Global similarity, and some key differences, in the metagenomes of Swedish varroa-surviving and varroa-susceptible honeybees

Abstract

There is increasing evidence that honeybees (Apis mellifera L.) can adapt naturally to survive Varroa destructor, the primary cause of colony mortality world-wide. Most of the adaptive traits of naturally varroa-surviving honeybees concern varroa reproduction. Here we investigate whether factors in the honeybee metagenome also contribute to this survival. The quantitative and qualitative composition of the bacterial and viral metagenome fluctuated greatly during the active season, but with little overall difference between varroa-surviving and varroa-susceptible colonies. The main exceptions were Bartonella apis and sacbrood virus, particularly during early spring and autumn. Bombella apis was also strongly associated with early and late season, though equally for all colonies. All three affect colony protein management and metabolism. Lake Sinai virus was more abundant in varroa-surviving colonies during the summer. Lake Sinai virus and deformed wing virus also showed a tendency towards seasonal genetic change, but without any distinction between varroa-surviving and varroa-susceptible colonies. Whether the changes in these taxa contribute to survival or reflect demographic differences between the colonies (or both) remains unclear.

Figure 1

Bacterial community composition. Distribution of the most common bacterial OTUs in the MR and MS colonies during the 2015 season. Each bar represents the mean relative abundance of reads from the 16S rRNA V2 hypervariable region assigned to different bacterial OTUs, averaged over the six colonies in each of the MR (red) and MS (blue) genetic groups. (A) Displays the full bacterial composition, including both core and minor taxa, with the names of the most abundant taxa shown beside their stacked segment. (B) Highlights the relative contribution of just the 11 minor taxa, accounting for < 1.5% of the total bacterial compositiion. The numbers beside each stacked histogram indicate how many of the six colonies in the MR or MS groups contributed reads for each of the minor taxa (identified by their first two letters) while the legend shows the bacterial taxa in the order in which they are stacked in the histogram, identified by colour.

Figure 2

Clustering analysis bacterial composition. Two-dimensional NMDS ordinations clustering the bacterial community structures of the six MR and six MS colonies for each of the five sampling occasions. Analysis of similarity (ANOSIM) revealed statistically distinct (P = 0.0001) communities of bacteria for the different sampling occasions, but with overlapping structures for the MR (red) and MS (blue) colonies at each sampling occasion (increasingly darker shades of colour from April to October).

Figure 3

Bacterial community diversity analyses. Shannon H-index estimates for bacterial community diversity in the MR (red) and MS (blue) colonies between April and October 2015. The size of the circles represents the standard deviation for the estimate at the nodes. Similar graphs were obtained with other diversity estimates.

Figure 4

Virus titre distribution. Virus titres in adult bees from April–October 2015 season for colonies in the MR and MS honeybee colonies, as determined by RT-qPCR (line drawings, averages and confidence intervals) and RNA sequencing (histograms). Asterisks indicate that the virus titre difference between the MR and MS populations at that sampling occasion was significant at P < 0.05, as determined by Welch's t-test. The RT-qPCR titres are marked on the primary Y-axis (left) while the genome equivalents of the read-count data are marked in the secondary Y-axis (right). The sixth panel shows the development of the varroa phoretic infestation rates of the MR and MS colonies between October 2014 and April 2016, together with the standard deviations.

Figure 5

Phylogenetic analyses viral genomes. Phylogenetic analyses of the DWV, SBV, BQCV, LSV and ARV-1 consensus sequences for the MR colonies (red) and MS colonies (blue) from April to October 2015, with the seasonal sequences represented by increasingly darker shades of colour.