LETTER Severe intergenerational reproductive conflict and the evolution of menopause Mirkka Lahdenpera ¨, 1 * ,† Duncan O. S. Gillespie, 2,† Virpi Lummaa 3,4 and Andrew F. Russell 5,6 Abstract Human menopause is ubiquitous among women and is uninfluenced by modernity. In addition, it remains an evolutionary puzzle: studies have largely failed to account for diminishing selection on reproduction beyond 50 years. Using a 200-year dataset on pre-industrial Finns, we show that an important component is between-generation reproductive conflict among unrelated women. Simultaneous reproduction by succes- sive generations of in-laws was associated with declines in offspring survivorship of up to 66%. An inclu- sive fitness model revealed that incorporation of the fitness consequences of simultaneous intergenerational reproduction between in-laws, with those of grandmothering and risks of dying in childbirth, were suffi- cient to generate selection against continued reproduction beyond 51 years. Decomposition of model esti- mates suggested that the former two were most influential in generating selection against continued reproduction. We propose that menopause evolved, in part, because of age-specific increases in opportuni- ties for intergenerational cooperation and reproductive competition under ecological scarcity. Keywords Cooperation, ecological constraints, grandmother hypothesis, inclusive fitness, kin selection, mother hypoth- esis, reproductive conflict hypothesis, reproductive skew, restraint, suppression. Ecology Letters (2012) INTRODUCTION Hunter–gatherers live in extended family-groups wherein reproduc- tion is monopolised by a minority of group members and non-breeders contribute to rearing the offspring of breeders (Hrdy 2009). In combi- nation, such features are unusual in the animal kingdom, but are common to all cooperative breeding societies (Wilson 1971; Emlen 1995; Russell 2004). Yet, among cooperative breeders, humans are atypical because a significant proportion of non-breeding helpers comprise post-menopausal women. While it remains unclear whether menopause arose following the evolution of prolonged lifespan or before, a key question is why there is a dramatic mismatch between age-specific survival and fertility in women. The Mother (Williams 1957) and Grandmother (Hawkes et al. 1998) Hypotheses suggest that the answer lies in the direct costs of late-life reproduction or in the indirect benefits of increasing the reproductive success of existing off- spring, respectively. Calculations have suggested that menopause could evolve if almost certain maternal death ensued from childbirth late in life (Pavard et al. 2008), particularly in conjunction with benefi- cial grandmother effects (Shanley et al. 2007). However, such extreme risks of childbirth seem improbable, given that the probability of dying seldom exceeds 3% before menopause in populations lacking modern medicine (Cant et al. 2009; Lahdenpera ¨ et al. 2011a). Conse- quently, although inclusive fitness theory (Hamilton 1964) has been successful in accounting for the evolution of sterile workers in social insects (Abbot et al. 2011) and non-sterile helpers in cooperative breeding vertebrates (Cornwallis et al. 2009), whether it can account for the evolution of human menopause is contentious (Hill & Hurta- do 1991; Rogers 1993; Cant et al. 2009). Previous inclusive fitness calculations, however, have ignored inevitable age-specific increases in reproductive overlap between generations of women. Cant & Johnstone (2008) showed using game-theoretic modeling that reproductive conflict between mother and adult offspring could select against late-life reproduction, poten- tially leading to selection against increasing ages of fertility in line with increases in longevity (hereafter the Reproductive Conflict Hypothesis). Simultaneous reproduction between mothers and daughters was unlikely to generate selection against continued reproduction in the elder generation because mothers and daughters will often be equally related to the offspring of the elder-generation mother (R = 0.5), whereas mothers from the elder generation will always be half as related to her daughter’s offspring as her own (R = 0.25 vs. 0.5). By contrast, while mothers-in-laws will be more related to their own offspring than those of their daughter-in-law (R = 0.5 vs. ∼0.25), daughters-in-law will be unrelated to the off- spring of mothers-in-law (R = 0.5 to own offspring vs. 0 to mother-in-law’s offspring). Thus, daughters-in-law will be under greater selection to ‘win’ any conflict over reproduction between the two parties; potentially selecting against continued reproduction in the elder generation (Cant & Johnstone 2008). Reproductive conflict between successive female generations is universal in cooperative animal societies; resulting in temporal or 1 Section of Ecology, Department of Biology, University of Turku, FIN-20014, Turku, Finland 2 Department of Biology, Stanford University, Stanford, CA, 94305, USA 3 Department of Animal & Plant Sciences, University of Sheffield, Sheffield, S10 2TN, UK 4 Wissenschaftskolleg zu Berlin, Institute for Advanced Study, Wallotstraße 19, D-14193, Berlin, Germany 5 Centre for Ecology & Conservation, College of Life & Environmental Sciences, University of Exeter, Penryn, TR10 9EZ, UK 6 Station d’Ecologie Expe ´ rimentale du CNRS USR 2936, 09200 Moulis, France *Correspondence: E-mail : mirkka.lahdenpera@utu.fi †These authors contributed equally to this work. © 2012 Blackwell Publishing Ltd/CNRS Ecology Letters, (2012) doi: 10.1111/j.1461-0248.2012.01851.x