Cockchafer larvae smell host root scents in soil

Abstract

In many insect species olfaction is a key sensory modality. However, examination of the chemical ecology of insects has focussed up to now on insects living above ground. Evidence for behavioral responses to chemical cues in the soil other than CO(2) is scarce and the role played by olfaction in the process of finding host roots below ground is not yet understood. The question of whether soil-dwelling beetle larvae can smell their host plant roots has been under debate, but proof is as yet lacking that olfactory perception of volatile compounds released by damaged host plants, as is known for insects living above ground, occurs. Here we show that soil-dwelling larvae of Melolontha hippocastani are well equipped for olfactory perception and respond electrophysiologically and behaviorally to volatiles released by damaged host-plant roots. An olfactory apparatus consisting of pore plates at the antennae and about 70 glomeruli as primary olfactory processing units indicates a highly developed olfactory system. Damage induced host plant volatiles released by oak roots such as eucalyptol and anisol are detected by larval antennae down to 5 ppbv in soil air and elicit directed movement of the larvae in natural soil towards the odor source. Our results demonstrate that plant-root volatiles are likely to be perceived by the larval olfactory system and to guide soil-dwelling white grubs through the dark below ground to their host plants. Thus, to find below-ground host plants cockchafer larvae employ mechanisms that are similar to those employed by the adult beetles flying above ground, despite strikingly different physicochemical conditions in the soil.

Figure 1. GC-MS/EAD chromatograms of CLSA-samples from oak roots.

A) undamaged oak roots The upper trace shows the electroantennographic response of a larval antenna (EAD), the lower trace shows the total ion current of the mass spectrometer (MSD). Compounds detected consistently (numbers as in Table 1): 4∶6-methyl-5-hepten-2-one, 7: furanoid trans-linalooloxide, 8: nonanal, 9/10: (R/S) camphor, 11: borneol, 12: decanal. Not consistently detected compounds are labeled by o. B) feeding damaged oak roots The upper trace shows the electroantennographic response of a larval antenna (EAD), the lower trace shows the total ion current of the mass spectrometer (MSD). Compounds detected consistently (numbers as in Table 1): 1: anisol, 3: octan-3-one, 5∶2-ethyl-hexan-1-ol, 6: eucalyptol, 7: furanoid trans-linalooloxide, 8: nonanal, 9/10: (R/S) camphor, 11: borneol. Not consistently detected compounds are labeled by o.

Figure 2. Design of an experimental unit of the dual choice bioassay.

The set-up consists of one large Petri dish (ID 14 cm) with two holes (diameter 24 mm each) in the lid, two small Petri dishes (ID 5 to 6 cm) and a cage made of steel wire (2.5 cm×1.5 cm). Larvae chose between the test compound and the control, or stayed in the neutral area (central bar of 3×14 cm, orthogonal to the connection line of the holes). First choice position in relation to the stimulus compound, the control or the neutral area was recorded.

Figure 3. Dose-response curves of M. hippocastani to different odors.

The selected test odors are released by oak-roots damaged by feeding larvae. Dose- response curves to anisol (N = 8), eucalyptol (N = 3), octan-3-one (N = 6), (1R)-camphor (N = 8), and furanoid trans-linalooloxide (N = 3); mean values of antennal responses (amplification factor 100). Lowest dilutions eliciting responses significantly different to baseline noise from at least 50% of all antennae are marked as full symbols.

Figure 4. SEM pictures of M. hippocastani apical and subapical antennomere.

A) Lateral view of the apical and sub-apical antennomere. B–D) Dorsal, apical, and ventral view of the apical antennomere, respectively. In B and D the multiporous olfactory sensilla (MOS) can easily be observed. In C the apical part of the antennomere is shown, with the dorsal (left) and ventral (right) MOS. E) MOS on the inner surface of the lateral protrusion of the subapical segment. F) Close up view of the MOS surface, pierced by numerous tiny cuticular pores. Bar scale: A, B, D: 200 µm; C:100 µm; E: 50 µm; F: 2 µm.

Figure 5. M. hippocastani apical antennomere.

A) Light microscopy cross section showing the dorsal (DMOS) and ventral MOS (VMOS). B) TEM cross section at the level of the dorsal MOS, showing two bundles of outer dendritic segments (ODS). C, D, F) Details of the dendritic branches (DB) filling the space below the porous cuticle (PC), pore tubules (PT) can also be observed. E) Two bundles of four dendrites taken at the level of the ODS. Bar scale: A: 50 µm; B: 10 µm; C, E: 2 µm; D: 500 nm; F: 200 nm.

Figure 6. M. hippocastani brain including the antennal lobes (AL), frontal views.

A) Maximum projection of 229 serial confocal images: Green codes for anti synapsin immunostaining, magenta for a dye (dextran) backfill from the antenna. B) 3D-reconstruction of A showing the brain outline (light gray) and selected brain areas: yellow, reconstructed from the antenna backfill; the other brain areas, including the contralateral AL (blue), the mushroom bodies (red), the central complex (darker green), the protocerebral bridge (lighter green), and remaining neuropil (gray) are reconstructed from the anti synapsin immunostaining which can be used to label neuropil areas in insects (see e.g. [10]). Arrowheads, projection to the subesophageal ganglion; arrows, cell bodies of two antennal motoneurons; star, projections to the lateral protocerebrum; AN, antennal nerve. C) Single confocal images of the image stack of the left antennal lobe in A, clearly showing many spheroidal structures, the so called olfactory glomeruli in the larval beetle brain. AL - labeled by the synapsin antibody (C1) and the backfill staining (C3). C2: Overlay of C1 and C3.

Figure 7. Behavioral data of M. hippocastani in dual choice tests in soil.

Numbers in the bars show the percentage of larvae orienting towards the odor source (black), of active larvae showing no decision (vertically hatched), of inactive larvae (diagonally hatched), and of larvae orienting away from the odor source (white). Numbers next to the bars indicate the total number of individuals in the different experiments. The Attractivity Index (Attr. I.) was calculated by relating the number of larvae attracted to the compound to the total number of larvae showing a decision. Statistical significance is indicated by *** (p<0.001), * (p<0.05) and n.s. (p>0.05, not significant, chi² test, α = 0.05).