Sexual activity and reproduction delay ageing in a mammal

A trade-off between fecundity and life expectancy is fundamental to most theories on the evolution of life histories and ageing. In fact, sexual reproduction has been shown to reduce lifespan in a large number of species across all biological taxa [1]. The only exception known thus far are eusocial hymenopterans, with reproducing queens usually living much longer than the sterile workers [2]. Here, we show that in a eusocial mammal, a nearly identical life history has evolved. Among mammals, eusociality has evolved only among the subterranean African mole-rats (Rodentia: Bathyergidae), namely in the naked mole-rat Heterocephalus glaber [3] and several species of the genus Cryptomys [4,5]. Eusociality in mole-rats is characterized by monopolization of reproduction by few individuals (usually a single breeding pair), monogamy of the breeding pair, long-lasting philopatry of the offspring resulting in an overlap of adult generations and cooperative brood care [3–5]. Social mole-rats are known for their remarkable longevity, more than 25 years for naked mole-rats [6], but a thorough study comparing ageing rates of breeders and non-breeders is still lacking. Using long-term breeding data (1984–2005) of Ansell ́s mole-rat Cryptomys anselli, we compared life expectancy of breeding and nonbreeding animals held in captivity. Due to strict inbreeding avoidance [4], adult offspring remained reproductively quiescent unless removed from their colonies and paired to an unfamiliar mate. This held true also when one or both breeding animals died. Thus, reproductive division of labor was clear-cut in our experimental stock, allowing assignment of each individual to either the ‘breeder’ or ‘nonbreeder’ group. Both male and female breeders lived approximately twice as long as non-breeders. No difference was found between the sexes in either reproductive caste. With the exception of a single female, all breeders reached an age of at least 6 years, irrespective of sex. The oldest female breeder was 14.9 years and still alive at the end of data collection. The two oldest male breeders reached ages of nearly 20 years. In contrast, in the non-breedergroup, deaths occurred at a constant rate after reaching maturity, and so far all nonbreeders died before their 8th birth date (Figure 1). This pattern was not caused by intrinsic quality differences between breeders and nonbreeders which might have resulted from an unintended bias towards high-class-individuals during the selection of breeders. Our pairing scheme was determined by the availability of unfamiliar mates of similar age rather than by any other factor, and neither female nor male future breeders grew faster than non-breeders in their juvenile phase (Figure 2A). Breeders and non-breeders differ in mean social rank, raising the possibility that social rather than reproductive status causes differential ageing rates. However, in captive Cryptomys anselli colonies there is no competition for mates and barely any for food. As a consequence, aggression levels are very low, and sociopositive behaviours such as allogrooming or huddling, which are shared among all group members, prevail. It is, therefore, unlikely that non-breeders die younger because they suffer from social stress. In support of this assumption, survival probability did not differ between dominant (i.e., early-born) and subdominant (late-born) non-breeders (Figure 2B). We thus conclude that social rank does not affect the pace of ageing in this species. Rate of living can also influence life expectancy. We,