Infra-population and -community dynamics of the parasites Nosema apis and Nosema ceranae, and consequences for honey bee (Apis mellifera) hosts

Abstract

Nosema spp. fungal gut parasites are among myriad possible explanations for contemporary increased mortality of western honey bees (Apis mellifera, hereafter honey bee) in many regions of the world. Invasive Nosema ceranae is particularly worrisome because some evidence suggests it has greater virulence than its congener N. apis. N. ceranae appears to have recently switched hosts from Asian honey bees (Apis cerana) and now has a nearly global distribution in honey bees, apparently displacing N. apis. We examined parasite reproduction and effects of N. apis, N. ceranae, and mixed Nosema infections on honey bee hosts in laboratory experiments. Both infection intensity and honey bee mortality were significantly greater for N. ceranae than for N. apis or mixed infections; mixed infection resulted in mortality similar to N. apis parasitism and reduced spore intensity, possibly due to inter-specific competition. This is the first long-term laboratory study to demonstrate lethal consequences of N. apis and N. ceranae and mixed Nosema parasitism in honey bees, and suggests that differences in reproduction and intra-host competition may explain apparent heterogeneous exclusion of the historic parasite by the invasive species.

Figure 1. Effect of Nosema infection on mortality of adult worker western honey bees (Apis mellifera).

Mortality is shown as the cumulative percentage of dead individuals from control, Nosema apis, Nosema ceranae and mixed N. apis/N. ceranae treatments each day. The experiment was terminated at 30 d post inoculation. Treatments with different letters had significant differences in mortality.

Figure 2. Effect of Nosema infection on adult worker western honey bee (Apis mellifera) nutritional demand.

Consumption is shown as volume of 50% (weight/volume) sucrose-water mixture per bee per week post inoculation for control, Nosema apis, Nosema ceranae, and mixed N. apis/N. ceranae treatments (Week 4 included only consumption from 22–25 d post inoculation). Boxplots show interquartile range (box), median (black or white line within interquartile range), data range (dashed vertical lines), and outliers (open dots); asterisks (black or white) represent means. No significant differences were observed among treatments for daily consumption per worker.

Figure 3. Level of Nosema infection in dead adult worker western honey bees (Apis mellifera) 28–30 d post oral inoculation for control, Nosema apis, Nosema ceranae and mixed N. apis/N. ceranae treatments.

Boxplots show interquartile range (box), median (black line within interquartile range), data range (dashed vertical lines), and outliers (open dots); asterisks (black) represent means. Horizontal square parenthesis under boxplots indicates a significant difference; controls were excluded from analyses because no infections were observed.

Figure 4. Levels of Nosema apis and Nosema ceranae DNA (square root-transformed) in adult worker western honey bees (Apis mellifera) that died between 28 and 30 d post inoculation in N. apis, N. ceranae or mixed N. apis/N. ceranae treatments (same workers shown in Fig. 3 ).

Boxplots show interquartile range (box), median (black or white line within interquartile range), data range (dashed vertical lines), and outliers (open dots); asterisks (black or white) represent means. No significant differences were observed in quantities among the four instances where we expected to find DNA (i.e., the boxes with means above 0).

Figure 5. Nosema infection intensities in live-sampled adult worker western honey bees (Apis mellifera) at 7, 14, and 21 d post oral inoculation in control, Nosema apis, Nosema ceranae, and mixed N. apis/N. ceranae treatments.

Boxplots show interquartile range (box), median (black or white line within interquartile range), data range (dashed vertical lines), and outliers (open dots); asterisks (black or white) represent means. Horizontal square parenthesis under boxplots indicates a significant difference; controls were excluded from analyses because no infections were observed.