Better red than dead: carotenoid-based mouth coloration reveals infection in barn swallow nestlings Nicola Saino1*, Paola Ninni2, Stefano Calza1, Roberta Martinelli1, Fiorenza De Bernardi1 and Anders Pape MÖller2 Dipartimento di Biologia, Sezione Zoologia Scienze Naturali, Universita© degli Studi di Milano, via Celoria 26, I-20133 Milano, Italy Laboratoire d'Ecologie, CNRS UMR 7625, Universite¨ Pierre et Marie Curie, 7 quai St Bernard, Case 237, F-75252 Paris Cedex 05, France 1 2
Nestling birds solicit food from their parents by displaying their open brightly coloured gapes. Carotenoids a¡ect gape colour, but also play a central role in immunostimulation. Therefore, we hypothesize that, by di¡erentially allocating resources to nestlings with more brightly coloured gapes, parents favour healthy o¡spring which are able to allocate carotenoids to gape coloration without compromising their immune defence. We demonstrated that, in the barn swallow Hirundo rustica, (i) parents di¡erentially allocate food to nestlings with an experimentally brighter red gape, (ii) nestlings challenged with a novel antigen (sheep red blood cells, SRBCs) have less bright gape colour than their control siblings, (iii) nestlings challenged with SRBCs but also provided with the principal circulating carotenoid (lutein) have more brightly coloured red gapes than their challenged but unsupplemented siblings and (iv) the gape colour of nestlings challenged with SRBCs and provisioned with lutein exceeds that of siblings that were unchallenged. This suggests that parents may favour nestlings with superior health by preferentially feeding o¡spring with the brightest gapes. Keywords: carotenoids; gape colour; Hirundo rustica; immunity; parental care; signalling main circulating carotenoid (lutein) in adult male barn swallows (Hirundo rustica) is negatively correlated with the concentration of immunoglobulins in the plasma and with counts of lymphocytes and heterophils (Saino et al. 1999). This suggests that healthy individuals with relatively low activation of the immune system have more circulating carotenoids. Hence, we hypothesize that carotenoids may be allocated to gape coloration at the expense of allocation to immune function. Infected nestlings may thus be forced to reduce the amount of carotenoids in the soft tissues of their gape, resulting in a duller coloration to be displayed to their parents. This hypothesis is corroborated by the observation that experimentally parasitized birds have duller carotenoid-based coloration of visible integumental tissues than parasite-free individuals (Bletner et al. 1966; Kowalski & Reid 1970; Ru¡ et al. 1974). Evolutionary theory of parent^o¡spring relationships predicts that o¡spring are selected to demand more parental resources than parents are selected to provide and a larger share of parental investment than their siblings (Trivers 1974; Parker et al. 1989; Godfray 1991). Models of the resolution of parent^o¡spring con£ict have shown that costly signals may have evolved which honestly reveal the bene¢ts accruing to needy o¡spring from gaining extra parental resources (Godfray 1991, 1995). O¡spring of many animals solicit food from their parents (Clutton-Brock 1991; Kilner & Johnstone 1997; Kilner et al. 1999). For example, nestling birds obtain food by displaying their open gape while producing begging calls. Parents of some passerine species have been shown to feed o¡spring with an experimentally exaggerated gape colour preferentially (GÎtmark & AhlstrÎm 1997)
1. INTRODUCTION
The coloration of soft tissues and in particular gape coloration in nestling birds is partly based on carotenoids, which produce yellow to red hues of the integumental and other tissues (Ficken 1965; Goodwin 1984). In birds, carotenoids can be stored in their feathers, skin, £anks, eyes, other organs such as the liver and in plasma (Goodwin 1984). Since carotenoids cannot be synthesized by animals and, thus, re£ect ingestion (Goodwin 1984), the intensity of the colour signal of nestling birds should re£ect the amount of carotenoids present in visible gape tissues. Carotenoids are thought to play important roles in immunoregulation and immunostimulation. Carotenoids enhance lymphocyte proliferation, stimulate e¡ector T-cell function, enhance macrophage and cytotoxic T-cell capacities and stimulate the production of various cytokines and interleukins in mammals (Bendich 1989; Chew 1993; Olson & Owens 1998). Lutein is known to enhance antibody production to Tcell-dependent antigens and speci¢cally enhance anti-sheep red blood cell (SRBC) antibody production in mice (Jyonouchi et al. 1995). Moreover, carotenoids have been shown to protect animals from experimentally induced tumours (e.g. Krinsky 1989; Chew 1993). Finally, carotenoids act as oxygen radical scavengers and protectors of lipids from peroxidation (Ames 1983). Since no experimental studies of birds have shown that the e¡ects of carotenoids are similar to those described for mammals, this is an untested assumption. However, the concentration of the *
Author for correspondence (
[email protected]).
Proc. R. Soc. Lond. B (2000) 267, 57^61 Received 2 September 1999 Accepted 7 October 1999
57
& 2000 The Royal Society
58
N. Saino and others
Infection and gape colour in nestling swallows
which reliably re£ects o¡spring need (Kilner 1997). However, parents should be selected to invest in their o¡spring di¡erentially whenever such investment enhances the ¢tness of parents as compared to an even allocation of resources among o¡spring (Howe 1976; Parker & Mock 1987). Parents should therefore assess nestling condition and this requires a mechanism enforcing honesty on o¡spring signal. In the present study, we analysed the e¡ect of the gape coloration of o¡spring on parental food allocation decisions among the progeny in the barn swallow by experimentally altering nestling gape colour. This manipulation is similar to that adopted in previous studies of the response of parental feeding to altered o¡spring signals (e.g. Lyon et al. 1994; GÎtmark & Olsson 1997; Kilner 1997). In addition, we tested whether an experimental challenge to the immune system of nestlings, simulating an infection by a pathogen, resulted in a duller gape colour. Finally, we tested whether the gape colour of experimentally infected nestlings could be restored by arti¢cially provisioning them with the prevailing circulating carotenoid. The barn swallow is an insectivorous passerine with biparental care. Nestlings show typical passerine begging behaviour, giving solicitation calls while opening their gape widely during feeding visits by their parents (MÖller 1994). Gape colour varies considerably among nestlings, ranging from greenish yellow to bright red (N. Saino and A. P. MÖller, personal observation). 2. METHODS The study was carried out during spring in 1997 and 1998 in nine colonies east of Milano, Italy and eight colonies at Kraghede, Denmark.
(a) Experiment 1
This experiment (see ¢gure 1) tested whether parents preferentially fed nestlings with relatively redder gapes. In this experiment, nestlings were marked with pen colours on their white breast feathers for individual identi¢cation and the parental feeding rates of individual nestlings aged 12 days were recorded for 1h around noon. The nestlings were subsequently randomly assigned to one of the following three treatments. (i) Two drops of red food colour (azorobin (E 122) dissolved in citric acid and sodium sorbate) in the gape of a nestling. (ii) Two drops of yellow food colour (ribo£avin (E 102) dissolved in citric acid and sodium sorbate) in the gape of a nestling (a control treatment). (iii) Two drops of physiologically bu¡ered saline (PBS) in the gape of a nestling (a second control treatment). Since more than three nestlings were present in all these broods, the others were left untreated. However, this does not cause any bias since the comparisons were made between nestlings receiving the three treatments. The nestlings were subsequently placed in their nest again and, after leaving them for 15 min to minimize the disturbance from handling, the feeding rates of individual nestlings were observed for 1h. The gape colour manipulation procedure was similar to that adopted in previous studies (GÎtmark & AhlstrÎm 1997; Kilner 1997). Twelve broods from Denmark were used in this experiment. We only used nests in good light conditions and observations were Proc. R. Soc. Lond. B (2000)
made by continuously focusing binoculars on the nest. The assignment of feeds to nestlings was readily observable because only a single nestling was fed with a single, large food bolus during each visit. Since the large bolus requires extensive swallowing movements by the nestlings, the identity of the nestling being fed could be readily ascertained. In a previous study (N. Saino, unpublished results), we made observations of nestlings being fed using two observers on a number of occasions. More than 96% of observed cases (n 31) of food delivery were recorded similarly by the two observers. These observations suggest that the assignment of food delivery to individual nestlings was reliable. Observations of 1h have been repeatedly shown to provide reliable estimates of the absolute and relative feeding rates during a single day. For example, we have shown that the correlation coe¤cients between the absolute feeding rates (i.e. feedings by both parents per unit time) in the morning, around noon and in the afternoon range from 0.71 to 0.76 (Saino et al. 1997; see also MÖller 1994; Saino & MÖller 1995).
(b) Experiment 2
We tested whether challenge of nestlings with a novel antigen (which is supposed to simulate infection) a¡ected gape coloration. In 1997 in Milano, six-day-old nestlings were randomly assigned in pre-established proportions to a control group, while the remaining nestlings were assigned to challenge with sheep red blood cells (SRBCs). In 1998 they were assigned to an untreated control group, a group of nestlings injected with PBS only (a second control) or challenge with SRBCs, after being provided with a numbered aluminium ring to allow later individual recognition. Overall, 94 broods were included in this experiment. Since the number of nests with nestlings receiving a given treatment di¡ered due to di¡erent designs in the two years, the sample sizes di¡er in the three pairwise comparisons (see ¢gure 2). Nestlings injected with SRBCs (n 168) received an intraperitoneal injection of 3.5 107 cells suspended in 30 ml PBS and sham-inoculated nestlings (n 62) received 30 ml of PBS injected intraperitoneally, whereas untreated control nestlings (n 160) were just handled for approximately the same time as the nestlings of the other groups and then put back in the nest. Intraperitoneal inoculation of SRBCs is a standard immunological test, with SRBCs acting as a multigenic antigen eliciting a humoral antibody response in our study organism as well as in other bird species (e.g. Roitt et al. 1996; Saino & MÖller 1996; Deerenberg et al. 1997; Pastoret et al. 1998). On day 12 all nestlings were identi¢ed by a student and ranked according to the colour of their gape without prior knowledge of their treatment. In each brood, the nestling with the most pale gape was given a rank of 1, the second least coloured one was ranked 2 and so forth for the remaining nestlings, allowing for tied scores when they appeared very similar in gape colour. Large colour rank values indicate a brighter gape. The colour ranks were signi¢cantly repeatable across scorers as determined by 58 nestlings being ranked independently by two students (F57,58 22.30, p50.0001 and R (repeatability according to Falconer (1981)) 0.957). A full format photograph of the open gape was taken for 60 nestlings using a circular £ash in a dark room. Nestlings within each brood were then ranked according to their hue and chroma values, respectively, which were obtained by colour analysis of the palate photographs by a portable spectrophotometer (Ocean Optics Europe, Dunedin, FL). The visually assigned colour ranks were strongly correlated with the hue and chroma ranks (unsigned Kendall's t-values 0.89
Infection and gape colour in nestling swallows
(c) Experiment 3
All the nestlings in this experiment were injected with SRBCs six days after hatching. We then tested whether nestlings provisioned with lutein had a brighter gape colour than their unprovisioned siblings in two control groups also injected with SRBCs. Lutein is the most abundant circulating carotenoid in the barn swallow with a mean concentration of 2.69 mg ml 1 (s.e. 0.14 and n 41) in the plasma (Saino et al. 1999). On the second day after hatching, nestlings in 47 broods in Milano were individually marked with a small ink patch on either posterior limb and assigned to one of the following treatments. (i) No food provisioning (group 1): starting on the second day after hatching, every second day these nestlings were taken out of their nests, handled for approximately 20 s and then put back in the nest (a control treatment). (ii) Provisioning with oil (group 2): starting on the second day after hatching, every second day nestlings were given peanut oil (5 ml on days 2 and 4, 15 ml on days 6 and 8 and 25 ml on day 10) (a second control treatment controlling for any e¡ects of the solvent used for the carotenoids). (iii) Provisioning with lutein (group 3): these nestlings were given lutein dissolved in peanut oil following the same time schedule as for group 2 (0.6 mg lutein in 5 ml oil on days 2 and 4, 1.8 mg in 15 ml on days 6 and 8 and 3 mg in 25 ml on day 10). The lutein doses were chosen arbitrarily. However, the maximum dose (on day 10 when nestlings weigh approximately as much as breeding adult males) corresponded approximately to the mean estimated amount in the entire plasma volume recorded in a sample of 41 adult males (2.7 mg assuming an average body mass of 18 g and a 10% blood to total body mass ratio). In broods containing two nestlings, one nestling was randomly assigned to group 3 while the other was assigned either to groups 1 or 2. In broods with three or more nestlings, randomly chosen nestlings were assigned to the experimental groups according to a pre-determined scheme, ensuring that all treatments were represented in each brood unless one or more nestlings died before colour evaluation. There were 56 nestlings in group 1, 50 nestlings in group 2 and 74 nestlings in group 3. The di¡erence in the sample sizes in the pairwise comparisons between groups resulted from one experimental group being unrepresented in some small broods. This gives rise to slightly di¡erent sample sizes in the pairwise comparisons (see ¢gure 3). The nestlings were ranked for gape colour on day 12 as described above.
(d) Experiment 4
The fourth experiment was designed to test whether lutein could restore the normal brightness of gape colour typical of uninjected nestlings by comparing nestlings injected with SRBCs and provisioned with lutein to their unprovisioned siblings which had not been injected with SRBCs. This experiment di¡ers from the previous one because in experiment 3 we Proc. R. Soc. Lond. B (2000)
59
15 before after feeding rate (no. h−1)
and 0.63, respectively and n 60 and p50.0001 in both cases) in the predicted direction, indicating that large visual ranks accurately re£ected the redness and, with less precision, saturation of the gape colour. We analysed the e¡ect of SRBC injection on gape colour rank by comparing injected nestlings with their sham-inoculated and unmanipulated siblings.
N. Saino and others
10
5
0 red
yellow
water
treatment Figure 1. Feeding rates of barn swallow siblings in relation to experimental manipulation of gape coloration. The values are the mean (+ s.e.) feeding rates of individual nestlings per hour. The nestlings received either two drops of red food colourant, two drops of yellow food colourant (a control treatment) or two drops of physiological water (a second control treatment) during the experiment. There was a statistically signi¢cant change in the feeding rate caused by the treatment (see main text). The sample sizes are 12 nestlings for each experimental group. tested whether lutein enhanced the gape colour in nestlings injected with SRBCs but did not provide information about the gape colour of injected nestlings provisioned with lutein in relation to the gape colour of uninjected and unprovisioned nestlings. After being individually marked as in experiment 3, a randomly chosen half of the nestlings from nine broods in Milano were given the same treatment as the nestlings in group 3 in experiment 3, whereas their siblings were considered unmanipulated controls and handled for approximately 20 s every second day, starting on the second day after hatching. There were 22 unprovisioned and uninjected controls and 20 nestlings injected with SRBCs and provisioned with lutein. Gape colour rank was assessed as in the other experiments. 3. RESULTS
Parent barn swallows allocated food to their nestlings in relation to the experimental intensity of the coloration of their gape, as demonstrated by experiment 1 (¢gure 1). There was a highly signi¢cant di¡erence in the change in feeding rate caused by the treatment (one-way ANOVA F2,33 80.54 and p50.001), with the red treatment di¡ering signi¢cantly from the other two (Fisher's protected least-signi¢cant di¡erence test p50.001). Nestlings which had their gape dyed with red food colour received more feedings than nestlings receiving the yellow control treatment or the water control treatment. Hence, parent barn swallows di¡erentially allocated food to nestlings with a more brightly coloured red gape. Experiment 2 tested whether a simulated infection (caused by the injection of SRBCs) a¡ected gape
N. Saino and others
4
gape colour rank
3.5
Infection and gape colour in nestling swallows 4
unmanipulated sham inoculated immunized with SRBC z = 2.51 N = 52 p = 0.012
z = 0.48 N = 52 n.s.
3
z = 3.39 N = 94 p = 0.0007
z = 3.57 N = 43 p = 0.0004
z = 0.42 N = 40 n.s.
2
2
treatment of immunized nestlings
treatment Figure 2. Mean (+ s.e.) gape colour ranks of 12-day-old barn swallow nestlings challenged at day 6 after hatching with SRBCs, sham inoculated with PBS or left untreated. The nestlings immunized with SRBCs had signi¢cantly less brightly coloured gapes than their siblings in the two control groups (Wilcoxon matched-pairs, signed-ranks test). When two or more siblings in a brood received the same experimental treatment, their mean gape colour rank was used in the analyses. Not all experimental groups existed in all broods. Therefore, the sample sizes di¡er between the pairwise comparisons. The sample sizes are pairs of nestlings (or pairs of groups of nestlings). The di¡erences remained signi¢cant after Bonferroni adjustment.
coloration. The gape colour of the nestlings challenged with SRBCs was less bright than that of the unmanipulated controls and nestlings which were sham inoculated with PBS. However, the gape colour rank of the unmanipulated nestlings was not signi¢cantly di¡erent from that of their sham-inoculated siblings (¢gure 2). Challenge of the immune system with SRBCs therefore reduced the intensity of gape coloration. In experiment 3 nestlings challenged with SRBCs and provisioned with lutein had more brightly coloured orange or red gapes than siblings which received only the solvent for the carotenoid (peanut oil) or which did not receive any solvent (¢gure 3). However, the gape colour of the nestlings that did not receive any solvent was ranked very similar to that of their control (oilprovisioned) siblings (¢gure 3). In experiment 4 we tested whether lutein restored the normal brightness of the gape colour in nestlings challenged with SRBCs. The gape colour of the nestlings injected with SRBCs and provisioned with lutein was even more brightly coloured than that of their unmanipulated siblings (mean s.e. for SRBC-injected nestlings provided with lutein 3.37 0.16 and for control nestlings 2.45 0.18; Wilcoxon matched-pairs, signed-ranks test, z 2.37, n 9 (pairs of groups of siblings) and p 0.018). The reduction in gape coloration caused by challenge with SRBCs was thus reversible by provisioning with additional carotenoids. Proc. R. Soc. Lond. B (2000)
z = 4.01 N = 42 p = 0.0001
3
2.5
2.5
unprovisioned fed with oil fed with oil and lutein
3.5 gape colour rank
60
Figure 3. Mean (+ s.e.) gape colour ranks of 12-day-old barn swallow nestlings immunized with SRBCs and given peanut oil, peanut oil with dissolved lutein or no peanut oil. High colour ranks indicate a redder and more saturated gape coloration. When there were two nestlings in the same experimental group, their mean visual colour rank was used in the analyses. One experimental group was not represented in a few broods. Therefore, the sample sizes di¡er between the pairwise comparisons. The sample sizes refer to the number of nestlings (or groups of nestlings). The nestlings which received lutein every two days had more brightly coloured gapes than siblings not provisioned with lutein, whereas the mean gape colour of the nestlings in the two control groups did not di¡er signi¢cantly (Wilcoxon matched-pairs, signed-ranks test). The di¡erences remained signi¢cant after Bonferroni adjustment.
4. DISCUSSION
Parent barn swallows preferentially allocated food to nestlings with a brighter red gape, consistent with ¢ndings from previous studies of other species (GÎtmark & AhlstrÎm 1997; Kilner 1997). Infection simulated by novel challenge to the immune system of nestling barn swallows reduced their gape coloration, which is at least partly based on a carotenoid (lutein). Moreover, dietary lutein was a limiting factor for gape colour expression since extra provisioning was necessary to restore a normal coloration in nestlings in which their colour had been depressed by challenge with SRBCs. The positive correlation between the intensity of coloration and parental e¡ort may seem puzzling in the absence of a mechanism preventing less needy o¡spring from begging at least at the same intensity as their needier siblings. Well-fed o¡spring should be able to beg more vigorously and acquire a disproportionate share of limiting resources than their siblings. Needier nestlings bene¢t di¡erentially from their less needy siblings and, thus, each o¡spring should signal at its optimal level (Godfray 1991, 1995). However, the incremental bene¢ts to parents of di¡erentially feeding o¡spring in better condition may under certain conditions di¡er from the bene¢ts of feeding more needy o¡spring.
Infection and gape colour in nestling swallows We have shown that lutein can enhance the brightness of the gape coloration of nestlings. The skin of poultry contains carotenoids and a reduction in the carotenoid content of their diet depresses skin coloration (Stone et al. 1970; Goodwin 1984). Furthermore, experimental infection of chickens with coccidia depresses their skin coloration and carotenoid plasma levels (e.g. Bletner et al. 1966; Ru¡ et al. 1974). Therefore, the reduction in pigment concentration in bird tissues occurs because (i) parasitism depresses nestling condition and, as a side-e¡ect, reduces the retention of carotenoids in the storage tissues, (ii) the direct uptake by parasites reduces the amount of pigments available to the host or (iii) adaptive mobilization of carotenoids from the storage tissues prevents the use of carotenoids for other functions, as suggested for sexual signals (Shyko¡ & Widmer 1996). In conclusion, this study has shown that gape colour may reveal the health status of barn swallow nestlings to their parents. The mechanism enforcing honesty on gape colour as a signal of health may be mediated by lutein, a carotenoid which is a biochemical determinant of gape colour. We assume that carotenoids in the barn swallow have immunomodulating and immunostimulating activities similar to those observed in mammals (Bendich 1989; Chew 1993). Furthermore, we assume that dietary carotenoids are available in limiting amounts, as shown in the experiment 2. Diseased o¡spring should be unable to allocate as great an amount of carotenoids to gape coloration as their healthy siblings owing to the competing demands for carotenoids of immune function and gape coloration. This enforces the reliability of gape colour as a signal of o¡spring quality. We are grateful to a large number of ¢eld assistants during the ¢eldwork. T. R. Birkhead, E. Danchin, H. C. J. Godfray, M. Pagel and J. Shyko¡ kindly commented on previous drafts of the manuscript. We are also grateful to G. Hill and an anonymous referee for valuable comments and suggestions. This study was supported by Italian Consiglio Nazionale delle Ricerche grants to N.S. and Centre National de la Recherche Scienti¢que (Atipe Blanche) grants to A.P.M. REFERENCES Ames, B. N. 1983 Dietary carcinogens and anticarcinogens. Oxygen radicals and degenerative diseases. Science 221, 1256^1264. Bendich, A. 1989 Carotenoids and the immune response. J. Nutr. 119, 112^115. Bletner, J. K., Mitchell Jr, R. P. & Tugwell, R. L. 1966 The e¡ect of Eimeria maxima on broiler pigmentation. Poultry Sci. 45, 689^694. Chew, B. P. 1993 Role of carotenoids in the immune response. J. Dairy Sci. 76, 2804^2811. Clutton-Brock, T. H. 1991 The evolution of parental care. Princeton University Press. Deerenberg, C., Apanius, V., Daan, S. & Bos, N. 1997 Reproductive e¡ort decreases antibody responsiveness. Proc. R. Soc. Lond. B 264, 1021^1029. Falconer, D. S. 1981 Introduction to quantitative genetics, 2nd edn. New York: Longman. Ficken, M. S. 1965 Mouth colour of nestling passerines and its use in taxonomy. Wilson Bull. 77, 71^75. Godfray, H. C. J. 1991 Signalling of need by o¡spring to their parents. Nature 352, 328^330. Godfray, H. C. J.1995 Signalling of need between parents and young, parent^o¡spring con£ict and sibling rivalry. Am. Nat. 142,1^24. Proc. R. Soc. Lond. B (2000)
N. Saino and others
61
Goodwin, T. W. 1984 The biochemistry of the carotenoids. II. Animals. London: Chapman & Hall. GÎtmark, F. & AhlstrÎm, M. 1997 Parental preference for red mouth of chicks in a songbird. Proc. R. Soc. Lond. B 264, 959^962. GÎtmark, F. & Olsson, J. 1997 Arti¢cial colour mutation: do red-painted great tits experience increased or decreased predation? Anim. Behav. 53, 83^91. Howe, H. F. 1976 Egg size, hatching asynchrony, sex, and brood reduction in the common grackle. Evolution 57, 1195^1207. Jyonouchi, H., Sun, S. & Gross, M. 1995 E¡ect of carotenoids on in vitro immunoglobulin production by human peripheral blood mononuclear cells: astaxanthin, a carotenoid without vitamin A activity, enhances in vitro immunoglobulin production in response to a T-dependent stimulant and antigen. Nutr. Cancer 23, 171^183. Kilner, R. 1995 When do canary parents respond to nestling signals of need? Proc. R. Soc. Lond. B 260, 343^348. Kilner, R. 1997 Mouth colour is a reliable signal of need in begging canary nestlings. Proc. R. Soc. Lond. B 264, 963^968. Kilner, R. & Johnstone, R. 1997 Begging the question: are o¡spring solicitation behaviours signals of need? Trends Ecol. Evol. 12, 11^15. Kilner, R. M., Nolle, D. G. & Davies, N. B. 1999 Signals of need in parent^o¡spring communication and their exploitation by the common cuckoo. Nature 397, 667^672. Kowalski, L. M. & Reid, W. M. 1970 Some e¡ects of roxarsone on pigmentation of chickens infected with coccidia. Poultry Sci. 49, 1405. Krinsky, N. I. 1989 Carotenoids and cancer in animal models. J. Nutr. 119, 123^126. Lyon, B. E., Eadie, J. M. & Hamilton, L. D. 1994 Parental choice selects for ornamental plumage in American coot chicks. Nature 371, 240^243. MÖller, A. P. 1994 Sexual selection and the barn swallow. Oxford University Press. Olson,V. A. & Owens, I. P. F. 1998 Costly sexual signals: are carotenoids rare, risky or required? Trends Ecol. Evol. 13, 510^514. Parker, G. A. & Mock, D. W. 1987 Parent ^ o¡spring con£ict over clutch size. Evol. Ecol. 1, 161^174. Parker, G. A., Mock, D. W. & Lamey, T. C. 1989 How sel¢sh should stronger sibs be ? Am. Nat. 133, 846^868. Pastoret, P., Gabriel, P., Bazin, H. & Govaerts, A. 1998 Handbook of vertebrate immunology. San Diego, CA: Academic Press. Roitt, I., Brosto¡, J. & Male, D. 1996 Immunology. London: Mosby. Ru¡, M. D., Reid, W. M. & Johnson, J. K. 1974 Lowered blood carotenoid levels in chickens infected with coccidia. Poultry Sci. 53, 1801^1809. Saino, N. & MÖller, A. P. 1995 Testosterone-induced depression of male parental behaviour in the barn swallow: female compensation and e¡ects on seasonal ¢tness. Behav. Ecol. Sociobiol. 36, 151^157. Saino, N. & MÖller, A. P. 1996 Sexual ornamentation and immunocompetence in the barn swallow. Behav. Ecol. 7, 227^232. Saino, N., Bolzern, A. M. & MÖller, A. P. 1997 Immunocompetence, ornamentation and viability of male barn swallows (Hirundo rustica). Proc. Natl Acad. Sci. USA 97, 579^585. Saino, N., Stradi, R., Ninni, P., Pini, E. & MÖller, A. P. 1999 Carotenoid plasma concentration, immune pro¢le and plumage ornamentation of male barn swallows (Hirundo rustica). Am. Nat. (In the press.) Shyko¡, J. A. & Widmer, A. 1996 Parasites and carotenoidbased signal intensity: how general should the relationship be ? Naturwissenschaften 83, 113^121. Stone, H. A., Collins, W. M. & Urban Jr, W. E. 1970 Evaluation of carotenoid concentration in chicken tissues. Poultry Sci. 49, 675^681. Trivers, R. L. 1974 Parent^o¡spring con£ict. Am. Zool. 14, 249^264.
