140 
THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 124. No. 1. March 2012 
provisioning to chicks across different hosts. 
Carotenoid pigments are derived entirely from 
diet, and their concentration is known to modulate 
nestling mouth color (Thorogood et al. 2008). 
Carotenoid concentration is widely hypothesized to 
indicate nestling quality, as demonstrated in House 
Sparrows ( Passerdomesticus) (Loiseau et al. 2008) 
and Bam Swallows ( Hirundo rusticci) (Saino et al. 
2000, 2003). 
We investigated whether cowbird nestlings 
differ in flange coloration when reared by one of 
two host species. Song Sparrows (Melospiza 
melodia) and Yellow Warblers (Setophaga pete¬ 
chia) using quantitative measures of coloration. 
Birds have a fourth violet- or ultraviolet-sensitive 
photoreceptor type, and human color perception is 
an insufficient proxy for avian color perception 
(Cuthill et al. 2000). We provide the first 
objective assessment of cowbird nestling flange 
colors based on measures of color using digital 
photographs and imaging software (Dale 2000). 
However, this remains a preliminary analysis 
because imaging software is designed for human 
vision and has limited value for avian perceptual 
studies (Stevens et al. 2007). 
The objective of our study was to test the 
hypothesis that rictal flange color of host and 
cowbird nestlings varies between nestlings and 
parasites of two sympatric hosts. We predicted 
measures of flange colors would differ among: (I) 
nestlings of different species (Yellow Warbler, 
Song Sparrow, and Brown-headed Cowbird), and 
(2) cowbird nestlings in nests of different host 
species. 
METHODS 
Study Site and Species. —This study was con 
ducted in the riparian corridors of four tributarie 
of Mono Lake (38° 1’ N, 119° 3' W) on th- 
eastern slope of the Sierra Nevada, California 
USA: Lee Vining, Mill, Rush, and Wilson creeks 
We located Song Sparrow' and Yellow Warble 
nests and monitored nests during the 200< 
breeding season lollowing Martin and Geupe 
(1993) and Ralph et al. (1993). The ranges of nes 
initiation dates for Yellow Warbler and Sonj 
Sparrows were similar in 2004 (Tonra et al. 2009) 
Nestling Photographs,- We photographed nest 
mgs on day 6 (hatching day = day 0) of the 
cowbird nestling cycle. This day was chose, 
because .t coincided with the age at which 
allowed f ^ S " tTicien,i >' ***& to band and 
allowed for incorporation of carotenoid pigments 
from host-provisioned diet into tissues. Photo¬ 
graphs were taken with a Hewlett Packard 
Photosmart 215 digital camera, set to ISO 200 
and 'fine' quality (1,280 X 960 pixels). We 
photographed all cowbirds in each nest and. if 
hosts were present, we also randomly selected one 
individual to photograph. Sixteen Brown-headed 
Cowbird. three Yellow Warbler, and three Song 
Sparrow nestlings were included in the analysis. 
Both cowbird and host chicks were photographed 
in two Song Sparrow and two Yellow Warbler 
nests, and one Yellow Warbler and two Song 
Sparrow nests each contained two cowbird 
nestlings. Each photograph was taken of the nght 
side of the head against a background of gray 
paper with a strip composed of six 1-cnr sections 
cut from paint store color sample cards (red. blue, 
green, white, yellow, and black) (Home Depot, 
Reno. NV, USA) as a color standard, and stored in 
a dark box between photography sessions. This 
allowed us to make direct comparisons of colors 
under varying light conditions in the field. The 
photographs were saved and subsequently ana 
lyzcd as jpeg images using the histogram function 
in Adobe Photoshop Elements 8.0 (Adobe Sys¬ 
tems Inc.. San Jose. CA, USA). Storing images as 
jpeg compresses both image and color data 
(Stevens et al. 2007); color compression in this 
software obscures rather than enhances differenc¬ 
es in color and would result in our failure to reject 
the null hypothesis (despite its falsehood: Type II 
statistical error). Thus, use of jpeg images made 
our analyses more conservative. 
Flanges were divided into three portions tor 
color measurement: A at the apex of the flange, 
site B at the fleshy middle, and site C at its most 
rostral point. Three replicate measures of red. 
green, and blue values were made at each flange 
site for each nestling. Red, green, and blue 
measures represent the intensity levels (satura¬ 
tion) for 24-bit color; these measures range in 
intensity from 0 (black) to 255 (white, totally 
saturated color). Measurements were also made 
from the center of the yellow standard present in 
each photograph to allow for direct comparison of 
the standard and biological colors. 
Data Analysis. —We compared red (R). g reen 
(G), and blue (B) values, while accounting lor 
variation in light conditions associated with held 
work in each photograph, by first scaling 
according to the RGB values of the yellow 
standard in that photograph i. Thus, F(A| * 
where I = R. G, or B value, F, = flange color 
