The evolution of parasite resistance has often been assumed to be governed by antagonistic selection pressures. Defense against pathogens, by mounting an immune response, confers evident benefits but may also incur costs, so that the optimal level of defense is expected to depend on the balance between benefits and costs. Although the benefits of immune surveillance are well known, estimates of costs are still equivocal. Here we studied the behavioral and physiological modifications associated with exposure to a nonreplicating antigen (lipopolysaccharide [LPS] of Escherichia coli) in a passerine species, the house sparrow (Passer domesticus). We further investigated whether the behavioral and physiological changes provoked by LPS induced measurable repercussions on life-history traits, such as the breeding effort and reproductive success. Finally, we tested whether the trade-off between immune activation and breeding effort was modulated by the workload required to feed the brood. Exposure to LPS reduced activity and increased body mass loss of captive individuals; similarly, LPS injection induced a dramatic drop in feeding rate and reproductive success of breeding females. However, this reduction depended on brood size, suggesting that the strength of the trade-off between immune activation and reproduction was affected by the workload required to feed the brood. Overall, this study stresses the magnitude of costs associated with mounting immune responses and the ecological and evolutionary consequences for natural populations.

Summary1. We tested the condition-dependent hypothesis of sexual advertisement in house sparrows (Passer domesticus). Male house sparrows have a bib of black feathers which serves as both a badge of social status and as a cue for female choice. We manipulated environmental conditions during the premoult period of juvenile house sparrows kept in outdoor aviaries. Birds were assigned to two treatments differing in the amount of dietary proteins, which are known to aect the expression of immune response in birds. We tested whether birds in the protein-rich group had better immune responses and developed larger bibs than birds reared on a protein-poor diet. We also checked whether immune response was a predictor of survival and parasite resistance. 2. Individuals with higher cellular immune response at capture had greater probability to survive during the 3 months of the experiment, and they had a higher probability to recover from infection with Haemoproteus sp. (a blood parasite). Conversely, birds with high immunoglobulin concentrations at capture had a higher probability of mortality. 3. Birds on the protein-rich diet had a higher cellular immune response compared to birds in the protein-poor treatment. Humoral immune response showed the opposite pattern, being higher for birds in the protein-poor treatment. We did not ®nd any eect of food quality on the development of the badge, assessed as the size of the trait and its colour properties. 4. In conclusion, our results support the view that immune defences are important for survival and parasite resistance in natural populations, and that they might be costly to produce. On the other hand, we did not ®nd support for the conditiondependent hypothesis of sexual advertisement, suggesting that the badge may not be a costly trait to produce. However, badge size could re¯ect other aspects of condition. The kind of pigments involved in colour signals may be the key factor determining the production costs of such traits.

Nestling birds produced later in the season are hypothesized to be of poor quality with a low probability of survival and recruitment. In a Spanish population of house martins (Delichon urbica), we first compared reproductive success, immune responses and morphological traits between the first and the second broods. Second, we investigated the effects of an ectoparasite treatment and breeding date on the recapture rate the following year. Due probably to a reverse situation in weather conditions during the experiment, with more rain during rearing of the first brood, nestlings reared during the second brood were in better condition and had stronger immune responses compared with nestlings from the first brood. Contrary to other findings on house martins, we found a similar recapture rate for chicks reared during the first and the second brood. Furthermore, ectoparasitic house martin bugs had no significant effect on the recapture rate. Recaptured birds had similar morphology but higher immunoglobulin levels when nestlings compared with non-recaptured birds. This result implies that a measure of immune function is a better predictor of survival than body condition per se.

Parasites are usually assumed to negatively affect fitness components of their hosts, particularly during adverse environmental conditions. We experimentally tested whether the ectoparasitic house martin bug (Oeciacus hirundinis) had larger negative effects on its house martin (Delichon urbica) host during second clutches. Conditions for reproduction decreased during the season, being less favourable for the second clutch than the first clutch. This was demonstrated by smaller clutch sizes, lower breeding success and smaller and lighter nestlings during the second clutch, as compared to the first, in the absence of parasites. Experimental manipulation of the intensity of parasitism resulted in the reproductive success of the host being reduced with increasing numbers of house martin bugs. This effect was much stronger during the second, as compared to the first, clutch. The house martin bug therefore had a greater detrimental effect on the reproduction of its host when conditions for reproduction were poor.

No abstract