Summary
We investigated the distinctiveness of males' song and the relationship between song variables and females' mate choice in the Golden-winged Warbler (Vermivora chrysoptera). We recorded the songs of 23 males and quantified them to examine performance-related traits and specific song content. We carried out discriminant function analysis to compare variation within and among males for characteristics of both song types I and II. We used microsatellite DNA to identify the parents of 62 nestlings from 17 nests. By assigning paternity to males in our population we could examine relationships between females' choice and variation in males' song by comparing characteristics of within- and extra-pair sires. While we were able to discriminate statistically between individual males' songs, we found no significant differences in song variables between the within- and extra-pair males at a given nest. Therefore, female Golden-winged Warblers do not appear to be using these differences in song in choosing a mate. Further research on the Golden-winged Warbler is needed to understand song function and to gain insight into this species' reproductive biology, both of which could have important conservation implications for this declining species.
Methodology
We conducted field work at the Queen's University Biological Station near Chaffey's Locks, Ontario (44 ° 34′ N, 76 ° 19′ W), from 28 April to 15 July 2005. The study population consisted predominantly of phenotypic Golden-winged Warblers; however, Blue-winged Warblers and hybrids have been observed in the area over the past 15 years (R. J. Robertson and T. Demmons, unpubl. data; Weir 1989), and the first Bluewinged Warbler nest was found in 2005 (R. Vallender, unpubl. data). Despite a short history of hybridization in this region, Vallender et al. (2007b) found that as many as 33% of phenotypic Golden-winged Warblers there are genetically introgressed, providing evidence for significant levels of cryptic hybridization. However, given that in this area the reproductive success of hybrids equals that of Golden-winged Warblers (Vallender et al. 2007a), and also because in songbirds song can be learned (Lachlan and Slater 1999), we did not take individual genetic status into account during our analyses.
We captured adult males by target mist netting using audio playback and a painted decoy, adult females by flushing them off the nest into a mist net after day 6 of incubation. In total, we captured, banded, and released 53 males and 23 females. All adults were banded with an aluminum Canadian Wildlife Service (CWS) leg band and a unique combination of plastic color leg bands to enable identification of individuals. We collected a small (∼20 µL) blood sample from the brachial vein and measured length of the right tarsus (mm) and mass (g). We aged adults as second year (SY) or after second year (ASY) by rectrix shape (Pyle 1997). More precise ages were calculated for birds that had been captured and marked in previous years (R. Vallender, unpubl. data). Our inclusion of age and body condition into our analyses was done under the assumption that both these factors may be correlated with a male's song: older males are more experienced and may sing a higher-quality song (Hasselquist et al. 1996, Gil et al. 2001), and larger males may be of better quality and thus more able to invest in singing (Gil and Gahr 2002).
We located nests by observing females carrying nesting material or by actively searching in suitable habitat at three study sites. Once a nest was located, we observed nest development and progress every second day through fledging or depredation. On the fourth day after hatching, we banded nestlings with a CWS band and took blood samples (∼15 µL) from the metatarsal vein.