Varied female and male courtship behavior facilitated the evolution of a novel sexual signal

Sexual selection can contribute to speciation when signals and preferences expressed during mate choice are coupled within groups, but come to differ across groups (generating assortative mating). When new  sexual signals evolve, it is important to investigate their roles in both  mate location and courtship contexts, as both signaling functions are critical in mate choice. In previous work, researchers identified two new male morphs (silent and purring) in Hawaiian populations of the Pacific field cricket, Teleogryllus oceanicus. These morphs likely evolved because they protect males from an acoustically orienting parasitoid, yet still obtain some reproductive success. But, it remains unknown how purring morphs function in close courtship encounters. We compared the relative success of the very recently evolved purring morph to that of the ancestral and silent morphs during courtship encounters. Purring males produce a novel courtship song and were not as successful in courtship as the ancestral type, but were mounted by females as often and as quickly as the obligately silent morph that arose and spread ~20 years ago. Purring males initiate courtship more quickly than other morphs, and females from populations where purring is common exhibit higher overall mounting rates. Thus, differences in the behavior of purring males and of females from populations where purring is common may have facilitated the origin of this novel sexual signal. We found no assortative mating between males of a given morph and females from their own population,  and so we hypothesize that multiple male types will be maintained within the species because each achieves fitness in different ways.

Methods
We collected Pacific field cricket eggs from three locations in the field (Hilo, Kalaupapa, and Wailua) and established lab colonies for each. We reared the crickets until adulthood, isolating virgin females individually upon their final molt (eclosion). To avoid the indiscriminate mate preferences that sometimes occurs in virgin female crickets, we mated females with males from the same population exactly twice before using them in a focal courtship trial.

In our focal courtship trials, we paired males and females from each population in a fully factorial design. Generally, males from Hilo are typical (producing the ancestral song), males from Kalaupapa are purring (recently evolved novel song that sounds like a cat's purr), and males from Wailua are silent (producing no song). These different song types are referred to as "morphs", and they are listed under the column "male_morph". Males were pulled from the stock population. Though females do not have morphs, they are grouped based on their population and are listed under the column "female_population".

We calculated the females age by counting the number of days between her eclosion date and her courtship trial date ("female_age" column). We measured cricket pronotums using Vernier digital calipers to the nearest 0.01mm and these are listed under the "male_pronotum" and "female_pronotum" columns. We measured the amount of time from the start of the trial until the male stridulated, or rubbed his wings together (which produces sound in typical and purring males—silent males also stridulate but do not produce sound), recorded in seconds as "latency_strid_sec". In the "mount" column we denote TRUE if the female mounted the male within 10 minutes after he stridulated and FALSE if she did not. The "latency_to_mount_sec" column is the amount of time in seconds between stridulation and mounting (calculated after the end of the trial). If we noticed female aggression towards the male, we denoted TRUE in the "female_aggression" column, and if the male courted for more than 50% of the trial, we denoted TRUE in the "male_consistent_courting" column. 

Following data collection, we cleaned our dataset by removing trials where the male did not stridulate (unless the female still mounted) and fixing any errors, such as typos or unusable trials (for example, one in which the male appeared to die and the female began to eat him). We also removed trials if the male morph, as determined by our ear, did not fit the majority morph for the population (typical from Hilo, purring from Kalaupapa, and silent from Wailua) because we wanted to represent one morph per population.