Anthropogenic noise negatively affects some animals more so than others and often in different ways. Female crickets reared in traffic noise are reported to be faster or slower to locate male song than those reared in silence depending on species. No study has considered whether observed differences were due to hearing or decision-making. We reared female Teleogryllus oceanicus in traffic noise and silence, and adult females located male song broadcast amidst traffic noise or silence. We recorded activity of two auditory interneurons in a subset of individuals under identical acoustic conditions. Regardless of rearing treatment, crickets were slower to leave their shelter when presented with male song in silence than in traffic noise, while crickets reared in traffic noise were also slower to leave overall. Crickets reared in traffic noise also had higher baseline AN2 activity, but rearing condition did not affect hearing thresholds or auditory response to male song. Our results demonstrate behavioural and auditory effects of long-term exposure to anthropogenic noise. Further, they support the idea that silence itself is a potentially aversive acoustic condition.

Behaviour 

To compare the effect of the two rearing treatments on female phonotaxis, we randomly assigned crickets at fourteen days after final moult to a playback condition of either a traffic noise or silence (speaker on but no sound) during male song playback. In the center of the experimental trial room (H: 2.3 m, W: 2.6 m, L: 4.6m), we placed a circular arena made of cork (dia.: 1.22 m, 6.25 mm thick) on the floor, the arena was marked with a circle of 1.2 m diameter, and the circle divided into 8 segments of equal arc (45 degrees; Fig. 1B). We randomly assigned two focal speakers (Ultrasonic Dynamic Speakers, Vifa, Avisoft, Berlin) to opposite sides of the arena for each trial, positioned on the dividing line between two segments. Each focal speaker was flanked by two paired speakers (Companion, Bose) 25 cm to either side of it along the circumference of the arena. All speakers faced the center of the arena.

For each trial, we placed an adult female cricket under an egg carton cup in the center of the arena and then covered the cup with a darkened plastic container to constrain the cricket. The four flanking speakers played either a 3-minute repeating segment of traffic noise, selected for its low variability in amplitude (68-78 dB SPL at the center of arena), or a silent sound file. In both acoustic treatments, we randomly selected one of the focal speakers to play the modified calling song of T. oceanicus (70 dB SPL at center of arena). The crickets were given 60 seconds to acclimatise to their physical and acoustic environment before the plastic container was removed, marking the beginning of the trial. All trials were conducted at ~21°C.

We recorded all trials at 15 frames per second using a camera (Analog CCTV Camera HD 1080P 4-in-1 Security Dome Camera) mounted 2.45 m overhead. Each trial lasted five minutes and was considered complete if the cricket reached the edge of the circle or failed to reach it by the five-minute mark. We first scored crickets based on whether they successfully reached the focal speaker playing the male song (i.e., exiting from one of the two segments on either side of the focal speaker). For crickets that exited the arena, we determined how long they took to (a) emerge from the egg carton shelter (start latency) and (b) to reach the edge of the arena (time to finish). We then calculated (c) search time, by subtracting (a) start latency from (b) time to finish. We weighed each cricket after the trial was finished to control for potential size-based variation in statistical analyses.

Auditory activity

We recorded neural responses to conspecific song under the same acoustic conditions used in the phonotaxis experiments. We first randomly selected 20 crickets used in the phonotaxis experiment, 10 of which had been reared under traffic noise and 10 under silence. To prepare crickets for neurophysiological recordings, we pinned them ventral side up to a block of modelling clay and removed the cuticle on the ventral surface between the head and thorax. This exposed the cervical connectives that contain the axons of the AN1 and AN2 auditory interneurons. We draped one of the cervical connectives over a hook-shaped tungsten recording electrode and placed a reference electrode in the abdomen. Electrodes were connected to a differential amplifier (model DP-301, Warner Instruments, USA) and the output was passed to a data acquisition board (UltrasoundGate 816H, Avisoft Bioacoustics) for digital recording on a computer.

Neurophysiological recordings were made in one of the acoustic chambers. We used the same arenas used for phonotaxis trials, the only exceptions being a lack of shelter in the center and only one pair of flanking speakers surrounding one focal speaker. We also placed a microphone (CM16, Avisoft Bioacoustics ) one meter from the center of the arena opposite the set of speakers. The microphone was connected to a different channel of the same data acquisition board as the differential amplifier. We placed the cricket neural preparation in the center of this arena, with the ear ipsilateral to the recorded connective facing towards, and 0.6 m away from, the focal speaker.

To measure AN2 thresholds and AN1 activity across frequencies (i.e. produce audiograms), we played increasingly loud sound pulses from 2-30 kHz at 2 kHz intervals from the focal speaker. For each frequency, sound pulses increased in amplitude from 30-80 dB SPL in 2 dB steps. Sound pulses were 20 ms in duration (including 1 ms rise/fall times) and broadcast at 500 ms intervals to avoid neural adaptation. Audiograms were repeated twice, in silence and with traffic noise from the flanking speakers calibrated to be 70 dB SPL at the cricket.

To measure interneuron activity in response to cricket song under the conditions crickets experienced during the phonotaxis experiment, we played T. oceanicus song from the focal speaker to the cricket neural preparations under two noise conditions: silence or traffic noise from the flanking speakers calibrated to be 70 dB SPL at the cricket. We first recorded neural activity in the background noise condition (silence or traffic noise) for one minute to establish a baseline response to the noise condition. After a minute, we began playback of the optimized T. oceanicus song from the focal speaker alongside the background noise condition for another minute.  Crickets were exposed to the two noise conditions in a random order.