Binary population biofilms by Maarten Alexander Siebel A thesis submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Civil Engineering Montana State University © Copyright by Maarten Alexander Siebel (1987) Abstract: Biofilm research has been restricted to studies of undefined mixed microbial populations and to investigations of (defined) mono microbial populations. In the first case, the organisms are considered as a homogeneous mass, biomass, ignoring the properties of individual species, the sum of which determines the observed phenomena. The second case concentrates on the properties and processes of one microbial species ignoring the influence of the broader environment, eg. other microbial species, on this species. Goal of this research was to study a defined mixed microbial biofilm and to determine possible interactive effects between the species comprising the biofilm. Biofilm experiments were conducted with mono populations of Klebsiella pneumoniae and Pseudomonas aeruginosa and with binary populations of K. pneumoniae and P aeruginosa . Process rates and stoichiometric coefficients, determined for the mono population biofilms were compared with those found in the binary population biofilm. Results indicate that the specific cellular product formation rate of K. pneumoniae or P. aeruginosa in the binary biofilm is not affected by the presence of the other species. Similarly, the glucose-oxygen stoichiometric ratio of K. pneumoniae or P. aeruginosa in the binary biofilm is not affected by the presence of the other species. Many processes at the cellular level are performed faster by K. pneumoniae than by P. aeruginosa: eg. biofilm specific product formation rate and the maximum specific growth rate of K. pneumoniae are 5 times that rate of P. aeruginosa. Nevertheless, K. pneumoniae cell mass does not dominate the biofilm, possibly because of its non-motility and its product formation properties.  BINARY.POPULATION BiOFILMS by Maarten Alexander Siebel A thesis submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Civil Engineering MONTANA STATE UNIVERSITY Bozeman, Montana September 1987 :P372> S M 41 APPROVAL of a thesis submitted by Maarten Alexander Siebel ii This thesis has been read by each member of the thesis committee and been found to be satisfactory regarding content, English usage, format, citations, bibliographic style and consistency and is ready for submission to the College of Graduate Studies. Dat e Chairperson, Graduate Committee Approved for Major Department / f j y / Date — i — — - - - - Head, Major Department Approved for College of Graduate Studies f - z y - f / Date Graduate Dean iii STATEMENT OF PERMISSION TO USE In presenting this thesis in partial fulfillment of the requirements for a doctoral degree at Montana State University, I agree that the Library shall make it available to borrowers under rules of the Library. I further agree that copying of this thesis is allowable only for scholarly purposes, consistent with the "fair use" as prescribed in the U .S . Copyright Law. Requests for extensive copying or production of this thesis should be referred to University Microfilms International, 300 North Zeeb Road, Ann Arbor, Michigan 48106, to whom I have granted "the exclusive right to reproduce and distribute copies of the dissertation in and from microfilm and the right to reproduce and. distribute by abstract in any format". Signature Date iv ACKNOWLEDGEMENT Completion of this study would have been impossible without the help, encouragement and love of numerous people. Realizing that completeness is impossible I like to mention the following: My friends at I .P .A . - Diane, Wendy, Pam, Shari, Kelly, Anne, Rob, Andy, Rich, Mukesh , Nick, Rune, Zbigniew, BjfSrn, Ewout, Dick, Al, Bob, Laura, Frank, Whon Chee, Joe, Kim, Jan, Brent, Dave, Dave, Chris : a group of people that provided the most stimulating environment one could wish to work in. My thesis committee - Keith Cooksey, Gordon McFeters, Hayden Ferguson, Karel Luyben, Al Cunningham. Those people who provided technical and analytical support - Gordon Williams, John Rompel, Stuart Aasgaard, Andy Blixt. The staff in Civil Engineering, Computing Services, Library, Montana Hall for patiently answering my questions. The staff of International Education for helping me with more than just the legal part of my stay. Niek Luijtjes for financial and 'computational' support and Vicky Thompson for proofreading the manuscript. Last but certainly not least those who supported and encouraged me on a day to day basis: The Characklii for giving me this 'Bozeman experience'. Your encouragement and stimulating excitement will guide me for many years to come. Gefrie and Edwin for your love and patience. Your invaluable moral and practical support was a necessary ingredient to complete this job. Financial support is acknowledged from the Office of Naval Research, National Science Foundation, Montana State University Experiment Station and the I .P .A . Associates Program. V vi TABLE OF CONTENTS Page LIST OF T A B L E S ......................................... ix LIST OF FIGURES.......... xi ABSTRACT.................................................. xvi INTRODUCTION........................................... I Goal and Objective .............................. 2 LITERATURE REVIEW ........... Biofilm Processes . . . . Biofilm Properties . . . . Microbial Species . . . . Klebsiella pneumoniae Pseudomonas aeruginosa Extracellular Products . . Microbial Interactions . MATHEMATICAL DESCRIPTION .............................. 12 Mass Balance Equations for the Biofilm Reactor . . 13 Mass Balance Equations for the Chemostat........ 18 Mass Balance Equations for the Batch Reactor . . . 19 EXPERIMENTAL APPROACH ................................ 20 Rototorque .............................. . . . . . 20 Preparation of Rototorque ................... 22 Preparation of Chemostat . .................... 23 Start-up and Operating Conditions ........... 23 Dilution Water ................................ 25 Sampling....................................... 25 C h e m o s t a t ......................................... 29 Preparation and Start-up ...................... 29 Operation and Sampling . ....................... 31 Batch R e a c t o r ..................................... 32 Principles of Operation ...................... 32 Monitoring, Operation and Sampling ........... 34 Nutrient M e d i u m ..................... 35 Nutrient Solution ............................ 35 Preparation of Plates ............. . . . . . 36 Microbial Species ................................ 36 Klebsiella pneumoniae ........................ 36 Pseudomonas aeruginosa . . . . . ............ 36 Analytical Methods ................. 36 Statistical Methods .............................. 40 k o o o m c n m f ' w w vii R E S U L T S ............. 42 Batch E x p e r i m e n t s ............................. 43 Yield Coefficients......................... 43 Chemostat Experiments ............................ 45 Growth Kinetic Coefficients ................. 46 Product Formation Kinetic Coefficients . . . . 48 Cell and Product Yield Coefficients ........ 50 Cell Dimensions as Function of Dilution Rate . 55 Biofilm Experiments . . . 57 Progression of Biofilm Carbon Components . . . 58 Specific Rates ................................ 58 Specific Cellular Detachment Rate .... 63 Biofilm Specific Cellular Growth Rate . . 63 Specific Product Detachment Rate ......... 63 Biofilm Specific Product Formation Rate . 64 Product Formation Coefficients ............... 64 Biofilm Specific Substrate Uptake Rate . . . . 69 Biqfilm Cell and Product Yield Coefficients . .70 Biofilm Glucose-Oxygen Stoichiometric Ratio . 74 Biofilm Thickness ............................ 78 Species Distribution .......................... 83 Photographic Illustrations ................... 83 DISCUSSION............................ 94 Product Formation ................................. 94 Product Formation in Suspension ............. 96 Product Formation in the Biofilm ............. 97 Glucose-Oxygen Stoichiometric Ratio ............. 99 Glucose-Oxygen Stoichiometric Ratio for K , pneumoniae.......... 99 YgQ for Binary Population Biofilms ........... 102 Influence of Diffusional Resistance on YgQ . . 103 Factors affecting Initial Adsorption ............. 106 Specific Cellular Glucose Uptake Ratio ........... 107 Growth Kinetics . .................................... Ill Suspended Growth .............................. Ill Mono Population Biofilms ...................... Ill Binary Population Biofilms ................... 115 Accumulation of a Binary Population Biofilm: a Conceptual Description................. . . . . . 117 CONCLUSIONS . ............................................. 123 REFERENCES............... 125 NOMENCLATURE ........................................... 131 viii APPENDICES.......... 135 APPENDIX A: Oxygen Diffusion Study ............... 136 APPENDIX B : Characteristics of the Rototorque . . 139 APPENDIX C : Raw Data Batch Experiments............. 140 APPENDIX D : Raw Data Chemostat Experiments . . . . 143 APPENDIX E : Comparison of Nutrient Compositions . 144 APPENDIX F : Sample Preparation Procedure ......... 145 Transmission Electron Microscopy (TEM) . . . . 145 Scanning Electron Microscopy (SEM) ........... 146 APPENDIX G : Raw Data and Parameters Logistic Equation..........................................147 Mono population biofilm - Ki. pneumoniae . . . 148 APPENDIX H : Raw Data and Parameters Logistic Equation..........................................156 Mono Population Biofilm - Pi, aeruginosa . . . 156 APPENDIX I : Raw Data and Parameters Logistic Equation....................................... 165 Binary Population Biofilm ................... 165 APPENDIX J : Diffusion through the Laminar Sublayer 174 APPENDIX K : Diffusion in the B i o film................176 ix LIST OF TABLES Page 1. Relevant characteristics of pneumoniae and P . aeruginosa. . 7 2. Composition of nutrient solution ................. 35 3. Summary of batch reactor experimental results . . 43 4. Means and standard deviations of yield coefficients in batch reactor experiments .................... 43 5. Summary of chemostat experimental results for K . pneumoniae.......... ...................... 45 6. Specific growth- and non-growth associated product formation coefficients for pneumoniae in the c h e m o s t a t ......................................... 50 7. Yield coefficients for pneumoniae chemostat experiments . ...................................... 55 8. Specific growth- and non-growth associated product formation coefficients in the biofilm for K . pneumoniae. P . aeruginosa and binary population . 69 9. Values for the growth- (A) and the non-growth (B) associated term in Eg. 30 of specific substrate uptake rate in the biofilm for pneumoniae. P . aeruginosa and binary population . . . ........... 70 10. Yield coefficients Y^g and Ypg in the biofilm for K . pneumoniae. P . aeruginosa and binary population 74 11. Glucose-oxygen stoichiometric ratio in the biofilm for Kj_ pneumoniae. P , aeruginosa and binary population...................................... 78 12. Summary of specific growth- Ckpcj) and non-growth- (kpn) associated product formation coefficients. . 98 13. Relation between steady state biofilm thickness and diffusion effectiveness factor E for pneumoniae and aeruginosa.......................... 105 14. Comparison of specific cellular glucose uptake ratios (Ag) for pneumoniae and Pi aeruginosa in suspension and in biofilm.......................... 109 15. Comparison of growth and product yield coefficients for pneumoniae and aeruginosa............... 16. Experimental data from oxygen diffusion study. . . 17. Parameter values for determination of diffusion effectiveness factor . ............................ 18. Relation between steady state biofilm thickness and diffusion effectiveness factor € for pneumoniae and P^ aeruginosa................................... X H O 138 179 179 xi 1. Schematic of a Rototorque reactor.................. 21 2. Schematic of a Rototorque experimental setup. . . 24 3. Overview of dilution water treatment.............. 26 4. Schematic of a chemostat............................ 30 5. Schematic of a batch reactor....................... 33 6. Overview of the sampling scheme and analytical procedures.......................................... 37 7 Typical progression of oxygen consumption by K. pneumoniae in a batch reactor ................... 44 8. Change in chemostat carbon concentrations with varying dilution rate (squares = influent glucose carbon, crosses = product carbon, triangles = cell carbon and stars = glucose carbon in effluent) . . 47 9. Change in chemostat effluent glucose carbon concentration with varying dilution rate ........ 49 10. Specific product formation rate vs. dilution rate for chemostat growth of Kju pneumoniae. Models for soluble (triangles, R2=O.95) and total product (crosses, R2=O.86) are represented by continuous lines....................... 51 11. Change in chemostat carbon concentrations with dilution rate (squares = influent glucose carbon, crosses = product carbon, triangles = cell carbon and stars = glucose carbon in effluent). Continuous lines represent models describing the change in carbon concentration with dilution rate .......... 53 12. Variation in yield coefficients with dilution rate (triangles = yield of product carbon, crosses = yield of cell carbon) ............................. 54 13. Variation in cell length and breadth with dilution rate. Length of bars is two times the standard e r r o r .............................. 56 LIST OF FIGURES Page xi i 14. Typical progression of organic carbon in a K . pneumoniae biofilm. Length of bars is two times the standard error ..................................... 59 15. Typical progression of organic carbon in a P ■ aeruginosa biofilm. Length of bars is two times the standard error . .................................... 60 16. Typical progression of organic carbon in a binary population biofilm. Length of bars is two times the standard e r r o r .......... 61 17. Comparison of organic carbon components in steady state biofilms of pneumoniae. P . aeruginosa and binary populations CL = grown on low glucose carbon concentration, 10 g C.m- ,^ H = grown on high glucose carbon concentration, 20 g C.m ; . . . . 62 18. Comparison of specific biofilm cellular growth rate (hatched) and biofilm product formation rate (cross-hatched) in steady state biofilms of K . pneumoniae. P . aeruginosa and binary populations (L = grown on low glucose carbon concentration, 10 g C.m~ , H = grown on high glucose carbon concentration, 20 g C.m d) . ...................... 65 19. Biofilm specific product formation rate vs. biofilm specific cellular growth rate for K. pneumoniae. Continuous line is regression model (R2=Q.74). . . 66 20. Biofilm specific product formation rate vs. biofilm specific cellular growth rate for P. aeruginosa. Continuous line is regression model (R2=O.31)... 67 21. Biofilm specific product formation rate vs. biofilm specific cellular growth rate for binary population. Continuous line is regression model (R2=O . 90)........................................... 68 22. Biofilm specific substrate uptake rate vs. biofilm specific cellular growth rate for pneumoniae. Continuous line is regression model (R2=O.72)... 71 23. Biofilm specific substrate uptake rate vs. biofilm specific cellular growth rate for P. aeruginosa. Continuous line is regression model (R2=O. 57)... 72 24. Biofilm specific substrate uptake rate vs. biofilm specific cellular growth rate for binary population. Continuous line is regression model (R2=O. 98)..................................... , - 73 Xiii 25. Biofilm specific glucose uptake rate vs. biofilm specific oxygen uptake rate for pneumoniae. Continuous line is regression model (R2=O.93). . . 75 26. Biofilm specific glucose uptake rate vs. biofilm specific oxygen uptake rate for F> aeruginosa. Continuous line is regression model (R2=O.34). . . 76 27. Biofilm specific glucose uptake rate vs. biofilm specific oxygen uptake rate for binary population. Continuous line is regression model (R2=O.94). . . 77 28. Progression of biofilm thickness for Ki. pneumoniae. Continuous line represents time smoothed means and means plus or minus the standard deviation of the means................................................ 79 29. Progression of biofilm thickness for Pi. aeruginosa. Continuous line represents time smoothed means and means plus or minus the standard deviation of the means. 80 30. Progression of biofilm thickness for binary population. Continuous line represents time smoothed means and means plus or minus the standard deviation of the means. . . .............. 81 31. Comparison of relative thickness of steady state biofilms of Ki. pneumoniae. P , aeruginosa and binary population. . 82 32. Comparison of biofilm thickness models for K , pneumoniae. P . aeruginosa and binary population. Note differences between duration of 'lag phase', slope of 'log phase' and level of 'plateau'. . . . 84 33. Progression of biofilm cell mass ratio for binary population biofilms. Continuous line represents time smoothed means and means plus or minus the standard deviation of the means................... 85 34. Accumulation of ICi pneumoniae on polycarbonate substratum (Nomarsky microscopic picture, taken 50 hours after inoculation, magnification 625 times). 87 35. Accumulation of K . pneumoniae on polycarbonate substratum (Nomarsky microscopic picture, taken 100 hours after inoculation, magnification 500 times). Cell clusters expand but uncolonized space between clusters still exists................... 88 xiv 36. Accumulation of pneumoniae on polycarbonate substratum (Nomarsky microscopic picture, taken 125 hours after inoculation, magnification 625 times). The colored rings indicate different elevations. . . . ................................ 37. Accumulation of P_^ aeruginosa on polycarbonate substratum (Nomarsky microscopic picture, taken 50 hours after inoculation, magnification 625 times). Cells are more.or less regularly distributed at the substratum.......................................... 38. Accumulation of P_j_ aeruginosa on polycarbonate substratum (Nomarsky microscopic picture, taken 65 hours after inoculation, magnification 625 times). Mono layer of closely packed cells................ 39. Accumulation of a binary population biofilm on polycarbonate substratum (Nomarsky microscopic picture, taken 123 hours after inoculation, magnification 500 times). A relatively smooth surface with peaks sticking out................... 40. Accumulation of a binary population biofilm on polycarbonate substratum (Nomarsky microscopic picture, taken 208 hours after inoculation, magnification 500 times). Surface roughness increases older. . . when binary population biofilms get 41. Comparison of product kinetic models for mono population biofilms of pneumoniae and P . aeruginosa and for binary population biofilms. Single lines represent attached, double lines represent suspended growth models. . . . ........ 42. Comparison of glucose-oxygen stoichiometric ratio for mono population biofilms of K^ _ pneumoniae and P . aeruginosa and for binary population biofilms. Curved lines represent 95% confidence interval. 43. Glucose-oxygen stoichiometric ratio for K . aerogenes grown in a chemostat under various nutrient limitations (Neijssel and Tempest, 1975) Bars indicate Yoq for growth on glucose carbon under resp. phosphorus, nitrogen, sulfur and carbon limitation. . . ............................... .. . 89 90 91 92 93 95 100 101 X V 45. 46. 47 . 48. 49. 50. 51 . 52 . 44 . Transmission Electron Micrograph of a microbial cell in a binary population biofilm. Comparison with a TEM of pneumoniae by Roth (1977) indicates that this picture probably shows a K . pneumoniae cell. ................................ Comparison of biofilm specific glucose uptake rate models for pneumoniae. P . aeruginosa and binary population biofilms. Curved lines represent 95% confidence interval............... ............... . Comparison of growth kinetic models for K . pneumoniae and aeruginosa....................... Comparison of specific cellular growth rate vs. substrate concentration for aeruginosa mono population biofilms (data points) with specific cellular growth rate vs. substrate concentration for suspended growth (continuous line) for P . aeruginosa.......................................... Comparison of specific cellular growth rate vs. substrate concentration for K_^ pneumoniae mono population biofilms (data points) with specific cellular growth rate vs. substrate concentration for suspended growth (continuous line) for K . pneumoniae.............. ............................ Comparison of specific cellular growth rate vs. substrate concentration for binary population biofilms (squares for 10 g C.m experiments, crosses for 20 g C.m experiments) with specific cellular growth rate vs. substrate concentration for suspended growth of pneumoniae and P , aeruginosa.......................................... Schematic representation of accumulation of mono population biofilms of ICl pneumoniae and P . aeruginosa.......................................... Schematic representation ■ of accumulation of a binary population biofilm consisting of K . pneumoniae and P^ aeruginosa cells. Time progression of the diffusion effectiveness factor for various thicknesses of the biofilm of K . pneumoniae and P_^ aeruginosa................... 108 112 113 114 116 119 122 104 180 xvi ABSTRACT Biofilm research has been restricted to studies of undefined mixed microbial populations and to investigations of (defined) mono microbial populations. In the first case, the organisms are considered as a homogeneous mass, biomass, ignoring the properties of individual species, the sum of which determines the observed phenomena. The second case concentrates on the properties and processes of one microbial species ignoring the influence of the broader environment, eg. other microbial species, on this species. Goal of this research was to study a defined mixed microbial biofilm and to determine possible interactive effects between the species comprising the biofilm. Biofilm experiments were conducted with mono populations of Klebsiella pneumoniae and Pseudomonas aeruginosa and with binary populations of pneumoniae and aeruginosa. Process rates and stoichiometric coefficients, determined for the mono population biofilms were compared with those found in the binary population biofilm. Results indicate that the specific cellular product formation rate of pneumoniae or P^ aeruginosa in the binary biofilm is not affected by the presence of the other species. Similarly, the glucose-oxygen stoichiometric ratio of K_^ pneumoniae or P^ aeruginosa in the binary biofilm is not affected by the presence of the other species. Many processes at the cellular level are performed faster by K . pneumoniae than by F\_ aeruginosa: eg. biofilm specific product formation rate and the maximum specific growth rate of pneumoniae are 5 times that rate of Pjl aeruginosa. Nevertheless, K^ pneumoniae cell mass does not dominate the biofilm, possibly because of its non-motility and its product formation properties. IINTRODUCTION A biofilm is a layer of microbial cells and inorganic debris held together in a polymeric matrix and firmly attached to a substratum. Accumulation of biofilm is encountered in many natural environments. Natural purification of surface waters depends to a large extent on the activity of adsorbed microorganisms to remove pollutants from the bulk liquid. In certain engineered environments, adsorption.of organisms is fundamental to the processes anticipated eg., fixed—film biological wastewater treatment. In other environments, adsorption of microorganisms is considered a nuisance resulting in energy losses, sudden deterioration of water quality and possible destruction of long-term exposed surfaces. A major dilemma in the study of biofilm processes is the extreme complexity of natural biofilms which makes distinguishing individual processes and obtaining process oriented information virtually impossible. This is in contrast with study of biofilms of one single microbial species, mono population biofilms. As far as we know, however, these films do not exist in natural systems and information exposed serves mainly to improve theoretical understanding. The focus of this research is accumulation of binary population biofilms, biofilms resulting from accumulation of two identified microbial species. The behaviors of different microbial species respiring in close proximity does not necessarily equal the sum of the behavior of the individual microbial species. Different interactions between organisms have been distinguished. However, most studies in this area have been conducted in suspended growth systems and results may not be directly applicable to systems with attached microbial growth. The scope of 2this study was to elucidate processes describing accumulation of a binary population biofilm by comparison with processes describing accumulation of mono population biofilms. Goal and Objective The goal of this research was to determine the effect of a mixed microbial population on biofilm accumulation processes and biofilm properties. The objective was to describe biofilm accumulation of a binary population of Klebsiella pneumoniae and Pseudomonas aeruginosa in terms of the accumulation of the mono population biofilms of K__ pneumoniae and F\_ aeruginosa. Process variables were monitored to determine the process kinetics and stoichiometry for mono population biofilms of each, species and for the two species combined. ? 3LITERATURE REVIEW Biofilm Processes Biofilm accumulation is the result of microbial, chemical and physical processes occurring in the liquid phase, both within the biofilm and at the substratum: Transport and adsorption of orcranic macromolecules and nutrients from the liquid phase to the substratum: Research indicates that adsorbed macromolecules may both enhance and inhibit microbial adsorption. Characklis and Cooksey (1983) conclude that the role of conditioning film in microbial adsorption to surfaces is not yet clear. Transport and adsorption of microorganisms from the liquid phase to the substratum: Microbial cells (0.5- 10 pm effective diameter) can be transported from the bulk fluid to the wetted substratum by (Brownian) diffusion, gravity, thermophoresis, taxis' and fluid dynamic forces like inertia, lift, drag, drainage and downsweep. The relative contribution of each of those forces depends on the size of the organism, its density, etc. Motility can contribute significantly to the rate of transport of a microbial cell to the substratum. Adsorption of microorganisms may be reversible or irreversible, closely related to the bonding energy between the substratum and the macromolecular conditioning film (Mitchell and Kirchman, 1981). Irreversible adsorption occurs following the production of extracellular fibers that allow the bacteria to overcome repulsive forces (Marshall et al. 1971) . Reactions within the biofilm: Once organisms are adsorbed energy is expended for growth or replication, for the formation of extracellular products (eg. polysaccharides, proteins, small molecules, peptides, 4antimicrobial agents), for cell maintenance, and for death or lysis of the cell. Detachment of biofilm and associated products: Parts of the biofilm may separate from the biofilm and become reentrained in the bulk liquid. Erosion occurs when detachment concerns individual cells. Detachment generally refers to separation of pieces of biofilm (cells and products). Desorption refers to the reentrainment of cells from the substratum. Reactions between the biofilm and the substratum: Reactions may occur between microbial products and the substratum and reactions may occur between locations of the substratum, covered with biofilm and locations not covered. The adsorbed cells grow and reproduce forming colonies which constitute physical anomalies on . a surface. Nonuniform or "patchy" colonization by microorganisms results in the formation of differential aeration cells where areas under respiring colonies are depleted of oxygen relative to surrounding, non-coIonized areas, forming differential surface chemistries. Biofilm Properties Biofilm properties can be distinguished in physical, chemical and biological properties. Relevant thermodynamic properties are volume (thickness seldom over 1000 pm) and mass (dry mass density varies from 10 - 50'kg.m-3 in fluid flow systems) (Characklis and Cooksey, 1983). Both values are highly dependent on hydraulic conditions and on the chemical regime to which the film is exposed. Transport properties of biofilms determine mass, heat and momentum transfer. Diffusion coefficients in biofilms are most probably related to biofilm density (Characklis and Cooksey, 1983); thermal conductivity is not significantly 5different from that of water (Characklis et al., 1981). Chemical properties of the biofilm vary with the chemical composition of the bulk liquid and probably affect the physical and biological structure of the film, eg., interspecies bonding strength is probably affected by calcium (Turakhia, 1986). In addition, inert suspended solids and corrosion products, when substratum is ferro- metallic, may accumulate in the biofilm matrix. Organic composition of biofilms is closely related to the energy, carbon and nutrient sources available for metabolism. For example, nitrogen limitation can result in a relatively large amount of carbon being devoted to production of extracellular microbial polysaccharide (EPS). In terms of macromolecular composition, Bryers (1979) has measured protein-to-polysaccharide mass ratios from 0 - 1 0 (the former in terms of casein equivalents, the latter in terms of glucose equivalents). In addition, the physiological state of organisms is of importance in determining the composition of a biofilm: stationary phase cells are generally EPS producers rather than log phase cells (Characklis and Cooksey, 1983). Biological properties of a biofilm strongly depend on the species colonizing the substratum in addition to the physical and chemical properties of the environment in which the biofilm accumulates. Initial microbial activity on the substratum results in small colonies of cells distributed randomly. In time, colonies may grow together forming a relatively smooth biofilm or grow as isolated colonies with bare substratum in between, forming a patchy biofilm. Viable cell numbers are relatively low in relation to the biofilm volume (IO1 ^ - IO14.m-3), occupying only from I to 10% of the biofilm volume in dilute nutrient solutions (Characklis, 1980; Trulear, 1983) . Cell densities are in 6the order of 1 0 ^ - 1 0 ^ cells.m ^. Microbial Species Klebsiella pneumoniae K , pneumoniae. also described as Klebsiella type I CErbing et al., 1976), is an organism in the coliform group. It is commonly found in soil and water and is present in 30 - 40 % of all warm-blooded animals, including humans. Individual densities range up to 10®.(gram of feces) . K^ pneumoniae is frequently implicated in hospital infections. Approximately 60 - 85% of all Klebsiella from feces and clinical specimens are K . pneumoniae. Strains that are positive by the fecal coliform test (ferment lactose with gas production at 44.5° C) are considered K^ pneumoniae (Geldreich and Rice, 1987). This organism is the rare cause of pneumonia in humans (Brock, 1979). Several investigations (Knittel, 1975; Brown and Seidler, 1973) suggest that the origin of K^ pneumoniae in surface waters could be human, animal or mixed sources (Campbell et al.. 1976). When growing anaerobically, Klebsiella strains fix a property not found among other enteric bacteria (Brock, 1979) . K_^ pneumoniae is generally a non-motile, gram negative, rod shaped bacterium. Relevant characteristics of K , pneumoniae are summarized in Table I. NB. K^ pneumoniae appears to be referred to by a variety of names including, Aerobacter aerocrenes . Klebsiella aeropenes . K , eduardi i . and K^ _ oxytocum (Brown and Seidler, 1973). ■ ■ z Pseudomonas aeruginosa P . aeruginosa is a polymer-forming bacterium, ubiquitous in nature and the cause of many infections and disease. The primary mode of growth is in polymer-enclosed / 7 microcolonies (Buchanan and Gibbons, 1977), attached to a wide variety of substrata. The polymer capsule is assumed to act as a protective layer.- This polymer layer is relatively diffuse, and is easily dispersed in the liquid phase. P . aeruginosa has been studied extensively in both suspension and in biofilms (Trulear, 1983; Mian et al, 1978; Robinson et al., 1984; Bakke et al., 1984; Turakhia, 1986). Kinetic and stoichiometric coefficients for this organisms have been determined. Relevant characteristics of P . aeruginosa are summarized in Table I . Table I. Relevant characteristics of K^ _ pneumoniae and P , aeruginosa. K . pneumoniae P . aeruginosa shape rod shaped rod shaped b breadth (pm) 0.3 - 1.5 h 0.5 - 0.8 J) length (pm) 0.6 - 6.0 I) 1.5 - 4.0 EPS formation + + EPS composition glucose, primarily fucose, mannuronic and glucuronic acid. guluronic acid 5) and pyruvic acid 4) motility non motile polar flagella I) respiration ' facultative obligate aerobe anaerobic )^ except when denitrification + 2) nitrate present 3) metabolism chemoorganotroph )^ chemoorganotroph gram stain negative “■) negative b optimal temperature 35-37° C )^ 35-37° C optimal pH 7.2 i) 6.8 3) h (Holt, 1977) Z) (Sutherland, 1977) 'T) (Buchanan and Gibbons, 1974) ’) (Erbing et al., 1976) 5) (Mian et al., 1978) 8Extracellular Products Extracellular Polymeric Substances (EPS) are by­ products of microbial metabolism produced mostly in the cytoplasm and transported through the cell wall. These, generally large chain macromolecules may either take the form of a discrete capsule located around the cell perimeter or may be present as extracellular slime apparently unattached to the bacterial surface, depending on the species (Sutherland, I977). Capsules are generally quite stable but stability depends on cell age, chemical composition of surrounding liquid, etc. Consequently, capsule layer thickness will vary from hardly discernible to extending 0.1 to 10 pm beyond the exterior of the cell wall (Sutherland, 1977). It should be noted that observations of capsular material very much depend on the sample preparation procedure. Several methods and resulting observations on pneumoniae. discussed by Roth (1977), show that capsule appearance can range from a clearly visible, layered structure with a well-defined edge to not visible. Cell growth rate and the nature of the compound limiting cell growth rate, eg, nitrogen or carbon, strongly influence the rate of product formation and the product yield. Under nitrogen limitation in chemostat growth, specific polysaccharide synthesis by F\_ aeruginosa increased from 0.27 g. (g of celI .h ) at a dilution rate of 0,05 . h-"*- to 0.44 g . (g of cell . h) at a dilution rate of 0.1.h-1. Polysaccharide yields, based on glucose used, ranged from 56 to 64% (Mian, I978). Polysaccharides were also produced under carbon limitation at a rate of 0.19 g.(g of cell.h)-1 at D=0.05.h-1 at a yield of 19%. (Mian et al., 1978). Similar results were obtained with Klebsiella aerogenes, Product formation 9increased with nitrogen relative to carbon as limiting nutrient (Sutherland, 1977). It was even observed that Klebsiella continued to produce considerable amounts of polysaccharide after cessation of cell growth despite.the presence of residual carbohydrate in the growth medium (Dudman, 1960). Neijssel and Tempest (1975), however, report that K^ _ aerogenes . growing at very high substrate concentrations going into the chemostat (10 g C.l-1 when carbon limited, 30 g C.l-1 when nitrogen, phosphate or sulphate limited) and at low dilution rate, 0.17.h~1, did not produce any polysaccharide or other metabolic products when growing under carbon limitations. In that case, cell yields were reportedly 0.45 (g cell-C.(g glucose-C)_1) for growth on glucose carbon: glucose-oxygen stoichiometric ratio was 2.83 (g ©2 . (g glucose-C) -"*-) . Oxygen has also been implicated in production of polymeric material. Dudman (1960) found that increased aeration led to decreased product yields in a variety of microorganisms but resulted in an increased polysaccharide production in coli and Klebsiella species. It seems possible that in bacteria, which are facultative anaerobes utilizing the Embden-Meyerhof pathway, optimal conditions of polysaccharide production include aeration, whereas in obligate aerobes this leads to reduced yield of extracellular polysaccharide (Sutherland, 1977) . Microbial Interactions Many examples of microbial interactions are described in the context of biodegradation, microbial communities degrading compounds. Slater and Lovatt (1984) give an overview of the significance of microbial communities in biodegradation and discuss why techniques for the isolation of strains might fail to reveal community interactions. 10 Batch culturing of an originally adsorbed community, plating and nutrient excess are examples of conditions that are liable to disturb and not reveal species interactions. Traditionally microbial interaction studies have been conducted in laboratory type chemostats CFrederickson, 1977) which allow the definition of the various types of interactions between microbial organisms. Generally interaction has a strong tendency to result in exclusion of one of the competitors. But often mitigating circumstances are present, resulting in 'balanced coexistence' (Van Gemerden, 1974) . Extrapolation of results from suspended growth interaction studies to attached growth cultures is virtually impossible. Homogeneity of chemostat conditions is not available in the attached growth system while characteristic aspects of the latter (transport through laminar sublayer, adsorption, growth on substratum, shear force, gliding, desorption, detachment) are not accounted for in the suspended growth system. Interestingly, discrepancy between the mathematical description of a predator - prey relation and observations could be removed by addition of a wall growth term in the. equation, and Frederickson (1977) concludes that 'variation of the ratio of chemostat wetted surface area to culture volume should be an important part of future studies on microbial predator - prey relations'. If interorganism distance is a variable of. importance in microbial interaction (Crisp, 1981), attached growth seems more liable to result in interaction than suspended growth. Based on data from biofilm experiments in Rototorque reactors (Trulear, .1983) , cell density is approximately 4 orders of magnitude higher in biofilms than in suspension (cell number per volume of biofilm or volume of suspension). <* I 11 Interactions between algal and bacterial species in a variety of environments have been documented (Schiefer and Caldwell, 1982; Haack and McFeters, 1982; Escher and Characklis, 1982; Sandbeck and Ward, 1981). Few have been described to occur on solid substrata. Holmes (1986) studied the colonization of vinyl by 2 algal species on open air swimming pool walls. Rate artd extent of algal colonization drastically increased in the presence of bacteria. Bacterial extracellular products (EPS) seemed to play a role in initial bacterial and in subsequent algal colonization. In a laboratory assembled (no growth took place during biofilm formation) algal bacterial biofilm (Murray et al., 1986), it was shown that algal exudates served as a sole carbon source for attached bacteria. Microbial interaction is mathematically approached by Wanner and Gujer (1985-a) describing biofilm formation by a heterotrophic and autotrophic bacteria. Frederickson (1977) is the first to denote a microbial system consisting of 2 identified species by a binary population. 12 MATHEMATICAL DESCRIPTION Microbial conversion of substrate into cell mass can be mathematically described using a mass balance approach. The equation describing a mass balance is of the following general form: net rate of accumulation net rate of transport net + rate of transformation Mass accumulation equals the sum of rates of mass transformation and rates of mass flow in or out of a control volume. For two microbial species in the system. the following sequence of events is considered: > dissolved substrate > microbial cell mass #1 microbial extracellular product #1 microbial cell mass #2 microbial extracellular product #2 Substrate carbon is used to provide microbial cell carbon and extracellular product carbon. COg is produced as a respiration product. Thus, microbial cell carbon is used for cell synthesis (cell mass) and respiration (energy). 13 Mass Balance Equations for the Biofilm Reactor Licruid phase substrate carbon Cl) : dsSl - DCSgi Sg1) A XM f -rS XM1 ' rPl+ dt V CO YPS net rate net rate of rate of rate of of substrate = substrate - substrate - substrate uptake accumulation where: infIow uptake by biofilm in liquid phase Sg-^ = substrate carbon concentration in liquid phase [MsL-3] t = time [t] D = dilution rate ft ] C= flow rate into divided by volume of reactor) 3Si = substrate carbon concentration in influent [MgL-3] ^Mf " cell carbon concentration in biofilm [M^L-3] A ' = substratum area [L-3] V = reactor volume [L-3] rg = biofilm specific substrate uptake rate [MgMjvJ-^t--*-] Xmi = cell carbon concentration in liquid phase [M^L-3] p. = specific cellular growth rate [t-"*"] Yms - yield of cell carbon from substrate carbon [M^Mg- ]^ rPl - specific product formation rate in liquid phase [MpMM-1t-1] Ypg = yield of product carbon from substrate carbon [MpMg-1] Liquid phase cell carbon (2) : dXMl dt A D(XMi-xMi) + XM f ■- ’rMd + X^i ■ M- net rate of cell mass accumulation net rate of rate of rate of cell cell mass + cell mass + growth in inflow detachment liquid phase 14 where: Xj^ji = cell carbon concentration in influent rMd " specific cell detachment rate [t~*] Licruid phase product carbon (3) : D(Spi-Sp1) + Spf.-.rpd + XM1 Tp1 net rate of product accumulation net rate of product + inflow rate of product detachment rate of product + formation in liquid phase where: Sp1 = product carbon concentration in liquid phase [MpL-3J Spi = product carbon concentration in influent [MpL-3] Spf = product carbon concentration in biofilm [MpL-3] rpd = specific rate of product detachment [t-*] Biofilm substrate carbon (4) : dsSf dt xMf-rS XM f •fD 1 rPl YPS net rate of net rate of rate of substrate = substrate - substrate accumulation uptake consumption for cell growth where: rate of substrate consumption for product .formation Sgf = substrate carbon concentration in biofilm [MgL 3J fp = effectiveness factor, for substrate diffusion [-] 15 Biofilm cell carbon (5): dxMf dt net rate of cell mass accumulation " xMf-rMd net rate of cell mass detachment + xMf -fD-I1 rate of + cellular growth Biofilm product carbon (6) : 3 dSpf dt 3Pf-rPd + xMf-rPf net rate of . product accumulation net rate of product detachment rate of + product formation where: rp^ = specific rate of product formation in biofilm [ M p M M ^ f 1] Oxygen (7): dsOl A — — = d s^Os-sOI-1 + — net rate of oxygen accumulation net rate of oxygen inflow in dilution water rate of + oxygen inflow by diffusion XMf•fD- A V rate of oxygen uptake for cell growth and product formation in biofilm rate of oxygen uptake for cell growth and product formation in liquid phase 16 where: Sq -^ = concentration of oxygen in liquid phase [Mq L- ]^ Sq s = oxygen saturation concentration in influent [Mq L- ]^ Nq = flux of dissolved oxygen into reactor system [M0L"2t_1] Yjvjq = yield of cell carbon from oxygen [MjvjMq - ]^ YpQ = yield of product carbon from oxygen [MpMQ-■*•] The following simplifying assumptions are made: 1. With long biofilm detention times, biofilm substrate concentration (Sgf) is zero. 2. Glucose is the only carbon and energy source in the influent. 3. Specific product formation rate is (Luedeking & Piret, 1959) growth associated and non-growth associated: rpi = kPg -P + kPn (8) where: kpg = growth associated product formation coefficient [MpMjvf1] kpn = non-growth associated product formation coefficient [MpMjvj *t ^] 4. Specific microbial growth rate is dependent on substrate concentration according to Monod (1949): P ^m-4 5Sl, Kq+So -1 (9) where: Pm = maximum specific cellular growth rate [t x] Kg = half saturation concentration [MgL-8] 17 5. The flux of oxygen into the reactor can be estimated as (Appendix A ) : V N0 = kc-csOS-sOl5 -- (A-2) A where: kc = dissolved oxygen specific mass transfer rate [t-1] With dilution water at saturation concentration, the oxygen diffusion component in Eg. 17 can be substituted by an expression similar to transport of oxygen by dilution water: V N0 .- = kc . (C0g-C01) (10) A 18 Mass Balance Equations for the Chemostat Liquid phase substrate carbon C U ) : - D(Ssi-Sg1) net rate net rate of rate of of substrate = substrate - substrate uptake accumulation inflow in liquid phase Liquid phase cell carbon (12) : dxMl ----- = d c xMI-xMI5 + XM1 -V- dt net rate of cellular accumulation net rate of cellular + inflow rate of cell growth in liquid phase Liquid phase product carbon (13) : D (Sn; —Sp-i ) XM1 .rp1 net rate of product accumulation net rate of product infIow rate of product + formation in liquid phase 19 Mass Balance Equations for the Batch Reactor Liquid phase substrate carbon (14): . dt M . rp YMS YPS net rate rate of of substrate = - substrate uptake accumulation ■Liquid phase cell carbon (15) : dxMl dt XM1 net rate of rate of cell cell mass = growth accumulation Liquid phase product carbon (16) : XM1 rPl net rate of rate of product product = formation accumulation Oxygen (17): + Rq net rate of oxygen accumulation rate of oxygen uptake rate of for cell growth and + oxygen product formation generation 2 0 EXPERIMENTAL APPROACH The experimental work can be divided in three parts according to the reactor used. Experiments conducted in the Rototorque resulted in stoichiometric and kinetic coefficients for biofilm processes for both the mono and binary populations. Chemostat experiments provided kinetic and stoichiometric coefficients for mono population processes in suspension. Batch experiments provided stoichiometric data on the consumption of oxygen. Rototorgue The Rototorque, essentially a Couette vessel, consists of two concentric cylinders, a stationary outer cylinder and a rotating inner cylinder (Figure I). Removable slides (4-12) which form an integral part of the inside wall of the outer cylinder permit sampling of the biofilm so that thickness, mass, and/or biofilm chemical composition can be determined. The reactor liquid phase is completely mixed by virtue of draft tubes bored through the solid inner cylinder (Figure I). The draft tubes are positioned at angles so that the rotation of the inner cylinder pumps the fluid through the tubes. By virtue of the complete mixing, effluent liquid samples represent the reactor liquid composition. The Rototorque is a continuous flow stirred tank reactor (CFSTR), an open reactor (i.e ., there are flows in and out) in which concentration gradients within the liquid volume are minimized. The CFSTR provides significant advantages for observing, separating and evaluating the kinetics and stoichiom&try of each biofilm process: PORTPORT CORKCORK RECIRCULATION — ^ TUBE g RECIRCULATION TUBE REMOVABLE SLIDE REMOVABLE SLIDE INNER - CYLINDER — OUTLET Drawn b y G e rrte Slebwl 1615 V. B e a l ls t r e e t B ozenaa MT 59715 (4 06 ) 5 8 7 -3 3 4 7 Figure I . Schematic of a Rototorque reactor. 2 2 1 . The liquid phase is homogeneous which simplifies sampling, chemical analysis, and mathematical modelling. 2. The mass transfer rate and shear stress in the annular geometry at different rotational speeds has been described mathematically CMizushina, 1971). 3. The annular geometry has been used in numerous experimental observations related to biofilm processes CTrulear & Characklis, 1982; Bakke et al. 1984; Turakhia, 1986). Material balance calculations for carbon permit a measure of biofilm activity. ' Oxygen balances can also be performed. Fluid shear stress at the wall is a function of rotational speed. Mean liquid residence time depends on dilution flow rate through the reactor. Thus fluid shear stress and residence time can be varied independently. Reactor residence time was maintained at approximately 10 minutes.so that suspended growth was negligible and all reactor activity can be attributed to the biofilm. The reactor has a high surface area-to-volume ratio and most of its surface area is exposed to a uniform shear stress. A summary of characteristics and dimensions of the Rototorque is presented in Appendix B . Preparation of Rototorque Biofilm experiments were prepared by initiating the standard cleaning procedure for reactor and tubing components. This consisted of brushing all reactor components in a solution of 2% active chlorine followed by rinsing with distilled water. After drying, the slides were inserted and the intake ports on top of the reactor were filled with loosely packed cotton plugs. The air vent was attached, the effluent line connected and closed with a 23 screw clamp. Parts of the slide were provided with a thin layer of resin for the preparation of biofilm Transmission Electron Micrographs (TEM). The entire assembly was then autoclaved for 15 minutes at 121°C . Tubing and flow meters between feed stock solutions and the reactor were flushed with distilled water and tube ends were sealed with aluminum foil and clamped. Tubing was autoclaved for 15 minutes at 121°C . Preparation of Chemostat A 500 ml solution of nutrient medium was made up, added to the clean chemostat and autoclaved for 15 minutes at 121° C . After cooling to room temperature, the chemostat was inoculated with I ml of the appropriate microbial suspension, previously grown in batch culture. The chemostat was operated in batch mode for 12 to 16 hours to obtain a cell density between IO12 and IO13.m~3 . Then a continuous flow of nutrients and substrate (40 g C.m 3) was started. Steady state conditions were assumed to have been reached after 6 residence times. Start-up and Operating Conditions The Rototorque setup was assembled. The reactor was then filled with dilution water and continuous flows started (dilution water 3.6*10"3 m3 .h-1) giving final concentrations of calcium 25 g Ca .m 3 , buffer 1.5 mMol of mono- and dibasic phosphate and substrate 8 or 20 g C.m The rotational speed was set at 200 rpm. After a minimum of 15 minutes, the Rototorques were inoculated by starting a flow of microorganisms (0.06*10 3 m3 . h , cell density 10^2-10-*"^ .m-3) from the chemostat at time zero. The inoculation period was 12 hours. A schematic of the Rototorque experimental setup is presented in Figure 2. CELLS BUFFER DILUTION SUBSTRATE CALCIUM WATER Drawn b y G e r r ie S tebel 1615 V . B e a l ls t r e e t B ozenaa MT 59715 (406 ) 5 8 7 -3 3 4 7 0 6 /0 2 /8 7 ROTDTQROUE SETUP to Figure 2 Schematic of a Rototorque experimental setup 25 Dilution Water The source of dilution water was Bozeman city tap water which was pretreated by fine sand filtration, activated carbon filtration, ion exchange and reverse osmosis. The product was stored in a flow-through tank. Additional micro-filtration, before entering the Rototorques, was provided by a 0.45 pm and a 0.2 pm high volume filter in series (Gelman Sciences Acroflow II) and a O <2 pm mini capsule filter (Gelman Sciences) (Figure 3). Sampling Samples were collected: 1) once or twice daily from the influent line directly before flows enter the Rototorque, representing the influent solution, 2) once or twice daily from the effluent line, representing the liquid phase of the biofilm reactor, and 3) once a day or once every other day from the slides, representing the biofilm phase. Influent samples: 20 ml of influent sample was collected in a test tube, that had been heated previously to 500° C to remove residual carbon, and samples were taken for Total organic carbon (TOC): 4 samples of 2 ml each were pipetted into TOC ampules. Ampules were sealed immediately (Oceanography International Corp., College Station, TX). Glucose: 2 samples of I ml each were pipetted into test tubes, previously heated to 5 00° C , sealed with parafilm and frozen. Effluent samples: 80 ml of effluent sample was homogenized (DuPont Instruments, Sorvall Omni Mixer) for I minute at 60% speed and the following volumes taken for: o w - i z m o ^TAP waters 6 SAND FILTRATION I n. d ep th 1.5 mm, dlom. CARBON FILTRATION 0.8 m. d ep th REVERSE OSMOSIS SOFTENING STORAGE I N D I V I D U A L T R E A T M E N T T 0.45 y tM 0.20 y tMCARTRIDGE CARTRIDGE LL FILTRATION FILTRATION ■-------- % AERATION CHAMBER o.2o y i M CAPSULE FILTER I REACTORS Br*an by Berrie Bebet MlS V. beeIUrtreet BenneA MT 31713 MOti 307-3347 1 7 /2 0 /1 7 VlTCT TbCTTMEKT Figure 3 Overview of dilution water treatment. 27 Total organic carbon (TOC) : 4 samples of 2 ml each were pipetted into TOC ampules. Ampules were sealed immediately (Oceanography International Corp., College Station, TX) . Acridine orange direct count (AODC): 4 ml sample was pipetted into 4 ml 4% sterile formaldehyde solution and stored. Plate count: I ml of sample was added to a sterile test tube filled with 9 ml of sterile water and mixed thoroughly. I ml was then transferred into a second sterile test tube filled with 9 ml of sterile water and again thoroughly mixed. Additional dilutions were added as appropriate to obtain a countable cell density on the plate. After mixing, 0.1 ml of sample from the 3 most appropriate dilutions was added to agar plates, spread, incubated for 24 hours or less if necessary and counted. Suspended mass: 50 ml of sample was filtered through a 0.42 jam Nuclepore filter, that had been dried previously at 103°C for I hour and preweighed. Filters were dried at 103°C for I hour and reweighed. Difference between mass before and after weighing is mass per sample volume. Filtrate of suspended solids measurement was collected for the following determinations: Soluble organic carbon (SOC): 4 samples of 2 ml of filtrate were pipetted into TOC ampules. Ampules were sealed immediately (Oceanography International Corp., College Station, TX) . Glucose: 2 samples of I ml of filtrate were pipetted into test tubes, previously heated to 500° C , sealed with parafilm and frozen. . 28 Dissolved oxygen (DO): 200 ml of effluent was collected in an Erlenmeyer flask and DO was determined with a DO-probe (model 54, Yellow Springs Instruments, Co. Inc.), that was previously standardized in oxygen saturated water. Biofilm samples: A slide was removed from the reactor and immediately immersed in a measuring glass filled with reactor effluent. Biofilm thickness: Thickness was measured by light microscopy, noting stage micrometer setting, by focussing on .the biofilm surface and on the biofilm-substratum interface (Trulear and Characklis, 1982). Biofilm was removed from a known surface area of the slide with a razor blade and added to 50 ml of autoclaved, filtered, carbon free water. The sample was homogenized for I minute at 60% speed and samples were taken: Total organic carbon (TOC): 4 samples of 2 ml each were pipetted into TOC ampules. Ampules were sealed immediately (Oceanography International Corp., College Station, TX). Acridine orange direct count (AODC): 4 ml sample was pipetted into 4 ml 4% sterile formaldehyde solution and stored. Plate count: I ml of sample was added to a sterile test tube filled with 9 ml of sterile water and mixed thoroughly. I ml was then transferred into a second sterile test tube filled with 9 ml of sterile water and again thoroughly mixed. Additional dilutions were added as appropriate to obtain a countable cell density on the plate. After mixing, 0.1 ml of sample from the 3 most appropriate dilutions was added to agar plates, spread, incubated for 24 hours or less if necessary and counted. 29 Adsorbed mass : 25 ml was filtered through a 0.42 |um Nuclepore filter, previously dried at 103°C for I hour and preweighed. Filter was dried again at 103°C for I hour and reweighed. Difference between mass before and after weighing is mass per sample volume. In addition, the following observations were made: Transmission electron micrographs (TEM): Parts of the slide with the adsorbed biofilm sample were immersed in a solution of 2.5% glutaraldehyde in phosphate buffer for a minimum period of 6 to 14 hours, then stored in buffer solution until TEM sample preparation. Optical photomicrographs (OPM): The biofilm slide was air dried and stored in a closed container for microscopic examination. Chemostat The chemostat was used for the determination of kinetic and stoichiometric coefficients. The chemostat is a constant volume, continuous flow stirred tank reactor (CFSTR) used for growth of microbial cells. It consists of a Pyrex beaker (0.45*10"" ^ m-3 voiume) with side arm and rubber stopper. The chemostat was continuously stirred so as to maintain homogeneous conditions throughout the liquid and contained a device for scraping the reactor walls to minimize biofilm accumulation (Figure 4). Preparation and Start-up A 450 ml solution of , nutrients and substrate (40 g glucose-C.m~^) was prepared and added to the chemostat. 30 Figure 4. Schematic of a chemostat. bellows "tubing s u b s t r a te In flu en t scraping disk mixing b a f f le s magnetic s tir r in g disk an tl-b a ck flow devices e f f lu e n t Drawn b y Gwrrlw Slwbel 1613 V . B w aU s irw w t B ozw nea HT 39713 (406) 3 8 7 -3 3 4 7 0 7 /2 8 /8 7 CHEMOSTAT 31 Chemostat tubing was flushed, ends were sealed with aluminum foil and clamped, and the entire assembly was autoclaved for 15 minutes at 121° C . After cooling, the chemostat was inoculated with I ml of the appropriate inoculum, previously cultured . in a batch reactor. The chemostat was operated in batch mode for 12 hours to obtain a cell density of between I O ^ to I O ^ cells.m~^. Then a continuous flow of nutrients was started. Initial flow rate was set at 0.12 I . h "*• (dilution rate 0.27 h- )^ . Operation and Samplincr Steady state conditions were assumed to have been reached 6 residence times (24 hours) after starting up the continuous nutrient flow. Samples were taken by homogenizing 80 ml of effluent (DuPont Instruments, Sorvall Omni Mixer) for I minute at 60 % speed for the following analyses: Total organic carbon (TOC): 4 samples of 2 ml each were pipetted into TOC ampules. Ampules were sealed immediately (Oceanography International Corp., College Station, TX) . ' Acridine orange direct count (AODC): 4 ml sample was pipetted into 4 ml 4% sterile formaldehyde solution and stored. Suspended mass: 50 ml was filtered through a 0.42 pm Nuclepore filter, previously dried at 103°C for I hour and preweighed. Filter was dried again at 103°C for I hour and reweighed. Difference between mass before and after weighing is mass per sample volume. The filtrate of suspended mass measurement was collected for the following determinations. Soluble organic carbon (SOC): 4 samples of 2 ml of filtrate were pipetted into TOC ampules. Ampules were sealed 32 immediately (Oceanography International Corp., College Station, TX). Glucose: 2 samples of I ml of filtrate were pipetted into test tubes, previously heated at 500°C. sealed with parafilm and frozen.. The chemostat was operated at the following dilution rates: 0.10, 0.13, 0.40, 0.50, 0.56, 0.67. 0.80, 1.07, 1.60 and 1.87 h-"*" . After each change in dilution rate, a period of 6 residence times elapsed before samples were taken. Batch Reactor Principles of Operation The batch reactor used is a respirometer (Oceanography International Corp., College Station, TX) designed to continuously monitor and replace oxygen consumed as a result of microbial activity. It consists of a sample bottle with a manometer and an electrolysis cell (Figure 5). When oxygen is consumed, the oxygen partial pressure decreases, raising the electrolyte level on the inside of the manometer and, consequently, lowering it on the outside. This breaks an electrical circuit through a switch electrode and initiates oxygen production on the DO-electrode simultaneously producing hydrogen on the Hg-electrode. The hydrogen is allowed to leave instantaneously while the oxygen pressure reduces the electrolyte level in the inside and raises it on the outside of the manometer. Oxygen is produced until the outside electrolyte level in the manometer touches the switch electrode. CO2 produced as a result of microbial activity is removed from the air space by KOH pellets. The time that oxygen is generated is monitored and as 33 Figure 5. Schematic of a batch reactor. +dc oxygen e le c tro d esw itch e le c tro d e -d c hydrogen e le c tro d e e le c t r o ly te odapto i— alkali con ta iner s t ir r in g magnet 1613 V . f l# a l ls - t r e # t B o ze n ea MT 39715 (4 0 6 ) 3 8 7 -3 3 4 7 BOD r * s p iro m e te r 34 oxygen production is proportional to the current flowing through the electrolysis cell, DO-production can be determined as a function of time. Monitoring-. Operation and Samplincr The amount of oxygen produced and, hence, of oxygen consumed is registered automatically as a function of time. When oxygen consumption (temporary) stalls, a plateau has been reached and liquid samples were taken by homogenizing 80 ml of suspension (DuPont Instruments, Sorvall Omni Mixer) for I minute at 60% speed for the following analyses: Total organic carbon (TOC): 4 samples of 2 ml each were pipetted into TOC ampules. Ampules were sealed immediately (Oceanography International Corp., College Station, TX). Acridine orange direct count (AODC): 4 ml sample was pipetted into 4 ml 4% sterile formaldehyde solution and stored. Suspended mass: 25 ml was filtered through a 0.42 p.m Nuclepore filter, previously dried at 103°C for I hour and preweighed. Filter was dried again at 103°C for I hour and reweighed. Difference between mass before and after weighing is mass per sample volume. Filtrate of suspended mass measurement was collected for the following determinations: Soluble organic carbon (SOC): 4 samples of 2 ml of filtrate were pipetted into TOC ampules. Ampules were sealed immediately (Oceanography International Corp., College Station, TX) . Glucose: 2 ml samples of I ml of filtrate was pipetted into test tubes', previously heated to 500° C , sealed with parafilm and frozen. I35 Nutrient Medium Nutrient Solution The composition, of the influent nutrient solution (Table 2) is identical to the one discussed by Turakhia (1986) with the exception that the calcium concentration is maintained at 25 g Ca.m ^ . Table 2. Composition of nutrient solution. batch reactor chemo- annular stat reactor units glueose-Carbon 40.0 40.0 8.0 -3g.m NH4CL 36.0 36.0 T . 2 -3 g.m 13 MgSO4 .TH2O 10.0 10.0 2.0 -3 g.m J (NH^)6M07O24.4 H20 .005 .005 .001 g .m-^ ZnSO4 .7H20 .5 .5 .1 — 3g.m MnSO4 -H2O .040 .040 .008 -3g.m CuSO4 .SH2O .010 .010 .002 -3g.m Na2B4O7 .IOH2O .005 .005 . 001 -3g.m FeSO4 -TH2O .560 .560 .112 -3g.m (HOCOCH2)3N 2.0 2.0 .4 -3g.m CaCO3-Ca - - 25.0 -3g.m (Na2HPO4) 4.0 4.0 1.5 mmol Na2HPO4 568.0 568.0 213.0 g.m"3 (KH2PO4) 4.0 4.0 1.5 mmol) KH2PO4 544.0 544.0 204.0 -3 g.m final pH 6.8 6.8 6.8 — 36 Preparation of Plates Agar plates of 2 different media composition were used, depending on the bacterial strain ~to be counted. Pseudomonas aeruginosa plates were prepared by adding 12 grams of Tryptone Glucose Extract Agar (Difco Laboratories, Detroit, MI) to 500 ml of distilled water. The solution was heated to dissolve the medium, autoclaved for 15 minutes at 121° C , allowed to cool and poured into plates. Klebsiella pneumoniae plates were prepared by adding 24 grams of Bacto Agar (Difco Laboratories, Detroit, MI) and 7.5 grams of m Endo Broth MF (Difco Laboratories, Detroit, MI) to 500 ml of distilled water. Mixture was heated. 10 ml of absolute ethanol was added just before boiling and the mixture was allowed to cool and poured into plates. Microbial Species Klebsiella pneumoniae The K_i_ pneumoniae strain used in this study was originally isolated in the water distribution system of New Haven, CT. and was obtained from Anne Kamper, Dept, of Microbiology, Montana State University, Bozeman, MT. Pseudomonas aeruginosa The F\. aeruginosa strain used for this study was obtained from the American Type Culture Collection, Denver, CO. Analytical Methods An overview of analytical procedures and their interrelations is shown in Figure 6. t o t a l -C (= ) ce ll-C (+ ) g lucose -C (+ ) po lym er-C ■ ru n by Oeme Oebel 1613 V. I w l l r t i w l WL HT 30713 (406) 307-3347 07/20/07 SMVUNB SOOC W Figure 6, Overview of the sampling scheme and analytical procedures. 38 Acridine orange direct count: Total number of cells in effluent . and resuspended biofilm samples were determined after staining with acridine orange according to Hobbie et al. (1977). Plate count: Cell density in liquid phase was determined by multiplying the cell number counted per plate by the dilution factor and dividing by the volume of diluted sample, applied per plate. Cell density in the biofilm was determined by multiplying the cell number counted per plate by the dilution factor, dividing by the volume of diluted sample applied per plate, dividing by the liquid volume into which the biofilm sample was suspended and dividing by the surface area of the biofilm removed. Mass: Mass in effluent and resuspended biofilm samples was determined as described in sections on Sampling. Glucose carbon: Glucose carbon CSg-^ ) was determined using the procedure of the enzymatic Glucose Analysis Procedure Sigma 510 (Sigma Chemical Co., St. Louis, MO) and modified by TruTear (1983) . Carbon: Carbon was determined with the ampule analysis module of the Oceanography International Carbon Analyzer (Oceanography International Corp., College Station, TX). Cell carbon: Cell carbon in liquid phase (X^1) and in the biofilm (Xjvlf) was. determined by multiplying the cell number obtained with the Acridine Orange I Direct Count by, successively, cell density (1.07*10^ g cell.m I) cell volume, 2) Doetsch et al., 1973; Bakken and Olsen, 1983), 3) dry cell mass to wet cell mass ratio (0.22 dry cell mass.(wet cell m a s s ) Bakken and Olsen, 1983) and 4) ratio of > carbon mass to dry cell mass (0.5 g cell carbon.(g 39 cell dry mass)-1, Doetsch et al., 1973). Licruid phase product carbon: Mass of product carbon in the liquid phase (Sp^) was determined by subtracting the mass of liquid phase glucose (Sgj) and cell carbon (Xj^ j) from the liquid phase total organic carbon mass: Sm — Sip^ r1I ~ S C11 ~ Xjvji (18) lass of product carbon in the determined by subtracting the (Xjyjp) from the biofilm Biofilm »P1 “ 0TOCl - °S1 product carbon: biofilm (Spg) was mass of biofilm cell carbon of liquid phase total organic carbon: Spf = ST0Cf - XMf (19) Licruid phase adsorbed product carbon; Mass product carbon adsorbed to the cell was determined by subtracting glucose carbon, soluble organic carbon and cell carbon from total organic carbon: sTOCl - Sc - SSOCl (20) Dissolved oxygen: Dissolved oxygen was determined with a DO-probe (model 54, Yellow Springs Instruments, Co. Inc .) . Biofilm thickness: Biofilm thickness was measured according to Trulear (1983) and calculated according to Bakke and Ollson (1986) . Biofilm thickness data are the mean of 8 observations along premarked points on the center line of the slide. Transmission electron micrographs (TEM): Samples stored in buffer solution were prepared for electron microscopy according Appendix F . Optical photo micrographs (PPM): Nomarsky microscopy on the air dried biofilm slides was used to follow progression of biofilm formation processes. 40 Statistical Methods Time progression of biofilm variables can generally be represented by a sigmoidally shaped curve, either starting from zero at time zero and increasing in magnitude to reach a steady state value or starting at a certain value at time zero and decreasing in magnitude to a steady state value. The sigmoidal shape can mathematically be represented by the logistical equation. Measured variables were fit to the logistic equation using the statistical package Logit (LOGIT, 1986). The logistic function is presented in the following version: A + Ag (X-X0) + A (X-X0)2 B + Bg (X-X0) + B (X-Xq)2 Y = ----------------- -------- + ------------------------ --- I (X-X0) r Q (X-X0Tl Q 1 I X 3 I £ > X I X I -------- 1 + e 2 D0 L_ -J -------- 1 + e 2D0 L_ - I 1+e 1+e I.......... A ......... i I...... . B ........ I With X being the independent variable, in this case, time, Y is some form of system response in time. The model consists of 3 regions. The first part of the equation (A) relates to the pretransition region of the model for which the intercept (A), the slope (As) and a quadratic term (Aq) define the pretransition asymptote. Similarly, the second part of the equation (B) relates to the posttransition region of the model for which the intercept (B), the slope (Bg) and a quadratic term (Bq) define the posttransition asymptote. The transition region of the model is described by the transition region width parameter (Dq) , the position of the transition region (Xq) and the asymmetry parameter (Q) which allows, the curvature in the pre- and post 41 transition ends of the transition region to vary. Asymptotes for both pre- and posttransition region were assumed to be described by the intercept terms only (As, Ag, Bg and Bg were made zero). As a result, the model is described by the intercept parameters A and B, the position parameter Xq and width parameter Dq of transition region and the asymmetry parameter Q . Determination of parameter values allowed analysis of experimental results by means of the model generated. Nonlinear regression using the saturation model was used to correlate TOC standards to TOC readings, and to correlate dilution rate to substrate effluent concentration in the chemostat experiments (BMDP, 1983). Linear regression was used to correlate glucose standards to glucose readings and to determine product formation coefficients (MSUSTAT, 1986). 42 RESULTS Batch reactor and chemostat experiments using Klebsiella pneumoniae were conducted for the determination of stoichiometric coefficients and coefficients describing the growth and product formation kinetics. Raw data from the batch experiments are listed in Appendix C, from the chemostat experiments in Appendix D . Rototorque experiments were conducted with Pseudomonas aeruginosa and Klebsiella pneumoniae. both in mono and in binary population experiments. Objectives were to determine the biofilm formation properties of K^ pneumoniae. to compare them with those of P_^ aeruginosa and to study those properties when both organisms form the biofilm. Raw data and parameter estimates from the Rototorque experiments are listed in Appendix G (K^ pneumoniae) , H (P^ aeruginosa) and I (binary population). Throughout this thesis cell mass, product mass and substrate mass data are reported as carbon equivalents. Glucose was used as the sole source of carbon and energy at concentrations of 40 g C.m ^ for the batch and chemostat experiments and at 10 g C.m ^ for the Rototorgue experiments. In two binary population experiments the influent substrate concentration was 20 g C.m- .^ When applicable, results are expressed in terms of mean ± standard error. K^ pneumoniae. P . aeruginosa and binary populations are, when space required, referred to by KP, PA and BP. 43 Batch Experiments Batch experiments were conducted to determine the oxygen consumption by pneumoniae (see Figure 7). Experimental results are presented in Appendix C , a summary is presented in Table 3. Table 3. Summary of batch reactor experimental results. Reactor Product Cell Product Product Glucose soluble mass adsorbed t otal oxygen ratio Spi XM1 5Pla 3Plt YSO # .... (g C.m-3) . CgC.gO-1) I 10.7 7.1 3.7 14.4 0.90 II 11.3 7.6 4.7 16.0 0.91 III 9.0 8.0 4.4 13.4 0.90 MEAN 10.3 7.6 4.3 14.6 0.90 STDEV 1.2 .9 2.0 2.3 0.01 Yield Coefficients Yield coefficients can be calculated from the results of the batch reactor experiments: Table 4. Means and standard deviations of yield coefficients in batch reactor experiments. Y Y Y Y MS SO i PSl2 PSt = 0,20 ± 0.02 gc .gc_r = 0.90 ± 0.01 9c'gO-I = 0.27 ± 0.03 9c'gC-I = 0.38 ± 0.06 g^.gQ I Yield of soluble substrate carbon. product carbon from 2 Yield of total substrate carbon. product carbon from 44 Figure 7 Typical progression of oxygen consumption by K . pneumoniae in a batch reactor. BATCH EXPERIMENTS oxygen uptake - Kp > 10 - 0 y X K X X X M M TIME (h) 45 Chemostat Experiments Relevant kinetic and stoichiometric information for K . pneumoniae was obtained by conducting a series of chemostat experiments and determining, for various dilution rates, steady state values of O cell carbon concentration, , O product carbon concentration, Sp^, O substrate carbon concentration in influent, Sg^, and in effluent , Sg-^ , and O biomass carbon concentration, X^ q . Results are presented in Appendix D, a summary is presented in Table 5. Table 5. Summary of chemostat experimental results for K , pneumoniae. Dilution rate (h_I) sSi SS1 XX1 .... (g XM1 C.m”3) CO Tl 3Pla 5Plt .00 36.4 .0 .0 .0 .0 .0 .0 . 10 37.4 .3 8.1 5.7 9.1 2.3 11.4 . 13 37.3 .3 10.d 7.8 9.6 2.2 11.9 .27 37.1 .7 13.8 7.4 8.5 6.4 14.9 .27 37.1 .4 13.4 6.1 8.6 7.4 16.0 .40 37.0 . .6 14.0 7.2 8.9 6.8 15.7 .40 37.0 .2 14.3 8.1 9.3 6.2 15.5 .50 36.5 .5 13.0 5.8 10.3 7.1 17.5 .50 36.5 1.3 12.2 7.2 11.0 5.0 16.0 .56 35.4 .2 12.6 6.3 11.4 6.3 17.7 .56 35.4 .6 12.7 11.0 11.3 1.7 13.0 .67 34 .5 .2 14.8 6.2 11.2 8.6 19.8 .67 34.5 .0 12.6 8.4 11.2 4.3 15.5 .80 35.9 .4 14.5 4.2 10.1 10.3 20.4 .80 35.9 .4 12.1 7.3 9.6 4.8 14.4 1.07 37.1 .8 ■11.1 6.6 10.1 4 .5 14.6 1.07 37.1 1.0 12.9 7.7 9.0 5.2 14.1 1.60 36.7 6.0 11.5 ■ 4.3 8.4 7.1 15.5 1.87 36.7 17.8 10.2 I .2 3.5 9.0 12.6 46 The effect of dilution rate on effluent concentration of substrate (stars), product (crosses) and cell carbon (triangles) is demonstrated in Figure 8. Also indicated is the influent substrate concentration (squares). Cell carbon concentration, rather stable over the range of dilution rates from 0.13 to 1.3 h— , decreases beyond this value and reaches zero around 2 h--*- . Product carbon behaves very similar to cell carbon and appears to indicate that product concentration is a function of growth rate. Simultaneously, glucose carbon concentration in the effluent is almost zero below dilution rates of I h--*-, but. increases rapidly hereafter. Growth Kinetic Coefficients A material balance on cell carbon in the chemostat is described b y : dXMl ------ — D + xMl'^ (12) dt With data collection in the chemostat experiment taking place in steady state conditions and no cell carbon entering the chemostat, Eq. 12 can be rewritten as: 0 = - D.X^i (21) or after dividing by Xjvq , D = n (22) indicating that at steady state, the dilution rate equals the specific cellular growth rate. As microbial growth rate is dependent upon substrate concentration (Monod, 1949), Eg. 22 can be written as t^ m-sSl Kg+Sgi D = M- (22) 47 Figure 8. Change in chemostat carbon concentrations with varying dilution rate (squares = influent glucose carbon, crosses = product carbon, triangles = cell carbon and stars = glucose carbon in effluent). CHEMOSTAT EXPERIMENT summary of results DILUTION RATE ( l / h ) Substrate concentration in the effluent vs. dilution rate (Figure 9) was modelled according to Monod kinetics for determination of Hm and Kg, using non-linear regression (BMDP, 1983). Parameter values are: Hm = 2.00 ± 0.29 h-1 and Ks = 1.43 ± 0.53 g C.m-3. Product Formation Kinetic Coefficients A material balance for product formation in the chemostat is given by Eg. 13: dSpi -------- — D(Sp^ —Sp^) + XMj rpj (13) dt ' 48 L Product growth and non-growth associated kinetic coefficients (Luedeking and Piret, 1959) can be determined by rearranging Eq1 13. Assuming steady state conditions., no product material in the chemostat influent, substituting Eq. 8 for the specific product formation rate, assuming that the specific cellular growth rate equals the dilution rate and dividing by the cell carbon concentration, the following expression arises: ----- — k-pg. D + kpn (23) XM1 representing a linear equation for which the intercept equals the non-growth associated product formation coefficient ^ p n) and the slope equals the growth associated product formation coefficient (kp^). Using linear regression (MSUSTAT, 1986), parameter values were 49 Figure 9. Change in chemostat effluent glucose carbon concentration with varying dilution rate. CHEMOSTAT EXPERIMENT growth kinetics 1.75- O 1.25- DILUTION RATE ( l / h ) 50 obtained for formation of unadsorbed product (calculated as the difference between SgQQ-^ and Sg^) and for formation of total product (unadsorbed + adsorbed product, the last calculated as the difference between S-j-QCl ~ 5SOCl ~ xMl-1 ■ Table 6. Specific growth- and non-growth associated product formation coefficients for K . pneumoniae in the chemostat. solubIe total ^ kPg kPn (g .g 1) (g.g™^™1) 1.45 ± 0.13 -.01 ± 0.06 2.33 ± 0.35 -.02 ± 0,20 ^ sum of soluble product and product adsorbed to cell mass. Values for the intercept were not different from zero at the 5% level of significance. A plot of dilution rate vs. specific (unadsorbed and total) product formation rate is shown in Figure 10 together with the lines describing the linear models. Under the conditions of this chemostat experiment, almost two-thirds, of the product formed appears unadsorbed while the remainder is adsorbed to the cell. Cell and Product Yield Coefficients When accepting a saturation model for the relation between effluent substrate concentration and dilution rate, it seems reasonable to assume that, considering the relation between substrate consumed and biomass produced, cell mass and product mass will follow a model that is related to the saturation model. Therefore, mass of cell carbon and product carbon were fitted to an inverse and displaced saturation model described by the following equation: SP EC IF IC P R O D U C T FO RM AT IO N RA TE ( g /g .h ) 51 Figure 10. Specific product formation rate vs. dilution rate for chemostat growth of pneumoniae. Models for soluble (triangles, R2=O.95) arid total product (crosses, R2=O.86) are represented by continuous lines. CHEMOSTAT EXPERIMENT product kinetics DILUTION RATE ( 1 /h ) 52 - A * ( X - C ) Y = ---------------- B - ( X - C ) where Y = response (eg. product carbon in effluent), X = dilution rate D, A = parameter describing maximum value of concentration, B = parameter comparable to Kg in the saturation equation, and C = parameter describing model displacement, comparable with nm in the saturation equation. The dilution rate at which no substrate is consumed can neither support growth of microbial cells nor can product be formed. Consequently, the parameter for the model displacement, Cf does necessarily assume the value of dilution rate at which the effluent glucose concentration equals the influent glucose concentration. Mass of cell carbon and product carbon vs, dilution rate were fitted to an inverse and displaced saturation model. The models are graphically , shown in Figure 11 together with the raw data. Also shown is the model curve for effluent substrate concentration. Yield coefficients for the conversion of substrate to cell mass and product mass appear not necessarily constant over the whole range of dilution rates (Figure 12). Mean values and standard deviations for the yield coefficients are summarized below. 53 Figure 11. Change in chemostat carbon concentrations with dilution rate (squares = influent glucose carbon, crosses = product carbon, triangles = cell carbon and stars = glucose carbon in effluent). Continuous lines represent models describing the change in carbon concentration with dilution rate. CHEMOSTAT EXPERIMENT summary of results RATE ( 1 /h )DILUTION 54 Figure 12. Variation in yield coefficients with dilution rate (triangles = yield of product carbon, crosses = yield of cell carbon). CHEMOSTAT EXPERIMENT yield coefficients DILUTION RATE ( 1 /h ) 55 Table 7. Yield coefficients for Kj. pneumoniae chemostat experiments. yMS = 0.19 ± 0.05 sc-sc"1 yPSI1 = 0.27 ± 0.04 sc-sc"1 Ypst2 = 0.45 ± 0.09 Sc-Sc"1 I Yield of substrate soluble carbon. product carbon from 2 Yield of substrate total carbon. product carbon from Cell Dimensions as Function of Dilution Rate Cell dimensions were measured during the cell count procedure using Image Analysis Microscopy. Cell length increases rapidly with increasing dilution rate until a dilution rate of approximately 0.6 to 0..7 h-1 after which it decreases. Cell length at maximum dilution rate is very similar to cell length at low dilution rate. In contrast, cell breadth remains, almost constant over the range of dilution rates ("Figure 13) . 56 Figure 13. Variation in cell length and breadth with dilution rate. Length of bars is two times the standard error. CHEMOSTAT EXPERIMENT cell length & breadth 57 Biofilm Experiments Biofilm experiments included 3 sets of 4 experiments conducted in the Rototorque. The first set featured K . pneumoniae. the second set featured aerucrinosa. both as mono populations. The last set of experiments was conducted with both organisms in a binary population. Initial conditions in all experiments were the same (Sg^ = 10 g C.m except for 2 binary population experiments (Sg^ = 20 g C.m . As a result, the only variable in these experiments is the microbial inoculum. The following parameters were determined: In influent: - Glucose - Dissolved Oxygen, Sf In effluent: - Cell mass, XMl - Product mass, Sp^ - Glucose, Sgj - Dissolved Oxygen Sqj In biofilm : - Cell mass, - Product mass, Sp^ - Biofilm thickness, 6 p . Results are presented for Kjl pneumoniae. for Pi aerucrinosa and for the binary population experiments. Raw data and parameter estimates for the biofilm experiments are summarized in Appendix G CKi pneumoniae) , H (Pi aerucrinosa) and I (binary population). • 58 Progression of Biofilm Carbon Components In a typical progression, the mass of biofilm cell carbon in a Ka. pneumoniae experiment (Figure 14) is one tenth of that in a aeruginosa experiment (Figure 15). However, the product mass is almost the same for both organisms, indicating a relatively high specific product formation rate for Ka. pneumoniae. By comparison, the combined mass of biofilm cell carbon in the binary population biofilm is a little higher than for the P. aeruginosa biofilm, and the mass of product carbon is about twice as great (Figure 16). Biofilm composition in terms of carbon for Ka. pneumoniae. P . aeruginosa and the binary populations (with low, 10 g C.m~3 and high, 20 g C.m“3 substrate influent concentration) is shown for steady state conditions in Figure 17. Specific Rates Information on biofilm specific rates may give an indication of biofilm activity and as such is a useful tool in the evaluation of biofilm performance. 59 Figure 14. Typical progression of organic carbon in a K . pneumoniae biofilm. Length of bars is two times the standard error. MONO POPULATION BIOFILM progression of carbon in biofilm KP O .2- tr .1 - TIME (h) 60 Figure 15. Typical progression of organic carbon in a P , aeruginosa biofilm. Length of bars is two times the standard error. MONO POPULATION BIOFILM - PA progression of carbon in biofilm M ll-C o .2- TIME (h) 61 Figure 16. Typical progression of organic carbon in a binary population biofilm. Length of bars is two times the standard error. BINARY POPULATION BIOFILM progression of carbon in biofilm 62 Figure 17. Comparison of organic carbon components in steady state biofilms of K_^ pneumoniae. P . aeruginosa and binary populations CL = grown on low glucose carbon concentration, 10 g C.m- , H = grown on high glucose carbon concentration, 20 g C.m d) . CARBON IN BIOFILM Bp-HBp-L product-C g Kp-Cell-C ^ Pa-Cell-C 63 Specific Cellular Detachment Rate. The specific cellular detachment rate (rdM'> obtained from the balance of liquid phase cell carbon CEq. 2). dxMl ---- - D 1-Xjvj1 - Xjv11-H - ' dt rMd = ---------------------- C24) : A x„f .- Biofilm Specific Cellular Growth Rate. Rewriting the mass balance of biofilm cell carbon (Eq. 5) by dividing by Xjvj^ , I dXjvjf -------- = - rMd + p f (25) Xjvjf dt permits the calculation of the biofilm specific cellular growth rate. Specific Product Detachment Rate. The specific rate of product detachment may be determined from the liquid phase product balance (Eq. 3): dSpi ---- - D 1-Spl - XjjllTp1 dt rpd = --------------------------- (26) When assuming that, the liquid phase cell concentration Xjj1 is negligible (liquid detention time in Rototorque is 10 minutes), Eq1 26 reduces to: 64 rdP (27) Biofilm Specific Product Formation Rate. Biofilm specific product formation rate is determined by dividing the biofilm product mass balance (Eg. 6) by, the biofilm cell mass Xjvjfi dSpf Spf ------ = — ---,rPd rPf (2 8) x M f - d t X j v j f Specific biofilm cellular growth and biofilm product formation rates for steady state conditions are summarized for Kjl pneumoniae. -P . aeruginosa and the binary populations with low (10 g C .rtf"3) and high (20 g C .m~3) substrate influent concentration (Figure 18). Product Formation Coefficients Product formation coefficients (Eg. 8) can be evaluated for, biofilm conditions according to Luedeking and Piret (1959) : rPf “ kPg-Pf + kPn (29) by performing linear regression (MSUSTAT, 1986) of rpf versus |_if (Figures 19. 20 and 21 for Kjl pneumoniae. P , aeruginosa and for binary populations), Table 8 summarizes values for kpg and kpn . . 65 Figure 18. Comparison of specific biofilm cellular growth rate (hatched) and biofilm product formation rate (cross-hatched) in steady state biofilms of pneumoniae. P . aeruginosa and binary populations (L = grown on low glucose carbon concentration, 10 g C.m , H = grown on high glucose carbon concetration, 20 g C.m ^). SPECIFIC RATES IN BIOFILM specific cellular growth and product formation rate .3-1--------------------------------------------------------------------------------------------------------------------------- Kp Pa Bp-L Bp-H SP EC IF IC P R O D U C T FO RM AT IO N RA TE ( g /g .h ) 66 Figure 19. Biofilm specific product formation rate vs. biofilm specific cellular growth rate for K , pneumoniae. Continuous line is regression model (R2=O.74). BIOFILM PRODUCT FORMATION - KR BIOFILM SPECIFIC CELLULAR GROWTH RATE ( 1 /h ) SP EC IF IC P R O D U C T FO RM AT IO N RA TE ( g /g .h ) 67 Figure 20. Biofilm specific product formation rate vs. biofilm specific cellular growth rate for P . aeruginosa. Continuous line is regression model (R2=O.31). BIOFILM PRODUCT FORMATION - PA BIOFILM SPECIFIC CELLULAR GROWTH RATE ( l / h ) SP EC IF IC P R O D U C T FO RM AT IO N RA TE ( g /g .h ) 68 Figure 21. Biofilm specific product formation rate vs. biofilm specific cellular growth rate for binary population. Continuous line is regression model (R2= O .90). BIOFILM PRODUCT FORMATION - BR BlOFlLW SPECIFIC CELLULAR GROWTH RATE ( 1 /h ) 69 Table 8. Specific growth- and non-growth associated product formation coefficients in the biofilm for pneumoniae. P . aeruginosa and binary population. . K , pneumoniae P . aeruginosa binary population kPg Cg.g 1) 1.72 ± 0.11 0.36 ± 0.07 0.75 ± 0.03 kpn Cg.g_1h_1) 0.20 ± 0.041 0.10 ± 0.071 0.08 ± 0.021 Values for kpn are different from zero at the 5% level of significance. Biofilm Specific Substrate Uptake Rate Biofilm specific substrate uptake rate can be evaluated from the material balance for biofilm phase substrate (Eg. 4). When assuming that there is no substrate in the biofilm, substituting Hf for the product of fp and p , kPg'^f + kPn for the product of fp and rp^, and after dividing by the concentration of biofilm cell carbon, , Eq. 14 changes to ^f rPf 1 kPg kPn rS + --- = Pf • + yMS YPS yMS' yPS . CA) YPS CE) in which rg is defined as the biofilm specific substrate uptake rate. rg can be determined from the liquid phase substrate balance CEg. I') when assuming that liquid phase cell mass is negligible: 70 rS = D(SSi-SSl) XMf A- V (31) Using linear regression of rg versus (Figures 22, 23 and 24 for pneumoniae. for P^ _ aeruginosa and binary populations) allows the evaluation of the biofilm specific substrate uptake rate expressed in a growth (A in Eg. 30) and a non-growth (B in Eg. 30) associated term (Table 9) . . Table 9. Values for the growth- (A) and the non-growth (B) associated term in Eg. 30 of specific substrate uptake rate in the biofilm for K , pneumoniae, P , aeruginosa and binary population. K . pneumoniae P . aeruginosa binary population (A) (g.g-1) .16,30 ± 1.07 0.10 ± 0.60 6.61 ± 0.10 (B) (g.g~1h~1) 0.41 ± 0.761 0.25 ± 0.191 0.20 ± 0.071 Values for the intercept are different from zero at the 5% level of significance. Biofilm Cell and Product Yield Coefficients YjvJg and Ypg can now be calculated from Eg. 30 (Table 10) . 71 Figure 22. Biofilm specific substrate uptake rate vs. biofilm specific cellular growth rate for K . pneumoniae. Continuous line is regression model (R2=O.72). BIOFILM SUBSTRATE UPTAKE - KR BIOFILM SPECIFIC CELLULAR GROWTH RATE ( 1 /h ) 72 Figure 23. Biofilm specific substrate uptake rate vs. biofilm specific cellular growth rate for P . aeruginosa. Continuous line is regression model (R2=O.57). BIOFILM SUBSTRATE UPTAKE - PA BIOFILM SPECIFIC CELLULAR GROWTH RATE ( 1 /h ) 73 Figure 24 . Biofilm specific substrate uptake biofilm specific cellular growth binary population. Continuous regression model (R -0.98). rate rate line BIOFILM SUBSTRATE UPTAKE - BR BIOFILM SPECIFIC CELLULAR GROWTH RATE ( 1 /h ) vs . for i s 74 Table 10. Yield coefficients Y^g and Ypg in the biofilm for pneumoniae. P . aeruginosa and binary population. K . pneumoniae P . aeruginosa binary population Ym s (g.g-1) 0.08 ± 0.04 Yps (g.g-1) 0.48 ± 0.90 0.24 ± 0.06 0.41 ± 0.41 0.22 ± 0.05 0.38 ± 0.43 Biofilm Glucose-Oxygen Stoichiometric Ratio The ratio of oxygen consumed and glucose consumed (Hg) can be evaluated by comparing the specific uptake rates of oxygen and glucose. The specific glucose uptake rate is given by Eg. 31. The specific oxygen uptake ' rate can be derived from Eg A-4 (Appendix A) by assuming liguid phase activity negligible, defining Tq as ^f rPf rO - ----- + ---- YM0 YP0 (32) and rearranging: rO (D+kc) .(S0s-Soi) dsOl dt (33) Using linear regression of rg versus r0 (Figures 25, 26 and 27 for pneumoniae. for P^ _ aeruginosa and binary populations) allows the evaluation of the glucose-oxygen stoichiometric ratio. Table 11 summarizes results. 75 Figure 25. Biofilm specific glucose uptake rate vs. biofilm specific oxygen uptake rate for K . pneumoniae. Continuous line is regression model CR2 = O. 93) . BIOFILM GLUCOSE-OXYGEN STOIC. RATIO - KR SPECIFIC OXYGEN UPTAKE RATE (g /g .h ) O 15 76 Figure 26. Biofilm specific glucose uptake rate vs. biofilm specific oxygen uptake rate for P . aeruginosa. Continuous line is regression model (R2=O.34). BIOFILM GLUCOSE-OXYGEN STOIC. RATIO - PA .8 1 1.6 2 2.4 3 3.2 SPECIFIC OXYGEN UPTAKE RATE (g /g .h ) 77 Figure 27. Biofilm specific glucose uptake rate vs. biofilm specific oxygen uptake rate for binary population. Continuous line is regression model (R2=O.94). BIOFILM GLUCOSE-OXYGEN STOIC. RATIO - BR SPECIFIC OXYGEN UPTAKE RATE (g /q .h ) 78 Table 11. Glucose-oxygen stoichiometric ratio in the biofilm for pneumoniae. P . aeruginosa and binary population. K . pneumoniae P . aeruginosa binary population Ag Cg.g-1) 2.90 ± 0.08 0.96 ± 0.18 1.46 ± 0.04 I (g.g-1h_1) 0.33 ± 0.371 0.07 ± 0.201 0.19 ± 0.131 *8 1 The intercepts CD are different from zero at the 5% level of significance. Biofilm Thickness Biofilm thickness was determined by measuring thickness at 8 locations on the biofilm slide. The mean value for the 8 data points represents the average biofilm thickness at that point and time. The variation in values is indicative of the roughness of the biofilm. Individual measuring points per sampling time and time smoothed curves (LOGIT, 1986) through the means of the points and through the means + o r - the standard deviation have been combined for K . pneumoniae. P , aeruginosa and binary population (Figures 28, 29 and 30) . Individual measuring points for ID. pneumoniae cover a wide range of thicknesses, indicative of a rough biofilm surface. This contrasts with the aeruginosa biofilm where the range of thicknesses, especially past the I00 hours mark, is relatively narrow. The binary population biofilm is relatively smooth in the period around 100 hours and widens after that. Comparison of the variation in biofilm thickness is facilitated (Figure 31) by contrasting the relative mean biofilm thickness and standard deviation at steady state for the three populations. 79 Figure 28. Progression of biofilm thickness for K . pneumoniae. Continuous line represents time smoothed means and means plus or minus the standard deviation of the means. PROGRESSION OF BIOFILM THICKNESS - KR 80 Figure 29. Progression of biofilm thickness for P , aeruginosa. Continuous line represents time smoothed means and means plus or minus the standard deviation of the means. PROGRESSION OF BIOFILM THICKNESS - PA TIME (h) 81 Figure 30. Progression of biofilm thickness for binary population. Continuous line represents time smoothed means and means plus or minus the standard deviation of the means. PROGRESSION OF BIOFILM THICKNESS - BR TIME (h) 82 Figure 31. Comparison of relative thickness of steady state biofilms of pneumoniae. P . aeruginosa and binary population. RELATIVE BIOFILM THICKNESS 300 T---------------------------------------- K. pneumoniae P. aeruginosa B. population 83 The extent of accumulation, the length of the period preceding accumulation and the rate at which accumulation takes place are compared in terms of biofilm thickness in Figure 32. pneumoniae accumulates to a lesser extent than F\_ aeruginosa (approx. 15 vs. 30 pm at steady state) or the binary population (approx. 40 pm). Biofilm accumulation for pneumoniae starts in a very early stage of the experiment relative to accumulation for P . aeruginosa (starting at 50 hours) and the binary population (starting at 100 hours), The last two have clearly distinguishable periods' in which no accumulation takes place. The maximum rate of increase of biofilm thickness for P^ aeruginosa (8 p m .h~^) is twice as high as that of the binary population (4 p m . h-"*-) . Maximum rate of thickness increase for pneumoniae is only a fraction of that (0.4 p m .h-1) . Species Distribution Distribution of .K^ pneumoniae and P_^ aeruginosa in the biofilm is of importance in the evaluation of properties of the binary population biofilm. Data points and time smoothed curves (LOGIT, 1986) for the means and the means + or - the standard deviation of pneumoniae cell mass fraction in the biofilm have been combined (Figure 33). Starting the experiments with both species having the same cell mass , steady state, conditions show that K_j_ pneumoniae cell mass occupies 26% of the biofilm. Photographic Illustrations Transmission Electron Micrographs were made from the binary population. In addition, Nomarsky Photo Micrographs (NPM's) were made of the dried surface of the K . pneumoniae. P , aeruginosa and the binary population biofilm. The differences between the three populations in 84 Figure 32. Comparison of biofilm thickness models for K . pneumoniae. P . aeruginosa and binary population. Note differences between duration of 'lag phase', slope of 'log phase' and level of 'plateau'. PROGRESSION OF BIOFILM THICKNESS TIME (h) 85 Figure 33. Progression of biofilm cell mass ratio for binary population biofilms. Continuous line represents time smoothed means and means plus or minus the standard deviation of the means. BIOFILM SPECIES DISTRIBUTION progression of cell moss ratio TIME (h) 86 biofilm accumulation can be illustrated with Figures 34 through 40. K , pneumoniae accumulates in little. well spaced clusters of cells (Figure 34) which rapidly multiply. Cells appear to stay together enlarging the cluster. This results in expansion parallel and perpendicular to the substratum (Figure 35) resulting in the formation of "microtowers11 (Figure 36) which can reach a height of 60 pm and more with bare substratum between the towers. Only gradually is the space between clusters being filled up. P , aeruginosa colonizes the substratum in a relatively r equal distribution (Figure 37). Cells grow and multiply, reducing intercellular distance. Gradually, a smooth biofilm builds up (Figure 38). A binary population biofilm consists of a relatively smooth surface with peaks sticking out (Figure 39). Gradually, surface roughness increases (Figure 40). 87 Figure 34. Accumulation of pneumoniae on polycarbonate substratum (Nomarsky microscopic picture, taken 50 hours after inoculation, magnification 625 times) . 88 Figure 35. Accumulation of K^ _ pneumoniae on polycarbonate substratum (Nomarsky microscopic picture, taken 100 hours after inoculation, magnification 500 times). Cell clusters expand but uncolonized space between clusters still exists. 89 Figure 36. Accumulation of pneumoniae on polycarbonate substratum (Nomarsky microscopic picture, taken 125 hours after inoculation, magnification 625 times). The colored rings indicate different elevations. 90 Figure 37. Accumulation of aeruginosa on polycarbonate substratum (Nomarsky microscopic picture, taken 50 hours after inoculation, magnification 625 times). Cells are more or less regularly distributed at the substratum. 91 Figure 38. Accumulation of aeruginosa on polycarbonate substratum (Nomarsky microscopic picture, taken 65 hours after inoculation, magnification 625 times). Mono layer of closely packed cells. 92 Figure 39. Accumulation of a binary population biofilm on polycarbonate substratum (Nomarsky microscopic picture, taken 123 hours after inoculation, magnification 500 times). A relatively smooth surface with peaks sticking out. 93 Figure 40. Accumulation of a binary population biofilm on polycarbonate substratum (Nomarsky microscopic picture, taken 208 hours after inoculation, magnification 500 times). Surface roughness increases when binary population biofilms get older. 94 DISCUSSION Processes in mixed population biofilms of Kju pneumoniae and aerucrinosa have been observed. The results will be discussed in terms of. observations in mono population biofilms. Product Formation Both K^ pneumoniae and F\_ aerucrinosa form extracellular products. Erbing et al. (1976) found that pneumoniae produced D-glucose, L—fucose, D—glucuronic and pyruvic acid in equal ratios. Neijssel and Tempest (1975) emphasized the effect of the limiting nutrient in the excretion of products. Yields of polysaccharide of almost 40% simultaneous with yields of various acids (appr. 30%) occurred with chemostat growth under nitrogen limitation. The range of products and the relative quantities in which they are produced is determined by the qualitative and quantitative composition of the nutrient solution (Neijssel and Tempest . 1975) . P_._ aeruginosa produces extracellular material in both batch and chemostat growth. In chemostat studies, P^ aeruginosa produced mannuronic and guluronic acid in a ratio of 4:1 (Mian et al. 1978). Growth- and non-growth associated product formation coefficients are summarized in Table 12. Nutrient limitation other than carbon, appears to increase the specific rate of growth-associated product formation (Mian et al. 1978). In this study, specific rates of product formation by both Kj_ pneumoniae and P^ aeruginosa are a function of the specific cellular growth rate, both in suspended and in attached growth (Figure 41). SP EC IF IC P R O D U C T F OR MA TI ON RA TE ( g/ g. h) 95 Figure 41. Comparison of product kinetic models for mono population biofilms of pneumoniae and P . aeruginosa and of binary population biofilms. Single lines represent attached, double lines represent suspended growth models. PRODUCT FORMATION -- = biofilm, = = suspended growth GROWTH RATE / MAX GROWTH RATE ( - ) 96 Product Formation in Suspension Specific product formation rates in suspension for K . pneumoniae (C-Iimited) are considerably higher than those measured by Trulear (1983) for aeruginosa (C-Iimited) but in the same range as those measured by Mian et al. (1978) for aeruginosa (N-Iimited) . Neijssel and Tempest (1975) , studying chemostat growth of K_^ aerogenes under carbon, sulphate, ammonia or phosphate limitations for four different carbon sources (glucose, glycerol, mannitol and lactate) found a wide range of specific product formation rates. Under carbon limitation, the specific product formation rate of K_^ aer ogenes (product concentration determined according to Herbert et al., 1971) was zero, but excess carbon (carbon in excess of the stoichiometric required amount) invariably resulted in excretion of ' overflow' metabolites'. The specific growth-associated product formation coefficients for K__ aerogenes grown in suspension on glucose under limitations other than carbon are in the range of 2.2 - 3.5 g product G .(g cell C)-^ (Neijssel and Tempest (1975). Specific growth-associated product formation coefficients found in this research are 1.4 g product C .(g cell C) ^ for suspended growth of K . pneumoniae (Table 12). This suggests that growth under non­ carbon limitation could be responsible for the high specific product formation rates found in this research. Comparison of the glucose mineral salts medium used in this study for the growth of pneumoniae and P_^ aeruginosa with the one used by Neijssel and Tempest (1975, specified .by Evans et al., 1970, Appendix E) indicates that the glucose mineral salts medium contains an excess of all nutrients but one, cobalt, required at 0.2 |ug.m~^. Batch and chemostat experiments were conducted using distilled water, the Rototorque experiments.used specially treated water (Bozeman city tap water followed by softening and 97 reverse osmosis). Both water sources may contain cobalt in insufficient concentrations, leaving growth media cobalt limited. It should be noted that in this study and in the study by Trulear (1983) product formation in suspension included soluble product and product - adsorbed to the cell while Neijssel and Tempest (1975) measured only soluble product. Product Formation in the Biofilm For the binary population biofilms, the growth- associated product formation coefficient is 0.75 ± 0.02 g product C , (g cell C)- .^ Based on a biofilm composed of 7495 P . aeruginosa and 26% iv_ pneumoniae and assuming no interaction between species, the specific growth-associated product formation coefficient in the binary population biofilm can be predicted by summing the products of the growth-associated product formation coefficient for each species found in the mono population biofilm experiments and the cell mass fraction in the binary population biofilm. The specific growth-associated product formation coefficient, based on species distribution is 0.72 ± 0.11 g product C .(g cell C)-1. At the 5% level of significance, there is no difference between the growth-associated product formation coefficient in the binary population biofilm determined experimentally and that predicted on the basis of species composition. Consequently, in the binary population biofilm, the specific product formation rate of K . pneumoniae or P_._ aeruginosa appears not to be affected by the presence of the other species. 98 Table 12. Summary of specific growth- (kp ) and non­ growth- (Itpn) associated product formation coefficients. Reactor kPg kPn 3Si Reference (g.g-1) (g.g-1h-1) (g C.m-3) P . aeruginosa chemost. chemost. biofiIm biofilm biofilm biofilm 3.29±0.53l 0.36±0.44 1.0 ±0.86 0.01±0.14 0.02± 0.13 0.36±0.07 0.13±0.04 0.03±0.IO1 2 34 0.25±0.36 0.09±0.04 0.01±0.03 0.10±0.07 8000 36.8±2.3 3.8±I . I 9.6± 0.1 16.7±0.3 9.8± 0.3 Mian et al. (1978) TrUlear (1983) Trulear (1983) Trulear (1983) Trulear (1983) This study K . pneumoniae chemost. chemost. chemost. chemost. chemost. chemost. biofilm 0 2.2 3.4 3.5 I . 44±0.06 2.2 9± 0.17 I t72± 0.11 4 5 6 7 : s 0.20±0.04 10,000 30.000 30.000 30.000 36.4±0,9 36.4± 0.9 10.8± 0 . I Neiissel and Tempest (1975) Neijssel and Tempest (1975) This study This study This study Binarv population biofilm 0.75±0.03 0.08±.0.025 8.9/20.5 This study 1 Mean ± standard deviation, except if otherwise noted. 2 Not different from zero at 5% level of significance. 3 Expressed as mean ± 95% confidence interval. 4 , 5 , 6 , 7 Determined under respectively carbon, sulphate, ammonia and phosphate limitations. 8 Product material in suspension (not attached to cells). 9 The sum of product material in suspension and attached to cells. 99 Glucose-Oxycren Stoichiometric Ratio Conversion of glucose into biomass can be described stoichiometrically as follows (Busch, 1971): C5h I2C6 + 2 .4602 + 0.YlNH2 — > 0.YlC5H7NO2 + 2 .46C02 + 0.46H20 (34) Thus , I g of oxygen i s required for the biochemical conversion of 0.92 g (= Yso) of glucose carbon into biomass. This equation was derived for the removal of glucose by an undefined mixed population in a batch reactor system, which was initially substrate saturated. Turakhia (1986) found Y^q = 0.89 ± 0,06 g glucose carbon consumed.(g oxygen consumed)- ^ for batch growth of F\_ aeruginosa and YgQ = 0,92 ± 0.02 g glucose carbon consumed.(g oxygen consumed) "*"*■ for F\_ aeruginosa biofilms. This study found a glue os e-oxygen stoichiometric ratio of 0.96 ± 0.18 g glucose carbon consumed, (g oxygen consumed) for P . aeruginosa biofilm experiments (Figure 42), well in agreement with the theoretical value (Eg. 34) and the other investigations. Glucose-Oxygen Stoichiometric Ratio for K. pneumoniae In this study YgQ" = 0.90 ± 0.01 g glucose carbon consumed.(g oxygen c o n s u m e d ) i n C-Iimited batch growth of K . pneumoniae. In contrast, Neijssel and Tempest (1975), studying chemostat growth of K^ _ aerogenes . reported YgQ varied between 2.8 (C-Iimited) and 4.5 (N-Iimited) g glucose carbon consumed. (g oxygen consumed) (Figure 43) . For the mono population biofilms of pneumoniae. YgQ = 2.90 . ± 0.08 g glucose carbon consumed,(g oxygen consumed) (Figure. 42) . This is the same value as found by G LU C O SE U PT A K E RA TE ( g /g .h ) Figure 42. Comparison of glucose—oxygen stoichiometric ratio for mono population biofilms of K . pneumoniae and Pi aeruginosa and for binary population biofilms. Curved lines represent 9596 confidence interval . 100 BIOFILM GLUCOSE-OXYGEN STOIC. RATIO SPECIFIC OXYGEN UPTAKE RATE (g/g.h) C O N S U M P T I O N RA TI O (g /g ) IOl Figure 43. Glucose-oxygen stoichiometric ratio for K . aerogenes grown in a chemostat under various nutrient limitations (Neiissel and Tempest, 1975) . Bars indicate YgQ for growth on glucose carbon under resp. phosphorus, nitrogen, sulfur and carbon limitation. GLUCOSE-OXYGEN STOIC. RATIO chemostat growth under various nutrient limitations carbon ammoniasulphate NUTRIENT LIMITATION phosphate 102 Neijssel and Tempest (1975) for suspended growth of K . aerogenes under glucose carbon limitation. Yg0 for Binary Population Biofilms The glucose-oxygen stoichiometric ratio, Yg0 , in the binary population biofilm can be determined by summincr the products of the glucose—oxygen stoichiometric ratio for each species found in the mono population biofilms and the cell mass fraction in the binary population biofilm. This results in a glucose—oxygen stoichiometric ratio, based on species distribution, of 1.47 ± 0 .20 g glucose carbon consumed.(g oxygen consumed)-1. Yg0 determined experimentally equals I .44 ± 0.02 g glucose carbon consumed.(g oxygen consumed)-1. At the 5% level of significance, there is no difference between Yg0 in the binary population biofilm determined experimentally and Yg0 determined on the basis of species composition. Consequently, in the binary population biofilm. the glucose-oxygen stoichiometric ratio of pneumoniae or P . aeruginosa appears not to be affected by the presence of the other species. Binary population biofilm experiments were conducted at substrate carbon concentrations of 10 and 20 g C.m-3. The extent of biofilm accumulation for the 20 g C.m-3 experiments is approximately twice the accumulation for the 10 g C.m ^ . Nevertheless, the glucose—oxygen stoichiometric ratio is the same for both experiments. This implies that both the glucose consumption .rate and the oxygen consumption rate for the 20 g C.m-3 is double that rate for the 10 g C.m 3 experiments. This is consistent with the observation of DO concentrations in the effluent of the 20 __ O g C.m experiments which are lower than those in the 10 g C.m . Lower DO concentrations in the effluent mean lower DO concentrations in the biofilm. Hence, anaerobic 103 microenvironments in the aggregates may more likely occur in the 20 g C.m-3 than in the 10 g C.m-3 experiments. As mono population biofilm experiments of pneumoniae may be substrate diffusion limited, diffusion limitation is also likely to occur in the binary population biofilm experiments. Consequently, microbial activity of both species is reduced in the deeper layers of the biofilm. This is consistent with the observation that the species distribution in the 20 g C.m-3 experiments was not different from the one in the 10 g C.m-3. Influence of Diffusional Resistance on Ygp The difference in glucose-oxygen stoichiometric ratio between aeruginosa and pneumoniae may be related to the respective physiology of the organisms or to differences in diffusional resistance within the respective biofilms. K . pneumoniae has a high specific growth rate and a high specific rate of product formation. In addition, this organism is non-motile. As a result , K^ _ pneumoniae biofilm cells aggregate to a high degree relative to P . aeruginosa. The aggregates can reach a height of .60 pm or more with a base width of approximately 30 pm. Cohesive forces between the cells in the aggregates are sufficient to resist the hydraulic pressure at the aggregates in a direction parallel to the substratum. Extracellular product (Figure 44) and cell density may impose a considerable diffvisional resistance for transport of dissolved compounds to and from the biofilm ■cells. A comparison of the effectiveness factor for diffusion of substrate into 104 Figure 44. Transmission Electron Micrograph of a microbial cell in a binary population biofilm. Comparison with a TEM of pneumoniae by Roth (1977) indicates that this picture probably shows a K , pneumoniae cell. I k 105 biofilms which differ only in the microbial species that has accumulated. K^ _ pneumoniae and aeruginosa ("Appendix K), suggests that this is indeed the case (Table 13). Table 13. Relation between steady state biofilm thickness and diffusion effectiveness factor € for K . pneumoniae and Pjl. aeruginosa. thickness (pm) K . pneumoniae P . aeruginosa 10 64% - 15 45% 92% 30 23% 78% Resistance against substrate diffusion in a biofilm of 30 p.m thickness . built up of pneumoniae. is 3 times as high as in a similar biofilm of P^ aeruginosa. Diffusional resistance.in aggregates of pneumoniae with a base width of 30 pm or more is, therefore, certain to affect substrate concentrations in the core of the aggregate significantly. 106 Factors affecting Initial Adsorption Initial adsorption at the substratum is considerably different for the two organisms (Figures 34 through 40). Various factors influence adsorption including the nature the wetted substratum (material of construction and surface roughness), the deposition of conditioning film, and the characteristics of the adsorbing cells. Cell suspensions pumped into the Rototorgue were observed microscopically for aggregation as this could affect both the rate of particle diffusion through the laminar sublayer and subsequent adsorption and growth. Aggregation was not observed for either K , oneumoniae and P ■ aeruginosa. The substratum in the biofilm studies was polycarbonate, a hydrophobic material. The cell wall of both organisms is hydrophobic at neutral pH (Buchanan and Gibbons, I974) but the extent of hydrophobicity may be different. Also, both organisms produce product, some of which is immediately released into the environment, possibly more so for aeruginosa than for pneumoniae . Observation of the dried substratum showed that in the early stages of the experiment polymeric material was deposited in mono population biofilm experiments of both organisms. Drying of the substratum resulted in cracking (microscopically observed) of the Pjl aeruginosa deposit but not of that of K . pneumoniae. This could indicate that the former deposit is thicker than the latter. Product deposited at the substratum or adsorbed to the cell may obscure the effects of hydrophobicity and, therefore, may affect cell adsorption. 107 Specific Cellular Glucose Uptake Ratio The specific cellular glucose uptake rate is proportional to the specific cellular growth rate. The proportionality coefficient (Hg) is the inverse of the observed yield coefficient Y0J33 and can be approximated by: n S I ^obs (35) Specific cellular glucose uptake ratios (Ag) for suspended and attached growth of Kjl pneumoniae and aeruginosa are summarized in Table 14. Ag^ for aeruginosa biofilms in this study and in Truleaf's (1983) are not different at the 5% level of significance. But Ag^ for the suspended growth is significantly different from the one for biofilms. Also, Ag for Kk_ pneumoniae is considerably higher than for P , aeruginosa (Figure 45). The main reason for Ag being higher in attached than in suspended growth is that the cellular growth yield is higher in suspension than in biofilms (Eg. 35) . This is true for both pneumoniae ■ and for P , aeruginosa (Table 15), ) 108 Figure 45. Comparison of biofilm specific glucose uptake rate models for K^ _ pneumoniae . P , aeruginosa and binary population biofilms. Curved lines represent 95% confidence interval. BIOFILM SUBSTRATE UPTAKE < ( 1 5 - < 10 - BIOFILM SPECIFIC CELLULAR GROWTH RATE (1 /h ) 109 Table 14. Comparison of specific cellular glucose uptake ratios (fig) for pneumoniae and aeruginosa in suspension and in biofilm.. suspended growth attached growth Reference p. aeruginosa 3.5 ± 0.2^ 4.6 ± 1.3 Trul ear (1983) 5.1 ± 0.6 This study K. pneumoniae 11.6 ± 1 . 7 16.3 ± 1.1 This study I Mean ± standard deviation. . - Table 15. Comparison coefficients aeruginosa. of growth for Ki. and product yield pneumoniae and P . suspended growth attached growth Reference P . aeruginosa Ym s Cg.g 1) 0.36±0.381 0.25±0.132 Trulear (1983) Yps Cg.g"1) 0.50±1.94 0.43±0.72^ yMS Cs-g-1) y ps cg.g""1) K . pneumoniae 0 .24±0.06 0.41±0.41 This study Yms Cg.g-1) 0.19±0.05 0.08±0.04 This study Yps Cg.g-1) 0.27±0.04 0.4 8± 0.90 Mean ± standard deviation. Mean of three values for Y^g and Ypg.2 H O For the binary population biofilm, the experimental D q 7^ — 1is 6.6 ± 0.1 glucose carbon.(g cell C) 1 (Figure 45). Based on cell mass distribution and assuming no interaction, the theoretical ^Sf equals 8.0 ± 1,0 glucose carbon, (g cell C) . The values for the theoretical and the experimental specific cellular glucose uptake ratio are different at the 5% level of significance. An experimental nSf which is less than Agf determined on the basis of species distribution would suggest species interaction. This contradicts earlier findings. Therefore, the specific cellular glucose uptake ratio, Agf, for the binary population biofilm appears to be underestimated. I l l Growth Kinetics Suspended Growth The saturation kinetic (Monod) model describes the relation between substrate, concentration and microbial growth rate in suspended growth systems. Robinson et al. (1984) report growth kinetic coefficients for P . aeruginosa: ^max = 0.40 ± 0.01 h-1 and Kg = 2.0 ± 0.5 g glucose carbon.m-3. In this study, growth kinetic coefficients for pneumoniae were: Hmax = 2.0 ± 0.3 h-1 and Kg = 1.4 ± 0.5 g glucose carbon.m 3 . A comparison of chemostat growth kinetics (Figure 46) emphasizes the large difference in maximum specific growth rate between K . pneumoniae and aeruginosa. Mono Population Biofilms Saturation kinetics also apply to f\_ aeruginosa biofilms (Bakke et al., 1984). In this study, biofilm specific cellular growth rate of mono population biofilms of P^ aeruginosa can also be described by the saturation equation (Figure 47) . The dependence of biofilm specific cellular growth rate on substrate concentration. in K^ _ pneumoniae biofilms can not be described by saturation kinetics (Figure 48). The observed behavior may be explained by diffusion limitation (Appendix K ) . Specific cellular growth rate for K . pneumoniae biofilm decreases rapidly during substratum colonization. The rapid formation of "microtowers" by this organism results in diffusional resistance so that most of the nutrients are consumed by the cells in the outer layer of the aggregate. Consequently, only a fraction of the cells is exposed. to the liquid phase substrate concentration. The majority of the cells 'see' a much lower 112 Figure 46. Comparison of growth kinetic models pneumoniae and aeruginosa. GROWTH KINETICS SUBSTRATE CARBON CONCENTRATION (g /m 3 ) for 113 Figure 47. Comparison of specific cellular growth rate vs. substrate concentration for F\_ aeruginosa mono population biofilms (data points) with specific cellular growth rate vs. substrate concentration for suspended growth (continuous line) for aeruginosa. SPECIFIC GROWTH RATE - PA < 1 . 5 - □ P SUBSTRATE CARBON CONCENTRATION (g /m 3 ) 114 Figure 48. Comparison of specific cellular growth rate vs. substrate concentration for Kj, pneumoniae mono population biofilms (data points) with specific cellular growth rate vs. substrate concentration for suspended growth (continuous line) for Kj, pneumoniae. SPECIFIC GROWTH RATE (KR) SUBSTRATE CARBON CONCENTRATION (g /m 3 ) 115 substrate concentration and exhibit a much lower specific cellular growth rate. With the increase of both size and number of aggregates, the specific biofilm cellular growth rate decreases with the bulk liquid. substrate concentration. Binary Population Biofilms Progression of biofilm specific cellular growth rate vs. reactor substrate concentration for a binary population biofilm (Figure 49) shows that at high reactor substrate concentration biofilm specific cellular growth rate decreases almost independently of reactor substrate concentration. With accumulation of biofilm, the decrease of specific cellular growth rate follows the decrease in liquid phase substrate concentration. The change of specific biofilm cellular growth rate from almost independent, to proportional to liquid phase substrate concentration coincides with the change in biofilm species composition, ie., from a biofilm composed of equal amounts of cell mass of each species to a biofilm dominated by P . aeruginosa. Comparison with the corresponding figure for K . pneumoniae (Figure 48) suggests that biofilm specific cellular growth rate is initially determined by cells in the rapidly growing . aggregates of pneumoniae. When biofilm cell mass becomes dominated by P^ _ aeruginosa. the specific biofilm cellular growth rate reflects the combined effect of growth of a P^ aeruginosa biofilm and growth of a K . pneumoniae biofilm. It should be noted that K_^ pneumoniae is the dominant species in the effluent. This finding is in agreement with the fact that, due to its high specific cellular growth rate, K. pneumoniae will rapidly produce cells in the 116 Figure 49. Comparison of specific cellular growth rate vs. substrate concentration for binary population biofilms (squares for 10 g C.m- ^ experiments, crosses for 20 g C.m experiments) with specific cellular growth rate vs. substrate concentration for suspended growth of K . pneumoniae and aeruginosa. SPECIFIC GROWTH RATE - BR < 1.5- SUBSTRATE CARBON CONCENTRATION (g /m 3 ) 117 .biofilm. Consequently, more K^ _ pneumoniae cells will detach. aeruginosa nevertheless dominates the biofilm cell mass because the rate of (re-)attachment is higher for P , aeruginosa than for pneumoniae (Appendix J). In addition, forces between cells which permit the formation of aggregates, suggest that detachment of K_^ pneumoniae is likely to occur in "lumps" of cells rather than by single cells. Diffusion of these lumps through the laminar sublayer is considerably less than diffusion of individual cells of K^ _ pneumoniae. Therefore, reattachment of K , pneumoniae is small compared to reattachment of P . aeruginosa. Accumulation of_a Binary Population Biofilm: a Conceptual Description This section is intended to provide an overall picture of the sequence of events in initial adsorption and subsequent accumulation of a binary population biofilm.. Binary population biofilm accumulation is the net result of the following events: 1. Transport to. the substratum from the liquid phase through the laminar sublayer: pneumoniae is not motile, Fb_ aeruginosa is a motile organism. Rate of transport of single cells of Fb_ aeruginosa through the laminar sublayer could be as high as 250 times the rate of transport of K^ _ pneumoniae cells. Assuming an equal number of cells of both species in the liquid phase, 250 times more cells of Pb, aeruginosa could reach the substratum as compared to pneumoniae cells. 2. Net adsorption (= adsorption to and desorption from the substratum): Probability of adsorption to the substratum is often related to extracellular product 118 formation (Marshall et al. 1971). After initial contact in which stage cells are assumed to remain at close proximity of the substratum only (reversible adsorption), cells either desorb or bridge the distance to the substratum by means of extracellular material (irreversible adsorption) . The specific rate of product formation by pneumoniae is almost 5 times the specific rate of product formation by aeruginosa. Consequently, probability that pneumoniae will remain at the substratum could be almost 5 times that of F\_ aeruginosa . . Following desorption, cells of F^ _ aeruginosa are 250 times as likely to return to the substratum as cells of K . pneumoniae. 3. Motility: Daughter cells of pneumoniae will remain close to the parent cells and form cellular aggregates which expand both horizontally and vertically (microtowers) resulting in a patchy structure. The relationship between motility, high specific product formation rate, and formation of aggregates has been related by Costerton et al. (1985) by proposing that 'daughter cells are trapped in a juxtaposition that results in the formation of microcolonies'. P . aeruginosa. a motile organism, has the ability to move over the substratum (Marshall, personal communication) . P . aeruginosa will form a relatively smooth layer of cells (Figure 50) . 4. Growth rate: Maximum specific growth rate of K , pneumoniae is 5 times the maximum specific growth rate of P^ aeruginosa. Therefore, the mass of pneumoniae will increase as much as 5 times as fast as that of P . aeruginosa. However. the high specific growth rate together with the high specific rate of product P s e u d o m o n a s a e r u g in o s a K le b s ie l la p n e u m o n ia e P s e u d o m o n a s a e r u g in o s a K le b s ie l la p n e u m o n ia e Figure 50. Schematic representation of accumulation of mono population biofilms of pneumoniae and P . aeruginosa. 119 1 2 0 formation may result in nutrient (oxygen, carbon) depletion in the core of the aggregate, significantly reducing (overall) biofilm specific cellular growth rate. With Pjl. aeruginosa spreading out over the substratum, depletion of nutrients may occur only in a later stage of biofilm accumulation. 5. Detachment: Considering the adhesive celI-substratum and the cohesive cell-cell forces necessary to resist the hydraulic pressure at the aggregate in the flow system, detachment may occur in "lumps" rather than by single cells. The likelihood of lumps reattaching may be small since particle diffusion rate through the laminar sublayer is inversely proportional to particle diameter. Detachment might also occur when oxygen concentration in the core of the aggregate approaches zero. In fact, anaerobic conditions in the core environment may initiate detachment. Detachment of cells of aeruginosa has been shown to occur as lumps (sloughing) and as single cells (Bakke, 1986). 6. Accumulation: The two species form a biofilm. apparently unaware of the presence of the other species. Ki. pneumoniae forms aggregates in which a P . aeruginosa cell may become incorporated by attachment. P , aeruginosa forms a smooth layer of cells. In time, both species expand, Pi. aeruginosa taking up space between the aggregates, Ki pneumoniae aggregates towering above the relatively smooth Pi aeruginosa film. With the surrounding structural support of the Pi aeruginosa biofilm,. Ki pneumoniae cell aggregates may develop to a greater height before detachment occurs than if this support were not present (Figure 51). Summarizing, the accumulation of a binary population biofilm of Ki pneumoniae end Pi aeruginosa is the net 121 result of cellular transport, growth, product formation and motility properties. The proliferation of species in this binary population biofilm is, therefore, determined by the physiology of the species present. 122 Figure 51. Schematic representation of accumulation of a binary population biofilm consisting of K . pneumoniae and P^ _ aeruginosa cells. B . JP O p 'L t l C L tXO TL B . J p o j p X L l c z t X O T L O O - c.y o u c> O 0 - '<11 CJ CJ 123 CONCLUSIONS The following conclusions, valid within the range of experimental conditions tested,. can be drawn from this study with K , pneumoniae and P . aerucinosa in binary population biofilms: 1. Specific product formation rate of K__ pneumoniae or P . aeruginosa in the binary population biofilm is not affected by the presence of the other species. 2. Specific product formation rate in suspended growth and in the biofilm for K_^ pneumoniae is 5 times the specific product formation rate in suspended growth and in the biofilm for P^ aeruginosa. 3. Specific product formation rate for pneumoniae and IL_ aeruginosa is linearly related to specific cellular growth rate. The slope of. the linear relation, the growth associated product formation coefficient, is the same for suspended and for attached growth for each species. 4. Glucose-oxygen uptake ratio of pneumoniae or Pj aeruginosa in the binary population biofilm is not affected by the presence of the other species. 5. The ratio of specific biofilm cellular glucose uptake rate and specific biofilm cellular oxygen uptake rate for K_,_ pneumoniae is 3 times that for P^ aeruginosa. 6. Specific growth rate for both pneumoniae and P . aeruginosa in suspended growth is described by saturation (Monod) kinetics. The maximum specific growth rate for K . pneumoniae is 5 times the maximum specific growth rate for R u aeruginosa. 124 7. Growth kinetics of F\_ aeruginosa in the biofilm have been described by the saturation model (Bakke, 1984). Growth kinetics of pneumoniae.in the biofilm follow an exponential model. 8. The ratio of specific biofilm cellular glucose uptake rate vs. specific biofilm cellular growth rate for K , pneumoniae is 3 times that for aeruginosa. REFERENCES 126 REFERENCES Atkinson, B and Davies, I .J . 1974. The overall rate of substrate uptake (reaction) by microbial films. Part I - A biological rate equation. Trans. Instn. Chem. Engrs. Vol 52: 248-259. Bakke, R., M.G. Trulear, J.A. Robinson and W.G. Characklis. 1984. Activity of Pseudomonas aeruginosa in Biofilms: Steady State. Biotech. Bioeng1 26': 1418-1424 . Bakke, R . 1986. Biofilm Detachment. Ph.D . thesis, Montana State University, Bozeman, Mt. Bakke, R . and P.Q. Ollsen. 1986. Biofilm thickness measurements by light microscopy. J . Microb. Met.5:1-6. Bakken, L.R. and R;A. Olsen. 1983. Buoyant densities and dry-matter contents of microorganisms: conversion of a measured biovolume into biomass. Appl. Environ. Microbiol. 45:1188-1195. Bevington, Ph. R . 1969. Data reduction and error analysis for the physical sciences. McGraw-Hill Book Company. BMDP Statistical Software. 1983. W.J. Dixon. (Ed.). University of California Press, Berkeley, CA. Brock, T .D . 1979. Biology of Microorganisms. Prentice-Hall. Inc11 Engelewood Cliffs, NJ. Brown, C . and Seidler, R.J. 1973. Potential pathogens in the environment: Kj, pneumoniae. a taxonomic and ecological enigma. Appl. Microbiol. June:900-904. Buchanan, R.E. and Gibbons, N.E. (Ed.'s). 1974. Sergey's manual of determinative bacteriology. The Williams and Wilkins Company, Baltimore, MD. Campbell, L.M., Michaels, G., Klein, R.D. and Roth, I .L . 1976. Isolation of ICj, pneumoniae from lake water. Can. J . Microbiol. 22:1762-1767. Characklis, W.G.., and K.E. Cooksey. 1983. Biofilms and microbial fouling. Adv. Appl1 Microbiol. 29:93-138. Characklis, W.G., Zelver, N . and Roe, F.L. 1986. Continuous on-line monitoring of microbial deposition on surfaces. In: Sixth international biodeterioration symposium. Barry, S . and Houghton, D.R. (Ed.’s). 127 Costerton. J.W., Marrie, T.J. and Cheng, K.-J. 1985. Phenomena in bacterial adhesion. 3-43. In: Bacterial adhesion. Savage, D.C. and Fletcher, M . (Ed.'s). Plenum Press, New York. Crisp, D . J . 1981. Overview of research on marine invertebrate larvae, 1940-1980. 103-126. In: Marine biodeterioration: an interdisciplinary study. Costlow, J .D . and Tipper, R ,C . (Ed.'s), Naval institute press. Annapolis, MD. Davies, D .G . and McFeters. G .A . 198.. Growth and comparative physiology of Klebsiella oxytoca attached to GAC particles and in liquid media. Microbial Ecology. Approved for publication. Dixon, W.J. and Massey, Jr., F .J . 1983. Introduction to statistical analysis . McGraw-Hill Book Company. Doetsch, R.N. and T.M. Cook. 1973. Introduction to bacteria and their ecobiology. University Park Press. Baltimore, MD. •1/ Erbing, C., Kenne, L., Lindberg, B . and Lonngren, J . 1976. Structural studies of the capsular polysaccharide from Klebsiella type I . Carbohydrate Research. 50:115-120. Escher, A. and Characklis, W.G. 1982. Algal-bacterial interactions within aggregates. Biotech. Bioeng. 24:2283- 2290 Escher, A .R . (1986) Bacterial Colonization of a Smooth Surface: An analysis with an Image Analyzer. Ph.D.-thesis, Montana State University, Bozeman, Mt. ’ Evans, C.G.T., Herbert, D . and Tempest, D.W. 1970. The continuous cultivation of micro-organisms. 2. Construction of a chemostat . 277-327. In.: Methods in microbiology. Vol. 2. Norris, J.R, and Ribbons, D.W. (Ed.'s). Frederickson, Av1G. 1977. Behavior of mixed cultures of microorganisms. Ann. Rev. Microbiol. 31:63-87. Geldreich, E .E . and E.W. Rice. 1987, Occurrence, significance and detection of Klebsiella in water systems. J AWWA. May=74-80. 1 2 8 Grady, C .P . 1982. Modeling of biological fixed films - A state of the art review. 344-404. In: Proceedings First International Conference on Fixed-Film Biological Processes. April 20-23. Kings Island. Ohio. Wu. Y .C ., Smith, E.D., Miller, R .D . and Patken, E.J.O. (Ed.'s). Haack, T .K . and McFeters, G .A . 1982. Nutritional relationships among microorganisms in an epilithic biofilm community. Microb. Ecol. 8:115-126. Herbert, D., Phipps, P.J. and Strange, R .E . 1971. Chemical analysis of microbial cells . 210-341. %n : Methods in microbiology, Vol. SB. Norris,J.R. and Ribbons, D .W . CEd.'s) . Hobbie, J .E ., R .J . Daley, and S . Jasper. 1977. Use of nuclepore filters for counting bacteria by fluorescence microscopy. Appl. Env. Microbiol. 35:858-862. Holmes, P .E . 1986. Bacterial enhancement of vinyl fouling by algae. Appl. Env. Microbiol. Dec.:1391-1393. Holt, H.G. 1977. The shorter Bergey7 s manual of determinative bacteriology. Eighth Edition. The Williams & Wilkins Company. Baltimore, MD. Jang, L .K . and T .F . Yen. 1985. A theoretical model of convective diffusion of motile and non-motiIe bacteria toward solid surfaces . 226-246. In.: Microbes and oil recovery, Vol. I. Zajic, J.E. and Donaldson, E .C . (Ed.'s). International bioresources journal. LOGIT. 1986. Logistic function data analysis program logit. Kirchhoff. W.H. Chemical Thermodynamics Division Center for Chemical Physics, National Bureau of Standards, Gaithersburg, MD 20899. Kissel, J 1C., P .L . McCarthy and R.L. Street. 1984. Numerical simulation of mixed-culture biofilm. J 1 Env. Eng. 110(2):393-411. Knittel , M . D . 1975 . Occurrence of K^ . pneumoniae in surface waters. Appl. Microbiol. May:595-597. Luedeking, R . and E.L. Piret. 1959. A kinetic study of the lactic acid fermentation. Batch process at controlled pH. J . of Biochem. and Microbiol. Technology and Engrg. I (4):393-412. ' . " 129 Marshall, K .C ., Stout, R . and Mitchell, R . 1971, Mechanism of the initial events in the sorption of marine bacteria to surfaces. J. Gen. Microbiol. 68:337-348. Matson, J.V. and Characklis. W.C. 1976. Diffusion into microbial aggregates. Water Res. 10: 877-885. Mian, F.A., Jarman, T.R. and Righelato, R.C. 1978. Biosynthesis of exopolysaccharide by Pseudomonas aeruginosa. J . Bacteriology. May: 418-422. Mitchell, R . and Kirchman, D . 1981. The microbial ecology of marine surfaces. 49-56. In: Marine biodeterioration: an interdisciplinary study. Costlow. J.D. and Tipper, R.C. (Ed.'s). Naval institute press. Annapolis, MD. iXMizushina, T . 1971. The electrochemical method in transport phenomena. Adv. Heat Transfer. 7:87-161. Monod, J . 1949. The growth of bacterial cultures. Annual Rev. of Microbiol. 3:371-394. MSUSTAT - Statistical analysis package. 1986. Montana State University, Bozeman, MT. Murray, R.E., Cooksey, K.E.. and Priscu, J.C. 1986. Stimulation of bacterial DNA synthesis by algal exudates in attached algal-bacterial consortia. Appl. Env1 Microbiol. Nov:1177-1182. X Neijssel. O.M. and Tempest, D.W. 1975. The regulation of carbohydrate metabolism in Klebsiella aerogenes NCTC418 organisms, growing in chemostat culture. Arch. Microbiol. 106:251-258. Rittmann, B.E. and P.L. McCarty. P.L. 1980-a. Model of steady-state biofilm kinetics. Biotech. Bioeng. 22:2343- 2357. Rittmann, B.E. and P.L. McCarty. 1980-b. Evaluation of steady-state biofilm kinetics. . Biotech. Bioeng. 22:2 359- 2373. Robinson, J.A., Trulear,. M.G. and Characklis . W.G. 1984. Cellular Reproduction and Extracellular Polymer Formation by Pseudomonas aeruginosa in Continuous Culture. Biotech. Bioeng. 26:1409-1417. Roth, I.L. 1977. Physical structure of surface carbohydrates. 5-26. In: Surface carbohydrates of the prokaryotic cell. Sutherland, I . (Ed.). Academic Press. 130 London. Sandbeck, K.A. and Ward, D.M. ■ 1981. Fate of immediate methane precursors in low-sulfate, hotspring algal- bacterial mats. Appl. Env. Microbiol. Mar:775-782. Schiefer, G.E. and Caldwell, D.E. 1982. Synergistic interaction between Anabaena and Zoogloea species in carbon dioxide-limited continuous culture. Appl1 Env. Microbiol. Jul:84-87. Slater, J.H. and Lovatt, D . 1984. Biodegradation and the significance of microbial communities, 439-485. In: Microbial degradation of organic compounds. Gibson, D.T. (Ed.). Dekker, New York, New York. Snedecor, G.W. and Cochran, W.G. 1967. Statistical methods. The Iowa State University Press, Ames, Iowa. Sutherland. I. 1977. Bacterial Exopolysaccharides - their nature and production. 27-96. In : Surface carbohydrates of the prokaryotic cell. Sutherland, I. (Ed.). Academic Press. London. Sutherland, I . 1982. Biosynthesis of microbial exopolysaccharides. 79-150. In.: Advances in Microbial Physiology. 23. Rose, A.H. and Morris, J.G. (Ed.'s). Topiwala, H.H. and Hamer, G . 1971. Effect of wall growth on steady-state 'continuous culture. Biotech. Bioeng.13:919- 922. Trulear, M.G. 1983. Cellular reproduction and extracellular polymer formation in the development of biofilms. Ph.D .- thesis, Montana State University, Bozeman, MT. Trulear, M.G. and W.G. Characklis. 1982. Dynamics of Biofilm Processes. J WPCF. 54 (9) :1288-1300. Turakhia, M.H. 1986. The Influence of Calcium on Biofilm Processes. Ph.D .-thesis. Montana State University, Bozeman, Mt . Vaituzis, Z . and. Doetsch, R.N. 1969. Motility tracks: Technique for quantitative study of bacterial movement. Appl. Microbiol. 17(4):584-588. Van Gemerden, H . 1974. Coexistence of organisms competing for the same substrate: an example among the purple sulfur bacteria. Microbial Ecology. 1^:104-119. 131 NOMENCLATURE 132 NOMENCLATURE substratum area [L- ]^ dilution rate [t- ]^ effectiveness factor for substrate diffusion [-] dissolved oxygen specific carbon transfer rate C f 1] growth—associated product formation coefficient [MpMm"1] non-growth associated product formation coefficient [MpM11-1V*1] half saturation concentration [MgL flux of dissolved . oxygen into reactor system specific cell detachment rate [t- ]^ biofilm specific oxygen uptake rate [MqMj1- t-^] specific rate of product detachment [t- ]^ specific rate of product formation in biofilm [MpMj1-It-I] specific rate of product formation in suspension [MpMm -1t-1] biofilm specific substrate uptake rate [MgMj1 ^t *] rate of oxygen generation [Mq .L .t x] concentration of oxygen in influent [Mq L concentration of oxygen in liquid phase [Mq L oxygen saturation concentration in influent [Mq L product carbon concentration in biofilm [MpL ^] product carbon concentration in influent [MpL product carbon concentration in liquid phase [MpL-3] adsorbed product carbon concentration in liquid phase [MpL-3] total product carbon concentration in liquid phase [MpL-3] 133 Sg^ substrate carbon concentration in biofilm [MgL-^] Sg^ substrate carbon concentration in influent [IigL- ]^ Sgj substrate carbon concentration in liquid phase [MgL-3] St o c I concentration of total organic carbon in liquid phase [MqL-3] 3SOCl concentration of soluble organic carbon in liquid phase [M(-.L-3] 3TOCf concentration of total organic carbon in biofilm [McL-3] t time [t] V Volume [L-3] Xpjf cell carbon concentration in film [MpjL 3] Xpjj cell carbon concentration in influent [M^L-3] Xjvjj cell carbon concentration in liquid phase [M^L-3] Ypjg yield of cell carbon from substrate carbon [MpjMg ]^ YpjQ yield of cell carbon from oxygen [MjvjMq *] Y0J33 observed yield coefficient [MpjMg ]^ Ypg yield of product carbon from substrate carbon [MpjMp-1 ] Ypgj Yield of soluble product carbon from substrate carbon [MpjMp ]^ Ypst Yield of total product carbon from substrate carbon [ MpjM p ] YpQ yield of product carbon from oxygen [MpjMp--*"] YgQ glucose-oxygen stoichiometric ratio [MgMQ- ]^ 134 Greek letters 6 ^ biofilm thickness [L] € effectiveness factor [-] Tjvjf cell carbon density in biofilm [MjvjL- ]^ Ag specific cellular glucose uptake ratio [MgMjvj-"*"] H specific cellular growth rate [t-"*"] |am maximum specific cellular growth rate [t- ]^ subscripts a adsorption, attachment d desorption, detachment i influent f biofilm phase I liquid phase s saturation M Microbial cell 0 Oxygen P Product S Substrate 135 APPENDICES 136 Objective of this study was the determination of the mass transfer coefficient of dissolved oxygen (DO) for the Rototorques. Diffusion of dissolved oxygen into the reactors can be determined by pumping dilution water of low dissolved oxygen concentration into the sterile reactors and monitoring the (DO) concentration in the effluent. The rate at which oxygen diffuses into the reactors is determined by a material balance across the reactor. Dissolved oxygen concentration in the influent was varied by purging the dilution water mixing tank with nitrogen gas. Varying the nitrogen gas flow rate for a constant flow rate of dilution water (0.06*10 ^m^.min "*".) resulted in different concentrations of dissolved oxygen entering the reactors. The reactors were pretreated as described in Experimental Approach. Rotational speed was 200 rpm. After changing the nitrogen gas flow, dissolved oxygen concentration in the effluent was monitored continuously to determine steady state DO concentration. DO concentration was measured with a YSI model-54 oxygen meter. A DO mass balance (Eg. 7 without DO consumption terms) was used to quantify the oxygen transfer rate in the reactor: dSoi A ---- = D . (Sq -^-Sq 1) + Nq .- • (A-I) dt V where: Sq 1 = dissolved oxygen concentration in effluent [Mq L Sq 1 = dissolved oxygen concentration in influent [MqL D = dilution rate [t Nq =flux of dissolved oxygen into the system [Mq L ^t "*■] APPENDIX A: Oxygen Diffusion Study 137 The dissolved oxygen flux into the reactor is defined by: V N0 = kc--- (A-2) A where: kc = DO specific mass transfer rate [t--*-] Sq s = DO saturation concentration [Mq L Substituting Eg. A-2 into Eg. A-I and assuming steady state conditions results in: 5OI-sOs The specific DO mass transfer rate for the experimental reactors was determined from 14 independent measurements (Table 16). The means and standard deviations are: kcl = 9.30 ± 1.06 h-1, kc2 = 9.44 ± 0.95 h-1. Biofilm experiments were designed such that dilution water was DO saturated in a mixing chamber prior to entering the reactor. Therefore, mass transfer of DO into the reactors could be accounted for in the DO material balance as follows: dS0 A ---- = d 8^Os-sOI5 + N0 - dt V A V - xMl YM0 + — YP0 (7) 138 Influent Effluent Reactor I Effluent Reactor 2 temp . DO temp DO kcl temp DO W O to (0C) iQ 3 I 00 V (0C) (g.m-3) (h-1) (0C) (g.m-3) (h-1) 20.8 .75 20.5 4 .70 9.77 20.3 4.75 10.07 20.7 .46 20.5 4.35 8.55 20.3 4.45 9.06 20.7 .40 20.4 4.65 10.32 20.3 4.68 10.51 20.8 .41 20.5 4.73 10.80 20.3 4.62 10.12 20.5 .42 20.5 4.65 10.28 20.3 4.55 9.69 20.5 .45 20.3 4.70 10.51 20.3 4.80 11.15 20.5 1.65 20.3 4.75 7.80 20.3 4.80 8.08 20.5 2.05 20.3 5.08 8.67 20.5 5.13 9.00 20.7 1.50 20.5 4.95 9.37 20.4 4.85 8.75 20.7 1.35 20.5 4.92 9.58 20.3 4.95 9.77 20.7 2.62 20.5 5.35 8.79 20.4 5.30 8.43 20.5 4.50 20.4 6.10 7.91 20.3 6.10 7.91 20.5 6.44 20.3 7.00 7.75 20.3 7.05 9.38 20.2 6.71 20.5 7.18 10.17 20.5 7.18 10.17 Table 16. Experimental data from oxygen diffusion study. With the time constant for microbial oxygen consumption greater than for the oxygen diffusion, the value for the specific oxygen mass transfer rate, derived for steady state conditions, can be used to describe oxygen diffusion in non-steady state conditions. This results in: dSo dt (D+kc) .(S0g-S01) (A-4) 139 APPENDIX B: Characteristics of the Rototorcrue Inner cylinder: - wetted height 177 mm - diameter 102 mm - wetted surface area (vertical) 56700 rnm^ - wetted surface area (horizontal) 16300 — total inner cylinder wetted surface area 73100 mm2 Draft tubes: - number 4 - diameter 10 mm - length 180 mm - angle of inclination 80 O — surface area .23100 mm2 Outer cylinder: - wetted height 220 mm - diameter 113 mm - wetted surface area (vertical) 78100 mm2 - wetted surface area (horizontal) 20100 mir/ - total outer cylinder wetted surface area 98200 mm ^ — number of slides 12 Slides: - number . 12 - thickness 0.5 mm - width 15 mm - wetted length 220 mm - wetted slide surface area 3300 mrrc - volume of I slide 1650 mm ^ - volume of 12 slides 20000 mm'3 Reactor: - wetted surface area (excl. tubes) 0.17 m2 - slide surface area (excl. tubes) 42 - total wetted surface area (incl . tubes) 0.19 nr*1 — slide surface area (incl. tubes) 37 %_ - liquid volume, (at 200 rpm) 0.52*10-3 m3_, - surface/volume ratio (excl. tubes) 0.29 mm ^ - surface/volume ratio (incl. tubes) 0.33 mm ^ - liquid residence time (at I m l .s dilution water flow rate, at 200 rpm) 594 s (9.9 min) - recycle flow rate^ (at 200 rpm) 500 mm .s - width of annular gap 5 .5 mm Flow rate through draft tubes as a result of inner cylinder rotation. I 140 APPENDIX C : Raw Data Batch Experiments Oxygen consumption Time Experiment I II H I (h) dissolved oxygen concentration Cg.m 3) .0 . .0 .0 .0 1.0 .0 .0 .0 2.0 .0 .0 .0 3.0 .0 .0 .0 4.0 .0 .0 .0 5.0 .0 .0 .0 6.0 .0 .0 .0 7.0 .0 .0 .0 8.0 1.3 .0 .6 9.0 2.2 .8 .6 10.0 2.9 2.3 1.3 11.0 5.1 6.3 2.7 12.0 10.0 12.2 5.8 13.0 15.8 16.2 10.9 13.9 18.1 18.4 15.5 14.0 18.7 18.4 15.5 14 .2 18.7 18.7 16.1 14.5 19.5 19.4 17.4 14.9, 21.1 20.7 18.6 15.0 21.1 20.7 18.8 15.3 21.8 21.3 19.4 15.5 21.8 22.0 20.0 15.8 22.5 22.8 20.7 16.0 23.1 . 23.5 21.4 16.3 23.7 24.2 22.0 16.5 24.6 25.0 22.7 16.8 25.3 25.8 23.3 17.0 25.8 26.5 23.9 17.3 26.5 27.2 24.6 17.5 27.2 27.2 25.2 17.8 27.8 28.0 25.9 18.0 ■ 28.5 28.6 26.4 18.3 29.1 28.6 27.1 18.5 29.8 28.6 27.7 18.8 30.3 29.2 28.4 141 (continued) Time Experiment I II III (h) dissolved oxygen concentration (g.m**3) 19.0 31.0 29.2 28.9 19.3 31.4 29.2 29.5 19.5 31.5 29.8 30.1 19.8 32.2 29.8 30.7 20.0 32.2 29.8 31.3 20.3 32.2 30.5 31.9 20.5 32.0 30.5 31.9 20.8 32.8 30.5 31.9 21.0 33.4 30.5 32.5 22.0 34.0 . 31.2 33.1 23.0 34.7 32.6 34.3 24.0 35.4 33.3 34.8 25.0 36.2 34.1 35.4 26.0 36.8 34.8 36.0 27.0 37.4 35.4 36.6 28.0 37.4 35.4 37.2 29.0 38.2 36.1 37.8 30.0 38.2 36.7 37.8 31.0 38.8 37.3 38.4 32.0 39.6 38.0 39.0 33.0 39.6 38.0 39.0 34.0 40.3 38.7 39.7 35.0 41.1 39.4 40.3 35.7 41.1 40.0 40.9 36.0 41.7 40.8 40.9 36.5 41 .7 40.8 41.6 37.0 41.7 40.8 41.6 37.2 41.7 40.8 41.6 37.7 41,7 40.8 41.6 38.0 ' ■ 41.7 40.8 41.6 38.5 41.7 40.8 41.6 39.0 41.7 40.8 41.6 39.5 41.7 40.8 41.6 40.0 41.7 40.8 41.6 41.0 41.7 40.8 41.6 42.0 41 .7 40.8 41.6 43.0 41.7 40.8 41.6 44.0 41 .7 . 40.8 41.6 45.0 . 41.7 40.8 41.6 142 Results of analysis Reactor # 2Si . 3Sl _ . . ' (a 5TOCi C.rrT3') .. sTOCl sSOCl I 38.5 0 39.1 • 21.5 ■ 10.7 II 38.5 0 39.1 23.6 11.3 III 38.5 0 39.1 21.4 9.0 MEANS 38.5 0 39.1 22.2 10.3 STD .0 .0 .0 1.2 1.2 Reactor Cell# Length Breadth Roundness # >% COIE# (tim) Cum) C-) I 9.331el3 1.20 .85 1500 II 9.453el3 1.25 .86 1489 III 8.037el3 1,32 .93 1485 MEANS 8.940el3 1.26 .88 1491 STD 7.845el2 .06 .04 8 Reactor Oxygen consumed Mass # 4 3 I W (g.m-3) I 42.7 31.5 II 42.3 31.1 III 42.6 38.8 MEANS 42.6 33.8 STD .2 4.3 143 APPENDIX D: Raw Data Chemostat Experiments Dilution rate Cl .h"1) 5Si sSl .... (a C 5TOCi rtT3') . . sTOCl 5SOCl .10 37.4 .32 38.1 17.4 9.4 .13 37.3 .29 38.1 19.9 9.9 .2.7 37.1 .66 39.3 23.0 9.2 .27 37.1 .35 39.3 22.4 8.9 .40 37.0 . .60 . 38.4 23.5 9.5 .40 37.0 .17 38.4 23.8 9.5 .50 36.5 .54 37.9 23.8 10.9 .50 36.5 1.28 37.9 24.6 12.3 .56 35.4 .23. 37.8 24 .2 11.6 .56 35.4 .61 37.8 24.6 11.9 .67 34.5 .23 37.2 26.2 11.4 .67 34.5 .04 37.2 23.9 11.3 .80 35.9 .41 36.5 25.0 10.5 .80 35.9 .41 36.5 22.1 10.0 1.07 37.1 .79 36.9 22.0 10.9 1.07 37.1 1.04 36.9 22.9 10.0 1.60 36.7 6.05 39.4 25.9 14.5 1.87 36.7 17.79 39.4 31.6 21.3 Dilution rate Cl . h 1J Mass Cg .m-3) Cell # C# .m~3) Length Cum) Breadth Cum) Roundness C-) .10 47.0 I .047el4 1.11 .75 1684 . 13 76.5 I .276el4 1.22 .76 1745 .27 50.0 I .176el4 1.43 .71 2155 .27 31.5 9.732el3 1.24 .76 2053 .40 30.5 I .023el4 1.51 .73 2121 .40 36.0 9,615 e 13 1.65 .77 2068 .50 24.5 7.145el3 1.54 .78 2043 .50 29.5 5.498el3 1.91 .89 1995 .56 53.0 4.852el3 2.11 . 84 2186 .56 30.5 6.91Oel3 2.28 .90 1834 .67 59.0 3.659el3 2.45 .89 2200 .67 54.5 5.293el3 2.27 .89 1945 .80 61.0 3.661el3 1.96 .82 2211 .80 50.0 5.072el3 2.24 .86 2114 1.07 57.5 9.2b3el3 1.45 .75 1934 1.07 50.5 6.52 8 e13 1.98 .83 2073 I .60 37.5 5.44 Oel3 1.47 .79 1993 1.87 23.5 2.382el 3' I .09 .73 2101 144 APPENDIX E: Comparison of Nutrient Compositions ,Evans et al . Cl970) This study original based on 10 g.C m~3 Carbon (g C.m 3000 10 10 Phosphorus 310 0.10 92.40 Nitrogen (g.m- )^ 1400 0.47 1.88 Sulphur (g.m- )^ 64 0.02 0.31 Magnesium (g.m- )^ 30.4 0.01 0.20 Molybdenum (g.m- )^ 0.0096 3.2e-6 4.7e-4 Iron Cg.m-3) 5.59 0.002 0.023 Cobalt (g.m-3) 0.59 0.0002 v — Copper (g.m-3) 0.32 0.0001 0.0005 Manganese (g.m-3) 2.78 0.0009 0.0026 Zinc (g.m-3) 0.33 0.0001 0.0228 Boron (g.m-3) 0.043 0.00001 0.00011 145 APPENDIX F: Sample Preparation Procedure Transmission Electron Microscopy (TEM) 1. Fix sample with 2;5% (v/v) glutaraldehyde in millonig's phosphate buffer (pH 7.2) for 6-8 hours (or overnight). 2. If necessary, store at 4°C in millonig's phosphate buffer (pH 7.2) till sample preparation - change weekly. 3. Wash 3 times in millonig's phosphate buffer (pH 7.2) for 20 minutes each. 4. Postfix in 1% (w/v) osmium tetroxide (OsO^) for I hour 5. Wash 3 times in millonig's phosphate buffer (pH 7.2) for 20 minutes each. 6. Dehydrate in ethanol: 15 minutes in 50% ethanol - 15 minutes in 70% ethanol; sample may be stored in 70% ethanol overnight if needed - 15 minutes in 95% ethanol - 3 times for 15 minutes each in 100% ethanol 7. Embed in Spurr's (1969)1 epoxy resin: - I hour in. mixture of (v/v) 2 (100% ethanol) : I (epoxy resin) - I hour in mixture of (v/v) I (100% ethanol) : I (epoxy resin) - 1 8 hours or overnight in pure epoxy resin. 8. Polymerize in 70°C oven for 14 hours in a capsule. Spurr, A.R. 1969. Low-viscosity epoxy resin embedding medium for electron microscopy. J . of Ultra Structure Research. 25:31-43. I 146 Scanning Electron Microscopy (SEMI 1. Fix with 2.5% (v/v) glutaraldehyde in millonig's phosphate buffer (pH 7.2) for 6-8 hours (or overnight). 2. If. necessary, store at 4° C in millonig"s phosphate buffer (pH 7.2) till sample preparation - change weekly. 3. Wash 3 times in phosphate buffer (pH 7.2) for 20 minutes each. 4. Dehydrate in ethanol: - 15 minutes in 30% ethanol - 15 minutes in 50% ethanol - 15 minutes in 70% ethanol - 15 minutes in 90% ethanol - twice for 10 minutes each in 100 ethanol. 6. Critical point dry. 7. Mount on sample holder with colloidal graphite. 8. Sputter coat with gold (or other good electron conductor). 147 APPENDIX G: Raw Data and Parameters Locristic Equation Mono Population Biofilm - K. pneumoniae Notation DO = Dissolved Oxygen concentration [g.m- ]^ glue = glucose carbon concentration [g C.m- ]^ TOC .= Total Organic Carbon [g C.m~^] SOC = Soluble Organic Carbon [g C.m^^] mass = mass of suspended solids [g.m~^] Kp = Kj, pneumoniae cell concentration [#.m~^] Pa = P^ aeruginosa cell concentration [#.m~^] A = pretransition region intercept B = posttransition region asymptote Xq = position of the transition region D0 = width of the transition region StDev = Standard deviation of overall fit 148 Mono population biofilm - K. pneumoniae Experiment I - licruid phase Raw data time DO glue TOC SOC mass Kp 'Pa Ch) Cg1Itf3) . . . . Cg C.rrf3) ..........) - 3 for $ i. I 2. for a predominantly diffusion-controlled region (K3): I $ tanh (kg .6 f) k2 .6f D .- — Itanh ($) — T or 5» 2 I where: $ = Thiele modulus [-], defined by: $ (kg6f) Ck3Sgi) -^ 2 ( I "^kgSg1 ) [k3SSl_ln a ^k3Sg1TJ _ % (K4) with kg6f 1 obs dimensionless thickness [-] % (K5)V M f KsDg obs effective diffusion coefficient within microbial mass [L^t observed yield coefficient [M^Mg defined as: I 1 , kPg I+ kpn ^obs YMS YPS Pm YPS (K6) 1 7 8 Ypg = yield of product carbon from substrate carbon [MmMp-1] kpg = growth associated product formation coefficient [MpMm"1] kpn = non-growth associated product formation coefficient [MpMjyj-1t_1] kgSgi = dimensionless substrate concentration [-] k3 = Ks"1 [L3Mg-1] Substituting Eg. K6 into Eg. K5 results in V M f 1 kPg 1 kPn+ + KsD* YMS YPS YPS H CK7) Time progression of € is determined for smooth biofilms of Kju pneumoniae and aeruginosa of egual cellular carbon density and of 10 |um (Kp) and 15 and 30 pm (Kp and Pa) thickness. Table 17 presents the parameter values used for the simulation. ■! 179 Table 17. Parameter values for determination of diffusion I effectiveness factor. K . pneumoniae P . aeruginosa tarn [h 1] Sgi ] [g C.m 3] Ym s ; [g.g-1] Yps [g.g ] Kg [g C.m-3] kPg [g.g-1] kPn [g.g-1h-1] kg [g C.m 3] rMf ^g C . m-3] 2.0 10 - 1.5 0.08 0.48 1.43 1.72 0.20 0.70 10.0 0.4 10 - 1.5 0.24 0.41 2.0 0.36 0.10 . 0.50 10.0 The effective diffusion coefficient of glucose into a mixed population biomass is taken from Matson and Characklis (1976) ('De = 0.6*10-9 m2 .s-1). Results of the simulation (Figure Kl) indicate that, under identical conditions, both biofilms are diffusion- limited. But the extent of the limitation is more severe for Ki. pneumoniae than for P i. aeruginosa (Table 18) . Table 18. Relation between steady state biofilm thickness and diffusion effectiveness factor € for Ki. pneumoniae and P . aeruginosa. thickness (jam) , K . pneumoniae P . aeruginosa 10 64% • — 15 45% 92% 30 23% 78% Figure 52. Time progression of the diffusion effectiveness factor for various thicknesses of the biofilm of K . pneumoniae and aeruginosa. 180 EFFECTIVENESS FACTOR FOR DIFFUSION P. aeruginosa (Pa) & K. pneumoniae (Kp) Pa SO (78%) < . .6 - Kp 30 (23%) rel. thlekm TIME (h) MONTANA STATE UNIVERSITY LIBRARIES