Summary
We examined the continuous variation in the breeding plumage of male yellow warblers (Dendroica petechia) to determine if the amount of brown streaking on the breast functions as a badge of relative status in territorial males. Conspecific mounts differing in the amount of breast streaking were presented as simulated intruders to a sample of territorial males. Individual males exhibited higher levels of total aggressive response toward brighter models, i.e. those with more brown streaking. This indicates that the streaking is used to assess prospective opponents prior to a response, and that bright intruding males are regarded as more of a threat to the resident than are duller males. The level of response was also correlated with the brightness of the resident males. Bright males responded more aggressively than duller residents to model presentations, and also allocated more time and effort to territory maintenance. We conclude from this that bright males are signalling social dominance, but suggest that the proximate source of high status is a costly increase in the level of aggressive motivation, and not simply an intrinsic superiority in fighting competence. Thus the amount of brown streaking would be a reliable signal of the level of aggressive effort a male is willing to expend in escalating contests and defending resources. We propose that the reliability of the signal is ultimately maintained by the use of different but, overall, equally successful reproductive strategies by males of different appearance. An important implication of this is that ‘cheating’ on signalled status may be neither possible nor necessary.
Methodology
This study was done in the vicinity of the Queen's University Biological Station, 50 km north of Kingston, Ontario, during May and June 1983. Resident males were mist-netted and colour-banded at the beginning of the season to permit identification of individual birds. During handling, the following three plumage characteristics were assessed: percentage brown coloration on the breast (~o brown), number of distinct brown stripes on the breast, and overall conspicuousness of the plumage. We calculated ~ brown by centring a clear plastic grid (2 x 2 cm), consisting of 25 smaller squares, on the breast. A standardized method of grid placement was used on each bird. The percentage of brown plumage was then estimated within each square and averaged over all 25 to obtain the overall estimate. This measurement was repeatable, with 95~o confidence limits of about +__ 2'5~o. The conspicuousness score was assigned on an integer scale of 0 to 10, and was based on a subjective assessment of the overall contrast between the brightness of the yellow plumage and the darkness of the brown streaks. We found little variation among males in either the overall conspicuousness or the number of distinct stripes. Since ~ brown was the most variable of the three plumage characters, and might therefore be expected to be the important component of a plumage signal (like size of the black patch in Harris' sparrows, Zonotrichia querula; Rohwer & Rohwer 1978), this score was used to assign relative rank on the basis of appearance. If individuals had very similar ~ brown scores, the other two scores were then used to ensure that a higher rank was assigned to the male whose appearance was most distinctive from that of female yellow warblers. Higher-ranked birds in these cases were more conspicuous, having either more distinct stripes or slightly greater contrast in colours. No two birds were found to be exactly alike. 'Dull' males are herein defined as those with low ~ brown scores (low plumage rank), while 'bright' males are those with high ~ brown scores (high plumage rank). Note that 'dull' and 'bright' are terms that represent two extremes on a continuum of plumage rank, and are used simply for convenience of presentation.
That this variation is not simply an age effect is further supported by a comparison of ~ brown scores for seven males who had been banded and scored in 1982 (Studd 1983) and who returned in 1983. Table I shows that there is no consistent or extensive change in this character within individuals between years, although there is an indication that some males may return slightly duller. While the exact age of these males is unknown, it seems unlikely that the plumage pattern in this short-lived bird changes enough between years to account for the extensive variation. For this reason, we were not concerned about the age of males in the selection of samples for study or in our analyses. Two different samples of territorial males were used in this study: one for model presentations, and one for analysis of territorial behaviour. Overall, there were 23 resident males in two different study areas. One sample of 11 males was randomly selected to be tested with model presentations during May. This included males from both study areas. Fourteen of the original 23 males were found within one of the two study areas. Analysis of territorial behaviour (see below) was restricted to this second group of 14 males to minimize any potential effect of habitat differences on the collection of observational data. Seven of the males in this sample were also included in the group presented with models. Within each of the two samples, a relative integer plumage rank on a scale of 1 (dullest) to N (brightest) could be assigned for each male. The ranks are displayed in the figures, since we were interested in relative appearance and since rank-based non-parametric statistics were used for analysis (Siegel 1956). Table II shows the estimated ~o brown (to the nearest 5~), and plumage rank for each male in each sample (e.g. the male assigned plumage rank 8 in the first sample of 11 males had a % brown score of about 35). Note that intermediate males tend to be more common than either bright or dull males.