Bird song is crucial for attracting mates and defending territories, but different types of song or different singing behaviours may be involved in acquiring or maintaining each resource. Furthermore, male songbirds may adjust when and where they sing throughout the breeding season, depending on their breeding stage. However, such relationships remain untested in several avian taxa. Here, we studied male Bermuda White-eyed Vireos (Vireo griseus bermudianus), a passerine with two distinct song types (discrete and rambling), to test the mate attraction and territory defence hypotheses. We compare song production and song perch height among different stages of the breeding season and during the non-breeding season. We show that male vireos produce both song types during the breeding and non-breeding seasons, suggesting dual roles in mate choice and territorial defence. Song production did not differ significantly between the breeding and non-breeding seasons, but, within the breeding season, males without nesting duties sang significantly more songs than males with nesting duties. Song perch height was higher during the breeding season versus non-breeding season, among males without nesting duties compared to males with nesting duties, and when males produced discrete versus rambling songs. Our findings suggest that male vireos increase their conspicuousness to prospecting females by increasing song production and song perch height, and that they sing during the breeding and non-breeding seasons to defend year-round territories. Collectively, our study supports the mate attraction and territory defence hypotheses of bird song.

We audio-recorded the daytime (0700–1200 h) singing behaviour of 12 marked male birds throughout most of the subspecies' breeding season (April 2019 – August 2019). Of these 12 males, we found and re-recorded 11 during the following non-breeding season (December 2019 – January 2020). As part of our sampling regime, we visited one of our two sites each day during favorable weather (i.e., no rain and little to no wind), alternating between sites each day. In total, we visited the Ferry Reach Park site 41 times during the breeding season and 5 times during the nonbreeding season. We visited the Spittal Pond site 44 times during the breeding season and 7 times during the nonbreeding season. While at a site, we recorded each male at the site in a random order during a separate 15-minute recording session throughout the morning.

Upon arriving at a subject's territory, we searched for him for ≤ 15 min. If we found him, we waited 2 min before commencing recording. If we did not see or hear the focal male after 15 min, we stood in the approximate center of his territory, waited an additional 2 min, and commenced recording. If a vireo began singing from what we thought was the inside of his territory, we immediately approached him while recording. If we located the singing male and confirmed that he was our focal subject, we included in our analysis all the songs recorded throughout the 15-min session, including those acquired before visually locating him. In the rare instances when the singing male we located was not the focal subject (e.g., a neighbour), we aborted the recording session and repeated it later that day.

We recorded subjects throughout their 15-min session with a digital audio recorder (Marantz PMD661 MK II Professional recorder; WAVE format; 44.1 kHz; 16 bits) and a shotgun microphone (Sennheiser ME66 with K6 power module; super cardioid pickup pattern; 40-20,000 Hz frequency response (± 2.5 dB)) fitted with a foam windscreen. Recordings were made by following the subject no closer than 5 m while pointing the microphone directly at him (or towards the source of the songs if we had not yet located him). For each song produced while the subject was visible, we spoke into the microphone and visually estimated his song perch height above the ground (estimated accuracy ± 1 m); very few trees across Bermuda Vireo territories were > 10 m; all height estimates were made by the same person. We noted any periods in which we lost visual contact with the subject, but always continued recording until the 15-min session expired.

During the breeding season (April 2019 – August 2019), we revisited subjects in the afternoons (1300–1700 h) to document their breeding activities. We spent a maximum of 30 min searching for a given male and usually found them on account of their loud vocalizations and small territories. We followed located males at a minimum distance of 5 m and categorized them into one of six breeding stages: (1) no nesting duties, (2) nest building, (3) egg stage, (4) nestling care, (5) fledgling care, or (6) non-breeding (non-breeding status was assumed for all males between December 2019 and January 2020). We defined nest building as the stage when vireos are adding materials to a nesting branch until a nest is completed. We defined the egg stage as the period after nest completion, when the vireos are engaged in egg laying and incubation; the egg stage terminates at hatching. Nestling care is the stage when nestlings are seen inside the nest cup and the parents are actively feeding or brooding them. Fledgling care is the stage when the young are outside the nest and being fed by their parents. During the breeding season, we considered males to have nesting duties if they were engaged in nest building, the egg stage, nestling care, or fledgling feeding, and to be without nesting duties if they were not engaged in any of the above nesting behaviours with a female.

We generated a waveform and spectrogram (Hamming window, FFT = 512 samples, 87.5% overlap) for all 15-min recordings using Raven Pro sound analysis software (v1.5; Cornell Laboratory of Ornithology, Ithaca, NY). On each spectrogram, we drew cursor boxes (hereafter, “annotated”) around vireo songs that were visible on the spectrogram and waveform (i.e., clear pulses in amplitude). In some instances, songs from non-focal males could be seen and heard in the background of the recording, but these were easily distinguished from the subject's songs either because they were relatively faint or because they did not match the known vocal repertoire of the subject. We defined songs as vocalizations comprising one or more elements, where elements of the same song are separated by < 0.5 s and those of different songs are separated by ≥ 0.5 s.

The R language and environment for statistical computing and graphics.