Correction to “Xenobiotic estradiol-17ß alters gut microbiota of hatchling American alligators (Alligator mississippiensis)”

Abstract

Murphy, K.M., Watkins, M.M., Finger, J.W., Kelley, M.D., Elsey, R.M., Warner, D.A., Mendonça, M.T. (2022) Xenobiotic estradiol-17ß alters gut microbiota of hatchling American alligators (Alligator mississippiensis). Environmental Microbiology, 24(12), 6336–6347. https://doi.org/10.1111/1462-2920.16222

In Figure 2B, we report no statistical significance between treatment groups (P > 0.05) using a non-metric multidimensional scaling plot (NMDS) followed by an analysis of similarity (ANOSIM). We later state “Bray–Curtis NMDS showed different patterns of clustering of samples across treatment groups (Figure 2B). However, bacterial community structure did not differ between groups (R = 0.026, p = 0.28), suggesting that microbiomes of each treatment group were similar.”

In the following discussion, we expand on this by stating “One of the most interesting results from our study shows that exposure to a relatively low concentration of E2 increases microbiota absolute abundances but does not influence microbiota diversity. This dose-dependent effect suggests that varying concentrations of E2 may influence microbial community composition, in terms of alpha and beta diversity, in different ways.”

Upon revisiting these analyses, we discovered that the metadata file used to generate the NMDS plot was out of order; meaning, the samples displayed in the published beta diversity plot are not correctly associated to treatment groups. When this file is in the correct order, we found a different pattern in that the experimental groups do significantly differ from the control group.

We would like to issue a correction to Figure 2B and to correct our statement in the results section to read “Bray–Curtis NMDS showed different patterns of clustering of samples across treatment groups (Figure 2B). Indeed, bacterial community structure differs between groups (R = 0.208, p = 0.002), suggesting that microbiomes of each treatment group were distinct from one another.” We would also like to correct the statement in the discussion section to read “One of the most interesting results from our study shows that exposure to rising concentrations of E2 increases microbiota abundances but decreases alpha diversity. Additionally, community composition was distinct between treatment groups. This dose-dependent effect suggests that varying concentrations of E2 may influence microbial community composition in different ways.”

We apologize for this error and even though this result becomes statistically significant after our re-analysis, it has minimal effect on the scientific conclusions of our work.