Elevated virulence of an emerging viral genotype as a driver of honeybee loss

Abstract

Emerging infectious diseases (EIDs) have contributed significantly to the current biodiversity crisis, leading to widespread epidemics and population loss. Owing to genetic variation in pathogen virulence, a complete understanding of species decline requires the accurate identification and characterization of EIDs. We explore this issue in the Western honeybee, where increasing mortality of populations in the Northern Hemisphere has caused major concern. Specifically, we investigate the importance of genetic identity of the main suspect in mortality, deformed wing virus (DWV), in driving honeybee loss. Using laboratory experiments and a systematic field survey, we demonstrate that an emerging DWV genotype (DWV-B) is more virulent than the established DWV genotype (DWV-A) and is widespread in the landscape. Furthermore, we show in a simple model that colonies infected with DWV-B collapse sooner than colonies infected with DWV-A. We also identify potential for rapid DWV evolution by revealing extensive genome-wide recombination in vivo The emergence of DWV-B in naive honeybee populations, including via recombination with DWV-A, could be of significant ecological and economic importance. Our findings emphasize that knowledge of pathogen genetic identity and diversity is critical to understanding drivers of species decline.

Keywords: decline; emerging infectious disease; pollinator; virulence.

Figure 1.

Test of DWV virulence under controlled conditions. (a) Fitted Cox proportional hazard survival curves (solid coloured lines) in days post-infection (p.i.) following exposure by injection of V. destructor-free, newly emerged adults. C = control (black); A = DWV-A (blue); B = DWV-B (green); M = mix (orange) and 95% CIs for each fitted curve (dashed coloured lines). Star/lines show significant differences between treatments (p < 0.05) based on post hoc pairwise comparisons of the final model in table 1. Median survival of control bees was 29 days, for DWV-A injected bees it was 18 days, and for DWV-B injected bees it was 13.5 days. (b) DWV-A and -B titres in A, B and M treatments from bees extracted at 9 and 13 days p.i. (c) Population dynamics over time of colonies infected with DWV-A or -B. Models were run in BEEHAVE [45] with Mite-model parameters adjusted to reflect the relative individual mortality of adults and pupae infected with either DWV-A or -B. Individual daily mortality rates were derived from laboratory experiment survival data (panel (a); described in Methods). Colony collapse events are indicated by a vertical red arrow.

Figure 2.

Prevalence of DWV-A and -B in honeybees collected during a structured sampling survey of GB (%). (a) Estimated prevalence of DWV-A (blue), DWV-B (green) and putative co-infections (orange) are indicated by variable circle areas at sampling locations. Sample sites with zero prevalence not indicated (see the electronic supplementary material; figure S3 for sample site information). (b) True prevalence estimates for each virus, including 95% CIs (%), with significant non-independence (χ² test) of the two virus genotypes indicated by an asterisk over the co-infection (orange) column.