Multimodal interaction in the insect brain

Abstract

Background: The magnitude of multimodal enhancement in the brain is believed to depend on the stimulus intensity and timing. Such an effect has been found in many species, but has not been previously investigated in insects.

Results: We investigated the responses to multimodal stimuli consisting of an odour and a colour in the antennal lobe and mushroom body of the moth Manduca sexta. The mushroom body shows enhanced responses for multimodal stimuli consisting of a general flower odour and a blue colour. No such effect was seen for a bergamot odour. The enhancement shows an inverse effectiveness where the responses to weaker multimodal stimuli are amplified more than those to stronger stimuli. Furthermore, the enhancement depends on the precise timing of the two stimulus components.

Conclusions: Insect multimodal processing show both the principle of inverse effectiveness and the existence of an optimal temporal window.

Keywords: Inverse effeciveness; Moth; Multmodal interaction; Superadditivity; Temporal window.

Fig. 1

Responses of the antennal lobe to different odour concentrations (BM) with and without the colour stimulus (n $=$ 18). Error bars show standard error of mean

Fig. 2

Responses of the mushroom body to different odour concentrations (BM) with and without the colour stimulus (n $=$ 21). Error bars show standard error of mean

Fig. 3

Responses of the antennal lobe to different odour concentrations (PAA) with and without the colour stimulus (n $=$ 9). Error bars show standard error of mean

Fig. 4

Activity patterns in AL example of the activity pattern in the antennal lobe with different odour concentration. Top Multimodal stimulus. Bottom Only odour

Fig. 5

Activity patterns in MB Example of the activity pattern in the mushroom body with different odour concentrations. Top Multimodal stimulus. Bottom Only odour

Fig. 6

Responses of the mushroom body to different odour concentrations (PAA) with and without the colour stimulus (n $=$ 22). Error bars show standard error of mean

Fig. 7

Inverse effectiveness of multimodal interaction in the mushroom body The multimodal enhancement is larger for the lower odour concentrations (n $=$ 22)

Fig. 8

No responses to visual stimulus. The presentation of a visual stimulus (V) does not elicit a measurable response in the MB when compared to not stimulus at all (N) (n $=$ 94)

Fig. 9

Response peak latency density plot of the latency of the response peak for three odour concentrations relative to the start of each trial for the data set in Fig. 6 (red multmodal, green odour only). There are no differences in the respons timing for any of the odour concentrations

Fig. 10

Stimulus timing five different temporal relations between the visual and the odour components of the multimodal stimulus stimulus used in the final experiment (a–e) and the blank stimulus (f)

Fig. 11

Temporal window the temporal window for multimodal interaction in the mushroom body for the highest (blue) and lowest (green) odour concentration (n $=$ 122). For both concentrations, the response was significantly enhanced, compared to the timing A, when the visual stimulus component coincided ($\mathrm{\Delta }t=0$ s) or came slightly after ($\mathrm{\Delta }t=0.5$ s) the odour. For the higher concentration, there was also an significant increase in the response when the visual stimulus preceded the odour with half a second ($\mathrm{\Delta }t=-0.5$ s). The different temporal relations are shown in Fig. 10. Error bars show standard error of mean