Biomass and porosity profiles in microbial granules used for aerobic wastewater treatment

AIMS: To obtain biomass and porosity profiles for aerobically grown granules of different diameters and to determine a suitable range of granule diameters for application in wastewater treatment. METHODS AND RESULTS: Microbial granules were cultivated in an aerobic granulated sludge reactor with model wastewaters containing acetate, or ethanol plus acetate, or glucose as the main carbon source. Granules were formed by retaining microbial aggregates using a settling time of 2 min. Sampled granules had diameters ranging from 0.45 to 3 mm. Microbial biomass in the granules was detected with the nucleic acid stain SYTO 9 and confocal laser scanning microscopy. The thickness of the microbial biomass layer was proportional to the granule diameter, and had a maximum value of 0.8 mm. The thickness of the microbial biomass layer correlated with the penetration depth of 0.1 microm fluorescent beads into the granule. CONCLUSIONS: The microbial biomass and porosity studies suggest that aerobically grown microbial granules should have diameters less than a critical diameter of 0.5 mm, if deployed for wastewater treatment applications. This critical diameter is based on the assumption that whole granules should have a porous biomass-filled matrix. SIGNIFICANCE AND IMPACT OF THE STUDY: This work could contribute to the development of aerobic granulation technology for effective biological wastewater treatment.     

The effect of hydraulic retention time on the stability of aerobically grown microbial granules

AIMS: The aim of this study is to evaluate the effect of hydraulic retention time (HRT) on the development of aerobically grown microbial granules. METHODS AND RESULTS: Five column-shaped sequential aerobic sludge blanket reactors (SASBRs) were seeded with aerobically grown microbial granules and operated in a cyclic mode at different HRTs. At the shortest HRT of 1 h, the strong hydraulic pressure triggered biomass washout and led to reactor failure. At the longest HRT of 24 h, which represented the weakest hydraulic selection in this study, aerobic granules were gradually substituted by bioflocs because of the lower frequency of volumetric exchange. Within the optimum range of HRTs from 2 to 12 h, however, aerobic granules became stabilized in the presence of adequate hydraulic selection in the reactors, with good mixed liquor volatile suspended solids (MLVSS) retention, high volumetric chemical oxygen demand (COD) removal, low sludge volume index (SVI) values, good effluent quality, low sludge production rate, stronger and more compact structures, high cell hydrophobicity and high ratios of extracellular polysaccharides (PS) to extracellular proteins (PN). CONCLUSIONS: HRTs between 2 and 12 h provided the hydraulic selection pressures favourable for the formation and maintenance of stable aerobic granules with good settleability and activity. SIGNIFICANCE AND IMPACT OF THE STUDY: This is the first systematic study on the effect of HRT on heterotrophic aerobic granules. The results of the investigation are useful in understanding how aerobic granules can be applied for wastewater treatment.     

Specific layers in aerobically grown microbial granules

AIMS: To determine the optimal size of aerobically grown granules for wastewater treatment by measuring specific layers within the granules. METHODS AND RESULTS: A variety of biological layers were detected by oligonucleotide probes, specific fluorochromes, and fluorescent microspheres. The channels in the granule matrix penetrated to depths of 900 microm. A layer of obligate anaerobic bacteria was detected at a depth of 800 microm below the granule surface. Dead cells were also observed in the granule interior. CONCLUSIONS: Aerobically grown granules contained layers of aerobic and anaerobic micro-organisms. SIGNIFICANCE AND IMPACT OF THE STUDY: The optimal diameter of the aerobic granule is less than 1600 microm. This is twice the distance from the granule surface to the anaerobic layer. This approach can be used to optimize the thickness of other microbial aggregates such as flocs, colonies and biofilms.     

Starter culture of <Emphasis Type="Italic">Pseudomonas veronii</Emphasis> strain B for aerobic granulation

Aggregation of bacterial cells is used in formation of microbial granules. Aerobically grown microbial granules can be used as the bio-agents in the treatment of wastewater. However, there are problems with start up of microbial granulation and biosafety of this process. Aim of this research was selection and testing of safe microbial strain with high cell aggregation ability to shorten period of microbial granules formation. Five bacterial strains with cell aggregation index higher than 50% have been isolated from the granules. Strain of Pseudomonas veronii species was considered as most probably safe starter culture for granulation because other strains belonged to the species known as human pathogens. The microbial granules were formed after 3 days of cultivation in case when P. veronii strain B was applied to start-up aerobic granulation process using model wastewater. The granules were produced from activated sludge after 9 days of cultivation. Microbial aggregates produced from starter culture of P. veronii strain B were more compact (sludge volume index was 70 ml/g) than those produced from activated sludge (sludge volume index was 106 ml/g). It is a first proof that application of selected safe starter pure culture with high cell aggregation ability can accelerate and enhance formation of microbial granules.     

固定化微生物处理模拟污染地表水

以聚乙烯醇和海藻酸钠为包埋剂、驯化后的活性污泥为包埋菌剂,制备固定化微生物颗粒,其中包埋剂与包埋菌剂的比例为2:1。将该固定化微生物颗粒按20%的填充率装填到自制反应器中,用于处理模拟污染地表水,研究该固定化微生物的性能特点及其对模拟污染地表水的净化效果。结果表明:固定化微生物反应器的最佳水力停留时间为10h,最佳进水COD负荷为1.15～1.85g·L-1·d-1。在水温为20～29℃、溶解氧为3～4mg·L-1、水力停留时间为10h的条件下,当进水COD浓度为70.58～91.76mg·L-1、铵氮浓度为13.68～17.82mg·L-1时,COD去除率>62.3%,铵氮去除率>90.6%,表明固定化微生物能够有效地去除污染地表水中的COD和铵氮。     

Presence of anaerobic bacteroides in aerobically grown microbial granules

Microbial granules were grown in a column-type sequential aerobic sludge blanket reactor inoculated with activated sludge flocs taken from a wastewater treatment plant and containing a medium with glucose as the main carbon source. The reactor selected for granules that could settle rapidly by employing a short settling time of 2 min. Matured granules with diameters between 2 and 3 mm were examined for anaerobic bacteria as their presence can signal the onset of diffusion limitation problems that can potentially diminish granule stability due to the bacterial production of fermentation gases and organic acids under anaerobic conditions. To detect the anaerobes in the granules, clones were constructed from 16S rRNA PCR amplicons. Two sequence types associated with a strict anaerobe Bacteroides spp. were identified from these clones. Fluorescence in situ hybridization (FISH) followed by confocal laser scanning microscopy (CLSM) demonstrated that cells of Bacteroides spp. were concentrated at a depth of approximately 800 mm below the surface of the granule. Cell enumeration using flow cytometry showed that the percentage of labeled cells of Bacteroides spp. compared to total bacterial cells in the granules was 0.56%. This is the first study to use a suite of culture-independent techniques to report the presence of a defined species of anaerobic bacteria in aerobically grown microbial granules.     

Influence of phenol on cultures of acetate-fed aerobic granular sludge

AIMS: This paper attempts to investigate the inhibition of phenol on the acetate utilization in acetate-fed aerobic granular sludge culture. METHODS AND RESULTS: Acetate-fed aerobic granules with a mean diameter of 1.0 mm were predeveloped in a column sequencing aerobic sludge blanket reactor. The present study looked into the utilization kinetics of acetate by acetate-fed aerobic granules in the presence of different phenol concentrations ranging from 0 mg l(-1) to 50 mg l(-1). For this purpose, batch experiments were conducted at 25 degrees C, while the initial biomass and acetate concentrations were in a range of 109-186 mg mixed liquor suspended solids (MLSS) l(-1) and 185-300 mg acetate-chemical oxygen demand (COD) l(-1). Results showed that the utilization of acetate in the presence of phenol was subject to a zero-order reaction kinetics. The relative phenol concentration in terms of the ratio of initial phenol concentration (C(p)) to initial biomass concentration (X(0)) was used to describe the real inhibitory strength of phenol imposed on acetate-fed aerobic granules. When the C(p)/X(0) ratio increased from 0 to 0.19 mg phenol mg(-1) MLSS, the zero-order reaction rate constant of acetate dropped from 1.15 mg l(-1) min(-1) to 0.38 mg l(-1) min(-1), and a similar trend was also observed in specific oxygen utilization rate. As compared to the control test without addition of phenol, the acetate-COD removal efficiency was reduced by nearly 50% at a C(p)/X(0) value of 0.19 mg phenol mg(-1) MLSS. It was found that biodegradation of phenol was negligible in acetate-fed aerobic granular sludge batch culture. CONCLUSIONS: It appears that phenol can seriously repress the utilization of acetate in the acetate-fed aerobic granular sludge batch cultures. A simple zero-order reaction model could adequately describe the utilization of acetate by acetate-fed aerobic granules in the presence of phenol. SIGNIFICANCE AND IMPACT OF THE STUDY: It is expected that this study would lead to a better understanding of the behaviour of acetate-fed aerobic granules in the presence of inhibitory organic compounds.     

Specific aerobic granules can be developed in a completely mixed tank reactor by bioaugmentation using micro-mycelial pellets of Phanerochaete chrysosporium

Aerobic granules were firstly developed in a completely mixed tank reactor (CMTR) by seeding micro-mycelial pellets (MMPs) of Phanerochaete chrysosporium. During phenol wastewater treatment, sludge granulation rate reached 67 % after 15-day operation. The granules in CMTR are different from aerobic granules described in literature in morphology, and a majority of them are rod-shaped or rodlike sludge besides spherical granules. The polymorphic granules, having no essential difference with aerobic granules previously reported, achieve advantages over conventional activated sludge in settling ability, biomass concentration, density, integrity coefficient and removal ability to phenol wastewater. The optimized parameters for sludge granulation in CMTR including temperature, inoculum quantity, rotary speed and superficial air upflow velocity are 30 °C, 5–7 g/l, 150 rpm, and 0.5 cm/s, respectively. Analysis on sludge granulation mechanism indicates that MMPs not only result in the formation of aerobic granules containing MMPs as nuclei, but also induce the formation of biogranules which do not have MMP at their cores. The work challenges the general belief that the homogenous circular flow pattern of microbial aggregates is necessary for aerobic sludge granulation.     

Biodegradation of high phenol containing synthetic wastewater by an aerobic fixed bed reactor

The continuous aerobic degradation of phenol, mixed with readily degradable synthetic wastewater was studied over a period of 400 days at 25+/-5 degrees C temperature in a fixed bed biofilm reactor using 'Liapor' clay beads as packing material. The phenol concentration added to the reactor ranged from 0.19 to 5.17g/l and was achieved by a gradual increase of phenol in wastewater, thus adapting the microbial flora to high contaminant concentrations. A maximal removal rate of 2.92g phenol/(ld) at a hydraulic retention time (HRT) of 0.95 days and a total organic loading rate (OLR) of 15.3g COD/(ld) with a phenol concentration of 4.9g/l was observed. However, this was not a stable rate at such high phenol loading. At the end of reactor operation on day 405, the phenol removal rate was 2.3g/(ld) at a influent phenol concentration of 4.9g/l. There were no phenol intermediates present in the reactor, as evident from corresponding COD, phenol removal and the absence of fatty acids. Omission of organic nitrogen compounds or of urea in influent feed was not favourable for optimal phenol removal. The phenol degradation profile that was studied in shake flasks indicated that the presence of a acetate which represent as an intermediate of phenol degradation retarded the phenol degradation. The highest phenol degradation rate observed in batch assays was 3.54g/(ld).     

Optimization of high-molecular-weight polycyclic aromatic hydrocarbons' degradation in a two-liquid-phase bioreactor

A microbial consortium degrading the high-molecular-weight polycyclic aromatic hydrocarbons (HMW PAHs) pyrene, chrysene, benzo[a]pyrene and perylene in a two-liquid-phase reactor was studied. The highest PAH-degrading activity was observed with silicone oil as the water-immiscible phase; 2,2,4,4,6,8, 8-heptamethylnonane, paraffin oil, hexadecane and corn oil were much less, or not efficient in improving PAH degradation by the consortium. Addition of surfactants (Triton X-100, Witconol SN70, Brij 35 and rhamnolipids) or Inipol EAP22 did not promote PAH biodegradation. Rhamnolipids had an inhibitory effect. Addition of salicylate, benzoate, 1-hydroxy-2-naphtoic acid or catechol did not increase the PAH-degrading activity of the consortium, but the addition of low-molecular-weight (LMW) PAHs such as naphthalene and phenanthrene did. In these conditions, the degradation rates were 27 mg l-1 d-1 for pyrene, 8.9 mg l-1 d-1 for chrysene, 1.8 mg l-1 d-1 for benzo[a]pyrene and 0.37 mg l-1 d-1 for perylene. Micro-organisms from the interface were slightly more effective in degrading PAHs than those from the aqueous phase.     

Properties of phenol-removal aerobic granules during normal operation and shock loading

The physical structure and activity of aerobic granules, and the succession of bacterial community within aerobic granules under constant operational conditions and shock loading were investigated in one sequencing batch reactor over ten months. While the maturation phase of the granulation process began on day 30, the structure of microbial community changed markedly until after three months of reactor operation under constant conditions with a loading rate of 1.5 g phenol L⁻¹ day⁻¹. A shock loading of 6.0 g phenol L⁻¹ day⁻¹ from days 182–192 led to divergence of bacterial community, an inhibition of the biomass activity, and a decrease in phenol removal rate in the reactor. However, phenol was still completely removed under this disturbance. After the shock loading, the mean sizes of aerobic granules increased, and the activity of the microbial population within the granules decreased, although there appeared highly resilient for the dominant bacterial community of aerobic granules which mainly included β-Proteobacteria. Correlation analysis suggested that biomass concentration and biomass loading were significantly related to the community composition of aerobic granules during the whole operational period. The development of a relatively stable bacterial community in aerobic granules implied that those distinct dominant microbes in aerobic granules were favorably selected and proliferated under the operational conditions.     

Microscopic observation of aerobic granulation in sequential aerobic sludge blanket reactor

AIMS: This paper attempts to provide visual evidence of how aerobic granulation evolves in sequential aerobic sludge blanket reactors. METHODS AND RESULTS: A series of experiments were conducted in two column-type sequential aerobic sludge reactors fed with glucose and acetate as sole carbon source, respectively. The evolution of aerobic granulation was monitored using image analysis and optical and scanning electron microscopy. The results indicated that the formation of aerobic granules was a gradual process from seed sludge to compact aggregates, further to granular sludge and finally to mature granules with the sequential operation proceeding. Glucose- and acetate-fed granules have comparable characteristics in terms of settling velocity, size, shape, biomass density and microbial activity. However, the microbial diversity of the granules was associated with the carbon source supplied. In this work, an important aerobic starvation phase was identified during sequential operation cycles. It was found that periodical aerobic starvation was an effective trigger for microbial aggregation in the reactor and further strengthened cell-cell interaction to form dense aggregates, which was an essential step of granulation. The periodical starvation-induced aggregates would finally be shaped to granules by hydrodynamic shear and flow. CONCLUSION: Aerobic granules can be formed within 3 weeks in the systems. The periodical starvation and hydrodynamic conditions would play a crucial role in the granulation process. SIGNIFICANCE AND IMPACT OF THE STUDY: Aerobic granules have excellent physical characteristics as compared with conventional activated sludge flocs. This research could be helpful for the development of an aerobic granule-based novel type of reactor for handling high strength organic wastewater.     

Degrading high-strength phenol using aerobic granular sludge

Aerobic granules were adopted to degrade high-strength phenol wastewater in batch experiments. The acclimated granules effectively degraded phenol at a concentration of up to 5,000 mg l⁻¹ without severe inhibitory effects. The biodegradation of phenol by activated sludge was inhibited at phenol concentrations >3,000 mg l⁻¹. The granules were composed of cells embedded in a compact extracellular matrix. After acid or alkaline pretreatment, the granules continued to degrade phenol at an acceptable rate. The polymerase chain reaction-denaturing gradient gel electrophoresis technique was employed to monitor the microbial communities of the activated sludge and the aerobic granules following their being used to treat high concentrations of phenol in batch tests.     

Effect of the nitrogen source on caffeine degradation by Aspergillus tamarii

AIMS: To evaluate caffeine degradation and nitrogen requirements during Aspergillus tamarii growth in submerged culture. METHODS AND RESULTS: Aspergillus tamarii spores produced on a coffee infusion agar medium added with sucrose were used. Several caffeine and ammonium sulphate concentrations (0-1 and 0-1.36 g l-1, respectively) were tested simultaneously on fungal biomass production and caffeine degradation. An additional caffeine pulse (4 g l-1) was added for all experiments after 48 h of fermentation. Results revealed that when using 0.90 g l-1 of caffeine and 0.14 g l-1 of ammonium sulphate, biomass production and caffeine degradation were enhanced. Highest biomass production (Xmax = 9.87 g l-1) with a specific growth rate (micro) of 0.073 h-1 and caffeine degradation rate of 0.033 g l-1 h-1, was observed under these conditions. CONCLUSIONS: Caffeine degradation as well as biomass production were characterized. SIGNIFICANCE AND IMPACT OF THE STUDY: These studies set the stage for future characterization studies of intracellular enzymes involved in caffeine degradation. Moreover, results observed may help in the biotreatment of residues from the coffee agroindustry.     

Performance characterization of a model bioreactor for the biodegradation of trichloroethylene by Pseudomonas cepacia G4.

Pseudomonas cepacia G4 grown in chemostats with phenol demonstrated constant specific degradation rates for both phenol and trichloroethylene (TCE) over a range of dilution rates. Washout of cells from chemostats was evident at a dilution rate of 0.2 h-1 at 28 degrees C. Increased phenol concentrations in the nutrient feed led to increased biomass production with constant specific degradation rates for both phenol and TCE. The addition of lactate to the phenol feed led to increased biomass production but lowered specific phenol and TCE degradation rates. The maximum potential for TCE degradation was about 1.1 g per day per g of cell protein. Cell growth and degradation kinetic parameters were used in the design of a recirculating bioreactor for TCE degradation. In this reactor, the total amount of TCE degraded increased as either reaction time or biomass was increased. TCE degradation was observed up to 300 microM TCE with no significant decreases in rates. On the average, this reactor was able to degrade 0.7 g of TCE per day per g of cell protein. These results demonstrate the feasibility of TCE bioremediation through the use of bioreactors.     

Changes in structure,activity and metabolism of aerobic granules as a microbial response to high phenol loading

Four column-type sequential aerobic sludge blanket reactors were fed with phenol as the sole carbon and energy source and operated at loading rates of 1.0, 1.5, 2.0 and 2.5 kg phenol m^–3 day^–1. The results indicated that phenol loading exerted a profound influence on the structure, activity and metabolism of the aerobic granules. Compact granules with good settling ability were maintained at loadings up to 2.0 kg phenol m^–3 day^–1, and structurally weakened granules with enhanced production of extracellular polymers and proteins and significantly lower hydrophobicities were observed at the highest loading of 2.5 kg phenol m^–3 day^–1. Specific oxygen uptake rate, catechol 2,3-dioxygenase (C23O) and catechol 1,2-dioxygenase (C12O) activities peaked at a loading of 2.0 kg phenol m^–3 day^–1, and declined thereafter. Granules degraded phenol completely in all four reactors, mainly through the meta cleavage pathway as C23O activities were significantly higher than C12O activities. At the highest loading applied, the anabolism and catabolism of microorganisms were regulated such that phenol degradation proceeded exclusively via the meta pathway, apparently to produce more energy for overstimulation of protein production against phenol toxicity. This work contributes to a better understanding of the ability of aerobic granules to handle high-strength industrial wastewaters containing chemicals that are normally inhibitory to microbial growth.     

Bacterial diversity and function of aerobic granules engineered in a sequencing batch reactor for phenol degradation

Aerobic granules are self-immobilized aggregates of microorganisms and represent a relatively new form of cell immobilization developed for biological wastewater treatment. In this study, both culture-based and culture-independent techniques were used to investigate the bacterial diversity and function in aerobic phenol- degrading granules cultivated in a sequencing batch reactor. Denaturing gradient gel electrophoresis (DGGE) analysis of PCR-amplified 16S rRNA genes demonstrated a major shift in the microbial community as the seed sludge developed into granules. Culture isolation and DGGE assays confirmed the dominance of beta-Proteobacteria and high-G+C gram-positive bacteria in the phenol-degrading aerobic granules. Of the 10 phenol-degrading bacterial strains isolated from the granules, strains PG-01, PG-02, and PG-08 possessed 16S rRNA gene sequences that matched the partial sequences of dominant bands in the DGGE fingerprint belonging to the aerobic granules. The numerical dominance of strain PG-01 was confirmed by isolation, DGGE, and in situ hybridization with a strain-specific probe, and key physiological traits possessed by PG-01 that allowed it to outcompete and dominate other microorganisms within the granules were then identified. This strain could be regarded as a functionally dominant strain and may have contributed significantly to phenol degradation in the granules. On the other hand, strain PG-08 had low specific growth rate and low phenol degradation ability but showed a high propensity to autoaggregate. By analyzing the roles played by these two isolates within the aerobic granules, a functional model of the microbial community within the aerobic granules was proposed. This model has important implications for rationalizing the engineering of ecological systems.     

Bacterial Diversity and Function of Aerobic Granules Engineered in a Sequencing Batch Reactor for Phenol Degradation

Aerobic granules are self-immobilized aggregates of microorganisms and represent a relatively new form of cell immobilization developed for biological wastewater treatment. In this study, both culture-based and culture-independent techniques were used to investigate the bacterial diversity and function in aerobic phenol- degrading granules cultivated in a sequencing batch reactor. Denaturing gradient gel electrophoresis (DGGE) analysis of PCR-amplified 16S rRNA genes demonstrated a major shift in the microbial community as the seed sludge developed into granules. Culture isolation and DGGE assays confirmed the dominance of β-Proteobacteria and high-G+C gram-positive bacteria in the phenol-degrading aerobic granules. Of the 10 phenol-degrading bacterial strains isolated from the granules, strains PG-01, PG-02, and PG-08 possessed 16S rRNA gene sequences that matched the partial sequences of dominant bands in the DGGE fingerprint belonging to the aerobic granules. The numerical dominance of strain PG-01 was confirmed by isolation, DGGE, and in situ hybridization with a strain-specific probe, and key physiological traits possessed by PG-01 that allowed it to outcompete and dominate other microorganisms within the granules were then identified. This strain could be regarded as a functionally dominant strain and may have contributed significantly to phenol degradation in the granules. On the other hand, strain PG-08 had low specific growth rate and low phenol degradation ability but showed a high propensity to autoaggregate. By analyzing the roles played by these two isolates within the aerobic granules, a functional model of the microbial community within the aerobic granules was proposed. This model has important implications for rationalizing the engineering of ecological systems.     

Anaerobic degradation of benzoate: batch studies

Response of benzoate along with phenol to different anaerobic inocula has been investigated in batch reactors. In Phase I, the anaerobic biodegradability of benzoate and phenol were evaluated using (a) washed acclimatized granular sludge (WAGS) collected from a passive phenol fed bench-scale up-flow anaerobic sludge blanket reactor (UASBR) and (b) unacclimatized flocculent sludge (UFS) from a UASB based sewage treatment plant (STP). The effect of varying concentrations of benzoate has been investigated in Phase II using acclimatized granular sludge (AGS) from a bench-scale UASBR. Extent of degradation of benzoate was more than the phenol. Increasing benzoate COD from 2500 to 11,700mg/L, resulted in decrease in (i) rate constant, k from 0.79 to 0.11/d and (ii) ultimate biochemical methane potential (microb, g CH4-COD formed/g benzoate COD) from 84% to 60%. Temporal trend conforming to logistic S-curve indicated stressed conditions at higher benzoate concentration. Benzoate degradation was found to be sensitive to nature as well as quantity i.e. food to microorganism (F/M) of inocula used.     

Large-scale (20 l) culture of surface-immobilized Catharanthus roseus cells.

Surface-immobilized C. roseus cell cultures were grown in a 20-l modified airlift bioreactor operated at 0.51 vvm (kLa approximately 8 h-1) under various gassing regimes [air, 2% (v/v) and 5% CO2]. Extracellular ammonium, phosphate, and nitrate ions as well as carbohydrate uptake and pH value of the medium were monitored together with on-line dissolved oxygen concentration, conductivity of the medium, and carbon dioxide production rate (CPR) of the cultures. Cultures supplemented with 2% CO2 showed higher nitrate (5.0-7.0 mM d-1) and carbohydrate (3.3 g l-1 d-1) uptake rates and biomass production (mu approximately 0.24 d-1, yield approximately 0.33 g dw g CHO-1 and 7.4 g dw L-1) as compared to air (3.6 mM d-1, 2.1 g l-1 d-1; 0.20 d-1, 0.25 g dw g CHO-1 and 5 g dw l-1) and 5% CO2 (2.0-3.6 mM d-1, 2.0 g l-1 d-1; 0.11 d-1, 0.20 g dw g CHO-1 and 5 g dw l-1) cultures and as reported previously for suspension cultures. In addition, air and 5% CO2 cultures displayed incomplete carbohydrate uptake and, more important, phosphate and ammonium ion release into the medium at the end, which was ascribed to loss of viability. This was not observed for 2% CO2 immobilized bioreactor as well as shake flask control suspension cultures, which suggests that sparged C. roseus surface-immobilized cell cultures require 2% CO2 supplementation of the gas phase for both maximum growth and retained viability. The maximum CPRs of all cultures were in the same range (2.1-2.8 mM CO2 l-1 h-1). However, the estimated maximum specific CO2 production rates of 2% CO2 and 5% CO2 immobilized cultures (0.6 mM g dw-1 h-1) were lower than those found for air-sparged immobilized cultures (1.0-1.3 mM g dw-1 h-1). These rates are significantly higher than those reported in the literature for C. roseus cell suspension cultures performed in bioreactors gassed with air (approximately 0.2-0.55 mM g dw-1 h-1).