Demographic Transition as Caused by Biological Effects of Social Dislocation:

Adaptive Reproductive-Suppression Triggered by ‘Crowding’ Stress of Males Transmitted Epigenetically to Female Offspring Multi-Generationally; Plus Out-Breeding Fertility Depression through Genetic Incompatibilities

Steve Moxon, 2014 stevemoxon3@talktalk.net

[Speculative paper intended and accepted for science conference presentation (unable to attend)]

 ABSTRACT

An explanation of demographic transition [DT] as an adaptive mechanism is thought unlikely or untenable, but is strongly suggested by the finding that close proximity of non-familiar conspecifics causing ‘crowding’ stress of males is epigenetically transmitted specifically to female offspring, triggering multi-generational fertility decline and aberrant reproduction-related behaviors. This has no apparent function other than plausibly population self-regulation (to which there is no theoretical objection given advances in modelling ‘mutualism’). Out-breeding depression would be a complementary additional biological root.

A biological aetiology (adaptive or otherwise) is not unlikely given causality testing in latest analyses of economic models suggests an as yet unidentified underlying driver. Similar is conceded regarding the ‘kin influence’ anthropological hypothesis. That this possibility is overlooked is inherent in all theorising placing resources centre-stage, with offspring falsely considered (‘inferior’) ‘goods’ rather than ends-in-themselves.

The need for a new theoretical approach to DT is apparent from the now confirmed findings that K-selection (for ‘quality’) does not outdo r-selection (for ‘quantity’) of offspring across generations, challenging DT modelling in terms of ‘Life History Theory’, which anyway generates predictions opposite to DT (bar reduced child mortality prompting ‘slow’ life history traits, but this is not supported by data).

With both reproductive-suppression and out-breeding depression driven by social dislocation, then migration would be a factor; notably the predominant rural-to-rural mode in under-developed countries – accounting for DT where there is no economic development. Though the impact of absent husbands and other confounds requires better data and analysis for the theory to be fully testable (falsifiable), there is the prospect that hitherto ever more complex DT modelling may become much more parsimonious.

KEYWORDS: demographic transition, adaptive, biological, epigenetic, reproductive suppression, ‘crowding’ stress, outbreeding depression

 

The seeming profoundly non-adaptive ‘demographic transition’ (DT) of falling fertility (whether to below or remaining above replacement level) has been recognised for the past century, and though originally occurring in developed nations (Europe) demographers generally agree that it is now globally ubiquitous (eg, Lee & Reher, 2011). That is, DT is understood to be apparent to some degree (if sometimes ‘stalling’) in all under-/un-developed nations, leaving ‘natural fertility’ remaining in extant forager (hunter-gatherer) communities. The heterogeneity of conditions in which DT takes place indicates the possibility of a more profound underlying basis than had been supposed. The literature on its aetiology is still, as ever, regularly complained of as being as messily contradictory as it is large (eg., Shenk et al, 2013). It was recognised 25 years ago that, after even then already over half a century of work, DT modelling was “notably lacking in such components of theories as a specifiable and measurable mechanism of ‘causation’” (Teitelbaum 1987, p421). A notable divide in the literature is between economics/ demography and (evolutionary) anthropology, with challenges to the essentially economic model of DT (as the interplay between falling child mortality and economic development) in a relatively new focus on social interaction. For example, and influentially, Bongaarts & Watkins (1996) argued that given a relatively small degree of development (and/or, they suspect, child mortality decline) then social forces alone could account for subsequent fertility declines. However, this is a modification only of DT theory as formulated by economists in that there is no replacement initial cause. Nevertheless, the now global reach of DT to encompass extremes of rural/urban and poor/rich puts further strain on economic modelling through the challenge this poses to economic growth as the putative key driver, opening up the possibility that the root causation may even be as yet entirely missed.

Adaptive evolutionary explanations of DT have long been given a poor outlook (Barkow & Burley, 1980), and the prospects would seem to have considerably worsened with DT being shown to have no adaptive value in terms of a quality versus quantity trade-off in offspring. K-selection (for quality of offspring) over r-selection (for quantity of offspring) produces no effect, or even a negative effect, on the total numbers of grandchildren and great-grandchildren (Goodman, Koupil & Lawson, 2012), confirming previous findings. This contradicts the notion that DT fertility falls can be explained (at least in the long-term) as restricting reproduction to a smaller number of offspring in some circumstances, in order that the second-generation can have more and/or better quality offspring of their own; thus conferring a pay-off in total number and/or quality of grand-offspring of the first generation. The reasoning is encapsulated within ‘Life History Theory’ (LHT), which is viewed by some evolutionary demographers as a promising theoretical basis of DT (Mathews & Sear, 2008). Yet LHT, as most cogently formulated by Ellis et al (2009), predicts ‘faster’ life history traits (such as earlier onset of puberty, younger age of first sexual intercourse and first birth, and reduced parental investment) when the environment is ‘harsh’ and ‘unpredictable’. This is hardly a putative basis of DT, being the opposite of the change in conditions typically occurring in nations undergoing DT. The occurrence of DT appears to be in circumstances where the environment has become more ‘harsh’ and ‘unpredictable’, not less; as in both under-developed nations recently and across Europe historically – when the profound socio-economic stability of extended-family based small-scale community gives way to internal migration and insecure wage labour in relatively anonymous urban concentrations or industrialised agriculture. LHT here would predict not the ‘slower’ traits as found in DT, but their ‘faster’ alternatives. This would not be the case if the child mortality decline observed in DT itself were to be considered a key environmental indicator, but Mathews & Sear (2008) in a rare test of a link between mortality perceptions and fertility preferences find that at an individual level mortality priming has no impact on females and only a slight one on males (and none at all re the costs/benefits of children; only at the 10% level of statistical significance re desired family size). That in any case K-selection is found not to trump r-selection in overall reproductive output down the generations leaves LHT an unlikely candidate to be a root cause of DT. ‘Slow’ versus ‘fast’ life history traits in response to environmental fluctuation would appear to be adaptive in the short-term, but not – judging by the occurrence of DT — in the long-term.

To try to square DT theory with evolutionary principles, alternatively it has been envisaged to occur as the result of traits that were at one time adaptive but which become mismatched with the environment; so that individuals attempt to maximise acquisition of resources at the expense of a reduction in fertility to levels much lower than would maximise fitness even over the medium-term (Borgerhoff Mulder 1998). This ‘adaptive lag’ presumably does not last into the long-term, because self-evidently selection pressures eventually would force a readjustment. However, Pritchett & Viarengo (2012) raise the issue that it would have to explain in the medium- and even short-term not merely lower fertility, but zero fertility: why a substantial minority (20% or even 30%) of women in some developed countries stay childless. No theory, evolutionary or economic,Pritchett & Viarengo claim, hitherto has accounted for this phenomenon.

Attempts at an evolutionary explanation nevertheless have continued, though (efforts to utilise LHT aside) they have become confined mainly to examining kin assistance (Turke, 1989) and more recently and importantly kin influence (Newson et al, 2005, 2007); both predicated on the ‘inclusive fitness’ principle in respect of genetic relatives. Kin influence is less an evolutionary than a sociological hypothesis in that it concerns the transmission of social norms of high-fertility amongst female kin. Its application to DT is in the relative absence of such norms (and/or, conversely, the transmission of low-fertility norms amongst non-kin) in the wake of the breakdown of the ‘traditional’ social context of the predominance of kin, where high-fertility social norms would have prevailed. Albeit that there are studies showing a seeming direct impact of kin presence on female fertility rates (eg., Tymicki, 2004), these are all correlational findings, which do not demonstrate a causal relationship. Newson et al concede in their 2005 paper an absence of research actually showing either that kin transmit more pro-natal messages than do non-kin or that this then produces changes in attitudes to family creation. Their 2007 study was an attempt to remedy this by a role-play experiment featuring hypothetical scenarios of women considering whether or not to try to conceive, with two supposedly an “easy” choice (reproductively propitious) but two others “difficult” (reproductively non-propitious), where subjects were asked to imagine themselves as either a friend (non-kin) or mother (kin) offering advice. Clearly lacking ecological validity, the study in any case found only weak kin influence; and the research is problematic in the extremely biased and self-selecting sampling, and the choosing of experimental conditions seemingly to suit the bias of the sample obtained.

Recent literature review of kin effects as a possible root of DT (Sear & Coall, 2011) reveals merely a “tentative” link to fertility, and even then this is of mixed direction – fertility may be either decreased or increased in any given set of circumstances, with different kin relationships often being opposed in their impact. In the most recent work on kin influence, incorporating a literature review re both kin assistance and kin influence in a new analysis of data, Mathews & Sear (2013) look at an individual (rather than a population) level, and find only partial support for kin influence (or kin assistance). They are unable to decide between kin influence and kin assistance, and not only find a very weak impact on fertility (merely claiming statistical significance, not any effect size), but also no particular close kin relationship predominant in terms of influence/assistance. This is contrary to what would be expected from the ‘inclusive fitness’ principle given the very different degrees of genetic relatedness corresponding to degree of kinship, and contrary to the authors’ previous findings of the predominance in kin influence by affinal grand-parents.

The problem here for the notion of kin influence is that if even affinal grand-mothers do not stand out in the data, and the data re the impact of kin on fertility in general is very weak, then the kin influence hypothesis hardly would appear to be a likely candidate prime driver of DT. Its key difficulty, though, remains that data is merely correlational, allowing the distinct possibility that any finding of supposed kin influence with respect to DT is artefactual, leaving the question of causation an open one. Along with any putative social interaction factor of DT, initial cause is anyway ceded to development – or, rather, the combination of falling child mortality and development – as driving changes in social interaction. It is hard to see, therefore, how (evolutionary) anthropologists can make a decisive contribution to solving the impasse of DT root causality. Indeed, Mathews & Sear (and Sear & Coall) themselves point to the potential for an alternative causation of some factor common to kin cohesion and greater fertility. They miss the obvious inversion, though: that fertility may be depressed not by a relative absence of kin influence and/or support but by the presence of some factor concerning social dislocation, of which the relative absence of kin is an indirect measure.

Causality testing using new analytic techniques is a very recent new direction in modelling DT in the economics literature, which is where the bulk of work on DT has been done; and therefore this is here now very briefly considered, to look at the context of the difficulties that have prompted the new focus.

To almost any claim made in DT literature there is serious question. This applies even to the principal putative causal factor of child mortality decline, in that not only did it long pre-date DT (eg., Malmberg, 2008), but according to some researchers was of too small a magnitude even potentially to explain it (Palloni & Rafalimanana, 1999). As Heuveline (Heuveline, 2004, p2) points out: “Sporadic attempts to precisely document mortality influences on fertility have generated more doubts than evidence about any such direct relationships“. As for the putative principal driver of economic growth, Heligman finds that in various nations the decline in fertility “can’t be explained by education or economic development” (Heligman, 2002) — and, indeed, education has been ruled out as surplus to explanation in a most recent comparative test across hypotheses (see below) (Herzer, Strulik & Vollmer, 2012), completely contradicting what has become an article of faith amongst anthropologists, that global population falls depend crucially on the education specifically of women – apparently an ideological imposition on science. It is well established that contraceptive technology cannot be a major factor given effective already long established traditional methods of controlling conception (eg., Dribe & Scalone, 2010; Santow, 1995). A persisting problem of DT theory is that it provides “a weak explanation at best” in respect of Asia, Africa and Latin America (the regions more recently to experience DT after Europe and the Anglophone nations) (Newbold, 2007, p11).

Progress has been made in integrating the differing emphases of demography and economics – respectively, falling child mortality and economic growth – in ‘unified growth theory’ (eg., Cervellati & Sunde, 2013). This development has borne fruit in at last producing an account for the long delay between the declines in child mortality and fertility (Strulik & Weisdorf, 2010) — fertility initially and for some time increasing before eventually falling – through additionally taking into consideration the fact that parents are concerned with the nutrition of offspring and not merely offspring survival. Yet it is something of a revelation of the paucity of modelling of DT that hitherto it could have been considered otherwise. It points up the problem of modelling being (essentially) economic, with DT theory subsumed within economics and distant from other disciplines. There is a major conceptual problem in economics that offspring are regarded not only as merely economic ‘goods’, but actually ‘inferior goods’ (goods for which demand declines with income); this necessarily in order to explain why socio-economic improvement apparently is linked with declining fertility (Pritchett & Viarengo, 2012). This is entirely at odds with a biological perspective of reproduction being an (indeed, the ultimate) end in itself. Self-evidently, parents do not regard their offspring as goods, but as the embodiment of their lives and future, in respect of which the acquisition of goods is instrumental. Consequently, changing costs of rearing children, the opportunity costs of having children, and the value of a child as labour, would not be expected necessarily or at all to translate into corresponding changes in fertility, as economists assume. Economics, being concerned essentially with trading, is predicated on the assumption that humans are invariably motivated to maximise acquisition of resources in all contexts and circumstances. Yet motivation in all species (with humans no exception) primarily (distally if not proximally) is reproduction, with all other motivation necessarily subsidiary to this, and the acquisition of resources being merely means to reproductive outcomes, both short- and long-term — from gaining access to sex, through to children and grand-children. This fundamental failure of understanding is compounded by and expressed in non-, indeed anti-scientific ideological constructs in various guises of Marxism / neo-Marxism, notably in ‘identity politics’ whereby the sexes are misconceived of as competing ‘power’ interests contesting over resources. Explanations of DT to date are not unlikely, therefore, to be rationalisations of an underlying biological cause, to which the major factors identified in models are but indirectly related as proximal rather than ultimate causes.

The problem of identifying root causation is not least through the number and inextricability of causes: the simultaneity (as the term is understood in econometrics) of multiple factors (Guinnane, 2011). As well as difficulty identifying and distinguishing between them and ascertaining which are the more important, even establishing causal direction has been elusive (eg., O’Sullivan, 2013). What are considered causal factors may be better understood as themselves products of DT or in a reciprocal relationship with it (eg., Canning, 2011). To tacklethe impasse in respect of causation, economists / demographers most recently have employed sophisticated comparative methods to examine across competing hypotheses. Shenk et al (2013, 2011) use a methodology derived from ‘likelihood theory’ to respectively weight models according to the data, so as to decide between models notwithstanding predictions that are not mutually exclusive. They consider all of the principal putative drivers of DT: (1) child mortality decline, (2) the economic costs of raising children and the benefits of investing in them, and (3) the transmission of low-fertility social norms (as in the kin influence hypothesis). Albeit finding most support for models emphasising economic factors, Shenk et al uncover multiple causal pathways: they fail to identify a root aetiology. An alternative very recent modelling — using panel co-integration estimators and Granger causality testing – likewise revealed only a circular causality: Herzer et al show that against the backdrop of mortality decline and economic growth, the fertility changes in DT are both cause and consequence. In footnote number one in the paper comes the suggestion of an as yet hidden underpinning to this circular causality: “Nevertheless it could be that Granger causality (testing) fails to identify true causality. It could be that the co-integrated variables are driven by another neglected process” (Herzer, Strulik & Vollmer, 2012, p367). What demographers and economists have identified may well be inter-related second-order processes that are a consequence of an underlying primary cause hitherto entirely missed. This is a similar impasse in respect of causality facing anthropologists with the ‘kin influence’ hypothesis (see above).

This reprises the prospect of a biological aetiology, which anyway looks likely from the afore-mentioned particularly poor fit of DT theory in respect of fertility declines in many undeveloped nations. Given fertility falls evident with very few exceptions everywhere across the world, then any DT model has to be capable of being applied throughout the continuum from extremes of urban-rich to rural-poor – to include places where there is little if any economic growth.

Returning, then, to trying to understand DT from an evolutionary perspective, and even as an adaptive phenomenon: there is a remaining possibility – that is, it is not inconceivable — that rather than reduced fertility being an inadvertent by-product, it may be that curtailment of reproduction per se (temporary complete cessation by some, many or all females) is adaptive in circumstances where there is a major threat to ecological viability, to the extent that local extinction otherwise would be the likely prospect. A mechanism which avoided this eventuality – even which merely tended to do so, and therefore being apparent only statistically — would be selected. And note that there is no theoretical objection to this proposition.

This warrants a brief diversion into evolutionary theory. For curtailment of reproduction per se to be adaptive there is no need to invoke ‘multi-level’ (‘group’) selection; either naïve or very recent re-formulations as most prominently by Novak et al (Novak, Tarnita & Wilson, 2010), although this is just what is still argued in some quarters specifically regarding reproductive restraint (Werfel & Bar-Yam, 2004). ‘Group’ selection, now understood in terms of two competing versions of ‘multi-level’ selection, is still generally considered to be applicable only in highly restricted scenarios (Okasha, 2008). However, there is very recent analysis that the theoretical opposing of ‘multi-level’ and ‘kin’ selection, with a too strict understanding of ‘kin selection’, ignores a real basis of evolution of co-operation through the interplay of genetic and population structuring (Lion, Jansen & Day, 2011). Other researchers arrived at a similar model (Powers, Penn & Watson, 2011). Another perspective is that lower individual fitness in the short-term can be more than compensated by an increase in fitness in the long-term, through the exploitation of an aspect of the selection process itself in ‘lineage selection’ (Nunney, 1999). With a number of mathematically equivalent rival models, then it is more a question of which model is preferred philosophically than which is empirically justified. As Clutton-Brock points out (personal communication, 2011): “mutualism works”, however it may be theorised; so the question as to which theory is required as explanation would be a meta exercise other than for those working in the field of the philosophy of biology. The point at issue is that theory is not an obstacle to positing fitness-enhancing temporary curtailment of reproduction in the local reproductive group to reduce population to a sustainable level.

Evidence for just such a mechanism would appear to have been provided, and strikingly so, by very recent research on a mammalian model species (the mouse), but hitherto its applicability to DT is an insight that has not been made. It had previously been established that social ‘crowding’ stress produces a distinct stress response in males but not females (Laviola et al, 2002); that there are epigenetic changes within the brains of stressed individuals (Hunter, 2012); that there is paternal transmission (Dietz et al, 2011); and that stress from crowding can affect reproductive activity even through the second generation (Marchlewska-Koj, 1997). Now, Saavedra-Rodríguez & Feig find that the presence of unfamiliar others causing social ‘crowding’ stress in males is then epigenetically transmitted to female offspring to produce physiological fertility decline and degradation of reproductive-related behaviours of great effect sizes; this being then further transmitted down the female line for at least another two generations (the male’s f1, f2 and f3 female offspring) (Saavedra-Rodríguez & Feig, 2012). The trigger, though described as social ‘crowding’ stress, Saavedra-Rodríguez & Feig suggest or indicate in their title and content that this is / may be an operational proxy for either a more general ‘social instability’, or, more specifically, one concerning male dominance hierarchy. Henceforth here the trigger will be referred to as ‘crowding’ stress / social instability, abbreviated to ‘CS/SI’.

It is unlikely that a fundamental mechanism of such large effect present in one mammal would not be present across all mammals, so it would not be expected to be absent or vestigial in humans. That this research is likely to apply well to humans is indicated by the congruence between human and ‘monogamous’ mouse models regarding the impact of male parental absence – which is itself an indicator of social instability. Bambico et al (2013) found abnormal social interactions and greater aggressiveness (associated with pre-frontal cortex changes) far more pronounced in female than in male male-parent-absent ‘monogamous’ mouse offspring; this being consistent with human studies of children raised without a father.

This is a profoundly counter-intuitive finding. A mechanism directly causing obligatory markedly lower reproduction is here heritable by daughters, grand-daughters and great-grand-daughters, despite this (the female) half of the lineage being the limiting factor in reproduction. The multi-generational transmission of obligatory reproduction-lowering traits is evolutionarily strange enough; but for it to be specifically to the female half of the lineage, where it would have most impact on reproductive output, is particularly unexpected. The suspicion must be of an adaptive mechanism to produce a sharp decline in population levels within the local reproductive group, presumably in response to a threat to ecological viability.

The transmission from male to female is the question posed as supposedly unfathomable in a commentary on Saavedra-Rodríguez & Feig (2013) (Champagne, 2013). Yet the function of this aspect is not difficult to establish. Male mice are more likely than are females to encounter and contest with stranger-conspecifics, and therefore males are more exposed to CS/SI. However, the high-ranking males (who would be responsible for the bulk of reproduction) will themselves experience little impact of glucocorticoid-mediated stress. This is because not only (in comparison to low-ranking males) do glucocorticoid stress hormones in high-rankers quickly revert to a low baseline (eg., Summers & Winberg, 2006), but the higher testosterone levels in high-ranking males will decrease glucocorticoid levels owing to mutual antagonism (Glenn, 2009) (presumably through testosterone ‘blocking’ the Type II glucocorticoid receptors). The male low-rankers by contrast are reproductively suppressed by glucocorticoid-mediated stress, leaving the high-rankers to take a still greater lion’s share of reproductive opportunities. Thus is the normal working of reproductive-skew apparently functioning to maximise reproductive efficiency within the local reproductive group, in what seems to be the standard basis of social system whereby species vary along a continuum with the extreme reproductive skew evident in ‘co-operative breeding’ species at one extreme, rather than ‘co-operative breeding’ species being a distinct form of sociality (Moxon, 2009). Indeed, humans are now considered to be a ‘co-operative breeding’ species (Sear & Coall, 2011). Consequently, notwithstanding exposure to CS/SI, male high-rankers seem ill-equipped to eschew maximising and instead to temporarily curtail their reproduction in response to an indicated threat to ecological viability. A mechanism is required whereby the registering of CS/SI by all males including high-rankers is somehow translated into lower reproduction; and the obvious solution would be to transmit this registering to females (by an epigenetic modification), to then leave the females – as the sex that is always the ‘limiting factor’ in reproduction — to self-suppress their own reproduction. The otherwise peculiar-seeming male-to-female transmission is thereby neatly explained.

It is not unreasonable, then, to posit a mechanism extending usual reproductive-skewing beyond maximising reproductive efficiency of the local reproductive group actually to reduce population. This would be adaptive if it could head off a precipitous population decline that otherwise would lead to a local extinction. Albeit that a tactical reduction in population risks the strategic failure of over-shooting to produce the very outcome it functions to avoid; if statistically such a mechanism more often results in preventing local extinction than its facilitation, then it would have adaptive value, and there would be selection of the underpinning genes through an evolutionary process supporting ‘mutualism’, as briefly discussed above. (Note that over-shooting would be expected to be inherent in the mechanism, because some degree of over-shooting in itself would be adaptive, given the likelihood that serious ecological degradation having occurred once will occur again in the very near future.)

There is a pedigree for this hypothesis, though it has long lain dormant. A stress-based population self-regulation mechanism was proposed well over half a century ago in the wake of the finding that snowshoe hares undergo population cycles, at the peak of which even slight stress results in death through hypoglycaemic shock (Green, Larson & Bell, 1939). The theory and relevant findings are recently summarised by Feldhamer (Feldhamer, 2007, pp471-478):

“Christian (1950) proposed that mammalian populations could be regulated by shock disease caused by exhaustion of the adrenal gland, following prolonged psychological stress from agonistic interactions at high population levels. This idea grew from Selye’s (1950) work on the general adaptation syndrome (GAS) … The phenomenon of death by adrenal exhaustion turned out to be an extreme case, and Christian (1978) subsequently modified his hypothesis … There is, in fact, a rise in adrenocortical output in response to increasing population density ….. growth and sexual maturation are inhibited by increased adrenocortical output, as are spermatogenesis, ovulation and lactation (Rivier et al, 1986). Some of these effects on reproduction persist even into subsequent generations, in spite of a reduction in population density. … Several mechanisms have been proposed by which populations of mammals could regulate their own numbers.”

Notwithstanding, then, the previous discovery of what would appear to be multi-generational transmission of stress to produce reproductive-suppression (complete with the expected over-shooting), hitherto this has not been considered evidence for an evolved mechanism of population self-regulation – though, of course, no evidence could be seen to suffice for those who do not accept the possibility of a theoretical framework that could contain it (the issue of whether or not there is a tenable mechanism for ‘mutualism’). The idea has had recent currency only in a popular science account (Morrison, 1999), in which adaptive tactical population crash is idiosyncratically termed as either ‘plague mode’ or ‘the general adaptation syndrome’ – borrowing Selye’s term for a multi-stage common stress response confusingly to have this more specific meaning. Morrison cites several-decades-old research, but rather than a critical review asserts a population self-regulation model as if it is accepted science. Clearly, this is not the case, but the new evidence re CS/SI and its epigenetic transmission should prompt a re-evaluation.

For CS/SI to be a candidate key underlying factor in DT, it would have to be plausible as ubiquitous, irrespective of local economic conditions. There would need to be psychologically salient CS/SI across the world not just in obviously crowded urban milieus but also in rural contexts – in all of the nations and regions within nations where fertility decline is in evidence; a genuine pan-global factor, which in this would be unlike any of the putative factors identified thus far, with the exception of falling child mortality. With child mortality decline not sufficient on its own to account for DT, and economic development not being sufficiently ubiquitous, then CS/SI could then be the missing linchpin in the aetiology. There would be a need for research to quantify the requisite conditions and impact of CS/SI in humans, but there is, at least, data on the presence of strangers (/ social instability) in large overwhelming proportion in comparison to familiar-others in various regions around the world, in the form of migration statistics. Examination of migration patterns reveals an unexpected major phenomenon of rural-to-rural as well as rural-to-urban. In poor countries, individual males especially seem actually more likely to be in a social milieu of strangers in rural areas than in urban centres, through male migration that is either seasonal or permanent to take up or seek paid work. So here there is a very large-scale basis of specifically male CS/SI in the very geographical locations where it would be least expected.

To very briefly review: in lower-income countries this is far more common than rural-to-urban migration (Lucas, 2007); with male labourers relocating from subsistence to plantation agriculture, and moving village between dry farming and irrigated areas. In Ghana, for example, rural-rural permanent and seasonal migration of men so as to send remittances home is the predominant pattern (Primavera, 2005). This pattern was evident several decades ago in Kenya, where the percentage of all migration that was rural-rural as against rural-urban was 40%/33% (Oucho, 1984). The split was 68%/25% more recently in Nepal, and similarly in India (Deshingkar & Grimm, 2004). Oucho argued that: “students of migration have shown unwarranted obsession with rural-urban at the expense of rural-rural migration in the developing countries” (Oucho, 1984, p123). Deshingkar & Grimm point out that rural-rural migration is: “the least visible because such migration is usually missed by official surveys (through) the inability to capture seasonal and part-time occupations; covering only registered migrants; and … owing to scattered locations of sending and receiving areas” (Deshingkar & Grimm, 2004, pp11, 20 & 21). Unfortunately, these sources of inaccuracy would compromise measuring both CS/SI (of which migration flows would be an indirect rough proxy) and direct female fertility reduction (caused by the afore-mentioned confound of seasonally absent husbands), thereby obscuring an overall picture; so better data likely will be required.

With CS/SI of males a viable candidate underlying cause of DT, future research is needed to extend the work on the mouse to investigate an expected homologous mechanism across mammalia to include primates and to not exclude humans. The state of the literature on human social ‘crowding’ stress at present is so contradictory, ill-defined and mostly out-of-date as hardly to repay a review – as evident in the most recent (Ramsden, 2009). Furthermore, most research re humans of crowding has concerned physical confinement within the home, which is with familiar others (and usually close kin) and therefore not relevant to the different sort of stress caused by the social proximity of strangers, or more general social instability. Studies have been in sociology / psychology, not biology; so the types of impact investigated have not included depressed reproduction-related behaviour and physiological fertility. The review most commonly cited (Yakov & Epstein, 1981) deals with crowding in terms of producing a sense of loss of ‘control’ and problems of group identification; which may mediate the phenomena here at issue but are not proxy measures of them.

In formulating an hypothesis there are several afore-mentioned considerations, beginning with the question as to the specific nature of the trigger mechanism at issue: whether it is social ‘crowding’ stress per se, or more specifically disruption of male dominance hierarchy, or — as Saavedra-Rodríguez & Feig suggest/indicate – a more general social instability. It may be that in whichever case a general, quantifiable measure of social dislocation in human populations suffices, but there would need to be control of the confounding obvious direct impact of social dislocation on fertility in the seasonal separation of couples through male migration preventing husband-wife conception for the duration of the male’s absence.

Another mechanism that would be consonant with social dislocation being the apparent driver of DT is an additional simple biological (rather than adaptive) phenomenon: out-breeding fertility depression. This previously has been suggested as a possible principal biological basis of DT (Helgason et al, 2008), and may be a consequence of social dislocation complementary to the proposal here of an adaptive response to ‘crowding’ stress. The relationship between genetic relatedness and fertility is an n-shaped curve, given that a very close genetic relatedness between the male and female of a couple tends to pair up harmful recessive alleles, whereas less close genetic relatedness produces hybrid vigour as harmful recessive genes are masked; but, conversely, as genetic relatedness becomes progressively ever more distant then there is exposed a disparity between the contexts in which particular genes and gene-complexes have co-evolved, causing incompatibility to the extent of corresponding progressively reduced fertility (eventually, with sufficient distance, sterility).

Helgason et al used the meticulous genealogical records of Iceland to reveal a very clear monotonic relationship between degree of relatedness of couples and the number of children and grand-children produced, even with fine distinctions in relatedness, and holding for those born in each successive 25-year period of the data from 1800 to 1965. The overall decrease in mean kinship between the partners in couples across the whole timespan was by a factor of ten. Fertility peaked at an average level of consanguinity between third and fourth cousins, before declining to that of fifth cousins. Given Iceland being notably socio-economically homogenous and with little change in social structure over the whole time-frame, and only minor variation in family size, contraceptive use, and marriage practices; the authors conclude that the phenomenon is a biological one, as a response to the change in Iceland from a rural to an urban society with the usual accompanying great population increase. This was, of course, a great social dislocation, so the data may be a measure not of out-breeding fertility depression but of the mechanism herein outlined: fertility decline through epigenetic effects consequent to ‘crowding’ stress. Alternatively, the data may reflect both mechanisms together impacting on fertility.

The findings from Iceland are supported by data from, respectively, Danish and Brazilian populations (Labouriau & Amorim, 2008; Weller & Santos, 2013), and is consonant with research in other species (Edmands, 2007), where lowered fertility with distance of genetic relatedness is found to be produced by the breakdown in co-adapted gene complexes. This is the standard explanation of ‘genetic mismatch’, which appears to be in particular that between nuclear and mitochondrial genes (eg., Chou et al, 2010). Incompatibility causing fertility decline has been discovered in the levels of proteins (that is, in mere quantitative differences in the same rather than different proteins) (Thomae et al, 2013), and a non-coding region of the genome has been identified as the locus of a mechanism of fertility reduction (Ferree & Barbash, 2009), which, interestingly paralleling what is found in the stress-triggered reproductive-suppression mechanism, impacts not male but only female embryos.

It would seem likely, or not unlikely, that both the here-outlined adaptive and this simple biological mechanisms are factors manifest in DT, and may be not merely factors but underlying drivers of the phenomenon. Future work is required to disentangle them from each other as well as from the confound of migration patterns, to establish their respective relative significance to confirm if either or together they are of sufficient strength to be the primary driver(s) of DT, leaving other factors as secondary. With the interplay of ultimate and proximate causes, there may be required better quality data and/or more advanced statistical analyses than have been available if an adaptive aetiology can be formulated as an hypothesis falsifiable not merely in principle but to be empirically testable. The new analytic techniques in economics to investigate causation across hypotheses presumably can encompass these new hypotheses, so it may well be in the economics rather than the (evolutionary) anthropology literature that, as has been the case recently, progress in modelling DT continues. It surely would be fruitful for the divide between the literatures to be bridged, and a key way forward would be for evolutionary/ biological anthropologists to adopt the new causation-testing models themselves.

The male ‘crowding’ stress hypothesis, on its own or together with the out-group fertility depression hypothesis, would be a parsimonious answer as to why hitherto it has proved protractedly so difficult to model DT simply in terms of economics/ demographics/ social norms. Modelling necessarily in future would then have to take account of a very different set of causal relations. However, with such a complex multi-factorial phenomenon as DT — given likely unforeseen and unquantifiable additional variables, with the problems of fully identifying and respectively weighting all relevant factors and understanding how they interact – it is unlikely that this would lead to modelling sufficiently robust as to have not merely descriptive but predictive value with any ecological validity. Confirming an adaptive and/or other biological aetiology as a fully scientific theory may prove elusive, but with the importance of DT to strategic planning everywhere around the world it is important to develop and integrate whatever lines of research arise, notwithstanding the difficulties.

REFERENCES

Bambico, F.R., Lacoste, B., Hattan, P.R. & Gobbi, G. 2013. Father Absence in the Monogamous California Mouse Impairs Social Behavior and Modifies Dopamine and Glutamate Synapses in the Medial Prefrontal Cortex. Cerebral Cortex. Published online December 4, 2013. doi:10.1093/cercor/bht310

Barkow,J.S. & Burley, N. 1980. Evolutionary Biology, Human Fertility, and the Demographic Transition. Ethology and Sociobiology, 1, 163-180

Bongaarts, J. & Watkins, S.C. 1996 Social interactions and contemporary fertility transitions. Population and Development Review, 22(4), 639-682

Borgerhoff Mulder, M. 1998. The demographic transition: are we any closer to an evolutionary explanation? Trends In Ecology, 13(7), 266-270

Champagne, F.A. 2013. Effects of stress across generations: Why sex matters. Invited commentary on: Saavedra-Rodriguez, L. & Feig, L.A. 2013. Chronic Social Instability Induces Anxiety and Defective Social Interactions Across Generations Biological Psychiatry, 73(1), 44-53

Canning, D. 2011. The Causes and Consequences of the Demographic Transition. PGDA Working Paper No. 79. Harvard School of Public Health http://www.hsph.harvard.edu/pgda/working.htm

Cervellati, M. & Sunde, U. 2013. The Economic and Demographic Transition, Mortality, and Comparative Development Discussion Paper DP9337, Centre for Economic Policy Research, London

Chou, J-Y., Hung, Y-S., Lin, K-H., Lee, H-Y. & Leu, J-Y. 2010. Multiple Molecular Mechanisms Cause Reproductive Isolation between Three Yeast Species. PLoS Biology, 8(7).

Christian, J.J. 1950. The adrenopituitary systemand population cycles in mammals. Journal of Mammalogy, 31, 247-259

Christian,  J.J.1978. Neurobehavioral endocrine regulation in small mammal populations. In Populations of Small Mammals under Natural Circumstances (Snyder DP ed), 143-158, University of Pittsburg Press

Deshingkar, P. & Grimm, S. 2004. Voluntary Internal Migration. An Update. Overseas Development Institute, London

Dietz, D.M., LaPlant, Q., Watts, E.L., Hodes, G.E., Russo, S.J., Feng, J., Oosting, R.S., Vialou, V. & Nestler, E.J. 2011 Paternal transmission of stress-induced pathologies. Biological Psychiatry 70(5) 408

Dribe, M. & Scalone, F. 2010. Detecting Deliberate Fertility Control in Pre-transitional Populations: Evidence from six German villages, 1766–1863. European Journal of Population, 26(4), 411-434

Edmands, S. 2007. Between a rock and a hard place: evaluating the relative risks of inbreeding and outbreeding for conservation and management. Molecular Ecology, 16 (463), 463-475

Ellis, B.J., Figueredo, A.J., Brumbach, B.H. & Schlomer, G.L. 2009. Fundamental dimensions of environmental risk: The impact of harsh versus unpredictable environments on the evolution and development of life history strategies. Human Nature, 20, 204-268.

Feldhamer, G.A., Drickamer, L.C., Vessey, S.H., Merritt, J.F. & Krajewski, C. 2007. Mammalogy: Adaptation, Diversity, Ecology. John Hopkins University Press

Ferree, P.M. & Barbash, D.A. 2009. Species-Specific Heterochromatin Prevents Mitotic Chromosome Segregation to Cause Hybrid Lethality in Drosophila. PLoS Biology, 7(10)

Glenn, A.L. 2009. Neuroendocrine Markers of Psychopathology. In Titsner ed, The Handbook of Neuropsychiatric Biomarkers, Endophenotypes and Genes. Volume III. Metabolic and Peripheral Biomarkers. Springer Science + Business Media

Goodman, A., Koupil, I. & Lawson, D.W. 2012. Low fertility increases descendant socioeconomic position but reduces long-term fitness in a modern post-industrial society. Proceedings of the Royal Society (Biological Sciences)

Green, R.G., Larson, G.L. & Bell, J.F. 1939. Shock disease as the cause of the periodic decimation of the snowshoe hare. American Journal of Hygiene, 30, 83-102

Guinnane, T.W. 2011. The Historical Fertility Transition: A Guide for Economists. Journal of Economic Literature, 49(3), 589-614

Helgason, A., Palsson, S., Guthbjartsson, D.F., Kristjansson, T. & Stefansson, K. 2008. An Association Between the Kinship and Fertility of Human Couples. Science, 319(5864), 813-816

Heligman, L. 2002. Speech for the UN Population Division, April 2002

Herzer, D., Strulik, H. & Vollmer, S. 2012. The long-run determinants of fertility: One century of demographic change 1900-1999. Journal of Economic Growth, 17(4), 357-385

Heuveline, P. 2004. Mortality and Fertility Interactions: New Insights from Recent Population Dynamics in Cambodia. Paper presented at the annual meeting of the American Sociological Association, San Francisco

Hunter, R.G. 2012. Epigenetic effects of stress and corticosteroids in the brain. Frontiers in Cellular Neuroscience, 6(18)

Labouriau, R. & Amorim, A. 2008. Human fertility increases with marital radius. Genetics, 178, 601

Laviola, G., Adriani, W., Morley-Fletcher, S. & Terranova, M.L. 2002. Peculiar response of adolescent mice to acute and chronic stress and to amphetamine: evidence of sex differences. Behavioral Brain Research, 130(1-2), 117-125

Lee, R.D. & Reher, D.S. 2011. Introduction to the Landscape of Demographic Transition. Population and Development Review, 37, Issue Supplement s1, 1-7

Lion, S., Jansen, V.A.A. & Day, T. 2011. Evolution in structured populations: Beyond the kin versus group debate. Trends in Ecology and Evolution, 26(4), 193

Lucas, R.E.B. 2007. Migration and rural development. Journal of Agricultural and Development Economics, 4(1), 99-122

Malmberg, B. 2008. Demography and the Development Potential of Sub-Saharan Africa. Institute For Future Studies (Part of the White Paper on Africa submitted to the Swedish Parliament)

Marchlewska-Koj, A. 1997. Sociogenic stress and rodent reproduction. Neuroscience & Biobehavioral Reviews, 21(5), 699-703

Mathews, P. & Sear, R. 2008. Life after death: An investigation into how mortality perceptions influence fertility preferences using evidence from an internet-based experiment. Journal of Evolutionary Psychology, 6 (3), 155-172

Mathews, P. & Sear, R. 2013. Family and Fertility: Kin Influence on the Progression to a Second Birth in the British Household Panel Study. PLoS ONE,  8(3), e56941

Morrison, R. 1999. The Spirit of the Gene: Humanity’s Proud illusion and the Laws of Nature. Cornell University Press

Moxon, S.P. 2009. 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. Medical Hypotheses, 73(1), 5-14

Moxon, S.P. 2013. Human pair-bonding as primarily a service to the female (in excluding other males of lower (but not higher) mate-value, and a buffer against her own age-related mate-value decline). New Male Studies, 2(2), 24-38

Newbold, K.B. 2007. Six Billion Plus: World population in the Twenty-First Century. Rowman & Littlefield

Newson, L., Postmes, T., Lea, S.E.G. & Webley, P. 2005. Why are modern families small? Towards an evolutionary and cultural explanation for the demographic transition. Personality And Social Psychology Review, 9(4), 360-375

Newson. L., Postmes, T., Lea, S.E.G., Webley, P., Richerson, P.J. & McElreath, R. 2007. Influences on Communication about Reproduction: The Cultural Evolution of Low Fertility. Evolution and Human Behavior, 28, 199-210

Novak, M.A., Tarnita, C.E. & Wilson, E.O. 2010. The evolution of eusociality. Nature, 466, 1057-1062

Nunney, L. 1999. Lineage selection: natural selection for long-term benefit. In Keller (ed) Levels of Selection in Evolution, 238-252

O’Sullivan, J. 2013. Revisiting demographic transition: correlation and causation in the rate of development and fertility decline. Paper presented at the 27th IUSSP International Population Conference, 26-31 August 2013, Busan, Korea.

Okasha, S. 2006. Evolution and the Levels of Selection. Clarendon Press

Oucho, J.O. 1984. Rural-Rural Migration Field in Kenya: The Case of Kericho Tea Estates Complex in a Regional Setting Geografiska Annaler. Series B, Human Geography, 66(2), 123-134

Palloni, A. & Rafalimanana, H. 1999. The effects of infant mortality on fertility revisited: new evidence from Latin America. Demography,36(1), 41-58

Powers, S.T., Penn, A.S. & Watson, R.A. 2011. The concurrent evolution of cooperation and the population structures that support it. Evolution,65(6), 1527-1543

Primavera, C. 2005. Rural-rural migration in Ghana. The effects of out-migration on the sustainability of agriculture in the Upper West Region, Ghana. Masters thesis, University of Amsterdam, Faculty of Social and Behavioural Sciences, Human Geography — Environmental Geography of Developing countries

Pritchett, L. & Viarengo, M. 2012. Why Demographic Suicide? The Puzzles of European Fertility. Population and Development Review, 38 (Supplement), 55-71

Ramsden, E. 2009. The urban animal: population density and social pathology in rodents and humans. Bulletin of the World Health Organisation, 87(2), 82

Rivier. C., Rivier. J. & Vale, W. 1986. Stress-induced inhibition of reproductive functions: Role of exogenous corticotropin-releasing factor. Science, 231, 607-609

Saavedra-Rodriguez, L. & Feig, L.A. 2013. Chronic Social Instability Induces Anxiety and Defective Social Interactions Across Generations. Biological Psychiatry, 73(1), 44-53

Santow, G. 1995. Coitus interruptus and the control of natural fertility. Population Studies, 49(1), 19-44

Sear, R. & Coall, D. 2011. How Much Does Family Matter? Cooperative Breeding and the Demographic Transition. Population and Development Review, 37, 81-112

Selye, H. 1950. Stress and the General Adaptation Syndrome. British Medical Journal, 4667, 1383-1392

Shenk, M.K., Kress, H.C. & Towner, M.C. 2011. Why Does Fertility Decline? Comparing Evolutionary Models of the Demographic Transition. Population Association of America, Annual Meeting, Washington DC.

Shenk, M.K., Towner, M.C., Kress, H.C. & Alam, N. 2013. A model comparison approach shows stronger support for economic models of fertility decline. Proceedings of the National Academy of Sciences USA, 110(20), 8045-8050

Strulik, H. & Weisdorf, J. 2010. How Child Costs and Survival Shaped the Industrial Revolution and the Demographic Transition: A Theoretical Inquiry Leibniz University at Hannover., Discussion Paper No. 442

Summers, C.H. & Winberg, S. 2006. Interactions between the neural regulation of stress and aggression. Journal of Experimental Biology,209, 4581-4589

Teitelbaum, M.S. 1975. Relevance of demographic transition theory for developing countries. Science, 188(4187), 420-425

Thomae, A.W., Schadem G.O.M., Padeken, J., Borath, M., Vetter, I., Kremmer, E., Heun, P. & Imhof, A. 2013. A Pair of Centromeric Proteins Mediates Reproductive Isolation in Drosophila Species. Developmental Cell, 27(4), 412-424

Turke, P.W. 1989. Evolution and the Demand for Children. Population and Development Review, 15(1), 61-90

Tymicki, K. 2004. Kin influence on female reproductive behavior: The evidence from reconstitution of the Bejsce parish registers, 18th to 20th centuries, Poland. American Journal of Human Biology, 16(5), 508-522

Weller, M. & Santos, S. 2013. A positive association between consanguinity and fertility in communities of Paraíba, Northeast Brazil. Annals of Human Biology, 40 (6), 527-530

Werfel, J. & Bar-Yam, Y. 2004. The evolution of reproductive constraint through social communication. Proceedings of the National Academy of Sciences, USA, 101(30), 11019-11024

Yakov, M. & Epstein, Y.M. 1981. Crowding Stress and Human Behavior. Journal of Social Issues, 37(1), 126–144