DEVELOPMENT OF A MODEL MARINE FISH: HIGH THROUGHPUT GENOME AND MICROBIOME ANALYSIS OF PACIFIC CHUB MACKEREL Scomber japonicus ACROSS SEASONS AND BODY SITES

Although the potential of developing marine aquaculture production in coastal areas is promising, the realization has been encumbered by concerns of environmental degradation, including genetic contamination of wild stocks through escapement, and indigenous fish disease transmission. Because of the high costs and feasibility challenges with carrying out the research to test these questions, the fate of marine offshore finfish production in the US and elsewhere is in limbo. High value species like yellowtail, Seriola lalandi, and bluefin tuna, Thunnus sp., are targets of marine pen culture, but few data are available on understanding the natural genomic variation of these fish. Since host-associated microbes make up over 50% of the cellular composition and 100 times the genomic content of most vertebrates, we set out to describe the hologenome (Fig 1a) of multiple body sites (gill, skin, digesta, gastrointestinal tract, and pyloric caeca) of the Pacific Chub Mackerel, Scomber japonicas, across 20 samplings over 7 months from the Scripps Institute of Oceanography (SIO) pier in San Diego (Fig 1b). Currently there are is no genomic assembly for Pacific Mackerel (Scomber japonicas) so we sequenced the genome with several technologies including long-read nanopore technology. Using 60x coverage of Illumina short read sequencing we have estimated the diploid Scomber japonicas genome to be 750 Mb with 100 Mb of high copy number repeats such as the ribosomal DNA and centromeres. Based on the current genome sequencing the individual we are sequencing has low (1%) heterozygousity. To assess the microbiome, we developed a low cost, miniaturized (5 ul), high-throughput (384-sample) microbiome library preparation method with the Echo-550 using this micobiome timecourse. Alpha diversity, as measured by microbial richness and eveness (Shannon entropy), were not significantly different within sample types across PCR methods, but were significant as compared across body sites (Kruskal-Wallis, P<0.0001) (Fig 1c). Gill samples were rich in microbial diversity but dominated by multiple Shewanella sp. sub-operational taxonomic units (sOTUs)and skin samples were enriched in Photobacterium and Rickettsia. Digesta had the highest eveness across body sites, a higher richness than the GI or pyloric cacae containing various sea water associated microbes like Synechococcus and Pirellulaceae along with putative GI associated Vibrio sp. and Photobacterium sp. The GI and pyloric cacae had the lowest alpha diversity (Fig 1c) across body sites and were both dominated by mycoplasma. The gill and skin microbiomes of the mackerel are most similar to sea water compared to GI-associated microbiomes (Fig 1d) yet each are enriched in unique microbes found in low numbers or absent from the ocean. While this finding has been suggested and shown in mammals, this is the first example in fish. Beta diversity was not significantly different when comparing the two methods (Fig 1e), but did differ significantly when comparing across sample types. Various microbes across the body sites (gill = 12 sOTUs, skin = 4 sOTUs, digesta = 8 sOTUs, pyloric caeca = 2 sOTUs) were correlated to seasonal variation of water temperature (Spearman correlation). Synechococcus and Rhodobacteraceae were positively associated with water temperature and prevalence on the gill, skin and digesta. No microbes were correlated with length, mass, or condition factor of the mackerel. We have described the Pacific mackerel wild type microbiome across multiple body sites and seasons demonstrating the abiotic and biotic effects driving the microbial communities and providing context for future disease transmission studies.