Anim. Behav., 1996, 51, 1165–1173 Why do male birds not discriminate between their own and extra-pair oﬀspring? BART KEMPENAERS* & BEN C. SHELDON² *Netherlands Institute of Ecology, The Netherlands ² Department of Z oology, Uppsala University, Sweden ( R eceived 17 February 1995; initial acceptance 11 A pril 1995; final acceptance 20 October 1995; M S. number: 4858) Abstract. A number of recent models of optimal paternal investment predict that males should alter their investment in oﬀspring in response to changes in paternity or certainty of paternity. One way in which it has been suggested that male birds might do this would be to recognize their own oﬀspring and to discriminate in their favour. Despite frequent statements that nothing is known about whether birds possess this ability or whether they exercise it, there is a considerable body of evidence suggesting that males do not discriminate against non-related oﬀspring. This evidence is reviewed and an attempt made to explain the absence of kin discrimination in males under these circumstances by considering potential mechanisms of kin discrimination. Although there is selection for males to discriminate in favour of their own oﬀspring it is argued that they are unable to do so because of conflicts between the male, female and oﬀspring over signalling identity, and because the circumstances associated with extra-pair paternity disrupt the operation of some mechanisms. Male birds might possess behavioural ‘rules of thumb’ which lead to behaviour that appears similar to oﬀspring recognition, but the sophistication of such rules is likely to be limited by the stochasticity inherent in fertilization.  1996 The Association for the Study of Animal Behaviour The realization that broods of socially monoga- mous birds may often contain nestlings fathered by diﬀerent males has provoked a great deal of research into the behavioural and evolution- ary consequences of mixed paternity broods (Birkhead & Møller 1992). One question that has received considerable theoretical (Whittingham et al. 1992; Owens 1993; Westneat & Sherman 1993) and empirical (e.g. Møller 1988; Westneat 1988; Lubjuhn et al. 1993; Whittingham et al. 1993; Wright & Cotton 1994) attention is the extent to which the share of parental investment performed by a male is influenced by the share of paternity (real and perceived) that a male achieves within a nest. Intuitively, because reproductive eﬀort is costly, it seems reasonable that males should invest less in the care of oﬀspring when they have fathered relatively fewer of them (Trivers 1972), although some models predict that paternal investment will not be a ﬀected by certainty of paternity, if certainty of paternity is fixed for individual males from one breeding attempt to another (Maynard Smith 1977; but see Xia 1992). Other models predict that a male’s optimal amount of parental investment should be influ- enced by certainty of paternity, although the shape of the relationship between the two may be determined by other factors, such as the relation- ship between paternal care and oﬀspring fitness (Whittingham et al. 1992; Westneat & Sherman 1993). The majority of models describing optimal paternal care in relation to certainty of paternity do not require that males are able to recognize, and discriminate against, non-kin. Instead, males are assumed somehow to ‘assess’ their certainty of paternity in a particular nesting attempt and respond by reducing their rate of feeding to the whole brood. Burke et al. (1989) and Dixon et al. (1994) have shown, for polyandrous dunnocks, Prunella modularis, and monogamous reed buntings, Emberiza schoeniclus, respectively, that Correspondence and present address: B. C. Sheldon Institute of Cell, Animal and Population Biology, University of Edinburgh, West Mains Road, Edinburgh EH9 3JT, U.K. B. Kempenaers is now at the Konrad Lorenz Institute for Comparative Ethology, Savoyen- strasse 1a, 1160 Wien, Austria. 0003–3472/96/051165+ 09 $18.00/0  1996 The Association for the Study of Animal Behaviour 1165