Microbial community structure of wastewater treatment subjected to high mortality rate due to ozonation of return activated sludge

AIMS

This study investigated the effects of return activated sludge (RAS) ozonation, on the bacterial community structure of pilot-scale wastewater treatment systems.

METHODS AND RESULTS

Two parallel activated sludge reactors were operated to treat real municipal wastewater for 98 days. The RAS of one of the reactors was subjected to increasing doses of ozone during the experimental period, which resulted in higher reduction in biosolids waste production and higher bacterial growth rate. The bacterial community structures were investigated by 16S rRNA gene amplicon high-throughput pyrosequencing and fluorescence in situ hybridization (FISH). The structures remained highly similar throughout the experiment despite the ozone treatment. Comparative analyses between pyrosequencing and FISH revealed clear discrepancies in the proportion of some bacterial populations.

CONCLUSIONS

The results suggest that RAS ozonation is not a main environmental factor structuring the community composition. Instead, the parallel drifts and slight convergence of the two community structures indicate that other environmental factors such as influent wastewater composition and temperature may be more important. Care should be exercised in interpreting the proportion of sequence reads as pyrosequencing may be biased as compared to FISH.

SIGNIFICANCE AND IMPACT OF THE STUDY

This study provides new insights on the importance of indiscriminate high mortality rates brought by external factors (here ozonation) on microbial community structures of activated sludge system.

Modelling using rRNA-structured biomass models

Models currently used have been developed to describe the storage response in the activated sludge process. In these models the distribution of the substrate flux between growth and storage is an empirical function. rRNA-structured biomass models are proposed to describe the metabolic status of cells in view of predicting the growth response (dmicro/dt) of cells in activated sludge process. The autocatalytic reaction rate of the synthesis of the PSS component (rRNA) can provide a mechanistic explanation for the growth response and the growth lag phase. The proposed models were able to describe and predict properly the growth response of the biomass in various types of reactor. Such models could be more widely applicable by using intrinsic model parameters. This would be a key improvement for as it would lead to improved models for design.

Effect of chemical and physical parameters on a pulp mill biotreatment bacterial community

The bacterial community composition in an activated sludge plant treatment from a bleached kraft pulp mill was monitored over a period of 209 days. Using DGGE and terminal-Restriction Fragment Length Polymorphism (t-RFLP) analysis we generated community DNA fingerprints over the time period. Both methods produce fingerprints that can be used to monitor stability in the system and generate fragments that can be associated with bacterial taxa. Chemical and physical parameters were also collected during that same time frame. We found a number of significant correlations with influent variables such as temperature, chemical oxygen demand (COD), Biochemical oxygen demand (BOD) and chloroform concentrations suggesting that these were the most likely parameters to influence the bacterial community structure. In addition several taxa correlated to important performance indicators such as COD/BOD removals and SVI. Multivariate analysis also confirmed the strong links between taxa variation and temperature, nutrient loads, chloroform and also one class of filaments. Establishing the identity of these taxa and their ecological preferences will greatly enhance our understanding and management of biological treatment systems.

Oligonucleotide probe hybridization and modeling results suggest that populations consuming readily degradable substrate have high cellular RNA levels

Analyses based on ribosomal RNA (rRNA)-targeted hybridization performed in our laboratory identified two types of bacterial populations: a population with a high RNA level per biomass and a population with a low level of RNA per biomass. To extend these descriptions, the diurnal dynamics of the RNA pool were monitored by rRNA-targeted oligonucleotide probe membrane hybridization. Under the typical diurnal variation in COD loading rate experienced by municipal wastewater treatment plants, the RNA level of the bacterial population with a high level of RNA per biomass varied with changes in the COD loading rate. Under the same conditions, the RNA level of the population with low RNA level per biomass remained constant. A structured biomass model was developed to describe these data. Substrate COD was divided into a readily biodegradable and a slowly biodegradable COD fraction. It was assumed that two specialized populations coexist in municipal activated sludge treatment systems. One population consumes readily degradable COD and the other consumes slowly degradable COD. According to model simulations, the population consuming readily degradable COD has a high level of RNA per biomass under variable substrate concentrations. Comparatively, the population consuming slowly degradable COD has a low level of RNA level per biomass. Furthermore, model simulations reproduced the two diurnal RNA profiles observed in a full-scale municipal activated sludge system. Therefore, we suggest that two populations can be distinguished in municipal activated sludge systems: a population consuming readily degradable substrate and a population consuming slowly degradable substrate.

Who eats what? Classifying microbial populations based on diurnal profiles of rRNA levels

Identifying the relationships between various bacterial populations and the substrates they consume is central to the understanding of population dynamics and to the development of process control in activated sludge. However, linking a heterotrophic population to its activity in situ is difficult because ribosomal RNA (rRNA) techniques, while allowing the rapid identification of populations, provide little information about their heterotrophic activity. Activated sludge models describe biodegradation kinetics by classifying substrates into two types: readily and slowly degradable substrates. Assuming that bacterial populations specialize in degrading one type of substrate, their growth rate should be affected differently if the COD loading rate varies diurnally as for a municipal activated sludge system. Modeling results suggested that the growth rates of populations consuming readily degradable substrates vary according to variations in COD loading rate. On the other hand, the growth rates of populations consuming slowly degradable substrates do not change despite the variation in COD loading rate. Since the cellular rRNA level is positively correlated with the growth rate, we hypothesized that the rRNA levels of some populations in municipal activated sludge should increase throughout the day, while they should stay constant for other populations. This hypothesis was verified by monitoring the rRNA level of Acinetobacter (a model population consuming readily degradable substrates) and Gordonia (a model population consuming slowly degradable substrates) in the mixed liquor of a full-scale municipal activated sludge reactor for three weeks.