1416 and 694 ASVs in gills and skin respectively were preserved in the taxonomic analysis after filtering. The predominant phylum in gills was Proteobacteria (~50%) in the control and hydrolyzed groups and reached 70.58% in the raw diet. The Bacteroidota phylum was the most represented and Firmicutes and Actinobacteria were close to 3% in all treatments in this tissue. However, in all three different treatments, the phyla abundance was similar in skin samples, ~90% of which corresponded to Proteobacteria. Bateroidota (5.5%–5.9%), Firmicutes (1.8%–2%) and Actinobacteria (<1%) constitute the other phyla. In gills, 13 significantly higher ASVs were obtained in the control (such as Achromobacter, Acidobacter) versus 4 whose abundance was higher in the hydrolyzed diet. The number of ASVs that differed significantly between the gill microbiota in the control vs. raw group amounted to 70, most of them being ASVs corresponding to the genus Shewanella (43) higher in the raw diet. Nevertheless, in the skin samples, Control group showed a significant increase of abundance related to Acinetobacter, Achromobacter, Pseudomonas, Shewanella, Vibrio and Sphingomonas among others. In the hydrolyzed group, the most significant abundance was associated with the genus Pseudomonas, Vibrio, Pseudoalteromonas, Ralstonia o Cutibacterium. In the case of the skin raw samples, there were ASVs corresponding to the genus Acinetobacter, Streptococcus or Pseudoalteromonas that were significantly different respect of control. Taking ASV abundance matrix for each diet and tissue, 6 co-occurrence networks were constructed. In all treatments in gills, Acinetobacter was a central genus in the network, and exhibited a negative correlation with Polaribacter. Besides, in the raw group, Polaribacter also showed negative correlation with Aeromonas, Pseudomonas and Francisellacea. This work shows that Acinetobacter has a key role in the balance of mucosa microbiota and was in co-exclusion with Polaribacter.