Low doses of neonicotinoid pesticides in food rewards impair short-term olfactory memory in foraging-age honeybees

Abstract

Neonicotinoids are often applied as systemic seed treatments to crops and have reported negative impact on pollinators when they appear in floral nectar and pollen. Recently, we found that bees in a two-choice assay prefer to consume solutions containing field-relevant doses of the neonicotinoid pesticides, imidacloprid (IMD) and thiamethoxam (TMX), to sucrose alone. This suggests that neonicotinoids enhance the rewarding properties of sucrose and that low, acute doses could improve learning and memory in bees. To test this, we trained foraging-age honeybees to learn to associate floral scent with a reward containing nectar-relevant concentrations of IMD and TMX and tested their short (STM) and long-term (LTM) olfactory memories. Contrary to our predictions, we found that none of the solutions enhanced the rate of olfactory learning and some of them impaired it. In particular, the effect of 10 nM IMD was observed by the second conditioning trial and persisted 24 h later. In most other groups, exposure to IMD and TMX affected STM but not LTM. Our data show that negative impacts of low doses of IMD and TMX do not require long-term exposure and suggest that impacts of neonicotinoids on olfaction are greater than their effects on rewarding memories.

Figure 1. Bees rewarded with 1 M sucrose solution containing IMD during massed (a) and spaced (b) conditioning had a slower rate of learning.

Bees rewarded with TMX solutions were largely unaffected during massed (c) and spaced (d) conditioning. The y-axis represents the probability of eliciting a conditioned PER to the odour prior to food presentation. Each treatment group for each neonicotinoid contained 60 responding subjects, therefore the overall sample size consisted of 480 honeybees.

Figure 2. Comparisons of the bees’ conditioned PER responses on the 6^th conditioning trial to their responses at 10 min (STM) and 24 h (LTM) after training.

Bees fed with rewards containing 10 nM IMD were less likely to respond overall during massed (a) and spaced (b) conditioning. Bees fed with TMX were least likely to respond during the 10 min test and most likely to respond during the 24 h test during both massed (c) and spaced (d) conditioning. Error bars are + SE. Letters indicate comparisons of the control group to each treatment group (subscripts: 1 = 6^th trial, 2 = 10 min, 3 = 24 h). *indicates P < 0.003 (Bonferroni-adjusted critical value) for comparisons within treatment groups. Each treatment group contained 60 responding subjects in the STM test at 10 min. The sample size for the 24 h test changed overnight because some bees died. For the massed trained group at the 24 h test, the sample sizes were as follows: IMD: control = 56, 0.1 nM = 53, 1 nM = 48, 10 nM = 53; TMX: control = 58, 0.1 nM = 60, 1 nM = 57, 10 nM = 56. For the spaced-trained group at the 24 h test, the sample sizes were as follows: IMD: control = 56, 0.1 nM = 48, 1 nM = 57, 10 nM = 52; TMX: control = 53, 0.1 nM = 58, 1 nM = 58, 10 nM = 60.