Mild chronic exposure to pesticides alters physiological markers of honey bee health without perturbing the core gut microbiota

Abstract

Recent studies highlighted that exposure to glyphosate can affect specific members of the core gut microbiota of honey bee workers. However, in this study, bees were exposed to relatively high glyphosate concentrations. Here, we chronically exposed newly emerged honey bees to imidacloprid, glyphosate and difenoconazole, individually and in a ternary mixture, at an environmental concentration of 0.1 µg/L. We studied the effects of these exposures on the establishment of the gut microbiota, the physiological status, the longevity, and food consumption of the host. The core bacterial species were not affected by the exposure to the three pesticides. Negative effects were observed but they were restricted to few transient non-core bacterial species. However, in the absence of the core microbiota, the pesticides induced physiological disruption by directly altering the detoxification system, the antioxidant defenses, and the metabolism of the host. Our study indicates that even mild exposure to pesticides can directly alter the physiological homeostasis of newly emerged honey bees and particularly if the individuals exhibit a dysbiosis (i.e. mostly lack the core microbiota). This highlights the importance of an early establishment of a healthy gut bacterial community to strengthen the natural defenses of the honey bee against xenobiotic stressors.

Figure 1

Effects of imidacloprid, difenoconazole and glyphosate on the establishment of gut microbiota. Gut community composition of colonized (CL) and microbiota-depleted (MD) honey bees following exposure to the fungicide (difenoconazole), herbicide (glyphosate) or insecticide (imidacloprid), alone or in a ternary mixture (Mix). (A) Stacked bar plots show the relative abundance of gut bacterial genera in control and pesticide-treated honey bees. Each column represents an individual bee. (B) Boxplots of total bacterial 16S rRNA gene copies estimated by qPCR in CL and MD bees. (C) Boxplots of total bacterial 16S rRNA gene copies in control bees and in bees exposed to the different pesticides, reported separately by gut microbiota colonization treatment. (D) Principal coordinate analysis of Bray–Curtis dissimilarities based on amplicon-sequence data normalized by total bacterial loads as obtained by qPCR. (E) Venn diagram showing the distribution of the 42 ASVs that differed in abundance between pesticide treatments in permutation ANOVAs in either the microbiota-depleted group, the colonized group, or in both treatment groups. Note that not all these ASVs have significant Tukey post hoc tests in pairwise comparisons between pesticide treatments and controls. See Table S1 and S2 for further details.

Figure 2

Physiological impacts of pesticides in colonized (CL) and microbiota-depleted (MD) honey bees. For 5 days, colonized (CL) and microbiota-depleted (MD) newly emerged honey bees were fed sucrose solutions containing no pesticides (C, Control), imidacloprid (I, Insecticide), glyphosate (H, Herbicide), difenoconazole (F, Fungicide) or the ternary mixture of these pesticides (Mix) at the concentration of 0.1 µg/L in food. The impact of the exposure to pesticides on the physiology of the surviving honey bees at day five was investigated through an analysis of three common markers in the head, abdomen and midgut (GST, G6PDH and LDH) and two specific markers in the midgut (ALP and POx). ANOVA or Kruskal–Wallis tests were applied to detect significant differences between treatments. Treatments with different letters are significantly different (p < 0.05). # indicates a significant difference in the marker levels between colonized honey bees exposed to pesticides and their colonized control (CL.C). * and # indicate a significant difference in the marker levels between microbiota-depleted honey bees exposed to pesticides and their control (MD.C) (* or #: p ≤ 0.05; ** or ##: p ≤ 0.01; *** or ###: p ≤ 0.001).

Figure 3

Effects of the pesticides and gut colonization on the physiological state of honey bees. The levels of physiological markers in colonized (CL) and microbiota-depleted (MD) newly emerged honey bees exposed or not to pesticides was submitted to a cluster analysis to assess, with an integrative approach, the effect of pesticide treatments and gut colonization status on the physiological markers analyzed in the head (h), abdomen (a) and midgut (m). Colonized (CL) and microbiota-depleted (MD) honey bees were fed sucrose solutions containing no pesticides (Control), imidacloprid (Insecticide), glyphosate (Herbicide), difenoconazole (Fungicide) or the ternary mixture (Mix) at concentration of 0.1 µg/L in food. The distance measured was Euclidian distance with UPGMA as the linkage rule for clusters. Data normalization was required to convert the mean of each treatment to the rate of variation compared with the average of the control (CL.Control). The intensity of modulation is illustrated by the range of colors, with green and red indicating respectively a decrease and an increase of the mean enzymatic activity in each treatment, by comparison with the mean value in the CL.Control. Black indicates no change by comparison with the mean value of CL.Control.