Attractiveness can be defined as the ability of being preferred as a mate, and thus attractive traits are expected to indicate the reproductive value of an individual (Rhodes, Simmons, & Peters, 2005). There are two main hypotheses for the evolution of mating preferences: a runaway process (Fisher, 1930), predicting that an initial random preference for a trait can eventually lead to an increase in its frequency if attractive individuals have more offspring, and the good genes hypothesis (Hamilton & Zuk, 1982), which predicts that traits genuinely reflecting genetic quality are preferred. In humans, attractiveness encompasses several physical and psychological traits (Buss, 1989 and Gangestad and Scheyd, 2005), and the facial phenotype, conveying critical information such as sex and health, is likely to play a key role in a mate choice context. There is also some evidence that facial attractiveness is more relevant than body attractiveness in hypothetical mate choice decisions (Currie & Little, 2009).
So far, three main properties have been suggested to influence facial attractiveness: symmetry, averageness, and sexual dimorphism (Rhodes, 2006). While several studies have found a relationship between these attractive traits and markers of mate quality (e.g., Rhodes et al., 2003 and Thornhill and Gangestad, 1999), a considerable number of studies have failed to detect any significant link (e.g., Rhodes and Simmons, 2007 and Zebrowitz and Rhodes, 2004). In addition, it is unclear whether markers of mate quality reflect genetic quality or are instead cues of the phenotypic condition of the individual, unrelated to genetic quality.
Finally, even if preferences towards facially attractive individuals have now been well established (Langlois et al., 2000), only a few studies have investigated whether these preferences are adaptive, i.e., whether facial attractiveness is heritable and is associated with higher fertility. Livshits & Kobyliansky (1989) found evidence for phenotypic similarity between parents and offspring for a combined measure of body and facial symmetry, while for the majority of specific facial measures, only a weak correlation was found. More recently, a study based on a sample of parent–offspring pairs from Scotland reported a significant positive correlation between parents' and daughters' facial attractiveness but not between parents and sons (Cornwell & Perrett, 2008). The association between facial attractiveness and reproductive success is also not clear. In the Ache of Paraguay, for instance, women with the most attractive faces were found to have 1.6 times higher fertility than those of average attractiveness (Hill & Hurtado, 1996: 312). In contrast, a study in Poland failed to find a correlation between women's facial attractiveness and the number of offspring (Pawlowski, Boothroyd, Perrett, & Kluska, 2008). However, according to a study based on a large sample of individuals from the USA, attractive men and women have, respectively, 12% and 16% more children than less attractive individuals (Jokela, 2009). Finally, a recent study found that attractive individuals in a sample of Slovak men are more likely to be married and consequently have higher number of children than unattractive individuals (Prokop & Fedor, 2011; but see Prokop, Obertová, & Fedor, 2010)
Besides the Ache study (Hill & Hurtado, 1996: 312), no other research has been conducted in a high-fertility and polygynous population with intense mating competition, although this socioecological setting is the most frequent among human societies (Marlowe, 2003). Moreover, in other studies, only short-term measures of reproductive success have been used (i.e., number of children). However, since the transition to low fertility rates observed in postindustrial populations is hypothesized to result from a higher demand of parental investment per child which possibly maximizes long-term reproductive success (Lawson & Mace, 2011), the use of multigenerational data may be required to investigate the fitness outcomes of facial attractiveness in modern populations.
The main objective of this study is to investigate whether facial attractiveness brings reproductive benefits in contemporary populations with low uptake of modern contraceptive methods. Indeed, modern birth control may not only lead to a disconnection between mating opportunities and fertility in men (Pérusse, 1993), but it is also possible that it affects mate preferences in both men and women (see Alvergne & Lummaa, 2010 for review). First, we used multigenerational historical data from the USA (men from the West Point Military academy born in the 1930s and graduating in the 1950s; note that the first birth control pill was available in 1960) to test whether facially attractive men have more children (completed fertility) and grandchildren. Second, we used data from contemporary rural Senegal where contraceptive prevalence is low [<15% (Wickstrom, Diagne, & Smith, 2006)] to investigate the link between facial attractiveness and age-specific reproduction in both men and women.
