Alpha diversity of zebrafish and carp RAS- and fish-associated microbiota. A) Chao1 index (ASV richness) of alpha diversity per sample type. Each dot corresponds to one sequenced sample, with bars indicating median values. Significance of species and sample type in the systems was tested using a two-way ANOVA. Sample type had a significant effect on Chao1 diversity index values (F (3, 29) = 23.84, p < 0.001). B) Shannon diversity index (ASV richness and evenness) of alpha diversity per sample type. Each dot corresponds to one sequenced sample, with bars indicating median values. Significance of the species and sample type was tested using a two-way ANOVA followed by multiple comparison testing. Sample type had a significant effect on the Shannon diversity index (F (3, 29) = 5.683, p = 0.0035), with an additional interaction effect observed between sample type and species (F (3, 29) = 112.0, p < 0.001). Carp and zebrafish samples were tested separately. Significant differences between sample types are indicated using different letters, with upper-case letters used for carp samples and lower-case letters for zebrafish samples.

Ordination plot of carp and zebrafish RAS- and fish-associated microbiomes. Weighted UniFrac distances were calculated from log-transformed abundance data to achieve homogeneity of variance between sample types. Colors indicate the sample types and species are distinguished by symbol. Significance of both factors was tested using a PERMANOVA test. Sample type and species significantly affected weighted UniFrac distances (F = 28.9, r2 = 0.58, p < 0.001 and F = 11.57, r2 = 0.078, p < 0.001, respectively).

Phylum-level composition of carp and zebrafish RAS- and fish-associated microbiomes. Major phylum level groupings are shown (>5 % relative abundance in any individual sample), with phyla below this value classified as ‘Other’ and with Proteobacteria split further to class level. A) Relative abundances of major phyla present in carp system compartments. B) Relative abundances of major phyla present in zebrafish system compartments. Abundances of phyla are shown as percentages of the total number of 16S rRNA gene copies recovered.

Unique and shared ASVs in RAS- and fish-associated microbiomes of carp and zebrafish. Venn diagrams indicate the set of ASVs found in each respective microbiome for carp (A) and zebrafish (B) systems separately. Shared ASVs had to be present in at least half of the samples of the type.

Maximum likelihood tree of ammonia monooxygenase subunit A (amoA) gene sequences obtained from RAS, water and gill microbiomes. amoA sequences obtained in this study are colored according to the source of the sequence, with biofilter samples in red, water samples in blue and gill samples in green and are printed in bold. Previously published gill-associated amoA sequences are also printed in bold. Bootstrap values are given for all nodes in the tree. Methyloprofundus pmoA (LC616792) was used as outgroup for the tree.