.J.W., G.W. Skyring, G.W. O'Brien, and D.T. Heggie. 1991. Heterotrophic bacterial activity and growth rates in sediments of the continental margin of eastern Australia. Deep Sea Research, 38: 693-712. Mueller, R.F. 1996. Bacterial transport and colonization in low nutrient environments. Water Research, 30: 2681-2690. Parsons, T.R., and J.D.H. Strickland. 1962. On the production of particulate organic carbon by heterotrophic processes in sea water. Deep Sea Research, 8: 211-222. Sawyer, T.E., and G . M . King. 1993. Glucose uptake and end product formation in an intertidal marine sediment. Applied and Environmental Microbiology, 59: 120-128. Stanley, P . M . 1983. Factors affecting the irreversible attachment of Pseudomonas aeruginosa to stainless steel. Canadian Journal of Microbiology, 29: 1493-1499. Stanley, P . M . , and J.T. Staley. 1977. Acetate uptake by aquatic bacterial communities measured by autoradiography and filterable radioactivity. Limnology and Oceanography, 22: 26-37.  127  Thut, R.N. 1989. Utilization of artificial marshes for treatment of pulp mill effluents. In: D.A. Hammer (ed.), Constructed Wetlands for Wastewater Treatment—Municipal, Industrial, and Agricultural. Lewis Publishers: Chelsea, MI. pp239-244. Toerien, D.F., and B. Cavari. 1982. The effect of temperature on heterotrophic glucose uptake, mineralization,  and turnover  rates in lake  sediments.  Applied  and  Environmental  Microbiology, 43: 1-5. Toerien, D.F., and M . C . Toerien. 1985. Microbial heterotrophy in an effluent treatment system using macrophytes. Agricultural Wastes, 12: 287-312. USEPA. 2000. Constructed Wetlands Treatment of Municipal Wastewaters. Manual of EPA/625/R-99/010, U.S. Environmental Protection Agency, Cincinnati, OH. ppl3, 36-37. van Loosdrecht, M . C . M . , J. Lyklema, W. Norde, and A.J.B. Zehnder. 1990. Influence of interfaces on microbial activity. Microbiological Review, 54: 75-87. Vanhaeke, E., J.P. Remon, M . Moors, F. Raes, D. DeRudder, and A . van Petighem. 1990. Kinetics of Pseudomonas aerugunosa adhesion to 304 and 316-L stainless steel: role of cell surface hydrophobicity. Applied and Environmental Microbiology, 56: 788-795. Wetzel, R.G. 1993. Limnology, 2nd ed. CBS College Publishing: New York, N Y . pp487-519. Wetzel, R.G., and G.E. Likens. 2000. Limnological Analyses, 3rd ed. Academic Press: San Diego, C A . pp289-300. White, D . C , R.D. Kirkegaard, R J . Palmer Jr, C A . Flemming, G. Chen, K.T. Leung, C.B. Phiefer, and A . A . Arrage. 1999. The biofilm ecology of microbial biofouling, biocide resistance and corrosion. In: C W . Keevil, A . Godfree, D. Holt and C. Dow (eds.), Biofdms in the Aquatic Environment. The Royal Society of Chemistry: Cambridge, U K . pp 120-130. Wright, R.T., and J.E. Hobbie. 1966. Use of glucose and acetate by bacteria and algae in aquatic ecosystems. Ecology, 47: 447-464.  128  6  PERFORMANCE EVALUATION AND EFFECTS OF HYDRAULIC RETENTION TIME AND MASS LOADING RATE ON TREATMENT OF WOODWASTE L E A C H A T E IN MESOCOSM WETLANDS*  6.1  Introduction  CWs have been applied for treatment of a variety of wastewaters for decades all over the world (Kadlec and Knight, 1996; Cole, 1998; USEPA, 2000). Modeling studies (Polprasert and Agrawalla, 1994; Polprasert and Khatiwada, 1998; Polprasert et al., 1998) incorporating activities of suspended and attached bacteria with hydraulics of CWs and ponds do not support wider application of kinetic-based C W design. Designers have to derive design parameters by aggregating performance data from existing wetlands, usually pooling all the treatment mechanisms into one first-order reaction. Most of the existing CWs are used for treating municipal wastewater and agricultural runoff (Cole, 1998; USEPA, 2000). Moreover, performance of existing CWs has varied widely due to the influences of diverse natural factors and design parameters, such as the type of wastewaters, HRT, mass loading rate, and climate (USEPA, 2000; Carleton et al., 2001). Performance data have been derived from both longterm and short-term monitoring. Extrapolation of these performance data to a specific C W would lead to uncertainties about the validity of the parameters. Surface flow CWs have become an attractive alternative for on-site wood leachate treatment. Pilot-scale studies (Hunter et a l , 1993; Frankowski, 2000; Masbough, 2002; Tao et al., 2005) have proved the effectiveness of CWs in removal of COD, B O D , T & L , VFAs, and acute toxicity of wood leachate. Taylor and Carmichael (2003) reported a decline of BOD in aspen leachate over 10-12 d of storing in a leachate catch basin. Nevertheless, wood leachate has been characterized as acidic, of very high oxygen demand, and toxic to aquatic organisms (Peters et al. 1976; Field et al. 1988; Taylor et al. 1996; Zenaitis et al. 2002; Taylor and  * A version of this chapter has been submitted for publication. Tao, W., K.J. Hall, and S.J.B. Duff. Performance evaluation and effects of hydraulic retention time and mass loading rate on treatment of woodwaste leachate  in surface  Engineering, accept upon revisions. 129  flow  constructed wetlands. Ecological  Carmichael, 2003; Woodhouse and Duff 2004). Derivation of design parameters from performance of CWs treating other types of wastewaters is questionable. The relationships between treatment performance and design parameters have to be addressed in order to facilitate empirical design and improve operation of CWs. Mesocosm wetlands have the advantage of a greater controllability over pilot- and full-scale CWs. The influent quality, inflow rate and water depth could be easily regulated to examine the effects of such design parameters as HRT and mass loading rate. CWs depend mainly on microorganisms to remove organic matter (Polprasert et al., 1998; USEPA, 2000). The rate of biological decomposition may be limited by a variety of environmental factors. Heterotrophic activities of bacteria in CWs treating woodwaste leachate were influenced by the availability of organic substrates, electron acceptors, and inorganic nutrients (Tao and Hall, 2004; Tao et al., 2005). Although the constituents that are responsible for toxicity of wood leachate to microorganisms are mostly unknown, a close relationship has been found between low-molecular-weight tannins and methanogenic toxicity (Field et al., 1990). The ability to degrade toxic and recalcitrant compounds depends primarily on the presence of appropriate microorganisms and acclimatization time (Metcalf & Eddy, 2003). The levels of bacterial substrates and toxic constituents in a given C W are related to design parameters, such as H R T and mass loading rate. Considering the recalcitrant, toxic nature of the woodwaste leachate, H R T and mass loading rate may play an important role in the performance of surface flow CWs for treatment of woodwaste leachate. A n appropriate microbial community acclimatized to the woodwaste leachate may be established and has adequate contact time to attack the recalcitrant and toxic compounds at a longer HRT. A short HRT, however, reduces the land requirement for wetland construction. The influent should be loaded at a concentration and rate to provide sufficient substrates for heterotrophic microorganisms, while avoiding inhibition on heterotrophic activity. To date, no correlations of HRT and mass loading rate to C W performance have been reported for leachate treatment. The main purpose of this study was to examine the effects of HRT and mass loading rate on performance of surface-flow CWs for treatment of woodwaste leachate. Parallel operation of four mesocosm wetlands at different HRTs, during two periods fed with different strengths of influent, generated eight sets of performance data to derive statistically the effects of HRT and loading rate on treatment performance and kinetics. Meanwhile, the vertical variation in 130  planktonic and epiphytic biomass was examined along with the depth profiles of environmental conditions in the mesocosm wetlands to evaluate the potential of stratification in microbial decomposition. Microbial biomass in water, epiphyton and sediment at initial operation was also followed to estimate the maturation time for microbial communities, or the time required for a C W to reach a steady state. Moreover, the disturbance to performance due to precipitation and evapotranspiration was discussed.  6.2 6.2.1  Materials and Methods Setup of mesocosm wetlands  Four wooden mesocosms were set up with a slope of 3%. Each had an internal dimension of 0.36 m wide, 1.98 m long and 0.58 m tall. The bottom and sides were lined with double plastics (0.2 mm thick polyethylene sheet). A 22-cm layer of sandy loam was laid on the bottom. The soil had 70% sand (0.05-2 mm) and 5.3% organics. Broad-leaved cattail (T. latifolia) rhizomes (20 bunches with tubers and roots) were transplanted evenly to each mesocosm on April 7, 2003 (Figure 6-1 left). The influent was supplied at a constant flow rate, and dripped to the mesocosm front end. Each mesocosm had an outlet on the other end at 4 cm above the soil layer. The effluent overflowed through an elbow.  Figure 6-1. Layout (blue influent tank, blue pumping chamber, newly planted mesocosms, and outflow-measuring buckets) of the mesocosm wetland treatment system (left) and established mesocosm vegetation (right).  131  The mesocosm wetlands were operated initially at a shallow water depth (5 cm) with raw sewage for six weeks for development of cattail shoots. Between May 19 and 29 water depth was increased gradually up to 23 cm. B y this time, each mesocosm had developed 28-32 new plants. On June 1; raw woodwaste leachate was directly added to each mesocosm at about 600 mg L" COD. Starting from June 2, the mesocosm wetlands were operated at a constant water 1  depth of 25 cm with dilute woodwaste leachate. The woodwaste leachate was collected monthly from the pool and stored at 4°C. Every five days a pre-set amount of leachate was diluted with tap water in two 220-L high-density polycarbonate tanks to serve as influent (Figure 6-1 right). The influent COD concentration was targeted at 600 mg L" during the first 1  operating period (June 2 to August 25, 2003), about 10% raw leachate in tap water, and 1800 mg L" during the second operating period (August 26 to November 2, 2003), about 30% raw 1  leachate in tap water. Four different flow rates were applied to the mesocosm wetlands, aiming at different HRTs and mass loading rates. This area is characterized as a coastal climate with colder, wet winters and warm, drier summers. According to the meteorological data of the nearby Vancouver International Airport meteorological station, B C , Canada (Environment Canada, 2004), the monthly air temperature during the study period (April-October) varied between 9.3°C in April and 19.1°C in July, and the monthly precipitation was between 4.1 mm in August and 248.2 mm in October.  6.2.2  Field measurement and chemical analysis  As wetlands are subject to diverse natural factors, environmental parameters may undergo diurnal and daily changes. This study undertook field measurement weekly, following sample collection. Influent and effluent samples were collected with  125-mL high-density  polyethylene bottles. No field measurements were made or effluent collected during the startup (April 7 to June 29) and transition (August 26 to September 28) periods. Field measurements were made at the center of each mesocosm to approximate the average conditions. Water temperature and DO were measured at the depths of 2 cm, 7 cm, and 15 cm. Redox potential was determined at the depths of 2 cm, 7 cm and 15 cm, and on the bottom. pH was measured at the depth of 7 cm. The instruments and methods for field measurement and analysis of COD, T & L , V F A s , ammonia, nitrate plus nitrite, and orthophosphate were the same as those described in Section 2.2.2 for woodwaste leachate. A volumetric method was employed to check inflow rate and measure the outflow rate weekly. A rainfall gauge was 132  installed beside the mesocosm wetlands to record the weekly precipitation. Weekly evapotranspiration was calculated as the difference between the inflow rate plus precipitation and outflow rate. The number of emergent plants in each mesocosm was counted, whenever there was an apparent change in vegetation. The perimeter of submerged stems and width of submerged leaves were selectively measured and the averages were estimated. Void fraction was estimated with the size and number of emergent plants. Total plant surface area of each mesocosm wetland was estimated with water depth, average sizes of submerged plant stems and leaves, and plant number.  6.2.3  Determination of microbial biomass  Water, sediment and epiphyton samples were collected from the beginning of leachate feeding to determine the maturation time of microbial communities in terms of cellular ATP. Water and epiphyton samples were usually collected at 7 cm deep. Additional water and mature (6-8 weeks of colonization) epiphyton samples were collected at depths of 2 cm and 15 cm in one of the mesocosm wetlands (#2) to evaluate the vertical profile. Extraction and assay of cellular A T P followed Karl (1993) (see Appendix 10 for details). Each sample of water, epiphyton and sediment was extracted in duplicate. Chlorophyll-a in water was determined with the same method and instrument as described in Section 4.2.2. Mesocosm water for A T P determination was collected weekly for six consecutive weeks from the center of the mesocosm wetlands, using a 60-mL syringe with a long needle. Water was transferred to 125-mL autoclaved polypropylene bottles through a 200-u.m nylon sieve to remove macrozooplankton, roots and litter. Planktonic microorganisms were concentrated by passing 5 mL water through a nitrocellulose membrane filter (24 mm diameter; 0.45 urn pore size). The filter was immediately immersed in boiling (95-100°C) Tris buffer (20 m M ; p H 7.4) for 5 min to extract cellular ATP. After cooling, the extract was made up to 5 mL with Tris buffer, transferred to a 7-mL scintillation vial, and stored frozen. Epiphyton samples were collected on 600-grit silicon carbide waterproof sandpaper discs (0.95 cm each) adhered to P V C duct tape wrapped around Perspex plates (12.5 cm by 8 cm) 2  and immersed at given depths in the middle of the mesocosm wetlands. Mature epiphyton samples were retrieved for eight consecutive weeks. Three discs were placed in 10-mL cold orthophosphoric acid (49 g L" , 4°C) in a 15-mL centrifuge tube to extract cellular ATP for 20 1  133  min at 4°C. The acid extract was adjusted to p H 7.4 with NaOH, separated from the solid phase by centrifugation (1400xg, 5 min), transferred to a 7-mL scintillation vial, and stored frozen. Sediment was sampled along with overlying water weekly for six consecutive weeks, using 20-mL plastic syringe barrels (lock end removed). Three cores were taken near the inlet, middle and outlet of each mesocosm. The cores were sealed with rubber stoppers, and kept in an upright position. Before ATP extraction, the overlying water was poured out. The top 2-cm sediment of the three cores was pushed out by syringe plungers into a flask, mixed with a spatula, and passed through a 1-mm nylon sieve to remove gravel and larger ATP-containing organisms and plant rhizomes. Sediment slurry (0.1 mL) was transferred to 10-mL cold orthophosphoric acid (49 g L " , 4°C) in a 15-mL centrifuge tube and mixed. After an 1  extraction period of 20 min at 4°C, the extract was separated from the solid phase by centrifugation (2800xg, 7 min) and transferred to a 7-mL scintillation vial. The pH of the acid extract was adjusted to 7.4, and stored frozen. Two replicates of the sediment slurry were taken to measure sediment density (dry weight at 103°C for 24 h) and organic content (weight lossat550°Cfor>10h). The extracts were thawed and analyzed for A T P by the luciferin-luciferase method with an Aminco Model J4-7441B Chem-Glow photometer equipped with a strip chart recorder. The enzyme preparation was made from lyophilized firefly lantern extract (Sigma FLE50). A T P standards (2, 5, 10, 20, 40, and 70 p.g L" ) were made of the disbdium salt recovered from 1  equine muscle (Sigma Chemical, St. Louis, MO) with 20-mM Tris buffer for water samples, and potassium phosphate buffer (8.1 g L" K2HPO4) for sediment and epiphyton samples. 1  A T P standard or sample extract (0.2 mL) was injected to the enzyme preparation (0.2 mL in a 6x50 mm culture tube) in the photometer's holder to record the luminescence peak. The light intensity is proportional to the concentration of A T P in solution. One extraction blank was processed with each set of water, epiphyton or sediment samples. Two additional subsamples were spiked with an A T P standard and extracted to assess recovery. A T P concentration of extracts was converted to sample concentration according to the quantity of water, epiphyton or sediment used for extraction, volume of sample extract, and ATP recovery. Instrumental detection limit was 0.3 ug L  _ 1  A T P with Tris and 0.6 ug L" A T P with phosphate buffer. 1  134  6.2.4  Calculations of operating parameters and statistical analysis  H R T was calculated in equation 1-1. Treatment performance was expressed as: Concentration reduction efficiency (%) = 100x(Q' - Ce)ICi  (6-1)  Mass reduction efficiency (%) = 100x(C/xg/- CexQe)l(CixQi)  (6-2)  Reduction rate (g m" d" ) = (Ci*Qi - CexQe)IV  (6-3)  3  1  where Ci = the influent concentration (mg L" ); 1  Ce = the effluent concentration (mg L" ); 1  Qi — the inflow rate (m d" ); 3  1  Qe = the outflow rate (m d" ); and 3  1  V= water volume (m ). 3  Models for both plug-flow and complete-mix reactors have been used in surface flow CWs (Kadlec and Knight, 1996; Crites and Tchobanoglous, 1998). Tracer studies (Woods, 1995) showed that concentration-time distributions of wetland systems lie somewhere between the distributions of plug-flow and complete-mix conditions. It is easier to incorporate the effects of precipitation and evapotranspiration into the complete-mix model. Assuming first-order reaction kinetics for removal of organic pollutants, the complete-mix model was modified as: dCe y* ^± = CixQi - CexQe - k* V*Ce dt  (6-4)  dCe At steady state —^—= 0, dt n  k = (CixQi - CexQe)/(VxCe)  (6-5)  where k is the first-order reduction rate constant (d ). The difference between means was assessed by single-factor A N O V A . L S D was calculated to identify which two means were significantly different (Townend, 2002). The Bonferroni method was used to correct the significance level for multiple comparisons. Spearman's rank correlation analysis was used to test the significance of a monotonic relationship between two variables, giving P value and coefficient r . When the scatter of two variables showed a trend s  of straight line, Pearson's product moment correlation analysis was employed to test the significance of a linear relationship, giving P value and coefficient of determination R . 2  Correlation and difference were considered significant at a P<0.05. 135  6.3 6.3.1  R e s u l t s Operating conditions  The influent strength and inflow rate (Table 6-1 and Table 6-2) were relatively stable (Appendix 11). The mass loading rates to the four mesocosm wetlands (Table 6-3 and Table 64) were significantly different (P = 0.00) during the two operating periods. The larger variation coefficient of influent V F A s concentration, 32-51% during the first period and 2932% during the second period, was caused by a larger variability of leachate VFAs concentration. The influent had very low concentrations of nutrients during the two periods, 0.06-0.17 mg L" N H - N , 0.07-0.24 mg L" NO;-N,'and 0.07-0.19 mg L" P0 "-P on average. 1  1  1  3  3  4  The effluent contained 0.05-0.14 mg L" N H - N , 0.05-0.27 mg L" N O > N , and 0.15-2.01 mg 1  1  3  L" PO 4~ -P on average. 1  Table 6-1. Operating conditions of the mesocosm wetlands fed with a weak influent during the first operating period Mesocosm wetland  #1  #2  #3  #4  245±3  151±14  90±3  63±1  . 215±7  118±16  57±8  33±7  Influent pH  4.7±0.5  4.4±0.5  4.4±0.5  4.4±0.5  Mesocosm water pH  5.5±0.1  5.6±0.1  5.6±0.1  5.6±0.2  628±107  617±66  615±66  618±66  472±33  451±25  437±31  435±70  178±17  179±18  177±16  174±15  164±12  163±14  152±11  144±12  37±12  45±20  39±20  38±20  0.8±1.5  1.2±1.8  0.7±1.2  0.7±1.1  Inflow rate, L week"  1  Outflow rate, L week"  1  Influent COD, mg L"  1  Effluent COD, mg L" Influent T & L , mg L"  1  1  Effluent T & L , mg L"  1  Influent V F A s , mg L"  1  Effluent V F A s , mg L"  1  Mean ± SD June 2-August 25, 2003 (n = 9 for effluent, pH and flow rate; n = 12 for influent).  136  Table 6-2. Operating conditions of the mesocosm wetlands fed with a strong influent during the second operating period #1  #2  #3  #4  246±4  143±7  90±6  64±3  Outflow rate, L week"  245±8  129±16  85±3  61±9  Influent pH  4.6±0.3  4.5±0.2  4.5±0.2  4.5±0.2  Mesocosm water pH  4.7±0.2  4.9±0.1  5.2±0.2  5.7±0.2  Mesocosm Inflow rate, L week"  1  1  Influent COD, mg L"  1  1911±244  1719±227  1720±254  1678±266  Effluent C O D , mg L '  1  1737±168  1567±270  1422±114  1424±123  665±75  591±68  607±49  601±50  594±80  569±54  514±87  518±36  217±62  225±71  232±72  227±69  169±27  128±21  99±26  100±30  Influent T & L , mg L"  1  Effluent T & L , mg L"  1  Influent V F A s , mg L"  1  Effluent V F A s , mg L"  1  Mean ± SD from August 26 to November 2, 2003 (n = 8-9 for influent concentration; n = 6 for inflow rate and pH; n = 4 for effluent, excluding the data for the two weeks of October 13 and October 20 when there was heavy rainfall occurred in one day).  The nominal H R T (Table 6-3) of each mesocosm wetland during the first period was longer than that during the second period (Table 6-4) because of less evapotranspiration and more rainfall during the second period (Figure 6-2), especially for the wetlands with a longer HRT. The influent pH was very close between the two periods and among the mesocosm wetlands. The mesocosm wetlands had a significant vertical variation in temperature (P = 0.00-0.05 during the first period), D O (P = 0.00), and redox potential (P= 0.00), except the insignificant vertical variation of temperature during the second period (P = 0.76-0.86) (Appendix 12). Water temperature, D O and redox potential decreased with increasing depth (Figure 6-3). The bulk water in all of the wetlands was mainly anaerobic, while there was an aerobic water surface.  137  Table 6-3. Performance of woodwaste leachate treatment in the mesocosm wetlands fed with a weak influent during the first operating period #1  #2  #3  #4  Hydraulic retention time, d  5.3±0.1  9.0±1.0  16.4±0.9  25.1±1.7  pH increase  0.8±0.4  1.2±0.5  1.2±0.4  1.2±0.4  COD loading rate, g m" d"  125.5±24.7  72.5±8.1  43.0±3.8  32.2±3.2  T & L loading rate, g m" d"  35.2±3.4  21.2±1.7  12.2±1.0  8.9±0.8  7.5±2.2  4.7±2.1  2.6±1.4  2.2±0.9  COD reduction rate, g m" d"  44.0±22.5  30.0±3.9  22.9±4.8  20.7±3.9  T & L reduction rate, g m" d"  6.9±3.5  5.9±2.7  5.3±1.2  5.0±1.1  7.4±2.3  4.6±2.2  2.5±1.4  2.2±0.9  COD reduction rate constant , d"  0.09±0.05  0.07±0.01  0.05±0.01  0.05±0.01  T & L reduction rate constant , d"  0.04±0.02  0.04±0.02  0.04±0.01  0.04±0.01  COD mass reduction efficiency, %  34±10  41±5  53±9  64±8  COD concentration reduction efficiency, %  24±12  25±4  27±6  31±13  T & L mass reduction efficiency, %  19±9  27±11  43±9  56±10  T & L concentration reduction efficiency, %  8±10  7±11  11±6  17±8  V F A s mass reduction efficiency, %  98±4  98±3  96±10  99±2  V F A s concentration reduction efficiency, %  98±4  92±17  94±14  98±3  Mesocosm wetland  3  1  3  1  V F A s loading rate, g m" d" 3  1  3  1  3  1  V F A s reduction rate, g m" d" 3  1  a  a  1  1  Mean ± SD (» = 9). a  lst-order reaction rate constant; no estimate for V F A s since very low effluent concentration.  Redox potential did not vary much over H R T or among mesocosm wetlands. However, the mean redox potentials were in the order of surface sediment (50-131 mV)<mesocosm water (217-253 mV)<influent (283-332 mV) during the first period and surface sediment (90-121 mV)<mesocosm water (210-267 mV)<influent (290-328 mV) during the second period. 138  Table 6-4. Performance of woodwaste leachate treatment in the mesocosm wetlands fed with a strong influent during the second operating period Mesocosm wetland  #1  #2  #3  #4  Hydraulic retention time, d  4.9±0.1  8.8±0.5  13.6±0.5  19.3±1.9  pH increase  0.2±0.2  0.5±0.3  0.8±0.3  1.3±0.1  COD loading rate, g m" d"  1  396.4±31.2 210.8±32.9 110.6±10.5 89.0±18.2  T & L loading rate, g m" d"  1  131.7±16.5 70.5±11.5  41.7±3.5  31.3±3.6  44.4±13.7  28.3±8.5  13.3±4.8  11.5=b4.1  COD reduction rate, g m" d"  55.9±18.4  31.5*22.1  16.4±6.9  13.7±9.0  T & L reduction rate, g m" d"  10.6±6.5  6.6±5.5  8.1±1.4  5.5±3.4  20.9±3.0  12.8±8.9  6.4±2.9  4.3±2.9  COD reduction rate constant , d"  0.03±0.01  0.02±0.02  O.OliO.OO  0.01±0.00  T & L reduction rate constant , d"  0.02±0.02  0.01±0.01  0.02±0.00  0.01±0.01  V F A s reduction rate constant , d"  0.13±0.09  0.11±0.07  0.06±0.02  0.05±0.04  COD mass reduction efficiency, %  14±5  16±14  14±5  15±8  COD concentration reduction efficiency, %  13±4  6±18  9±9  10±9  T & L mass reduction efficiency, %  7±9  10±7  19±5  17±9  T & L concentration reduction efficiency, %  6±10  -1±8  13±13  12±11  V F A s mass reduction efficiency, %  40±35  44±26  48±6  43±28  V F A s concentration reduction efficiency, %  38±35  35±33  44±8  41±23  3  3  V F A s loading rate, g m" d" 3  1  3  1  3  1  V F A s reduction rate, g m" d" 3  1  3  1  a  1  3  1  Mean ± SD (« = 4 except n = 6 for loading rate). 3  lst-order rate constant.  The mesocosm wetlands had similar (P = 0.74) sediment densities of 0.72-0.75 g (dry weight) mL"  1  and similar (P = 0.08) sediment organic contents of 5.2-6.0% (Appendix 13). The  organic contents remained close to that of the native soil throughout the two periods. 139  B Precipitation • Evapotranspiration  S o  s o  2 2  Hydraulic loading rate  co S 60 &, -3 B  2 —'  o 3 5 -aS pq  •T —'  J~l  I 60  CN  OO fN  60  3  60  3  60  CN  Figure 6-2. Weekly precipitation and evapotranspiration in comparison to the hydraulic loading rate to mesocosm wetlands #1 (shortest HRT) and #4 (longest HRT) in 2003.  Dense emergent plants were established in mid-August, and started to die off gradually from late August. Around each submerged cattail stem were two or more separate leaves. The mesocosm wetlands had similar densities of emergent plants (Appendix 13). There were fewer, bigger emergent plants in July (24-30 per mesocosm; 9.0-10.5 cm of perimeter) than in October (46-48 per mesocosm; 7-9 cm perimeter). There were similar void fractions in the mesocosm wetlands, 0.95-0.97. Specific plant surface area of the mesocosm wetlands was estimated at 9.0-10.6 m m" water during the first period and 10.6-12.7 m m" water during 2  3  2  3  the second period. Numerous mosquito larvae were observed in the summer. Similarities of the mesocosm wetlands in sediment and vegetation justify the comparison across the mesocosms for HRT effects.  6.3.2  Temporal and vertical variations of biomass  Algal assemblages were visible in the water surface when the mesocosm wetlands were fed with raw sewage. The algae gradually disappeared after starting to feed woodwaste leachate. Chlorophyll-a concentration in mesocosm water was 0.35-1.36 u,g L ' on average after one 1  week of leachate feeding, and decreased to 0.07-0.27 u,g L" after 5 weeks. Assuming 1.5% 1  chlorophyll-a in the dry weight of planktonic algae ( A P H A et al., 1999) and 2.4 ug ATP mg"  1  dry weight organic matter (Karl, 1993; A P H A et al., 1999), algae constituted <9% total microbial biomass in the first two weeks and <2% in the next three weeks. While preparing 140  samples, macrozooplankton, roots and litter had been removed from samples. A T P measurements should reflect mainly the biomass of bacteria.  • depth of 2 cm S depth of 7 cm • depth of 15 cm  • depth of 2 cm S depth of 7 cm • depth of 15 cm  60  6  I  60  >, X  rfittTrlTll  o  II  —I  u  >  14 Weak influent  Strong influent  19  Hydraulic retention time (d)  500  • depth of 2 cm • depth of 7 cm • depth of 15 cm T_  mim  14  Strong influent  Weak influent  Hydraulic retention time (d)  IT  at  16 25  I  I  X-  c o X  o U  E3 influent • superficial sediment  400 300  I  ^  ^ ^  JL ^ ^ ^  200 100  Pi  0 16 25 Weak influent  Strong influent  Weak influent  Hydraulic retention time (d)  19 Strong influent  Hydraulic retention time (d)  Figure 6-3. Depth profiles of water temperature, D O and redox potential in the mesocosm wetlands with different HRTs during the first operating period with a weak influent (n = 8) and during the second period with a strong influent (n = 5-6). Error bar = SD.  Figure 6-4 presents the weekly variations of cellular A T P concentrations in water, epiphyton and sediment of the mesocosm wetlands after switching the influent from domestic sewage to diluted woodwaste leachate. Planktonic A T P concentration did not show a trend of development in the first six weeks (-0.37<r <0.84, P>0.05). Epiphytic A T P concentration did 5  141  not show a trend of development (r = -0.20, P>0.05) in the mesocosm with the longest H R T s  over eight weeks, and showed a significant increase (r = 0.81-0.94, P O . 0 5 ) in the mesocosm s  with the shortest HRT over 6 weeks of microbial colonization. Cellular A T P concentration in sediment showed a significant increase (r = 0.89, P O . 0 5 ) in the mesocosms with a HRT of 5 s  d and 25 d in the first six weeks, and appeared to increase (r = 0.77-0.80, T^O.05) in the s  mesocosms with a HRT of 9 d and 16 d over the first six weeks.  Elapsing weeks of operation  3500 3000 g? 2500 a  Number of weeks for colonization  HRT = 5 d —O^HRT = 9d -6—HRT=16d —o— HRT = 25 d  Sediment  —a—  /  ST 2000 1500 1000 500  • — ij —d 0 I—*"—i— — i — —i 1 5 1 2 3 4 Elapsing weeks of operation  6  Figure 6-4. Variation of cellular A T P in mesocosm wetlands with different HRT after switching influent from sewage to diluted woodwaste leachate. Average of two subsamples.  No statistically significant difference was found between any pair of the means of microbial biomass at water depths of 1 cm, 7 cm and 15 cm in water during the first six weeks of leachate feeding (P = 0.95) and in mature epiphyton during the following six weeks (P = 0.37). 142  6.3.3  Variations of treatment performance and kinetics  Table 6-3 and Table 6-4 summarize the treatment performance and kinetics of the mesocosm wetlands respectively during the two operating periods. V F A s were almost depleted and pH increased by about one unit in all of the mesocosms when the dilute leachate (10%) with a lower V F A s concentration was fed during the first period. When the less dilute leachate (30%) with a higher V F A s concentration was fed during the second period, pH was raised more by the mesocosm wetlands with a longer HRT. Mass reduction efficiency was apparently higher during the first period than the second period. The efficiency was significantly different across the mesocosm wetlands for removal of COD (P = 0.00, L S D = 10%) and T & L (P = 0.00, L S D = 12%) during the first period; the mesocosm wetlands with a longer HRT were more efficient. The efficiency was not significantly different between any pair of the mesocosm wetlands either for removal of COD (P = 0.99) or T & L (P = 0.12) during the second period. The reduction efficiency for COD was higher than or similar to that for T & L on average of individual mesocosms. V F A s were removed much more efficiently in a shorter H R T than C O D and T & L . When fed with 10% leachate during the first period, all of the mesocosm wetlands almost depleted the VFAs. When fed with 30% leachate during the second period, the mesocosm wetlands removed only 40-48% VFAs, and had similar (P = 0.97) mass reduction efficiencies. There was no significant difference in T & L reduction rate between any pair of the mesocosm wetlands during the two periods (P = 0.37 and 0.69 respectively). Reduction rate was significantly different across the mesocosm wetlands during the first period for removal of COD (P = 0.00, L S D = 15.7 g m" d" ) and V F A s (P = 0.00, LSD = 2.4 g m" d ), and during 3  1  3  _1  the second period for removal of COD (P = 0.01, LSD = 34.5 g m" d" ) and VFAs (P = 0.02, 3  1  LSD = 11.3 g m" d" ). There were not obvious differences in COD and T & L reduction rates 3  1  between the two periods. The reduction rate constants for COD and T & L were much larger during the first period than during the second period. There was not a significant difference in the T & L reduction rate constant between any two of the mesocosms during the first period (P = 0.65) and the second period (P = 0.85). The reduction rate constant of V F A s during the second period was not significantly different across the mesocosm wetlands (P = 0.18). The reduction rate constant of COD was significantly different across the mesocosm wetlands during the first period (P = 143  0.01, L S D = 0.04 d" ) and the second period (P = 0.05, L S D = 0.02 d" ). The COD reduction 1  1  rate constant was larger in the mesocosm wetlands with a shorter HRT. The ratio of T & L ThOD to COD during the first period increased significantly (P = 0.00-0.01) from influent of 0.35-0.36 on average to effluent of 0.42-0.45. The ratio of V F A s ThOD to COD decreased significantly (P = 0.00) from influent of 0.09-0.11 on average to effluent of 0.00. The ratio of effluent T & L ThOD to C O D (0.43-0.46 on average) during the second period remained (P = 0.23-0.87) as high as that of influent (0.43-0.45), and the ratio of V F A s ThOD to COD decreased significantly (P = 0.00-0.05) from influent of 0.19-0.21 on average to effluent of 0.12-0.14. In comparison to the hydraulic loading rates applied to the mesocosm wetlands in this study, evapotranspiration during the first period and precipitation during the second period (Figure 62) had profound impacts on removal efficiency and its variation. There was an apparent difference between the concentration reduction efficiency and mass reduction efficiency (Table 6-3 and Table 6-4). Moreover, the performance was more variable during the second period due to heavy rainfall events.  6.4 6.4.1  Discussion Time for microbial maturation  For the treatment wetlands that rely on microbiological processes to remove contaminants, performance can only be properly evaluated after initial maturation of a microbial community. Development of the microbial biomass in the media of a subsurface flow C W system typically requires three to six months (USEPA, 2000). A previous study (Tao and Hall, 2004) found that acetate gross uptake rate and mineralization percentage of the epiphytic community in a surface flow C W treating woodwaste leachate at an influent COD of 2475-6563 mg L" and an 1  HRT of >7 d increased considerably over three weeks of microbial colonization. The present study found differences of maturation time among water, epiphyton and sediment of the surface flow mesocosm wetlands. The mesocosm wetlands appeared to require a longer maturation time for the attached microbial community than the suspended community. Microorganisms normally accumulate within surface environments due to simultaneous attachment and growth (Caldwell, 1987). Either the epiphytic and sedimentary communities  144  were mainly established by a slow attachment process, or the attached bacteria had a growth rate lower than the suspended bacteria in the mesocosm wetlands. The maturation times of the epiphytic and sedimentary communities varied across the mesocosm wetlands. The rate of microbial colonization and biofilm formation is influenced by the abundance of planktonic bacteria, bacterial properties, surface roughness of substrata, and environmental parameters (Stanley, 1983; Vanhaeke et al.,' 1990; Mueller, 1996; Hunt and Parry, 1998). There was no significant difference (P = 0.50) in planktonic ATP between any pair of the mesocosm wetlands. The quality of the influent to the mesocosm wetlands was almost the same. The difference among the mesocosm wetlands operated in parallel was in HRT and loading rates. However, the variations of maturation time across the mesocosm wetlands could not be interpreted with the differences in H R T and mass loading rate. The difference in maturation time between epiphytic and sedimentary communities was likely due to the differences in phylogenic composition and physiology.  6.4.2  Disturbance of weather to performance  This study demonstrated a substantial difference between concentration reduction efficiency and mass reduction efficiency due to the net impacts of precipitation and evapotranspiration on the mesocosm water quality and outflow rate. On one hand, evapotranspiration concentrates the C W water, while precipitation dilutes C W water. On the other hand, evapotranspiration extends H R T due to reduced outflow rate and subsequently improves effluent quality, while precipitation reduces H R T and tends to flush bacterioplankton out of CWs, and subsequently deteriorates effluent quality. The magnitude of the net impacts will depend on the local climate character and the C W design parameters. Only mass balancebased performance data can be more confidently referred for design of CWs at sites with a different climate, and used for comparative evaluation across different sites and over time. The water quality models using a simply averaged flow rate may bring out significant errors. Unlike evapotranspiration that changed gradually, rainfall happened in durations (Figure 6-2) much shorter than the HRTs. Relative to the variability of influent strength and inflow rate, the treatment performance calculated on the basis of weekly sampling was not stable, especially in the second operating period due to disturbance of heavy rainfall events. The stochastic character of rainfall and the diurnal periodicity and seasonal fluctuation in evapotranspiration contribute toward much of the variability in wetland effluent concentrations (USEPA, 2000). 145  6.4.3  Implications of vertical variation  The color of woodwaste leachate changes from amber at pH 4 to dark brown at pH>6. The sunlight could not penetrate far through the dark mesocosm water (Tao et al., 2005). Therefore, the field measurements showed a significant vertical variation of temperature. According to the chlorophyll-a concentration of the inland water trophic classification system (Ryding and Rast, 1989), the mesocosm water was ultra-oligotrophic, suggesting negligible algal photosynthesis. The higher DO in the water surface was likely due to surface reaeration and macrophytic photosynthesis. Biological oxidation in sediment was likely responsible for the decreasing redox potential toward the sediment. In an investigation of the vertical distribution of heterotrophic bacteria in an alder swamp, Kjoller et al. (1985) found a decreasing number of bacteria per g dry litter with depth from the surface layer to 25 cm deep. Bacterial growth rates of epiphyton in a surface flow CW treating secondary municipal effluent were significantly higher near the water surface, but with no difference between the depths of 1.5 cm and 7.5 cm (Pollard et ah, 1995). Nevertheless, there was no significant vertical variation of microbial biomass in the mesocosm wetlands treating woodwaste leachate, despite the vertical variations of temperature, DO and redox potential. The physiology or activity of microorganisms in the aerobic water surface was probably different from that in the deep water, though there was no significant vertical variation in biomass. Layering of physiologically different microorganisms occurs due to a gradient of redox conditions near the interface of oxygen-containing and oxygen-free waters (Pedros-Alio et a l , 1993).  6.4.4  Effect of hydraulic retention time  Degradation rates in natural wetlands depend upon the ratios of major polymers, such as lignin, cellulose, and hemicellulose (Moran et al., 1989). Hemicellulose is generally degraded at an equal or faster rate than cellulose in most wetlands (Hedges et al., 1985). Long-term exposure is needed to induce and sustain the enzymes and bacteria required for degradation of the recalcitrant, toxic organic compounds (Metcalf & Eddy, 2003), such as T & L . The HRTs applied to the mesocosm wetlands for treatment of woodwaste leachate (5-25 d) exceeded the recommended values (USEPA, 2000) for treatment of municipal wastewater and spanned the general operating values of existing CWs (USEPA, 2000; Carleton et al., 2001) for treatment of runoff. 146  The similarities of the environmental conditions (pH, D O , sediment, vegetation, etc.) among the mesocosm wetlands justified the derivation of H R T effect by comparing the mesocosm wetlands. The mass reduction efficiencies for C O D and T & L increased significantly and linearly with H R T (R - 0.99, P = 0.00) during the first period. The slight increase of mass 2  reduction efficiency with increasing HRT during the second period was likely due to the more recalcitrant influent character. The negative relations of H R T to reduction rate constants for COD, T & L , and V F A s confirmed that a longer H R T resulted in a more recalcitrant effluent. The depletion of V F A s even in the mesocosm wetland with the shortest HRT and the increasing efficiency for C O D and T & L removal with increasing H R T suggest that fermentation was likely the limiting stage for treatment of woodwaste leachate. Fermentation of less diluted woodwaste leachate would generate V F A s in addition to influent input, resulting in an insignificant change of V F A s reduction efficiency with longer HRTs. Consequently, woodwaste leachate may be better treated in a series of C W cells. The first C W cell can aim at substantially removing V F A s and raising pH at a shorter HRT; the following cells can be used for further treatment of such recalcitrant contaminants as T & L at longer HRTs. It has been shown that a number of cells in series can consistently produce a higher quality effluent (USEPA, 2000). This kind of layout minimizes the short-circuiting effect. To reduce 90% of the raw woodwaste leachate T & L would require 7 and 13 mesocosms respectively at a reaction rate constant of 0.02 d" and 0.01 d" , with a H R T of 20 d for 1  1  individual mesocosm wetlands. 6.4.5  Effect of mass loading rate  Mass loading rate is an important parameter for empirical design and operation. A higher loading rate is desirable for increasing microbial production and respiration, while severe or prolonged overloading will inhibit bacterial activity. A n upper limit of 6-11 g m" d" B O D 2  1  loading to surface flow CWs has been suggested (WPCF, 1990; USEPA, 2000) for treatment of municipal wastewater by aerobic oxidation. However, a C O D loading rate of 0.4-19.7 kg m" 3  d" has been applied for landfill leachate treatment in anaerobic systems (Forgie, 1988; 1  Young, 1991). A moderate range of mass loading rates was applied to the mesocosm wetlands for treatment of woodwaste leachate. The loading rates to the mesocosm wetlands were significantly correlated to the reduction rates of COD (P = 0.02, R = 0.62), T & L (P = 0.00, R 2  2  = 0.81) and V F A s (P = 0.00, R = 0.90). The positive, moderate-strong linear correlations of 2  147  mass loading rate to reduction rate suggested no inhibitory effect of woodwaste leachate at an organic loading of up to 0.4 kg m" d  C O D and 0.13 kg m" d" T & L . A near-linear  3  3  1  relationship of COD removal rate (51.7-124.2 g m" d" ) to loading rate was also reported (Jing 3  1  et al., 2002) for treatment of municipal wastewater in CWs with HRTs of 1-4 d. Whereas, mass reduction efficiency of the mesocosm wetlands had a significant, negative relationship with the loading rates of COD (r, = -0.78, P<0.05), T & L (r = -0.92, P<0.01), and V F A s (r = s  s  -0.85, P<0.02). The optimum mass loading rate is determined by a compromise between reduction rate and reduction efficiency. The significant increase of T & L reduction rate with increasing loading rate suggested the presence of a microbial community that could break down recalcitrant, toxic T & L in the mesocosm wetlands. The toxic and recalcitrant compounds may be degraded by serving as a growth substrate or an electron acceptor for heterotrophic bacteria, or adsorbed onto microorganisms (Metcalf & Eddy, 2003). A relatively constant supply of toxic and recalcitrant organic compounds can lead to better biodegradation performance than intermittent additions (Metcalf & Eddy, 2003). The COD and T & L loading rates to the mesocosm wetlands varied in a narrower range, with a coefficient of variation of <20%. However, heavy rainfall events during the second period disturbed the supply of toxic, recalcitrant T & L to the microorganisms in the mesocosm wetlands in addition to possible washout of the suspended bacteria, especially in those mesocosms with longer HRTs. More rainfall and lower temperatures were likely the other factors that resulted in less efficient removal of COD, T & L and V F A s , and smaller reduction rates for C O D and T & L through the mesocosm wetlands with a longer HRT during the second period than during the first period.  6.5  Conclusions  Woodwaste leachate could be fed to surface flow CWs without inhibition at loading rates as high as 0.4 kg m" d" COD and 0.13 kg m" d" T & L . Higher reduction efficiency was attained 3  1  3  1  by increasing H R T up to 25 d. Treatment performance may also vary with temperature and influent strength. Because of substantial impacts of precipitation and evapotranspiration and a considerable development process of the microbial communities, only mass balance-based data of CWs with a mature microbial community could provide a proper base for performance assessment.  148  Moderate reductions of COD and T & L were found for treatment of woodwaste leachate in CWs, although the V F A s at a loading rate of up to 7.5±2.2 g m" d" could almost be depleted. 3  1  Layout of C W cells with different HRTs in series may be tested for fast removal of V F A s and further treatment of recalcitrant contaminants.  6.6  References  A P H A et al. 1999. Standard Methods for the Examination of Water and Wastewater, 20th edition. American Public Health Association, American Water Works Association, and Water Environment Federation. Caldwell, D.E. 1987. Microbial colonization of solid-liquid interfaces. In: M . L . Shuler and W.A. Weigand (eds.), Annals of the New York Academy of Sciences, Vol. 506. The New York Academy of Sciences: New York, N Y . pp274-280. Carleton, J.N., T.J. Grizzard, A . N . Godrej, and H.E. Post. 2001. Factors affecting the performance of stormwater treatment wetlands. Water Research, 35: 1552-1562. Cole, S. 1998. The emergence of treatment wetlands. Environmental Science and Technology, 32: 218A-223A. Crites, R.W., and G. Tchobanoglous. 1998. Small and Decentralized Wastewater Management System. WCB—McGraw-Hill: New York, N Y . Environment  Canada.  2004.  National  Climate  Data  and  Information  Archive.  http://www.climate.weatheroffice.ec.gc.ca/Welcome_e.html, updated on January 27, 2004. Field, J., G. Lettinga, and L.H.A. Habets. 1990. Measurement of low molecular weight tannins:  indicators of methanogenic  toxic tannins.  Journal  of Fermentation and  Bioengineering, 69: 148-153. Field, J.A., M.J.H. Leyendeckers, R. S. Alvarez, G. Lettinga, and L . H . A . Habets. 1988. The methanogenic toxicity of bark tannins and the anaerobic biodegradability of water soluble bark matter. Water Science and Technology, 20: 219-240. Forgie, D.J.L. 1988. Selection of the most appropriate leachate treatment methods Part 1: a review of potential biological leachate treatment methods. Water Pollution Research Journal of Canada, 23: 308-328.  149  Frankowski, K . A . 2000. The Treatment of Wood Leachate Using Constructed Wetlands. M.A.Sc. thesis, The University of British Columbia, Vancouver, Canada. Hedges, J.I., G.L. Cowie, and J.R. Ertel. 1985. Degradation of carbohydrates and lignins in buried woods. Geochim. Cosmochim. Acta, 49: 701-727. Hunt, A.P., and J.D. Parry. 1998. The effect of substratum roughness and river flow rate on the development of a freshwater biofilm community. Biofouling, 12: 287-303. Hunter, R., A . E . Birkbeck, and G. Coombs. 1993. Innovative marsh treatment systems for control of leachate and fish hatchery wastewaters. In: G . A . Moshiri (ed.), Constructed Wetlands for Water Quality Improvements. Lewis Publishers: Baca Raton, F L . pp477-484. Jing, S.-R., Y . - F . Lin, T.-W. Wang, and D.-Y. Lee. 2002. Microcosm wetlands for wastewater treatment with different hydraulic loading rates and macrophytes. Journal of Environmental Quality, 31:690-696.  ~ .  Kadlec, R.H., and R.L. Knight. 1996. Treatment Wetlands. Lewis Publishers: Boca Raton, F L . Karl, D . M . 1993. Total microbial biomass estimation derived from the measurement of particulate adenosine-5'-triphosphate. In: P.F. Kemp, B.F. Sherr, E.B. Sherr and J.J. Cole (eds.), Handbook of Methods in Aquatic Microbial Ecology. Lewis Publishers: Boca Raton, F L . pp359-372. Kjoller, A . , S. Struwe, and K . Vestberg. 1985. Bacterial dynamics during decomposition of alder litter. Soil Biology and Biochemistry, 17: 463-468. Masbough, A . 2002. The Effectiveness of Constructed Wetlands for Treatment of Wood Leachate. M.A.Sc. thesis, The University of British Columbia, Vancouver, Canada. Metcalf & Eddy. 2003. Wastewater Engineering: Treatment and Reuse, 4th ed. McGraw-Hill: New York, N Y . Moran, M . A . , R. Benner, and R.E. Hodson. 1989. Kinetics of microbial degradation of vascular plant material in two wetland ecosystems. Oecologia, 79: 158-167. Mueller, R.F. 1996. Bacterial transport and colonization in low nutrient environments. Water Research, 30: 2681-2690.  150  Pedros-Alio, C , J.G. Cantizano, and J.L. Calderon. 1993. Bacterial production in anaerobic water columns. In: P.F. Kemp, B.F. Sherr, E.B. Sherr and J.J. Cole (eds.), Handbook of Methods in Aquatic Microbial Ecology. Lewis Publishers: Boca Raton, FL. pp519-530. . Peters, G.B., H.J. Dawson, B.F. Hrutfiord, and R.R.Whitney. 1976. Aqueous leachate from western red cedar: effects on some aquatic organisms. Journal of Fisheries Research Board Canada, 33: 2703-2709. Pollard, P.C., J.A. Flood, and N.J. Ashbolt. 1995. The direct measurement of bacterial growth in biofilms on emergent plants of an artificial wetland. Water Science and Technology, 32: 251-256. Polprasert, C , and B . K . Agrawalla. 1994. A facultative pond model incorporating biofilm activity. Water Environment Research, 66: 725-732. Polprasert, C , and N.R. Khatiwada. 1998. A n integrated kinetic model for water hyacinth ponds used for wastewater treatment. Water Research, 32: 179-185. Polprasert, C , N.R. Khatiwada, and J. Bhurtel. 1998. A model for organic matter removal in free water surface constructed wetlands. Water Science and Technology, 38: 369-377. Ryding, S.-O., and W. Rast. 1989. The Control of Eutrophication of Lakes and Reservoirs. U N E S C O , Paris, France. Stanley, P . M . 1983. Factors affecting the irreversible attachment of Pseudomonas aeruginosa to stainless steel. Canadian Journal of Microbiology, 29: 1493-1499. Tao, W., and K . J . Hall. 2004. Dynamics and influencing factors of heterotrophic bacterial utilization of acetate in constructed wetlands treating woodwaste leachate. Water Research, 38: 3442-3448. Tao, W., K.J. Hall, and S.J.B. Duff. 2005. Heterotrophic bacterial activities and treatment performance of surface flow constructed wetlands receiving woodwaste leachate. Water Environment Research, 78: in press. Taylor, B.R., J.S. Goudey, and N . B . Carmichael. 1996. Toxicity of aspen wood leachate to aquatic life: Laboratory studies. Environmental Toxicology and Chemistry, 15: 150-159. Taylor, B.R., and N . B . Carmichael. 2003. Toxicity and chemistry of aspen wood leachate to aquatic life: field study. Environmental Toxicology and Chemistry, 22: 2048-2056. 151  Townend, J. 2003. Practical Statistics for Environmental and Biological Sciences. John Wiley & Sons: Chichester, England. USEPA. 2000. Constructed Wetlands Treatment of Municipal Wastewaters. Manual of EPA/625/R-99/010, U.S. Environmental Protection Agency, Cincinnati, OH. Vanhaeke, E., J.P. Remon, M . Moors, F. Raes, D. DeRudder, and A . Van Petighem. 1990. Kinetics of Pseudomonas aerugunosa adhesion to 304 and 316-L stainless steel: role of cell surface hydrophobicity. Applied and Environmental Microbiology, 56: 788-795. Woods, A . 1995. Constructed wetlands in water pollution control: fundamentals to their understanding. Water Science and Technology, 32: 21-29. WPCF. 1990. Natural Systems for Wastewater Treatment. Manual of Practice FD-16, Water Pollution Control Federation, Alexandria, V A . Woodhouse, C , and S.J.B. Duff, 2004. Treatment of log yard runoff in an aerobic trickling filter. Water Quality Research Journal of Canada, 39: 232-238. Young, J.C. 1991. Factors affecting the design and performance of upflow anaerobic filters. Water Science and Technology, 24: 133-155. Zenaitis, M . G . , H . Sandhu, and S.J.B. Duff. 2002. Combined biological and ozone treatment of log yard runoff. Water Research, 36: 2053-2061.  152  7  EFFECTS LOADING  OF HYDRAULIC RATE  ON  RETENTION  TIME  MICROORGANISMS  IN  AND MASS MESOCOSM  WETLANDS* 7.1  Introduction  CWs have become a cost-effective alternative for on-site treatment of woodwaste leachate (Frankowski, 2000; Hunter et al., 1993; Masbough, 2002). Laboratory studies have demonstrated that microbial degradation is the major mechanism of CWs to remove C O D , T & L , and V F A s from woodwaste leachate. In the pilot-scale wetlands treating woodwaste leachate, a substantial positive effect of ammonium nitrate amendment has been found on heterotrophic activities. The heterotrophic bacteria in CWs might be composed of variable, mixed species established during a long residence time under variable substrate and ambient conditions (Westermann, 1993). The biomass and activity of heterotrophic bacteria can be influenced by a variety of factors, such as organic substrate concentration, inorganic nutrient level, electron acceptor supply, grazing pressure, and toxicity. In CWs, these factors are closely related with such design parameters as HRT and mass loading rate. Mesocosm wetlands provide a more controllable facility than pilot-scale CWs to examine the effect of individual factors, such as H R T and mass loading rate, on treatment performance and microbial activity. Simultaneous determination of treatment performance as well as microbial biomass, heterotrophic production, and substrate assimilation and mineralization would assist in understanding the kinetics of biological carbon removal processes in CWs. A previous study (Flood et al., 1999) has found that the abundance, productivity, respiratory activity, and phylogenic  composition of the  epiphytic community responded  differently to  the  environmental conditions in two CWs receiving secondary municipal effluent. Microbial biomass, heterotrophic productivity and heterotrophic activity were examined in the present  * A version of this chapter has been drafted for publication. Tao, W., K.J. Hall, and S.J.B. Duff. Biomass and heterotrophic activities of microorganisms in surface flow mesocosm wetlands: responses to hydraulic retention time and mass loading rate and correlation with treatment performance. 153  study respectively by measuring cellular ATP, H-leucine incorporation into bacterial protein, and heterotrophic assimilation and mineralization of C-acetate in four mesocosm wetlands 14  treating woodwaste leachate. The three microbiological techniques were utilized in the present study to provide complementary profiles of the microbial communities in the wetlands. The mesocosm wetlands were operated at different HRTs during two periods with different strengths of influent, yielding eight mass loading rates. The effects of HRT and mass loading rate on treatment performance have been discussed in the preceding chapter. One of the main purposes of this chapter is to evaluate the effects of H R T and mass loading rate on biomass, heterotrophic production, and acetate assimilation and mineralization of the microbial communities in surface flow CWs treating woodwaste leachate. The heterotrophic bacteria in surface flow CWs may inhabit three components, water column, plant surface and sediment. Attention has been paid to the role of bacterioplankton and benthic bacteria in the decomposition of organic matter in natural aquatic environments. A few studies (Flood et al., 1999; Kozub and Liehr, 1999; Pollard et al., 1995) have investigated the dynamics of either the bacterioplankton or the epiphyton in relation to the environmental conditions of surface flow CWs. Moreover, different relative contributions of water, epiphyton and sediment to microbial activities have been reported for CWs (Tao and Hall, 2004; Tao et al., 2006; Toerien and Toerien, 1985; Toet et al., 2003). Simultaneous determination of mass balance-based treatment performance and microbial dynamics in the three C W components has provided the present study with a more reliable database to connect treatment performance with microbial activities in CWs. The current trend in wetland modeling is the application of simple mass balance models (USEPA, 2000), which integrate many complex reactions in a treatment wetland into one overall biological reaction. This chapter statistically correlates the treatment performance with the heterotrophic activities of the three microbial communities in water, epiphyton and sediment of the mesocosm wetlands. It bridges C W design parameters and treatment performance with activities of the microbial communities.  7.2 7.2.1  Materials and Methods Sample preparation  The setup and operating conditions of the mesocosm wetlands have been presented in Section 6.2.1 and Section 6.3.1. In six weeks after switching the influent from sewage to diluted 154  woodwaste leachate, the mesocosms were left without microbial sampling for development of the microbial communities in water, epiphyton and sediment. More than 4 weeks were left for a transition from the weak influent (10% raw leachate) to the strong influent (30% raw leachate) during the second operating period. Sampling procedures for microbial populations were devised to minimize exposure to the oxygen, light and temperature conditions prevailing at the surface. Water was collected at 7 cm in the center of the mesocosm wetlands, using a 60-mL syringe with a long needle. The water samples were transferred to 125-mL autoclaved polypropylene bottles through a 200-urn nylon sieve to remove macroorganisms and detritus. To avoid any loss of the limited number of cattails due to epiphyton sampling by removing parts of a cattail stem, microorganisms were allowed to colonize on 600-grit silicon carbide waterproof sandpaper discs (0.95 cm  2  each), which were adhered to P V C duct tape, wrapped around Perspex plates (12.5 cm by 8 cm) and immersed 7 cm deep in the middle of the mesocosm wetlands. Sandpaper discs (mature epiphyton) were retrieved after 6 weeks of colonization. Sediment was sampled along with overlying water, using 20-mL plastic syringe barrels (lock end removed). Three cores were taken near the inlet, middle and outlet of each mesocosm. The cores were sealed with rubber stoppers at both ends, and kept in an upright position. Before microbiological examination, the overlying water was poured out. The top 2-cm sediment was pushed out by a syringe plunger into a flask, mixed with a spatula, and passed through a 1-mm nylon sieve to remove coarse material and larger ATP-containing organisms. Two mL of the sediment slurry was diluted with 40 mL filter-sterilized mesocosm water to make sediment suspension. One mL aliquots of the sediment slurry were taken in duplicate to measure sediment density (dry weight at 103°C for 24 h) and organic content (weight loss at 550°C for >10 h).  7.2.2  Determination of cellular ATP  Cellular A T P concentrations of water, mature epiphyton, and sediment were determined using the same methods and instruments as described in Section 6.2.3.  7.2.3  Determination of heterotrophic production  This research mainly followed the approach used by Jorgensen (1992) and Ward and Johnson (1996) to track H-leucine incorporation into protein for determination of heterotrophic bacterial production rate in mesocosm water, sediment and epiphyton (see Appendix 14 for 155  detailed procedures). Mesocosm water (9-mL) in 20-mL plastic syringes, epiphyton (1 disc in 9-mL sterile mesocosm water) in 20-mL plastic syringes, or sediment (1 mL suspension plus 8 mL sterile mesocosm water) in 20-mL scintillation vials was added with 0.9 mL or 2.4-2.8 uCi L-[4,5- H]leucine (Amersham Biosciences U K Ltd., Buckinghamshire, U K ; specific radioactivity of 157 C i mmol" ) solution and 0.1 mL of 1.31 mg L" L-leucine solution. The 1  1  exact radioactivity of the H-leucine solution was determined by counting DPMs in 0.1 mL of the isotope solution with the Beckman LS6500 multi-purpose scintillation counter. For each sample, one killed control and two live subsamples were prepared. One mL of 25% w/w glutaraldehyde solution was used to fix the controls and terminate incubation of the live subsamples. Isotope dilution experiments were conducted initially to estimate leucine saturation levels for bacteria in water, sediment and epiphyton samples (Appendix 15). It appeared that addition of more than 39 iig L" leucine (final) would change the kinetics. It was estimated that leucine 1  incorporation rates of water, epiphyton and sediment were approaching the maximum at leucine concentrations of 13.1 ug L" . The addition of 13.1 ug L" non-labeled leucine (final) 1  1  was assumed to saturate bacterial uptake without causing increased kinetics and to reduce the risk of isotope dilution. Leucine incorporation at concentrations up to 13.1 ug L" has not been 1  reported to enhance bacterial growth within a few hours of incubation (Jorgensen, 1992; Tuominen, 1995). During the first operating period, the samples were incubated at the in situ water temperatures for 30 min. During the second period, water and epiphyton were incubated for 45-60 min and sediment was incubated for 60-90 min. Time course experiments (Appendix 15) were conducted initially with water, epiphyton and sediment samples to set appropriate incubation times. The incubation times were short enough to prevent any change of population and activity, while long enough to bring out a clear difference between the live subsamples and the control. After incubation, the fixed epiphyton samples were sonicated for 3 min in an Aquasonic Model ultrasonic cleaner (50/60 Hz, average sonic power of 45 W) to detach bacteria from the substratum (Schaule et al., 2000). The incubation solution of water and epiphyton samples in syringes was injected to 20-mL scintillation vials. Bacterial protein was extracted with 2.5 mL of 20% cold T C A for 15 min at 4°C. The extracted protein was collected on polycarbonate 156  filters with a pore size of 0.2 urn (see Appendix 15 for filter selection) through vacuum filtration, followed by ice-cold 5% TCA washing and 80% ethanol washing (see Appendix 15 for the usefulness of ethanol washing). Each polycarbonate filter was placed in 5 mL Ecolite scintillation cocktail (ICN, Costa Mesa, CA). DPMs of H-leucine incorporated in bacterial 3  protein were counted by the scintillation counter. The counter was operated in the external standard mode to correct automatically for counting efficiency and color quench. Leucine incorporation rate was calculated with equation 4-6. Intracellular isotope dilution was omitted with regard to the addition of non-labeled leucine. The measured leucine incorporation rate was corrected for extracellular isotope dilution, by multiplying the following factor: _ . . . ... . j-» (labeled leucine) + (non - labeled leucine) Extracellular dilution factor = -— (labeled leucine) 1  ._ (7-1)  In order to compare the relative contributions of water, epiphyton and sediment toward the total heterotrophic production of an entire mesocosm wetland, the areal incorporation rate of epiphyton (VI, ug m" h" epiphyton) and the volumetric rate of sediment (VI, u,g L" h" 2  1  1  1  sediment) were converted to water equivalent rate (17, ug L" h" water equivalent) in the 1  1  following equations: U= Vl*a/l000, for epiphyton  (7-2)  U= m2/25, for sediment  (7-3)  where a = specific surface area (m plant surface m" water); 2 = assumed microbially active surface sediment layer (cm); and 25 = water depth of the mesocosm wetlands (cm). 7.2.4  Determination of heterotrophic assimilation and mineralization  There were high concentrations of VFAs in woodwaste leachate, so C-acetate was used as a 14  radioactive tracer for assimilation and mineralization of organic substrates. With the separately measured acetate concentration of wetland water, this study determined the actual acetate assimilation and mineralization rates with a method (Appendix 16) similar to that used by Tao and Hall (2004) and Chidthaisong et al. (1999). For each sample, one killed control and two live subsamples were prepared with mesocosm water (9 mL) in 20-mL plastic syringes, 157  epiphyton (1 disc in 9 mL sterile mesocosm water) in 20-mL plastic syringes, or sediment (1 mL suspension in 8 mL sterile mesocosm water) in 25-mL flasks. One mL of 37% w/w formaldehyde was added to fix the controls. One mL of [l- C]acetate (Dupont Nen Products, 14  Boston, M A ; specific activity of 56 C i mol" ) solution was added to each syringe or flask, 1  giving a final concentration of 20-25 p.g L" or 0.2 uCi per syringe or flask. The exact 1  radioactivity added to each sample was determined by counting 0.1 mL working isotope solution. Acetate in the mesocosm water samples was analyzed by gas chromatography (HPGC 5880A; Supelco, Inc GC Bulletin 751G). Incubation with C-acetate was conducted in closed syringes and flasks in the dark at the in 14  situ water temperature for 15 min during the first operating period with lower concentrations of acetate in mesocosm water, and 2-3 h during the second period with higher acetate concentrations. Each flask was sealed with a serum cap, through which the rod of a small plastic cup was inserted before replacing the cap. The cup was used to hold a folded glass filter (24 mm in diameter) and was set in the flask's headspace. In order to set an appropriate incubation time for the mesocosm samples, experiments were done initially to assess the time course of acetate uptake by the bacterioplankton, epiphytic bacteria, and sedimentary bacteria in 4 h (Appendix 17). At the end of incubation, one mL of 37% w/w formaldehyde was added to the live subsamples to terminate incubation. The epiphyton samples were sonicated for 3 min in the ultrasonic cleaner to detach bacteria. The incubation solution in the syringes was transferred to 25-mL flasks, which were immediately sealed with a serum cap. Through the serum cap, each flask was injected with 0.2 mL of 0.2N  H2SO4  to release respired CO2. Phenethylamine (0.15 mL)  was injected into the folded glass filter in the small cup to absorb CO2 for 40 min, during which the flasks were swirled several times. Bacterial particles were collected by vacuum filtration through a nitrocellulose membrane filter (0.45 um pore size). Each particle-retaining or C02-trapping filter was placed in 5 mL Ecolite scintillation cocktail. DPMs of the C 1 4  acetate assimilated (on particle-retaining filter) and respired (in C02-trapping filter) by bacteria were counted with the scintillation counter. Acetate uptake rate was calculated in equation 5-1. When the measured DPMs of the C 1 4  acetate assimilated and respired are input separately, this equation gives acetate assimilation rate and mineralization rate respectively. The gross uptake rate is the sum of assimilation rate 158  and mineralization rate. Substituting acetate uptake rates for leucine incorporation rate in equation 7-2 and equation 7-3, the areal epiphyton and volumetric sediment uptake rates could be converted to water equivalent rate.  7.2.5  Statistical analysis  The differences between the two sampling periods and between the mesocosm wetlands were assessed by two-way A N O V A . LSDs were calculated to identify the pairs that were significantly different (Townend, 2002). The Bonferroni method was used to correct the significance level for multiple comparisons. Spearman's rank correlation analysis was used to test the significance of a monotonic relationship between two variables, giving P value and coefficient r„. Multiple regression analysis was employed to test the linear relationship between one dependent variable and three independent variables, giving P value and coefficient of determination R . Correlations and differences were considered significant at a P <0.05.  7.3  Results  7.3.1  Variation of microbial biomass  Figure 7-1 shows the variations of cellular A T P concentration with HRT during the two periods of operation at different influent strengths (original data in Appendix 18). Except for a significant difference in sedimentary A T P concentration between the two periods (P = 0.00), there were usually insignificant differences in cellular ATP concentration of water, epiphyton or sediment between the two periods (P = 0.64-0.94) or any pair of the mesocosm wetlands (P = 0.10-0.81). Cellular A T P concentration was not correlated to the loading rates of COD, T & L , or V F A s (0.63<r <0.59), except for the significant correlations of sedimentary ATP concentration with 5  T & L loading rate (r = -0.76, P<0.05) and V F A s loading rate (r, = -0.85, P<0.02). s  Sedimentary bacteria accounted for the majority of the total cellular A T P (Table 7-1). Bacterioplankton and epiphytic bacteria contributed nearly equally to the remaining A T P . Bacterioplankton and epiphyton contributed a little more to the total cellular ATP during the second period than during the first period.  159  10  • Weak influent • Strong influent  Water 60  mm  OH  CD  U  5(5)  9(9)  16(14)  25(19)  Hydraulic retention time (d) 1.0  • Weak influent • Strong influent  Epiphyton  0.8 60  •s  0.6  OH  0.4  a  3  "3 U  0.2 0.0 5(5)  9(9)  16(14)  25(19)  Hydraulic retention time (d) 4000  • Weak influent • Strong influent  Sediment  T V T 3200 00  OH  5 ii  u  2400 1600 800  i  i i 5(5)  9(9)  16(14)  25(19)  Hydraulic retention time (d)  Figure 7-1. Variation of cellular A T P concentration over mesocosm wetlands with different HRTs during two operating periods at different influent strengths. Bracketed value on x-axis = H R T for strong influent. Error bar = SD. n = 5.  160  Table 7-1. Relative contributions of water, epiphyton and sediment to the total microbial biomass and heterotrophic activities of mesocosm wetlands treating woodwaste leachate Cellular ATP  Influent strength  10% raw leachate  3  30% raw leachate  3  Leucine  Acetate  Acetate  Acetate gross  incorporation  assimilation  mineralization  uptake  3:3:94  46:23:31  45:3:52  46:4:50  46:4:50  7:9:84  43:50:7  50:11:39  69:14:17  63:13:24  Percentages of water:epiphyton:sediment on average of the four mesocosm wetlands. a  10% during the first operating period; 30% during the second period.  7.3.2  Variation of heterotrophic production  Figure 7-2 shows the variations of heterotrophic leucine incorporation rate with H R T during the two operating periods at different influent strengths (original data in Appendix 18). H R T did not have a significant effect on the incorporation rates of water, epiphyton, and sediment LP = 0.43-0.94). There were significantly lower bacterioplanktonic incorporation rates LP = 0.00), lower epiphytic incorporation rates LP = 0.00), and higher sedimentary incorporation rates LP = 0.01) at the low influent strength (10% leachate) than with the high influent strength (30% leachate). HRT did not affect the differences between the two periods (P = 0.57-0.94). The  mass loading rates of C O D , T & L and V F A s were significantly correlated with  bacterioplanktonic leucine incorporation rate (r = 0.88-0.95, P<0.01) and sedimentary leucine s  incorporation rate (r = -0.91 to -0.93, P<0.01). Epiphytic incorporation rate was significantly 5  correlated to V F A s loading rate (r = 0.76, P<0.05), but moderately correlated to the loading s  rates of COD and T & L (r, = 0.60-0.67, P>0.05). No significant correlation of leucine incorporation rate of water, epiphyton or sediment was found with cellular A T P concentration of water, epiphyton or sediment (r$ = 0.01-0.71). Planktonic bacteria accounted for nearly a half of the total heterotrophic leucine incorporation rate (Table 7-1). Sedimentary bacteria contributed a little more than epiphyton at the low influent strength and much less with the high influent strength.  161  7.5  • Weak influent • Strong influent  0)  6.0  e o  4.5  o _ &  .a  '43  S 3.0  g o  1.5  3  <u J  0.0  •mi 5(5)  c o  1o  6  9(9)  16(14)  25(19)  0.8 i  Hydraulic retention time (d) • Weak influent Epiphyton • Strong influent  0.6-  j  1.0  s  Water  &  8  M  .S -8 g o 3 u  AAAA  0.2 H 0.0  5(5)  9(9)  16(14)  25(19)  Hydraulic retention time (d) 40 u  32  •4-»  2  • Weak influent • Strong influent  Sediment  d  .2 —  "e  24  I •« 8 «  16  2  •S U  _g G  3 t>  •J  £  I  IiiM l 5(5)  9(9)  16(14)  25(19)  Hydraulic retention time (d)  Figure 7-2. Variation of heterotrophic leucine incorporation rate over mesocosm wetlands with different HRTs during two operating periods at different influent strengths. Bracketed value on x-axis = HRTs for strong influent. Error bar = SD. n = 5.  162  7.3.3 Variations of heterotrophic acetate assimilation and mineralization Figure 7-3 shows the variations of heterotrophic acetate uptake with H R T during the two periods (original data in Appendix 18). Table 7-2 summarizes the results of analysis of variance. The epiphytic assimilation and mineralization rates were significantly higher at the high influent strength, resulting in a higher gross uptake rate during the second period. The sedimentary assimilation and mineralization rates were significantly lower at the high influent strength, resulting in a lower gross uptake rate during the second period. The significantly higher bacterioplanktonic assimilation rate and the significantly lower mineralization rate at the low influent strength, however, resulted in an insignificant difference of gross uptake rate between the two periods. Mesocosm #1 with the shortest H R T during the first period had significantly lower acetate assimilation rates of water, epiphyton, and sediment. The differences of acetate assimilation between the two periods were usually affected by HRT, mainly due to the slow acetate assimilation in mesocosm #1 during application of the weak influent. Besides, the differences of sedimentary acetate assimilation, mineralization, and gross uptake between the two periods were affected by HRT, mainly due to the slow acetate use by sediment in mesocosm #3 (HRT = 14 d) during application of the strong influent.  Table 7-2. Two-way analysis of variance of heterotrophic acetate utilization in mesocosm wetlands with different HRTs during two operating periods at different influent strengths Acetate uptake Acetate assimilation rate  Acetate mineralization rate  Acetate gross uptake rate  Component  Period  HRT  Interaction  Water  0.00 (28)  0.00 (55)  0.02 (95)  Epiphyton  0.01 (0.4)  0.02 (0.8)  0.89  Sediment  0.00 (357)  0.03 (697)  0.00(1192)  Water  0.00 (48)  0.49  0.33  Epiphyton  0.00 (0.5)  0.76  0.43  Sediment  0.00 (380)  0.01 (743)  0.00 (1272)  0.21  0.48  0.30  Epiphyton  0.00 (0.9)  0.27  0.70  Sediment  0.00 (658)  0.01 (1286)  0.00 (2200)  Water  Value outside bracket = P value; bracketed value = L S D in the same units as Figure 7-3; n = 5. 163  750  • Mineralization 600 H • Assimilation  e  3  "a  Water  450  i l l  Weak influent  Strong influent  Hydraulic retention time (d) 10 8  e  • Mineralization • Assimilation  Epiphyton  o 3  o  J3 60  £  Weak influent | Strong influent Hydraulic retention time (d)  Hydraulic retention time (d) Figure 7-3. Variation of heterotrophic acetate uptake over mesocosm wetlands with different HRTs during two operating periods at different influent strengths. Error bar = SD. n = 5.  164  Table 7-3 summarizes the significance of correlations of acetate utilization with mass loading rates. There were significant, positive correlations of bacterioplanktonic mineralization rate with  the  loading rates of C O D , T & L and  VFAs,  and  negative correlations  of  bacterioplanktonic assimilation rate with C O D , T & L and V F A s loading rates. As a result, bacterioplanktonic gross uptake rate was not significantly correlated to the mass loading rates. The cellular A T P concentration of water, epiphyton or sediment was insignificantly correlated to acetate assimilation rate, mineralization rate and gross uptake rate of water, epiphyton or sediment  (-0.59<^<0.56). There  was  usually  no  significant  correlation  of  leucine  incorporation rate of water, epiphyton or sediment with the rate of acetate assimilation, mineralization or gross uptake rate of water, epiphyton or sediment (-0.59<r.s<0.56, P>0.05), except a significant correlation between bacterioplanktonic leucine incorporation rate and bacterioplanktonic acetate mineralization rate (r = 0.90, P<0.01). s  Table 7-3. Coefficients of Spearman's rank correlation of acetate utilization with mass loading rates to the mesocosm wetlands Acetate uptake  Wetland  Loading rate  component  Chemical oxygen Tannin and lignin demand  Gross uptake rate  Mineralization rate  Assimilation rate  Volatile fatty acids  Water  0.10  0.14  0.31  Epiphyton  0.40  0.52  0.67  Sediment  -0.64  -0.67  -0.71  Water  0.74*  0.76*  0.86*  Epiphyton  0.60  0.69  0.81*  Sediment  -0.57  -0.60  -0.67  Water  -0.79*  -0.74*  -0.64  Epiphyton  -0.05  0.12  0.24  Sediment  -0.64  -0.67  -0.71  Significant, based on the critical r of 0.738 at n = 8. s  165  Planktonic and sedimentary bacteria contributed approximately equally to the majority of acetate assimilation and mineralization at the low influent strength, while epiphytic bacteria played a minor role. The relative importance of bacterioplankton and epiphytic bacteria increased somewhat during the application of the strong influent, and the relative importance of sedimentary bacteria decreased substantially (Table 7-1).  7.3.4 Correlation of treatment performance with heterotrophic activities The operating conditions and treatment performance have been presented in Section 6.3.1 and Section 6.3.3. The cellular A T P concentrations of water, epiphyton, and sediment were not correlated to the reduction rates of COD, T & L and V F A s and reduction rate constants of COD and T & L (-0.69<r <0.71, P>0.05). The cellular A T P concentrations of water, epiphyton and 5  sediment together were usually not significantly correlated to the reduction rates and reduction rate constants (R = 0.73-0.89, P = 0.02-0.13). 2  Leucine incorporation rate of water, epiphyton or sediment was usually not correlated to the reduction rates and reduction rate constants. However, the reduction rate of COD, T & L or V F A s exhibited a significant correlation with the leucine incorporation rates of water (positive), epiphyton (negative) and sediment (negative) together (R = 0.83-0.88, P = 0.022  0.05); and the reduction rate constants of COD and T & L had a significant correlation with the leucine incorporation rates of water (positive), epiphyton (negative) and sediment (positive) together (R = 0.83-0.93, P = 0.01-0.05). 2  There was no significant correlation of the reduction rates to acetate assimilation, mineralization or gross uptake rates of water, epiphyton, sediment (-0.76<r.s_0.62), or the three C W components together {R <0.56). The reduction rate constant of COD or T & L was usually not correlated to the acetate uptake rate of water, epiphyton or sediment, but was significantly correlated with the three components together, either for assimilation, mineralization, or gross uptake (R = 0.81-0.97, P = 0.00-0.05). 2  When the significant multiple correlations of reduction rates or reduction rate constants with microbial activities were connected with the significant Spearman's rank correlations of microbial activities to mass loading rates or influent strengths, a conceptual model (Figure 74) was created for organic carbon removal by the three microbial communities in treatment wetlands. 166  Heterotrophic production Mass \ + loading rate ^^+?  + /  f  +  Bacterioplankton 46% (43%) Epiphytic bacteria 23% (50%)  N.  Influent\ strength i  j  (Reductior rate  \ +  Sedimentary bacteria 31% (7%)  Rate constant,  Heterotrophic assimilation Bacterioplankton 45% (50%) Epiphytic bacteria 3% (11%)  1  )  Sedimentary bacteria 52% (39%)  —  \  —  \  Heterotrophic mineralization Bacterioplankton 46% (69%) Epiphytic bacteria 4% (14%)  1  V  Sedimentary bacteria 50% (17%)  Figure 7-4. Conceptual model of wetlands for removal of COD, T & L , and V F A s from woodwaste leachate. "+" and " - " = statistically significant correlations (r^>0.74; n = 8); "+?" = moderate correlations with C O D (r = 0.60) and T & L (r = 0.67) and a significant s  s  correlation with V F A s (r = 0.76). The percentages represent the relative contributions of s  water, epiphyton and sediment to the entire wetland fed with weak and strong (in brackets) influents. 167  7.4 7.4.1  Discussion Establishment of microbial communities and effect of hydraulic retention time  Leucine comprises 7.3% of bacterial protein, and bacterial carbon production is 0.86 times bacterial protein production (Simon and Azam, 1989). The ATPxarbon ratio of microbial community is close to 0.4% by weight (APHA, 1999; Karl, 1993). Based on the average rate of leucine incorporation and cellular ATP concentration of water, suspended bacteria in the mesocosm wetlands had a generation time of 0.8-2.2 d. The HRTs applied to the mesocosm wetlands were much longer than the generation times to avoid washout of suspended bacteria, giving sufficient time to establish a mature microbial community. A longer HRT provides sufficient detention time for establishment of abundant bacterioplankton. Therefore, cellular A T P concentration and heterotrophic activities of bacterioplankton were usually not affected by HRT, after a startup period for microbial development. Similarly, the generation time of epiphytic bacteria was estimated at 3.1-6.3 d on average for individual mesocosm wetlands during the application of the weak influent and 0.7-1.0 d during the application of the strong influent. The epiphytic generation times were much shorter than the time for epiphytic community maturation in terms of cellular A T P concentration. Both reproduction of epiphytic bacteria and attachment of bacterioplankton could contribute to the development of the epiphytic community in the mesocosm wetlands. The epiphytic generation time of the mesocosm wetlands was much longer than that (4.5-7.0 h) of the wetlands receiving secondary sewage effluent (Flood et al., 1999). Compared to the woodwaste leachate, the sewage effluent had a much lower oxygen demand and higher influent ammonia and phosphorus concentrations. Sedimentary bacteria had a generation time of 71-169 d on average for individual mesocosm wetlands. In 6 weeks of initial operation, however, the sedimentary community reached maturity, suggesting that attachment of bacterioplankton to sediment played a dominant role in establishment and maintenance of the sedimentary community in the mesocosm wetlands.  7.4.2  Effects of mass loading rate and influent strength  Heterotrophic production and respiration in surface flow CWs were influenced by availability of organic substrates and electron acceptors (Kozub and Liehr, 1999; Tao and Hall, 2004; Tao et al., 2006; Toet et al., 2003). Temperature influence in surface flow CWs may not be as 168  profound as it is for some of the more conventional processes (Wittgren and Maehlum, 1996) since wetlands may be more substrate-limiting than temperature-stressed. The influent to the mesocosm wetlands during the second period was about 3 times as strong as that during the first period. Consequently, the strong influent during the second period fostered higher leucine incorporation rates for bacterioplankton and epiphyton than the weak influent during the first period. The lower sedimentary leucine incorporation rate during the second period was likely caused by the lower temperature and the inhibitory effect of the strong influent. Different levels of toxic effect of wood leachate on microorganisms have been reported (Field et al. 1988; Taylor and Carmichael 2003; Taylor et al. 1996). The toxicity of wood leachate is usually attributed to low pH and phenolic compounds, such as tannins, which were exaggerated  in the mesocosm wetlands with the strong influent. According to the  measurements of redox potential (Figure 6-3), sedimentary bacteria of the mesocosm wetlands were likely facultative and strict anaerobes (Westermann, 1993), which are more sensitive to low temperature and toxic substances (Metcalf & Eddy, 2003). The bacteria incorporating leucine might not be the same as those oxidizing acetate. Unlike the leucine incorporation rate, the acetate assimilation rates of bacterioplankton were significantly lower during the period of high strength influent application. The lower bacterioplanktonic assimilation rates as well as the lower sedimentary assimilation and mineralization rates during this period were likely again due to the lower temperature and inhibitory effect of the strong influent. The eight mass loading rates were different combinations of two influent strengths and four HRTs. The significant effect of influent strength plus occasionally significant interactions of influent strength with H R T resulted in different variations of microbial activities of the bacterioplanktonic, epiphytic and sedimentary communities with loading rates (Figure 7-4). The positive correlations between leucine incorporation of bacterioplankton and epiphyton with mass loading rates were likely due to more bacterial substrates supplied at higher loading rates. The higher influent strengths were usually accompanied by higher loading rates. The negative correlations between sedimentary leucine incorporation rate and mass loading rates as well as the negative correlation between acetate assimilation and mass loading rates could be the same reasons as the negative responses of sedimentary leucine incorporation and bacterioplanktonic acetate assimilation to the strong influent and lower temperature. 169  Low organic substrate conditions are beneficial to attached bacteria (Mueller, 1996; Tao and Hall, 2004; Tao et al., 2006; White et al., 1999), and subsequently affect the relative contributions of water, epiphyton and sediment to heterotrophic activities. Sedimentary bacteria in the mesocosm wetlands had less importance to the entire wetland system at higher V F A s during the second period relative to the first period (Table 7-1). Bacterioplankton and epiphyton usually played a larger role during the application of the strong influent. Greater contribution of epiphytic bacteria during the application of the strong influent suggested little restriction of substrate diffusion through the epiphyton. 7.4.3  Microbial diversity  Heterotrophic bacterial communities consist of heterogeneous populations in various physiological states (Wetzel and Likens, 2000). Not all bacteria incorporate leucine (Kirchman et al., 1985; Tabor and Neihof, 1982; Ward and Johnson, 1996). The relative contributions of water, epiphyton and sediment to the entire wetland system were different in terms of cellular A T P concentration, leucine incorporation rate arid acetate uptake rates. Bacterial productivity could not alone explain the observed numbers of bacteria. Several mechanisms for bacterial turnover are known, such as cell death, removal by grazing, and viral attack (Hobbie and Ford, 1993; Pace, 1988). Subsequently, natural systems are always in a dynamic state. No correlations were found between cellular A T P concentration and heterotrophic activities in the mesocosm wetlands in the present study, implying the potential importance of protozoan grazing and other factors in controlling microbial biomass. A small loss or net gain of inorganic nitrogen and orthophosphate through the mesocosm wetlands (Table 7-4) implied that there were usually no nitrogen and phosphorus limitations for bacterial growth in the mesocosm wetlands. The balances between loading and discharge of orthophosphate and inorganic nitrogen were usually less than the estimates of bacterial consumption, possibly due to nutrient remineralization by fermentative bacteria, nutrient leaching from senescent plants and detritus, nutrient release from sediment under anaerobic conditions, and nitrogen fixation by bacteria inside the mesocosm wetlands (Reddy and Graetz, 1988; USEPA, 2000). Amendment of ammonium nitrate to a pilot-scale C W treating woodwaste leachate has revealed that electron acceptors like nitrate might be more limiting than nutrients for heterotrophic production and mineralization (Section 4.4.2). Providing that bacteria are not grazed excessively and are not limited by essential nutrients, they can be 170  expected to respond to substrate increases by increasing in population size (Wright, 1988). The insignificant differences in the cellular A T P concentrations of water and epiphyton between the two periods suggested that grazing pressure, rather than substrate and nutrients, might regulate the active biomass of the mesocosm wetlands treating woodwaste leachate.  Table 7-4. Nutrient balance of mesocosm wetlands treating woodwaste leachate in 2003 #1  #2  #3  #4  5  9  16  25  17(11)  20(9)  15(4)  9(5)  17(19)  12(11)  7(4)  4(4)  PO ," loading (discharge) , mg P wk"  17(32)  21(33)  13(81)  8(66)  Nitrogen consumption , mg wk"  14-18  12-15  9-12  11-14  3-4  2-3  2-3  2-3  5  9  14  19  15(15)  9(8)  5(4)  2(3)  1  52(66)  32(26)  20(23)  15(11)  1  34(37)  27(31)  17(31)  12(24)  25-32  25-32  26-34  26-34  5-6  5-6  5-7  5-7  Mesocosm wetland First operating period (June 2 to August 25) Hydraulic retention time, d NH3 loading (discharge) , mg N wk" a  1  N O ~ loading (discharge) , mg N wk" a  3  a  1  1  b  1  Phosphorus consumption , mg wk' b  1  Second operating period (August 26 to November 2) Hydraulic retention time, d NH3 loading (discharge), mg N wk"  1  N O ; loading (discharge), mg N wk" PO4"" loading (discharge), mg P wk" Nitrogen consumption, mg wk"  1  Phosphorus consumption, mg wk" a  b  1  Derived from mean concentration and flow rate (n = 6-9). Estimated with leucine incorporation rates and typical nutrient requirements for bacterial  growth in anaerobic processes (Metcalf & Eddy, 2003).  7.4.4  Microbial bridges of pollutant loading and wetland performance  CWs rely mainly on microorganisms to remove organic pollutants (USEPA, 2000). Usually, the water quality models for CWs integrate all reactions to one overall rate constant. A newer model for organic carbon removal has been introduced to surface flow CWs (Polprasert et a l , 171  1998) to predict treatment performance with separate first-order reaction rate constants for suspended and attached bacteria. Nevertheless, no studies have been reported so far on the type of microbial degradation kinetics and its variation with C W design and operating parameters, such as influent strength and mass loading rate. Bacterioplankton, epiphyton, and sedimentary bacteria in the mesocosm wetlands treating woodwaste leachate responded differently to influent strength and mass loading rates. The distribution of heterotrophic activities among water, epiphyton and sediment varied with influent strength. The multiple correlation of reduction rate or reduction rate constant of COD, T & L or V F A s with heterotrophic production, assimilation or mineralization rate was not always positive (Figure 7-4). Overall, heterotrophic production, acetate assimilation, and acetate mineralization transferred respectively a negative, negative, and positive correlation between influent strength and reduction rate constant. Both heterotrophic production and substrate mineralization contribute to contaminant reduction. Further efforts need to be made to integrate the three modules into a single model so that the conceptual model can be verified by data sets of influent strength and reduction rate constant.  7.5  Conclusions  Cellular A T P concentrations in water, epiphyton and sediment of the mesocosm wetlands did not relate to the variations of HRT, influent strength, and mass loading rate, and was likely regulated by other factors such as protozoan grazing. A H R T value of longer than 5 d was adequate to establish and maintain the microbial communities in water, epiphyton and sediment of mesocosm wetlands for treatment of woodwaste leachate. The anaerobic sedimentary bacteria were likely more sensitive to the lower temperature and possible inhibitory effect of the strong influent and higher loading rates, while the heterotrophic activities of bacterioplankton and epiphyton were influenced by availability of organic substrates and electron acceptors. Consequently, bacterioplankton and epiphyton dominated heterotrophic activities at the higher influent strength, and the role of sedimentary bacteria increased at the lower influent strength. The microbial communities of CWs consist of a variable mixture of bacterial species. Heterotrophic production, assimilation and mineralization respond differently to the variations 172  of HRT, influent strength and mass loading rates. Treatment performance is a combined result of heterotrophic production, assimilation and mineralization by the microbial communities in water, epiphyton and sediment.  7.6  References  A P H A et al. 1999. Standard Methods for the Examination of Water and Wastewater, 20th edition. American Public Health Association, A W W A , and WEF. Chidthaisong, A . , B . Rosenstock, and R. Conrad. 1999. Measurement of monosaccharides and conversion of glucose to acetate in anoxic rice field soil. Applied and Environmental Microbiology, 65: 2350-2355. Field, J.A., M.J.H. Leyendeckers, R. S. Alvarez, G. Lettinga, and L . H . A . Habets. 1988. The methanogenic toxicity of bark tannins and the anaerobic biodegradability of water soluble bark matter. Water Science and Technology, 20: 219-240. Flood, J.A., N.J. Ashbolt, and P.C. Pollard. 1999. Complementary independent molecular, radioisotopic and fluorogenic techniques to assess biofilm communities in two wastewater wetlands. Water Science and Technology, 39: 65-70. Frankowski, K . A . 2000. The Treatment of Wood Leachate Using Constructed Wetlands. M.A.Sc. thesis, The University of British Columbia, Vancouver, Canada. Hobbie, J.E., and T.E. Ford. 1993. A perspective on the ecology of aquatic microbes. In: T.E. Ford (ed.), Aquatic Microbiology: An Ecological Approach. Blackwell Scientific Publications: Boston, M A . ppl-14. Hunter, R., A . E . Birkbeck, and G. Coombs. 1993. Innovative marsh treatment systems for control of leachate and fish hatchery wastewaters. In: G . A . Moshiri (ed.), Constructed Wetlands for Water Quality Improvements. Lewis Publishers: Baca Raton, FL. pp477-484. Jorgensen, N.O.G. 1992. Incorporation of H-leucine and H-valine into protein of freshwater 3  3  bacteria: uptake kinetics and intracellular isotope dilution. Applied and Environmental Microbiology, 58: 3638-3646. Karl, D . M . 1993. Total microbial biomass estimation derived from the measurement of particulate adenosine-5'-triphosphate. In: P.F. Kemp, B.F. Sherr, E.B. Sherr, and J.J. Cole  173  (eds.), Handbook of Methods in Aquatic Microbial Ecology. Lewis Publishers: Boca Raton, F L . pp359-372. Kirchman, D.L., E. K'Nees, and R. Hudson. 1985. Leucine incorporation and its potential as a measure of protein synthesis by bacteria in natural aquatic  systems. Applied and  Environmental Microbiology, 49: 599-607. Kozub, D.D., and S.K. Liehr. 1999. Assessing denitrification rate limiting factors in a constructed wetland receiving landfill leachate. Water Science and Technology, 40: 75-82. Masbough, A . 2002. The Effectiveness of Constructed Wetlands for Treatment of Wood Leachate. M.A.Sc. thesis, The University of British Columbia, Vancouver, Canada. Metcalf & Eddy. 2003. Wastewater Engineering - Treatment and Reuse, 4th edition. McGrawHill: New York, N Y . p984, 990. Mueller, R.F. 1996. Bacterial transport and colonization in low nutrient environments. Water Research, 30: 2681-2690. Pace, M . L . 1988. Bacterial mortality and the fate of bacterial production. Hydrobiologia, 159: 41-49. Pollard, P.C., J.A. Flood, and N.J. Ashbolt. 1995. The direct measurement of bacterial growth in biofilms on emergent plants of an artificial wetland. Water Science and Technology, 32: 251-256. Polprasert, C , N.R. Khatiwada, and J. Bhurtel. 1998. A model for organic matter removal in free water surface constructed wetlands. Water Science and Technology, 38: 369-377. Reddy, K.R., and D.A. Graetz. 1988. Carbon and nitrogen dynamics in wetland soils. In: D.D. Hook, W.H. McKee Jr., H.K. Smith et al. (eds.), The Ecology and Management of Wetlands, Vol. I. Timber Press: Portland, OR. pp307-318. Schaule, G., T. Griebe, and H.C. Flemming. 2000. Steps in biofilm sampling and characterization in biofouling cases. In: H.-C. Flemming, U . Szewzyk, and T. Griebe (eds.), Biofilms: Investigative Methods & Applications. Technomic Publishing Company, Inc.: Lancaster, PA. ppl-21. Simon, M . , and F. Azam. 1989. Protein content and protein synthesis rates of planktonic marine bacteria. Marine Ecology Progress Series, 51: 201-213. 174  Tabor, P.S., and R.A. Neihof. 1982. Improved microautoradiographic method to determine individual microorganisms active in substrate uptake in natural waters. Applied and Environmental Microbiology, 44: 945-953. Tao, W., and K.J. Hall. 2004. Dynamics and influencing factors of heterotrophic bacterial utilization of acetate in constructed wetlands treating woodwaste leachate. Water Research, 38: 3442-3448. Tao, W., K . Hall, and S. Duff. 2006. Heterotrophic bacterial activities and treatment performance of surface flow constructed wetlands receiving woodwaste leachate. Water Environment Research, 78: in press. Tao, W., K . J . Hall, A . Masbough, K . Frankowski, S J . B . Duff. 2005. Characterization of leachate from a woodwaste pile. Water Quality Research Journal of Canada, 40: in press. Taylor, B.R., J.S. Goudey, and N . B . Carmichael. 1996. Toxicity of aspen wood leachate to aquatic life: Laboratory studies. Environmental Toxicology and Chemistry, 15: 150-159. Taylor, B.R., and N . B . Carmichael. 2003. Toxicity and chemistry of aspen wood leachate to aquatic life: field study. Environmental Toxicology and Chemistry, 22: 2048-2056. Toerien, D.F., and Toerien, M . C . 1985. Microbial heterotrophy in an effluent treatment system using macrophytes. Agricultural Wastes, 12: 287-312. Toet, S., L.H.F.A. Huibers, R.S.P. Van Logtestijn, J.T.A. Verhoeven. 2003. Denitrification in the periphyton associated with plant shoots and in the sediment of a wetland system supplied with sewage treatment plant effluent. Hydrobiologia, 501: 29-44. Townend, J. 2003. Practical Statistics for Environmental and Biological Sciences. John Wiley & Sons: Chichester, England. Tuominen, L . 1995. Comparison of leucine uptake methods and a thymidine incorporation method for measuring bacterial activity in sediment. Journal of Microbiological Methods, 24: 125-134. USEPA. 2000. Constructed Wetlands Treatment of Municipal Wastewaters. Manual of EPA/625/R-99/010, U.S. Environmental Protection Agency, Cincinnati, OH, September. Ward, A . K . , and M . D . Johnson. 1996. Heterotrophic microorganisms. In: F.R. Hauer and G.A. Lamberti (eds.), Methods in Stream Ecology. Academic Press: New York, N Y . pp254-260. 175  Westermann, P. 1993. Wetland and swamp microbiology. In: T.E. Ford (ed.), Aquatic Microbiology: An ecological approach. Blackwell Scientific Publications: Boston, M A . pp215-239. Wetzel, R . G . and G.E. Likens. 2000. Limnological Analyses, 3rd ed. Springer-Verlag: New York, N Y . White, D.C., R.D. Kirkegaard, R.J. Palmer Jr. et al. 1999. The biofilm ecology of microbial biofouling, biocide resistance and corrosion. In: C.W. Keevil, A . Godfree, D. Holt and C. Dow (eds.), Biofilms in the Aquatic Environment. The Royal Society of Chemistry: Cambridge, U K . pp-120-130. Wittgren, H.B., and T. Maehlum. 1996. Wastewater constructed wetlands in cold climates. Water Science and Technology, 35: 45-53. Wright, R.T. 1988. Methods for evaluating the interaction of substrate and grazing as factors controlling planktonic bacteria. Archives Hydrobiologie Beih. Ergebn. Limnologie, 31: 229242.  176  8  TREATMENT OF WOODWASTE L E A C H A T E IN MESOCOSM WETLANDS:  EFFECTS  OF  INFLUENT  STRENGTH  AND  VEGETATION* 8.1  Introduction  Wood leachate may be generated in woodpiles, log yards, woodwaste piles, and debarking process by precipitation, sprinkling water, and process water. Wood leachate is usually dark, acidic, of very high oxygen demand, nutrient-poor, and very toxic to microorganisms (Field et al., 1988; Hunter et a l , 1993; Tao et al., 2005; Taylor et al. 1996; Taylor and Carmichael 2003; Woodhouse and Duff, 2004; Zenaitis et al., 2002). Conventional biological treatment and ozonation are efficient in removing organic matter from wood leachate (Field et al., 1988; Woodhouse and Duff, 2004; Zenaitis and Duff, 2002; Zenaitis et al., 2002). Because of lower operating cost, simple management requirement, and tolerance to variable influent quantity and quality, vegetated and open CWs have become attractive alternatives for treatment of wood leachate (Frankowski, 2000; Hunter et al., 1993; Masbough, 2002; Tao et a l , 2006). Microbial degradation is the major mechanism in surface flow CWs for the removal of COD, T & L and V F A s from woodwaste leachate (Section 3.5). Benner et al. (1984) confirmed anaerobic lignin degradation in a salt marsh, an acidic freshwater marsh, and a mangrove swamp. Due to the concern of inhibition of a strong influent on microbial activity, woodwaste leachate has been treated after dilution in pilot-scale CWs (Frankowski, 2000; Masbough, 2002; Tao et al., 2006). Inhibitory compounds leaching from plant material (e.g., plant tannins) may have an important influence on bacterial activity in marshes (Moran and Hodson, 1992). However, lower leachate dilution is preferred to higher dilution in regard to treatment capacity. A strong influent may also provide bacteria with more substrates, electron acceptors and nutrients. Further studies are needed to evaluate the effects of influent strength on  * A version of this chapter has been submitted for publication. Tao, W., K.J. Hall, and S.J.B. Duff. Treatment of woodwaste leachate in surface flow mesocosm wetlands: effects of influent strength and vegetation. Environmental Technology, in review. 177  performance of CWs in order to choose an appropriate dilution factor for woodwaste leachate treatment. Dilution of woodwaste leachate before feeding the pilot-scale CWs was also intended to maintain healthy aquatic plants. Vegetation, mainly emergent and submergent plants, provides several benefits for water quality improvement through CWs (Brix, 1997; USEPA, 2000). Nevertheless, vegetation made no significant difference in organic carbon removal through the pilot-scale CWs receiving a high-VFAs influent of woodwaste leachate (Masbough, 2002; Tao et al., 2006). During the later phases of the present study, V F A s of the woodwaste leachate decreased to very low concentrations (Tao et al., 2005) that would favor attached bacteria compared to suspended bacteria (Mueller, 1996; Tao and Hall, 2004; Tao et al., 2006; White et al., 1999). The effect of vegetation on treatment performance of CWs fed with a low-VFAs influent could be different from that of CWs fed with a high-VFAs influent. This study operated four vegetated mesocosm wetlands with different dilutions of the woodwaste leachate during a period of 12 weeks. During another period of 13 weeks, two vegetated mesocosm wetlands were fed with the raw leachate and diluted leachate respectively. One vegetated mesocosm wetland and one open-water wetland were operated in series with the mesocosm wetland fed with raw leachate during the second period. Deployment of mesocosm wetlands during the two periods allowed this study to evaluate the effects of influent strength, leachate characteristics and vegetation, and the effectiveness of wetland cells in series for treatment of recalcitrant woodwaste leachate. The other associated factors influencing kinetics and treatment performance of wetlands for treatment of woodwaste leachate are also discussed.  8.2 8.2.1  Materials and Methods Woodwaste leachate and mesocosm wetlands  Woodwaste leachate was collected monthly from the leachate pool in Mission, B.C. and stored in 220-L high-density polycarbonate tanks at 4°C until use. The woodwaste leachate from April through June 2004 had a p H of 5.4, COD 4810 mg U\ T & L 1270 mg L" , V F A s 319 mg 1  L" , ammonia 2.7 mg N L" , nitrate and nitrite 0.29 mg N L" , and phosphate 1.8 mg P L" on 1  1  1  1  average (Tao et al., 2005). The woodwaste leachate from July to October 2004 had a higher  178  pH (6.4) and more ammonia (18.3 mg N L" ) and orthophosphate (4.0 mg P L" ), while lower 1  1  C O D (2510 mg L" ), T & L (872 mg L" ) and V F A s (15 mg L ' ) on average. 1  1  1  The four mesocosm wetlands previously had been operated from June to November 2003, for treatment of woodwaste leachate, and then had been idle in the winter. From 25 March 2004, when cattail shoots began to emerge, the mesocosm wetlands were fed with raw sewage for 6 weeks. The water depth was gradually increased to 25 cm as the cattails grew. A clear plastic cover was put on a frame over the cattails to shed rainfall from the mesocosms. On May 7, 7 L of sludge (collected in the leachate storing tanks) plus woodwaste leachate was directly added to each mesocosm to provide bacterial seed and bring initial mesocosm water to the targeted influent T & L concentrations—1200 mg L" in mesocosm #1, 800 mg L" in mesocosm #2,400 1  1  mg L" in mesocosm #3, and 200 mg L" in mesocosm #4. Every seven days from May 10, 1  1  four different dilutions of influent were made by diluting pre-set amounts of woodwaste leachate with tap water in four 220-L tanks. The mesocosm wetlands in parallel were continuously fed at the same flow rate during the first operating period from May 10 to August 2 (Figure 1). On August 2, the cattails in mesocosm #3 were removed. Mesocosm #2 and #3 were operated in series with mesocosm #1, with each receiving equal amounts of mesocosm #1 effluent during the second operating period from August 2 to November 1 (Figure 8-1). Water depth was maintained at 25 cm during the two operating periods.  8.2.2  Field measurement, sampling and laboratory analysis  Field measurement was undertaken weekly, except during the startup (May 10 to June 13) and transition (August 2 to September 12) periods. Field measurements were made at 7 cm depth in the center of each mesocosm to approximate the average conditions in the water. Water temperature, D O , pH, and redox potential were measured with the methods and instruments described in Section 2.2.2. Redox potential was also determined for the water-sediment interface. From May 31 to October 25, the influent was prepared on Mondays, measured for pH and redox potential in the mixing tanks on Thursdays, and sampled after measurement. A volumetric method was employed to check inflow rate and measure the outflow rate weekly. The number of emergent plants in each mesocosm was counted, whenever there was an apparent change in vegetation. The average perimeter of submerged stems and the average width of submerged leaves were estimated. Void fraction and nominal HRT were estimated as mentioned in Section 6.2.2 and Section 4.2.2 respectively. The total plant surface area of each 179  mesocosm wetland was estimated with the water depth, average sizes of submerged stems and leaves, and plant number. Influent and effluent samples were collected with 50-mL high-density polyethylene bottles, and stored at 4°C until analysis. C O D , T & L , V F A s , ammonia, nitrate plus nitrite, and orthophosphate were analyzed with the methods and instruments as described in Section 2.2.2.  Mesocosm#l: HRT•=i3 d :  Mesocosm#2: HRT= 13d •Strength 3  Mesocosm#3: HRT=13 d  -Strength 4 Mesocosm#4: HRT= =13 d  Y  y Y Y  Period 1: May 10 - August 2  Strength 2r Mesocosm#l: HRT= 12 d  Y  Mesocosm#2: HRT=29 d  T  Mesocosm#3: HRT=29 d •Strength 4f  Mesocosm#4: HRT==12 d  Y  Period 2: August 2 - November 1  Figure 8-1. Treatment of woodwaste leachate in vegetated (T) and open mesocosm wetlands fed with different strengths of influent during two operating periods in 2004 (see text for mesocosm dimensions).  180  8.2.3 Data processing and statistical analysis The large difference between inflow and outflow rates due to evapotranspiration during the long HRTs required this study to employ a mass-based approach to derivation of treatment performance and kinetics in equation 6-2, equation 6-3, and equation 6-5. The differences in operating conditions and performance between mesocosm wetlands were assessed by singlefactor A N O V A . Data were log transformed when normal distributions were not met. LSD was calculated to identify which two means were significantly different (Townend, 2002). The Bonferroni method was used to correct the significance level for multiple comparisons. Differences were considered significant at a P<0.05.  8.3 Results and Discussion 8.3.1  Operating conditions  Each mesocosm wetland had 28-41 emergent plants in July and 40-51 in October (Appendix 19). The specific plant surface area in July was 12.0, 9.3, 8.8 and 8.2 m plant surface per m water in mesocosms #l-#4 respectively. The specific plant surface area in October was 15.5 and 13.5 m plant surface per m water in mesocosms #1 and #4 respectively. The vegetated 2  3  mesocosm wetlands had similar void fractions, 0.96-0.98. Duckweed (Lemna spp.) appeared in mesocosm #4 from late July and covered the entire surface from late August till the end of operation. Numerous mosquito larvae were observed from late May to August, with increasing abundance in the mesocosm wetlands fed with a stronger influent. Water temperature of the mesocosm wetlands during the first operating period (May 10 August 2, 2004), 18.1-18.4°C, was higher than that during the second period (August 2 November 1, 2004), 11.3-11.7°C. The open mesocosm (#3) during the second period was aerobic, with D O at 1.4±0.4 mg L" on average. The bulk water in all of the vegetated 1  mesocosm wetlands had a low DO concentration, 0.4-0.7 mg L" on average. There was not a 1  significant difference of redox potential among the mesocosm wetlands during the two periods (P = 0.55-0.96). However, surface sediment had an apparently lower redox potential (-6 to 86 mV) than did the water column (211-243 mV). Table 8-1 and Table 8-2 summarize the operating conditions during the two periods respectively (original data in Appendix 20). The influent was introduced with little variation in  181  flow rate. The outflow rates were smaller than the associated inflow rates due to evapotranspiration and evaporation. The mesocosm wetlands had similar nominal HRTs (13 d) during the first period. Mesocosms #1 and #4 had the same H R T (12 d) during the second period. Although evaporation in the open mesocosm #3 was much less than evapotranspiration in the vegetated mesocosm #2, the two mesocosm wetlands had similar HRTs (29 d) during the second period.  Table 8-1. Operating conditions of mesocosm wetlands receiving different strengths of raw and dilute woodwaste leachate during the first operating period  a  #1  #2  #3  #4  105±1  105±1  104±1  104±1  84±5  84±5  83±6  83±6  Hydraulic retention time , d  12.8±0.3  13.0±0.3  12.9±0.4  13.0±0.4  Influent pH  5.9±0.4  6.0±0.4  5.9±0.5  6.0±0.4  Mesocosm water pH  6.7±0.2  6.7±0.2  6.4±0.1  6.3±0.1  Mesocosm wetland Inflow rate, L week"  1  Outflow rate, L week"  1  b  Influent C O D concentration, mg L"  1  3740±786  2060±159  1340±191  586±65  Influent T & L concentration, mg L"  1  1210±76  751±46  437±43  222±24  95±95  47±51  25±38  12±17  7.1±8.4  3.6±5.4  1.8±3.0  0.9±1.6  0.21±0.05  0.14±0.08  0.11±0.04  0.08±0.02  1.83±1.21  0.97±0.82  0.49±0.38  0.22±0.18  3120±170  2140±180  1420±224  479±65  1120±78  824±59  451±41  20U16  ' 15±25  1.6±2.5  0.6±0.5  0.6±0.5  8.8±6.0  6.7±3.7  4.3±1.6  4.8±0.9  0.25±0.05  0.20±0.04  0.13±0.02  0.08±0.02  4.17±0.51  4.03±0.52  2.91±1.01  2.05±0.94  Influent V F A s concentration, mg L"  1  Influent NH3 concentration, mg N L"  1  Influent N O ; concentration, mg N L"  1  Influent ?0\~ concentration, mg P L"  1  Effluent C O D concentration, mg L" Effluent T & L concentration, mg L"  1  1  Effluent V F A s concentration, mg L"  1  Effluent NH3 concentration, mg N L"  1  Effluent N O ; concentration, mg N L"  1  Effluent P O 4 " concentration, mg P L" a  1  Mean ± SD of the influent during May 31-July 26 and the effluent during June 14-August 2,  2004 (n = 8). b  Based on the averages of inflow and outflow rates. 182  Table 8-2. Operating conditions of mesocosm wetlands receiving refractory woodwaste leachate during the second operating period Mesocosm wetland  a  #1  #2  #3  #4  104±1  45±2  45±2  104±1  95±5  38±4  42±1  95±5  Hydraulic retention time , d  12.2±0.3  29.2±1.7  28.7±0.6  12.2±0.3  Influent pH  6.5±0.2  7.0±0.2  7.0±0.2  6.3±0.3  Mesocosm water pH  6.9±0.1  6.9±0.1  7.0±0.1  6.2±0.1  2320±131  2510±206  2510±206  502±36  770±36  798±76  798±76  184±23  2.1±1.8  1.1±0.9  1.1±0.9  1.5±0.8  19.2±7.4  22.3±1.4  22.3±1.4  4.7±1.2  Inflow rate, L week"  1  Outflow rate, L week"  1  b  Influent COD concentration, mg L" Influent T & L concentration, mg L"  1  1  Influent V F A s concentration, mg L"  1  Influent NH3 concentration, mg N L"  1  Influent N O ; concentration, mg N L"  1  0.34±0.21  0.40±0.16  0.40±0.16  0.10±0.04  Influent PO \~ concentration, mg P L"  1  3.17±0.66  4.29±0.56  4.29±0.56  0.87±0.28  2240±184  2370±232  2170±157  415±58  732±42  786±64  720±53  143±11  1.6±0.8  1.8±0.6  1.8±0.5  1.4±0.6  19.6±5.4  16.0±1.3  14.8±1.3  3.4±1.1  Effluent C O D concentration, mg L" Effluent T & L concentration, mg L"  1  1  Effluent V F A s concentration, mg L"  1  Effluent NH3 concentration, mg N L"  1  Effluent N O ; concentration, mg N L"  1  0.26±0.18  0.37±0.21  0.38±0.21  0.11±0.07  Effluent PO ~ concentration, mg P L"  1  3.72±0.52  4.09±0.24  4.32±0.18  0.75±0.32  3  4  a  Mean ± SD of the influent during August 16-October 25 and the effluent during September  13-November 1,2004 (« = 8). b  Based on the averages of inflow and outflow rates.  The influent concentrations of COD and T & L were relatively constant over time for individual mesocosm wetlands (Table 8-1 and Table 8-2), with coefficients of weekly variation at 6-21% for C O D and 5-13% for T & L . A substantial decrease of V F A s and increases of ammonia and phosphate occurred in the woodwaste leachate from early June to early August, 2004 (Figure 2-3), resulting in great temporal variations in influent V F A s and nutrients during the first operating period as well as a clear difference between the influents during the two periods (Table 8-1 and Table 8-2). The influent of mesocosm #1 was either raw leachate or was only 183  slightly diluted leachate. The influent COD and T & L concentrations of mesocosm #1 during the second period were close to those of mesocosm #2 during the first period. Mesocosm #4 during the two periods had similar influent strengths in terms of COD and T & L . The average mass loading rates to the mesocosm wetlands were 42-315 g m" d" COD, 15-102 g m" d" 3  1  3  1  T & L , and 0.1-8.0 g m" d" V F A s . The mass loading and discharge of nutrients are presented 3  1  in Table 8-3.  Table 8-3. Nutrient balance of mesocosm wetlands treating woodwaste leachate in 2004 Mesocosm wetland  #2  #3  #4  746  378  187  96  739  563  357  398  22  15  11  8  21  17  11  8  192  102  51  23  350  338  242  170  #1  First sampling period from May 31 to August 2 (n =7) NH3 loading, mg N wk"  1  NH3 discharge, mg N wk" N O " loading, m g N wk"  1  1  NO ~ discharge, mg N wk" PO 4" loading, mg P wk"  1  1  P O 4 " discharge, mg P wk"  1  Second sampling period from August 16 to November 1 (n = 8) NH3 loading, mg N wk"  1  NH3 discharge, mg N wk" N O " loading, mg N wk"  1  1  N O ^ discharge, mg N wk" P O 4 " loading, mg P wk"  1  PO 4" discharge, mg P wk"  8.3.2  1  1  1997  995  995  489  1862  602  628  323  35  18  18  10  25  14  16  10  330  191  191  90  353  154  183  71  Effect of influent strength on treatment performance  Significantly different strengths of influent were applied to the mesocosm wetlands during the first period (P = 0.00-0.02) and to mesocosms #1 and #4 during the second period (P = 0.00) except for very low concentrations of V F A s to both mesocosms #1 and #4. Because of the low influent and effluent concentrations (Table 8-1 and Table 8-2), treatment performance and 184  kinetics were not estimated for V F A s . Due to the concentration effect of evapotranspiration and evaporation, the less efficient mesocosms #2 and #3 exhibited effluent C O D and T & L concentrations that were higher than the influent concentrations during the first period. Influent pH was similar in all mesocosm wetlands during the first period (P = 0.90) and in mesocosms #1 and #4 during the second period (P = 0.38). Mesocosm water pH increased somewhat during the first period, but there were no differences between mesocosm wetlands. The pH increases were likely due to V F A s removal. COD and T & L reduction rates (Figure 8-2) were significantly different among the mesocosm wetlands during the first period (P = 0.00) and between mesocosms #1 and #4 during the second period LP ~ 0.02 and 0.04). C O D and T & L reduction efficiencies and reduction rate constants (Figure 8-2) were significantly different among the mesocosm wetlands during the first period (P = 0.00-0.01) and between mesocosms #1 and #4 during the second period LP = 0.00-0.05). As Figure 8-2 shows, higher COD and T & L reduction rates were achieved with stronger influents, especially in mesocosm #1. The larger reduction rate constants, higher reduction efficiencies, and higher reduction rates of mesocosm #1 during the first period were likely due to its higher influent V F A s concentration and particularly the high effluent V F A s concentration relative to the other mesocosms (Table 8-1). V F A s , which are readily available bacterial substrates, have been identified as one of the major factors controlling heterotrophic activities in CWs treating this woodwaste leachate (Tao and Hall, 2004; Tao et al., 2006). The larger reduction rate constants and higher reduction efficiencies of mesocosm #4 during the two periods were probably due to the minimal inhibitory effect at lower T & L concentrations. The toxic effects of wood leachate were usually attributed to phenolic compounds, such as tannins. Borga et al. (1996) found that bacterial toxicity of the wastewater from watersprinkled Scotch pine and Norway spruce timber appeared to decline in parallel with dissolved organic carbon. The 50% inhibitory concentration of bark tannins to methanogens averaged approximately 600 mg L" C O D or 350 mg L* tannin solids (Field et al. 1988), which was 1  1  greater than the T & L concentrations in mesocosm #4. Moreover, duckweed has a very high rate of primary production, which could provide bacteria with more substrates in mesocosm #4, enhancing microbial degradation. There was not a clear difference in treatment performance and kinetics between the second period and the warmer first period, probably because the microbial degradation processes in the mesocosm wetlands treating recalcitrant woodwaste leachate were more substrate-limited than temperature-stressed. 185  Chemical oxygen demand "S  oo  <  '#1  e G  .2 "  #2,  J  #4  0  o 3 -o  _> _« 1000  2000  Of!  3000  4000  5000  300  600  900  1200  1500  Influent concentration (mg L"')  Influent concentration (mg L" ) 1  N»  >,  o G u 'o  m u G O o 3 •a  1000  2000  3000  4000  5000  300  Influent concentration (mg L" )  600  900  0.07  Tannin & lignin  0.06 G a  •*-» CO  G o o u  0.05 0.04  1 0.03 G o •«-» o CCJ  U  0.00  #4<  3000  4000  5000  4 #3  0.01  1 300  Influent concentration (mg L" ) 1  L,  '#4  0.02 0.00  2000  1500  Influent concentration (mg L )  0.10  1000  1200  r  1  #2« 600  900  1200  1500  Influent concentration (mg L" )  Figure 8-2. Variations of treatment performance and kinetics of mesocosm wetlands with influent strength. Black dot = mean during the first operating period; open circle = mean during the second period; error bar = ± SD (n = 8). 186  8.3.3  Effect of vegetation on treatment performance  The effluent of mesocosm #1 provided mesocosms #2 and #3 with little V F A s during the second period. The difference between the vegetated mesocosm #2 and open mesocosm #3 reflected the effect of vegetation on treatment of recalcitrant woodwaste leachate. There was no significant difference between the two mesocosm wetlands (Figure 8-3) in COD and T & L reduction rates (P - 0.58 and 0.66), reduction efficiencies (P = 0.46 and 0.59) and reduction rate constants (P = 0.91 and 0.87). There was no significant difference in pH increase between the two mesocosm wetlands (P = 0.59). Insignificant effect of vegetation was also reported for treatment of high-VFAs woodwaste leachate in pilot-scale CWs (Masbough, 2002; Tao et al., 2006). More than 98% of solids and COD in the woodwaste leachate are soluble (Tao et al., 2005). Therefore, performance could not be improved by means of sedimentation and flocculation under the quiescent conditions in vegetated CWs. However, it has been reported that microorganisms benefit from attaching on emergent plants, especially under substrate-limiting conditions (Kjelleberg et al., 1982; Mueller, 1996; USEPA, 2000; White et al., 1999) similar to the mesocosm wetlands during the second period. Our studies (Tao and Hall, 2004; Tao et al., 2006) indicate that bacteria attached on emergent plants usually play a minor role in the overall heterotrophic activities in comparison to the bacteria in water and sediment of the surface flow CWs treating woodwaste leachate. Emergent plants minimize the concentration of DO by restraining atmospheric reaeration at the water-air interface. Bezbaruah and Zhang (2005) found very little oxygen released from bulrush (Scirpus validus) roots, 2.3 mg O2 m" wetland surface d* , comparing to the oxygen demand of treatment wetlands. The open 1  mesocosm wetland had a significantly higher concentration of DO due to atmospheric aeration. The insignificant difference in treatment performance between the vegetated and open mesocosm wetlands suggested that the tradeoff between bacterial attachment and restrained surface aeration due to vegetation was not significant.  8.3.4  Treatability of woodwaste leachate in wetlands  The reduction efficiency of the mesocosm wetlands varied with influent strength (Figure 8-2). At an influent C O D concentration similar to that of mesocosm #3 during the first period, Hunter et al. (1993) found a C O D reduction efficiency of 29% in small-scale CWs treating wood leachate at a HRT of 10 d. Tao et al. (2006) reported a similar T & L reduction efficiency 187  (12-18%) and higher C O D reduction efficiency (23-29%) through pilot-scale CWs with similar HRTs at an influent C O D concentration similar to and V F A s concentration much higher than mesocosm #3 during the first period.  50 'E  60  C O  40 30 20  u  3  •a  (4  10  50  • Chemical oxygen demand • Tannin & lignin  • Chemical oxygen demand • Tannin & lignin  40  iii  c a  30  c o .2  20  3 —I  a  10  w  12.2 d  29.2 d  28.7 d  Wetland#l  Wetland#2  Wetland#3  a  2  12.2 d  ^ 0.020 CO  1 0.015 u <u  2 0.010 c _o | 0.005 —i u  * 0.000  28.7 d  Wetland* 1 Wetland#2  Wetland#3 Hydraulic retention time (d)  Hydraulic retention time (d)  0.025  29.2 d  • Chemical oxygen demand • Tannin & lignin  Mi 12.2 d  29.2 d  Wetland#l  Wetland#2  28.7 d  Wetland#3 Hydraulic retention time (d)  Figure 8-3. Comparison of treatment performance and kinetics of mesocosm wetlands in the vegetated mesocosm #1 - vegetated mesocosm #2 series and the vegetated mesocosm #1 open mesocosm #3 series during the second operating period. Error bar = SD (n = 8).  A small loss or net gain of inorganic nutrient mass through the mesocosm wetlands (Table 83) implied insignificant nutrient limitation for microbial degradation of the woodwaste leachate. In the CWs treating secondary sewage effluent, bacterial growth rates were not  188  correlated to nutrient loadings, while the population structure changed markedly to adapt to the nutrient conditions (Flood et al., 1999). The ThOD of recalcitrant T & L accounted for 39-47% of COD in the influents and 40-53% in the effluents of the mesocosm wetlands. The ThOD of V F A s only comprised 0-4% of COD in the influents and <1% in the effluents. The recalcitrance of the woodwaste leachate requires a long exposure time for bacteria to break down the contaminants such as T & L . Mesocosm #1 with a HRT of 12 d was less efficient than the following mesocosms #2 and #3 with a H R T of 29 d during the second period (Figure 8-3). The overall C O D reduction efficiency was 46% through mesocosm #1 - mesocosm #2 series and 45% through mesocosm #1 - mesocosm #3 series. The overall T & L reduction efficiency was 39% through mesocosm #1 - mesocosm #2 series and 36% through mesocosm #1 - mesocosm #3 series. C W cells in series would provide recalcitrant woodwaste leachate with a long detention time to treat biologically degradable constituents, while minimizing odor problems related with anaerobic reactions.  8.4  Conclusions  The woodwaste leachate could be treated without dilution in surface flow CWs for removal of organic carbon. Emergent plants made no difference in performance for treatment of woodwaste leachate. A better overall treatment performance could be achieved through wetland trains at long HRTs. A greater availability of bacterial substrates with a strong influent may offset the inhibitory effect on microorganisms for treatment of woodwaste leachate. Improved surface aeration in open wetlands may compensate for the lack of plant surfaces for microbial attachment.  8.5 References A P H A , A W W A , and WEF. 1999. Standard Methods for the Examination of Water and Wastewater, 20th edition. American Public Health Association, American Water Works Association, and Water Environment Federation. Benner, R., A . E . Maccubbin, and R.E. Hodson. 1984. Anaerobic biodegradation of the lignin and polysaccharide components of lignocellulose and synthetic lignin by sediment microflora. Applied and Environmental Microbiology, 47: 998-1004. 189  Bezbaruah, A . N . , and T. C. Zhang. 2005. Quantification of oxygen release by bulrush (Scirpus validus) roots in a constructed treatment wetland. Biotechnology and Bioengineering, 89: SOSSIS. Borga, P., T. Elowson, and K . Liukko. 1996. Environmental loads from water-sprinkled softwood timber: 2. Influence of tree species and water characteristics on wastewater discharges. Environmental Toxicology Chemistry, 15: 1445-1454. Brix, H . 1997. Do macrophytes play a role in constructed treatment wetlands? Water Science and Technology, 35: 11-17. Field, J.A., M.J.H. Leyendeckers, R. S. Alvarez, G. Lettinga, and L.H.A. Habets. 1988. The methanogenic toxicity of bark tannins and the anaerobic biodegradability of water soluble bark matter. Water Science and Technology, 20: 219-240. Flood, J.A., N J . Ashbolt, and P.C. Pollard. 1999. Complementary independent molecular, radioisotopic and fluorogenic techniques to assess biofilm communities in two wastewater wetlands. Water Science and Technology, 39: 65-70. Frankowski, K . A . 2000. The Treatment of Wood Leachate Using Constructed Wetlands. M.A.Sc. thesis, University of British Columbia, Vancouver, Canada. Hunter, R., A . E . Birkbeck, and G . Coombs. 1993. Innovative marsh treatment systems for control of leachate and fish hatchery wastewaters. In: G . A . Moshiri (ed.), Constructed Wetlands for Water Quality Improvements. Baca Raton, F L : Lewis Publishers. pp477-484. Kjelleberg, S., B . A . Humphrey, and K . C . Marshall. 1982. Effect of interfaces on small, starved marine bacteria. Applied and Environmental Microbiology, 43: 1166-1172. Masbough, A . 2002. The Effectiveness of Constructed Wetlands for Treatment of Wood Leachate. M.A.Sc. thesis, University of British Columbia, Vancouver, Canada. Moran, M . A . , and R.E. Hodson. 1992. Contributions of three subsystems of a freshwater marsh to total bacterial secondary productivity. Microbial Ecology, 24: 161-170. Mueller, R.F. 1996. Bacterial transport and colonization in low nutrient environments. Water Research, 30: 2681-2690.  190  Tao, W., and K.J. Hall. 2004. Dynamics and influencing factors of heterotrophic bacterial utilization of acetate in constructed wetlands treating woodwaste leachate. Water Research, 38: 3442-3448. Tao, W., K . Hall, and S. Duff. 2006. Heterotrophic bacterial activities and treatment performance of surface flow constructed wetlands receiving woodwaste leachate. Water Environment Research, 78: in press. Tao, W., K . J . Hall, A . Masbough, K . Frankowski, S.J.B. Duff. 2005. Characterization of leachate from a woodwaste pile. Water Quality Research Journal of Canada, 40: in press. Taylor, B.R., J.S. Goudey, and N . B . Carmichael. 1996. Toxicity of aspen wood leachate to aquatic life: Laboratory studies. Environmental Toxicology Chemistry, 15: 150-159. Taylor, B.R., and N . B . Carmichael. 2003. Toxicity and chemistry of aspen wood leachate to aquatic life: field study. Environmental Toxicology Chemistry, 22: 2048-2056. Townend, J. 2003. Practical Statistics for Environmental and Biological Sciences. Chichester, England: John Wiley & Sons. USEPA. 2000. Constructed Wetlands Treatment of Municipal Wastewaters. Manual of EPA/625/R-99/010, U.S. Environmental Protection Agency, Cincinnati, OH. White, D . C , R.D. Kirkegaard, R.J. Palmer Jr. et al. 1999. The biofilm ecology of microbial biofouling, biocide resistance and corrosion. In: C W . Keevil, A . Godfree, D. Holt and C. Dow (eds.), Biofilms in the Aquatic Environment. Cambridge, U K : The Royal Society of Chemistry. ppl20-130. Woodhouse, C , and S.J.B. Duff. 2004. Treatment of log yard runoff in an aerobic trickling filter. Water Quality Research Journal of Canada, 39: 232-238. Zenaitis, M . G . , and S.J.B. Duff. 2002. Ozone for removal of acute toxicity from logyard runoff. Ozone: Science and Engineering, 24: 83-90. Zenaitis, M . G . , H . Sandhu, and S.J.B. Duff. 2002. Combined biological and ozone treatment of log yard runoff. Water Research, 36: 2053-2061.  191  9  MICROBIAL COMMUNITY STRUCTURE AND HETEROTROPHIC ACTIVITIES  OF  MESOCOSM  WETLANDS:  EFFECT  OF  INFLUENT STRENGTH* 9.1  Introduction  A dark, acidic wood leachate with a high oxygen demand is generated when rainfall and runoff percolate through woodwaste disposal sites (Hunter et al., 1993; Peters et al., 1976; Tao et al., 2005). Wood leachate can be toxic to microorganisms (Field et al., 1988; Taylor et al. 1996; Taylor and Carmichael 2003; Woodhouse and Duff 2004; Zenaitis et al. 2002). Inhibitory compounds, such as plant tannins, may have an important influence on bacterial activity in marshes (Moran and Hodson, 1992). A surface flow C W is an attractive alternative for treatment of woodwaste leachate (Hunter et al., 1993; Masbough, 2002; Tao et al., 2006). Organic carbon removal in CWs is mainly achieved by heterotrophic microorganisms (USEPA, 2000). Due to concerns with the inhibitory effect of a strong influent on microbial activity, woodwaste leachate has commonly been diluted before treatment in CWs. However, lower leachate dilution is preferred to higher dilution in regard to treatment capacity. A strong influent may also provide bacteria with more substrates, electron acceptors and nutrients. Heterotrophic bacterial activities in surface flow CWs are influenced by the levels of organic substrates, electron acceptors and inorganic nutrients (Kozub and Liehr, 1999; Pollard et al., 1995; Tao and Hall, 2004; Tao et a l , 2006; Toet et al., 2003), which are associated with the strength of woodwaste leachate fed to CWs. Heterotrophic bacterial activities in the mesocosm wetlands treating woodwaste leachate have been found to be significantly different at two strengths of influent T & L , 174-179 mg L" and 591-665 mg L" (Chapter 7). Bacteria may 1  1  inhabit three components (water, epiphyton and sediment) of a surface flow CW. Planktonic, epiphytic and sedimentary bacteria respond differently to environmental conditions of CWs (Tao and Hall, 2004; Tao et al., 2006; Toerien and Toerien, 1985; Toet et al., 2003). The  * A version of this chapter has been drafted for publication. Tao, W., K.J. Hall, and W. Ramey. Microbial community structure and heterotrophic activities of surface flow mesocosm wetlands receiving woodwaste leachate: effect of influent strength. 192  overall response of an entire C W system may be a complicated function of heterotrophic activity in the three components and influent strength. To improve biological treatment of woodwaste leachate in CWs, the effect of influent strength on heterotrophic activities needs to be examined in a wider range. Microorganisms within an aquatic system may include bacteria, algae, and protozoa. Algae convert inorganic carbon into organic carbon. Bacteria assimilate and mineralize organic matter. Protozoa, including heterotrophic flagellates and ciliates, feed on bacteria and algae. Bacteria, algae and protozoa form a microbial loop for carbon flow in aquatic systems (Kisand and Noges, 2004; Steenbergen et al., 1993; Wetzel, 2001). Many acidic environments like woodwaste leachate are well populated by acidophiles, including bacteria, algae, and protozoa (Kushner, 1993). Nevertheless, the microbial loop has not been investigated in surface flow CWs. Because of the dark color of woodwaste leachate, chlorophyll-a in both the pilot-scale CWs (Section 4.3.1) and the mesocosm wetlands (Section 6.3.2) has been found to be present at ultra-oligotrophic levels that are negligible relative to bacterial biomass. The insignificant correlations of cellular A T P concentration with heterotrophic activities of the mesocosm wetlands (Sections 7.3.2 and 7.3.3) implied the possibility of selective protozoan grazing in controlling the biomass of respiring bacteria (del Giorgio et al., 1996; Pernthaler et al., 1996). Investigation into microbial community structure in the CWs receiving different dilutions of woodwaste leachate would facilitate the understanding of the dynamics of heterotrophic activities. A previous study (Flood et al., 1999) found that bacterial abundance, productivity, respiratory activity, and phylogenic composition of an epiphytic community responded differently to the environmental conditions in the CWs receiving secondary municipal effluent. The present study investigated the effects of influent strength on heterotrophic activities, using H-leucine 3  as a radioactive tracer for bacterial production and C-acetate as a radioactive tracer for 14  bacterial substrate assimilation and mineralization in water, epiphyton and sediment of surface flow mesocosm wetlands. Meanwhile, cellular A T P concentration and the abundances of respiring bacteria and protozoa were determined in order to understand some of the interactions between heterotrophic activities and influent strength.  193  9.2 Materials and Methods 9.2.1 Sample preparation for microbiological examinations The setup and operating conditions of the mesocosm wetlands have been described in Sections 8.2.1  and 8.3.1. Samples were taken weekly from the four mesocosms after the startup period  during the first operating period and from mesocosms #1 and #4 after the transition period during the second operating period. Cellular ATP, leucine incorporation, acetate assimilation, and acetate mineralization were determined for water, epiphyton, and sediment. Protozoa and respiring bacteria were counted during the second period. Samples were prepared using the procedures described in Section 7.2.1.  9.2.2 Enumeration of protozoa This study employed the live counting technique (Gasol, 1993; Massana and Gude, 1991) to enumerate flagellates and ciliates in water and sediment (Appendix 21). Water (20 uL) was pipetted to an etched circle (diameter 1 cm) on a Gold Seal fluorescent antibody microslide (0.93 mm thick). The slides were loaded on a Motic B 3 Professional series biological microscope under the 10x phase contrast objective to take two 5-second videos (100* magnification). Eight videos for each sample were captured at randomly selected points in four circles. Nanoflagellates (2-20 um), larger flagellates (20-200 jam) and ciliates (20-200 um) in water were counted separately in the videos by their characteristic motility and cell size. Fresh surface sediment (0.5 mL) was diluted with 5 mL filtered (0.45 um pore size) wetland water, and allowed to sit for a while for settling of coarse particles. Sediment suspension (20 uL) was pipetted to an etched circle. Flagellates and ciliates in four circles were directly counted from the eyepieces (10x) under the 1 0 phase contrast objective. The x  abundances of protozoa in water and sediment samples were estimated from the average count, sample volume (20 uL) pipetted, observation area, circle area, and sediment dilution factor.  9.2.3 Enumeration of respiring bacteria This study employed a modification of the procedure developed by Rodriguez et al. (1992) to count respiring bacteria (Appendix 22). Subsamples of wetland water (5 mL) and sediment (1 mL suspension plus 4 mL sterile wetland water) were prepared in 7-mL test tubes. Epiphyton  194  (1 disc plus 5 mL sterile wetland water) were prepared in 20-mL scintillation vials. Half a milliliter of 13.1 g L" CTC (Poly-Sciences, Warrington, PA; final 3.5 mM) was added to each 1  subsample and incubated in the dark for 4 h at the in situ water temperature with gentle agitation in an Innova 4230 incubator shaker (New Brunswick Scientific, Edison, NJ). The incubation time and final CTC concentration were chosen after trials of 2-6 h of incubation with final C T C concentrations of 2.0-3.5 m M with reference to previous studies (Rodriguez et al., 1992; Schaule et al., 1993). With each live subsample, one fixed subsample was used to control abiotic CTC-reduction. Formaldehyde (0.3 mL of 37% w/w solution) was used to fix the controls and terminate live incubation. A combination of chemical and physical treatments (Velji and Albright, 1993) was used to disperse bacteria from their attached sites and aggregated forms in sediment. Each sediment subsample was treated with 50 uL of 16 g L" tetrasodium pyrophosphate (Fisher Scientific 1  Co., Fair Lawn, New Jersey) for 20 min, and sonicated in the ultrasonic cleaner for 1 min. One mL of the incubation solution was pipetted into 4 mL sterile dilution water. One mL of the diluted incubation solution was filtered by vacuum filtration on a wet black Millipore polycarbonate membrane filter (25 mm diameter; 0.2 um pore size), which was mounted on a damp nitrocellulose backing filter. Epiphyton on sandpaper discs was directly retrieved after incubation. Air-dried black filters and discs were mounted over a drop of low-fluorescence immersion oil (Cargille type B) on a Fisherbrand precleaned microscope slide (38x75x1 mm). A Fisherfinest premium cover glass was put on the filter or disc with a drop of immersion oil. When nonfluorescent, oxidized C T C is reduced by electron transport activity of active bacteria, insoluble fluorescent formazans are deposited intracellularly. The formazans of CTCrespiring bacteria were counted with an Axioskop Routine Microscope (Carl Zeiss, Oberkochen, Germany) equipped with a 50W H B O A C arc lamp and #15 filter set (BP546/12 FT580 LP590) under the Plan-Neofluar 100x/1.30 oil immersion objective. 6-15 fields per subsample were counted, depending on abundance. Respiring bacteria abundance was estimated with the mean net field count, volume of incubation solution filtered, quantity of samples incubated, dilution factors, and the ratio of grid to effective black filter (or disc) areas. 9.2.4  Determination of cellular ATP  Cellular ATP concentrations of water, mature epiphyton, and sediment were determined in the same methods and instruments as described in Section 6.2.3. 195  9.2.5  Determination of heterotrophic production  The procedures and calculations were the same as described in Section 7.2.3 (details in Appendix 14), except the different incubation times (20 min, 40 min and 60 min respectively for water, epiphyton and sediment). Time course experiments (Appendix 15) were conducted initially to set these incubation times.  9.2.6  Determination of heterotrophic assimilation and mineralization  The procedures and calculations were the same as described in Section 7.2.4 (details in Appendix 16), except the different incubation time (15 min). Time course experiments (Appendix 17) were conducted initially to set this incubation time. In order to compare the relative contributions of water, epiphyton and sediment to the total acetate assimilation or mineralization of an entire CW, the measured areal rate of epiphyton and the volumetric rate of sediment were converted to water equivalent rate as described in Section 7.2.4. Similarly, the plant surface areal and sediment volumetric measurements could be converted to water equivalent values for leucine incorporation rate, flagellate abundance, and respiring bacteria abundance mentioned above.  9.2.7  Statistical analysis  Single-factor A N O V A was performed to assess the difference between means. Spearman's rank correlation analysis was used to test the significance of a monotonic relationship between two variables, giving P value and coefficient r$. Regression analysis was employed to test the linear relationship between a dependent variable and one or more independent variables, giving P value and coefficient of determination R . Correlations and differences were 2  considered significant at a P<0.05.  9.3 9.3.1  Results Operating conditions of mesocosm wetlands  The vegetation, water temperature, D O and redox potential in the mesocosm wetlands have been described in Section 8.3.1. The top 2-cm sediment of the mesocosm wetlands (Appendix 19) had similar densities of 0.65-0.73 g (dry weight) mL" (P = 0.47) and organic contents of 1  5.3-6.7% (P = 0.14) during the first operating period. Organic content was significantly higher (P = 0.00) in mesocosm #1 (8.0+0.7%) than in mesocosm #4 (5.8±0.4%), while sediment 196  density was significantly lower (P = 0.00) in mesocosm #1 (0.57±0.06 g mL" ) than in 1  mesocosm #4 (0.71±0.06 g mL" ) during the second period. Due to evaporation and 1  evapotranspiration, C O D and T &