Pyropia yezoensis
is commercially the most important edible red alga in China, and red rot disease is viewed as one of the major constraints for its cultivation. Microbes within the oomycetic genus
Pythium
have been reported as the causative agents for this disease; however, little is known about the interactions between the disease and the epiphytic and planktonic bacterial communities. In the present study, bacterial communities associated with uninfected, locally infected, and seriously infected thalli collected from cultivation farms, and within seawater adjacent to the thalli, were investigated using in-depth 16S ribosomal RNA (rRNA) gene sequencing in conjunction with assessing multiple environmental factors. For both thalli and seawater, uninfected and infected communities were significantly different though alpha diversity was similar. Phylogenetic differences between epiphytic bacterial communities associated with
P. yezoensis
were mainly reflected by the relative changes in the dominant operational taxonomic units (OTUs) assigned as genus
Flavirhabdus
, genus
Sulfitobacter
, and family Rhodobacteraceae. The prevalent OTUs in seawater also differed in relative abundance across the communities and were affiliated with diverse taxa, including the phyla Actinobacteria, Verrucomicrobia, and Bacteroidetes, and the classes Alpha- and Gamma-proteobacteria. The differentiation of bacterial communities associated with
P. yezoensis
and seawater was primarily shaped by reactive silicate (RS) content and salinity, respectively. In particular, 14 potential indicators (two OTUs on
P. yezoensis
and twelve OTUs in seawater) were identified that significantly differentiated
P. yezoensis
health statuses and correlated with environmental changes. Overall, the present study provides insights into the alterations of bacterial communities associated with
P. yezoensis
and surrounding seawater co-occurring with red rot disease. Observed changes were closely associated with health status of algal host, and highlight the potential of using community differentiation to forecast disease occurrence.