Figure 1.
The two playback experimental setups.
In the first set of playback bioassays (Type 1) the sounds emitted by Myrmica scabrinodis queens and workers and pre- and post-adoption Maculinea alcon and M. teleius larvae (stimuli) were tested against a white noise (control). In the second experimental setup (Type 2) we tested simultaneously two acoustic stimuli: the sounds of pre-adoption larvae of the two parasite species (M. alcon pre vs. M. teleius pre), the acoustic emissions produced by integrated parasite larvae (M. alcon post- vs. M. teleius post-adoption) and the sound of M. scabrinodis castes (Queen vs. Worker).
Figure 2.
Sound emissions of Maculinea larvae and Myrmica ants.
Example waveforms (upper traces) and spectrograms (lower traces) of sounds emitted by pre-adoption and post-adoption larvae of the two parasites (Maculinea alcon and M. teleius) and stridulations produced by Myrmica scabrinodis queens and workers. Mean ± SD of the four pulse parameters are also reported for each insect category.
Figure 3.
Principal Components Analysis (PCA) on the four sound parameters.
(a) Two-dimensional plot of the first two factors extracted by principal components analysis over all individual measurements of the four sound parameters (peak frequency, peak power, IQRBW and pulse length) for each insect category - Myrmica scabrinodis queens and workers and Maculinea alcon and M. teleius pre- and post-adoption caterpillars. Ellipses indicate 95% confidence intervals; squares show the centroids for each category. (b) The component loadings extracted by PCA from the four sound parameters are reported in the table. ANOVA based on the 6 groups of samples are also reported.
Table 1.
Average Euclidean distances between sounds emitted by parasite larvae and ants.
Figure 4.
Worker ant reactions to sound stimuli and white noise.
When the sound stimuli (color bars) were played simultaneously against the white noise (white bars), they always elicited higher behavioural responses on worker ants (experimental setup Type 1). Comparing the sound stimuli, we found that stridulations produced by queens caused stronger reactions in M. scabrinodis ants than those emitted by workers (red vs. yellow bars). Workers reacted more frequently to sounds produced by M. alcon integrated larvae than those emitted by M. alcon pre-adoption caterpillars (dark green vs. pale green) while on the contrary, the sounds emitted by M. teleius pre-adoption larvae caused more frequent reactions in workers than those produced by post-adoption larvae (light blue vs. dark blue) significantly for “antennating”. Different letters indicate significantly different behavioural responses elicited by sound stimuli (Chi square Yates' correction). Significantly (p<0.05) different behavioural frequencies between white noises and sound stimuli are indicated by asterisks. ns = statistically not significant.