A sublethal dose of a neonicotinoid insecticide disrupts visual processing and collision avoidance behaviour in Locusta migratoria

Abstract

Neonicotinoids are known to affect insect navigation and vision, however the mechanisms of these effects are not fully understood. A visual motion sensitive neuron in the locust, the Descending Contralateral Movement Detector (DCMD), integrates visual information and is involved in eliciting escape behaviours. The DCMD receives coded input from the compound eyes and monosynaptically excites motorneurons involved in flight and jumping. We show that imidacloprid (IMD) impairs neural responses to visual stimuli at sublethal concentrations, and these effects are sustained two and twenty-four hours after treatment. Most significantly, IMD disrupted bursting, a coding property important for motion detection. Specifically, IMD reduced the DCMD peak firing rate within bursts at ecologically relevant doses of 10 ng/g (ng IMD per g locust body weight). Effects on DCMD firing translate to deficits in collision avoidance behaviours: exposure to 10 ng/g IMD attenuates escape manoeuvers while 100 ng/g IMD prevents the ability to fly and walk. We show that, at ecologically-relevant doses, IMD causes significant and lasting impairment of an important pathway involved with visual sensory coding and escape behaviours. These results show, for the first time, that a neonicotinoid pesticide directly impairs an important, taxonomically conserved, motion-sensitive visual network.

Figure 1

Percent mortality at 48 hours for male and female locusts after injection with imidacloprid (IMD) solutions ranging in concentrations from 10 to 10,000 ng/g (plotted on log scale), fitted with iterative non-linear regressions (solid lines). Points represent proportions from groups of 6 to 12 locusts, normalized to the mortality of vehicle control groups used on each testing day. Regression lines revealed an LD50 of approximately 2,500 ng/g for males. The LD50 for females was beyond the highest concentration used.

Figure 2

Effect of IMD on escape behaviours. (a) Proportion of animals responding to a looming stimulus by jumping or twitching hind legs at 2 and 24 hours after treatment with the vehicle, or 10 ng/g or 100 ng/g IMD. (b) Proportion of animals responding to looming stimulus while flying (R), flying but not responding to stimulus (NR), or not flying (NF) at 2 (left columns) and 24 (right columns) hours after treatment with vehicle or IMD.

Figure 3

Joint inter-spike interval (ISI) distribution heatmaps comparing one ISI (y-axis, ms) with the following ISI (x-axis, ms). Heatmaps in column 1 were compiled from 5 stimulus presentations per animal either before treatment (1A,B, n = 25), or 24 hours after treatment with the vehicle (1C,D, n = 10). 2–3A,B were recorded from 80 to 110 minutes after treatment with 10 ng/g (n = 5) or 100 ng/g (n = 20) IMD. 2–3C,D were recorded 24 hours after treatment with 10 ng/g (2C,D, n = 10) and 100 ng/g (3C,D, n = 10).

Figure 4

Rasters and peristimulus time histograms (PSTHs) constructed from DCMD responses to looming stimuli. (a) Raw continuous data from an extracellular recording of the left ventral nerve cord during presentation with the image of a 7 cm looming disk to the right eye (top). DCMD spikes were easily differentiated by their large amplitude. The projected time of collision (TOC) is marked with a vertical red line. Bursts, which comprise a minimum of two spikes occurring within 8 ms of each other, are highlighted with a vertical green line to signal the start, and horizontal lines to show the duration. PSTHs show the firing rates of the full DCMD response, spikes within bursts only, isolated spikes only, and bursts, smoothed with a 50 ms Gaussian filter. PSTHs response profile parameters included: peak firing rate (f _p) and the time of the peak relative to TOC (pt), denoted by an asterisk; peak width at half maximum (PW½M), rise phase, from the last time the histogram crosses the 99% confidence interval with a positive slope (t99) to the peak, and decay phase from the peak until it had decreased to 15% (t15). (b–e) Show rasters (top) and PSTHs (bottom) at four time points after injection with 100 ng/g IMD, 5, 17.5, 25, and 110 minutes. These panels are separated by rate type: full DCMD (b), burst spikes only (c), isolated spikes only (d), and bursts (e).

Figure 5

Response parameters (mean ± SEM) of the four DCMD rates (full DCMD, burst spikes, isolated spikes, and bursts) at 110 (S background) and 120 minutes (FF background) after treatment with an IMD dose (ng/g) or the vehicle (V). Data plotted as a percent of pre-treatment values within each animal. Grey backgrounds indicate a significant effect of one or more IMD dose for stimuli presented over both S and FF backgrounds, yellow indicates a significant effect within FF background only, and green for S background only. Significant results of post hoc analyses are indicated with letters colour coded to the respective background type.

Figure 6

Response parameters (mean ± SEM) of the four DCMD rates (full DCMD, burst spikes, isolated spikes, and bursts) 24 hours after treatment with an IMD dose (ng/g) or the vehicle. Gray backgrounds indicate a significant effect of one or more IMD dose for stimuli presented against both S and FF backgrounds, and green indicates a significant effect with S background only. Significant results of post hoc analyses are indicated with letters colour coded to the respective background type.

Figure 7

Time of maximum firing rate relative to collision (peak time; mean ± SEM) across doses and background types at 2 and 24 hours after treatment. Significant results of post hoc analyses are indicated with letters colour coded to the respective background type.

Figure 8

Effect of IMD on burst spike peak firing rate (f _p) and correlation with observed behavioural deficits. (a) Burst spike peak f _p at 2 and 24 hours after treatment, for stimuli presented against simple (S) and flow field (FF) backgrounds. (b) Burst spike f _p correlated with flight behaviour at 24 hours after treatment with IMD or vehicle: NF, not flying; NR, flying but not responding; R, responding with a turn or glide, with a Pearson correlation coefficient of 0.872. Each data point represents measurements from a single animal, and letters denote significant differences between means. (c) Percent inhibition with effects at 2 and 24 hours combined for each background type, plotted as percent reduction from the mean of the vehicle treatment group. Colour coded letters denote significant differences between treatments within each background, and asterisks denote significant differences between S and FF within treatment.