Sexual signals are often transmitted through multiple modalities (e.g., visual and chemical), and under selection from both intended and unintended receivers. A nuanced understanding of sexual signal evolution is important since they play a critical role in diversification. We recently documented percussive substrate-borne vibrations in the Pacific field cricket (Teleogryllus oceanicus), a species that uses airborne acoustic and chemical signals to attract and secure mates. The airborne signals of Hawaiian T. oceanicus are currently undergoing rapid evolution; at least five novel male morphs have arisen in the past 20 years. Nothing is yet known about the newly discovered percussive substrate-borne vibrations, so we ask ‘how’ they are produced, ‘who’ produces them (e.g., population, morph), ‘when’ they produce them (e.g., whether they are plastic), and ‘why’ (e.g., do they play a role in mating). We show that the vibrations are produced exclusively by males and during courtship via foreleg drumming. One novel morph, purring, produces quieter airborne songs and is more likely to drum than the ancestral morph. However, drumming behavior is also contextually plastic for some males; when we removed the ability of males to produce airborne song, ancestral males became more likely to drum, whereas two novel morphs were equally likely to drum regardless of their ability to produce song. Opposite our prediction, females were less likely to mate with males who drummed. We discuss why that might be and describe what we can learn about complex signal evolution from this newly discovered behavior.

Between June 2018 and November 2021, we collected ~30 adult male and ~30 adult female Teleogryllus oceanicus from each of seven populations (Kapa’a and Wailua on Kaua’i, La’ie and Manoa on O’ahu, Kalaupapa on Moloka’i, Hilo on Hawai’i, and Mo’orea in French Polynesia). We used airborne audio recordings to identify males as one of three morphs: ancestral, rattling, purring. At field stations, we conducted one-on-one courtship experiments where males could produce airborne sound and when they could not (silenced). We recorded whether the male drummed, producing substrate-borne vibrations, as well as female behaviors (spreadsheet 1). The second spreadsheet uses a subset of those male where we simultaneously recorded airborne and substrate-borne audio to calculate drumming rate (seconds) per airborne courtship song.