Dominance as Adaptive Stressing and Ranking of Males, Serving to Allocate Reproduction by Differential Self-suppressed Fertility: Towards a Fully Biological Understanding of Social Systems.

A Theory that Dominance (Hierarchy) Functions Not to Ameliorate Conflict, But to Create it;

and that this is Not Over Resources But the Goal of Reproduction

Moxon SP (2009) Medical Hypotheses 73(1) 5-14


Steve Moxon (independent -- non-affiliated researcher)
1 / 589 Crookesmoor Road, Broomhill, Sheffield S10 1BJ (UK)


Dominance is a biological concept of an asymmetric 'power' relationship between (any pair of) individuals, as a result of previous encounters with others biasing likelihood of contesting. That this requires dedicated neural structure shows that dominance is adaptive; and it's usually thought that fitness is increased through dominance (hierarchy) minimising mutually unproductive contest over resources, and/or determining access to or control over resources. But highly inconsistent data indicates that this operational definition is too wide, and given clear evidence that dominance is invariably same-sex, it would seem instead to function primarily to allocate reproduction.
    Dominance contest exposes individual differences in metabolic vigour especially, but also in various other, including sophisticated attributes; and by a self-organising process there is ranking of same-sex individuals in a hierarchy. But this achieves nothing in itself without an integral mechanism of corresponding individual variable self-suppression of the physiology re reproduction – and mate-choice with rank as the criteria. Reproductive suppression would appear to vary along a continuum, from in some species (most 'cooperative breeders') a 100% reproductive skew with total suppression of all individuals bar the sole breeder to, in most others, a gradient down the length of the dominance hierarchy. The mechanism in most species is directly either hormonal or pheromonal, on top of an indirect consequence of the stress caused by relatively low rank.
    Dominance would seem to have evolved as a major instrument of the proposed 'genetic filter' function of the male, whereby in effect accumulated deleterious genetic material is 'quarantined' in the male half of the lineage from where it is purged, so as to keep this source of reproductive logjam away from females, thereby to avoid amplifying the problem of the female being necessarily the limiting factor in reproduction.
    The theory makes predictions mutually exclusive of the consensus model, that dominance/ DH is:
*   same-sex only;
*   present whenever, within one or both sexes, there is potential conflict over reproduction, and there is no mechanism to preclude this, but otherwise is absent;
*   always associated with some degree of differential physiological reproductive suppression.

This new conceptualization of dominance has major implications for the social as well as biological sciences, in that resource-competition models of the basis of sociality will have to give way to a thoroughgoing biological understanding that places centre-stage not resources but reproduction; with consequent radical revision of notions of 'power'.

1. Dominance and how it's misconceived

The current conceptualization of dominance and dominance hierarchy (DH) is problematic. As with any phenomenon, until it's more fully understood there's only a provisional (operational) definition; this being informed by the practical difficulties of how data is collected, and is necessarily circular.[1]
    The ’working definition’ of dominance at its narrowest is simply the victory of an individual in a single contest, irrespective of what the contest is over. Less narrowly, dominance is a more general relationship of an individual to others, gauged by looking at different sorts of encounters and with a number of others (though always dyadic, even though it's not clear that dominance is not established through more complex interaction). However, given that the phenomenon is only vaguely conceived, then it's likely that an observed interaction is one where dominance is only one or not the main factor and is therefore only indirectly and inaccurately measurable; or where dominance is not in play at all. It's possible, therefore, to obtain data that is mostly or even entirely spurious in terms of the real phenomenon lying at the heart of its merely operational understanding.
    There's recent research demonstrating this very problem. [2] An attempt to ascertain DH in a social group of a particular species by the various accepted methods, usually results in several different DHs, with individuals ranked in different orders and the DHs varying in linearity to the point possibly of being web-like rather than pyramidal. This is what we would expect if dominance in fact concerns only one subset of what has been taken to be a wide range of types of encounters appropriate to study.
    An encounter is often taken to be one of dominance simply because of an apparent victory, without checking for signalled submission/ dominance, or even if any sort of contest took place. But submission/ dominance signalling itself cannot be taken as conclusive given that co-option for courtship signalling is evident in many species; from mammals down to the most primitive that exhibit DH, including crayfish. [3] Simple avoidance of conflict is taken as evidence of submission: an animal may be said to have run away, when actually it just declined to initiate engagement. And in many species even approach-avoidance occurs insufficiently often for this dubious inference of dominance to be the basis of measuring it.
    These problems lead to researchers themselves in effect creating the very interactions they measure as dominance. In many studies, animals are held captive in non-species-typical social organisation and provided with artificial food-competition scenarios. Linda Fedigan (who reviews the issues raised here and various definitional problems of dominance) remarks:

'In fact, it has been found that food tests do not generalize to other conflict situations in any consistent way. Not only does the test-situation only exist in the artificial laboratory-test setting, it has been found that priority to food does not necessarily correlate with priority to other incentives, and that dominance determined through dyadic tests does not generalize to dominance relationships for the same individuals within the group as a whole. ... Rather than peeling away the layers of the behavioral onion, to arrive at the core of an underlying “real” dominance rank or dominance relationship, it can be argued that the experimenter has in fact created the dominance relationship.'

Spurious instances of dominance interactions are not filtered out because all purported instances are valid according to the operational definition. With the conception the broadest possible, this obscures the possibility of a narrower one that may usefully give clues re a theory of mechanism and function that can lead to testable prediction. So we've a likely false consensus that dominance serves to minimize mutually unproductive conflict over resources.
    Of course, up to a point dominance is bound to so function, but this in no way implies that this is the primary function. Ritualized signalling, staged escalation and transitive reasoning are refinements that appear to have evolved for this 'purpose' [transitive reasoning [4, 5] is where males are able to ’fill in the gaps’, as it were: that, for example, if A is dominant to C, and B is sub-dominant to A, then B must be dominant to C; young boys are capable of doing this in a social context but not in any other], indicating that minimising mutually unproductive conflict is more likely how dominance subsequently became additionally adaptive rather than why it was so originally.
    That dominance is an adaptation is beyond doubt in that to rank in a DH an individual requires a neural structure re 'competition memory' -- the ability to store data on the outcomes of past encounters with a range of individuals according to the use of 'winner' and/or 'loser' effects [6] -- to bias the likelihood of future engagement in conflict. Computer modelling [7] confirms these minimum requirements and also shows that nothing beyond the behavior of the individual  participants is required. [The DH that arises from dominance interactions is epiphenomenal, in that it self-organizes without any supra-individual level of organisation beyond dyadic interaction of participants being necessary.] It's therefore appropriate to also approach dominance as an attribute of the individual – both as a genetically based predisposition subject to environmental impact, and in terms of the hormonal signatures (cortisol, serotonin and testosterone) of the rank attained. [8] Indeed, it's the focus on dominance as a relationship that tends to exclude consideration of what gives rise to it, thus serving to so strongly retain the wide operational definition, which as Robin Dunbar [9] has pointed out, provides us with mere description instead of explanation.
    There's one particular facet of how dominance currently is conceived that in being revealed as spurious provides the insight as to the real nature of the phenomenon.

2. Dominance is not inter-sexual

If dominance contests as maintained in the consensus position are simply contest over any resource; then all individuals would contest freely with all other individuals, regardless of sex. Yet the most obvious facet of dominance across the animal phylogeny as far back as its possibly earliest manifestation (the pre-insect phyla of crustacea) is of the predominance of male intra-sexual contest. Crayfish in particular and also shrimps have been well studied to confirm that in the wild males form DHs, with males exhibiting different behavior to the opposite sex. [10] Dominance hierarchy is notable as sex-typically male social organisation, although there may well be female DH as well; and in species where there's no male sociality, DH may be exclusively female.
    It's regularly observed across the animal kingdom that female DH (where it occurs) is separate and usually much weaker than the male equivalent. Furthermore, it's generally agreed, whereas mate-value clearly corresponds to contested dominance rank for males, female mate-value often appears to be based on fertility, which is not an attribute that can be acquired through contest, being a 'given' (a product of mainly heritable factors and age). Consequently, female rank rather than being the result of contest is in many species simply inherited from the mother.
    All this points to an obvious possible narrower function of dominance specifically to do with differentially allocating access to sex/ reproduction – as is indicated in so many species by the classic trials of strength preceding or coinciding with the start of the mating season. Here it would make no sense for males to contest with females or vice-versa, but instead to have only same-sex conflict, [11, 12] as is now apparent in analysis of human competition. [13, 14] This is what we would expect in that -- for the obvious reason that the female is the limiting factor in reproduction and the smaller, weaker sex, so should not risk injury through agonistic encounter; and males should not give sexual selection opportunities to rivals by allowing them to offer to females protection from such attack. This drives male deference to females, which seems to be an entirely different phenomenon to submission (albeit that the two terms are frequently taken to be synonymous): either a simple non-engagement in recognition of the female’s interests (that are not in conflict with, or are tied up with a male's own), or the signalling of an instrumental yielding. The contrast between intra-sexual dominance-submission and inter-sexual (male-to-female) deference was common parlance in human society until recently, and this is evident in data from analysis of human behavior. [15]

There has been much study of what is assumed to be dominance between the sexes where males ‘give up’ food to females. This is a type of encounter between the sexes where researchers have consistently made a firm conclusion that dominance is at issue. It's routinely asserted that one sex (in the great majority of species the male, but in several species the female) is always dominant to the other. That is, all individuals of one sex supposedly are dominant to all individuals of the other. A moment‘s reflection on why it would be that dominance in whichever direction is 100% shows a fatal flaw in this notion. A total one-way asymmetry in dominance between two very large normal variant sub-groups of an animal species that must to some extent overlap in relevant criteria (such as body size between males and females) must be due to some other factor, not dominance.
    Researchers reveal in their contradictory statements that they are confused. It's said, for example, that chimpanzees have separate male and female DHs, but also that young males first dominate females in a series of encounters and only then join the male DH at its foot and seek to climb it. The ring-tailed lemur is also said to have separate male and female DH, but that the females are dominant to the males. The reality though, must be that either the DHs are separate and males and females do not have cross-sex dominance interactions, or the DH is as one and there's inter-sexual dominance. It can't be both.
    The most interesting cases are those species, like the ring-tailed lemur, where ‘female dominance’ is supposed. In these cases, this may be despite the fact that there's sexual dimorphism in size in favor of the male, making the interpretation in terms of dominance still more unlikely. For the very reason it's supposedly ‘female dominant’, the gray mouse lemur recently has been closely researched (by Ute Radespiel [16]). Only a single interaction in an entire study was judged an exception to what Radespiel termed 'female dominance' (99.9%). Radespiel also notes that only in a small number of instances did both animals show any aggressive behavior: the males were very rarely aggressive to the females, even when the opportunity was there to reciprocate.
    'Continuous spatial proximity between the sexes so far has not been observed in the field', Radespiel admits. So the inter-sexual interactions observed in the study may be a wholly artificial scenario. This possibility fits with the data from the study that most conflict occurred not as usually expected between the sexes in connection with food, but simply 'in the spatial context'. The definition of agonistic conflict had to be defined down to include simple approach/ avoidance. Radespiel admits that in the wild: 'the submissive partner can more easily prevent an escalation of conflicts by fleeing and avoiding the dominant partner'. The simpler interpretation is that any encounter is simply avoided.
    The ring-tailed lemur is a classic case of a species where there are inter-sexual encounters during feeding. The species is very well documented in this respect. [17, 18] When a male is feeding near to females, a female might come and take food from out of his hands. Far from an agonistic encounter, the female hardly so much as snatches the food, whilst the male not only does not resist but remains quite placid and unresisting. It's not even a case of giving the food upon request, because there's no active yielding. The encounter is simply a non-interaction.
    There has been a lot of discussion about ‘female dominance’ that is apparent only in the context of feeding, being actually male deference in female ‘feeding priority’. This was the conclusion by Kappeler [19] re cross-sex aggression in lemur species generally. White [20] concluded that male ring-tailed lemurs defer to females because they thereby become preferred mates. Pochron and Wright [21] summarized their work on a species of lemur known as the Milne-Edwards’ Sifaka, that: 'despite behavioral similarities between intra- and inter-sexual aggression, dominance within sexes differs fundamentally from dominance between sexes'. Their insight was that the probability of winning an intra-sexual dominance contest was for mammals generally higher with age, so if the nature and function of cross-sex aggression was similar, then outcomes should show the same relation to age. For sifakas, victories indeed do increase according to age in male-male encounters, but there's no relationship whatsoever in the case of cross-sex interactions. 'This supports the hypothesis that intra- and inter-sexual aggression play markedly different roles in lemur society.'
    If what has been categorised in common as dominance in the intra- and inter-sexual scenarios apparently 'differs fundamentally' and plays 'markedly different roles', then these are distinct forms of aggressive interaction with insufficient in common to be regarded as the same phenomenon. In line with the converging evidence, only the intra-sexual form should be termed dominance. The picture is interestingly complicated in the Mongoose lemur, where there are high levels of female-on-male aggression (but little male-on-female) apparently related less to feeding priority than to mate monopolization by females. [22, 23, 24] There's scope for much of what usually little inter-sexual aggression there is in any species to be explained not by dominance but by ‘mate-guarding’.
    The conclusion that male ‘yielding’ of food to females is not dominance but female ‘feeding priority’ fits with the now well researched phenomenon of resource partitioning as the basis of sexual dimorphism, [25] whereby the sexes of many mammals and birds show a marked divergence in either the types of foods they eat or the modes of foraging for it, or even changed morphology. Some bird species have different types of beak according to sex, so that the sexes have no option but to forage in different ways or for different foods. The most well-known case of this in behavioral form is that of the giraffe, where the males put out their long necks vertically to forage only in the high branches, whereas females (whose necks are just as long) take leaves at heights no more than they can reach by putting out their necks horizontally.

To avoid the possibility of an inappropriate dominance contest and aggressive behavior, the evolutionary process evidently has produced a mechanism to engage such behavior only if the individual encountered is same-sex. A single gene (TRP2) in mice crucial for sexing any other conspecific individual encountered, when ‘knocked out’ causes males to behave as if all others encountered are females. Instead of trying to repel males, the male mouse tries to mount them just as if they were females. [26, 27] This reveals that default behavior is not dominance-submission but sexual; and that sexing is necessary before dominance-submission behavior may be initiated. The same is true of females. [28] The only possible explanation for this is that one sex only – the same sex -- is to be engaged with in dominance-submission terms. This is very strong evidence indeed that dominance is always a same-sex phenomenon, in turn strongly implying that dominance is an adaptation functioning to allocate access to sex.
    In further support of this, there is the profound sex-differential physiology re testosterone and cortisol to do with competition. There is not the space here to review the literature on this; suffice to say that a merely quantitative sex difference would not be inconsistent with the consensus model of dominance, because a predominance of male intra-sexual competition would be expected given that the female is the limiting factor of reproduction. But it is evident that the sex difference is very clearly a qualitative one, and of a whole system with several components.
    That the phenomenon of dominance turns out to be more profound than hitherto had been realised, is what would be expected of an adaptation that could not have evolved without very substantial 'policing' (see below). It's hard to see how the benefits of ameliorating competition over resources would constitute sufficient selection pressure, given that resources (essentially food and shelter) are merely instrumental to homeostatic control of the body, with demand for them therefore inherently self-limiting – whereas (for the male) sex / reproduction is not. There's little that animals and specifically even our own ancestors could do to usefully and successfully hoard resources. Indeed, the one major resource exploited by extant hunter-gatherer societies – large animal prey -- is hunted for subsequent equitable sharing across the whole social group. [29] With its proxy of sex, reproduction is the ultimate goal; but if instead sex / reproduction were to be envisaged not as a goal but as a resource, then it's the only resource that for males can never be satiated and remains in perennial short supply.

3. The evolution of sex and of the male sheds light on the function of dominance hierarchy

The origin and function of dominance as primarily a means of allocating reproduction and especially in the male, would appear to be illuminated by an examination of the function of the male mating type, which entails considering aspects of the evolution of sex.
    The standard explanation of sex, that compared to asexual reproduction it increases genetic variation and consequently adaptability, still appears on the face of it to be very broadly correct once there's taken into account the complexities in natural populations of finite size, spatial structure, and genetic drift; and that sex appears to be favored in some but not all conditions. [30, 31, 32, 33, 34]
Even so, the standard explanation has long been found wanting; hence the continued debate on the subject, which is currently at a complicated impasse. [35] I will briefly discuss aspects salient to the function of the male, which (I will argue, after Atmar [36]) appears to be key to understanding sex.
    Sex would seem to solve two problems. The first and surely most crucial is the problem that besets any system of iterated replication: a steady build-up of replication errors. This is not so much of a problem in species with a relatively simple genome where reproduction is fairly cost free, which is why many simpler species are obligate asexual reproducers. Whole lineages can die, leaving the fitter ones extant. For species with more complex genomes, however, reproduction is costly and at a far lower rate; so there had to evolve a more nuanced means of 'purging' replication error. As will emerge from the discussion, this is at root why sexual reproduction evolved: most primitively, only periodically after several generations of asexual reproduction, but then obligate sexual reproducing species arose.
    The contribution of sex is not mere recombination, which does not prevent lineage extinction. We now know that recombination is present in mitosis as well as in meiosis; so that in diploid asexual populations, deleterious genetic material is unmasked through recombination, reducing complementation of alleles. Recombination also leads to the evolution of self-promoting genetic material and hence intra-genomic conflict. It would seem that loss of complementation and the unmasking of recessive deleterious (especially self-promoting) genes is what meiosis evolved to counter. [37]
    The more comprehensive recombination that meiosis produces, repeatedly 'dilutes' replication errors and renders them less synergetic, thereby reducing the impact of the relentless build-up in the lineage of replication errors. But paradoxically, this is at the expense of tending actually to retain replication errors more than in the absence of iterated recombination, because the reduced assortment of deleterious genetic material correspondingly reduces the proportion of individuals sufficiently non-viable re sexual assortment or survival to be 'purged'. [38] This problem is compounded by the coincidence, found in computer modelling, [39] that a system with built-in robustness to disruption (such as the genome necessarily has against both replication errors and the shuffling of genetic material in sexual recombination), engenders increased interactivity of components. This will work antagonistically to sexual recombination to render deleterious material less synergetic. In the end, then (modelling confirms), 'dilution' doesn’t work; not on its own. An additional mechanism is required to (in effect) concentrate deleterious material in just a subset of individuals in the lineage.
    This perhaps counter-intuitive insight is behind recent re-thinking about in-breeding. [40]
Evidence of any negative impact of in-breeding being hard to find, it was realized that in-breeding actually provides a benefit in increasing homozygosity, whereby recessive genes are much more likely to be unmasked and thereby ‘purged‘.
    A problem would arise if some of those individuals 'purged' were female (the female being for reasons all too apparent the 'limiting factor' in reproduction). The obvious solution is to somehow effectively 'quarantine' deleterious material in the male half of the lineage, from where it can be discarded through the reproductive oblivion of a minority or even a majority of individuals without having a significant impact on the reproductive potential of the lineage.
    This would capitalize on the existence of the two separate mating types of male and female, that arose from the inevitable polarisation of the slightest anisogamy, [41] then driven by intra-genomic conflict, notably in respect of mitochondria. [42, 43, 44] This explains why sex evolved to be not between hermaphrodites (obligately out-breeding): it became dioecious. Sexual reproduction thus would appear to be distinguished from mere meiosis by differentiation of mating type to give a distinct role to the male. And this is not only to highlight and eliminate deleterious genetic material, but complementarily also to highlight and retain genetic material that by mutation and/or recombination somehow enhances the genome. [45] And here the synergy found in computer modelling of systems robust to disruption works not antagonistically but itself synergetically with the thoroughgoing recombination in sex.
    This is the second problem that sex solves: of how to better optimize to the changing environment. The reduction of deleterious genetic material is the other side of the coin of retaining genetic material that is beneficial. Taking these together, it can be seen why it's said that sex is explained by increasing genetic variation. Another take on this is the ‘red queen’ notion of the changing environment requiring the genome to keep pace. All theories of the evolution of sex in essence stem from the genome being out of kilter with the environment; whether through changes in the genome (replication error or recombination) or in the environment.
    The very important selective advantage of a process that capitalizes on mutational and recombinational change is the basis of theories that sex evolved so as to win an ‘arms race’ with parasites. But there's no need to restrict to this one source of selection pressure. If you combine with general optimization in the context of a changing environment, and place together with the still more important need to eliminate deleterious genetic material; then there's clearly more than ample selection pressure to account for the evolution of sex and of the role of the male. Theory had been too myopic. A divisiveness in search of mutually exclusive prediction necessarily is how scientific enquiry proceeds, but all of the theories for the evolution of sex seem to be complementary, being subsumable under this overarching understanding; as echoed by recent papers calling for a pluralistic approach. [46, 47]
    The male acts as a 'filter': metaphorically speaking as a sieve with a mesh of an
appropriate size to catch and so retain genetic material for the furtherance of evolutionary development, as well as to allow to fall through and so reject newly acquired and recurring defects.

The male fulfils his role in several obvious ways. A clear case of 'quarantining' is the Y chromosome, which is a very large part of the genome in simple organisms. But mostly the male's role is accomplished not by any 'quarantining' as such, but by ways that produce more selection in males than in females. There's more rapid and overwhelmingly greater gamete formation in the male than in the female. Many of these male gametes will be non-viable from the outset, and the rest have to negotiate the hostile environment of the female genital tract. A vanishingly small fraction only ever get near the ovum. The sperm will have been severely tested for their metabolic vigor and against other performance criteria. There are mechanisms not just of intra-male but also of inter-male sperm competition (which are a main locus of intra-sexual competition in multi-mating species such as bonobos).
    However, sperm competition places only a fraction of the genome under test, whereas in adult males genetic defect (or enhancement) across a much larger portion of the genome is exposed in being more genetically heterogeneous than are females in sex chromosome heterozygosity (XY): recessive genes that are hidden by female homozygosity (XX) are unmasked.
    For the genes on the autosomes, of course, this cannot be the case. As Atmar realized, to differentially expose this material in males necessitates the behavioral adaptation of male intra-sexual competition, so as to test under conditions of metabolic stress, just as are sperm. Behavioral adaptations of male intra-sexual competition can be not just in the mating process but in anticipation of this throughout the life cycle. Most primitively, male intra-sexual competition is confined just to the courtship antecedent to mating, as in the bee, where the queen simply leads her suitors on a chase, yielding to the male who catches her first.
    Such crude testing of males in the mating process could be broadened to encompass behavior much wider than the mating process proper. Social organisation has become a precursor to mating in sorting males by degree of viability, in both 'positive' and 'negative' senses. Most obviously this manifests as DH (and/or territoriality: this being very closely related to DH). The less viable males are marked out through the 'competition memory' that facilitates DH formation.

If the function of the male indeed is as outlined here, then in all species the combined impact of natural and sexual selection should be significantly greater on the male half of the lineage than on the female. This is obvious in the great bulk of species, and it's becoming clear that this is the case even in seeming anomalous cases; notably bird species where there's an apparent ‘sex reversal’ in males assuming the sole parental role. [48]

The understanding of sex I've outlined here is interestingly consonant with that re mitochondria, [49] which, as previously mentioned, point to an intra-genomic origin of sex in the wake of inevitable anisogamy. Mitochondria are always passed down the female half of the lineage. This is a partitioning away from the ‘filtration’ system of the male side of the lineage, involving large numbers of extra copies of genes so that almost no level of mutation will prevent perfect replication. This 'extreme redundancy' solution to the problem of replication-error accumulation would be an inordinately expensive mechanism if applied to the whole genome, so this illuminates why the male 'genetic filter' function' is needed. This function however does further spur male intra-sexual competition, thus exacerbating mutually unproductive conflict; but DH can ameliorate this as its secondary function. Thus, DH becomes adaptive in an additional manner in respect of dealing with unwanted by-product of what DH has itself produced, thereby making more efficient the function of the male as 'genetic filter' and for the allocation of reproduction.

4. The association with physiological reproductive suppression

In the new conceptualization, the ranking of males in a DH achieves nothing in itself. If DH arose in order to rank and grade males in terms of fitness, then it should be inextricably associated with one or more mechanisms to translate this into corresponding reproductive outcome. The obvious 'add-on' mechanism is to key females into the process by them becoming capable of assessing the overall relative fitness of individual males in their mate choice. This seems near ubiquitous across fauna, and there's abundant scientific evidence, [50] including in humans. [51, 52] However, the simplest, least expensive 'add-on' mechanism is for males to self-suppress their fertility to some degree corresponding to their rank. Males in any case must in effect self-assess their own rank when they calibrate their propensity to engage in future conflict based on the results of past encounters in order to rank in a DH. It's a simple extension to express this also in terms of a similar propensity to reproduce: by differential physiological reproductive suppression.
    Why it would make sense for this to evolve is just as it makes sense for males to accept other than high or at least middling rank in a DH: having even the lowliest rank is strategically valuable compared to being outside the DH – outside the reproductive group. Any propensity to tactically subvert is obviated by the evolution of mechanisms of 'policing' (by which is meant any mechanism that reduces 'cheating', not merely the behavioral – the behavioral morphing into psychology that obviates other behavioral expression) through a mode of selection that acts indirectly in reducing the frequency of short-term 'cheating' thereby to favour adaptations advantageous in the longer term: lineage selection. [53] [Selection acts across timescale as well as level (respectively lineage and multi-level selection [54]) and within the confines of a finite local reproductive group. An overall view has to be simultaneously ‘gene-eyed’ and from the vantage of the whole local reproductive group -- population genetics. The imperative in biology is the maximisation of gene replication over time, and the context of this necessarily is neither the infinitely large gene pool nor the simple and short-term interaction of theory. An abstract snapshot view is of little use. The old 'group selection' controversy has collapsed -- as is now conceded in very recent papers confirming that 'policing' explains eusociality in species where individuals are unrelated (see below).]

Mechanism appears to be both indirect and direct. The lower the rank of an individual, the higher is the sustained level of cortisol -- the second-level stress hormone secreted to deal with sustained competition that diverts resources away from non-immediate priorities, notably reproduction. Primitive species aside, cortisol is generally thought to be the key to a mechanism of reproductive suppression, and it's not in dispute that sustained elevated levels of cortisol have this effect, in all higher animals; and that this is characteristic of lower status males.
    This physiology is common to animals as phylogenetically primitive as fish. [55] In cichlids, once a hierarchy is established, the cortisol levels of the more dominant fish drops precipitously (despite their repeatedly chasing away males and attempts to attract females), whereas the non-dominant males’ cortisol rises to very high levels and stays there. Dr. Sabrina Burmeister [56] has studied the gene mechanisms of this and she confirms that: 'The basic mechanisms that control reproduction in fish and in humans are the same and may be in all vertebrates'.
    Given the useful by-product, as it were, of the stress of competition to indirectly suppress fertility, we would expect the evolutionary process to have enhanced this to make a more direct impact on fertility. It has. Cortisol is intimately connected to sex hormone production through its antagonist, dehydroepiandrosterone (DHEA), and cortisol acts directly (and only in the breeding season, when sexual behavior is so restricted) on the male pituitary gland to suppress Luteinising Hormone secretion and thereby to effect reproductive suppression. [57, 58] Social status and serotonin (the level of which appears to reflect status) also have direct effects. [59, 60]

5. Demonstrating a tight linkage between DH ranking and a gradient of reproductive suppression

Although reproductive suppression is a phenomenon existing across fauna, it's almost exclusively studied in 'cooperative breeder' species, because this is where the phenomenon is so starkly apparent; usually there being just a single breeding pair within the social group of closely related individuals (colony), with the fertility of all others completely suppressed (a 100% reproductive skew). No consideration has been given to the possibility that 'cooperative breeders' are simply at one end of a continuum of direct reproductive suppression existing across the animal kingdom, with at the other extreme species in which all individuals within a reproductive group have some degree of fertility, varying in proportion to rank. Certainly it has not been proposed that a reproductive suppression continuum may be related specifically to DH and to be integral to it, with whatever degree of differential reproductive enhancement/ suppression of individuals according to rank: the alpha allotted a full or possibly even enhanced reproductive role and lower ranks relatively and progressively more suppressed.
    Mechanism put forward re 'cooperative breeders' has been either control by the alpha – a 'dominant control' model -- or by individuals inhibiting their own fertility/mating effort autonomously -- a 'self-restraint' model. 'Dominant control' might seem to make sense for 'cooperative breeders', but this looks less apposite applied across the phylogeny in the continuum model, where the usual case is a reproductive suppression gradient. Here, a single individual at the apex of a DH somehow would have to be able to calibrate the reproductive state of others to varying degrees. A game-theoretic 'commitment model of self-inhibition' [61] has been proposed, which is a 'social contract' between the (supposed) 'sub-dominants' choosing either to commit to complete self-inhibition of reproduction or to do so partially, in exchange for a concession from the (supposed) dominant not to impose other costs of sub-dominance. This would involve assessing compliance and the possibility of 'cheating', necessitating non-breeders or relatively suppressed breeders monitoring the situation to check if things have changed such that it would be more appropriate to commence or increase breeding; and that breeders were doing so only up to an allowed ceiling.
    Such a complicated mechanism would be costly and difficult to achieve, albeit that there are countless instances in evolution of selection operating on simple decision rules leading to sophisticated context-dependent behavior. For the contrasting 'self-restraint' model, all that is required is a simple autonomous neuro-hormonal mechanism within each individual.
    In reality there's no 'tug-of-war' or 'reproductive bribing', as another conceptualization has it: reproductive self-suppression is a mechanism of the reproductive group congruent with individual 'interest' once 'policing' has evolved (this being, as afore-mentioned, the very recently matured thinking about how 'altruism' evolves in the absence of close-relatedness, that has put paid to the 'group selection' controversy.) [62] Reflecting this, there's data showing the opposing models ('dominant control' versus 'self-restraint') not to be mutually exclusive. [63]
    A range of mechanism is likely to be apparent, reflecting both general level of evolutionary development and adaptation through specific ecology; but a phylogenetic trend towards an autonomous automatic mechanism would be expected for its parsimony (from co-opted dominance-submission signalling, through pheromonal signalling, to autonomy). The more gradation -- the closer the correspondence between DH rank and level of reproductive suppression/ enhancement -- the greater would be the efficiency of the mechanism, but this would be subject to practical and cost considerations against the diminishing returns of progressively finer tuning, and the extent to which factors other than DH rank contribute to determining reproductive success (notably male signalling of 'reliability'). Rough correspondence only may be required given female mate-choice having become such a strong mode of translating male status into sex/ reproduction.

Clear instances in nature of a gradient of reproductive suppression corresponding to DH ranking have been found in invertebrate 'co-operative breeders': bee, wasp and ant species. What would seem to be a problem for the new conceptualization here though is that the DH is for females and not for males; and that this is in such evolutionarily 'primitive' animals. However, all-female sociality found in many social insects is a special case generally agreed to be an evolutionary sideline that evolved from normal sexual reproduction. [64]
    These social systems are based on haplodiploidy, whereby males have only a single allele per gene (haploidy) whereas the female has two (diploidy), and consequently deleterious genes are far more exposed to selection in the male, and much more powerfully so than in the XY chromosome system, to better enhance the role of the male as genetic ‘filter’. The upshot is that there's little point in the males having a DH as well; so that beyond the mating process itself there's no male sociality.
    Amongst the females in these species there's a sole breeder, but there are also others waiting in the wings to take over. Although it appears that reproductively it's 'winner-takes-all', rank determines the degree of suppression of ovarian activity compared to the full functioning of the sole breeder. The level of fertility of each individual in the DH is signalled pheromonally from the body surface by cuticular hydrocarbons (CHCs), of which the topmost ranked individual and egg layer has the most and widest range; this profile decreasing as the DH descends from alpha down to delta or epsilon. Below this, other individuals are unranked and sterile and have no CHC signature. [65] This works as ‘honest signalling’ -- that is, it's accurate because it would be 'against the interests' of all concerned for it to be otherwise and instead for 'cheating' to occur. The system on the face of it might seem to be one of 'dominant control', but can be equally or better interpreted as one of self-calibration of fertility on the 'self-restraint' model, given that individuals are merely responding to information about the fertility of others.
    An important point is that the 'honest signalling' here is not owing to the very close-relatedness of the individuals within the reproductive group in these species and the consequent inclusive fitness considerations. This is merely the basis of the extreme skew in reproductive suppression, not the fact of differential reproductive suppression itself. Presumably, without very close-relatedness producing a stark contrast between one or a few breeders and all others being infertile, differential reproductive suppression would be far less apparent, if not empirically invisible.
    Given the contemporary understanding of selection and it's context of the local reproductive group, and the operation of 'lineage selection' to produce 'policing'; then a mechanism whereby an individual calibrates his (or her) own fertility according to dominance rank could evolve in any reproductive group of any species, even if there's low or no close-relatedness. [66, 67]
    The tight linkage between DH and reproductive suppression is illustrated by the fact that whereas, for example, queenless ants [68] establish a DH agonistically and then signal their dominance and consequent reproductive state 'honestly' by pheromones; honey bee workers do not contest physically at all, but by using pheromones that actually mimic those produced by queens to signal to others to suppress their ovarian function. [69] This results in the losers of contests rendering themselves sterile. The two-stage process of establishing rank and subsequent reproductive suppression are here wedded as one.
    This neatly reveals that ranking and reproductive suppression are conjoined in the DH mechanism, and is also interesting with respect to an evolutionary path towards an internal mechanism. In the honey bee, success in agonistic contest has been replaced by outcome of another form of competition to manifest as 'mate value' which combines with signalling inclusive of the reproductive suppression process. Given that all is 'honest signalling', then there's no reason why in turn signalling cannot be completely replaced by an internalized signal, as it were -- a hormone rather than a pheromone.

What is most revealing about the very short DH in these species is that the exact length of the DH can be predicted from knowing the value of just two factors: the number of individuals in the particular colony, and the degree of relatedness of colony members. [70] The lower these values, then the longer is the DH. By extension, it would be expected that in species other than 'cooperative breeders', where individuals are not very closely related (or, alternatively, intensely 'policed'), then the DH would extend to include most or even all members of the reproductive group. Without either close-relatedness or evolved intense 'policing', a 'winner takes all' reproductive system whereby there's a sole breeder would not be adaptive, and instead a graded reproductive suppression tied to rank indeed would be.
    Further illumination is provided by those 'cooperative breeder' species where there's more than one breeding individual in a DH: the alpha plus the beta, or the alpha, beta and the gamma, etc. Here we've a simple clear case of a gradient of reproductive suppression so that there's some correspondence between rank and physiological fertility other than simply the all-or-none situation of the sole breeder and same-sex alloparents.
    This really does reveal a grey area between species that are 'cooperative breeders' and species that are the more usual 'social but unrelated' (as it were), where there's a gradient of physiological reproductive suppression. This indicates that reproductive suppression indeed is likely to be a phenomenon that exists as a continuum across the animal kingdom with a degree of shared mechanism.  Albeit that (some) 'cooperative breeders' may in some details have different mechanisms, the current usual consideration of them as being a distinct sub-group of fauna seems to be false. They seem instead to belong within a parsimonious account of reproductive suppression as a ubiquitous phenomenon (excepting species where there's not DH).

That DH and reproductive suppression are connected is revealed by other lines of evidence. The mediation of this connection by key neurochemicals has been demonstrated. Juvenile Hormone integrates dominance position with reproductive status in primitively social wasps and bumble bees, [71] and higher up the phylogeny, serotonin provides a direct relationship between dominance and reproductive enhancement by the stimulating of release of Luteinizing Hormone-Releasing Hormone (LHRH). [72] This is interesting in that it shows that high-ranking individuals may have not merely a non-suppressed but a boosted potential to reproduce. So conversely, sub-dominance by this mechanism will by hormonal cascade result in low levels of Luteinizing Hormone, which studies have shown is key to physiological reproductive suppression. Serotonin we know is the chemical that mediates DH rank, even in animals more lowly than insects, such as crayfish; and it's perfectly antagonistic to octopamine, [73] which as well as being behind experience-dependent plasticity of aggression -- 'competition memory', if you will -- also triggers the production of Juvenile Hormone in insects.

6. Evidence of DH-related separate same-sex reproductive suppression in humans: the epidemiology of stress

At the other end of the putative continuum of reproductive suppression from 'cooperative breeder' species, is a species with a markedly low reproductive skew: humans. Despite what we would expect consequently to be far less clear evidence of DH-related reproductive suppression, the evidence is very apparent, albeit indirect.
     There's no reason not to suppose that sustained elevated cortisol in humans just as in other animals has a direct impact on fertility apparently by evolved 'design' as well as an indirect impact as a by-product of serious health impairments caused by prolonged elevated cortisol. This indirect impact is more easily seen and measured – it causes a range of serious disease and early death -- than is any curtailment of fertility, and so is a good proxy measure of relative reproductive suppression in any individual.
    The initial chronic effect of cortisol is ‘insulin resistance‘, which leads to the ‘metabolic syndrome’ [74, 75] and the associated insulin resistance, entailing the range of serious disease to various organ systems in middle-age. Epidemiological work on this began in 1967 with the ‘Whitehall’ study [76] of Civil Service men. This revealed that men in the lowest employment grades were much more likely to die prematurely than men in the higher grades. 'Whitehall II' [77] was started in 1985 to find out what underlies this 'social gradient' (as it was dubbed) in death/disease, and to include women.
    A striking 'social gradient' of a constellation of ill-health that Sir Michael Marmot terms the ‘status syndrome’, Marmot concluded must be related to the control that men felt over what they did in the job, what demands were stacked upon them, and what rewards accrued. But in not considering how the men first got into and then failed to escape these lowly jobs, Marmot appears to have mistaken cause and effect. In the key paper [78] Marmot reveals that the self- and independently assessed criteria for 'control' showed no correlation, and his claim that there was no confound with job grade was qualified and left open to question in his pointing out that the dataset was too small to analyse.
    That it's not something inherent in the job that causes ill-health but in the individual who holds it, is what was found to be the case with men in 'top jobs' -- who were once thought to be the ones most at risk of stress-related illness. High status alters physiology so that stress is experienced positively (the mechanism for this is not clear, but presumably is through cortisol being bound to substances that thus prevent it from then binding to receptors). It's adopting this error of former analysis to claim that for men at 'the bottom' it's their job that's the problem.
    The ’social gradient’ applied only to men. For women, it related not to their grade at work but to their situation at home. This fits the model of the sex-separation of DH and the sex-distinct criteria for mate-choice, with only men competing for status. Regarding an hierarchical work environment where status is key, then we would expect a relationship between health and rank only to apply to men; except inasmuch as the criteria for female ranking in their own DH (if indeed human female sociality could be so characterised, rather than a personal network) would be correlated weakly with job status through their male partners, given that female attractiveness assorts with male status. Just such a weak correlation, and only this, was evident. [79] Previously, the best fit that could be obtained for women was by placing them according to the husbands’ occupation, [80] but endorsing other researchers, Marmot [81] writes: 'A measure that reflects general social standing best predicts the social gradient in women'.
    Furthermore, in women, the proxy of health impact overstates the corresponding reproductive suppression, because -- as Valerie Grant pointed out to me (personal communication, 2008) – the fact that in women increased cortisol leads to not decreased testosterone as in men, but to increases, is the very opposite of a stress-based mechanism of reproductive suppression. Men are subject to differential reproductive suppression from even relatively normal levels of cortisol, whereas for women there have to be extreme stress levels for there to be an impact. This fits with the model of sex-separate and an essentially male mechanism of reproductive suppression.

7. DH and reproductive suppression always go together in cooperative breeders whenever there's potential reproductive conflict

If the new conceptualization of dominance here is valid, then the integral nature of reproductive suppression and DH should be evident across a range of species. But it's fruitful only to examine those species where reproductive suppression is so pronounced that it has to be direct rather than possibly merely indirect through stress. These are the vertebrate 'cooperative breeder' species. As I will now outline, the variety of arrangements in these species provides a matrix to illuminate a general rule.
    The most well-known higher animal 'cooperative breeding' species is the wolf. It's becoming clear that contrary to the usual view that wolf groups have an alpha male and alpha female, in fact dominance (usually) plays no role. This is the result of the most comprehensive long-term and up-close research on wolves in the literature, by L David Mech, [82] who concurs with what several other researchers had noted in the 1940s and 1970s. Almost all recent research on wolves has been of captive packs made up of adult animals from various sources, but Mech found that in the wild:

'... young members constitute a temporary portion of most packs, and the only long-term members are the breeding pair ... Because most wolves disperse before two years of age, and almost all before three years of age, there would be no source of sexual competition within most packs'.

The notion of an alpha male and female here is, in Mech's own words, 'particularly misleading', given that 'dominance contests with other wolves are rare, if they exist at all'. Certainly there are no agonistic encounters.
    Mech found 'active submission and food-begging indistinguishable', and indeed that it also acted as 'a food-gathering motivator'; that all individuals tried to retain any food they possessed irrespective of which other wolf challenged it, the supposed 'alpha' included – all wolves freely so challenging any other wolf.  After large prey kills: 'pack members of all ranks (ages) gather around a carcass and feed simultaneously, with no rank privilege apparent'. Indeed, young pups were fed
preferentially by both their parents and older siblings.
    Dominance/ submission signalling in the wolf apparently has been co-opted as signalling to aid affiliative and division-of-labor behavior (just as dominance/submission signalling has been co-opted in courtship; and, conversely, just as in some species pseudo-copulatory behavior is employed to signal dominance/ submission). It's not that wolves don't retain a dominance/ submission behavioral repertoire, because interestingly wolves do have the facility to form a DH in the artificial circumstance of captivity when several unrelated sexually mature individuals of one sex are forced to live together; and it occasionally happens that unrelated animals join a family group in the wild. Generally, however, in the natural state an affiliative family social organisation pertains with dispersal upon sexual maturity, obviating the need for DH.
    That there's (usually) no competition over reproduction in wild wolf packs means that there would be no purpose served by a DH. We also know that there's no direct physiological suppression of reproduction. It's not needed given that all animals other than the breeding pair are not sexually mature. All this fits with the new model.

As with the wolf, the naked mole-rat [83,  84] is a 'cooperative breeder' with a sole breeding pair in a family group, but there's a crucial difference in that animals do not disperse upon attaining sexual maturity, and research shows that there's no 'in-breeding avoidance mechanism' (in most species where this occurs, simply an inability to engage in sex with individuals who share the colony's odour). So there would be nothing to stop intra-sexual competition over reproduction. Consequently, the naked mole-rat does have recourse to DH, and there is physiological reproductive suppression; and in both sexes.
    What happens when only one sex disperses is seen in the dwarf mongoose [85] and the meerkat. [86] Meerkat males must disperse to become a (sole) breeder in a colony, and they stay within their natal colony to bide their time until they become fully mature; in the meantime possibly breeding with extra-colony females. An 'in-breeding avoidance mechanism' precludes mating with same-colony females: if a newborn male is removed and reared outside the colony until after reaching sexual maturity, then the alpha female will not mate with him. [87] There's no evidence that males are reproductively suppressed, and only the female sole-breeder undergoes morphological, physiological and behavioral changes that serve to maintain dominance. [88] Females remain in the natal group, so there must be female intra-sexual competition, and we know that they are subject to physiological (as well as behavioral) reproductive suppression, and that females form a DH. The working of any male DH is not described in the literature (albeit that by usual convention the male breeder is termed a dominant), and males don't fight for dominance, because, it would seem, there would be no point: the breeding female mates only with the male with whom she founded the colony or, if he dies, with a roving male she chooses to be the successor. [89] All this is consistent with the male not requiring a DH, as in the new model.
    The further complication of only partial dispersal is evident in the Damaraland mole-rat, [90] and it's sex differential: twice as many males disperse as do females. That this limited dispersal is not sufficient to preclude competition for breeding in females is shown by female (and female-only) physiological reproductive suppression [91] – together with DH. Males by contrast are subject to caste differentiation, which is known as a means of allocating reproduction in social insects, so this is an alternative mechanism to DH. [92] They are polarized by size, with the larger individuals specialising as aggressive defenders of the colony, and the more likely to disperse.
    Given the scope here for misinterpretation, and with insufficient detail in the published research, it's unclear whether or not there's male DH; but it looks likely that there's not a hierarchy of individuals but instead a three-tiered stratification: sole-breeder/ 'defender' caste/ 'feeder' caste. The picture is made more complex still by what is thought to be an 'in-breeding avoidance' mechanism for males.
    There's even a case of a 'cooperative breeder' where the separate systems of familial organisation and DH co-exist: the alpine marmot. This species forms extended family groups but due to territory take-overs these usually contain unrelated individuals. The adults of both sexes contest only with unrelated same-sex individuals, which unlike individuals belonging to the extended family group are reproductively suppressed. So here, parallel social systems emerge whereby the individuals of one system are physiologically reproductively suppressed, whereas individuals of the other system are not; notwithstanding that all individuals from both systems co-reside. Hacklander et al [93] identified sub-dominants as opposed to extended family members, and were thus able to tease out the parallel systems.

I've here outlined the main factors at play but it's difficult to be sure that all factors involved in a species' breeding system have been identified. Consequently, interpretations of reproductive behavior systems in the discussion sections of studies are liable to be confused or false. In particular, the recent questioning of a stress-based mechanism of reproductive suppression in some species is itself questionable through the assumption that a DH operates when in fact there isn't one for that species (or it is only partial through there being just a short DH, or two-tier through the phenomena of 'super-dominance'); and then there is the mechanism that reduces the effect of cortisol on high-rankers. All this further helps to explain why a likely generic link between DH and physiological reproductive suppression had not been spotted. Nevertheless, as I've just now demonstrated, reproductive systems of various species can be analysed in terms of a number of dimensions and (with caveats) shown to express and abide by the rule than DH and differential physiological reproductive suppression indeed do go hand in hand: the data is consistent -- at least not inconsistent -- with the new model.

8. A new conceptualization of dominance: mutually exclusive predictions

The function of dominance (hierarchy) is to 'deliberately' create intra-sexual conflict so as to allocate ranks amongst the individual males within a reproductive group, that translate into corresponding differential reproductive auto-suppression of fertility. This mechanism would seem to be a major part of the overall function of the male to filter out deleterious genetic material from the local gene pool (and, conversely, to retain beneficial new mutation and combination); serving to maximise gene replication over time in the context of the reproductive group. Although this would seem to be an adaptation re males, it can also apply separately to the females of the same reproductive group, albeit usually more weakly; or even exclusively to females in species where there is no male sociality.
    By contrast, the current understanding is that dominance (hierarchy) serves merely to ameliorate conflict, and thereby to minimise mutually unproductive competition, and/or to more efficiently allocate whatever resource seems to be at issue. That dominance (hierarchy) functions in this way is not a point that distinguishes between the old and new models, except that in the new conceptualisation this is secondary, not primary: dominance (hierarchy) once in play is bound to also thus function.
    The new conceptualization of dominance outlined here is testable given that it generates predictions mutually exclusive of the consensus model (of minimizing mutually unproductive contest and in effect to allocate resources generically). These are that dominance/ DH is:
*    same-sex only
*    present whenever, within one or both sexes, there's potential conflict over reproduction, and there's no mechanism to preclude this; but otherwise is absent
*    always associated with some degree of differential physiological reproductive suppression
Evidence has been presented in this paper to establish all of the above.
    The new understanding of a goaded battling at the level of same-sex individuals is in the service of what at other levels of analysis is a 'collective effort' within each sex and within the reproductive group as a whole to express the biological imperative of maximising gene replication over time. It's a profound if not quite beautiful paradox that (same-sex) individuals proximally are driven to compete so that distally there's more perfect co-operation.
    This proper understanding of dominance belies the usual general assumption that control of resource-competition is the basis of sociality (more complex than simple herding in mutual defence against predators). The reproductive system of a species now can be seen as even more centre-stage than hitherto appreciated, with resources regarded as possible factors only, making for species-specific further complication of social organisation more fundamentally arising from reproductive considerations. There will need to be a complete change in understanding of social systems in terms of how we conceive of 'power'.

1 Fedigan LM (1992) Dominance and Alliance: Chapter 7 of Primate Paradigms: Sex Roles and Social Bonds University of Chicago Press.
2 Lanctot RB & Best LB (2000) Comparison of methods for determining dominance rank in male and female prairie voles. Journal of Mammalogy v81n3 pp734-745.
3 Issa F & Edwards D (2006) Ritualized Submission and the Reduction of Aggression in an Invertebrate. Current Biology 16 pp2217-2221.
4 Smith PK (1988) The Cognitive Demands of Children’s Social Interactions With Peers. Pp94-110 in Byrne RW & Whiten A (eds) Machiavellian Intelligence: Social Expertise and the Evolution of Intellect in Monkeys Apes and Humans. Oxford University Press.
5 Cummins DD (2000) How the social environment shaped the evolution of mind. Synthese 122 pp3-28.
6 Dugatkin LE & Earley RL (2004) Individual recognition, dominance hierarchies and winner and loser effects. Proceedings of the Royal Society of London, biological sciences 271 pp1537-1540.
7 Eg; Hemelrijk C (1999) An individual-oriented model of the emergence of despotic and egalitarian societies. Procedures of the Royal Society, Biological Sciences 266 pp361-369.
8 Grant VJ (2005) Dominance, testosterone and psychological sex differences. Pp1-28 in Psychology of Gender Identity ed Janice W Lee. Nova Science Publications, New York.
9 Dunbar RIM (1988) Primate Social Systems. Ithaca, New York: Cornell University Press
10 Villanelli F & Gherardi F (1998) Breeding in the crayfish, Austropotamobius pallipes: Mating patterns, mate choice and inter-male competition Freshwater Biology 40 pp305-315.
11 Peeke HSV, Sippel J & Figler MH (1995) Prior residence effects in shelter defense in adult signal crayfish: Results in same- and mixed-sex dyad. Crustaceana 68 pp873-881.
12 Peeke HSV, Figler MH & Chang ES (1998) Sex differences and prior residence effects in shelter competition in juvenile lobsters. Journal of Experimental Maritime Biological Ecology 229 pp149-156.
13 Gneezy U, Niederle M & Rustichini A (2003) Performance in competitive environments: gender differences. Quarterly Journal of Economics pp1049-1074.
14 Gneezy U & Rustichini A (2004) Gender and competition at a young age. American Economic Review Papers and proceedings pp377-381.
15 Gregory SW & Gallagher KJ (2002) Spectral analysis of candidates' nonverbal vocal communication: Predicting US presidential outcomes. Social Psychology Quarterly v65n3 pp298-308.
16 Radespiel U & Zimmerman E (2001) Female dominance in captive gray mouse lemurs (Microcebus murinus). American Journal of Primatology v54n4 pp181-92.
17 Eg; Keith-Lucas T, White FJ, Keith-Lucas L, Vick LG (1999) Changes in behavior in free-ranging Lemur catta following release in a natural habitat. American Journal of Primatology v47n1 pp15-28.
18 BBC (2002) Wildlife On One February 2002.
19 Kappeler PM (1993) Female dominance in primates and other mammals in Perspectives in Ethology, Behavior and Evolution (ed Bateson PPG, Klopfer PH, & Thompson NS) v10 pp143-158. New York, Plenum.
20 White FJ, Overdorff DJ, Keith-Lucas T, Michele A. Rasmussen MA, W. Eddie Kallam WE & Forward Z (2006) Female Dominance and Feeding Priority in a Prosimian Primate: Experimental Manipulation of Feeding Competition. American Journal of Primatology 69 pp295-304.
21 Pochron ST, Fitzgerald J, Gilbert CC, Lawrence D, Grgas M, Rakotonirina G, Ratsimbazafy R, Rakotosoa R & Wright PC (2003) Patterns of female dominance in Propithecus diadema edwardsi of Ranomafana national park, Madagascar. American Journal of Primatology v61n4 pp173-185.
22 Anzenberger G (1992) Monogamous social systems and paternity in primates. Pp203-244 in Paternity in Primates: Genetic Tests and Theories (ed Karger).
23 Curtis DJ (1997) The Mongoose Lemur (Eulemur mongoz): A Study in Behavior and Ecology. Unpublished PhD thesis, University of Zürich.
24 Young AL, Richard AF & Aiello LC (1990) Female dominance and maternal investment in strepsirhine primates. The American Naturalist 135 pp473-488.
25 Eg; Robbie A McDonald (2002) Resource partitioning among British and Irish mustelids. Journal of Animal Ecology 71 pp185-200.
26 Stowers L, Holy TE, Meister M, Dulac C & Koentges G (2002) Loss of sex discrimination and male-male aggression in mice deficient for trp2. Science 295 pp1493-1500.
27 Leypold BG, Yu CR, Leinders-Zufall T, Kim MM, Zufall F & Axel R (2002) Altered sexual and social behaviors in trp2 mutant mice. PNAS v99n9 pp6376-638.
28 Kimchi T, Xu J & Dulac C (2007) A functional circuit underlying male sexual behavior in the female mouse brain. Nature advance online publication
29 Eg; Kent, Susan (1993) Sharing in an Egalitarian Kalahari Community Man, New Series v28n3  pp479-514. Royal Anthropological Institute of Great Britain & Ireland.
30 Agrawal AF (2006) Evolution of sex: Why do organisms shuffle their genotypes? Current Biology v16n17 pp696-704.
31 De Visser JA & Elena SF (2007) The evolution of sex: Empirical insights into the role of epistasis and drift. Nature Reviews Genetics 8 pp139-149.
32 Otto SP & Gerstein AC (2006) Why have sex? The population genetics of sex and recombination. Biochemical Society Transactions v34n4 pp512-522.
33 Misevic D, Ofria C & Lenski RE (2005) Sexual reproduction reshapes the genetic architecture of digital organisms.  Proceedings of the Royal Society: Biological Sciences Published online.
34 Jaffe K (2002) On the adaptive value of sex. Proceedings of the fourth international conference on complex systems. New England Complex Systems Institute. On-line publication.
35 Otto SP & Lenormand T (2002) Resolving the paradox of sex and recombination. National Review of Genetics v3n4 pp252-261.
36 Atmar W (1991) On the role of males. Animal behavior 41 pp195-205.
37 Archetti M (2003) A selfish origin for recombination. Journal of Theoretical Biology 223 pp335–346.
38 Paland S & Lynch M (2006) Transitions to asexuality result in excessive amino acid substitutions. Science 311 p5763.
39 Azevedo RBR, Lohaus R, Srinivasan S, Dang KK & Burch CL (2006) Sexual reproduction selects for robustness and negative epistasis in artificial gene networks. Nature 440 pp87-90.
40 Ekblom R (2000) Inbreeding avoidance through mate choice.
41 Parker GA, Baker RR & Smith VGF (1972) The origin and evolution of gamete dimorphism and the male-female phenomenon. Journal of Theoretical Biology 36 pp529-553.
42 Cosmides LM & Tooby J (1981) Cytoplasmic inheritance and intra-genomic conflict. Journal of Theoretical Biology 89 pp83-129.
43 Lane, Nick (2005) Power, Sex Suicide: Mitochondria and the Meaning of Life. Oxford University Press.
44 Ridley, Matt (1994) The Red Queen: Sex and the Evolution of Human Nature. Penguin.
45 Atmar W (2007) Email correspondence loop/ personal communication.
46 West SA, Lively CM & Read AF (1999) A pluralist approach to sex and recombination. Journal of Evolutionary Biology 12 pp1003-1012.
47 Birky Jr CW (1999) An even broader perspective on sex and recombination. Journal of Evolutionary Biology v12n6 pp1013-1016.
48 Reeve HK & Pfennig DW (2003) Genetic biases for showy males: Are some genetic systems especially conducive to sexual selection? PNAS
49 Lane N (2005) Power, Sex, Suicide -- Mitochondria and the Meaning of Life. Oxford University Press.
50 Eg: Klinkova E, Hodges JK, Fuhrmann K, de Jong T & Heistermann M (2005) Male dominance rank, female mate choice and male mating and reproductive success in captive chimpanzees. International Journal of Primatology v26n2 pp357-484.
51 Buss DM (2003) The Evolution of Desire. Basic Books, New York
52 Okami P & Shackelford TK (2001) Human sex differences in sexual psychology and behavior. Annual Review of Sex Research 12 pp186-241.
53 Nunney L (1999) Lineage selection: natural selection for long-term benefit. Pp238-252 in Keller L (ed) Levels of selection in evolution (1999) Princeton University Press.
54 Keller L (ed) (1999) Levels of Selection in Evolution. Princeton University Press.
55 Fox HE, White SA, Kao MHF, Fernald RD (1997) Stress and dominance in a social fish. Journal of Neuroscience v17n16 pp6463–6469.
56 Burmeister SS, Jarvis, ED & Fernald, RD (2005) Rapid behavioral and genomic responses to social opportunity. PloS Biology v3n11 e363.
57 Stackpole CA, Clarke IJ, Breen KM, Turner AI, Karsch FJ & Tilbrook AJ (2006) Sex Difference in the Suppressive Effect of Cortisol on Pulsatile Secretion of Luteinizing Hormone in Sheep. Endocrinology v147n12 pp5921-5931.
58 Dubey AK & Plant TM (1985) A suppression of gonadotropin secretion by cortisol in castrated male rhesus monkeys (Macaca mulatta) mediated by the interruption of hypothalamic gonadotropin-releasing hormone release. Biology of Reproduction v33 pp423-421.
59 Eg; Abbott DH, Hodges JK & George LM (1988) Social status controls LH secretion and ovulation in female marmoset monkeys. Journal of Endocrinology 117 pp329–339.
60 Hery M, Francois-Bellan AM, Hery F, Deprez P, Becquet D (1997) Serotonin directly stimulates luteinizing hormone-releasing hormone release from GT1 cells via 5-HT7 receptors. Endocrine v7n2 pp261-265.
61 Hamilton IM (2004) A commitment model of reproductive inhibition in cooperatively breeding groups. Behavioral Ecology v15n4 pp585-591.
62 Wilson DS & Wilson EO (2007) Rethinking the Theoretical Foundation of Sociobiology. Quarterly Review of Biology 82 pp327-348.
63 Eg; Langer P, Hogendoorn K, Schwarz MP & Keller L (2006) Reproductive skew in the Australian allodapine bee Exoneura robusta. Animal Behavior v71n1 pp193-201.
64 Normark BB (2003) The evolution of alternative genetic systems in insects. Annual Review of Entomology v48 pp397-423.
65 Monnin T (2006) Chemical recognition of reproductive status in social insects. Annales Zoologici Fennici 43 pp515-530.
66 Brandvain Y & Wade MJ (2007) The evolution of competition and policing: opposing selection within and among groups. BMC Evolutionary Biology v7n203. Published on-line.
67 Ratnieks FLW & Wenseleers T (2007) Altruism in insect societies and beyond: voluntary or enforced? Trends in Ecology and Evolution v23n1 pp45-52.
68 Cuvillier-Hot V, Lenoir A, Crewe R, Malosse C & Peeters C (2004) Fertility signalling and reproductive skew in queenless ants. Animal Behavior v68n5 pp1209-1219.
69 Moritz RFA, Lattorff HMG & Crewe RM (2004) Honeybee workers (Apis mellifera capensis) compete for producing queen-like pheromone signals. Proceedings of the Royal Society, Biological Sciences 271 pp98-100.
70 Monnin T, Ratnieks FLW & Brandão CRF (2004) Reproductive conflict in animal societies: hierarchy length increases with colony size in queenless ponerine ants. Biomedical and Life Sciences v54n1 pp1-79.
71 Hartfelder K (2000) Insect juvenile hormone: from 'status quo' to high society. Brazilian Journal of Medical Biological Research v33n2 pp157-177.
72 Hery M, Francois-Bellan AM, Hery F, Deprez P, Becquet D (1997) Serotonin directly stimulates luteinizing hormone-releasing hormone release from GT1 cells via 5-HT7 receptors. Endocrine v7n2 pp261-265.
73 Stevenson PA, Dyakonova V, Rillich J & Schildberger K (2005) Octopamine and Experience-Dependent Modulation of Aggression in Crickets. The Journal of Neuroscience v25n6 pp1431-1441.
74 Sapolsky RM (2004) Social status and health in humans and other animals. Annual Review of Anthropology v33 pp393-418.
75 Sapolsky, RM (2005) The influence of social hierarchy on health. Science 308 pp648-652.
76 The Whitehall study summarized at There are several papers.
77 Marmot MG, Smith GD, Stansfeld S, Patel C, North F, Head J, White I, Brunner E, Feeney A (1991) Health inequalities among British civil servants: the Whitehall II study. The Lancet 337 pp1397-1393.
78 Bosma M, Marmot M, Hemingway H, Nicholson A, Brunner E & Stansfield S (1997) Low job control and risk of coronary heart disease in the Whitehall II study. British Medical Journal 314 pp558-564.
79 Sacker A, Firth D, Fitzpatrick R, Lynch K & Bartley M (2000) Comparing health inequality in men and women: prospective study of mortality 1986-96. British Medical Journal  320 pp1303-1307.
80 Moser K, Pugh H & Goldblatt P (1988) Inequalities in women's health in England and Wales: mortality among married women according to social circumstances, employment characteristics and life cycle stage. Subsequently published in Genus vXLVI pp71-84.
81 Marmot M (2004) The status syndrome: how social standing affects our health and longevity. Times Books.
82 Mech LD (1999) Alpha Status, Dominance, and Division of Labor in Wolf Packs. Canadian Journal of Zoology 77 pp1196-1203.
83 Faulkes CG & Bennet NC (2001) Family values: group dynamics and social control of reproduction in African mole-rats. Trends in ecology and evolution v16n4 pp184-190.
84 Sherman PW, Jarvis JUM & Alexander RD (1991) The Biology of the Naked Mole-Rat. Monographs in Behavior and Ecology Princeton University Press.
85 Davidson College animal behavior website, mongoose pages.
86 Griffin AS, Pemberton JM, Brotherton PNM, McIlrath G, Gaynor D, Kansky R, O’Riain J & Clutton-Brock TH (2003) A genetic analysis of breeding success in the cooperative meerkat (Suricata suricatta). Behavioral Ecology v14n4 pp472-480.
87 BBC TV/ Kalahari Meerkat Project, University of Cambridge (2005) Meerkat Manor
88 Clutton-Brock TH, Hodge SJ, Spong G, Russell AF, Jordan NR, Bennett NC, Sharpe LL & Manser MB (2006) Intrasexual competition and sexual selection in cooperative mammals. Nature 444 pp1065-1069.
89 Kalahari Meerkat Project, University of Cambridge (2007) FAQ: Meerkat biology and behavior.
90 Burland TM, Bennett NC, Jarvis JUM &. Faulkes CG (2004) Colony structure and parentage in wild colonies of cooperatively breeding Damaraland mole-rats suggest incest avoidance alone may not maintain reproductive skew. Molecular Ecology v13n8 p2371.
91 Molteno AJ & Bennett NC (2000) Anovulation in non-reproductive female Damaraland mole-rats. Journal of Reproduction and Fertility 119 pp35-41
92 Bennett NC & Jarvis JUM (1998) The Social Structure and Reproductive Biology of Colonies of the Mole-Rat, Cryptomys damarensis. Journal of Mammalogy v69n2 pp293-302.
93 Hacklander K, Mostl E & Arnold W (2003) Reproductive suppression in female alpine marmots. Animal Behavior v65n6 pp133-140.