Drosophila media microbiome

Since the dawn of agriculture, humans selected for desirable animal and plant traits using quantitative genetic approaches. Yet, many traits have low genetic heritability and do not respond to artificial selection. Recent progress in microbiology and the realization that the microbiome affects host phenotypes suggested another approach for such traits – selecting on the microbiome instead. “Microbiome engineering” that indirectly selects on the microbiome via desired host traits promises to change host phenotypes, potentially improving crop yields, targeting metabolic diseases, etc. However, only a few studies have actually attempted to carry out such experiments, and exclusively in plants. Using fruit flies as an animal model, and selecting on development time, we found that selection on the host phenotype was insufficient to overcome independent evolution by the microbiome occurring simultaneously. Independent microbial evolution may limit the efficacy of microbiome engineering and must be accounted for by appropriate experimental controls.