vol. 156, no. 6 the american naturalist december 2000 Begging and Parental Care in Relation to Offspring Need and Condition in the Barn Swallow (Hirundo rustica) Nicola Saino, 1,* Paola Ninni, 2 Michele Incagli, 1 Stefano Calza, 1 Roberto Sacchi, 3 and Anders Pape Møller 2 1. Dipartimento di Biologia, Universita ` degli Studi di Milano, via Celoria 26, I-20133 Milano, Italy; 2. Laboratoire d’Ecologie, Conseil National du Recherche Scientifique Unite Mixte de Recherche 7625, Universite ´ Pierre et Marie Curie, Batı ˆment A, 7e `me e ´tage, 7 quai St. Bernard, Case 237, F-75252 Paris Cedex 05, France; 3. Dipartimento di Biologia Animale, Universita ` degli Studi di Pavia, Piazza Botta 9, I-27100 Pavia, Italy Submitted October 14, 1999; Accepted July 20, 2000 abstract: Parents are selected to maximize their fitness by allo- cating care among their progeny in relation to the differential re- productive value of offspring. Nestlings have been hypothesized to signal need for parental care reliably through their begging behavior, but offspring condition as reflected by their reproductive value may likewise affect begging and hence provisioning. We assessed the rel- ative importance of need and condition in determining begging be- havior and feeding rate of nestling barn swallows (Hirundo rustica) through short-term starvation, a challenge to their immune system with a foreign antigen negatively affecting condition, and brood size manipulation. Food deprivation but not condition or brood size manipulation increased nestling begging rate. Parents fed offspring depending on both need and condition but only when feeding broods that were reduced or of normal size. In enlarged broods, offspring received less food per capita than in reduced broods, and parents did not discriminate among nestlings relative to their need or con- dition. Thus, nestlings signal their need by increased solicitation. Parents allocate food to offspring dependent on both need and con- dition, with these effects depending on parental workload as deter- mined by experimental brood size. Keywords: begging, health status, Hirundo rustica, parental care, sig- naling, starvation. Reproduction and parental care entail individuals with costs in terms of survival or future reproduction (Linde ´n and Møller 1989; Partridge 1989; Clutton-Brock 1991; * To whom correspondence should be addressed; e-mail: n.saino@ mailserver.unimi.it. Am. Nat. 2000. Vol. 156, pp. 637–649.  2000 by The University of Chicago. 0003-0147/2000/15606-0006$03.00. All rights reserved. Lemon 1991; Roff 1992; Stearns 1992; Saino et al. 1999a). A fundamental consequence of sexual reproduction is that parents share only half of their nuclear genetic material with their offspring, as does, on average, each offspring with any of its full siblings, paving the way for the oc- currence of a conflict of interests between parents and their progeny, as well as among the offspring (Trivers 1974; Dawkins and Krebs 1979; MacNair and Parker 1979; God- fray 1991, 1995a). Individual offspring are selected to ob- tain larger parental investment than parents are selected to provide, and individual offspring are selected to obtain a larger share of parental investment than their siblings do. Evolutionary theory of parental investment therefore predicts that natural selection should have favored the evolution of the ability of parents to assess offspring qual- ity, since different fitness rewards may result from allo- cation of the same amount of resources to offspring dif- fering in quality. Offspring may reveal their needs because features of the phenotype (e.g., body mass) directly reflect their condition. Alternatively, parents may have to infer indirectly the condition through signals produced by the offspring. Parents are predicted to base their decision on signals reliably reflecting offspring reproductive value, more valuable nestlings presumably being those in better condition. In addition, parents may vary their decision about allocation of food according to its availability and offspring need of care (Haig 1990). Models of offspring signals directed toward parents have attempted to resolve the question whether such signals should reliably reflect the need of offspring (Godfray 1995a, 1995b; reviews in Kilner and Johnstone 1997; Mock and Parker 1997). These models have analyzed this prob- lem by assuming that signalers such as offspring produce a behavior to which another individual may respond. Thus, there is no temporal element in such models. Re- cently, McNamara et al. (1999) have shown that incor- poration of responses into models of evolutionary games can drastically affect the evolutionarily stable strategy, that is, the evolutionarily stable negotiation rules. The reason for such changes in behavior is that negotiations between the chick and the parent and the outcome may differ from