Evolution of the human pygmy
phenotype
George H. Perry1
and Nathaniel J. Dominy2
1
Department of Human Genetics, University of Chicago, 920 E. 58th Street, Chicago, IL 60637, USA
2
Department of Anthropology and Department of Ecology and Evolutionary Biology, University of California, 1156 High Street,
Santa Cruz, CA 95064, USA
Small human body size, or the ‘pygmy’ phenotype, is
characteristic of certain African, Southeast Asian and
South American populations. The convergent evolution
of this phenotype, and its strong association with tropical
rainforests, have motivated adaptive hypotheses
that stress the advantages of small size for coping with
food limitation, warm, humid conditions and dense forest
undergrowth. Most recently, a life-history model has
been used to suggest that the human pygmy phenotype
is a consequence of early growth cessation that evolved
to facilitate early reproductive onset amid conditions of
high adult mortality. As we discuss here, these adaptive
scenarios are not mutually exclusive and should be
evaluated in consort. Findings from this area of research
are expected to inform interpretations of diversity in the
hominin fossil record, including the purported smallbodied
species Homo floresiensis.
The human pygmy phenotype
Body size is central to the biology of all living organisms,
affecting basic caloric requirements, basal metabolism,
foraging opportunities and the risk of predation [1]. Among
humans, there exists considerable intra- and interpopulation
variation in body size, an important component of
which can be attributed to genetic factors [2]. Although the
evolutionary ecology of human body-size diversity has long
intrigued anthropologists, the recent discovery of $18 000year-old
skeletal remains of an individual purported to be
$106 cm tall and representative of a new hominin species,
Homo floresiensis [3], has intensified the collective interest
in this topic [4–6].
Of modern human populations, Efe hunter-gatherers in
the Ituri rainforest (Democratic Republic of Congo) are the
shortest on record, with mean adult female and male
statures of 136 and 143 cm, respectively [7]. Historically,
the Efe and other populations with mean adult male
statures of <150, 155 or 160 cm have been termed ‘pygmies’
[6,8]. Such a broad classification tends to de-emphasize
the genetic, geographic and cultural distinctiveness of
these populations and, thus, might obscure the potential
ecological significance of small body-size evolution in
humans. On these grounds, it might be more appropriate
to limit the use of the term ‘pygmy’ to discussions of the
small body-size phenotype.
With few exceptions, the human pygmy phenotype is
associated with populations that traditionally hunted and
gathered food in tropical rainforest environments
(Figure 1). An understanding of how these challenging
habitats might have influenced the origins of this phenotype
could shed light on the evolutionary implications of
body-size diversity not only among modern humans but
also potentially of that within the hominin fossil record.
With this goal in mind, here we examine recent findings
that inform both classic and modern hypotheses of pygmy
phenotype evolution.
Genetics or growth stunting?
The small body sizes of rainforest hunter-gatherers might
partly reflect stunted growth from poor childhood nutrition
[9]; however, three lines of evidence suggest that this
phenotype is determined principally by genetic, rather
than environmental, factors. First, although the specific
DNA mutations have not yet been identified, genetic disruptions
of the growth hormone (GH) and insulin-like
growth factor I (IGF1) pathway are likely to have etiological
roles (Box 1). Second, the childhood growth rates of
some rainforest hunter-gatherer populations are surprisingly
fast; for example, growth rates for 3- to 10-year-old
Biaka (Cameroon) girls might even exceed those for USA
children (7.1 cm yÀ1
versus 6.5 cm yÀ1
, respectively) [10].
Therefore, for at least some populations, small adult body
sizes are a reflection of relatively slow growth in adolescence
rather than childhood [8,10], which is inconsistent
with a simple model of stunted growth from poor nutrition.
Indeed, other populations that also endure frequent episodes
of nutritional stress still achieve adult heights that
are greater than those of rainforest hunter-gatherers [8].
Third, the offspring of Efe mothers and Lese (agriculturalist)
fathers have statures intermediate to those of the two
parental populations [11].
Evolutionary theories
It is important to consider whether the genetically determined
component of the pygmy phenotype reflects adaptation
by natural selection or by neutral evolution (i.e.
genetic drift) [12]. Based on mitochondrial DNA (mtDNA)
and Y chromosome haplotype data, African, Southeast
Asian and South American rainforest hunter-gatherers
are more closely related to other populations from their
respective continents than they are to each other [13].
Therefore, it is likely that genetic mutations for the pygmy
Review
Corresponding authors: Perry, G.H. (gperry@uchicago.edu); Dominy, N.J.
(njdominy@ucsc.edu).
218 0169-5347/$ – see front matter ß 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.tree.2008.11.008 Available online 24 February 2009
phenotype occurred and have been maintained at least
three times in humans. The convergent evolution of this
phenotype in generally similar environments suggests that
it has been favored by natural selection, and several hypotheses
have been proposed to explain the mechanisms
under which this might have occurred. Here we present the
four hypotheses that, in our view, remain most tenable in
light of recent research.
When evaluating these (nonmutually exclusive) hypotheses,
it is crucial to consider the likelihood that at least
some of these populations have migrated or were displaced
from the environments in which the pygmy phenotype
evolved. Caution is therefore advised before rejecting a
hypothesis simply because it does not apply to the ecologies
of all modern small-bodied populations. Moreover, we
should not necessarily expect exactly the same scenarios
for each independent origin of the pygmy phenotype. This
notion might be especially relevant for those few smallerbodied
populations that inhabit savanna-woodlands rather
than tropical rainforests (Figure 1).
Food limitation
Rainforests are food-limited environments for human hunter-gatherers.
For example, substantive plant foods are
scarce for nearly half of each year in the Ituri forest, leaving
only small game (including fish), honey and insect
resources of a combined marginal nutritional and caloric
value [14]. Small body size might have conferred a selective
advantage to hunter-gatherers in these habitats by reducing
the total caloric intake necessary for survival [12]. In
effect, this model considers rainforests as ‘ecological
islands,’ and there are numerous examples of rapid size
decreases of larger land mammals following their appearance
on resource-limited islands (Box 2).
Figure 1. Association of the human pygmy phenotype with tropical rainforest habitats. (a) Approximate locations of small-bodied hunter-gatherer populations discussed in
this article, with mean adult male stature estimates [6,10,65,78,79]. The smallest modern human statures (mean adult male height < 155 cm) are always associated with
tropical rainforests (red circles). Some hunter-gatherer populations occupying savanna-woodlands (black circles) are also relatively small, such as the Hiwi of the
Venezuelan llanos, the Hadza of Tanzania and the !Kung San of Botswana and Namibia. Precipitation data are from the Tropical Rainfall Measuring Mission (Goddard Space
Flight Center, National Aeronautics and Space Administration; http://trmm.gsfc.nasa.gov). (b) Yanomamo¨ male, Venezuela (photograph by Raymond Hames, with
permission). (c) Efe male, Democratic Republic of Congo (photograph by William Wheeler, with permission from the National Anthropological Archives, Smithsonian
Institution). (d) Batek male, Malaysia, with white-handed gibbon (Hylobates lar) hunted by blowdart (photograph by Kirk Endicott, with permission).
Review Trends in Ecology and Evolution Vol.24 No.4
219
Cavalli-Sforza [11] rejected this hypothesis based on the
observation that African rainforest hunter-gatherers spend
only approximately half of each day obtaining and cooking
food, suggesting a surfeit of adequate resources. However,
this activity distribution might not reflect that under which
the pygmy phenotype evolved. Today, these African groups
exchange hunted forest game for cultivated foods with
neighboring agriculturalists, as do Southeast Asian rainforest
hunter-gatherers [15]. In fact, there is the open question
of whether full-time occupation of the rainforest is even
possible for independent hunter-gatherers (i.e. without
trade for cultivated goods), given the food limitations discussed
above [14–17]. If not, then these groups might have
originally inhabited rainforest-edge environments, where
food resources are generally more stable, before being displaced
to deeper forest habitats by the farming populations
with whom they now trade [14,15].
However, some archaeological evidence does support
human occupation of African and Asian rainforest environments
long before the spread of agriculture in these regions
[16,18]. In this event, any early rainforest inhabitants
(perhaps at lower population densities than today) are
likely to have experienced severe nutritional and caloric
stress, which in turn could explain the adaptive evolution
of the pygmy phenotype.
Thermoregulation
For humans, sweat production and evaporation is a crucial
component of thermoregulation. However, this mechanism
is inefficient in tropical rainforests, where small differences
between air and skin relative humidities are combined
with little air movement. Cavalli-Sforza [11]
suggested that, in the absence of effective evaporative
cooling, smaller bodies (potentially including relatively
decreased muscle masses, as suggested by calf circumference
measurements [11]) would benefit thermoregulation
by generating less internal heat during activity (Figure 2).
Interestingly, Biaka men (from Cameroon) were unexpectedly
efficient energetically relative to taller European men
when running under rainforest conditions (temperature
= 28.7 8C; relative humidity = 70%) [19]. Although
the European men had lower energetic costs at walking
speeds [19], the Biaka might have benefited from relatively
Box 1. The GH1-IGF1 pathway and the pygmy phenotype
As one of three endocrine systems regulating somatic growth and
stature, the GH1-IGF1 pathway has been an attractive candidate for
understanding the underlying physiological mechanisms of the
pygmy phenotype. Perturbations of this pathway have been
reported in rainforest hunter-gatherer populations from both Africa
(Biaka, Efe and Mbuti) and Southeast Asia (Aeta, Ati, Mamanwa,
Manni and Mountain Ok) [65–68].
Whereas GH1 plasma concentrations are normal in individuals
from each of these populations compared to controls, downstream
metabolic responses to GH1 are low. This GH1 resistance could be
mediated at the level of the GH1 receptor (GHR) or its ectodomain,
the GH binding protein (GHBP) [66]. Indeed, low circulating levels of
GHBP have been reported for individuals from each population.
Interestingly, with few exceptions, serum concentrations of IGF1 are
relatively low in rainforest hunter-gatherers from Africa, Thailand
and the Philippines, but not in the Mountain Ok from Papua New
Guinea, suggesting differences in underlying genetic mechanisms
[65–68]. Given that height is a complex trait influenced by many
genes, such differences are not unexpected. Such a finding would
follow examples of phenotypic convergence in other organisms that
are driven by different genetic mechanisms, such as the light coat
colors of beach mice from the Florida Atlantic and Gulf coasts [69]
and pigmentation changes among Drosophila species [70]. Relatively
decreased expression of the IGF1 receptor (IGF1R) has also
been observed in Efe-derived immortalized cell lines [66], but, to our
knowledge, this experiment has not been conducted for any other
population.
Because chronic undernutrition can lead to low concentrations of
GHR, GHBP and IGF1 [9], it is important to assess whether the
biochemical profiles linked to pygmy size are genetic in origin, or
the result of epigenetic effects. General health measurements and
plasma-based nutritional indices (e.g. albumin, prealbumin, zinc,
iron, etc.) are germane to this issue. For example, despite
differences in adult stature, the average skinfold thickness (a
measure of fatness) of Efe and taller Lese (agriculturalist) women
did not differ, indicating a comparable level of nutritional health in
the two populations [71]. Among the Ati (Philippines), plasma-based
indices revealed some minor nutrient deficiencies, but no energy
restriction overall [67]. Similarly, the serum proteins of the
Mountain Ok (Papua New Guinea) showed no evidence of chronic
or acute malnutrition [66]. As discussed above, levels of IGF1 among
Mountain Ok are similar to taller controls, in contrast to the low
levels of IGF1 reported in the Aeta, Ati, Efe, Manni and Mamanwa
[65–68]. This finding argues against a shared etiology for the pygmy
phenotype, further suggesting that the observed perturbations of
the GH1-IGF1 pathway are largely genetically determined. However,
until the specific mutations involved (probably regulatory ones) are
identified, the specific detailed mechanisms underlying the pygmy
phenotype will remain unknown.
Box 2. Foster’s Island Rule: insular dwarfism
Foster’s Island Rule is the phenomenon of miniaturization of large
animals (insular dwarfism) and gigantism of small animals on
oceanic islands and near-shore archipelagos. This rule applies to a
range of vertebrate lineages in the fossil and modern records,
including tortoises, lizards, birds, bats and non-volant mammals
such as primates, sloths and mammoths [72,73], although by no
means is it universal for even these taxa [74]. The strength of the
Island Rule tends to vary as a function of island size and isolation,
and among mammals it is greatest in herbivores and weakest in
terrestrial carnivores [72,75]. Several mechanisms (mainly competitive
release, resource limitation, dispersal ability and lighter
predation pressure on islands, as well as a general physiological
advantage of modal size) have been advanced to explain this
pattern. Recent life-history models of island ecology have stressed
the combined effects of decreased extrinsic mortality and competition
for resources on individual growth rates and body size [75,76].
The extent to which the mechanisms driving insular dwarfism are
also applicable to humans in tropical rainforests is unknown.
Similar to small islands, rainforests (for humans) can be considered
food-limited environments; but given that insular dwarfism has little
or no effect on mammalian carnivores, it is valid to consider
whether the same exception applies to humans, who can hunt for
food by both land and water.
Insular dwarfism might still be the most appropriate model to
explain the extremely small body size of Homo floresiensis (stature
$106 cm; mass $16–36 kg) [3,51]). However, insular dwarfism
normally results in a proportional reduction in brain size [50]. The
diminutive cranial capacity of H. floresiensis ($417 cm3
; at least as
measured from the holotype specimen, LB1) is substantially smaller
than expected given its body size [53]. This scaling discrepancy
could be associated with ecological differences between Flores and
other islands. Specifically, it has been suggested that Flores cannot
support a mammalian predator larger than 5 kg [77]. Intriguingly,
the cranial capacity of H. floresiensis scales to a hominin with a
body mass near this theoretical limit, $6 kg [50]. It is plausible that
selection for dwarfism on an island the size of Flores favored the
peculiar decoupling of brain size and body size in a species that was
both predator and herbivore.
Review Trends in Ecology and Evolution Vol.24 No.4
220
better thermoregulation during more rigorous exercise.
Therefore, the pygmy phenotype could confer a selective
advantage in tropical rainforest habitats by mitigating the
fitness-reducing effects of heat stress [11].
This hypothesis might fail, however, to explain the
variation in height among all small-bodied rainforest hunter-gatherer
populations, at least in terms of levels of
humidity and heat in their contemporary environments
[20]. In particular, humidity and body size are not clearly
negatively correlated; the Efe (the shortest population in
the world) inhabit one of the least wet tropical rainforests
(Figure 1), although the Ituri forest might have been more
humid 10 000 years ago [21]. In addition, other smallbodied
populations live in higher-elevation rainforests that
are relatively cool (although still very wet; e.g. the Mountain
Ok in Papua New Guinea). This hypothesis also has
little relevance for the few relatively small bodied populations
living outside rainforests.
Mobility
In open environments, longer limb lengths significantly
reduce the metabolic cost of locomotion [22,23]. This is
unlikely to be true in dense forests if taller individuals are
regularly forced into bent or crouched postures that markedly
increase energy expenditure [24]. The frequent use of
such postures would also inhibit rapid travel. Accordingly,
Diamond [20] hypothesized that the pygmy phenotype
might have adaptively improved mobility and increased
foraging efficiency. Although the appeal of this hypothesis
is readily grasped during an off-trail hike in the rainforest,
at present there are only anecdotal observations regarding
the density of understorey vegetation or other impediments
faced by human hunter-gatherers.
The ability to climb trees of incredible height for food
gathering (e.g. for honey, the Efe climbed as high as 51.8 m;
mean 19.1 m; SD = 9.7 m; n = 34 [25]) is another potential
mobility-related adaptive benefit of the pygmy phenotype
[20]. These food resources, especially honey, are crucial
nutritional components for African, Southeast Asian and
South American rainforest hunter-gatherers [14,16,26],
but their collection represents considerable energetic cost
to larger-bodied foragers [27,28]. Additionally, owing to the
possibility of falls, the overall fitness risk of tree climbing
for larger-bodied primates is expected to be high [29]; this
risk is likely to be positively correlated with body size.
Even among contemporary small-bodied rainforest huntergatherers,
accidents during the course of tree climbing for
honey and palm products (e.g. pith and cabbage [16])
remain a significant cause of mortality; 7 of the 106 (7%)
deaths recorded for young adult and adult Biaka males
were the result of such falls [30].
We note the potential relevance of this hypothesis to the
few non-rainforest small-bodied populations. Specifically,
the Hiwi (Colombia and Venezuela), Hadza (Tanzania) and
!Kung San (Botswana and Namibia) all rely heavily on
tubers and small- to medium-sized animal prey in dense
bush-land portions of their savanna-woodland habitats.
Recent reports of !Kung San persistence hunts (chasing
an animal to its exhaustion) emphasize the importance of
locomotor efficiency amid dense vegetation [31]. It is therefore
conceivable that smaller body sizes conferred a mobility-related
selective advantage in both savannawoodlands
and rainforests [20], even if the habitats and
specific activities invoking selection were quite different.
Life history
By analyzing detailed growth, fertility and mortality data
in a life-history framework, Walker et al. [10] and Migliano
et al. [8] have formulated a simple yet elegant hypothesis to
explain the evolution of the pygmy phenotype. Specifically,
to maximize fitness under conditions of limited lifespan,
small-bodied populations have an early cessation of growth
or a diminished growth spurt to facilitate a relatively early
age of first reproduction. In effect, this hypothesis predicts
an indirect relationship between adult life expectancy and
body size (Figure 3).
Tropical rainforests are challenging habitats for
humans, not only in terms of foraging, mobility and thermoregulation
but also because these environments harbor
many parasites and infectious disease pathogens [32].
Perhaps as a consequence, infant and child mortality rates
among these groups are high relative to those of other
human populations; for example, rates for African rainforest
hunter-gatherers are $20% for infants and 27–40%
for children less than 5 years old, approximately twice the
regional rates [33]. This pattern extends into adulthood:
among six rainforest hunter-gatherer populations from
Africa and Southeast Asia, the chances of surviving to
Figure 2. Thermographic image of Batek hunter-gatherers after a 1 h hike in their
rainforest habitat (Taman Negara National Park, Malaysia). Internal and external
body temperatures can be estimated using a thermographic camera. Analysis of
pixels from the open mouth provides a good estimate of core body temperature.
The open-mouth pixels of the taller man (a) ($164 cm) show an elevated core
temperature of 38.0 8C, whereas those of the shorter man (b) ($156 cm) show a
normal core temperature of 36.9 8C. An estimate of the skin surface temperature
based on facial pixels excluding the open mouth did not differ substantially
between them (mean values of 36.1 8C and 36.4 8C, respectively). Although not
representing a formal test of the thermoregulatory hypothesis as an explanation
for pygmy phenotype evolution, the higher core body temperature of the taller
man following physical activity illustrates the plausibility of its underlying
premise—that in tropical rainforests, where the thermoregulatory mechanism of
sweat production and evaporation is inefficient, smaller bodies might confer a
fitness advantage by generating less internal heat during activity. Images were
captured with a ThermaCAM SC640 (FLIR Systems, Boston, MA, USA) under
photographic license number 395108 issued by the Department of Wildlife and
National Parks, Malaysia.
Review Trends in Ecology and Evolution Vol.24 No.4
221
age 15 are 30–51%, versus 59–76% for one pastoralist and
two non-rainforest hunter-gatherer populations [8]. Once
reaching 15 years of age, ensuing life expectancy estimates
for these groups are 20–32.5 and 41.5–46.6 years, respectively
[8].
Under these circumstances of high adult mortality, any
widening of the otherwise small reproductive window
(thereby increasing the number of opportunities to raise
offspring that reach adulthood) is expected to be adaptive.
Rainforest hunter-gatherers might have maximized their
fitness in this respect with earlier menarche and ages of first
reproduction relative to other populations [8,10,34]. In the
proposed life-history model of pygmy phenotype evolution
[8,10], early maturation is achieved at the expense of body
size. Indeed, growth curves level off for rainforest huntergatherer
females at around age 12, when for other populations
the adolescent growth spurt is just beginning [8,10].
In anextensionof thismodel,Walkerand Hamilton[34] also
consider the effects of population density on human bodysize
variation, because at least within similar environments
(e.g. across tropical rainforests) population density might be
positively correlated with both resource constraint and
mortality. These observations are concordant with recent
life-history theory and research on other organisms [35,36].
Based on their data, Migliano et al. [8] concluded that
early growth cessation is the primary (although not necessarily
exclusive) mechanistic cause of the human pygmy
phenotype. This result seems somewhat discordant with
analyses of detailed growth data from the Efe which
showed increasing stature differences between this population
and neighboring agriculturalists from birth to age 5
[7]. It is possible that the Efe achieve small body size via a
different mechanism from other small-bodied populations
(i.e. a change in growth rate rather than timing). Alternatively,
growth might be both slow in childhood and stop
early in the Efe. Because longitudinal data were not collected
into adulthood for this population [7], such a possibility
cannot yet be evaluated.
Migliano et al. [8] reasoned that life-history factors
alone probably explain the evolutionary origins of the
human pygmy phenotype, to the exclusion of direct selection
for small body size itself under the food limitation-,
thermoregulation- and mobility-associated adaptive
scenarios discussed above (beyond what they might contribute
to high mortality). However, given that early
growth cessation could be a mechanism of small body-size
evolution regardless of the ultimate evolutionary pressures,
it might be too soon to distinguish among these
hypotheses. Although the observation that the pygmy
phenotype is often associated with unusually high adult
mortality [8] is certainly a promising line of evidence in
support of the life-history hypothesis, mortality rates can
be difficult to estimate accurately depending on the underlying
data source [37] and detailed mortality rates have
thus far been examined for only a few populations. It is not
yet clear the extent to which these mortality rates are
representative of those for other rainforest hunter-gatherers,
or even whether they appropriately reflect those
from the same populations in pre-modern times. For
example, one of the populations examined by Migliano
et al. [8], the Agta, have experienced considerable cultural
change in the past 50 years [38] that corresponds to a
marked increase in adult mortality [37]. Moreover, the lifehistory
hypothesis does not explain the relatively small
body sizes of all the savanna-woodland populations mentioned
previously: whereas the Hiwi experience high adult
mortality [39] comparable to that of the small-bodied rainforest
hunter-gatherers studied by Migliano et al. [8], the
Hadza and !Kung San do not [8,39].
It could be beneficial to consider the combined effects of
habitat-specific ecology and high adult mortality on pygmy
phenotype evolution. For example, in open environments,
the locomotor efficiency of taller statures could partly offset
any life-history pressures favoring early growth cessation.
By contrast, the strongest directional selection for the
pygmy phenotype might have been driven by high adult
mortality in habitats absent of such advantages for larger
body sizes or in which small body sizes are even beneficial
(Figure 4). Detailed growth and mortality data from the
Efe might be germane to this issue, given the evidence that
small stature in this population is not entirely due to early
growth cessation [7].
Antiquity of the pygmy phenotype
Although we must be careful not to confuse entirely population
histories with the issue of pygmy phenotype evolution,
insights from one might inform the other. Among
Figure 3. Relationship between adult mortality and body size. The life-history
hypothesis of human pygmy phenotype evolution predicts an indirect relationship
between adult mortality and body size; that is, small stature might be a
morphological consequence of selection for early cessation of growth and early
reproduction to offset limited adult lifespan under conditions of high adult
mortality. The predicted relationship between stature and adult life expectancy
(ensuing lifespan for individuals reaching 15 years of age) is generally observed, at
least for the few small-scale societies for whom relatively high quality mortality
data are available and have been standardized [37]. Male and female stature
estimates are from Ref. [10] (populations are labeled at the male values; female
values are directly below). It is not possible to control for all confounding variables;
for example, poor nutritional availability could negatively affect both stature (via
stunted childhood growth) and adult life expectancy. As a result, the observed
relationship is consistent with but does not prove a selection-driven effect. The
Yanomamo¨ are unexpectedly short (mean adult heights: male = 152 cm;
female = 142 cm) given the reported adult life expectancy (Yanomamo¨ Mucajai,
Brazil, 41.3 years), which could partly reflect variation that exists among different
Yanomamo¨ groups with respect to both stature (e.g. mean adult male statures
range from 147 to 157 cm among Venezuelan Yanomamo¨ communities [80]) and
potentially adult mortality (e.g. adult life expectancy for the Brazilian Yanomamo¨
Xiliana is only 28.3 years, albeit based on more limited data [37]).
Review Trends in Ecology and Evolution Vol.24 No.4
222
African rainforest hunter-gatherers, mtDNA haplotype
profiles of the Mbuti (Democratic Republic of Congo) and
populations from Western Africa (Cameroon, Gabon and
the Central African Republic, including the Biaka) have
little overlap [40,41]. This level of differentiation (although
partly reflecting genetic drift) exceeds that observed between
Western African rainforest hunter-gatherers and
Bantu agriculturalist populations, which itself is thought
to represent a divergence of 30 000–70 000 years and only
limited gene flow [41,42]. The Mbuti and Biaka are also
readily distinguished in analyses of nuclear DNA singlenucleotide
polymorphism genotypes [43,44].
Together, these data suggest a relatively ancient population
divergence between rainforest hunter-gatherers
from the two African regions (Western and more Eastern).
If the pygmy phenotype pre-dates this divergence, then it
could be old, certainly older than the emergence and spread
of agriculture in Africa $5000 years ago [41]. Alternatively,
the introgression of any more recent, highly advantageous,
mutations might have required only limited gene flow
between these groups. Finally, there could have been
multiple, independent origins of the pygmy phenotype
even within Africa.
Ultimately, identification of the specific causative
mutations for the pygmy phenotype (Box 1) will help us
to distinguish among these scenarios, and potentially to
estimate the ages of the mutations themselves. Similar
analyses will also be important for studying the evolutionary
ecology of Southeast Asian rainforest hunter-gatherers
for whom, interestingly, similar issues exist. Specifically,
the mtDNA profiles of Malaysian and Andaman Island
populations do not overlap, and each of these different sets
of haplotypes was estimated to have diverged >50 000
years ago from other populations in Asia [13,45–47].
Paleoanthropological relevance
Knowledge of the adaptive basis of modern human bodysize
variation, including the pygmy phenotype, could
benefit interpretations of the hominin fossil record. Currently,
the evolution of major shifts in body sizes is poorly
understood. Moreover, as discussed earlier, recent
analyses have shown that distinct modern populations
achieve small body sizes by the early cessation of growth;
thus, it now seems clear that life history can evolve surprisingly
rapidly in humans (i.e. within the last 30 000–
50 000 years and perhaps considerably more recently).
Extending this notion to a broader view of human evolution,
we should anticipate the possibility of considerable
life-history diversity among fossil hominins.
Flores
The description in 2004 of $18 000-year-old skeletal
remains of a $106 cm tall hominin from the island of
Flores, Indonesia [3], generated considerable excitement
and debate [5,6,48,49]. In particular, the designation of
Homo floresiensis as a new species that is more closely
related to Homo erectus than to Homo sapiens has been
criticized on the grounds that pathologies could account for
distinctive traits such as a relatively small cranial capacity
[50,51]. Although not all questions of pathology and taxonomy
have been resolved, long bone skeletal material
from a second Flores individual suggests that small stature,
at least, was characteristic of the population, rather
than restricted to the holotype specimen, LB1 [52]. If H.
floresiensis does prove to be distinct from modern humans,
then it would represent another independent origin of the
pygmy phenotype in recent hominin evolution.
Several studies on the taxonomic validity and other
aspects of H. floresiensis have used modern human rainforest
hunter-gatherers as a referential model [4–6,52–55].
However, such a model should be applied with some caution.
Specifically, H. floresiensis (as presently construed) is
characterized by diminutive cranial capacity [53], which
represents a marked departure from the overall phenotype
of modern small-bodied rainforest hunter-gatherers [50].
The extremely small body size of H. floresiensis was
perhaps driven largely by a special case of insular dwarfism
(Box 2), although Jacob et al. [5] emphasize the high
humidity and dense vegetation of Flores. Therefore, the
adaptive advantages of improved thermoregulation and
mobility of small body size that were potentially involved
in the evolution of the modern human pygmy phenotype
should also be considered for H. floresiensis.
Based on a new appreciation of life-history diversity
among modern humans and the apparent association of this
diversity with body size [8], it becomes reasonable to ask
whether the life-history pattern of H. floresiensis was relatively
accelerated compared to H. erectus and other fossil
Figure 4. Integrated model of pygmy phenotype evolution. This model proposes
that the overall strength of selection on small body size in humans might be driven
by the combination of habitat- and adult mortality-level variables. Specifically, the
intensity of selection for small body size might be highest for tropical rainforest
populations that also experience high adult mortality, in which case the pygmy
phenotype might confer fitness benefits related to both ecology (e.g. food
limitations, high temperature and humidity, dense forest undergrowth) and life
history (i.e. early growth cessation could facilitate an early age of first
reproduction). Otherwise, selection coefficients for small body size might be
lower or negligible. For example, the life-history-related benefits of early growth
cessation could be mitigated by the locomotor advantages of larger body sizes in
open environments.
Review Trends in Ecology and Evolution Vol.24 No.4
223
hominins. Such a question could be answered with microscopic
analyses of fossil teeth. Specifically, enamel is
secreted in a circadian manner and contains short- and
long-period incremental features that enable reconstructions
of dental growth rates and molar eruption schedules
[56]. Additionally, the timing of major life events, such as
birth, weaning and illness, manifest as discrete disruptions
in enamel deposition; as a result, the overall pace of dental
growth can be compared to somatic growth schedules [57].
These techniques have been used to describe dental development
rates or life-history profiles for australopithecines,
H. erectus, Neandertals (Homo neanderthalensis) and
anatomically modern humans dated to as early as 160 000
years ago [57–62]. We hope that similar analyses can eventually
be performed on the remains of H. floresiensis.
Palau
Other potentially interesting skeletal remains were
recently recovered from small islands in the Republic of
Palau. These remains, with estimated adult body sizes
similar to, or smaller than, those of modern Onge (Andaman
Islands), have been dated to $1000–3000 years ago
and attributed to H. sapiens [63]. If these body-size estimates
are confirmed (for a critical view, see Ref. [64]), then
it will be important to test whether this Palauan population
was closely related to extant Southeast Asian rainforest
hunter-gatherers and thus was probably small
bodied before arriving on Palau, or whether they independently
evolved small body size by insular dwarfism.
Concluding remarks
Small body size is a characteristic that is shared among
multiple human populations that hunt and gather food in
tropical rainforests. The convergent origins of this phenotype
in Africa, Southeast Asia and South America can
probably be explained as an adaptive response to the
similar challenges posed by rainforest or rainforest-edge
habitats, whether those challenges were related to food
limitation, high temperature and humidity, structural
properties of the forest itself, unusually high adult
mortality or some combination thereof.
Future research on this topic might benefit from the
detailed study of rainforest hunter-gatherer behavioral
ecology and non-body-size physiology. Specifically, these
populations might have numerous cultural and genetic
adaptations to the tropical rainforest that are not yet
described or considered in this framework. By understanding
the evolutionary context of such adaptations, and
thereby isolating features of rainforest habitats that have
strongly affected fitness in the histories of these populations,
we will be better positioned to identify the evolutionary
pressures that favored the pygmy phenotype.
Ultimately, by considering modern human and broader
hominin body-size diversity in the context of life-history,
mortality, habitat and other measured and reconstructed
ecological variables, we stand to shed considerable light on
our evolutionary history.
Acknowledgements
We thank Nurul Fatanah, Zanisha Man and the Batek for their
hospitality, and Luis Barreiro, Geoffrey Benjamin, John Hart, Kim Hill,
Kirk Endicott, Raymond Hames, Tuck-Po Lye, Edith Mirante, Etienne
Patin, Gary Schwartz, Paul Verdu and Robert Walker for helpful
comments and discussions. This work was supported by a grant to
N.J.D. from the Packard Foundation. G.H.P. is supported by NIH
fellowship F32GM085998.
References
1 Blanckenhorn, W.U. (2000) The evolution of body size: what keeps
organisms small? Q. Rev. Biol. 75, 385–407
2 Visscher, P.M. (2008) Sizing up human height variation. Nat. Genet. 40,
489–490
3 Brown, P. et al. (2004) A new small-bodied hominin from the late
Pleistocene of Flores, Indonesia. Nature 431, 1055–1061
4 Argue, D. et al. (2006) Homo floresiensis: microcephalic, pygmoid,
Australopithecus, or Homo? J. Hum. Evol. 51, 360–374
5 Jacob, T. et al. (2006) Pygmoid Australomelanesian Homo sapiens skeletal
remains from Liang Bua, Flores: population affinities and pathological
abnormalities. Proc. Natl. Acad. Sci. U. S. A. 103, 13421–13426
6 Richards, G.D. (2006) Genetic, physiologic and ecogeographic factors
contributing to variation in Homo sapiens: Homo floresiensis
reconsidered. J. Evol. Biol. 19, 1744–1767
7 Bailey, R.C. (1991) The comparative growth of Efe pygmies and African
farmers from birth to age 5 years. Ann. Hum. Biol. 18, 113–120
8 Migliano, A.B. et al. (2007) Life history trade-offs explain the evolution
of human pygmies. Proc. Natl. Acad. Sci. U. S. A. 104, 20216–20219
9 De Souza, R.G. (2006) Body size and growth: the significance of chronic
malnutrition among the Casiguran Agta. Ann. Hum. Biol. 33, 604–619
10 Walker, R. et al. (2006) Growth rates and life histories in twenty-two
small-scale societies. Am. J. Hum. Biol. 18, 295–311
11 Cavalli-Sforza, L.L. (1986) African pygmies: an evaluation of the state
of research. In African Pygmies (Cavalli-Sforza, L.L., ed.), pp. 361–426,
Academic Press
12 Shea, B.T. and Bailey, R.C. (1996) Allometry and adaptation of body
proportions and stature in African pygmies. Am. J. Phys. Anthropol.
100, 311–340
13 Thangaraj, K. et al. (2003) Genetic affinities of the Andaman Islanders,
a vanishing human population. Curr. Biol. 13, 86–93
14 Hart, T.B. and Hart, J.A. (1986) The ecological basis of hunter-gatherer
subsistence in African rain forests: the Mbuti of Eastern Zaire. Hum.
Ecol. 14, 29–55
15 Bailey, R.C. et al. (1989) Hunting and gathering in tropical rain forest:
is it possible? Am. Anthropol. 91, 59–82
16 Endicott, K. and Bellwood, P. (1991) The possibility of independent
foraging intherainforestofpeninsularMalaysia.Hum.Ecol.19,151–185
17 Headland, T.N. and Bailey, R.C. (1991) Introduction: have huntergatherers
ever lived in tropical rain forest independently of
agriculture? Hum. Ecol. 19, 115–122
18 Mercader, J. (2002) Forest people: the role of African rainforests in
human evolution and dispersal. Evol. Anthropol. 11, 117–124
19 Minetti, A.E. et al. (1994) Pygmy locomotion. Eur. J. Appl. Physiol.
Occup. Physiol. 68, 285–290
20 Diamond, J.M. (1991) Why are pygmies small? Nature 354, 111–112
21 Weijers, J.W. et al. (2007) Coupled thermal and hydrological evolution
of tropical Africa over the last deglaciation. Science 315, 1701–1704
22 Bramble, D.M. and Lieberman, D.E. (2004) Endurance running and
the evolution of Homo. Nature 432, 345–352
23 Pontzer, H. (2005) A new model predicting locomotor cost from limb
length via force production. J. Exp. Biol. 208, 1513–1524
24 Carey, T.S. and Crompton, R.H. (2005) The metabolic costs of ‘bent-hip,
bent-knee’ walking in humans. J. Hum. Evol. 48, 25–44
25 Bailey, R.C. (1991) The Behavioral Ecology of Efe Pygmy Men in the
Ituri Forest, Zaire. Museum of Anthropology, University of Michigan
26 Chagnon, N.A. (1997) Yanomamo. Harcourt Brace
27 Elton, S. et al. (1998) Habitual energy expenditure of human climbing
and clambering. Ann. Hum. Biol. 25, 523–531
28 Hanna, J.B. et al. (2008) The energetic cost of climbing in primates.
Science 320, 898
29 Pontzer, H. and Wrangham, R.W. (2004) Climbing and the daily energy
cost of locomotion in wild chimpanzees: implications for hominoid
locomotor evolution. J. Hum. Evol. 46, 317–335
30 Hewlett, B.S. et al. (1986) Causes of death among Aka pygmies of the
Central African Republic. In African Pygmies (Cavalli-Sforza, L.L.,
ed.), pp. 45–63, Academic Press
Review Trends in Ecology and Evolution Vol.24 No.4
224
31 Liebenberg, L. (2006) Persistence hunting by modern huntergatherers.
Curr. Anthropol. 47, 1017–1026
32 Guernier, V. et al. (2004) Ecology drives the worldwide distribution of
human diseases. PLoS Biol. 2, e141
33 Ohenjo, N. et al. (2006) Health of indigenous people in Africa. Lancet
367, 1937–1946
34 Walker, R.S. and Hamilton, M.J. (2008) Life-history consequences of
density dependence and the evolution of human body size. Curr.
Anthropol. 49, 115–122
35 Stearns, S.C. et al. (2000) Experimental evolution of aging, growth, and
reproduction in fruitflies. Proc. Natl. Acad. Sci. U. S. A. 97, 3309–3313
36 Reznick, D. et al. (2001) Life history evolution in guppies. VII. The
comparative ecology of high- and low-predation environments. Am.
Nat. 157, 126–140
37 Gurven, M. and Kaplan, H. (2007) Longevity among hunter-gatherers:
a cross-cultural examination. Popul. Dev. Rev. 33, 321–365
38 Early, J.D. and Headland, T.N. (1998) Population Dynamics of a
Phillipine Rain Forest People: The San Ildefonso Agta. University
Press of Florida
39 Hill, K. et al. (2007) High adult mortality among Hiwi hunter-gatherers:
implications for human evolution. J. Hum. Evol. 52, 443–454
40 Destro-Bisol, G. et al. (2004) The analysis of variation of mtDNA
hypervariable region 1 suggests that eastern and western pygmies
diverged before the Bantu expansion. Am. Nat. 163, 212–226
41 Quintana-Murci, L. et al. (2008) Maternal traces of deep common
ancestry and asymmetric gene flow between pygmy huntergatherers
and Bantu-speaking farmers. Proc. Natl. Acad. Sci. U. S.
A. 105, 1596–1601
42 Batini, C. et al. (2007) Phylogeography of the human mitochondrial L1c
haplogroup: genetic signatures of the prehistory of Central Africa. Mol.
Phylogenet. Evol. 43, 635–644
43 Jakobsson, M. et al. (2008) Genotype, haplotype and copy-number
variation in worldwide human populations. Nature 451, 998–1003
44 Li, J.Z. et al. (2008) Worldwide human relationships inferred from
genome-wide patterns of variation. Science 319, 1100–1104
45 Macaulay, V. et al. (2005) Single, rapid coastal settlement of Asia
revealed by analysis of complete mitochondrial genomes. Science
308, 1034–1036
46 Thangaraj, K. et al. (2005) Reconstructing the origin of Andaman
Islanders. Science 308, 996
47 Hill, C. et al. (2006) Phylogeography and ethnogenesis of aboriginal
Southeast Asians. Mol. Biol. Evol. 23, 2480–2491
48 Diamond, J. (2004) The astonishing micropygmies. Science 306, 2047–
2048
49 Tocheri, M.W. et al. (2007) The primitive wrist of Homo floresiensis and
its implications for hominin evolution. Science 317, 1743–1745
50 Martin, R.D. et al. (2006) Flores hominid: new species or microcephalic
dwarf? Anat. Rec. A Discov. Mol. Cell. Evol. Biol. 288, 1123–1145
51 Niven, J.E. (2007) Brains, islands and evolution: breaking all the rules.
Trends Ecol. Evol. 22, 57–59
52 Morwood, M.J. et al. (2005) Further evidence for small-bodied hominins
from the Late Pleistocene of Flores. Indonesia. Nature 437, 1012–1017
53 Falk, D. et al. (2005) The brain of LB1, Homo floresiensis. Science 308,
242–245
54 Larson, S.G. et al. (2007) Homo floresiensis and the evolution of the
hominin shoulder. J. Hum. Evol. 53, 718–731
55 Gordon, A.D. et al. (2008) The Homo floresiensis cranium (LB1): size,
scaling, and early Homo affinities. Proc. Natl. Acad. Sci. U. S. A. 105,
4650–4655
56 Smith, T.M. (2008) Incremental dental development: methods and
applications in hominoid evolutionary studies. J. Hum. Evol. 54,
205–224
57 Dean, M.C. (2006) Tooth microstructure tracks the pace of human lifehistory
evolution. Proc. R. Soc. Lond. B Biol. Sci. 273, 2799–2808
58 Macchiarelli, R. et al. (2006) How Neanderthal molar teeth grew.
Nature 444, 748–751
59 Smith, T.M. et al. (2007) Earliest evidence of modern human life history
in North African early Homo sapiens. Proc. Natl. Acad. Sci. U. S. A. 104,
6128–6133
60 Smith, T.M. et al. (2007) Rapid dental development in a Middle
Paleolithic Belgian Neanderthal. Proc. Natl. Acad. Sci. U. S. A. 104,
20220–20225
61 Guatelli-Steinberg, D. and Reid, D.J. (2008) What molars contribute to
an emerging understanding of lateral enamel formation in
Neandertals vs. modern humans. J. Hum. Evol. 54, 236–250
62 Robson, S.L. and Wood, B. (2008) Hominin life history: reconstruction
and evolution. J. Anat. 212, 394–425
63 Berger, L.R. et al. (2008) Small-bodied humans from Palau. Micronesia.
PLoS ONE 3, e1780
64 Fitzpatrick, S.M. et al. (2008) Small scattered fragments do not a dwarf
make: biological and archaeological data indicate that prehistoric
inhabitants of Palau were normal sized. PLoS ONE 3, e3015
65 Ishida, T. et al. (1998) Preliminary report on the short stature of
Southeast Asian forest dwellers, the Manni, in southern Thailand:
lack of an adolescent spurt in plasma IGF-I concentration. Southeast
Asian J. Trop. Med. Public Health 29, 62–65
66 Jain, S. et al. (1998) Insulin-like growth factor-I resistance. Endocr.
Rev. 19, 625–646
67 Clavano-Harding, A.B. et al. (1999) Initial characterization of the GHIGF
axis and nutritional status of the Ati Negritos of the Philippines.
Clin. Endocrinol. (Oxf.) 51, 741–747
68 Davila, N. et al. (2002) Growth hormone binding protein, insulin-like
growth factor-I and short stature in two pygmy populations from the
Philippines. J. Pediatr. Endocrinol. Metab. 15, 269–276
69 Steiner, C.C. et al. (2009) The genetic basis of phenotypic convergence
in beach mice: similar pigment patterns but different genes. Mol. Biol.
Evol. 26, 35–45
70 Wittkopp, P.J. et al. (2003) Drosophila pigmentation evolution:
divergent genotypes underlying convergent phenotypes. Proc. Natl.
Acad. Sci. U. S. A. 100, 1808–1813
71 Dietz, W.H. et al. (1989) Nutritional status of Efe pygmies and Lese
horticulturists. Am. J. Phys. Anthropol. 78, 509–518
72 Lomolino, M.V. (2005) Body size evolution in insular vertebrates:
generality of the island rule. J. Biogeogr. 32, 1683–1699
73 Bromham, L. and Cardillo, M. (2007) Primates follow the ‘island rule’:
implications for interpreting Homo floresiensis. Biol. Lett. 3, 398–400
74 Meiri, S. et al. (2008) The island rule: made to be broken? Proc. Biol. Sci.
275, 141–148
75 Raia, P. and Meiri, S. (2006) The island rule in large mammals:
paleontology meets ecology. Evolution Int. J. Org. Evolution 60,
1731–1742
76 Palkovacs, E.P. (2003) Explaining adaptive shifts in body size on
islands: a life history approach. Oikos 103, 37–44
77 Burness, G.P. et al. (2001) Dinosaurs, dragons, and dwarfs: the
evolution of maximal body size. Proc. Natl. Acad. Sci. U. S. A. 98,
14518–14523
78 Schebesta, P. and Lebzelter, V. (1928) Anthropological measurements
in Semangs and Sakais in Malaya (Malacca). Anthropologie 6, 183–251
79 Pandey, A.K. (2004) Anthropometry of male Onges of Little Andaman.
South Asian Anthropol. 4, 135–140
80 Hames, R. and Kuzara, J. (2004) The nexus of Yanomamo growth,
health, and demography. In Lost Paradises and the Ethics of Research
and Publication (Salzano, F.M. and Hurtado, A.M., eds), pp. 110–145,
Oxford University Press
Review Trends in Ecology and Evolution Vol.24 No.4
225