Brain neurons form auditory feature detector circuit for song pattern recognition in acoustically communicating crickets.

New communication signals can evolve by sensory exploitation if signaling taps into preexisting sensory biases in receivers [1, 2]. For mate attraction, signals are typically similar to attractive environmental cues like food [3-6], which amplifies their attractiveness to mates, as opposed to aversive stimuli like predator cues. Female field crickets approach the low-frequency calling song of males, whereas they avoid high-frequency sounds like predatory bat calls [7]. In one group of crickets (Eneopterinae: Lebinthini), however, males produce exceptionally high-frequency calling songs in the range of bat calls [8], a surprising signal in the context of mate attraction. We found that female lebinthines, instead of approaching singing males, produce vibrational responses after male calls, and males track the source of vibrations to find females. We also demonstrate that field cricket species closely related to the Lebinthini show an acoustic startle response to high-frequency sounds that generates substrate vibrations similar to those produced by female lebinthine crickets. Therefore, the startle response is the most likely evolutionary origin of the female lebinthine vibrational signal. In field crickets, the brain receives activity from two auditory interneurons; AN1 tuned to male calling song controls positive phonotaxis, and AN2 tuned to high-frequency bat calls triggers negative phonotaxis [9, 10]. In lebinthine crickets, however, we found that auditory ascending neurons are only tuned to high-frequency sounds, and their tuning matches the thresholds for female vibrational signals. Our results demonstrate how sensory exploitation of anti-predator behavior can evolve into a communication system that benefits both senders and receivers.

BackgroundAuditory mate or prey localisation is central to the lifestyle of many animals and requires precise directional hearing. However, when the incident angle of sound approaches 0° azimuth, interaural time and intensity differences gradually vanish. This poses a demanding challenge to animals especially when interaural distances are small. To cope with these limitations imposed by the laws of acoustics, crickets employ a frequency tuned peripheral hearing system. Although this enhances auditory directionality the actual precision of directional hearing and phonotactic steering has never been studied in the behaviourally important frontal range.Principal FindingsHere we analysed the directionality of phonotaxis in female crickets (Gryllus bimaculatus) walking on an open-loop trackball system by measuring their steering accuracy towards male calling song presented at frontal angles of incidence. Within the range of ±30°, females reliably discriminated the side of acoustic stimulation, even when the sound source deviated by only 1° from the animal's length axis. Moreover, for angles of sound incidence between 1° and 6° the females precisely walked towards the sound source. Measuring the tympanic membrane oscillations of the front leg ears with a laser vibrometer revealed between 0° and 30° a linear increasing function of interaural amplitude differences with a slope of 0.4 dB/°. Auditory nerve recordings closely reflected these bilateral differences in afferent response latency and intensity that provide the physiological basis for precise auditory steering.ConclusionsOur experiments demonstrate that an insect hearing system based on a frequency-tuned pressure difference receiver achieves directional hyperacuity which easily rivals best directional hearing in mammals and birds. Moreover, this directional accuracy of the cricket's hearing system is reflected in the animal's phonotactic motor response.

The singing behavior of male crickets allows analyzing a central pattern generator (CPG) that was shaped by sexual selection for reliable production of species-specific communication signals. After localizing the essential ganglia for singing in Gryllus bimaculatus, we now studied the calling song CPG at the cellular level. Fictive singing was initiated by pharmacological brain stimulation. The motor pattern underlying syllables and chirps was recorded as alternating spike bursts of wing-opener and wing-closer motoneurons in a truncated wing nerve; it precisely reflected the natural calling song. During fictive singing, we intracellularly recorded and stained interneurons in thoracic and abdominal ganglia and tested their impact on the song pattern by intracellular current injections. We identified three interneurons of the metathoracic and first unfused abdominal ganglion that rhythmically de- and hyperpolarized in phase with the syllable pattern and spiked strictly before the wing-opener motoneurons. Depolarizing current injection in two of these opener interneurons caused additional rhythmic singing activity, which reliably reset the ongoing chirp rhythm. The closely intermeshing arborizations of the singing interneurons revealed the dorsal midline neuropiles of the metathoracic and three most anterior abdominal neuromeres as the anatomical location of singing pattern generation. In the same neuropiles, we also recorded several closer interneurons that rhythmically hyper- and depolarized in the syllable rhythm and spiked strictly before the wing-closer motoneurons. Some of them received pronounced inhibition at the beginning of each chirp. Hyperpolarizing current injection in the dendrite revealed postinhibitory rebound depolarization as one functional mechanism of central pattern generation in singing crickets.

Mechanoelectrical transduction of acoustic signals is the fundamental process for hearing in all ears across the animal kingdom. Here, we performed in vivo laser-vibrometric and electrophysiological measurements at the transduction site in an insect ear (Mecopoda elongata) to relate the biomechanical tonotopy along the hearing organ to the frequency tuning of the corresponding sensory cells. Our mechanical and electrophysiological map revealed a biomechanical filter process that considerably sharpens the neuronal response. We demonstrate that the channel gating, which acts on chordotonal stretch receptor neurons, is based on a mechanical directionality of the sound-induced motion. Further, anatomical studies of the transduction site support our finding of a stimulus-relevant tilt. In conclusion, we were able to show, in an insect ear, that directionality of channel gating considerably sharpens the neuronal frequency selectivity at the peripheral level and have identified a mechanism that enhances frequency discrimination in tonotopically organized ears.

The neural mechanisms underlying cricket singing behavior have been the focus of several studies, but the central pattern generator (CPG) for singing has not been localized conclusively. To test if the abdominal ganglia contribute to the singing motor pattern and to analyze if parts of the singing CPG are located in these ganglia, we systematically truncated the abdominal nerve cord of fictively singing crickets while recording the singing motor pattern from a front-wing nerve. Severing the connectives anywhere between terminal ganglion and abdominal ganglion A3 did not preclude singing, although the motor pattern became more variable and failure-prone as more ganglia were disconnected. Singing terminated immediately and permanently after transecting the connectives between the metathoracic ganglion complex and the first unfused abdominal ganglion A3. The contribution of abdominal ganglia for singing pattern generation was confirmed by intracellular interneuron recordings and current injections. During fictive singing, an ascending interneuron with its soma and dendrite in A3 depolarized rhythmically. It spiked 10 ms before the wing-opener activity and hyperpolarized in phase with the wing-closer activity. Depolarizing current injection elicited rhythmic membrane potential oscillations and spike bursts that elicited additional syllables and reliably reset the ongoing chirp rhythm. Our results disclose that the abdominal ganglion A3 is directly involved in generating the singing motor pattern, whereas the more posterior ganglia seem to provide only stabilizing feedback to the CPG circuit. Localizing the singing CPG in the anterior abdominal neuromeres now allows analyzing its circuitry at the level of identified interneurons in subsequent studies.

Males of the bushcricket bear paired titillators that are spiny genital structures supposedly functioning as copulatory courtship devices. During copulation, the male inserts its titillators into the female's genital chamber, where they rhythmically tap on the sensilla-covered dorsal surface of the genital fold. Here, we investigated the stimulatory function of male titillators during mating in Tracer backfills of presumptive mechanosensory sensilla at the female genital fold revealed a thick bundle of sensory axons entering the last unfused abdominal ganglion (AG-7). Electrophysiological recordings of abdominal nerves demonstrated that females sense mechanical stimulation at their genital fold. The mechanosensory responses, however, were largely reduced by the insecticide pymetrozine that selectively blocks scolopidia of internal chordotonal organs but not campaniform and hair sensilla on the outer cuticle surface. In mating experiments, the females showed resistance behaviours towards males with asymmetrically shortened titillators, but the resistance was largely reduced when mechanoreceptors at the female's genital fold were either pharmacologically silenced by pymetrozine or mechanically blocked by capping with UV-hardened glue. Our findings support the hypothesis that the male titillators in these bushcrickets may serve as copulatory courtship devices to mechanically stimulate the female genitalia to reduce resistance behaviour.

Crickets use their long antennae as tactile sensors. Confronted with obstacles, conspecifics, or predators, antennal contacts trigger short-latency motor responses. To reveal the neuronal pathway underlying these antennal-guided locomotory reactions we identified descending interneurons that rapidly transmit antennal-tactile information from the head to the thorax in the cricket Gryllus bimaculatus. Antennae were stimulated with forces approximating those of naturally occurring antennal contacts. Responding interneurons were individually identified by intracellular axon recordings in the pro-mesothoracic connective and subsequent tracer injection. Simultaneous with the intracellular recordings, the overall spike response in the neck connectives was recorded extracellularly to reveal the precise response-timing of each individual neuron within the collective multiunit response. Here we describe four descending brain neurons and two with the soma in the subesophageal ganglion. All antennal-touch elicited action potentials apparent in the neck connective recordings within 10 ms after antennal-contact are generated by these six interneurons. Their dendrites ramify in primary antennal-mechanosensory neuropils of the head ganglia. Each of them consistently generated action potentials in response to antennal touching and three of them responded also to different visual stimulation (light-off, movement). Their descending axons conduct action potentials with 3-5 m/s to the thoracic ganglia where they send off side branches in dorsal neuropils. Their physiological and anatomical properties qualify them as descending giant fibers in the cricket and suggest an involvement in evoking fast locomotory reactions. They form a fast-mediating cephalo-thoracic pathway for antennal-tactile information, whereas all other antennal-tactile interneurons had response latencies exceeding 40 ms.