- Email: [email protected]
Plumage brightness as an indicator of parental care in northern cardinals
Anim. Behav., 1998, 55, 119–127
Plumage brightness as an indicator of parental care in northern cardinals SUSAN U. LINVILLE, RANDALL BREITWISCH & AMY J. SCHILLING Department of Biology, University of Dayton (Received 7 January 1997; initial acceptance 5 March 1997; final acceptance 24 April 1997; MS. number: 7810)
Abstract. Good parent and differential allocation models predict relationships between degree of sexual ornamentation and parental care, but relatively few studies have tested these models. The northern cardinal, Cardinalis cardinalis, is a sexually dichromatic species in which both sexes are ornamented. Males have red plumage, and females have tan plumage with limited areas of red feathering. Cardinals were used to address the two models and determine whether plumage brightness signals level of parental care by both sexes. Absolute effort in feeding nestlings by males was not correlated with male breast plumage colour, but effort by females was positively correlated with female underwing plumage colour. Absolute feeding effort by females was also inversely related to brightness of the mate’s breast colour. As a consequence, the proportion of a pair’s total feedings provided by the male was positively correlated with male breast colour. Proportion of total feedings provided by the female was positively correlated with female wing colour. Feeding efforts (both per nest and per nestling) were correlated between mates, but birds did not mate assortatively in relation to colour. These results support the good parent hypothesis, suggesting colour brightness is a signal of parental care. The results also indicate that ornamentation of both members of the pair may be important determinants of relative efforts in ? 1998 The Association for the Study of Animal Behaviour provisioning nestlings by parent birds.
Sexual selection theory is evoked to explain the often conspicuous secondary sex traits of male animals, but selective mechanisms behind conspicuous ornaments are still debated. Selection models predict that females prefer mates with relatively exaggerated secondary sexual characteristics (Darwin 1871; Fisher 1930; Hamilton & Zuk 1982; Kodric-Brown & Brown 1984; Andersson 1986), and females in at least several avian species choose mates on the basis of plumage colour (Johnson 1988; Hill 1990; Sundberg & Larsson 1994), tail length (Andersson 1982; Møller 1988), comb and eye colour (Zuk et al. 1990) and leg band colour (Burley 1981). If females select mates for exaggerated characteristics, it is important to determine the fitness accrued by such choices. Females may obtain indirect benefits for their offspring, including ‘good’ genes that result in healthier or more attractive offspring. Direct benefits for females include receiving food from the mate, obtaining a good territory or obtaining a mate that is a good parent. Correspondence: R. Breitwisch, Department of Biology, University of Dayton, Dayton, OH 45469-2320, U.S.A. (email: [email protected]). 0003–3472/98/010119+09 $25.00/0/ar970595
Two models address the relationship between degree of sexual ornamentation and parental care. The good parent model assumes that expression of the trait positively reflects parenting ability (Kirkpatrick 1985; Heywood 1989; Hoelzer 1989). Hence, females should benefit directly through their choice of highly ornamented mates by obtaining those that invest more in parental care than would other potential mates. The differential allocation model (Burley 1986) assumes that females choose mates for indirect benefits (good genes for the offspring), and predicts a negative relationship between parental care and ornamentation. As a consequence, females mated to highly ornamented males should invest more in parental care than females mated to less ornamented males. Carotenoid-based plumage colour is a secondary sexual characteristic commonly expressed in avian species. The expression of this trait results from numerous physiological and behavioural processes that determine carotenoid intake, absorption, transport and deposition in feathers, bills or exposed fleshy structures (Brush 1990). As such, carotenoid-based colour may function as an indicator of genetic quality (Hamilton & Zuk
? 1998 The Association for the Study of Animal Behaviour
119
120
Animal Behaviour, 55, 1
1982; Hudon et al. 1989), as well as foraging ability (Hill 1992), presence of parasitic and other diseases (Schaeffer at al. 1988) and hormone levels (Brush 1967; Temple 1974). The northern cardinal, Cardinalis cardinalis, is a sexually dichromatic, socially monogamous species. Both sexes have orange-red bills and dull red-brown remiges and rectrices that are nearly invariant in colour (R. Breitwisch, unpublished data). Males also have carotenoid-based bright red plumage with especially intense red colour on the breast. Although much duller in colour, females might also be considered ornamented. Body coverts are tan overall, but females display variable red colour on underwing median and lesser secondary coverts, and variable numbers of red feathers on the crest, face, upper breast and flank. The underwing coverts of females always display the brightest red colour in female plumage. In field studies in a northeastern U.S. population, the brightest red male cardinals paired with the earliest breeding females, obtained territories with the densest vegetation and produced more fledglings than duller males (Wolfenbarger 1996). In our southwestern Ohio population, males fed nestlings at higher rates than did females, but the relationship between level of provisioning and ornamentation was unknown (Filliater & Breitwisch 1997). The aim of this study was to investigate the relationship between both male and female ornamentation and parental care, measured as effort in provisioning nestlings. The good parent model predicts that brighter individuals invest more in parental care than do duller individuals. Brighter males and females are expected to provide more care than do duller individuals, independent of the mate’s level of ornamentation. Alternatively, the good parent model allows for the possibility that the mate’s colour influences a parent’s level of care, with individuals feeding less if they are mated to bright birds. The differential allocation model predicts the opposite with regard to male investment in parental care. Under this hypothesis, bright males are expected to feed less than duller males, while females mated to bright males are expected to feed proportionately more than females mated to dull males. The differential allocation model does not make a prediction with regard to female colour because only males are assumed to allocate their reproductive effort
between mating and parental efforts based on their level of ornamentation. We examined two hypotheses to determine whether plumage colour functions as an indicator of parental care, as proposed by the two models. The first hypothesis is that an individual’s level of ornamentation predicts its absolute parental care effort. The second hypothesis is that an individual’s level of ornamentation predicts its relative parental care effort in relation to its mate’s effort.
METHODS We conducted this study at Aullwood Audubon Center and Farm (39)52*N, 84)16*W), 15 km northwest of Dayton, Ohio, U.S.A., during the 1995 and 1996 breeding seasons. The property consists of approximately 80 ha of primarily reclaimed farmland with a variety of habitats, including an abundance of woody, disturbed habitat ideal for cardinals. We observed feeding in 22 pairs, 15 in 1995 and seven in 1996. We observed two males and one female both years, but used only the more complete set of observations from 1 year to represent them. We were unable to obtain plumage scores for two additional females. To be consistent throughout, we used 17 complete pairs for all data analyses. We conducted duplicate analyses where possible on the larger samples of 20 males and 19 females, and results were similar to those reported here based on 17 pairs. We compared feeding rates, feeding rates per nestling, proportion of feedings, brood size and colour measurements between years. There was no difference between years for any comparison (Mann–Whitney U-tests: all P>0.05), and we combined years in analyses. There was no difference in the distribution of hatching dates in the two years (Kolmogorov–Smirnov two-sample test: D=0.30, N1 =10, N2 =7, P>0.10). For the 2 years combined, we compared hatching date of each nest with the combined feeding efforts of the parents to determine seasonal effects, but found none (Spearman rank correlation: rs = "0.39, N=17, P=0.11). We captured adults with potter traps or mist nets only early in the season or during incubation to avoid disturbance during nest building, egg laying and feeding. Each was banded with a U.S. Fish and Wildlife Service aluminum band and three coloured celluloid bands in unique
Linville et al.: Cardinal parental care and colour combinations. We did not use red colour bands, because bands similar in colour to a species’ ornament influence reproductive success in several other species (Burley et al. 1982; Hagan & Reed 1988; Metz & Weatherhead 1991). We determined an individual’s colour by matching plumage colour to the most similar Munsell chip colour. We recorded breast colours for males and underwing colours for females. Although we did not recapture birds the same season, we tested colour scoring for reliability with museum skins at The Dayton Museum of Natural History and found scoring to be highly repeatable (rS =0.93, N=22, P<0.001). Since colour is a three-dimensional categorization, we converted Munsell scores to a onedimensional continuum similar to that in Burley & Coopersmith (1987). Colour value was weighted primary, however, with hue secondary, and chroma tertiary, to prevent the possibility of a highly saturated pink outranking a darker, less saturated red-orange. Although any such hierarchy is subjective, this ordering of colour chips produced a bright to dull ‘redness’ continuum that agreed with our perception. For male breast colour, scores ranged from 6.0 (Munsell 7.5 4/16) to 0.5 (Munsell 8.75 5/16). For female underwing colour, scores ranged from 13.5 (Munsell 8.75 4/16) to 1.0 (Munsell 10 6/13). Female cardinals also have variable numbers of red body coverts. We determined female ‘body colour’ by summing the scores for relative numbers of red feathers in the crest, cheek, eyebrow, upper breast and flank. We scored these in cheek and flank regions as 0 (red feathers absent) or 1 (a few red feathers present) and in eyebrows, crest and breast as 0 (red feathers absent), 1 (a few red feathers present) or 2 (several to many red feathers present). Males and females also have bright orange to red bills. Even though bill colour may be a sexually selected trait, we found it difficult to quantify bill colour because of its subtly mottled appearance. Thus, we did not examine bill colour as an ornament. The nestling period for cardinals is 10 days, with day 0 representing hatching. We conducted feeding observations on days 2–8 of nestling life. We did not include days 0 or 1 because females spend long periods brooding and, when not brooding, are often fed by males off the nest and out of sight of the observer. Such feeding may be viewed as indirect provisioning of nestlings by
121
males, and females sometimes deliver this food to nestlings. Day 9 was usually the last day nestlings remained in the nest, but fledging sometimes occurred on that day. We conducted observations for 1 h each day and recorded the number of feedings made by both members of the pair. Most observations were conducted before noon, but time of day has been shown not to influence feeding rates (Filliater & Breitwisch 1997). Fewer than 7 days of observations were completed if either the nest was found after day 2 or chicks were preyed upon. Two pairs are represented by 7 days of observation, eight pairs by 6 days, four pairs by 5 days, and three pairs by 3 days. We used three measures of nestling provisioning. First was the number of feedings/h, which we assumed to be an absolute measure of effort. However, effort per nestling is also important to both a nestling and to the mate and partially controls for differences in brood size. Accordingly, we also recorded effort as feedings/ nestling/h. Last, the proportion of a pair’s total feeding effort undertaken by a male or female is a measure of the response of each individual to the needs of the nestlings in relation to the response by the mate. It is unknown which of these three measures is most likely to have the greatest effect on the fitness of both the individual and its mate. Consequently, we report all three. We attempted to measure load size as well as feeding frequency. Cardinals carry food deep in the bill, and food quantity cannot be determined by direct observation. We were unsuccessful in our attempts to measure load sizes by the use of pipe cleaner collars. Foods delivered by parents were frequently mashed, and food boluses slid past the point of constriction. We standardized feeding rates by comparing individual rates to the mean feeding rates of all other individuals sampled on a given day of nestling life. An individual’s standardized feeding rate was the positive or negative difference between the the individual’s feeding rate and the mean feeding rate for that sex. We then totalled these standardized rates per day for an individual to give an overall standardized feeding rate for all of the days sampled. This was divided by the number of days sampled for the individual to yield a mean standardized feeding rate compared with all others of the same sex in the study population. This method controls for the increase in feeding
Animal Behaviour, 55, 1
frequency with nestling age and incomplete sampling of pairs. We calculated relative feeding effort as the proportion of feedings by a male or female based on all of the feedings by the pair. Although the male proportion of feedings changed over the days sampled (see Results), incomplete sampling for some pairs was not biased with regard to sampling within the 7-day period (Kruskal–Wallis one-way ANOVA: H6 =9.35, P=0.97); thus, calculation of proportions simply based on all data for a pair is reasonable. Parental care increases with age and experience in some species. In this population, very few birds banded as nestlings remain in subsequent years. Therefore, we do not know the age of individual cardinals. Banded adults tend to remain in subsequent years on or near the territory where they were banded, and we measured years at the study site as an indicator of local experience and, indirectly, of age. We used these data to assess the relationship between experience and parental care. We used Kendall rank correlations (T reported) to relate absolute feeding efforts and colour (hypothesis 1). We used Kendall partial rank correlations (Txy · z reported) to eliminate the effects of the mate’s colour in relating feeding proportion to plumage colour (hypothesis 2). We used Wilcoxon signed-ranks tests to compare feeding levels by mates and colour changes between years. We used the sequential Bonferroni technique (Rice 1989) to evaluate the significance of correlations when multiple comparisons (i.e. two female ornaments) were used to test the same hypothesis. We report the significance level associated with the outcome of each statistical analysis. In cases where we accept the null hypothesis and the significance level is between 0.05 and 0.20, we also report the statistical power of the test at detecting a significant difference at an á of 0.05. RESULTS General Feeding Pattern Males contributed more care to nestlings than females with a mean feeding effort of 2.2 feedings/h compared with 2.0 feedings/h, but these differences were not significant (W=32, N=17, P=0.33), in contrast to results of Filliater
0.8 Male proportion of feedings
122
0.7 0.6 0.5 0.4 0.3 0.2
1
2
3
4 5 6 Nestling age
7
8
9
Figure 1. Mean& proportion of feedings by the male parent as a function of nestling age for days 2–8. N=8–15 pairs per day (Kendall rank correlation: T=0.67, N=7 days, P=0.0001).
& Breitwisch (1997). Feeding efforts were correlated between mates for both feedings/h (T=0.40, N=17, P=0.013) and feedings/nestling/h (T=0.41, N=17, P=0.011), in agreement with Filliater & Breitwisch (1997). Furthermore, the proportion of feedings by males increased with nestling age (T=0.67, N=7 days, P=0.0001; Fig. 1). Absolute Feeding Effort and Ornamentation Male feeding effort as feedings/h was not correlated with male breast colour (T=0.09, N=17, P=0.31). Male feeding effort as feedings/ nestling/h was also not correlated with male breast colour (T=0.21, N=17, P=0.12, power=80%). Female wing and body colour were not correlated (T= "0.01, N=17, P=0.29). Female feeding effort/h was not correlated with either underwing colour (T=0.24, N=17, P=0.090, power=68%) or body colour (T= "0.08, N=17, P=0.33). Similarly, there was no correlation between feedings/ nestling/h and body colour (T= "0.07, N=17, P=0.39). There was, however, a positive correlation between feedings/nestling/h and underwing colour (T=0.40, N=17, P=0.012). This significant correlation remained even after reducing the testwise á to 0.012 by the Bonferroni technique, based on multiple tests (4) of the same hypothesis. We also determined whether an individual’s absolute feeding level was correlated with its
Linville et al.: Cardinal parental care and colour 0.8 Female proportion of feedings
Male proportion of feedings
0.8 0.7 0.6 0.5 0.4 0.3 0.2 2.5
123
3.5 4.5 5.5 Male breast colour
6.5
0.7 0.6 0.5 0.4 0.3 0.2
4
5
6
7 8 9 10 11 Female wing colour
12
13
Figure 2. Correlation between the proportion of male feedings and male breast colour (T=0.41, N=17, P=0.011).
Figure 3. Correlation between proportion of female feedings and female underwing colour (T=0.46, N=17, P=0.005).
mate’s colour. Female feeding effort as feedings/h was not correlated with the mate’s breast colour (T= "0.26, N=17, P=0.074, power=69%), but there was a negative correlation between female feedings/nestling/h and the mate’s breast colour (T= "0.38, N=17, P=0.017). Because of multiple tests (2), the Bonferroni technique reduces the acceptable á value to 0.025. This relationship remains significant, with females mated to bright males feeding less per nestling than females mated to duller males. Neither male feedings/h (T= "0.18, N=17, P=0.16, power=64%) nor male feedings/nestling/h (T= "0.19, N=17, P=0.14, power=67%) was correlated with female underwing colour.
male breast colour was held constant, female wing colour remained positively correlated with female feeding proportion (Txy · z =0.41, N=17, P=0.011). Even when an apparent ‘outlier’ from the brightest female is eliminated, the significant correlation remains.
Relative Feeding Effort and Ornamentation The male proportion of feedings was positively correlated with male breast colour (T=0.41, N=17, P=0.011; Fig. 2). When the potential influence of female wing colour was held constant, male breast colour remained positively correlated with male feeding proportion (Txy · z =0.35, N=17, P=0.025). Even when an apparent ‘outlier’ from the dullest male is eliminated, the significant correlation remains. The female proportion of feedings was positively correlated with female wing colour (T=0.46, N=17, P=0.005; Fig. 3), but not with body colour (T= "0.27, N=17, P=0.065, power=75%). When
Feeding Efforts and Local Experience Of 17 males on the study site, one had been present for 5 years, three for 4 years, one for 3 years, seven for 2 years, and five for 1 year. Experience was not correlated with feedings/h (T=0.08, N=17, P=0.33), feedings/nestling/h (T=0.07, N=17, P=0.35) or proportion of feedings (T= "0.09, N=17, P=0.31). Of 17 females on the study site, four had been present for 4 years, seven for 2 years, and six for 1 year. Experience was not correlated with feedings/h (T=0.11, N=17, P=0.27), feedings/ nestling/h (T=0.07, N=17, P=0.35) or proportion of feedings (T= "0.07, N=17, P=0.35). Age and Colour Nine males from 1994 and 1995 and eight males from 1995 and 1996 were measured to determine whether breast plumage brightness changed with age. There was no change from 1994 to 1995 (W= "18.0, N=9, P=0.30), but there was an increase in brightness from 1995 to 1996 (W=31,
Animal Behaviour, 55, 1
124
N=8, P=0.02). One individual was represented in both samples. Eight females from 1994 and 1995 and six females from 1995 and 1996 were measured to determine whether underwing plumage brightness changed with age. There was no change from 1994 to 1995 (W= "15.0, N=8, P=0.22), but there was a marginal increase from 1995 to 1996 (W=15.0, N=6, P=0.06). Two individuals were represented in both samples. Ornaments in Pairs There was no significant correlation between male breast colour and female underwing colour in pairs (T= "0.22, N=17, P=0.11, power= 60%), or between male breast colour and female body colour in pairs (T= "0.27, N=17, P=0.065, power=75%).
DISCUSSION Ornamentation and Feeding Effort Colour as ornamentation in male cardinals did not serve as a predictor of absolute feeding effort, and the first hypothesis, that an individual male’s ornamentation predicts absolute parental care effort, was refuted. This hypothesis was supported for females, however. Female underwing colour was positively correlated with female per-nestling feeding effort. Within the context of particular pairs, where the colours of both sexes were known, the relative feeding effort by both males and females was predicted by their own colours. Thus, the second hypothesis, that an individual’s ornamentation predicts relative parental care effort, was also supported. Furthermore, the absolute feeding level by females was inversely related to the mate’s colour; females mated to bright males fed less per nestling than did females mated to dull males. Therefore, we suspect that the pattern within pairs of relative feeding effort and colour is due, in part, to a female behavioural response to the mate’s colour. Investigators of sexual selection hypotheses concentrate on male ornamentation, for obvious reasons. Even when testing indicator models that predict provisioning levels, investigators typically test for correlations between male ornament expression and absolute provisioning level by
males. Three studies have shown positive relationships (Grant & Grant 1987; Hill 1991; Palokangas et al. 1994), supporting the good parent model. On the other hand, four studies have shown negative relationships (Studd & Robertson 1985; Burley 1988; de Lope & Møller 1993; Sundberg & Larrson 1994), supporting the differential allocation model. No one has yet explained why different species of ornamented, biparental birds lend support to one or the other model. Our finding, that accounting for female ornamentation helps explain the pattern of nestling provisioning in cardinals, suggests that in future studies of parental effort in biparental birds, both female ornamentation (where present) and female feeding effort should also be examined. Female cardinals are active participants in decisions made that pertain to the provisioning of nestlings, as argued for the general case by Breitwisch (1989) and Gowaty (1996). Females mated to bright males may increase their reproductive success by decreasing their feeding contribution while preparing physiologically for the successive nesting attempt. Between 70 and 80% of all cardinal nests in this population are lost to predators (Filliater et al. 1994). Pairs do not strongly defend nests and appear to have adopted re-nesting as a strategy to deal with predation (Nealen & Breitwisch 1997). Since only females build nests, incubate and brood, replenishing energy reserves for the successive nesting may be a better strategy than increasing feeding effort to nestlings. We lack complete nesting records for the birds in this sample and cannot yet test this hypothesis. A remaining question is how male cardinals apportion reproductive effort between parental effort and mating effort in seeking extra-pair fertilizations. Investigators have found either that support for the good parent model is accompanied by a low incidence of extra-pair fertilizations not associated with male ornamentation (Hill et al. 1994), or that support for the differential allocation model is accompanied by a pattern of extra-pair fertilizations favouring highly ornamented males (Møller 1994; Burley et al. 1996; Sundberg & Dixon 1996). The relationship between male ornamentation and extra-pair fertilizations in cardinals is more complicated, with brighter males losing paternity more than duller males (S. U. Linville & R. Breitwisch, unpublished data).
Linville et al.: Cardinal parental care and colour Female Ornamentation The sequestering by females of bright red pigmentation in underwing feathers might be considered evidence that carotenoids in female diet inevitably appear as bright red plumage. That is, female underwing colour may be a compromise between strong selection for ornamentation in males and crypticity in females rather than a sexually selected trait in females (Lande 1980; Muma & Weatherhead 1989). We found a significant correlation, however, between female underwing colour and absolute feeding rate, suggesting that colour is an indicator of female quality and thus may be maintained by sexual selection. We have also observed that this patch of red plumage on females is quite visible in flight, but we know of no display by females that clearly exposes and presents the underwing surface to either males or other females (Lemon 1968). We lack data on whether female ornamentation is related to success in intra-sexual interactions, although female cardinals are quite aggressive towards other females (S. U. Linville & R. Breitwisch, personal observation). Female expression of ornaments elaborated in males of dichromatic species has been little studied. Muma & Weatherhead (1989, 1991) found no support for a function of red epaulets in female red-winged blackbirds, Agelaius phoeniceus, in either female interactions or male mate choice. Hill (1993) found that male house finches, Carpodacus mexicanus, preferred the most brightly plumaged females in choice experiments but that female colour was not correlated with any of Hill’s measures of female quality in a wild population. Our results suggest that coloration of female cardinals serves as an indicator of quality, in contrast to the findings of these other studies. Age and Ornamentation Several investigators have found that ornamentation serves as an age indicator (Grant & Grant 1987; Sundberg & Larsson 1994), but others have found this not to be so (Hill 1990; Møller 1991). The advantages to an individual choosing an older mate may relate to parental or other experience that increases with age or the demonstration of survival influenced by genotype. We obtained mixed results pertaining to the question of whether ornamentation in male cardinals is an age
125
indicator. The breast colour of individual males increased in brightness from one breeding season to the next in one of the 2 years of the study. Wolfenbarger (1996) found in a different population of cardinals that male breast colour brightness increased with age. The dullest areas of red breast plumage increased in brightness, although the brightest red breast plumage did not change colour. The availability of carotenoids on our study site may have contributed to the absence of a change in one of 2 years, as we discuss elsewhere (Linville & Breitwisch, 1997). There has been almost no study of female ornamentation and age in birds. Hill (1993) found that first-year female house finches were more colourful than older females. Muma & Weatherhead (1989) found that epaulets of female red-winged blackbirds were brighter when the birds were recaptured a second year. We found no indication that coloration of female cardinals is related to their age. Ornamentation of Mated Pairs Absolute feeding efforts by mates were correlated, and individual ornamentation predicted the relative feeding effort by both sexes. The question that arises is why cardinals do not mate assortatively. Might the nestlings of highly ornamented pairs receive even more food? Ornamentation in cardinals is probably an imprecise predictor of parental care, because colour is directly influenced by an individual’s diet (thus, indirectly by territory quality) and, also, may increase with age. Cardinal pairs in this population tend to remain together year after year (R. Breitwisch & S. U. Linville, unpublished data), and the experience gained with the mate in long-term relationships may influence provisioning decisions as much as ornamentation predicts the pattern of provisioning nestlings. ACKNOWLEDGMENTS We thank Pat Weatherhead and an anonymous referee for helpful comments on this manuscript. For permission to conduct this research at Aullwood, we thank the Director, Charity Krueger, and Head of Research, John Ritzenthaler. We also thank all of the Aullwood staff for their continuing support. For field assistance, we thank Josh Banks, Nancy Beckman,
Animal Behaviour, 55, 1
126
Jeffrey Biessman, Tim DeMarco, Brittany Linville, Steve Ramey and Sharlene Regan. We also thank Gary Coovert at the Dayton Museum of Natural History for use of museum skins. This research was supported by a grant from the Ohio Biological Survey, and University of Dayton Graduate and Faculty Summer Fellowships. This project was conducted under USFWS Banding Permit No. 22351 and ODNR Banding Permit No. 5-57-04, both issued to R. Breitwisch.
REFERENCES Andersson, M. 1982. Female choice selects for extreme tail length in a widowbird. Nature, Lond., 299, 818– 820. Andersson, M. 1986. Evolution of condition-dependent sex ornaments and mating preferences: sexual selection based on viability differences. Evolution, 40, 804– 816. Breitwisch, R. 1989. Mortality patterns, sex ratios, and parental investment in monogamous birds. Curr. Ornithol., 6, 1–50. Brush, A. H. 1967. Pigmentation in the scarlet tanager, Piranga olivacea. Condor, 69, 549–559. Brush, A. H. 1990. Metabolism of carotenoid pigments in birds. FASEB J, 4, 2969–2977. Burley, N. 1981. Sex ratio manipulation and selection for attractiveness. Science, 211, 721–722. Burley, N. 1986. Sexual selection for aesthetic traits in species with biparental care. Am. Nat., 127, 415–445. Burley, N. 1988. The differential-allocation hypothesis: an experimental test. Am. Nat., 132, 611–628. Burley, N. & Coopersmith, C. B. 1987. Bill color preferences of zebra finches. Ethology, 76, 133–151. Burley, N., Krantzberg, G. & Radman, P. 1982. Influence of colour-banding on the conspecific preferences of zebra finches. Anim. Behav., 30, 444–455. Burley, N. T., Parker, P. G. & Lundy, K. 1996. Sexual selection and extrapair fertilization in a socially monogamous passerine, the zebra finch (Taeniopygia guttata). Behav. Ecol., 7, 218–226. Darwin, C. 1871. The Descent of Man and Selection in Relation to Sex. London: John Murray. Filliater, T. S. & Breitwisch, R. 1997. Nestling provisioning by the extremely dichromatic northern cardinal. Wilson Bull., 109, 145–153. Filliater, T. S., Breitwisch, R. & Nealen, P. M. 1994. Predation on northern cardinal nests: does choice of nest site matter? Condor, 96, 761–768. Fisher, R. A. 1930. The Genetical Theory of Natural Selection. Oxford: Clarendon Press. Gowaty, P. A. 1996. Field studies of parental care in birds: new data focus questions on variation among females. Adv. Study Behav., 27, 477–531. Grant, B. R. & Grant, P. R. 1987. Mate choice in Darwin’s finches. Biol. J. Linn Soc., 32, 247–270.
Hagan, J. M. & Reed, J. M. 1988. Red colour bands reduce fledging success in red cockaded woodpeckers. Auk, 105, 498–503. Hamilton, W. D. & Zuk, M. 1982. Heritable true fitness and bright birds: a role for parasites? Science, 218, 384–387. Heywood, J. S. 1989. Sexual selection by the handicap mechanism. Evolution, 43, 1387–1397. Hill, G. E. 1990. Female house finches prefer colourful males: sexual selection for a condition-dependent trait. Anim. Behav., 40, 563–572. Hill, G. E. 1991. Plumage coloration is a sexually selected indicator of male quality. Nature, Lond., 350, 337–339. Hill, G. E. 1992. Proximate basis of variation in carotenoid pigmentation in male house finches. Auk, 109, 1–12. Hill, G. E. 1993. Male mate choice and the evolution of female plumage coloration in the house finch. Evolution, 47, 1515–1525. Hill, G. E., Montgomerie, R., Roeder, C. & Boag, P. 1994. Sexual selection and cuckoldry in a monogamous songbird: implications for sexual selection theory. Behav. Ecol. Sociobiol., 35, 193–199. Hoelzer, G. A. 1989. The good parent process of sexual selection. Anim. Behav., 38, 1067–1078. Hudon, J., Capparella, A. & Brush, A. H. 1989. Plumage pigment differences in manakins of the Pipra erythrocephala superspecies. Auk, 106, 34–41. Johnson, K. 1988. Sexual selection in pinyon jays. I. Female choice and male–male competition. Anim. Behav., 36, 1038–1047. Kirkpatrick, M. 1985. Evolution of female choice and male parental investment in polygynous species: the demise of the ‘sexy son’. Am. Nat., 125, 788–810. Kodric-Brown, A. & Brown, J. H. 1984. Truth in advertising: the kinds of traits favored by sexual selection. Am. Nat., 124, 309–323. Lande, R. 1980. Sexual dimorphism, sexual selection, and adaptation in polygenic characters. Evolution, 34, 292–305. Lemon, R. E. 1968. The displays and call notes of cardinals. Can. J. Zool., 46, 141–151. Linville, S. U. & Breitwisch, R. 1997. Carotenoid availability and plumage coloration in a wild population on northern cardinals. Auk, 114, 796–800. de Lope, F. & Møller, A. P. 1993. Female reproductive effort depends on the degree of ornamentation of their mates. Evolution, 47, 1152–1160. Metz, K. J. & Weatherhead, P. J. 1991. Color bands function as secondary sexual traits in male red-winged blackbirds. Behav. Ecol. Sociobiol., 28, 23–27. Møller, A. P. 1988. Female choice selects for male sexual tail ornaments in the monogamous swallow. Nature, Lond., 332, 640–642. Møller, A. P. 1991. Sexual selection in the monogamous barn swallow (Hirundo rustica). 1. Determinants of tail ornament size. Evolution, 45, 1823–1836. Møller, A. P. 1994. Sexual Selection and the Barn Swallow. Oxford: Oxford University Press. Muma, K. E. & Weatherhead, P. J. 1989. Male traits expressed in females: direct or indirect selection? Behav. Ecol. Sociobiol., 25, 23–31.
Linville et al.: Cardinal parental care and colour Muma, K. E. & Weatherhead, P. J. 1991. Plumage variation and dominance in captive female red-winged blackbirds. Can. J. Zool., 69, 49–54. Nealen, P. & Breitwisch, R. 1997. Northern cardinal sexes defend nests equally. Wilson Bull., 109, 269–278. Palokangas, P., Korpimaki, E., Hakkarainen, H., Huhta, E., Tolonen, P. & Alatalo, R. V. 1994. Female kestrels gain reproductive success by choosing brightly ornamented males. Anim. Behav., 47, 443– 448. Rice, W. R. 1989. Analyzing tables of statistical tests. Evolution, 43, 223–225. Schaeffer, J. L., Tyczkowski, J. K. & Hamilton, P. B. 1988. Depletion of oxycarotenoid pigments in chickens and failure of aflatoxin to alter it. Poult. Sci., 67, 1080–1088. Studd, M. V. & Robertson, R. J. 1985. Sexual selection and variation in reproductive strategy in male yellow
127
warblers (Dendroica petechia). Behav. Ecol. Sociobiol., 17, 101–109. Sundberg, J. & Dixon, A. 1996. Old, colourful male yellowhammers, Emberiza citrinella, benefit from extra-pair copulations. Anim. Behav., 52, 113–122. Sundberg, J. & Larsson, C. 1994. Male coloration as an indicator of parental quality in the yellowhammer, Emberiza citrinella. Anim. Behav., 48, 885–892. Temple, S. A. 1974. Plasma testosterone titers during the annual reproductive cycle of starlings (Sturnus vulgaris). Gen. Comp. Endocrinol., 22, 470–479. Wolfenbarger, L. L. 1996. Fitness effects associated with red coloration of male northern cardinals (Cardinalis cardinalis). Ph.D. thesis, Cornell University. Zuk, M., Johnson, K., Thornhill, R. & Ligon, J. D. 1990. Mechanisms of female choice in red jungle fowl. Evolution, 44, 477–485.
Plumage brightness as an indicator of parental care in northern cardinals SUSAN U. LINVILLE, RANDALL BREITWISCH & AMY J. SCHILLING Department of Biology, University of Dayton (Received 7 January 1997; initial acceptance 5 March 1997; final acceptance 24 April 1997; MS. number: 7810)
Abstract. Good parent and differential allocation models predict relationships between degree of sexual ornamentation and parental care, but relatively few studies have tested these models. The northern cardinal, Cardinalis cardinalis, is a sexually dichromatic species in which both sexes are ornamented. Males have red plumage, and females have tan plumage with limited areas of red feathering. Cardinals were used to address the two models and determine whether plumage brightness signals level of parental care by both sexes. Absolute effort in feeding nestlings by males was not correlated with male breast plumage colour, but effort by females was positively correlated with female underwing plumage colour. Absolute feeding effort by females was also inversely related to brightness of the mate’s breast colour. As a consequence, the proportion of a pair’s total feedings provided by the male was positively correlated with male breast colour. Proportion of total feedings provided by the female was positively correlated with female wing colour. Feeding efforts (both per nest and per nestling) were correlated between mates, but birds did not mate assortatively in relation to colour. These results support the good parent hypothesis, suggesting colour brightness is a signal of parental care. The results also indicate that ornamentation of both members of the pair may be important determinants of relative efforts in ? 1998 The Association for the Study of Animal Behaviour provisioning nestlings by parent birds.
Sexual selection theory is evoked to explain the often conspicuous secondary sex traits of male animals, but selective mechanisms behind conspicuous ornaments are still debated. Selection models predict that females prefer mates with relatively exaggerated secondary sexual characteristics (Darwin 1871; Fisher 1930; Hamilton & Zuk 1982; Kodric-Brown & Brown 1984; Andersson 1986), and females in at least several avian species choose mates on the basis of plumage colour (Johnson 1988; Hill 1990; Sundberg & Larsson 1994), tail length (Andersson 1982; Møller 1988), comb and eye colour (Zuk et al. 1990) and leg band colour (Burley 1981). If females select mates for exaggerated characteristics, it is important to determine the fitness accrued by such choices. Females may obtain indirect benefits for their offspring, including ‘good’ genes that result in healthier or more attractive offspring. Direct benefits for females include receiving food from the mate, obtaining a good territory or obtaining a mate that is a good parent. Correspondence: R. Breitwisch, Department of Biology, University of Dayton, Dayton, OH 45469-2320, U.S.A. (email: [email protected]). 0003–3472/98/010119+09 $25.00/0/ar970595
Two models address the relationship between degree of sexual ornamentation and parental care. The good parent model assumes that expression of the trait positively reflects parenting ability (Kirkpatrick 1985; Heywood 1989; Hoelzer 1989). Hence, females should benefit directly through their choice of highly ornamented mates by obtaining those that invest more in parental care than would other potential mates. The differential allocation model (Burley 1986) assumes that females choose mates for indirect benefits (good genes for the offspring), and predicts a negative relationship between parental care and ornamentation. As a consequence, females mated to highly ornamented males should invest more in parental care than females mated to less ornamented males. Carotenoid-based plumage colour is a secondary sexual characteristic commonly expressed in avian species. The expression of this trait results from numerous physiological and behavioural processes that determine carotenoid intake, absorption, transport and deposition in feathers, bills or exposed fleshy structures (Brush 1990). As such, carotenoid-based colour may function as an indicator of genetic quality (Hamilton & Zuk
? 1998 The Association for the Study of Animal Behaviour
119
120
Animal Behaviour, 55, 1
1982; Hudon et al. 1989), as well as foraging ability (Hill 1992), presence of parasitic and other diseases (Schaeffer at al. 1988) and hormone levels (Brush 1967; Temple 1974). The northern cardinal, Cardinalis cardinalis, is a sexually dichromatic, socially monogamous species. Both sexes have orange-red bills and dull red-brown remiges and rectrices that are nearly invariant in colour (R. Breitwisch, unpublished data). Males also have carotenoid-based bright red plumage with especially intense red colour on the breast. Although much duller in colour, females might also be considered ornamented. Body coverts are tan overall, but females display variable red colour on underwing median and lesser secondary coverts, and variable numbers of red feathers on the crest, face, upper breast and flank. The underwing coverts of females always display the brightest red colour in female plumage. In field studies in a northeastern U.S. population, the brightest red male cardinals paired with the earliest breeding females, obtained territories with the densest vegetation and produced more fledglings than duller males (Wolfenbarger 1996). In our southwestern Ohio population, males fed nestlings at higher rates than did females, but the relationship between level of provisioning and ornamentation was unknown (Filliater & Breitwisch 1997). The aim of this study was to investigate the relationship between both male and female ornamentation and parental care, measured as effort in provisioning nestlings. The good parent model predicts that brighter individuals invest more in parental care than do duller individuals. Brighter males and females are expected to provide more care than do duller individuals, independent of the mate’s level of ornamentation. Alternatively, the good parent model allows for the possibility that the mate’s colour influences a parent’s level of care, with individuals feeding less if they are mated to bright birds. The differential allocation model predicts the opposite with regard to male investment in parental care. Under this hypothesis, bright males are expected to feed less than duller males, while females mated to bright males are expected to feed proportionately more than females mated to dull males. The differential allocation model does not make a prediction with regard to female colour because only males are assumed to allocate their reproductive effort
between mating and parental efforts based on their level of ornamentation. We examined two hypotheses to determine whether plumage colour functions as an indicator of parental care, as proposed by the two models. The first hypothesis is that an individual’s level of ornamentation predicts its absolute parental care effort. The second hypothesis is that an individual’s level of ornamentation predicts its relative parental care effort in relation to its mate’s effort.
METHODS We conducted this study at Aullwood Audubon Center and Farm (39)52*N, 84)16*W), 15 km northwest of Dayton, Ohio, U.S.A., during the 1995 and 1996 breeding seasons. The property consists of approximately 80 ha of primarily reclaimed farmland with a variety of habitats, including an abundance of woody, disturbed habitat ideal for cardinals. We observed feeding in 22 pairs, 15 in 1995 and seven in 1996. We observed two males and one female both years, but used only the more complete set of observations from 1 year to represent them. We were unable to obtain plumage scores for two additional females. To be consistent throughout, we used 17 complete pairs for all data analyses. We conducted duplicate analyses where possible on the larger samples of 20 males and 19 females, and results were similar to those reported here based on 17 pairs. We compared feeding rates, feeding rates per nestling, proportion of feedings, brood size and colour measurements between years. There was no difference between years for any comparison (Mann–Whitney U-tests: all P>0.05), and we combined years in analyses. There was no difference in the distribution of hatching dates in the two years (Kolmogorov–Smirnov two-sample test: D=0.30, N1 =10, N2 =7, P>0.10). For the 2 years combined, we compared hatching date of each nest with the combined feeding efforts of the parents to determine seasonal effects, but found none (Spearman rank correlation: rs = "0.39, N=17, P=0.11). We captured adults with potter traps or mist nets only early in the season or during incubation to avoid disturbance during nest building, egg laying and feeding. Each was banded with a U.S. Fish and Wildlife Service aluminum band and three coloured celluloid bands in unique
Linville et al.: Cardinal parental care and colour combinations. We did not use red colour bands, because bands similar in colour to a species’ ornament influence reproductive success in several other species (Burley et al. 1982; Hagan & Reed 1988; Metz & Weatherhead 1991). We determined an individual’s colour by matching plumage colour to the most similar Munsell chip colour. We recorded breast colours for males and underwing colours for females. Although we did not recapture birds the same season, we tested colour scoring for reliability with museum skins at The Dayton Museum of Natural History and found scoring to be highly repeatable (rS =0.93, N=22, P<0.001). Since colour is a three-dimensional categorization, we converted Munsell scores to a onedimensional continuum similar to that in Burley & Coopersmith (1987). Colour value was weighted primary, however, with hue secondary, and chroma tertiary, to prevent the possibility of a highly saturated pink outranking a darker, less saturated red-orange. Although any such hierarchy is subjective, this ordering of colour chips produced a bright to dull ‘redness’ continuum that agreed with our perception. For male breast colour, scores ranged from 6.0 (Munsell 7.5 4/16) to 0.5 (Munsell 8.75 5/16). For female underwing colour, scores ranged from 13.5 (Munsell 8.75 4/16) to 1.0 (Munsell 10 6/13). Female cardinals also have variable numbers of red body coverts. We determined female ‘body colour’ by summing the scores for relative numbers of red feathers in the crest, cheek, eyebrow, upper breast and flank. We scored these in cheek and flank regions as 0 (red feathers absent) or 1 (a few red feathers present) and in eyebrows, crest and breast as 0 (red feathers absent), 1 (a few red feathers present) or 2 (several to many red feathers present). Males and females also have bright orange to red bills. Even though bill colour may be a sexually selected trait, we found it difficult to quantify bill colour because of its subtly mottled appearance. Thus, we did not examine bill colour as an ornament. The nestling period for cardinals is 10 days, with day 0 representing hatching. We conducted feeding observations on days 2–8 of nestling life. We did not include days 0 or 1 because females spend long periods brooding and, when not brooding, are often fed by males off the nest and out of sight of the observer. Such feeding may be viewed as indirect provisioning of nestlings by
121
males, and females sometimes deliver this food to nestlings. Day 9 was usually the last day nestlings remained in the nest, but fledging sometimes occurred on that day. We conducted observations for 1 h each day and recorded the number of feedings made by both members of the pair. Most observations were conducted before noon, but time of day has been shown not to influence feeding rates (Filliater & Breitwisch 1997). Fewer than 7 days of observations were completed if either the nest was found after day 2 or chicks were preyed upon. Two pairs are represented by 7 days of observation, eight pairs by 6 days, four pairs by 5 days, and three pairs by 3 days. We used three measures of nestling provisioning. First was the number of feedings/h, which we assumed to be an absolute measure of effort. However, effort per nestling is also important to both a nestling and to the mate and partially controls for differences in brood size. Accordingly, we also recorded effort as feedings/ nestling/h. Last, the proportion of a pair’s total feeding effort undertaken by a male or female is a measure of the response of each individual to the needs of the nestlings in relation to the response by the mate. It is unknown which of these three measures is most likely to have the greatest effect on the fitness of both the individual and its mate. Consequently, we report all three. We attempted to measure load size as well as feeding frequency. Cardinals carry food deep in the bill, and food quantity cannot be determined by direct observation. We were unsuccessful in our attempts to measure load sizes by the use of pipe cleaner collars. Foods delivered by parents were frequently mashed, and food boluses slid past the point of constriction. We standardized feeding rates by comparing individual rates to the mean feeding rates of all other individuals sampled on a given day of nestling life. An individual’s standardized feeding rate was the positive or negative difference between the the individual’s feeding rate and the mean feeding rate for that sex. We then totalled these standardized rates per day for an individual to give an overall standardized feeding rate for all of the days sampled. This was divided by the number of days sampled for the individual to yield a mean standardized feeding rate compared with all others of the same sex in the study population. This method controls for the increase in feeding
Animal Behaviour, 55, 1
frequency with nestling age and incomplete sampling of pairs. We calculated relative feeding effort as the proportion of feedings by a male or female based on all of the feedings by the pair. Although the male proportion of feedings changed over the days sampled (see Results), incomplete sampling for some pairs was not biased with regard to sampling within the 7-day period (Kruskal–Wallis one-way ANOVA: H6 =9.35, P=0.97); thus, calculation of proportions simply based on all data for a pair is reasonable. Parental care increases with age and experience in some species. In this population, very few birds banded as nestlings remain in subsequent years. Therefore, we do not know the age of individual cardinals. Banded adults tend to remain in subsequent years on or near the territory where they were banded, and we measured years at the study site as an indicator of local experience and, indirectly, of age. We used these data to assess the relationship between experience and parental care. We used Kendall rank correlations (T reported) to relate absolute feeding efforts and colour (hypothesis 1). We used Kendall partial rank correlations (Txy · z reported) to eliminate the effects of the mate’s colour in relating feeding proportion to plumage colour (hypothesis 2). We used Wilcoxon signed-ranks tests to compare feeding levels by mates and colour changes between years. We used the sequential Bonferroni technique (Rice 1989) to evaluate the significance of correlations when multiple comparisons (i.e. two female ornaments) were used to test the same hypothesis. We report the significance level associated with the outcome of each statistical analysis. In cases where we accept the null hypothesis and the significance level is between 0.05 and 0.20, we also report the statistical power of the test at detecting a significant difference at an á of 0.05. RESULTS General Feeding Pattern Males contributed more care to nestlings than females with a mean feeding effort of 2.2 feedings/h compared with 2.0 feedings/h, but these differences were not significant (W=32, N=17, P=0.33), in contrast to results of Filliater
0.8 Male proportion of feedings
122
0.7 0.6 0.5 0.4 0.3 0.2
1
2
3
4 5 6 Nestling age
7
8
9
Figure 1. Mean& proportion of feedings by the male parent as a function of nestling age for days 2–8. N=8–15 pairs per day (Kendall rank correlation: T=0.67, N=7 days, P=0.0001).
& Breitwisch (1997). Feeding efforts were correlated between mates for both feedings/h (T=0.40, N=17, P=0.013) and feedings/nestling/h (T=0.41, N=17, P=0.011), in agreement with Filliater & Breitwisch (1997). Furthermore, the proportion of feedings by males increased with nestling age (T=0.67, N=7 days, P=0.0001; Fig. 1). Absolute Feeding Effort and Ornamentation Male feeding effort as feedings/h was not correlated with male breast colour (T=0.09, N=17, P=0.31). Male feeding effort as feedings/ nestling/h was also not correlated with male breast colour (T=0.21, N=17, P=0.12, power=80%). Female wing and body colour were not correlated (T= "0.01, N=17, P=0.29). Female feeding effort/h was not correlated with either underwing colour (T=0.24, N=17, P=0.090, power=68%) or body colour (T= "0.08, N=17, P=0.33). Similarly, there was no correlation between feedings/ nestling/h and body colour (T= "0.07, N=17, P=0.39). There was, however, a positive correlation between feedings/nestling/h and underwing colour (T=0.40, N=17, P=0.012). This significant correlation remained even after reducing the testwise á to 0.012 by the Bonferroni technique, based on multiple tests (4) of the same hypothesis. We also determined whether an individual’s absolute feeding level was correlated with its
Linville et al.: Cardinal parental care and colour 0.8 Female proportion of feedings
Male proportion of feedings
0.8 0.7 0.6 0.5 0.4 0.3 0.2 2.5
123
3.5 4.5 5.5 Male breast colour
6.5
0.7 0.6 0.5 0.4 0.3 0.2
4
5
6
7 8 9 10 11 Female wing colour
12
13
Figure 2. Correlation between the proportion of male feedings and male breast colour (T=0.41, N=17, P=0.011).
Figure 3. Correlation between proportion of female feedings and female underwing colour (T=0.46, N=17, P=0.005).
mate’s colour. Female feeding effort as feedings/h was not correlated with the mate’s breast colour (T= "0.26, N=17, P=0.074, power=69%), but there was a negative correlation between female feedings/nestling/h and the mate’s breast colour (T= "0.38, N=17, P=0.017). Because of multiple tests (2), the Bonferroni technique reduces the acceptable á value to 0.025. This relationship remains significant, with females mated to bright males feeding less per nestling than females mated to duller males. Neither male feedings/h (T= "0.18, N=17, P=0.16, power=64%) nor male feedings/nestling/h (T= "0.19, N=17, P=0.14, power=67%) was correlated with female underwing colour.
male breast colour was held constant, female wing colour remained positively correlated with female feeding proportion (Txy · z =0.41, N=17, P=0.011). Even when an apparent ‘outlier’ from the brightest female is eliminated, the significant correlation remains.
Relative Feeding Effort and Ornamentation The male proportion of feedings was positively correlated with male breast colour (T=0.41, N=17, P=0.011; Fig. 2). When the potential influence of female wing colour was held constant, male breast colour remained positively correlated with male feeding proportion (Txy · z =0.35, N=17, P=0.025). Even when an apparent ‘outlier’ from the dullest male is eliminated, the significant correlation remains. The female proportion of feedings was positively correlated with female wing colour (T=0.46, N=17, P=0.005; Fig. 3), but not with body colour (T= "0.27, N=17, P=0.065, power=75%). When
Feeding Efforts and Local Experience Of 17 males on the study site, one had been present for 5 years, three for 4 years, one for 3 years, seven for 2 years, and five for 1 year. Experience was not correlated with feedings/h (T=0.08, N=17, P=0.33), feedings/nestling/h (T=0.07, N=17, P=0.35) or proportion of feedings (T= "0.09, N=17, P=0.31). Of 17 females on the study site, four had been present for 4 years, seven for 2 years, and six for 1 year. Experience was not correlated with feedings/h (T=0.11, N=17, P=0.27), feedings/ nestling/h (T=0.07, N=17, P=0.35) or proportion of feedings (T= "0.07, N=17, P=0.35). Age and Colour Nine males from 1994 and 1995 and eight males from 1995 and 1996 were measured to determine whether breast plumage brightness changed with age. There was no change from 1994 to 1995 (W= "18.0, N=9, P=0.30), but there was an increase in brightness from 1995 to 1996 (W=31,
Animal Behaviour, 55, 1
124
N=8, P=0.02). One individual was represented in both samples. Eight females from 1994 and 1995 and six females from 1995 and 1996 were measured to determine whether underwing plumage brightness changed with age. There was no change from 1994 to 1995 (W= "15.0, N=8, P=0.22), but there was a marginal increase from 1995 to 1996 (W=15.0, N=6, P=0.06). Two individuals were represented in both samples. Ornaments in Pairs There was no significant correlation between male breast colour and female underwing colour in pairs (T= "0.22, N=17, P=0.11, power= 60%), or between male breast colour and female body colour in pairs (T= "0.27, N=17, P=0.065, power=75%).
DISCUSSION Ornamentation and Feeding Effort Colour as ornamentation in male cardinals did not serve as a predictor of absolute feeding effort, and the first hypothesis, that an individual male’s ornamentation predicts absolute parental care effort, was refuted. This hypothesis was supported for females, however. Female underwing colour was positively correlated with female per-nestling feeding effort. Within the context of particular pairs, where the colours of both sexes were known, the relative feeding effort by both males and females was predicted by their own colours. Thus, the second hypothesis, that an individual’s ornamentation predicts relative parental care effort, was also supported. Furthermore, the absolute feeding level by females was inversely related to the mate’s colour; females mated to bright males fed less per nestling than did females mated to dull males. Therefore, we suspect that the pattern within pairs of relative feeding effort and colour is due, in part, to a female behavioural response to the mate’s colour. Investigators of sexual selection hypotheses concentrate on male ornamentation, for obvious reasons. Even when testing indicator models that predict provisioning levels, investigators typically test for correlations between male ornament expression and absolute provisioning level by
males. Three studies have shown positive relationships (Grant & Grant 1987; Hill 1991; Palokangas et al. 1994), supporting the good parent model. On the other hand, four studies have shown negative relationships (Studd & Robertson 1985; Burley 1988; de Lope & Møller 1993; Sundberg & Larrson 1994), supporting the differential allocation model. No one has yet explained why different species of ornamented, biparental birds lend support to one or the other model. Our finding, that accounting for female ornamentation helps explain the pattern of nestling provisioning in cardinals, suggests that in future studies of parental effort in biparental birds, both female ornamentation (where present) and female feeding effort should also be examined. Female cardinals are active participants in decisions made that pertain to the provisioning of nestlings, as argued for the general case by Breitwisch (1989) and Gowaty (1996). Females mated to bright males may increase their reproductive success by decreasing their feeding contribution while preparing physiologically for the successive nesting attempt. Between 70 and 80% of all cardinal nests in this population are lost to predators (Filliater et al. 1994). Pairs do not strongly defend nests and appear to have adopted re-nesting as a strategy to deal with predation (Nealen & Breitwisch 1997). Since only females build nests, incubate and brood, replenishing energy reserves for the successive nesting may be a better strategy than increasing feeding effort to nestlings. We lack complete nesting records for the birds in this sample and cannot yet test this hypothesis. A remaining question is how male cardinals apportion reproductive effort between parental effort and mating effort in seeking extra-pair fertilizations. Investigators have found either that support for the good parent model is accompanied by a low incidence of extra-pair fertilizations not associated with male ornamentation (Hill et al. 1994), or that support for the differential allocation model is accompanied by a pattern of extra-pair fertilizations favouring highly ornamented males (Møller 1994; Burley et al. 1996; Sundberg & Dixon 1996). The relationship between male ornamentation and extra-pair fertilizations in cardinals is more complicated, with brighter males losing paternity more than duller males (S. U. Linville & R. Breitwisch, unpublished data).
Linville et al.: Cardinal parental care and colour Female Ornamentation The sequestering by females of bright red pigmentation in underwing feathers might be considered evidence that carotenoids in female diet inevitably appear as bright red plumage. That is, female underwing colour may be a compromise between strong selection for ornamentation in males and crypticity in females rather than a sexually selected trait in females (Lande 1980; Muma & Weatherhead 1989). We found a significant correlation, however, between female underwing colour and absolute feeding rate, suggesting that colour is an indicator of female quality and thus may be maintained by sexual selection. We have also observed that this patch of red plumage on females is quite visible in flight, but we know of no display by females that clearly exposes and presents the underwing surface to either males or other females (Lemon 1968). We lack data on whether female ornamentation is related to success in intra-sexual interactions, although female cardinals are quite aggressive towards other females (S. U. Linville & R. Breitwisch, personal observation). Female expression of ornaments elaborated in males of dichromatic species has been little studied. Muma & Weatherhead (1989, 1991) found no support for a function of red epaulets in female red-winged blackbirds, Agelaius phoeniceus, in either female interactions or male mate choice. Hill (1993) found that male house finches, Carpodacus mexicanus, preferred the most brightly plumaged females in choice experiments but that female colour was not correlated with any of Hill’s measures of female quality in a wild population. Our results suggest that coloration of female cardinals serves as an indicator of quality, in contrast to the findings of these other studies. Age and Ornamentation Several investigators have found that ornamentation serves as an age indicator (Grant & Grant 1987; Sundberg & Larsson 1994), but others have found this not to be so (Hill 1990; Møller 1991). The advantages to an individual choosing an older mate may relate to parental or other experience that increases with age or the demonstration of survival influenced by genotype. We obtained mixed results pertaining to the question of whether ornamentation in male cardinals is an age
125
indicator. The breast colour of individual males increased in brightness from one breeding season to the next in one of the 2 years of the study. Wolfenbarger (1996) found in a different population of cardinals that male breast colour brightness increased with age. The dullest areas of red breast plumage increased in brightness, although the brightest red breast plumage did not change colour. The availability of carotenoids on our study site may have contributed to the absence of a change in one of 2 years, as we discuss elsewhere (Linville & Breitwisch, 1997). There has been almost no study of female ornamentation and age in birds. Hill (1993) found that first-year female house finches were more colourful than older females. Muma & Weatherhead (1989) found that epaulets of female red-winged blackbirds were brighter when the birds were recaptured a second year. We found no indication that coloration of female cardinals is related to their age. Ornamentation of Mated Pairs Absolute feeding efforts by mates were correlated, and individual ornamentation predicted the relative feeding effort by both sexes. The question that arises is why cardinals do not mate assortatively. Might the nestlings of highly ornamented pairs receive even more food? Ornamentation in cardinals is probably an imprecise predictor of parental care, because colour is directly influenced by an individual’s diet (thus, indirectly by territory quality) and, also, may increase with age. Cardinal pairs in this population tend to remain together year after year (R. Breitwisch & S. U. Linville, unpublished data), and the experience gained with the mate in long-term relationships may influence provisioning decisions as much as ornamentation predicts the pattern of provisioning nestlings. ACKNOWLEDGMENTS We thank Pat Weatherhead and an anonymous referee for helpful comments on this manuscript. For permission to conduct this research at Aullwood, we thank the Director, Charity Krueger, and Head of Research, John Ritzenthaler. We also thank all of the Aullwood staff for their continuing support. For field assistance, we thank Josh Banks, Nancy Beckman,
Animal Behaviour, 55, 1
126
Jeffrey Biessman, Tim DeMarco, Brittany Linville, Steve Ramey and Sharlene Regan. We also thank Gary Coovert at the Dayton Museum of Natural History for use of museum skins. This research was supported by a grant from the Ohio Biological Survey, and University of Dayton Graduate and Faculty Summer Fellowships. This project was conducted under USFWS Banding Permit No. 22351 and ODNR Banding Permit No. 5-57-04, both issued to R. Breitwisch.
REFERENCES Andersson, M. 1982. Female choice selects for extreme tail length in a widowbird. Nature, Lond., 299, 818– 820. Andersson, M. 1986. Evolution of condition-dependent sex ornaments and mating preferences: sexual selection based on viability differences. Evolution, 40, 804– 816. Breitwisch, R. 1989. Mortality patterns, sex ratios, and parental investment in monogamous birds. Curr. Ornithol., 6, 1–50. Brush, A. H. 1967. Pigmentation in the scarlet tanager, Piranga olivacea. Condor, 69, 549–559. Brush, A. H. 1990. Metabolism of carotenoid pigments in birds. FASEB J, 4, 2969–2977. Burley, N. 1981. Sex ratio manipulation and selection for attractiveness. Science, 211, 721–722. Burley, N. 1986. Sexual selection for aesthetic traits in species with biparental care. Am. Nat., 127, 415–445. Burley, N. 1988. The differential-allocation hypothesis: an experimental test. Am. Nat., 132, 611–628. Burley, N. & Coopersmith, C. B. 1987. Bill color preferences of zebra finches. Ethology, 76, 133–151. Burley, N., Krantzberg, G. & Radman, P. 1982. Influence of colour-banding on the conspecific preferences of zebra finches. Anim. Behav., 30, 444–455. Burley, N. T., Parker, P. G. & Lundy, K. 1996. Sexual selection and extrapair fertilization in a socially monogamous passerine, the zebra finch (Taeniopygia guttata). Behav. Ecol., 7, 218–226. Darwin, C. 1871. The Descent of Man and Selection in Relation to Sex. London: John Murray. Filliater, T. S. & Breitwisch, R. 1997. Nestling provisioning by the extremely dichromatic northern cardinal. Wilson Bull., 109, 145–153. Filliater, T. S., Breitwisch, R. & Nealen, P. M. 1994. Predation on northern cardinal nests: does choice of nest site matter? Condor, 96, 761–768. Fisher, R. A. 1930. The Genetical Theory of Natural Selection. Oxford: Clarendon Press. Gowaty, P. A. 1996. Field studies of parental care in birds: new data focus questions on variation among females. Adv. Study Behav., 27, 477–531. Grant, B. R. & Grant, P. R. 1987. Mate choice in Darwin’s finches. Biol. J. Linn Soc., 32, 247–270.
Hagan, J. M. & Reed, J. M. 1988. Red colour bands reduce fledging success in red cockaded woodpeckers. Auk, 105, 498–503. Hamilton, W. D. & Zuk, M. 1982. Heritable true fitness and bright birds: a role for parasites? Science, 218, 384–387. Heywood, J. S. 1989. Sexual selection by the handicap mechanism. Evolution, 43, 1387–1397. Hill, G. E. 1990. Female house finches prefer colourful males: sexual selection for a condition-dependent trait. Anim. Behav., 40, 563–572. Hill, G. E. 1991. Plumage coloration is a sexually selected indicator of male quality. Nature, Lond., 350, 337–339. Hill, G. E. 1992. Proximate basis of variation in carotenoid pigmentation in male house finches. Auk, 109, 1–12. Hill, G. E. 1993. Male mate choice and the evolution of female plumage coloration in the house finch. Evolution, 47, 1515–1525. Hill, G. E., Montgomerie, R., Roeder, C. & Boag, P. 1994. Sexual selection and cuckoldry in a monogamous songbird: implications for sexual selection theory. Behav. Ecol. Sociobiol., 35, 193–199. Hoelzer, G. A. 1989. The good parent process of sexual selection. Anim. Behav., 38, 1067–1078. Hudon, J., Capparella, A. & Brush, A. H. 1989. Plumage pigment differences in manakins of the Pipra erythrocephala superspecies. Auk, 106, 34–41. Johnson, K. 1988. Sexual selection in pinyon jays. I. Female choice and male–male competition. Anim. Behav., 36, 1038–1047. Kirkpatrick, M. 1985. Evolution of female choice and male parental investment in polygynous species: the demise of the ‘sexy son’. Am. Nat., 125, 788–810. Kodric-Brown, A. & Brown, J. H. 1984. Truth in advertising: the kinds of traits favored by sexual selection. Am. Nat., 124, 309–323. Lande, R. 1980. Sexual dimorphism, sexual selection, and adaptation in polygenic characters. Evolution, 34, 292–305. Lemon, R. E. 1968. The displays and call notes of cardinals. Can. J. Zool., 46, 141–151. Linville, S. U. & Breitwisch, R. 1997. Carotenoid availability and plumage coloration in a wild population on northern cardinals. Auk, 114, 796–800. de Lope, F. & Møller, A. P. 1993. Female reproductive effort depends on the degree of ornamentation of their mates. Evolution, 47, 1152–1160. Metz, K. J. & Weatherhead, P. J. 1991. Color bands function as secondary sexual traits in male red-winged blackbirds. Behav. Ecol. Sociobiol., 28, 23–27. Møller, A. P. 1988. Female choice selects for male sexual tail ornaments in the monogamous swallow. Nature, Lond., 332, 640–642. Møller, A. P. 1991. Sexual selection in the monogamous barn swallow (Hirundo rustica). 1. Determinants of tail ornament size. Evolution, 45, 1823–1836. Møller, A. P. 1994. Sexual Selection and the Barn Swallow. Oxford: Oxford University Press. Muma, K. E. & Weatherhead, P. J. 1989. Male traits expressed in females: direct or indirect selection? Behav. Ecol. Sociobiol., 25, 23–31.
Linville et al.: Cardinal parental care and colour Muma, K. E. & Weatherhead, P. J. 1991. Plumage variation and dominance in captive female red-winged blackbirds. Can. J. Zool., 69, 49–54. Nealen, P. & Breitwisch, R. 1997. Northern cardinal sexes defend nests equally. Wilson Bull., 109, 269–278. Palokangas, P., Korpimaki, E., Hakkarainen, H., Huhta, E., Tolonen, P. & Alatalo, R. V. 1994. Female kestrels gain reproductive success by choosing brightly ornamented males. Anim. Behav., 47, 443– 448. Rice, W. R. 1989. Analyzing tables of statistical tests. Evolution, 43, 223–225. Schaeffer, J. L., Tyczkowski, J. K. & Hamilton, P. B. 1988. Depletion of oxycarotenoid pigments in chickens and failure of aflatoxin to alter it. Poult. Sci., 67, 1080–1088. Studd, M. V. & Robertson, R. J. 1985. Sexual selection and variation in reproductive strategy in male yellow
127
warblers (Dendroica petechia). Behav. Ecol. Sociobiol., 17, 101–109. Sundberg, J. & Dixon, A. 1996. Old, colourful male yellowhammers, Emberiza citrinella, benefit from extra-pair copulations. Anim. Behav., 52, 113–122. Sundberg, J. & Larsson, C. 1994. Male coloration as an indicator of parental quality in the yellowhammer, Emberiza citrinella. Anim. Behav., 48, 885–892. Temple, S. A. 1974. Plasma testosterone titers during the annual reproductive cycle of starlings (Sturnus vulgaris). Gen. Comp. Endocrinol., 22, 470–479. Wolfenbarger, L. L. 1996. Fitness effects associated with red coloration of male northern cardinals (Cardinalis cardinalis). Ph.D. thesis, Cornell University. Zuk, M., Johnson, K., Thornhill, R. & Ligon, J. D. 1990. Mechanisms of female choice in red jungle fowl. Evolution, 44, 477–485.