Differential investment in visual and olfactory brain areas reflects behavioural choices in hawk moths

Abstract

Nervous tissue is one of the most metabolically expensive animal tissues, thus evolutionary investments that result in enlarged brain regions should also result in improved behavioural performance. Indeed, large-scale comparative studies in vertebrates and invertebrates have successfully linked differences in brain anatomy to differences in ecology and behaviour, but their precision can be limited by the detail of the anatomical measurements, or by only measuring behaviour indirectly. Therefore, detailed case studies are valuable complements to these investigations, and have provided important evidence linking brain structure to function in a range of higher-order behavioural traits, such as foraging experience or aggressive behaviour. Here, we show that differences in the size of both lower and higher-order sensory brain areas reflect differences in the relative importance of these senses in the foraging choices of hawk moths, as suggested by previous anatomical work in Lepidopterans. To this end we combined anatomical and behavioural quantifications of the relative importance of vision and olfaction in two closely related hawk moth species. We conclude that differences in sensory brain volume in these hawk moths can indeed be interpreted as differences in the importance of these senses for the animal's behaviour.

Figure 1. Lower-order visual and olfactory neuropils in M. stellatarum and D. elpenor.

Posterior view of 3D reconstructed neuropils of the diurnal (a) and nocturnal (b) species. Lower-order structures in colour, higher-order neuropils in grey, remaining central brain neuropil in light grey. Absolute (c) and relative (d,e) neuropil volumes of the total brain (c) and visual (d) and olfactory (e) structures, respectively. Grey (diurnal) and black (nocturnal) circles show individual measurements, shaded areas interquartile ranges and horizontal bars medians (colour coded as in a,b). Significance tests with Mann-Whitney-U test: *p < 0.05, **p < 0.01, ***p < 0.001. (f) The grade shift index (gsi) illustrates scaling differences in neuropil volume between species (positive values: M. stellatarum > D.elpenor, negative values: vice versa), while the slope ratio (si) shows a divergence of the respective neuropil from isometric scaling with respect to the central brain. Significance is indicated by the colour. For full results and statistics see Supplementary Tables S1, S2.

Figure 2. Higher-order visual and olfactory neuropils in M. stellatarum and D. elpenor.

Anterior (left) and posterior (right) view of 3D reconstructed central neuropils of the diurnal (a) and nocturnal (b) species. Higher-order sensory structures in colour, other neuropils in grey, remaining neuropil in light grey (c). Close-up of the anterior optic tubercle (visual). (d,e) Olfactory projections from the antennal lobes identified the lateral horn (d, blue line), and the olfactory input regions to the mushroom body calyx (calyx inner zone, (e). Lobula injections revealed visual input regions in the mushroom body calyx (e). Scale bars in (c–e) 100 μm. (f–h). Relative neuropil volumes of higher-order visual (f), olfactory (g) and multimodal (h) brain areas: grey (diurnal) and black (nocturnal) circles show individual measurements, shaded areas interquartile ranges and horizontal bars (colour code as in a,b) medians. Significance tests with Mann-Whitney-U test: *p < 0.05, **p < 0.01, ***p < 0.001. (i) The grade shift index (gsi) and the slope ratio (si). Colour as in Fig. 1. For full results and statistics see Supplementary Tables S1, S2.

Figure 3. The relative importance of visual and odour cues in a foraging task.

(a) Animals were trained to a rewarded combination of visual and olfactory cues (yellow and honeysuckle odour (waves)) and the unrewarded (blue and bergamot odour (arrows)) and tested with this combination (control) or a conflict between cues. (b) Choices (touching of the feeders with the proboscis, percentage of animals) when presented with the control and conflict condition, visual cues only in M. stellatarum, and olfactory cues only in D. elpenor. (c) Feeding motivation of moths scored as touching of the feeder with the proboscis out of all animals flying, normalized to the control condition for each species. Significance tests in (b) and (c), Fisher’s exact test: *p < 0.05, **p < 0.01, ***p < 0.001. For full results and statistics see Supplementary Table S3.