Inbreeding, the mating of related individuals, can represent a significant threat to the health and viability of both individuals and populations. The deleterious effects of mating with ones kin are legion, including: Increased infant mortality (Packer 1979; Ralls and Balou 1982; Smith 1995); increased risk of cancers (Crawford and O'Rourke 1978); Increased rates of spontaneous abortion (Smith 1995b); abusive behaviour towards offspring (Jones 1995); low birth weight (Wildt et al 1982) and lower resistance to infection (O'Brien and Evermann 1988). The basis of these effects, termed inbreeding depression, is genetic, alleles for deleterious effects are usually recessive (Moore and Ali, 1984), thus it is more likely that two copies of a recessive allele will be passed to an infant when two relatives mate than if two random individuals do, given that related individuals can share copies of an allele by common descent. However, the severity of inbreeding necessary before depression is seen can vary between species. This statement is based on the premise that inbreeding, with resultant inbreeding depression, will act to eliminate many deleterious alleles from the gene pool, thus necessitating a closer relationship between mated individuals before any effect will be seen (Ralls and Balou 1988). Smith (1995a) discussed this proposition in the context of an inbred colony of rhesus macaques (Macaca mulatta); he showed that where individuals had a co-efficient of relatedness of less than 0.125 no depression was seen. The coefficient of relatedness, given the symbol r, is a measure of how likely two individuals are to share alleles via common descent; an r-value of 0.125 corresponds to two half-siblings. This level was much greater than other species he examined, leading Smith to suggest that M.mulatta was especially inbred at some point in its evolutionary history.
From the preceding evidence one can clearly see that inbreeding is something to be avoided wherever possible, a process that can be undertaken in two ways:
with their relatives. (Itani, 1972)
And/or
The first point, sex-biased dispersal, is extremely common, with males dispersing at sexual maturity while females remain philopatric in a wide range of species (Clutton-Brock and Harvey, 1976; Greenwood, 1980). An "incest taboo" as the second method is termed (Imanishi 1966), is more subtle and difficult to detect, despite this researchers have found that a strong inhibition, preventing mating between maternal kin, is common, being found in: Papio cynocephalus hamadryas (Kummer, 1968; Alberts and Altmann, 1995); Papio cynocephalus anubis (Packer, 1979); Macaca fuscata; Perry and Manson 1995); Macaca sylvanus (Paul and Kuester, 1985); M.mulatta (Missakian 1973; Smith 1995b) Cercopithecus aethiops (Clutton-Brock, 1988). One can surmise that this inhibition is based on familiarity than some intrinsic knowledge of ones r-value with an individual. This supposition is based on a number of studies, Paul and Kuester (1994) found that female M.sylvanus strongly avoided mating with males who had acted as alloparents for them during their infancy yet who were unrelated, suggesting that familiarity early in life is the key to avoidance. Perloe (1992) found a similar pattern in a M.fuscata group, the alpha and beta males never mated over two mating seasons, Perloe (1992) felt that these two males were victims of their own success, having dominated the group for an unusually long time thus they were familiar to all the sexually active females without having actually sired any of them.
Over the years debate has, if not raged, then certainly simmered as to the ultimate cause of this sexually biased dispersal pattern amongst mammals, and especially amongst the primates. Given the wealth of populations and captive colonies available for study, the focus soon fell upon the Cercopithcines, and more specifically upon those living in a polygynous mating system, the macaques, baboons and a solitary species of Cercopithecus, the vervet monkey C.aethiops. In order for Itani's (1972) theory of inbreeding avoidance to hold it would be necessary to show a cost to non-dispersal, Craig Packer achieved this in 1979. In January of that year he reported the result of a study of the less famous residents of Tanzania's Gombe National Park, the olive baboons (P.anubis).
Whilst examining male dispersal patterns a fortuitous accident occurred. Several years earlier a male, Bramble, had been born into B troop, as he grew so did his troop until it eventually fissioned, giving A and B troops. Bramble eventually left B troop, as is normal for his species at about 7 years old, the troop he migrated into however, was A troop, one containing significantly more related individuals than if he had joined another group. Realising this Packer examined the survivorship of infants to one month old of infants sired by Bramble compared to those sired by other males, this period being chosen as the infant is still dependent on its mother and thus less likely to meet an accidental death and also that any congenital problems are likely to take their effect within this time. First Packer found that, when excluding Bramble from the analysis, infant mortality to one month was uniform across all groups, then on including Bramble a huge difference was found. Amongst infants sired by males other than Bramble all but 15.8% survived to one month of age while only 50% of Brambles sires did. This was pointed to as quantitative evidence of inbreeding depression in a wild population; in 1982 Ralls and Ballou followed this with their paper showing that inbreeding increased infant mortality to one month old in Macaca fascicularis and Macaca nigra.
Moore and Ali (1984) argued that male natal dispersal was a result of intra-sexual competition and that any associated avoidance of inbreeding was, at best, coincidental. This position was based on Wade's (1979) discussion of kin selection, wherein he argues that inbreeding, in that it increases the r value of individuals in a group, makes one more able to benefit from kin selection, thus promoting altruistic sociality. One can see the logical basis of this argument, Hamilton's 1966 theory of kin-selection could indeed create "beneficial inbreeding" (Moore and Ali, 1984, p95). Moore and Ali take their position further, arguing for costly outbreeding, stating that it would act to decrease homozygosity, break up genotypes that are adapted to the local environment, and incur the costs usually associated with dispersal i.e. increased predation risk, reduced foraging efficiency and so on. Their adoption of intra-sexual competition as the ultimate cause of male dispersal was, the felt, supported by the aggression directed towards males of an emigrant age (adolescents) by older males. Packer (1984) soon rebuffed this argument; he pointed out that although adolescent males did receive considerable aggression from older males prior to emigration, they in fact faced far greater aggression from the males of the group they tried to join subsequently. This being the case, and allowing for the costs of dispersal, Packer (1984) stated that there must be some major benefit to dispersal with which to offset these costs. Paul and Kuester (1999) were able to show just what this benefit was, examining their captive colony of M.sylvanus at Affenberg, they showed that a male leaving a group containing a high number of female relatives would happily join a group containing a higher number of males than his original group, thus incurring higher intra-sexual competition but avoiding inbreeding at the same time. Further, Paul and Kuester (1999) showed that the decision to emigrate was not based solely on the ratio of females to males, rather it was a function of the ratio of unrelated females to males. By the mid 1980's Itani's (1972) inbreeding avoidance hypothesis had been largely accepted as the ultimate cause of male natal dispersal (but see Moore 1995 for an opposing view).
Debate continued as to the cause of secondary dispersal, a phenomenon again largely associated with males (Greenwood 1980) that being dispersal from a non-natal group. Largely it was felt that males would leave if another group with a higher female to male ratio was available (Packer, 1984). This was the dominant view until 1988 when Tim Clutton-Brock published the results of a small study in Nature. Having surveyed 19 different species across a range of mammalian taxa Clutton-Brock found, in all but one case, that the tenure of a male within a group was less than the age at which females of the species attained sexual maturity. The significance of these findings is that a father will be out of a group before his daughters are old enough to breed, a very effective way of avoiding father-daughter inbreeding. Clutton-Brocks findings have since been re-inforced, Alberts and Altmann's (1995) study of the migratory habits of the P.cyncocephalus of Amboseli showed a peak of secondary dispersal in the first year and a longer peak between the sixth and eighth years of residence. The authors found the initial peak to be caused by young males who did not immigrate successfully and were soon ejected, the second peak however, coincides with the onset of sexual maturity amongst females, and thus inbreeding is avoided. These findings combined with those of Packer (1979) and others pointed to inbreeding avoidance as the ultimate cause of male dispersal, both natal and secondary.
Since this discovery numerous factors have been found that affect the timing of natal and secondary dispersal, the choice of group emigrated to and so on. Sprague (1992) found that male M.fuscata chose their new group on the basis of sex ratio and of their likely position in the new group. That same year E.O.Smith (1992) found that P. Anubis and P.hamadryas chose their new groups on the basis of the number of same age males within the group, seeking to minimise competition. Alberts and Altmann (1995) examined P.cyncocephalus migration in great detail, they found that males dispersed from the natal group sooner if their mother was old, or if the mother died, a similar pattern to that seen in C.aethiops (Cheney 1983). Most intriguingly they found that the sons of high-ranking mothers emigrated significantly sooner than males born to low ranked females. The explanation of this rests on Altmann et al's 1988 observation that high-ranking P.cyncocephalus mothers produced more daughters than sons and vice versa for low ranking females. This being the case high ranking sons will have a lower unrelated female to male ratio and leave sooner than low ranking males who will have a higher unrelated sex ratio and thus can stay longer and exploit this. This ties in with Paul and Kuester's 1999 finding that unrelated sex ratio was a key determinant of the timing of natal dispersal in M.sylvanus. Provisioning has been seen to delay both natal and secondary dispersal, in both Tibetan macaques (M.thibetana Zhao, 1994) and Japanese macaques (M.thibetana, Sprague 1992) the food providing an increased cost to dispersal, yet not resulting in males remaining in their groups, merely leaving later than normal, showing that inbreeding avoidance must provide a strong impetus to emigration. The M.thibetana males studied by Zhao (1994) tended to join neighbouring groups to which other group members had previously emigrated, a phenomenon common to other species (C.aethiops: Cheney and Seyfarth 1983; Cheney 1981. for review see Melnick and Pearl 1988). Nozawa et al (1982) found that M.fuscata dispersal tended to be within one selection of groups, termed a local concentration of groups, which was effectively a circle of 100 km radius, thus groups over 100 km apart exchanged individuals extremely rarely. A similar picture is seen in M.fascicularis (Kawamoto et al 1984). This then raises the question of how these groups manage to remain outbred if they exchange males only between one another? Cheney and Seyfarth (1983) discuss this problem as it relates to the Amboseli vervet population, they argue that this localised dispersal will not be problematic as long as a few individuals always go further away to groups containing no relatives. In contrast to these studies Melnick et al's 1984 study of wild Indian M.mulatta found that natal dispersal was a random procedure, this contrasts with Drickamer and Vessey (1973) who found natal dispersal to be significantly affected by sex ratio, however this last study was of the provisioned M.mulatta of Cayo Santiago, Boelkins and Wilson (1972) having previously pointed out that a comparison between this population and a wild one would not be valid due to the unique nature of the Cayo Santiago population. Kuester et al (1994) found natal dispersal to be tied to the size of ones natal group, the larger it is the more likely it is that there are unrelated females within, thus one can stay longer and not incur the same costs found in a smaller group.
However, despite a strong incest taboo and the "drive" to leave the natal group some individuals do remain in their natal group and breed there. In most reports of this it is impossible to tell whether actual inbreeding has occurred rather than simply mating with an unrelated individual who happens to be in your natal group. Natal mating has been seen in a variety of species, including: P.cynocephalus (Altmann et al 1988); P.anubis (Packer, 1979); P.ursinus (Bulger and Hamilton, 1988); P.hamadryas (Sigg et al 1982); Theropithecus gelada (Dunbar 1984); M.mulatta (Missakian 1973; Chapais 1983; Colvin 1986); M.fuscata (Sugiyama 1976; Huffman 1987); M.sylvanus (Paul and Kuester 1985) and C.aethiops (Cheney et al 1988). The vast majority of these studies were purely observational, males were seen to mate in their natal group yet it was all but impossible to find out if they had actually reproduced. Until 1988 paternity could only be assessed by recording matings on the day of peak fertility, a problematic situation given that many female primates are highly promiscuous at this time, immunological assays or blood protein analysis, both of which methods can be very ambiguous given that most primates have low levels of variability in these systems (Inoue et al 1991; Rogers 1992). However, in 1988, Weiss et al utilised the great advances that had been made in DNA extraction and analysis to examine primate paternity. The discovery of minisatellites, regions of DNA containing a variable number of repeated codons by Bell et al (1982) provided the basis for the 1985 creation, by a team lead by Alec Jeffreys, of probes that bound to a common core sequence found in each human minisatellite (Jeffreys 1985). If cut with a restriction enzyme and then run through an electrophoretic gel these probes will deliver the individuals DNA fingerprint, a pattern of bands unique to that individual. Weiss et al (1988) took this human technology and applied it to four Old World monkeys: Macaca silenus; M.fuscata; Erythrocebus patas and Colobus guereza. Human probes were found to hybridise happily with the DNA from these species, thus minisatellites became accepted as a powerful tool in assigning primate paternity.
Rapidly minisatellite paternity analysis began to be used in the investigation of reproductive success in polygynous primates. It had long been assumed that reproductive success was rank dependent, the alpha male produced more offspring than the beta male and so on down the hierarchy, in purely behavioural terms this did seem to be the case. Perloe (1992) found the higher a male M.fuscata ranked, the more consortships it had, these consortships lasted longer and occurred closer to the females peak of fertility. A similar pattern was seen in: the Affenberg M.sylvanus population (Paul et al 1993); another M.fuscata group (Inoue et al 1993); M.mulatta (Berard et al 1993); M.arctoides (Bauers and Hearns 1994); M.fascicularis (de Ruiter et al 1994) and P.anubis (Packer 1979).
Weiss et al's (1988) discovery led the way for some very interesting results indeed. Inoue et al (1991; 1992; 1993) used a combination of behavioural observation and minisatellite analysis to assign paternity in a M.fuscata group. As stated above, from observational data the aforementioned dominance-reproductive success relationship was found, yet, when the DNA data was examined the picture was very different indeed. The number of offspring produced was unrelated to rank the opposite to what was expected. In 1993 Berard et al, working on a small group of the Cayo Santiago M.mulatta found no link between rank and reproductive success. A study of the Affenberg M.sylvanus, performed in 1993 by Paul et al found a similar pattern, using blood samples from all individuals taken over a five year period they discovered that of seventy five infants sired by the thirty three sexually mature males of the colony the alpha male did indeed produce the most offspring and a rank-offspring relationship was found. However, Paul et al point out that this relationship only exists because of the non-existent reproductive success of the sexually mature, yet not sexually active sub-adult males. This observation fits well with the Bercovitch-McMillan hypothesis, a synthesis of the findings of two papers (Bercovitch 1986; McMillan 1989) both of which found that the inclusion of sub-adult males produced a false-positive relationship between rank and reproductive success. McMillan (1989) explains that the low rank sub-adult males mate far less than fully adult males, even in the absence of any competition, and that their exclusion from analysis removes any rank-reproductive success relationship seen previously. This hypothesis has been confirmed in a number of studies (de Ruiter et al 1992; Paul et al 1993). Bauer and Hearns (1994), however, examined reproductive success in relation to rank in M.arctoides, and after excluding sub-adult males they still found a very strong relationship, wherein the alpha male was the only one who sired any offspring over the course of two years of study. This relationship is due to the nature of the reproductive physiology of the species, females have non-synchronous oestrus, thus allowing the alpha male to enjoy a 100% monopoly on mating during the fertile period of each female and also the females extreme preference for mating with the alpha male and the alpha male alone as a defence against infanticide.
We must now ask why no rank-reproductive success relationship exists. The relationship between rank and peak fertility matings has already been well established, so why does this not translate to offspring numbers? In answering this question one must consider a number of issues, male sexual strategies, sperm competition and depletion being amongst them
Paul et al (1993) explains their lack of a rank-reproductive success relationship in three ways. There are a large number of males in the study population between whom little asymmetry in fighting existed, perhaps as a result of this males were more inclined to form alliances. These males were competing for females undergoing highly synchronised oestrus and who are very promiscuous indeed, mating every twenty to thirty minutes at peak fertility. The authors conclude that if there were a greater asymmetry in fighting ability then perhaps a relationship would be found, this prediction seems to be borne out by the findings of Bear and Hearns study mentioned above.
The very attempt to monopolise matings actually lessens a males likelihood of a higher reproductive success, the more a male mates the lower quality his sperm becomes, thus the more the alpha male mates the higher the likelihood that a lower ranking male can sneak a mating and use his higher quality sperm to inseminate the female.