Composition and abundance of nitrifiers in engineered systems: Molecular and community ecology approaches

Abstract

Nitrification is of central importance in engineered systems for drinking water production and wastewater treatment because of its key role in ammonium removal from these systems. Incomplete nitrification can lead to the release of ammonium residues in the finished water that can pose human and environmental health risks. Therefore, efficient drink-ing water and wastewater nitrifying community are of high importance. Several studies have revealed vast phylogenetic and functional diversity of nitrifier guilds in engineered systems. Nitrifiers are grouped as ammonia-oxidizing prokaryotes (AOP) and nitrite oxidizing bacteria (NOB). AOP further consists of canonical ammonia oxidizing bacteria (AOB), complete ammonia oxidizers (comammox) and ammonia-oxidizing archaea (AOA). Microbial diversity and composition are of central importance in engineered systems because highly diverse com-munities can contain a greater pool of physiological and genetic traits which provide them the capacity to cope with environmental perturba-tions and ultimately contribute to better system performance. In addi-tion to the whole microbial community composition, knowing the con-tribution of guild members along with their compositional and ecophys-iological information can allow us to introduce preferred microbial eco-types as per the engineered system’s requirement. Yet, what controls nitrifiers composition in engineered systems is largely unknown. Histor-ically, the emphasis has been put on the role of selection (i.e. fitness dif-ference between and within the guilds) but new developments in com-munity ecology are available that include processes such as immigration and stochastic growth (‘neutral’ processes). This can be relevant be-cause drinking water and wastewater treatment systems are open sys-tems with influent water that contain nitrifiers. The overall aim of this Ph.D. project was to assess and implement key molecular and community ecology approaches for describing the abun-dance and composition of nitrifiers. Then, to identify the determinants of nitrifiers assembly in engineered systems for drinking water produc-tion and wastewater treatment. To quantify canonical AOB, researchers routinely employ qPCR target-ing ammonia monooxygenase (amoA) gene for canonical AOB or ca-nonical AOB specific 16S rRNA gene. However, these two approaches were not typically compared, and it was unclear whether they were equally good at estimating AOB abundance. In this comparison, incon-sistencies were found based on the difference in primer pair selectivity combined with the compositional differences of the canonical AOB in drinking water biofilters. Therefore, we suggested that the primer set selection for canonical AOB quantification should be carefully made as the results can be heavily primer and nitrifier composition dependent. Next, I attempted to develop new primer sets (based on the currently available amoA nucleotide sequence data for canonical AOB) that cover most AOB and amplify large enough amplicons to make them suitable for both quantifications (qPCR), and phylogenetic/compositional analysis (amplicon sequencing). Due to the large divergence between AOB genera, I utilized the maximum coverage degenerate primer design method and successfully generated new primer sets targeting amoA with better coverage than the ones presently used. But these primer sets showed unspecific amplification in the in-vitro analysis therefore, I choose the most suitable primer sets amongst the existing canonical AOB primers for the rest of my work. Next, I wanted to determine to which degree the universal 16S rRNA gene amplicon sequencing of environmental communities provides an accurate description of nitrifying guilds. I compared universal 16S rRNA gene amplicon sequencing to the functional gene (amoA AOB and nxrB Nitrospira (NOB)) based targeted sequencing approaches for assessment of nitrifiers relative abundance, diversity, and composition. The universal 16S rRNA gene sequencing approach provided accurate estimates of nitrifier composition and clustered samples consistently with their origin. It also provided relative abundance from the two ap-proaches within ~1.2 orders of magnitude of them, but with a measurable bias that should be considered when comparing estimates from both approaches. The richness and diversity estimations were found to be likely limited by the sequencing depth. Overall, the univer-sal approach works well when guilds of interest are dominant in an en-vironment (for example, Nitrospira in drinking water biofilters) or when the goal is to estimate guild relative abundance. Further, I compared the neutral community assembly model (which takes into account the frequency of presence and absence of taxa) and differential abundance based approaches in their ability to identify the selected members of the DWTP nitrifying community. Overall, the neu-tral model always predicted fewer positively selected taxa compared to the differential abundance based approach. The combined abundance of selected members of Nitrospira contributed majorly to the total abun-dance of Nitrospira for one drinking water treatment plant (DWTP) but not the other, indicating that the Nitrospira community at one DWTP was largely neutrally assembled while at other DWTP neutral processes played a smaller role compared to selection. Highlighting the pros and cons of both methods I suggested that detection of selection in microbial communities should be addressed using a combination of approaches covering both frequency and abundance data of the taxa. Lastly, the evaluated molecular and community ecology methods were implemented for describing the effect of releasing resource limitation on nitrifying communities in malfunctioning full-scale drinking water biofil-ters. We showed that releasing copper limitation involved changes in the relative abundance between nitrifying guilds and had no effect on the relative abundance of other microbial guilds or potentially pathogenic microbes. Mainly, the relative abundance of Nitrosomonas increased by almost one order of magnitude upon releasing copper limitation. The relative abundance of Nitrospira (including comammox Nitrospira) also increased, but this was true for only one plant. Also, no changes within nitrifying guild composition were observed which indicated that there were no fitness differences amongst nitrifying guild members. Taken together these findings suggested that it is possible to enhance the biological stability and key process performance in complex microbial communities by influencing the abundance of specific microbial groups through selective nutrient dosing, i.e., by releasing elemental nutrient limitation. Overall, this Ph.D. project presented a first systematic evaluation and implementation of molecular and community ecology approaches for describing specific microbial (nitrifying) community constituents and processes driving their assembly in complex microbial communities in drinking water and wastewater treatment systems.