TRENDS in Ecology & Evolution Vol.16 No.4 April 2001http://tree.trends.com0169–5347/01/$ – see front matter © 2001 Elsevier Science Ltd. All rights reserved. PII: S0169-5347(01)02116-4184ReviewReviewChris Faulkes*School of BiologicalSciences, Queen Mary,University of London,Mile End Road, London,UK E1 4NS.*e-mail:[email protected] BennettDept of Zoology andEntomology, University ofPretoria, Pretoria 0002,South Africa.A characteristic of COOPERATIVELY BREEDING mammals(see Glossary), where individuals other than theparents help care for offspring, is that reproduction ispartitioned unequally among members of the socialgroup (REPRODUCTIVE SKEW), with socially dominantindividuals monopolizing breeding opportunities. Anincreasing body of work is now revealing that thePROXIMATE MECHANISMS underlying this reproductivedivision of labour might differ between even closelyrelated species, and debate continues about the exactrole the breeding female plays in maintaining levels ofreproductive skew. Theoretical models of ‘incompletecontrol by dominants’1and ‘concession’ theoryinvolving ‘peace’ and ‘staying’ incentives2,3(Box1),have been proposed to explain inter- and intraspecificdifferences in reproductive skew. The idea, however,that the dominant breeder exerts, or attempts to exert,some kind of reproductive control over subordinatenonbreeders in the group is implicit in both models.This control could manifest itself in different ways,ranging from, for example, infanticide of the offspringof a subordinate, or interference by dominants withsubordinate mating attempts, to actual suppression ofsubordinate reproductive physiology4. As mostcooperatively breeding mammals live in extendedfamily groups, in many cases, INCEST AVOIDANCE alonecould be argued to be sufficient to maintainreproductive skew, if the only unrelated individuals inthe group are the founding parents.Almost all animals have evolved mechanisms thatprevent them from breeding with close relatives. Thisprotects them against the fitness costs of harmfulrecessive traits and decreased heterozygosity thatmight be manifest in offspring arising from suchmatings5. As molecular genetic techniquesincreasingly enable elucidation of the detailed kinstructure of social groups, the relative roles of differentproximate mechanisms that control reproductive skewcan be properly understood. When does ‘SELF RESTRAINT’in the form of incest avoidance give way to the moreextreme DOMINANT CONTROL that could be manifest asthe suppression of reproductive physiology innonbreeders, or even infanticide of their offspringshould they breed? Are there common mechanismsmediating physiological suppression among divergenttaxa, such as rodents, viverrids, canids and primates6(Table 1), and how might ecological factors influencethe proximate control of reproductive skew?Monopoly of reproduction in African mole-ratsThe African mole-rats of the family Bathyergidae arean ideal taxonomic group with which to tacklequestions about both ultimate factors and proximatemechanisms that control reproduction in a socialspecies, because they have a range of cooperativebreeding strategies and differing degrees ofreproductive skew. Mating preferences encompassboth facultative INBREEDING and obligate outbreeding,and result in clear differences between mole-rat speciesin the way their social systems are maintained.Sociality in African mole-ratsOf the 18 or more species of African mole-rats, four aresolitary. The other species in the genera CryptomysandHeterocephalus exhibit varying degrees of socialityand cooperative breeding. A characteristic feature ofthe social species is a reproductive division of labour,resulting in high reproductive skew and, consequently,large variation in the lifetime reproductive success ofindividuals. Two species, the naked mole-rat(Heterocephalus glaber) and the Damaraland mole-rat(Cryptomys damarensis), fit the classic definition ofEUSOCIALITY derived from social insects7,8. Other mole-rats and cooperatively breeding vertebratesmight also fit this definition9,10. Unlike other species,both Damaraland and naked mole-rats exhibit abehavioural division of labour among the nonbreedinghelpers. In the Damaraland mole-rats, smallerindividuals form a ‘frequent worker’ group, and innaked mole-rats, the frequency of ‘work’ in the form of‘colony maintenance’ behaviour shows a negativecorrelation with increasing body mass11,12. In addition,To exploit ecological niches where constraints have favoured selection forgroup living and cooperation, both vertebrates and invertebrates have evolvedelaborate social systems. In mammals,numerous divergent taxa haveconverged at similar solutions to these ecological challenges (such as fooddistribution and predator avoidance),culminating in the social insect-likebehaviour of the naked mole-rat.Characteristically, breeding is partitionedunequally in such groups, resulting in a ‘reproductive skew’. New researchlinking studies of physiology,behaviour and molecular ecology in African mole-rats is helping us to elucidate why different proximate mechanisms thatcontrol groups of cooperative breeders might have evolved.Family values:group dynamics andsocial control of reproduction inAfrican mole-ratsChris G.Faulkes and Nigel C.BennettTRENDS in Ecology & Evolution Vol.16 No.4 April 2001http://tree.trends.com185Reviewthere is solid evidence for the existence ofmorphologically distinct castes in naked mole-rats,including a disperser morph among males13,14, anddifferences in the vertebrae of the breeding queen15.The ARIDITY FOOD DISTRIBUTION HYPOTHESIS proposes thateusociality and cooperative breeding in African mole-rats could have evolved in response to rainfall patterns,its effects on food distribution and the subsequent costsand risks of foraging and dispersal16,17.Mating system and reproductive skewAll cooperatively breeding mole-rats studied to datehave skewed reproduction, with a single femalenormally breeding with one, but perhaps on occasiontwo or three, males. However, in terms of lifetimereproductive success, skew might differ considerablybetween species. Plural breeding in females appears tobe rare9. The social mole-rats studied to date in thegenus Cryptomysall have a mating system thatinvolves obligate outbreeding, whereas naked mole-ratsare unusual within the Bathyergidae and amongmammals in general because they can inbreed to a highdegree. These incestuous tendencies were