Anthropogenic noise, particularly traffic noise, disrupts animal communication. While electric vehicles (EVs) have been assumed to be quieter than gasoline vehicles (GVs), the relative impacts of EV and GV noise on wildlife remain unclear. This study examined how EV and GV noise affect the behavioral responses of Asian tits (Parus cinereus) to their recruitment calls, considering road surface and driving speed. We hypothesized that Asian tits would respond more quickly and approach more closely to the calls under EV noise, and this effect would become clearer at slower speeds and on smooth surfaces.

Using the eight types of traffic noise, which are the combinations of EV/GV, speeds (20/60 km/h), and road surfaces (rough/smooth), we conducted playback experiments on free-ranging tits in Tomakomai Experimental Forest. In each experimental trial, we broadcasted one of the noises with recruitment calls to a focal tit and observed the presence/absence of approach, latency to approach, minimum distance to the playback speaker, and presence/absence of vocal response.

The results showed that EVs significantly reduced noise only at 60 km/h on smooth surfaces, and this reduction was effective in lowering noise impacts on tits’ behaviors. Unexpectedly, however, the effects of noise on tits’ behaviors did not differ between EV and GV in other cases; for example, whereas decreased GV noise owing to reduced speeds resulted in shorter minimum distances to the speaker, such speed-dependent reduction of noise effects was not detected for EVs. Importantly, EV noise at 20 km/h delayed the approach and increased the distance to the speaker to a similar extent as GV noise at 60 km/h.

To clarify this negative effect of EVs, we conducted additional experiments using two edited noises specific to EVs at 20 km/h: one containing only the high-frequency (~10 kHz) band and another without this band. The EV noise without ~10 kHz band improved the latency to approach but not the minimum distance and ~10 kHz band noise did not fully explain the negative effects on both responses, suggesting that the impact of EVs at 20 km/h can partially be explained by the EV-specific high-frequency noise components but was attributed to being overall acoustic properties of EVs such as increased distractions or avoidance of unfamiliar high-frequency sounds to the focal species. Our study highlights the need for another mitigation considering sound frequencies, not just sound pressure levels, for future traffic noise pollution.