CORRESPONDENCE predominantly from the maternal tissues3 (unlike in animals, where multiple embryos originate from cleavage of the sexual zygote). In 78% of these polyembryonic plants, multiple embryos emerge from the integument and/or nucellus surrounding the sexual embryo; in the remaining 22%, they are derived from the cleavage of the sexual embryo or from the endosperm. Based on such predominance of maternal origin of these additional embryos, it has been proposed recently that polyembtyony in plants represents a maternal counter-strategy of compensating for the loss in brood size brought about by the sibling rivalry among the developing seeds3. It has been argued that, as in several birds, plant offspring (seeds) developing in a fruit (brood) compete among themselves for maternal resources; in extreme situations, the dominant among them may starve the neighbouring subordinate sibs to deaths. In certain situations, the offspring derive additional benefits if they are the lone survivors in the fruit. For instance, in species where fruits are dispersed whole via wind, water or animals, the offspring gain dispersal advantage by being the lone member in the fruits-a. In such situations, offspring may be selected to kill their neighbouring sibs in the fruit, both to gain dispersal advantage and to reduce post-dispersal sibling competition for resources. Any such rivalry among sibs leading to brood reduction (seed abortion) causes fitness loss to the mother, and hence such brood reduction might not be in the interest of the mother. In fact, inclusive fitness models have shown that brood reduction is frequently in the interest of the offspring rather than the mother. Using inclusive fitness and genetic models, Ganeshaiah et a/.3 examined the conditions under which polyembryony would be favoured by the maternal parent and the offspring. Their analysis suggested that in contrast with that for brood reduction, the maternal parent favours the production of additional embryos for relatively small gains of producing them, while offspring do not. That is, polyembryony can evolve more as a maternal than as an offspring strategy. Thus, in plants also, there is an active conflict between the mother and the offspring over the brood size (seed number per fruit). Offspring is selected to reduce the brood size while the maternal parent is selected to increase it; the former seems to be manifested in the form of sibling rival@, and the latter through the production of additional embryos or polyembryony3. The maternal parent seems to easily attain her ends by inducing her own tissues, such as integument and nucellus, to produce additional embryos. This argument implies that polyembryony would be selected under situations of intense sibling rivalry leading to intra-fruit seed abortion. Ganeshaiah et al.3 have offered empirical evidence in support of this. The number of polyembryos produced per seed in several species of Citrus is positively correlated with the extent of seed abortion in their fruits. Furthermore, polyembryony is also shown to occur more frequently than expected in species that exhibit a high rate of intra-fruit seed abortion. Thus, in the war over offspring numbers, the maternal parent in plants appears to deploy polyembryony as a counter-strategy of making good the loss in the fitness that she encounters owing to the sibling-driven brood reduction. It might be interesting to see if such a sociobiological view offers a better explanation for the evolution of polyembryony in animals. R. Uma Shaanker K.N. Ganeshaiah Dept of Genetics and Plant Breeding, University of Agricultural Sciences, G.K.V.K. Bangalore 560 065, India References 1 Craig, SF., Slobodkin, L.B. and Wray, G. (1995) Trends Ecol. Evol. 10,371-372 2 Hardy, I.C.W. (1995) Trends Ecol. Evol. 10,372 3 Ganeshaiah, K.N., Uma Shaanker, R. and Joshi, N.V.(1991) J. Genet. 70,103-127 4 Tisserat, B., Esau, E.B. and Murashige, T. (1979) Hortic. Rev. 1, l-78 5 Uma Shaanker, R., Ganeshaiah, K.N. and Bawa, KS. (1988)Annu. Rev.Ecol. .Syst.19,177-205 6 Ganeshaiah, K.N. and Uma Shaanker, R. (1988) Oeco/ogia77,135-139 7 Bawa, K.S., Hegde, S.G., Ganeshaiah, K.N. and Uma Shaanker, R. (1989) Nature 342,625-626 6 Hegde, S.G., Ganeshaiah, K.N. and Uma Shaanker, R. (1991) Oikos 60,20-26 Age and breeding performance in monogamous birds: the influence of pair stability Breeding performance in birds, especially in long-lived species, increases with agel12. This phenomenon is usually explained either by the optimization of reproductive effort or by age-related improvement of competence2. In the latter case, the increase in breeding success with age may result either from the improvement of foraging skills or social status, or from the increase in breeding experience (or a combination of these factors). In their recent TREEarticle, Forslund and Pitt2 have provided a stimulating discussion about the role of breeding experience, particularly pointing out two current problems in the interpretation of such studies. First, most studies assume an absence of correlation between age of first breeding and phenotypic quality, although earlier breeding by birds of higher quality will induce a positive relationship between breeding success and experience. Second, breeding experience generally increases with age, and statistical or experimental control is needed to separate the two effects2,3. We would like to draw attention to a third problem, not mentioned by the authors, but in our opinion highly relevant to the investigation of the role of breeding experience in the increase of reproductive success in birds. In many studies, breeding experience has only been considered as a property of the individual, independent of its partner. However, breeding success in monogamous birds results from the joint effort of both reproductive partners in incubation duties, brood defence and provision of offspring. In species with long-lasting pair bonds, individuals may improve their breeding performance through repeated breeding attempts with the same partner4. Indeed, improved coordination in breeding activities with the duration of pair-bonds has been proposed as a major benefit of avian monogamy. Previous studies of long-lived bird species generally support this idea of mate retention as an important component of breeding success independent of ages-g, although exceptions have been reported in the manx shearwater (Puffinus puffinus) and the barnacle goose (Branta /eucopsis)ll. Very few studies of age-related breeding success have so far statistically controlled for the duration of the pair-bond. In short-tailed shearwaters (Puffinus tenuirostris) breeding success was found to depend on both breeding experience and the duration of the pair-bond, with the latter appearing to be the stronger effect4. Another promising line would be to contrast effects of breeding experience on breeding performance in species with and without mate fidelity. For instance, breeding experience has little effect on reproductive success in Brandt cormorants (Phalacrocorax pencillatus) with low mate-fidelity13 and no effect in greater flamingos (Phoenicoptefus ruber roseus) with no mate-fidelity14. For some species, a decline in reproductive performance has been observed in older age classesl. Such a decline may result partly from an over-representation of new pairs in older age classes, following the replacement of mates when the first partner dies. Such an effect has been tested only in the barnacle goose, where the decline in breeding success in older individuals was independent of mate retentionl2. We therefore suggest that studies testing for an effect of breeding experience on reproductive performance in species with high mate-fidelity should systematically control for the duration of the pair bond. Frank CCzilly Station Biologique de la Tour du Valat, Le Sambuc, F-13200, Arles, France Ruedi G. Nager CEFE,CNRS, Montpellier, France References 1 S&her, B-E. (1990) in Current Ornithology (Vol. 7) (Power, D.M., ed.), pp. 251-283, Plenum 2 Forslund, P.and PBrt,T. (1995) Trends Ecol. Evol. 10,374-378 3 PBrt,T. (1995)Proc. R. Sot. London Ser. 8360, 113-117 4 Bradley, J.S., Wooller, R.D. and Skira, I.J. (1995) J. Anim. Eco/.64,31-38 5 Couslon,J.C. (1966)J.Anim. Ecol.35, 269-279 6 Coulson, J.C. and Thomas, C. (1985) in Behavioural Eco/o&(Sibly, R.H. and Smith, R.M., eds), pp. 489-503, Blackwell 7 Ollason, J.C. and Dunnet, G.M. (1988) in Reproductive Success (Glutton-Brock, T.H., ed.), pp. 268-278, University of Chicago Press 6 Bradley, J.S., Wooller, R.D., Skira, I.J. and Serventy,D.L.(1990) J.Anim.Ecol.59,487-496 9 Emslie, S.D., Sydeman, W.J. and Pyle, P. (1992) Behav. Ecol. 3,189-195 10 Brooke, M. (1990) The Manx Shea/water, Poyser 11 Forslund, P. and Larsson, K. (1991) Behav. Ecol. 2,116-122 I2 Black, J.M. and Owen, M. (1995) J. Anim. Ecol. 64,234-244 13 Boekelheide, R. and Ainley, D.G. (1989) Auk 106, 389-401 14 CBzilly, F. and Johnson, A.R. Ibis (in press) 27