Introduction: The frequency of cyanobacterial blooms seems to have increased globally in recent decades due to human induced eutrophication and climate change. Cyanobacterial blooms can produce several groups of toxins, among which microcystin-LR (MC-LR) is one of the most abundant. Effects of MC-LR on avian microbiome have not been studied and studies in laboratory murines have been limited to metabarcoding of prokaryotes.

Methods: Using RNA shotgun sequencing, we compared the richness and composition of metabolically active prokaryotes, expressed virulence factors, antimicrobial resistance genes, metabolic pathways, Gene Ontology terms, enzymes, and proteins in mallards (Anas platyrhynchos) that were orally exposed to a sublethal dose of MC-LR for one week and unexposed birds.

Results: Richness and composition of all compared features did not differ between exposed and control birds and none were differentially expressed between exposure groups. However, richness and/or composition of all features except virulence factors and Carbohydrate Active enzymes had multiple-fold greater dispersion in exposed birds than in controls. This effect was especially pronounced in expressed metabolic (MetaCyc) pathways.

Discussion: Our results suggest that MC-LR exposure had a stochastic (rather than deterministic) effect on cecal microbiota, especially its function. Observed disturbance of the microbiota homeostasis is consistent with the Anna Karenina Principle. This principle has been documented in a wide range of eukaryotes using primarily microbial community metabarcoding. Although stochastic disturbance of microbiota function has been hypothesized, our study seems to be the first to demonstrate this in an experimental study.