Human Behavioural Ecology

Monique Borgerhoff Mulder. Encyclopedia of Life Sciences: Supplementary Set. Volume 23, Wiley, 2007.

Human behavioural ecology emerged in the mid-1970 s as a result of applying the theory of evolution by natural selection to the study of human behaviour. Using explicit models to derive hypotheses that are tested with quantitative data primarily drawn from traditional human communities, it offers a natural science of sociocultural diversity.

Optimality and Human Behaviour

Human behavioural ecology (HBE) can be defined as the evolutionary ecology of human behaviour. Its central focus is how the behaviour of modern humans reflects our species’ history of natural selection. The field has grown rapidly over the last twenty years, in anthropology and other social and behavioural sciences. It passes under many names, including Darwinian (or evolutionary) anthropology, human evolutionary ecology, evolutionary biological anthropology, human ethology, sociobiology, socioecology and biosocial (or biocultural) science. Most researchers generally shun using the controversial term sociobiology, for three reasons: the inaccurate equation of sociobiology with kin selection (which is but one of its models), the general but erroneous view that sociobiologists see behaviour as genetically determined, and to distance themselves from popular but highly speculative works under that label.

Behavioural ecology (including HBE) sprang from a growing emphasis within evolutionary biology and animal behaviour throughout the 1960 s and 1970 s on individual level selection. Already by 1956 the British evolutionary biologist J. B. S. Haldane had argued that behavioural differences could be analysed as the responses of human beings with basically similar genetic compositions to varying environments. This idea was developed and given prominence by E. O. Wilson’s path-breaking book Sociobiology: The New Synthesis, in what turned out to be an inflammatory last chapter on humans. It was subsequently explored empirically in a number of edited volumes. Beginning as an embattled group of believers strongly committed to exploring the relevance for humans of the ideas of prominent biologists such as Richard Alexander, Eric Charnov, William D. Hamilton, John Maynard Smith, Gordon Orians, Robert Trivers, George Williams and Edward Wilson, HBE has now established itself by successfully adapting evolutionary ecology theory and methods to a wide range of topics important to anthropology and archaeology (Winterhalder and Smith, 2000).

Assumptions and Models

The principal assumption of HBE is that people have been selected to respond flexibly to environmental conditions in ways that enhance their fitness. Critical here is the idea that a history of natural selection endows our species with the ability to weigh the costs and benefits of adopting particular strategies, and that these decision rules (and the cognitive and physiological machinery behind them) are the focus of selection. As such HBE frames the study of adaptive design in terms of decision rules, for example in context X do A and in context Y do B, such that behavioural variation arises as people match their conditional strategies to their particular sets of circumstances. Accordingly, large differences in behaviour among individuals with no correlated genetic differences can result from adjustment to varying opportunities and constraints, as in many other species studied by behavioural ecologists. Since fitness is difficult to determine, proximate outcomes (reproductive success, food income, or even status) are often used as proxy measures, bringing evolutionary models often very close to quantitative models developed in microeconomics, demography, and some schools within sociology and anthropology. A key component of this highly adaptationist approach emphasizing the current utility of traits is the phenotypic gambit. Under this set of assumptions the details of how a trait is inherited do not seriously constrain adaptive responses to ecological variation. In other words behavioural ecologists assume extreme phenotypic plasticity, a wide array of feasible strategies, and the ability of the actor to assess payoffs and/or learn the best alternative under any given set of circumstances.

On the basis of these straightforward but obviously quite controversial assumptions, HBE uses formal models to derive testable hypotheses from either graphical or mathematically explicit models anchored in basic principles of evolution by natural selection. To achieve generality these models seek to identify the bare essentials of an adaptive problem, and are therefore particularly useful in highlighting the key trade-offs that individuals face, for example between increasing the number of offspring born or ensuring the survival and quality of those already existing (optimal clutch size model). These models are drawn from larger bodies of theory, specifically from optimal foraging, life history, sexual selection and sex allocation theory. They predict optimal behavioural strategies given certain critical constraints and conditions and, where predictions are not met, can be revised to include other constraints or trade-offs, as identified either deductively or inductively. With increasing sophistication optimality models can incorporate the behaviour of other optimizing individuals (game theoretical models), as well as the consequences of previous optimizing decisions (dynamic state models), and can use simulations to estimate the payoffs associated with different alternatives.

The assumptions and models outlined above have generated a very productive first cut through analyses of foraging and food production, food sharing, territoriality and spatial use, reproduction, marriage, family organization and intergenerational inheritance, generating specific hypotheses to explain variation both within and between different human societies. To date most empirical work has been done in contemporary so-called traditional societies, ones only marginally impacted by globalization, where customary behavioural and subsistence patterns still obtain and where modern contracepting technology is largely absent; valuable studies have also been conducted in historical populations.

Foraging

Initial explorations of HBE were in the field of foraging subsistence, drawing explicitly from optimal foraging theory (OFT). This was in part because OFT was already quite sophisticated and testable by the early 1980 s, and in part because so much of the history of our species was spent as foragers (the first evidence of farming is at most 10 000 years old, much more recent in most places). Contemporary foragers offer natural experiments for human behavioural variability. If modern people who forage for a living are constrained by features of local ecology, then variation in these constraints, the trade-offs they impose, and the solutions adopted by individuals differing in age, sex and reproductive status are open to direct ethnographic observation. With a general understanding of the relationships between constraints, trade-offs and behavioural variability hypotheses can be posed both for contemporary variation and for possible scenarios for the evolution of the distinctly human life history and lifestyle.

Optimal foraging theory consists of a family of models addressing resource selection, time allocation and habitat movement (patch choice), with the diet breadth (or prey choice) model the most widely used in human studies. In accordance with this model human foragers are shown in general to select food resources that maximize mean rate of nutrient acquisition, by trading off the search and handling times associated with prey species of different profitability. Foragers routinely bypass resources yielding relatively low post-encounter rates when more profitable items are common, but take a broader array of prey when those items are rare (Kaplan and Hill, 1992). Specific applications explain shifts in subsistence patterns over time in response to such factors as changes in technology, climatic fluctuations, and the availability of imported substitutes. Thus the adoption of new technology can in some cases expand, and in other cases contract optimal diet breadth, depending on whether search or handling costs are most affected. There are even applications to archaeological deposits. For example, deposits associated with communities on the brink of adopting agriculture show increasing exploitation of locally abundant but previously unused resources, notably seeds and other plant food that require extensive processing. As such the diet breadth model suggests that agriculture emerged (many times) in response to a marked decline in encounter rates with higher ranked prey items, which may itself have resulted from terminal Pleistocene climatic change, human population increase, and human-induced habitat change. Finally, OFT can be extended to examine the dynamics of horticultural, pastoral and aquatic production systems, such as the impact of information sharing on search patterns in modern fisheries. In these areas HBE converges closely with microeconomic theory used by economists and others.

Failures to support model predictions have been just as enlightening as the successes. For example, despite the broad convergence of foraging behaviour with maximizing nutrient acquisition, men often favour large animal prey, ignoring plant food and other smaller game species that are profitable enough to increase their mean acquisition rates, and women frequently do just the opposite, taking plants and other small predictable prey rather than large animals. These observations have generated two alternative hypotheses. The first attributes the discrepancies to constraints—that men maximize nutrient gain by paying attention to a currency that gives higher weight to protein (particularly lipids) than to carbohydrates and calories, or that women are constrained in the resources they can take by the associated costs in child welfare. The second proposes that men favour prey items that are large and irregularly obtained, thus providing windfall resources that attract many claimants. According to this latter hypothesis the question of why men engage in the risky and often unproductive business of hunting and then share the meat group-wide hinges on the social benefits they derive from the attention that they draw to themselves with their apparent largesse and/or skill, and not from direct nutritional benefits. As such this latter hypothesis, derived from the study of contemporary foragers, is an alternative to the long-standing view that men hunt primarily to provision their wives and offspring (Bird, 1999).

Exciting extensions have emerged from applying foraging models to humans, such as in the field of conservation biology. The picture of traditional communities living in ecological harmony with their environments and protecting their resources from over-exploitation has faded under the bright light of behavioural ecology. Because OFT models identify precisely the behaviours expected by short-term optimization, they provide the null hypothesis against which such a traditional ‘conservation ethic’ can be evaluated. The evidence shows that for the most part traditional foragers select prey species in accordance with the predictions of shorter-term optimization models, irrespective of the vulnerability of these species to local depletion. In addition, by incorporating optimal foraging strategies into models that track the population of human predators and their prey, it is possible to predict how the addition or subtraction of species with given profitability affects the long-term prospects for other prey animals. For example, and counter-intuitively, encouraging foraging populations to keep fast breeding and productive domestic species can increase the vulnerability of wild species (as a result of a growth in the size of the human population).

Marriage’ Raising Offspring’ and Life History Allocations

An early and enduring focus in HBE is why human communities exhibit such variable culturally-sanctioned mating patterns, formalized as marriage systems. Given consistent evidence from historical and traditional societies that an individual’s access to or control of resources, both material and symbolic (for example, prestige), positively affects his or her reproductive success, resource-based models were lifted from the avian and mammalian literature to generate predictions from sexual selection theory about how resource distributions affect optimal strategies of males and females. The most general finding is that polygyny occurs where males can monopolize resources critical to female survival and reproduction. The critical mechanism here is the choice that women make between a wealthy married man and a poorer bachelor, a dilemma formalized in the polygyny threshold model. Attention has also been given to the much rarer marriage system, polyandry, which occurs most commonly in materially inhospitable environments such as the Himalayas. Limited inheritances (due to the scarcity of cultivable land) and the need for extremely high labour inputs encourage a younger brother to join his elder brother’s marriage as a secondary husband rather than strive for a monogamous marriage on his own. In both cases marital decisions can be modelled as a trade-off between different options, predicated on the recognition that the interests of the sexes do not necessarily coincide.

These models have been very productive in terms of linking marriage patterns to salient features of the environment. They also offer an explanatory framework for the very different patterns of intergenerational inheritance of material or status resources seen across different societies. Notably polygyny often is associated with inheritance to sons (who can turn this inheritance into plentiful grandchildren through their own polygynous marriages), whereas monogamy is, at least in stratified societies, associated with dowry payments (to daughters), facilitating upward mobility for girls and potentially greater returns with respect to grandparental fitness.

Equally well studied are patterns of parental investment. Human offspring require extensive and extended parental care, beginning in utero and ending in some cases as late as the execution of a will. This investment affects a child’s health, survival, future mating success, and hence the parent’s inclusive fitness (own fitness plus effects of its action on fitness of relatives weighted by their average relatedness). Most attention has been given to the adaptive reasons for differential investment in offspring. Parental fitness payoffs depend on three sets of factors, the genealogical relatedness between caregiver and child, the effect of investment on the reproductive value of the child (their expected future reproductive output), and the effects of caregiving on the caregiver’s own reproductive value. An extremely complicated model would be needed to examine the effects of each of these factors in the context of a parent beset by a group of offspring of different age, sex, ability, and perhaps even likely relatedness to the parent. To date, simple models have been adopted from nonhuman behavioural ecology that identify certain key variables that might bias parental investment in offspring, for the most part testing them separately or in small combinations. Many of the predictions from evolutionary based models have been supported (Winterhalder and Smith, 2000), suggesting the power of behavioural ecological analyses at the level of what goes on inside the family. There are a number of exciting new lines of investigation regarding investment. First, paternal provisioning may function more to attract and/or maintain a relationship with a mate than to provide offspring with resources. For example, men in several societies put more time and resources into stepchildren who are offspring of their current mates than they do into stepchildren from former relationships. This pattern would not be predicted if paternal care is parental investment per se, but would be predicted if paternal care was designed to attract mates. Second, the distinct objectives concerning parental care with which men and women enter a marriage are being explored (e.g. Blurton Jones et al., 2000), generating considerable overlap with what economists and demographers call marital bargaining theory.

The adaptive trade-offs at the core of life history studies are central to HBE. Organisms face two major allocation decisions, the first between growth and reproduction, and the second between the number of offspring produced and the amount to be invested in each. As such, life history studies are concerned with the evolution of maturation rates, reproductive rates and timing, dispersal patterns, mortality patterns and senescence. As regards the latter allocation dilemma, a highly influential early study used David Lack’s optimal clutch size model to show how the famous Kung San birth interval of 4 years is optimal with respect to the production of surviving offspring in the harsh Kalahari environment. Subsequent analyses of variation in huntergatherer fertility levels lend some support to the hypothesis that variations in fertility reflect the costs of child raising, which are themselves a consequence of the character and distribution of resources and associated age-and sex-specific foraging practices. As regards the allocation dilemma between growth and reproduction, perhaps the most successful study to date is that of Hill and Hurtado (1996) who examine from a life history perspective patterns of growth, fertility and mortality in a pre-contact foraging population (through painstaking reconstruction); for example, they adapt a life history model of reproductive timing to successfully predict variation in age of reproductive maturation of women in three different populations, ranging from foragers to urban industrial North Americans.

Comparing human life histories to those of other primates and mammals reveals at least four distinctive characteristics: a very large brain, an exceptionally long lifespan, an extended period of juvenile dependence, and support of reproduction by older post-reproductive individuals; a more controversial fifth feature is male support of reproduction through the provisioning of females and their offspring. Recent attempts to link these features through a behavioural ecological analysis of how these traits may have coevolved have generated two new narratives. The first views these five traits as coevolved adaptive responses to a dietary shift towards high-quality, nutrientdense and difficult-to-acquire food resources which occurred only with the emergence of our species. The second view questions this scenario and presents an alternative hypothesis, the grandmother hypothesis. In this model, the appearance of arid seasonal environments 1.8-1.7 million years ago put pressure on Homo ergaster mothers to find new sources of food to feed themselves and their offspring. To obtain sufficient food they turned to low-ranked food resources such as tubers that were widespread but demanded adult strength to harvest, and more work for the mother. According to the grandmother hypothesis, this constraint provided an opportunity for the ageing female relatives of these stressed mothers to offset the declines in their own fitness by stepping in to provision grandchildren, thus selecting for a postreproductive lifespan. The debate is as yet unresolved (Hawkes et al., 2001).

As in most other areas of HBE inquiry, simple models have sparked interest not just in the past but also the present. An example here is the study of the dramatic decline in fertility that has been occurring across the world over the last 150 years. The marked declines in fertility across human societies as they gradually became richer, and the specific tendency for the rich to reduce their family size before such adjustments among the poor, were taken by many critics of HBE as evidence that humans do not behave in ways that are adaptive, that they do not consistently use resources to augment their fitness, and thus that evolutionary approaches to human behaviour are not useful, even wrong. This position characterizes nicely some common misconceptions about an evolutionary approach—that all behaviour should be adaptive, that only panhuman patterns are evidence of evolved behaviour, and that models supported in nonhuman systems (e.g. strong correlations between dominance and reproductive success) should necessarily also be evidenced in humans. In fact the disruption of the common association between rank, wealth and fitness that is observed in different societies at different times over the recent demographic transition merely challenges HBE to examine more carefully the environmental, social and physiological factors that influence optimal fitness-maximizing strategies. A key idea here is that rather than maximizing the number of offspring raised, humans adjust fertility in ways that maximize longer-term fitness. This latter can be measured as the number of grandchildren, or in more complex analyses as a weighted product of children and wealth, a fertility function adjusted by risk, or a simulated function. These explanations build on the trade-off between offspring quality and off-spring quantity, highlighting the possibility that particularly high levels of offspring quality are required in societies dominated by labour markets that reward human capital. Without analyses of the long-term fitness consequences of a variety of different fertility strategies in different kinds of environments, it is probably still premature to conclude that the low current levels of fertility in much of the developed world are maladaptive. Nevertheless this area of research represents one of the most integrative lines of enquiry to emerge from HBE (Kaplan, 1996).

Individual and Collective Interests

Almost invariably humans belong to groups—families, broader kinship groups, villages, lineages, communities, tribes and nations, groups that may (or may not) be residential, and may (or may not) be inclusive of each other. Some of these groupings are relatively egalitarian in organization, whereas others are stratified, with certain individuals enjoying more power or prestige than others. The individual selectionist perspective of behavioural ecology highlights the conflict of interests among individuals at each level as well as the potential for cooperation within and between groups, as do some schools within economics, political science and sociology. As such, it explores the inherent costs and benefits entailed in group living, and the implications of these costs and benefits for the emergence of hierarchies and other structures (Boone, 1992). At present there is little theory development within HBE with respect to the emergence of complex social institutions, particularly insofar as these might arise from adaptively based individual decision-making. Here both theoretical and empirical integration with other social sciences is sorely needed. In certain domains, however, behavioural ecological analyses are suggesting intriguing ideas that may have relevance for higher levels of social organization.

The first is food sharing. At present, most of the hypothesis development for human sociality has centred on resource transfers within foraging groups, usually measured as food sharing. Commonly human foragers take resources that are larger than the harvester can consume, for example a medium-sized ungulate. Furthermore, if only one member of the group has been successful in the hunt and others come back empty-handed, portions of the packet will be differently valued by different group members, depending on how hungry they are. A large number of hypotheses addressing the evolutionary mechanisms that maintain food sharing have been proposed, and current evidence strongly suggests that in different circumstances different mechanisms may be important (Winterhalder and Smith, 2000). Thus individuals in control of a food resource may share it to provision their kin, to trade the resource for another valued scarce item, to ensure (through reciprocity) against the risk of being without food at some future point in time when hunting was unsuccessful, or to advertise quality or seek attention among those who have been unable to procure resources (a signalling hypothesis where the benefits may perhaps be reaped through sexual selection). The potential for exploring these ideas in other areas of sociality is enormous.

The second is cooperation. Despite the predictions of narrowly self-interested behaviour predicated on the ultimately competitive nature of selection, humans for the most part live in stable cooperative groups, and come together to produce valuable collective goods—goods that could not be produced by any one individual alone. Indeed this is probably one of the hallmarks of being human, shared only with the eusocial insects. Kin selection and reciprocity are commonly invoked as solutions to cooperative dilemmas, but many human groups are made up of non-kin and reciprocity is always vulnerable to cheaters. There have been many recent game theoretical attempts to address this problem, but one particularly fruitful empirical line of research lies in the field of experimental economics. Behavioural ecologists and others, administering experiments in many different cultural contexts, are finding that people often cooperate in anonymous one-shot games where non-cooperation would lead to a higher payoff for any individual player but cooperation leads to highest mean payoff for all the players. These findings suggest that models based on short-term material self-interest may need to be seriously revised when dealing with human sociality.

Conceptual Challenges

Evolutionary studies of behaviour ideally incorporate investigations of mechanism, development and phylogeny, as well as those of functional (or adaptive) significance. The focus of behavioural ecology (both human and other) is principally on function, and the environmental factors that shape variations in behaviour through the process of optimization. Less attention has been paid (until recently) to the precise mechanistic, developmental or historical causes of variation. Indeed, the phenotypic gambit is invoked ever more broadly to suggest that not only genetic, but also phylogenetic and cognitive mechanisms do not seriously constrain human adaptive responses to ecological variation. In other words a behavioural ecologist predicts a certain matching between behavioural strategy and context (as defined in formal models) irrespective of whether humans reach this adaptive strategy as a consequence of genes, psychological mechanism or the learning of culture. As such, this gambit becomes increasingly controversial, stimulating some truly novel developments that can only enrich the field insofar as behavioural ecology becomes more attuned to the special challenge of explaining what may be uniquely human.

First, in line with this expanded phenotypic gambit, behavioural ecological models generally assume the decision-maker has perfect information. Various sorts of models have been developed to deal with the shortcomings of this assumption. Specifically cultural evolutionary theorists (Boyd and Richerson, 1985) explore mathematically how certain learning (or transmission) biases may have evolved (through the normal processes of natural selection) to reduce the costs of arriving at a locally adaptive optimum. In certain circumstances a naïve individual does better to imitate the behaviour of successful individuals in his or her group, or simply do what everyone else is doing rather than experiment with all the (perhaps costly) strategic alternatives. With such evolved learning biases in place, there arises the possibility for traits to spread that are not adaptive in the strict sense of enhancing individual fitness. Though the extent to which cultural traits parasitize human minds is arguably quite limited, cultural evolutionary processes may well account for apparently maladaptive behaviour. Furthermore they may be able to shed much light on the more complex aspects of human sociality, such as institutions. Thus the true nature of cognitive mechanisms may contradict the phenotypic gambit, and thus interfere with current predictions derived from behavioural ecological models.

Second, behavioural ecological models often assume perfectly rational processing of this information. But how can anyone be rational in a world where knowledge is limited and expensive, and where time is often pressing? Evolutionary cognitive psychologists are beginning to explore more psychologically plausible ways in which decisions are made in the real world, with a focus on identifying, through simulation, ‘fast and frugal heuristics’ (Gigerenzer et al., 1999). In this respect evolutionary psychology has a real contribution to make to behavioural ecology, by drawing attention more closely to the decision-making apparatus and its constraints. Furthermore it is not only the decision-making processes that need closer scrutiny, the issue of currencies also needs disentangling; for example, dynamic modelling can be used to determine empirically what currencies humans appear to be maximizing in different domains of their lives.

A third often less explicit assumption in the logic of a behavioural ecological approach is that subjects are in evolutionary equilibrium—only under this condition could selection account for the observable covariation between behaviour and ecology. In other words, it is assumed that our ancestors experienced similar environments over a long enough span of time that rules governing optimal behavioural patterns have emerged. In the study of humans this raises the thorny issue of modern environments, specifically whether there has been sufficient continuity between present and past environments, at least with respect to the critical cues stimulating (adaptive) behavioural responses. Evolutionary psychologists argue that modern selective environments are very different from those in which human adaptations were forged, and hence that contemporary outcomes are often maladaptive (victims of adaptive lag); they therefore emphasize heavily the concept of an ‘environment of evolutionary adaptedness’ against which the supposed adaptiveness of past behaviour can be gauged. Human behavioural ecologists claim that what happens in the present informs our understanding of the past, just as much as vice versa.

Because of the problem of modern environments (most acute for humans, but not absent from most other species that experience anthropogenic environmental deterioration), HBE must distinguish different ways in which selection pressures affect, or have affected, traits with respect to our current observations (table 1). An adaptation is a trait that has been shaped by a history of natural selection; if the selection pressures that favoured it in the past are still acting it is a current adaptation, but if the selection pressures have changed such that the trait no longer has a selective advantage it is a past adaptation. Alternatively the trait may not have been selected in the past and either has no beneficial effect in the present (dysfunctional byproduct) or provides some new benefit in the novel environmental conditions of today (exaptation); note that an exaptation can be transformed into an adaptation in the future. Viewing the historical process of adaptation more broadly in this way highlights the need for HBE to coordinate its research agendas across a range of other social science disciplines, from experimental psychology to palaeoan-thropology.

Fourth and finally, behavioural ecologists have not yet entirely grappled with how to deal with the history of the traits under study. In conducting comparative study of the function of traits, biologists now standardly use phylogenically-based comparative methods to control statistically for the common origins of correlated traits, to avoid the problem of statistical dependence (or double counting as adaptations events that only happened once). The method has been introduced quite successfully into HBE. However, as an appropriate method for conducting comparative analyses it begs many questions as to how cultural traits spread between different groups and persist over time, and the extent to which these patterns reflect adaptive processes.

Most social scientists remain sceptical if not hostile to the evolutionary perspective embodied in HBE. This stems in part from the relativistic and postmodern agenda, from fashionable anti-science negativism, and from an acute awareness of past abuses of Darwinism (social Darwinism). At the same time it seems HBE has spilled so broadly into various subfields in a diverse set of disciplines—economics, psychology, political science, philosophy, demography, reproductive biology, law, and conservation biology, as well as the allied fields of archaeology and palaeoanthropology—that its profile is growing enormously. Furthermore, while its simplifying assumptions have been key to the development of useful theoretical models that are revolutionizing our understanding of human behaviour, both past and present, its shortcomings are stimulating new developments in parallel fields.

Table 1 The difference between adaptive behaviour and adaptations (Laland and Brown, 2002)

Is the behaviour adaptive?
Yes No
Is the behaviour an adaptation? Yes Current adaptation Past adaptation
No Exaptation Dysfunctional byproduct