Variation in honey bee gut microbial diversity affected by ontogenetic stage, age and geographic location.

Abstract

Social honey bees, Apis mellifera, host a set of distinct microbiota, which is similar across the continents and various honey bee species. Some of these bacteria, such as lactobacilli, have been linked to immunity and defence against pathogens. Pathogen defence is crucial, particularly in larval stages, as many pathogens affect the brood. However, information on larval microbiota is conflicting. Seven developmental stages and drones were sampled from 3 colonies at each of the 4 geographic locations of A. mellifera carnica, and the samples were maintained separately for analysis. We analysed the variation and abundance of important bacterial groups and taxa in the collected bees. Major bacterial groups were evaluated over the entire life of honey bee individuals, where digestive tracts of same aged bees were sampled in the course of time. The results showed that the microbial tract of 6-day-old 5th instar larvae were nearly equally rich in total microbial counts per total digestive tract weight as foraging bees, showing a high percentage of various lactobacilli (Firmicutes) and Gilliamella apicola (Gammaproteobacteria 1). However, during pupation, microbial counts were significantly reduced but recovered quickly by 6 days post-emergence. Between emergence and day 6, imago reached the highest counts of Firmicutes and Gammaproteobacteria, which then gradually declined with bee age. Redundancy analysis conducted using denaturing gradient gel electrophoresis identified bacterial species that were characteristic of each developmental stage. The results suggest that 3-day 4th instar larvae contain low microbial counts that increase 2-fold by day 6 and then decrease during pupation. Microbial succession of the imago begins soon after emergence. We found that bacterial counts do not show only yearly cycles within a colony, but vary on the individual level. Sampling and pooling adult bees or 6th day larvae may lead to high errors and variability, as both of these stages may be undergoing dynamic succession.