Insects and Insecticides

Written evidence submitted by Dr Robert Paxton

SUMMARY

• Drs Vincent Doublet and Robert Paxton of Queen’s University Belfast/Martin-Luther-University Halle-Wittenberg have undertaken laboratory experiments on interactions between a neonicotinoid insecticide, thiacloprid, and pathogens for juvenile honey bee health.

• Both viruses and pesticides have a detrimental effect on honey bee brood development and survival.

• When viruses and pesticides are experimentally administered simultaneously to honey bee larvae at sub-lethal doses, they interact additively and sometimes synergistically, hindering larval development and enhancing larval/pupal mortality.

Reporting text:

As part of the BBSRC (Insect Pollinators Initiative) project ‘Impact and mitigation of emergent diseases on major UK insect pollinators’ (BB/l000100/1) and the EU funded research project BeeDoc (Bees in Europe and the Decline of Honeybee Colonies; 244956 CP-FP), Drs Vincent Doublet and Robert Paxton of Queen’s University Belfast/Martin-Luther-University Halle-Wittenberg have undertaken laboratory experiments on the interactions between a neonicotinoid insecticide, thiacloprid, and pathogens for honey bee health.

Our aim was to examine experimentally and in vitro how viral infection and pesticides affect individual larval and pupal bees, and the interactions between viruses and pesticides, so as to identify the main ‘driving processes’ that cause honey bee mortality.

This research has become all the more relevant because two recent papers have highlighted the role of neonicotinoid pesticides, systemic plant insecticides of growing importance to agriculture, in bee mortality (Henry et al. 2012; Whitehorn et al. 2012). Other recent papers have also suggested a major role for pesticides, both neonicotinoids and acaricides commonly used by beekeepers to control V. destructor mites inside the hive, in exacerbating the effects of honey bee pathogens (Alaux et al. 2010; Vidau et al. 2011; Locke et al. 2012; Pettis et al. 2012).

This report details our research aimed at uncovering if and how two pesticides interact with the commonest viral pathogen of honey bees transmitted by V. destructor mites, deformed wing virus (DWV), to cause brood mortality and other developmental aberrations. As pesticides, we employed: (i) t-fluvalinate, a synthetic pyrethroid commonly used by beekeepers inside the hive to kill V. destructor mites; and (ii) thiacloprid, a neonicotinoid commonly sprayed on oilseed rape and the commonest of this class of insecticide found as a residue inside European beehives. In addition to DWV, we also extended our analyses to examine the effects of the second most common virus in honey bees, black queen cell virus (BQCV), and its interactions with pesticides for honey bee health.

Experimental Protocol

To examine the interaction between pesticides and pathogens, we inject DWV into white-eyed pupae as our DWV treatment. We also undertook a series of parallel experiments in which we fed DWV to larvae on day 2 of larval age as our means of DWV treatment. This had the advantage that DWV is naturally acquired by feeding and its natural site of infection is likely the alimentary canal (ventriculus) of bees. This treatment therefore adds an extra dimension to our experiments on the interactions between DWV and pesticides for honey bee health.

We additionally investigated the impact of BQCV on honey bee larval/pupal health in a further set of replicate experiments. In this case, we fed BQCV directly to 2-day old larvae. BQCV is relatively stable, compared to DWV, facilitating its experimental manipulation and use.

For all experiments described herein, we employed standard methods for honey bee larval/pupal rearing, as described in Aupinel et al. (2007). In short, honey bee eggs in brood combs were transferred to a 340C incubator at 95% relative humidity. As they hatched, eggs were transferred to individual wells of a 48 well microtitre plate and kept in the same conditions as described above. For each treatment (including each control treatment), we used 48 larvae/pupae per treatment. We replicated entire experiments 3 times using honey bees derived from 3 colonies i.e. each replicate used bees from one colony (total 154 larvae/pupae per treatment). A statistical power analysis suggested that these sample sizes would allow us to detect more subtle effects of pesticide-viral treatments than would otherwise have been the case. Mortality of larvae was recorded every day.

After entering the prepupal stage one week after hatching from the egg (see Fig. 1), microtitre plates were held at 350C and 80% relative humidity till the start of the pupal stage (see Fig. 1). Pupation success and mortality were recorded through to the end of pupal development and emergence of adults.

RESULTS

Experiment 1. Virus (BQCV) + neonicotinoid (thiacloprid) fed to honey bee larvae.

Figure 1 shows the % mortality of larvae/pupae when fed different doses of BQCV two days after hatching and transfer to 48 well microtitre plates. On the upper part of the figure we also give the developmental stage of honey bees to help interpretation. Figure 1 shows that a quantity of 109 BQCV causes high mortality. Lower doses of BQCV have no effect on larval/pupal mortality.

In Figure 2, we see the effects of BQCV on development (pupation success). In this case, 109 BQCV causes high developmental abnormality (lack of pupation); 107 BQCV causes moderate developmental abnormality (reduced pupation success); 104 BQCV does not cause developmental abnormality (pupation success is as good as control bees).

Varying doses of t-fluvalinate and thiacloprid were fed directly to larvae across the entire larval period (5 days). In summary, we found sublethal doses of these two pesticides to be:

t-fluvalinate: 1 mg/kg larval food

thiacloprid: 0.1 mg/kg larval food

and we used these concentrations in further experiment, both with BQCV and DWV.

In Figure 3 we show the effect of t-fluvalinate, thiacloprid, 109 BQCV and interactions among the three on larval mortality when one or other pesticide is administered with BQCV. There is a clear additive effect of a pesticide with BQCV on larval mortality. If lower doses of BQCV are used in treatments instead of 109 BQCV, there is a corresponding drop in larval mortality, as seen in Figure 1, with little additional effect of pesticide + BQCV on larval mortality beyond treatment with either pesticide or BQCV alone (Figure 4).

In Figure 5, we see a similar response of pupae (successful pupation) to treatment with BQCV and pesticides as we saw with respect to larvae and larval survival. In essence, both BQCV and pesticides reduce pupation success, and they seem to act additively. Additivity is particularly marked for the treatment 107 BQCV + thiacloprid (Figure 5).

Experiment 2. Virus (DWV) + neonicotinoid (thiacloprid) fed to honey bee larvae.

Figure 6 shows the % mortality of larvae/pupae when fed different doses of DWV two days after hatching and transfer to 48 well microtitre plates. A quantity of 109 DWV causes high mortality. Lower doses of DWV have no effect on larval/pupal mortality.

In Figure 7, we see the effects of DWV on development (pupation success). In this case, 109 DWV causes high developmental abnormality (lack of pupation); 107 DWV causes moderate developmental abnormality (reduced pupation success); 104 DWV does not cause developmental abnormality (pupation success is as good as control bees). We note that controls for this experiment exhibited slightly elevated mortality.

As explained above, t-fluvalinate and thiacloprid were fed directly to larvae (t-fluvalinate: 1 mg/kg larval food and thiacloprid: 0.1 mg/kg larval food). In Figure 8 we show the effect of t-fluvalinate, thiacloprid, 109 DWV and interactions among the three on larval mortality when one or other pesticide is administered with DWV. There is a clear effect of a pesticide with DWV on larval mortality, and the data suggest the effect is synergistic (more than additive) in relation to DWV + either pesticide. If lower doses of DWV are used in treatments instead of 109 DWV, there is a correspondingly lower larval mortality, as seen in Figure 6, with no effect of pesticide + DWV on larval mortality beyond treatment with either pesticide or DWV alone i.e. additive effect, if at all and interactive effect (Figure 9).

In Figure 10, we see a similar response of pupae (successful pupation) to treatment with DWV and pesticides as we saw with respect to the DWV treatment of larvae and larval survival. In essence, both DWV and pesticides reduce pupation success, and they seem to act additively. Additivity is particularly marked for the treatment 107 DWV + t-fluvalinate.

Experiment 3. Virus (DWV) injected into + neonicotinoid (thiacloprid) fed to honey bee pupae.

Figure 11 shows the frequency of honey bees with wing deformity after emergence when injected with 103 viral particles of DWV and fed with or without pesticides during larval development. Pupae were injected at day 11 post-hatching. Honey bees were considered as emerged when ready to walk out of the experimental chamber (rearing plate). All treatments, including the injection of 103 viral particles of DWV, led to a high frequency of honey bees with deformed wings compared to treatments where bees were injected with a control solution. The effect of pesticides on the frequency of wing deformity when bees were injected DWV is low, though beyond that of controls. The interaction between DWV and pesticide is generally additive and never synergistic or multiplicative.

Conclusions with respect to the neonicotinoid: thiaclporid

BQCV and DWV have profound effects on their hosts, developing honey bee larvae, causing developmental abnormalities and mortality with increasing pathogen loads. A neonicotinoid pesticide (thiacloprid), when experimentally administered at sub-lethal doses to larvae or pupae, generally interacted additively with these two viruses, DWV and BQCV, to elevate mortality and developmental abnormalities. There is even a potentially synergistic interaction between DWV and the pesticide when the virus is fed at high but biologically realistic doses to larvae.

30 October 2012

Prepared 5th November 2012