Reply to Obadia et al.: Effect of methyl paraben on host–microbiota interactions in Drosophila melanogaster

Obadia et al. (1) suggest that variation in the concentration of the fly media fungicide methyl paraben (mp) (also known as Tegosept or Nipagin) can restrict gut microbial growth and diversity (2, 3), and that this could provide insight into conflicting data on the role of the gut microbiome in generating positive assortative mating by diet in Drosophila melanogaster (4–6).

mp is reported to reduce the culturable bacterial diversity of wild Drosophila to a level comparable with established laboratory populations (2), although the effect of mp on the microbial diversity of long-established laboratory populations is not yet clear. This is consistent with the greater gut microbiome diversity of wild in comparison with laboratory-reared Drosophila (7, 8). mp >0.1% is also reported to severely inhibit the growth of Acetobacter (3). However, this contrasts with data from our (5) and Sharon et al.’s (4) assortative mating experiments. In these tests, gut microbiomes from cornmeal–molasses–yeast (CMY) diet (0.1% mp) flies were dominated by Acetobacter. These species also represented a greater proportion of the gut microbiome in CMY compared with starch diet (0% mp) flies. Hence, variation in Acetobacter was influenced more strongly by other factors (e.g., diet and sugar availability) than by differences in mp between diets (9). Obadia et al. (1) also show that mp of up to 0.3% has little effect on the growth of Lactobacillus plantarum, the suggested putative causal agent of positive assortative mating in the host (4). Hence, variation in mp between studies cannot itself be the cause of differential effects on L. plantarum with the potential to generate host assortative mating.

As fly gut microbiota are generally transient in the laboratory (e.g., ref. 3), variation in mp in diets before the start of mate choice tests should not affect their outcome. This suggestion is supported by the finding that our (5) and Sharon et al.’s (4) analyses of microbial community composition of CMY and starch diet gut microbiomes showed convergence, despite different starting diets and mp concentrations before the initiation of the studies.

We agree that gut microbes could represent food or contribute nutrients to their fly hosts (1). Hence, quantification of absolute microbial loads would be useful. Microbiome “food” could influence a variety of host life history traits, including mate choice (e.g., if it influenced host body size). Given their transient nature, it is possible that nutrients from gut microbiomes could contribute to immediate proximate effects on hosts. However, they are unlikely to shape long-term evolutionary responses in host mating behavior.

Additional studies using wild isolates of the fly microbiota would also be useful to resolve emergent patterns. For example, a recent report of greater colonization ability of wild over laboratory fly bacteria (3) could suggest that host–microbe associations are stronger in the wild. However, these results currently contrast with studies of wild caught Drosophila, of a flexible diet-determined microbiome (8, 10, 11).

Overall, these growing insights are useful in increasing the resolution of discussions into the significance of diet-associated mate choice versus diet-induced divergent microbiota.