| Abstract [eng] |
Cyanobacterial blooms in eutrophic freshwater systems drive predictable succession in the surrounding microbial community, yet the temporal dynamics of these responses remain poorly characterised at the genome-resolved level. Standardised, reproducible pipelines for nanopore-only long-read metagenomics of freshwater communities are also lacking. This thesis aimed to characterise cyanobacteria-driven microbial community dynamics in a eutrophic freshwater pond using a reproducible nanopore-only metagenomics approach, with objectives covering pipeline development, metagenome-assembled genome (MAG) and viral sequence recovery, co-occurrence network construction, guild-level temporal analysis, and viral community characterisation. Water samples were collected from Simnas Fish Hatchery Pond at ten timepoints between August and September 2023. Bacterial and viral DNA fractions were sequenced on an Oxford Nanopore MinION using R10.4 chemistry. Assembly was performed with metaFlye, polished with Racon and Medaka, and binned using MetaBAT2 and MaxBin2 consolidated with MetaWRAP. MAG quality and taxonomy were assessed with CheckM2 and GTDB-Tk v2 respectively. Bacterial community profiles were generated with Kraken2, co-occurrence networks inferred with SPIEC-EASI, and communities detected using the Leiden algorithm. Viral sequences were identified with VIBRANT, quality-assessed with CheckV, and annotated with geNomad. Sequencing yielded 757,511 reads and 3.10 Gbp of data. Thirteen medium-quality MAGs were recovered across four phyla, with Bacteroidota predominating. Nine ecologically interpretable bacterial guilds were identified, showing structured temporal dynamics consistent with bloom senescence succession. Contextualisation against AFA cell count data revealed distinct temporal responses among bacterial families, with Verrucomicrobiaceae tracking peak bloom biomass and Sphingobacteriaceae responding to senescence-derived organic matter. Viral analysis yielded 1,986 sequences, predominantly lytic, dominated by Autographiviridae and Kyanoviridae, with putative auxiliary metabolic genes identified in several genomes. Nanopore-only metagenomics at modest sequencing depth can yield ecologically meaningful signal from complex freshwater communities. The publicly available pipeline and temporally resolved guild analysis represent the primary contributions of this work. |
