Host-associated microbiota structure is not always predicted by environment (e.g. bacteria in food) despite evidence that microbiota structure outside of hosts commonly is. This incongruence could be due to variation in important host genetic traits that influence microbiota assembly (e.g. immunity-related genes). We characterized gut microbiota structure among six wild threespine stickleback populations in Oregon and tested how well this structure was explained by population level genetic distance, physical distance, and environment.
We collected 120 fish, including 20 each from two estuaries and two nearby freshwater ponds on the coast, and two freshwater ponds >430 river miles inland. We sequenced restriction site associated DNA from fins and the V4 region of bacterial 16s rDNA from guts. We transformed 40K RAD markers per sample into a matrix of average genetic distance among populations using STACKS and >1 million 16s sequences per gut into 20K bacterial variants using the DADA2 pipeline. In R statistics software we visualized microbiota composition and ran a permutational multivariate analysis of variance on a Bray distance matrix based on microbiota structure (using adonis in the package ‘vegan’) with fish population genetic distance and distance among sites as factors nested within environment (freshwater or estuary).
Results/Conclusions
We found that gut microbial communities partitioned due to average genetic distance between populations of fish (PERMANOVA R2=0.10 P=0.001), distance in water miles among sites (PERMANOVA R2=0.05 P=0.001), and environment (PERMANOVA R2=0.05 P=0.001). Gut microbial community also differed significantly between 2015 and 2017 (ANOSIM R=0.176 P=0.001), hence PERMANOVA factors were stratified by time. In addition, gut microbial communities in fish collected at one of the estuary were distinct from all other sites in terms of composition. This was primarily driven by an over-representation of bacteria in the genera Vibrio, Photobacterium, and Endzoicomonas.
Our findings conclude that microbiota assembly among wild populations of threespine stickleback in Oregon is influenced by multiple factors. Of the factors we measured, time followed by host population genetic distance seem to have the strongest influence on microbiota structure. In addition, variation in host associated microbiota can be driven by a few abundant genera in a community of 20,000 bacterial variants. This should inform researchers to further study microbiota assembly in wild populations focusing their efforts on testing host genetic effects over time.