Most praying mantids possess a single tympanal ear located in the ventral midline between the metathoracic legs. The auditory system is generally most sensitive to ultrasound in the 25–50 kHz range. Flying males exhibit a short-latency, stereotyped, multi-component response to ultrasound that allows them to escape from attacking bats. This study describes a small subset of species that differs in three major respects from the majority of mantis species: (1) their auditory tuning is 1.5–2 times broader; (2) they are sensitive to frequencies above 60 kHz (up to 130 kHz in some species) with thresholds as low or lower than at 25–50 kHz; (3) the behavioural response of the broadly tuned (BT) species includes 10–50 times more flight cessations and can be far less stereotyped, i.e. more ‘evitable’, than that of narrowly tuned (NT) species. However, BT species do not differ from NT species in overall sensitivity. Two species from one subfamily, the Amelinae (family Mantidae), stand out because they are among the least sensitive of any hearing mantids so far tested. Although the two amelines differ from one another in tuning curve shape, they are both more broadly tuned than most mantids. The occurrence of BT species does not follow any obvious phylogenetic pattern; they are patchily distributed among the mantis families, and both BT and NT species can be found in the same subfamily or tribe. We suggest that BT species are responding to a shared ecological pressure. Based on their tuning, the nature of their behavioural response, and their geographic distribution, we hypothesize that high duty cycle (HDC) bats (Rhinolophidae and Hipposideridae) pose a special danger to BT mantids in addition to the threat that all flying mantids face from the more common and widely distributed low duty cycle (LDC) bats.

This study examines the behavioural responses to ultrasound in 52 species from the family Cicindelidae using both tethered flight and non-flight assays. Tethered flying tiger beetles respond to trains of bat-like ultrasonic pulses with a short-latency, multi-component behaviour. There was no variation in the nature of the behavioural responses regardless of geographical distribution or phylogenetic position. Lowest mean behavioural thresholds lie predominantly between 30 and 45 kHz. Sensitivity, however, varies widely, and several species do not respond at all in these assays. The lowest thresholds for responders are most often between 70 and 80 dB SPL. Almost all subgenera of North American tiger beetles in the genus Cicindela have at least some species with low-threshold acoustic behaviour. The single exception is the large subgenus Cicindela where all species are completely unresponsive or have very high thresholds. There was little relationship between habitat and responsiveness to ultrasound, but there is a strong correlation with seasonal activity – species with adults active in the spring and autumn (all in the subgenus Cicindela) do not show ultrasound-triggered behaviour whereas summer-active species generally do. Superimposing these data on a current phylogeny of the North American tiger beetles suggests that acoustic behaviour (and hearing) is a shared primitive trait among the taxa examined here and that there have been at least five independent losses of this character.