Annu. Rev. Anthropol. 2004. 33:585­623
doi: 10.1146/annurev.anthro.33.070203.143955
Copyright c 2004 by Annual Reviews. All rights reserved
First published online as a Review in Advance on June 21, 2004

THE EVOLUTION OF HUMAN SKIN AND
SKIN COLOR

Nina G. Jablonski
Department of Anthropology, California Academy of Sciences, San Francisco,
California 98103; email: njablonski@calacademy.org

Key Words pigmentation, melanin, UV radiation, thermoregulation, race

Abstract

Humans skin is the most visible aspect of the human phenotype. It is
distinguished mainly by its naked appearance, greatly enhanced abilities to dissipate
body heat through sweating, and the great range of genetically determined skin colors
present within a single species. Many aspects of the evolution of human skin and skin
color can be reconstructed using comparative anatomy, physiology, and genomics.
Enhancement of thermal sweating was a key innovation in human evolution that allowed
maintenance of homeostasis (including constant brain temperature) during sustained
physicalactivityinhotenvironments. Dark skin evolved paripassu with the loss of body
hair and was the original state for the genus Homo. Melanin pigmentation is adaptive
and has been maintained by natural selection. Because of its evolutionary lability, skin
color phenotype is useless as a unique marker of genetic identity. In recent prehistory,
humans became adept at protecting themselves from the environment through clothing
and shelter, thus reducing the scope for the action of natural selection on human skin.
INTRODUCTION
When humans visualize a body, they see mostly skin. The skin is the body's
direct interface with the physical environment, conveying a state of health and
personal identity. The skin comprises a sheet-like investiture that protects the
body from attack by physical, chemical, and microbial agents. It is the organ that
regulates body temperature through control of surface blood flow and sweating and
detects critical information about the ambient environment and objects touched.
The largest and most massive of the organs of the body, the skin of the average
adult human exceeds 2 m2
yet is generally no thicker than 2 mm (Odland 1991).
The skin also provides a forum for advertising. It provides information about a
person's age, health, and some aspects of ancestry, and furnishes a placard upon
which further information is placed through temporary and permanent decoration.
Research on the evolution of human skin and skin color has not been com-
mensurate with the importance of skin in human evolution. Skin is generally not
preserved in the fossil record and so details of its evolution can be gained only from
0084-6570/04/1021-0585$14.00 585
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
586 JABLONSKI
comparative anatomical and physiological evidence. Skin has also been overlooked
as a topic of research interest in anthropology and human biology in recent decades
because of the social sensitivity surrounding discussions of skin color and because
of the use and misuse of skin color in biological and social concepts of race.
The goal of this review is to provide a comprehensive yet economical survey of
the biology, evolution, and culture of human skin and skin color, with an emphasis
on new research--especially on the evolution of skin color. The review begins
with an overview of the basic biology of skin itself, followed by discussions of the
evolution of skin and skin color, and of skin color and race.
THE STRUCTURE AND FUNCTIONS OF HUMAN SKIN
The skin serves as an effective physical barrier because its laminar structure ren-
ders it relatively resistant to abrasion, puncture, and percutaneous absorption, and
because its immune cells mount a first line of defense against pathogens coming
in contact with the body. Lacking adequate protection from hair, human skin has
undergone numerous adaptive structural changes that give it strength, resilience,
and sensitivity (Montagna 1981). The skin of humans, like that of all tetrapods,
acts as a sun shield to protect the body from most solar UV radiation (UVR) and
is the locus for the initiation of the important, UVR-driven process of vitamin D
production in the body.
Epidermis
The laminar structure of human skin comprises two major tissue layers, a thinner
outer layer, the epidermis, and a thicker and more internally complex inner layer,
the dermis (Figure 1). The epidermis is a stratified keratinizing epithelium with
a smooth, abrasion-resistant surface that is interrupted only by hair follicles and
the pores of sweat glands. The barrier properties of the skin are predicated on the
integrity of the stratum corneum (Elias et al. 2003, Taylor 2002). Keratinocytes are
the principal cell type found in epidermis and are composed largely of filamentous
proteins known as keratins, which are imbedded in an amorphous matrix. The
skin's elasticity and resistance to physical and chemical attack can be attributed
to the high elastic modulus and unique amino acid composition of the keratinized
layer of the epidermis (Marks 1991, Odland 1991). The epidermis also contains
populations of three types of immigrant dendritic cells: melanocytes, Langerhans
cells, and Merkel cells. Melanocytes produce the skin's primary pigment, melanin,
and are discussed in greater detail below. Langerhans cells are specialized cells
of the immune system that present and respond to antigens coming in contact
with the skin, and Merkel cells are associated with nerve terminals that together
function as slow-adapting mechanoreceptors for touch; they are most common on
the glabrous skin of the fingertips (Chu et al. 2003, Kripke & Applegate 1991,
Lynn 1991, Odland 1991). The epidermis is subdivided into four layers from
deep to superficial: the stratum basale (the germinative layer of keratinocytes), the
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
SKIN AND SKIN COLOR 587
Figure 1 Schematic rendering of a cross-section of human skin, showing its
laminar structure, main cell types, and appendages.
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
588 JABLONSKI
stratum spinosum, the stratum granulosum, and the stratum corneum. The stratum
corneum consists of flattened, nonviable keratinocytes. In darkly pigmented or
heavily tanned individuals, these keratinocytes contain specks of melanin "dust"
(Kollias 1995a). The stratum corneum acts as a barrier to the unrestrained passage
of water and solutes through the skin, defends against invasion by microorganisms
and the penetration of toxic substances, and protects against most mechanical
injury caused by friction, abrasion, pricks, or arthropod bites (Marks 1991). These
functions are successfully served despite the epidermis being in a constant state of
turnover, as the outermost cornified cells of the stratum corneum are shed as they
are replaced from below.
Differences between human groups in epidermal structure and thickness have
been reported, but most studies of this topic have been based on small samples
with poorly controlled experimental designs, as reviewed elsewhere (Taylor 2002).
Considerable variation in epidermal thickness exists within human populations and
is likely related to age and history of sun exposure. The stratum corneum of darkly
pigmented or heavily tanned people is more compact and consists of more cornified
cell layers than that of lightly pigmented people; these characteristics enhance the
barrier protection functions of the skin (Taylor 2002).
In all primates, the epidermis of the volar surfaces of the hands and feet exhibit
well-developed epidermal ridges or dermatoglyphics, which impart greater resis-
tance against friction and help to insure secure purchase on locomotor substrates
and on objects being gripped or manipulated. Dermatoglyphics are also found on
the ventral surfaces of the tails of prehensile-tailed New World monkeys and on
the knuckle pads of chimpanzees and gorillas (Ellis & Montagna 1962, Montagna
1971).
The melanocytes of the epidermis warrant close attention because of their
role in the production of the skin's primary pigment or chromophore, melanin.
Melanocytes are specialized dendritic cells that reside in the stratum basale of
the epidermis and in the matrix portion of the hair bulb. They originate in the
neural crest as melanoblasts proliferate and migrate to the epidermis during the
eighteenth week of embryonic development (Rawles 1948). Melanocytes produce
melanins in specialized cytoplasmic organelles called melanosomes, which vary in
size and degree of aggregation depending on skin type and pigmentation (Figure 2)
(Szabo et al. 1969). The density of melanocytes varies over the surface of the body,
and the number of active (melanin-producing) melanocytes varies with age and
can be increased by exposure to UVR (Halaban et al. 2003, Jimbow et al. 1991,
Quevedo et al. 1975). The total number of melanocytes is relatively invariant from
one person to another, however, and is not related to variation between human
groups in skin pigmentation (Fitzpatrick et al. 1961, Jimbow et al. 1991, Robins
1991, Young & Sheehan 2001). MacKintosh, following Wasserman, has recently
advanced the hypothesis that melanocytes, melanosomes, and melanin together
functionaspartoftheimmunesystemagainstinvadingmicroorganismsandthatthe
more darkly pigmented skins of the indigenous peoples of the tropics have evolved
primarily to serve this function (MacKintosh 2001; Wassermann 1965b, 1974).
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
SKIN AND SKIN COLOR 589
Figure 2 Schematic rendering of cross-sections of lightly and darkly pigmented
human skin, showing differences in stratum corneum structure and in the size and
aggregation of melanin-containing melanosomes.
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
590 JABLONSKI
Melanocytes project their dendrites into keratinocytes where they then trans-
fer mature melanosomes (Figure 2). Melanosomes are ellipsoidal, membrane-
bound organelles containing melanin. After melanosomes have been transferred
to keratinocytes, they become aggregated and surrounded by a membrane in a
melanosome complex (Jimbow et al. 1991, Szabo et al. 1969). In darkly pig-
mented skin, melanosomes are large and are not clumped in aggregations, whereas
in lightly pigmented skin these organelles are smaller and aggregated (Szabo et al.
1969). Intensity of skin coloration is determined by many factors: (a) The total
number of melanosomes in the keratinocytes and melanocytes, and their degree
of dispersion; (b) the rate of melanin production (melanogenesis); (c) the degree
of melanization of melanosomes; (e) the rate of transport and type of incorpo-
ration of melanosomes into keratinocytes; (f) the degradation of melanosomes
within the keratinocytes; and (g) a person's chronological age because the number
of metabolically active melanocytes decreases over time (Halaban et al. 2003,
Jimbow et al. 1976, Ortonne 1990, Parker 1981). Larger melanosomes break
down more slowly in keratinocytes and contribute to higher levels of pigmentation
(Sulaimon & Kitchell 2003).
MELANIN PIGMENTATION AND ITS MEASUREMENT Human skin derives most of
itspigmentationfrommelanin,anextremelydense,virtuallyinsoluble,highmolec-
ular weight polymer that is attached to a structural protein (Jimbow et al. 1991,
Ortonne 2002, Parker 1981, Sulaimon & Kitchell 2003). Human skin contains the
two types of melanin found in all mammals, the brownish-black eumelanin and
the reddish-yellow pheomelanin (Thody et al. 1991). Higher concentrations of
eumelanin characterize darker skin phenotypes including tanned skin. Concentra-
tions of pheomelanin in the skin vary considerably from individual to individual
within any given human group, but pheomelanin-rich skin phenotypes are more
common among red-haired northern Europeans, as well as East Asians and Na-
tive Americans (Rana et al. 1999, Thody et al. 1991). Melanin is synthesized by
oxidation of tyrosine via the enzyme tyrosinase (Fitzpatrick et al. 1950, Jimbow
et al. 1976, Ortonne 2002). Eumelanins and pheomelanins arise from a common
metabolic pathway in which dopaquinone is the key intermediate (Ortonne 2002).
As is discussed in greater detail below, production of melanins is regulated by
pigmentation genes, hormones, and UVR (Fitzpatrick & Ortonne 2003, Sulaimon
& Kitchell 2003, Thody & Smith 1977). A balance of many regulatory factors is
essential for normal pigment production in the melanocyte, and derangements of
these factors can lead to anomalies of cutaneous pigmentation such as albinism
piebald spotting, and various types of hyperpigmentation (Robins 1991, Sulaimon
& Kitchell 2003, Thody & Smith 1977).
The optical and chemical properties of melanins have been studied in detail
(Ito 2003, Kollias et al. 1991, Ortonne 2002, Prota 1992c), but detailed chemical
characterization of the compounds has been difficult to obtain because melanin
polymers are composed of many different units connected through strong carbon-
carbon bonds (Ito 2003). The optical properties of natural melanin in vivo are
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
SKIN AND SKIN COLOR 591
related to its abilities to absorb, scatter, and reflect light of different wavelengths
(Kollias et al. 1991, Ortonne 2002). The melanins in human skin are a hetero-
geneous mixture of melanin polymers, precursors, and metabolites, character-
ized by a continuous absorption capacity in the UV range and exponentially
declining absorption capacity from the UV to the visible range (Kollias 1995b,
Sarna & Swartz 1998). Natural protection against sunburning (photoprotection)
is due to the absorption and scattering of UVR by melanin (Kaidbey et al. 1979;
Kollias 1995a,b). Both processes are influenced by the density and distribution
of melanosomes within keratinocytes (Figure 2), with the larger, singly dispersed
and heavily melanized melanosomes of darkly pigmented skin absorbing more
energy than the smaller, less dense, and lightly melanized melanosomes of lightly
pigmented skin (Kaidbey et al. 1979).
Melanin was long considered to act as a passive screening filter against UVR,
but it is by no means inert (Fitzpatrick et al. 1961). Photodegradation (photolysis)
and/or oxidative polymerization of melanin may occur when it absorbs photons
(Ortonne 2002). Recent evidence indicates that the photoprotective role of melanin
in darkly pigmented skin may be augmented by its ability to scavenge oxygen-
derived radicals (reactive oxygen species), such as superoxide anion and hydrogen
peroxide, which are cytotoxic compounds generated by the interaction of UV pho-
tons with membrane lipids and other cellular components (Ortonne 2002, Prota
1992c, Sulaimon & Kitchell 2003, Young & Sheehan 2001). At the physiological
level, the protective role of melanin pigmentation against UVR exposure derives
from its ability to prevent direct and indirect (oxidative) damage to DNA at wave-
lengths where it is most vulnerable (Cleaver & Crowley 2002, Kielbassa et al.
1997, Shea & Parrish 1991).
Melanin pigmentation in human skin is considered as either constitutive skin
color or facultative skin color (Quevedo et al. 1975). Constitutive skin color is
the amount of genetically determined cutaneous melanin pigmentation that is gen-
erated without any influence of solar radiation (Jimbow et al. 1976, Quevedo
et al. 1975). Facultative skin color or "tan" constitutes the short-lived, immedi-
ate, and delayed tanning reactions elicited by exposure to UVR (Jimbow et al.
1991, 1976; Quevedo et al. 1975). Lighter constitutive pigmentation is associ-
ated with a higher sunburn response, a lower tanning response, and a greater
susceptibility to skin cancers (Kollias et al. 1991, Sturm 2002, Wagner et al.
2002).
Objective and reproducible assessment of melanin pigmentation has long been
a goal of anthropology and dermatology. In anthropology, verbal descriptions of
skin colors ("white," "yellow," "black," "brown," and "red") were replaced by
color-matching methods during the early twentieth century (Olivier 1960, von
Luschan 1897). The most popular of these methods was the von Luschan scale,
based on the use of colored tablets or tiles of different colors and hues with which
the colors of unexposed skin were matched. These and similar matching meth-
ods could not be consistently reproduced, however, and were swiftly abandoned
when reflectance spectrophotometry was introduced in the early 1950s (Lasker
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
592 JABLONSKI
1954, Wassermann 1974). Reflectance spectrophotometry remains the method of
choice for the objective study of skin pigmentation, color definition, and the spec-
tral reflectance curves of skin because the incident light used and the distance
between the light source and the subject are invariable and because subjective
factors inherent in the visual matching methods are excluded (Wassermann 1974).
All instrumental approaches to skin color evaluation depend on the illumination
of the skin site by a standard light source at a fixed relative angle that minimizes
the reflected light from the stratum corneum. The detector collects light re-emitted
by the skin site from a particular angle and with a chosen color filter (Kollias
1995a). Because of the importance of assessing constitutive skin color on a part
of the body that is not routinely exposed to sun, the inner (medial) surface of
the upper arm has long been the standard reference site for studies of skin color.
Portable reflectance spectrophotometers came into use with Weiner's (1951) study,
with two types of instruments being commonly employed in anthropology dur-
ing the latter part of the twentieth century. The instrument manufactured by the
Evans Electroselenium Company (EEL) has been the most widely used, espe-
cially in studies of the skin colors of Old World peoples (Wassermann 1974),
whereas that made by the Photovolt Corporation was more widely used in studies
of New World peoples. Unfortunately, the skin reflectance measurements obtained
by these two instruments are not directly comparable, requiring conversion for-
mulae to make them so (Lees & Byard 1978). Research is now underway that may
make possible the conversion of skin color assessments made by von Luschan
color tablets to values comparable with those derived from reflectance spectropho-
tometry (M. Henneberg, personal communication).
In clinical medicine, constitutive skin color and skin sensitivity has been clas-
sified commonly according to skin phototypes or sun-reactive skin types, from
Type I (very sensitive, easily burned, with little or no potential for tanning)
to Type VI (insensitive, never burns, and deeply pigmented) (Fitzpatrick 1988,
Fitzpatrick & Ortonne 2003, Jimbow et al. 1991). Skin type does not correspond
well to constitutive skin color, however, and has limited applicability with respect
to the responses of moderately or deeply pigmented skin (Kollias et al. 1991, Prota
1992c, Taylor 2002, Wagner et al. 2002, Westerhof et al. 1990). Despite these
limitations, skin phototyping has been widely embraced by many clinicians be-
cause assessments can be made without instrumentation. In recent years, highly
sensitive diffuse reflectance spectrophotometers such as the DermaSpectrometer
and the Datacolor International Microflash as well as chromaticity meters have
been used increasingly to measure skin pigmentation and skin response to UVR
(Kollias 1995a, Wagner et al. 2002).
The photoprotective benefits of melanin have been assessed using several dif-
ferent measures including minimal erythemal dose (MED), DNA damage, and
incidence of skin cancer (Kollias et al. 1991). The MED represents the minimum
amount of UVR necessary to bring about a slight visible reddening of lightly pig-
mented skin. It is the easiest and most common method of assessing skin reactions
to UVR but is difficult to determine for deeply pigmented individuals in whom
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
SKIN AND SKIN COLOR 593
visual redness is difficult to assess (Kaidbey et al. 1979, Ortonne 2002, Shono et al.
1985).
Exposure of human skin to UVR results in a profound alteration of the meta-
bolism, structure, and function of epidermal cells. These activities include in-
creased activation of melanocytes, augmentation of melanosome production, an
increase in the size of melanosome complexes incorporated within keratinocytes,
and initiation of vitamin D synthesis (Parker 1981, Prota 1992a, Urbach 2001).
The erythema response or sunburn reaction is related to constitutive skin color:
Dark-skinned individuals can tolerate longer sun exposure than light-skinned indi-
viduals can. The skin of individuals with dark constitutive pigmentation exhibits a
sun protection factor (SPF) of 10­15, whereas that of moderately pigmented peo-
ple (e.g., from the circum-Mediterranean) achieves an SPF of only 2.5 (Kaidbey
et al. 1979, Kollias et al. 1991, Ortonne 2002). In vitro studies of the reactions of
human melanocytes to UVR have shown that heavily pigmented melanocytes have
a greater capacity to resume cell division after irradiation with short wavelength
UVR (UVB) than do their lightly pigmented counterparts, which suggests that they
suffered less damage to their DNA (Barker et al. 1995). In contrast, UVB damages
the immune system of the skin regardless of constitutive pigmentation by depleting
both heavily and lightly pigmented skin of Langerhans cells (Cleaver & Crowley
2002, Kripke & Applegate 1991). The protective role of melanin in connection with
skincancerthusderivesfromitsroleinpreventingdamagetoDNAinthefirstplace,
not in protecting against damage to the cutaneous immune system (Vermeer et al.
1991). Tanning or facultative pigmentation induced by UVR is photoprotective to
some degree against the deleterious effects of further UVR exposure, but it does not
significantly increase the SPF of individuals with light constitutive pigmentation
or protect the DNA of their skin from UVR-induced damage (Kaidbey et al. 1979,
Ortonne 2002). Although repeated exposure of tanned skin to UVR increases the
number of metabolically active melanocytes and the intensity of melanogenesis
(Lock-Anderson et al. 1998), the increased concentration of melanin in the tanned
skin of inherently lightly pigmented people does not approach the photoprotection
conferred by natural melanin in intrinsically darker-skinned people (Kaidbey et al.
1979). Individuals with lightly to moderately pigmented skin, who are repeatedly
exposed to UVR, experience premature aging (photoaging) of the skin, which is
characterized by wrinkling and anomalies of pigmentation (Chung 2001, Fisher
et al. 2002, Kollias et al. 1991). This process is initiated by the photochemical
generation of reactive oxygen species causing degradation of structural proteins
in the dermis that confer strength and resiliency to the skin (Fisher et al. 2002).
Dermis
The dermis is a thick, dense fibroelastic connective tissue composed of collagen
fibers, elastic fibers, and an interfibrillar gel composed of glycosaminoglycans,
salts, and water. The primary cells of the dermis are collagen-rich fibroblasts.
Collagen, which constitutes 77% of the fat-free dry weight of skin, largely accounts
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
594 JABLONSKI
for the tensile strength of the skin's fabric and for some of the ability of the dermis
to scatter visible light (Kollias 1995a, Shea & Parrish 1991). Interwoven with the
collagen is a network of abundant elastic fibers that restore the skin to its normal
configuration after stretching. The dermis is equally thick in people with dark or
light constitutive pigmentation (Taylor 2002).
The dermis encloses a widely ramifying network of blood vessels, an extensive
nerve network, sweat glands, and a pilosebaceous complex of hair follicles and
sebaceous glands (Figure 1). Of these, only the sweat glands are addressed in detail
in this review because of their importance in thermoregulation.
The rich vascular supply of the skin is responsible for supplying the needs of the
sweat glands, hair follicles, and rapidly multiplying epidermal cells in the stratum
basale. The density of cutaneous blood vessels varies throughout the body's surface
and is related to temperature and blood pressure regulation and the relative amounts
of intermittent physical pressure different parts of the body must withstand, with
the highest concentrations found in the skin covering the head, nipples, palms,
soles, and ischial tuberosities (Edwards & Duntley 1939). The perineal skin of
female macaques, baboons, and chimpanzees is richly suffused with blood vessels
(Montagna 1967, Montagna 1971) that create large sexual swellings advertising
the female's state of reproductive receptivity and lifetime reproductive potential
(Domb & Pagel 2001). The oxygenated and deoxygenated forms of hemoglobin
carried in the skin's blood vessels are some of the skin's main pigments, with
a person's skin color determined mainly by the skin's melanin and hemoglobin
content (Edwards & Duntley 1939). The erythema or strongly red appearance of
the skin caused by exposure to UVR is the result of increases in the number and
diameter of vascular capillaries through which blood is flowing and an increase in
the blood flow through each capillary (Kollias 1995a). Sunburned skin feels hot to
the touch because of the increased vascularization of the skin and the inflammatory
response mounted by the skin as it works to repair UVR-induced damage (Ryan
1991, Shea & Parrish 1991).
The nerve supply of the skin is highly complex because the skin is a ma-
jor sensory surface that contains varied types of receptors sending signals to the
central nervous system about the external environment and the internal state of
the skin (Chu et al. 2003, Lynn 1991). These receptors include two types of
temperature sensors, diverse mechanoreceptors associated with both hairy and
glabrous skin, and an important group of cutaneous sensory cells (nociceptive
afferents) specialized for the detection of tissue-threatening stimuli or the pres-
ence of injury or inflammation (Lynn 1991). The glabrous skin of the hands and
feet of primates is densely packed with sensory nerve endings that permit highly
sensitive tactile discrimination and exquisite differentiation of temperature and
texture (Chu et al. 2003, Lynn 1991, Martin 1990). These attributes greatly en-
hance the manipulative functions of these appendages, especially the hand (Martin
1990).
Numerous hairs, which grow from hair follicles located in the dermis, are as-
sociated with mechanoreceptors and sebaceous glands. Hair performs a range of
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
SKIN AND SKIN COLOR 595
functions from insulation, to protection against the sun, enhancement of cutaneous
sensation to communication of emotion (through piloerection), and ornamenta-
tion (Lavker et al. 2003; Montagna 1967, 1971; Wheeler 1984, 1985). Humans
are unique among primates in possessing effectively naked skin, except on the
scalp, the male chin, the axilla, and the groin. Although human skin bears mil-
lions of hairs, most of them are so small as to be nearly invisible (Montagna
1981).
SWEAT GLANDS Human dermis contains two main types of sweat glands, eccrine
and apocrine. The former are widely distributed throughout the surface of the body,
whereas the latter are concentrated in the axilla, perineum, and external auditory
canal. Eccrine glands are tubular in form (Figure 1) and lie in the outer portion of
the dermis. They produce copious amounts of dilute, watery fluid expressed to the
surface of the skin through an individual pore. Humans have two to four million
eccrine glands on the surface of their bodies, with an average distribution ranging
from 150­340/cm2
(Folk & Semken 1991, Goldsmith 2003). Both apocrine and
eccrine sweat glands are stimulated by the sympathetic division of the autonomic
nervous system and produce sweat in response to thermal stimulation (thermal
sweating). In contrast, the eccrine glands of the palms and soles respond only to
emotional stimuli, whereas those of the face and axilla respond to both (Folk &
Semken 1991, Zihlman & Cohn 1988).
Considerable attention has been placed on comparisons of the quantity, struc-
ture, and function of sweat glands between human groups. The number of strictly
controlled comparisons between members of different populations after equivalent
periods of deliberate acclimatization is quite small (Weiner 1977). The results of
most rigorous comparative study of sweat gland densities in humans (Knip 1977)
indicate that only small differences in the total number and average density of sweat
glands exist between disparate human populations. As yet it has proven virtually
impossible to design studies that can determine conclusively whether differences
in sweating performance between human groups are due to genetic influences or
environmental adaptations.
The Skin in Thermoregulation
Dissipation of heat is the function that most conspicuously distinguishes human
skin from that of all other animals (Montagna 1981). The reasons for the evolution
of this unique capacity are discussed in the following section. Humans encounter
heat stress more or less year round in equatorial areas and for varying lengths
of time in the rest of the world except for circumpolar and alpine environments.
Heat stress is exacerbated by prolonged or rigorous exercise. Maintenance of
homeostasis requires that the body's core temperature remain close to a neutral
point, which varies from about 36.8 to 37.2
C, in order to permit uninterrupted
functioning of the temperature-sensitive cells of the human central nervous system.
If the rates of production or loss of heat are excessively out of balance, core
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
596 JABLONSKI
temperature can quickly increase or decrease to dangerous levels (Kraning 1991,
Wenger 2003).
Temperature regulation in humans includes involuntary (physiologic) and vol-
untary (behavioral) activity (Wenger 2003). Voluntary temperature regulation in-
volves the conscious actions taken by people to maintain thermal comfort, includ-
ing the seeking of shade and shelter, and the wearing or shedding of clothing.
Involuntary temperature regulation in the skin has been studied in great detail in
the past 50 years by both physiologists and anthropologists, and only a superficial
summary of this corpus of work is presented here. Regulation of temperature by the
skin is accomplished through its roles in (a) perceiving and transmitting its own
temperature to the central nervous system; (b) regulating heat transfer between
the body's core and the skin through the cutaneous circulation; (c) serving as a
superficial casing through which body heat is conducted from the vascular layers
to the surface; (d) acting as an interface for the loss or gain of heat to or from the
environment by radiation, convection, or conduction; and (e) acting as a surface for
the spreading of sweat necessary for evaporative cooling (Frisancho 1981, Kraning
1991). The relative role of the four avenues of heat loss (radiation, convection,
conduction,andevaporation)dependsontheinteractionoftheambienttemperature
and humidity (Chaplin et al. 1994; Frisancho 1981; Wenger 2003; Wheeler 1984,
1991b). The ability of sweat glands to respond to heat stress is adversely affected
by sunburn (Pandolf et al. 1992). Protection of the integrity of sweat glands against
damage caused by UVR, therefore, has been of great importance during the long
course of human habitation of the tropics.
Experimental studies and simulations undertaken to determine how thermal
homeostasis is maintained under the stressful environmental conditions of the
tropics have shown that heat loss is maximized in people with a high ratio of
skin surface area to body weight, such as Nilotic tribespeople, the Kung San, and
Australian Aborigines (Frisancho 1981; Wheeler 1991a,b, 1992). This relationship
supports Allen's Rule in mammals, which states that mammals living in cold
regions will minimize the size and surface area of their extremities, whereas those
inhabiting hot areas will increase the relative size of appendages.
The Role of the Skin in Vitamin D Biosynthesis
Synthesis of vitamin D in the skin of vertebrates is the only unanimously agreed
positive effect of UVR exposure. Vitamin D3 is the form of vitamin D that is syn-
thesized by vertebrates, whereas vitamin D2 is the primary form found in plants
(Holick 2003). Vitamin D3 is more accurately characterized as a prosteroid hor-
mone than as a vitamin because, in mammals, it is derived from a cholesterol-like
precursor (7-dehydrocholesterol) found in the skin (Holick 2003). Vitamin D is a
unique natural product thought to have first occurred on Earth as a photosynthetic
product in marine phytoplankton more than 750 mya (Holick 1995). Although the
physiological role of vitamin D in plants and invertebrates is not clear, vitamin
D was essential for the evolution of terrestrial vertebrates (Holick 1991, 1995).
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
SKIN AND SKIN COLOR 597
Holick has reasoned that early tetrapods depended on vitamin D for the efficient use
of scarce dietary calcium to preserve their rigid calcified skeletons (Holick 1995).
Vitamin D can be synthesized only by a photochemical process, so early tetrapods
could only satisfy their body's vitamin D requirements by exposing themselves
to sunlight to photosynthesize vitamin D in their own skin or by ingesting foods
containing vitamin D (Holick 1995).
Vitamin D3 synthesized in the skin requires successive hydroxylations in the
liver and kidney to be converted to its biologically active form, 1, 25-dihydroxy-
vitamin D3 (Holick 1991, Jones et al. 1998). This functionally active form is
important for the regulation of calcium and phosphorus metabolism, skeletal de-
velopment and mineralization, the regulation of normal cell growth, and the inhi-
bition of cancer cell growth (Holick 1991, 2001). The production of vitamin D3
is optimally stimulated by UVR wavelengths of 295­300 nm, in the UVB range
(MacLaughlin et al. 1982). High-energy UVB photons penetrate the skin and are
absorbed by the 7-dehydrocholesterol in the keratinocytes of the epidermis (espe-
cially of the strata basale and spinosum) and fibroblasts of the dermis, catalyzing
the formation of previtamin D3 (Holick 2001, Webb et al. 1988). Once formed in
the skin, previtamin D3 can undergo isomerization to vitamin D3 at body temper-
ature and then undergo further chemical conversions to 1, 25-dihydroxyvitamin
D3. The conversion of previtamin D3 or vitamin D3 to the functionally active
form is rate-limited, however. In the presence of biologically sufficient amounts
of 1, 25-dihydroxyvitamin D3 in the circulation, previtamin D3 and vitamin D3
are transformed by UVA or UVB into a variety of inert byproducts, thus averting
overproduction of the biologically active form and subsequent "vitamin D intox-
ication" (Holick 2001, Holick et al. 1981). This finding disproves the hypothesis
that dark constitutive skin pigmentation evolved in the tropics as an adaptation to
protect against the overproduction of 1, 25-dihydroxyvitamin D3 (Loomis 1967).
Melanin pigments are highly effective at absorbing and scattering the UVB
wavelengths that catalyze vitamin D3 synthesis. Thus, high concentrations of
melanin in the skin result in a decrease in the efficiency of conversion of 7-
dehydrocholesterol to previtamin D3; pigmentation slows but does not prevent
cutaneous production of the vitamin (Holick et al. 1981, Webb et al. 1988). In-
dividuals with very deep constitutive pigmentation often require 10 to 20 times
longer exposure to sunlight than those of lighter pigmentation in order to promote
an adequate synthesis of vitamin D3 (Holick et al. 1981). This finding explains
why dark-skinned individuals living at high latitudes with low levels of envi-
ronmental UVB are at greater risk of vitamin D3­deficiency diseases than are
light-skinned people (Clemens et al. 1982, Holick 2001, Mitra & Bell 1997). The
evolutionary significance of this observation is discussed further below. The pho-
toconversion of 7-dehydrocholesterol to previtamin D3 in the skin is also adversely
affected by increasing age (Holick 1995), the wearing of clothing (Matsuoka et al.
1992), and by the use of topical sunscreens, which block the UVB wavelengths
responsible for both sunburn and vitamin D3 production (Holick 1997, Webb et al.
1988).
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
598 JABLONSKI
THE EVOLUTION OF MODERN HUMAN SKIN
Reconstruction of the evolution of human skin relies on evidence provided by
comparative anatomy and physiology, as well as study of the evolution of the
genes and gene complexes that determine the function and pigmentation of skin.
Using basic principles of historical morphology, one can reconstruct the major
steps in the evolution of human skin by utilizing a well-established phylogeny to
examine historical transformations of structure and function (Jablonski & Chaplin
2000). This method leads to the reconstruction of the probable appearance of the
skin in the last common ancestor of the human and chimpanzee lineages as being
lightly pigmented and covered with dark hair, like most catarrhine primates today
(Jablonski & Chaplin 2000).
The skin of modern humans is distinguished from that of other primates mainly
by its naked appearance, its greatly enhanced abilities to dissipate body heat
through sweating, and by the great range of genetically determined skin colors
present within a single species. Most investigators have considered these attributes
to be adaptations forged by natural selection.
The Evolution of the Thermoregulatory Properties
of Human Skin
Human skin is not hairless, but--as discussed above--the hairs over most of the
body's surface are so fine and present at a sufficiently low density that the skin
appears essentially naked. Explanations for the evolution of human hairlessness
have been many, varied, and often highly creative. The most cogent explanations
are based on the importance of a functionally naked skin in maintaining body
temperature in hot environments.
Many animals, including primates, which live in hot environments, have heavy
coats of insulating fur or feathers. In the heat caused by strong sunlight, such
insulation reduces environmental heat gain (Folk & Semken 1991, Walsberg 1988).
This is the case even for black coats, which absorb short-wave radiation near or at
the surface of the fur and reradiate large amounts of long-wave radiation before it
reaches the skin (Dmi'el et al. 1980). The effectiveness of fur insulation in reducing
environmental heat gain is lessened by sweating. The most efficient evaporative
cooling occurs at the skin's surface; in heavily furred animals, water vapor is
transferred through the fur to the atmosphere (Folk & Semken 1991). If the fur is
wet from sweating, however, maximum evaporation occurs at the surface of the
fur, and heat from the blood vessels cannot be transferred as efficiently to the site of
evaporation (Folk & Semken 1991). Under these circumstances, much more water
must be used for evaporative cooling. Thermal sweating as a method of cooling
becomes more important as environmental temperatures rise or as activity levels
increase because the lower gradient between core and environmental temperatures
restricts the amount of heat loss that can be achieved by radiation, convection,
and conduction (Frisancho 1981, Wheeler 1991b). Removal of excess heat is,
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
SKIN AND SKIN COLOR 599
therefore, greatly facilitated by the loss of body hair because it increases thermal
conductance and permits additional heat loss through sweating (Wheeler 1985,
Zihlman & Cohn 1988).
A strong case can be made for the evolutionary loss of apocrine sweat glands
in humans because these sweat glands are most common in heavily furred animals
(Folk & Semken 1991). The African apes exhibit a ratio of approximately 40%
apocrine sweat glands to 60% eccrine; the great preponderance of eccrine sweat
glands in modern humans probably evolved under the strong influence of natu-
ral selection, following the loss of the apocrine component to sweating (Folk &
Semken 1991, Montagna 1981, Zihlman & Cohn 1988). This process was proba-
bly propelled by increases in body size and activity levels associated with modern
limb proportions and striding bipedalism, which occurred in the transition from the
primitive hominins of the late Miocene to the genus Homo of the Plio-Pleistocene
(Chaplin et al. 1994; Folk & Semken 1991; Jablonski & Chaplin 2000; Montagna
1981; Schwartz & Rosenblum 1981; Wheeler 1984, 1996).
The importance of body cooling through the skin in modern humans has been
emphasized repeatedly by both physiologists and anthropologists because of the
primacy of preventing hyperthermia and attendant damage to the central nervous
system (Cabanac & Caputa 1979, Falk 1990, Wheeler 1984, Zihlman & Cohn
1988). The temperature of the brain closely follows arterial temperature, requir-
ing that the temperature of the circulating blood be carefully regulated (Nelson &
Nunneley 1998). This process became increasingly important as activity levels and
brain size increased in the genus Homo through the Pleistocene. Simulations and
experimental studies have confirmed that maintenance of stable core temperature
under conditions of increased environmental heat load or exercise is best accom-
plished via recruitment of a whole-body cooling system, involving cooling of
peripheral blood vessels through sweating (Desruelle & Candas 2000, Nelson &
Nunneley 1998). A recently mooted hypothesis that human hairlessness evolved
late in human evolution as a result of the adoption of clothing and the need to reduce
the load of external parasites (Pagel & Bodmer 2003) finds no support in light of
the overwhelming evidence of the importance of hairlessness in thermal sweating
and whole-body cooling in maintaining stable core temperature and homeostasis.
The Evolution of Human Skin Pigmentation
RECONSTRUCTION OF SKIN COLOR IN EARLY HOMO The early members of the
genus Homo from the late Pliocene and Early Pleistocene of Africa exhibited
larger bodies, relatively larger brains, and relative longer lower limbs than did
their australopithecine predecessors (McHenry & Berger 1998; Ruff et al. 1993,
1997). The higher activity levels and larger day ranges reconstructed for them
(Wheeler 1991a, 1992) would have required that their skin be functionally naked
and endowed with a high density of eccrine sweat glands in order to facilitate
heat loss (Jablonski & Chaplin 2000, Wheeler 1984). This situation created a new
physiological challenge for human skin: protection of a naked integument against
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
600 JABLONSKI
UVR. Dense hairy coats protect the skin of mammals from UVR-induced damage
to the skin because the hairs themselves absorb or reflect most short-wavelength
solar radiation. In mammals with sparse coats of hair, however, 3%­5% of inci-
dent UVR is transmitted to the skin (Walsberg 1988). Nonhuman mammals that
are active in hot, sunny environments exhibit sparse coats because they facilitate
passive heat loss; they also display highly melanized skin on their exposed (dorsal)
surfaces to effectively block the UVR transmitted to the skin (Walsberg 1988). This
evidence clearly indicates that hair loss in the human lineage was coupled with
increased melanization of the skin as activity levels in hot environments increased.
The early members of the genus Homo, the ancestral stock from which all later
humans evolved, were, thus, darkly pigmented (Jablonski & Chaplin 2000). This
interpretation has recently been supported by genetic evidence demonstrating that
strong levels of natural selection acted about 1.2 mya to produce darkly pigmented
skin in early members of the genus Homo (Rogers et al. 2004).
Heavily pigmented skin does not, in fact, perceptibly increase the body's heat
load under conditions of intense solar radiation (Baker 1958, Walsberg 1988).
This is because for half of the solar radiation reaching the Earth's surface--in the
infrared--there is essentially no difference in absorption between dark and light
skin (Baker 1958, Daniels 1964). This evidence negates the claim by Blum (1961)
and others (Morison 1985) that heavily melanized pigmentation in humans could
not be adaptive in the hot tropics because of the increased heat load caused by
greater amounts of absorbed solar radiation.
SKIN PIGMENTATION IN MODERN HUMAN POPULATIONS Many of the accounts of
travelers and explorers from the fifteenth century onward include reports of the
skin color of the peoples they encountered. As natural historians and human
geographers--mostly from Europe--ventured into Asia, Africa, Australia, and the
Americas and began to study the indigenous human populations in detail, maps
depicting the worldwide distribution of human skin color were slowly assembled.
The best known of these maps is that composed by the Italian geographer Renato
Biasutti, which was based on the von Luschan skin color scale. This map has
gained broad circulation in several widely distributed publications (Barsh 2003,
Lewontin 1995, Roberts 1977, Walter 1971), despite the fact that, for areas with
no data, Biasutti simply filled in the map by extrapolation from findings obtained
in other areas (Robins 1991). A more accurate and exhaustive compilation of the
skin colors of indigenous peoples based only on published skin reflectance mea-
surements is now available (Jablonski & Chaplin 2000). Both maps show similar
trends, with darkly pigmented peoples found near the Equator and incrementally
lighter ones found closer to the Poles. A larger percentage of people with dark skin
is found in the Southern Hemisphere as compared with the Northern Hemisphere
(Relethford 1997) because of a latitudinal bias in the distribution of land masses
(Chaplin & Jablonski 1998).
The data compiled by Jablonski and Chaplin also provide conclusive evidence
of sexual dimorphism previously observed in human skin pigmentation (Frost
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
SKIN AND SKIN COLOR 601
1988, van den Berghe & Frost 1986), with females being consistently lighter than
males in all populations studied (Jablonski & Chaplin 2000).
One of the major problems encountered in assembling data on the distribu-
tion of human skin color in indigenous populations is determining exactly what
an indigenous population represents. For most anthropologists and human geog-
raphers, an arbitrary cutoff date of 1500 has been adopted to distinguish native
or indigenous peoples from immigrant populations. This date is reasonable with
respect to the inauguration of the modern era of European colonization but fails to
recognize the several major movements of human groups within continents (such
as the so-called Bantu expansion within Africa) that occurred before 1500. These
movements, along with European colonization and the increasingly rapid and dis-
tant migrations of human populations through time, have fundamentally altered
the human landscape established in prehistoric times. This has made the interpre-
tation of geographically and biologically significant trends in human populations
much more difficult.
ENVIRONMENTAL CORRELATES OF HUMAN SKIN COLOR Theskinpigmentationof
indigenous human populations shows remarkable regularity in its geographic dis-
tribution. Darker skins occur in more tropical regions and lighter skins in tem-
perate, although the gradient is less intense in the New World as compared to
the Old World. Even within Africa, the continent with the largest equatorial land
mass, there is considerable heterogeneity of skin color, with the deepest colors
occurring not in the lowest latitudes but in the open grasslands (Chaplin 2001,
Roberts 1977). The strong latitudinal signal in skin color led most early work-
ers to conclude that skin pigmentation represented an adaptation to sunlight or
other solar-driven phenomena such as temperature. Walter (1958, 1971) was the
first researcher to suggest that the pigmentation gradient observed was linked to
the intensity of UVR, and he established this relationship by calculation of cor-
relation coefficients between skin color (as measured on the von Luschan scale)
and estimated UVR. The relationship between skin color and environment was
further explored by studies in which the relationship of skin color, as measured
by reflectance spectrometry, to latitude, temperature, and humidity was studied
by correlation and regression analyses (Roberts 1977, Roberts & Kahlon 1976).
These analyses showed the dominant association of skin reflectance with latitude,
which was then deduced to be an effect related to the intensity of UVR (Roberts
1977, Roberts & Kahlon 1976).
In recent years, studies of the relationships between morphological and physio-
logical variation and attributes of the physical environment have been advanced by
the availability of remotely sensed data on levels of UVR, total solar radiation, tem-
perature, humidity, precipitation, and other environmental variables at the Earth's
surface. These data, which were not widely available to workers before 1990,
have permitted correlation, regression, and other analyses of skin reflectance to be
conducted against actual measurements, rather than estimates, of environmental
variables (Chaplin 2001, 2004; Jablonski & Chaplin 2000).
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
602 JABLONSKI
Using data on the minimal erythemal dose of UVR (UVMED) at the Earth's
surface collected by the NASA TOMS 7 satellite, Jablonski & Chaplin were able to
establish a conclusive correlation between latitude and annual average UVMED,
and thence between annual average UVMED and skin reflectance (Jablonski &
Chaplin 2000). This publication was followed by a study in which the influence of
minimum, maximum, and seasonal levels of UVR, as well as other directly mea-
sured environmental variables, relative to skin reflectance were studied (Chaplin
2001, Chaplin 2004). This study showed that skin reflectance was correlated with
autumn levels of UVMED, and that skin reflectance could be almost fully mod-
eled as a linear effect of this variable alone (Chaplin 2001, 2004). This study
also showed that the relationship between summer levels of UVMED and skin
reflectance appeared to reach a threshold past which higher levels of UVR were
not correlated with incrementally lower skin reflectance (darker pigmentation)
(Chaplin 2001, 2004).
Low reflectance values for human skin (dark pigmentation) are primarily a func-
tion of UVMED (Jablonski & Chaplin 2000), with regression analysis demonstrat-
ing that autumn UVMED levels have the strongest effect. This indicates that skin
color is more strongly correlated with UVA, which is consistently higher through-
out the year at all latitudes, than with UVB (Chaplin 2001, 2004). Maximum
UVMED had the next most significant effect (Chaplin 2001, 2004). Winter lev-
els of precipitation have the opposite effect, being positively correlated with high
reflectance values (light pigmentation) (Chaplin 2001, 2004). Multiple regression
formulae relating skin reflectance to these environmental parameters can then be
used to derive a map of predicted human skin colors, with the colors shown being
realistic approximations of the true color of skin (Chaplin 2001, 2004) (Figure 3).
This map depicts an idealized situation in which humans worldwide are assumed
to have inhabited their respective regions for the same lengths of time, and have
followed similar cultural practices that could affect skin color (e.g., diet, activity
schedules, use of clothing and shelter).
NATURAL SELECTION AND THE EVOLUTION OF HUMAN SKIN PIGMENTATION The
geographical distribution of human skin colors has invited many explanations,
most of which have claimed melanin pigmentation to be an adaptation to some
attribute of the physical environment that varies primarily by latitude. Ever since
the harmful effects of UVR began to be appreciated by scientists, explanations for
the evolution of deeply melanized skin have centered on the importance of resis-
tance to sunburn, solar degeneration, and skin cancer (Daniels et al. 1972). Equally
popular has been the vitamin D hypothesis, which stated that lightly pigmented
skins were necessary outside of the tropics in order to permit vitamin D biosyn-
thesis in the skin by low levels of UVR, whereas darkly pigmented skin afforded
protection against production of toxic doses of vitamin D in equatorial regions
(Loomis 1967). Lightly pigmented skin has also been explained as an adaptation
to resist cold injury, on the basis of experimental and epidemiological data that
have documented more severe injuries incurred by pigmented skin exposed to
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
SKIN AND SKIN COLOR 603
freezing conditions (Post et al. 1975, Steegmann 1967). Other explanations have
imputed highly melanized skin as providing effective concealment in habitats such
as tropical forests with differing light intensities and environmental illumination
(Cowles 1959, Morison 1985), and still others have reasoned that tropical diseases
and parasites rather than tropical climate were the major selective forces lead-
ing to the evolution of differential pigmentation in humans (MacKintosh 2001;
Wassermann 1981, 1965a).
Although adaptive explanations for human pigmentation have dominated the
literature, others have downplayed or discounted the role of adaptation by natural
selection. Some workers have emphasized the role of sexual selection, especially
by way of explaining the lighter constitutive pigmentation of females relative to
males (Aoki 2002, Frost 1988). Deol claimed that differences in skin color between
human populations were the pleiotropic byproducts of natural selection on other
functions of pigmentation genes (Deol 1975). Others have simply discouraged
the "amusing pastime" of adaptive reconstruction in the absence of data on the
differential survival and reproduction of varying skin pigmentation phenotypes
(Blum 1961, Lewontin 1995). Adaptive explanations "for" any given phenotypic
trait require demonstration that the trait increases the real or probable reproductive
success of the organism. Although such evidence is often difficult to muster in
the case of traits borne by long-lived mammals, it is incumbent that adaptive
reconstructions be tethered by this responsibility.
In the past, adaptive explanations for different levels of melanin pigmentation
in human skin have suffered from an inability to demonstrate probable or real
differences in survivorship and reproduction of different skin color phenotypes
under the same environmental conditions. Blum introduced this mode of critical
appraisal of competing hypotheses when he drew attention to the fact that dark skin
pigmentation could not have evolved primarily as adaptive protection against skin
cancer because such cancers rarely cause death during peak reproductive years
(Blum 1961, Jablonski & Chaplin 2000). Other adaptive explanations for light
or dark skin pigmentation (e.g., protection against cold injury; camouflage) have
similarly failed to demonstrate real or probable increases in reproductive success
as a result of possession of these phenotypes.
MELANIN AS A REGULATOR OF THE PENETRATION OF UVR INTO THE SKIN Recen-
tly, Jablonski & Chaplin (2000) published a new adaptive hypothesis for the evo-
lution of human skin pigmentation stating that melanin pigmentation evolved to
regulate the penetration of UVR into the skin in order to prevent the photolysis of
photo-labile compounds while permitting the photosynthesis of others. This hy-
pothesis was based on two equally important observations: (a) that the B vitamin
folate is destroyed by long wavelength UVR (UVA), and that folate deficiencies
can markedly reduce individual reproductive success by adversely affecting cell
division; and (b) that vitamin D3 is synthesized in the skin by short wavelength
UVR (UVB) and that severe vitamin D deficiencies adversely affect reproduc-
tive success by interfering with normal calcium metabolism (Jablonski & Chaplin
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
604 JABLONSKI
2000). Natural selection has produced two opposing clines of skin pigmentation.
The first is a cline of photoprotection that grades from darkly pigmented skin at
the Equator to lightly pigmented skin near the Poles. The second is a cline of
vitamin D3 photosynthesis that grades from lightly pigmented near the Poles to
darkly pigmented at the Equator. In the middle of the two clines we find peoples
with enhanced abilities to develop facultative pigmentation according to seasonal
UVR levels.
THE FOLATE HYPOTHESIS The potential importance of dark skin pigmentation in
protecting folate from UVR-induced photolysis was first recognized upon discov-
ery that folate undergoes photolysis in vitro when subjected to UVA (360 nm) and
that serum folate levels of human subjects dramatically declined when humans
underwent long-term exposure (minimum of 3 months) to the same wavelength,
for 30­60 min once or twice a week (Branda & Eaton 1978). The potential sig-
nificance of the finding to the evolution of human skin pigmentation was echoed
later, but a causal mechanism was not mooted (Zihlman & Cohn 1988).
Few nutrients compare with folate (folic acid) for its impact on health. Adequate
folate status is vital for the synthesis, repair, and expression of DNA, and therefore
for all processes involved in cell division and homeostasis (Kesavan et al. 2003,
Lucock et al. 2003, Suh et al. 2001). The subtle influence of folate on the cell's
genomic machinery has led to the realization that even marginal folate deficien-
cies may have significance in developmental disorders and degenerative diseases
associated with high morbidity and mortality (Lucock et al. 2003). Now that folate
deficiency is widely acknowledged as a risk factor for neural tube defects, recurrent
early pregnancy loss, and other complications of pregnancy, the maintenance of
adequate folate status in women of reproductive age has become a primary public
health concern (Bower & Stanley 1989, Fleming & Copp 1998, Minns 1996, Suh
et al. 2001). Folate's importance in spermatogenesis also highlights its important
role in maintaining male reproductive competence (Cosentino et al. 1990, Mathur
et al. 1977).
The recognition of folate's pivotal roles in DNA synthesis and repair--and thus
most processes associated with reproductive success in both sexes--has underlined
the importance of protecting the body's folate stores from physical or chemical
degradation. Because folate is susceptible to oxidative damage as a result of expo-
sure to UVR and ionizing radiation (Branda & Eaton 1978, Hirakawa et al. 2002,
Kesavan et al. 2003), the primary evolutionary function of melanin in regions re-
ceiving high annual UVR is to protect folate from photodegradation (Jablonski
& Chaplin 2000). Photolysis of folate has been experimentally demonstrated at
340 nm and 312 nm, in the UVA and near-UVA wavelengths (Hirakawa et al. 2002,
Lucock et al. 2003). With skin reflectance being most closely correlated with au-
tumn levels of UVMED dominated by UVA, one can conclude that the longer
wavelengths of UVR, which are capable of penetrating deep into the dermis of
the skin, have been the most important agents of natural selection in connection
with the evolution of skin pigmentation (Chaplin 2001) (Figure 4). The results of
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
SKIN AND SKIN COLOR 605
Figure 4 The effects of UV radiation on the skin. Different wavelengths of UVR
penetrate to different thicknesses in the skin, with UVA penetrating more deeply than
UVB. UVC generally does not penetrate the Earth's atmosphere.
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
606 JABLONSKI
a recent study (Gambichler et al. 2001) did not confirm the photolytic effect of
UVA on serum folate levels in a small number of human volunteers. This finding
runs counter to the results of previous in vivo and in vitro studies demonstrating
profound photodegradation of folate upon exposure to UVR (Hirakawa et al. 2002,
Lucock et al. 2003) and to X- and -irradiation (Kesavan et al. 2003). A true and
statistically robust test of the folate hypothesis would require a case-control study
involving a large number of human volunteers experiencing long-term (once or
twice a week for a minimum of three months) exposure to UVR, with measurement
of more labile folate species such as specific red cell folate coenzymes (Lucock
et al. 2003).
SKIN PIGMENTATION AND VITAMIN D BIOSYNTHESIS In the millennia prior to
about 1.6 mya, the earliest members of the genus Homo appear to have been
restricted in their distribution to the high-UVR regimes of equatorial Africa. Un-
der these environmental conditions, possession of highly melanized skin was
critical for survival. As populations of early Homo moved both northward and
southward, they began to experience different schedules and intensities of UVR
exposure.
UVR levels at the Earth's surface are affected by latitude, altitude, season,
moisture content, cloud cover, the depth of the ozone column, orbital parameters,
and other factors (Hitchcock 2001, Madronich et al. 1998). Short wavelength
UVR (UVB, 280­315 nm) is more effectively absorbed by atmospheric ozone
than are longer wavelengths (UVA, 315­400). Thus, as one moves away from the
Equator and the angle of solar elevation decreases, the thickness of the atmosphere
(including the ozone layer), through which sunlight must pass, increases. This
results in a greater attenuation of UVR, especially of UVB, by scattering and
absorption by ozone, and consequently very low levels of UVB in high-latitude
ecosystems (Caldwell et al. 1998). Very small increments or decrements of UVB
lead to substantial biological effects (Madronich et al. 1998); thus, it is highly
biologically significant that regions north and south of 50
latitude receive only
tiny doses of UVB, and only then at the peak of summer (Caldwell et al. 1998,
Chaplin 2001, Johnson et al. 1976, Neer 1985).
As discussed earlier, deeply melanized skin confers excellent protection against
the deleterious effects of UVR, but it also greatly slows the process of vitamin
D3 synthesis in the skin. As hominins moved out of the tropics, their exposure
to UVR--especially to vitamin D­inducing UVB--was dramatically reduced.
Levels of UVR at the Earth's surface are not thought to have been appreciably
different in the Pleistocene as compared to today because similar conditions of
solar emissivity and orbital parameters existed at the time, and similar levels of
UVR have been reconstructed from biological proxies (Rozema et al. 2002). Even
before remotely sensed data on UVB levels outside of the tropics were available,
theorists surmised that early humans living in high latitudes with deeply pigmented
skin would have not been able to produce sufficient vitamin D3 in their skin to meet
their physiological demands and that strong selective pressure for depigmentation
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
SKIN AND SKIN COLOR 607
of the skin had been exerted in order to facilitate photosynthesis of vitamin D3
(Loomis 1967, Murray 1934, Neer 1975).
Using known values of UVMED at the Earth's surface (Herman & Celarier
1996) and the precise dosage of UVB necessary to catalyze vitamin D synthesis in
human skin at a specific latitude (Webb et al. 1988), researchers can calculate the
worldwide potential for vitamin D3 synthesis for lightly pigmented skin (Jablonski
& Chaplin 2000) (Figure 5). Zone 1 (shown without hachure in Figure 5) corre-
sponds closely to the tropics and represents an area in which there is adequate
UVR throughout the year to catalyze vitamin D3 synthesis in the skin (Jablonski
& Chaplin 2000). Zone 2 (area covered by vertical hachure in Figure 5) represents
the region in which there is insufficient UVR during at least one month of the year
to produce vitamin D3, and Zone 3 (cross-hatched area of Figure 5) represents that
in which there is insufficient UVR, as averaged over the entire year, to photosyn-
thesize vitamin D3 in the skin (Jablonski & Chaplin 2000). The configuration of
vitamin D synthesis zones for darkly pigmented skin differs markedly from this de-
piction, with Zone 1 being greatly reduced in area, and Zones 2 and 3 significantly
expanded because of the attenuation of UVB absorption by dark melanin pig-
mentation and concomitant prolongation of the length of UVB exposure required
for vitamin D3 biosynthesis (Jablonski & Chaplin 2000). This analysis clearly
demonstrates the profound impact of constitutive pigmentation on the potential
for vitamin D3 synthesis in the skin. An empirical demonstration of this was re-
cently provided by a school population of darkly pigmented and albino children in
South Africa, in which the former group of children required a significantly higher
dietary intake of vitamin D3 to attain the same levels of vitamin D3 and plasma
calcium than did the albinos (Cornish et al. 2000). The importance of the synthesis
and physiological activity of vitamin D have been further born out by studies of
the worldwide polymorphism in the vitamin D­binding protein (or group-specific
component, Gc) that show a clear relationship between the frequency of specific
Gc alleles and levels of sunlight (OMIM 2003).
Vitamin D3 insufficiency and deficiency can exert sinister effects on the body
throughout life and have the demonstrated potential to reduce fitness when they
afflict children and adolescents. The most serious and notorious of the vitamin
D3­deficiency diseases is rickets, caused by a failure of mineralization in the
cartilaginous matrix of developing bones as a result of calcium and phosphate
malabsorption (Shaw 2003, Wharton & Bishop 2003). Comprehensive clinical de-
scriptions of rickets (Bereket 2003, Holick 1995, Shaw 2003, Wharton & Bishop
2003) catalog the devastating osseous and nonosseous effects of the disease on
children and adolescents, including the delayed closure of fontanelles, bowing of
the lower limb bones, and narrowing of the pelvic outlet in females, which can
cause obstructed labor and a high incidence of infant and maternal morbidity and
mortality. Vitamin D3 deficiency in adults produce osteomalacia, a softening of
the bone matrix, but inadequate vitamin D3 status in pregnant women contributes
to hypocalcemia and rickets in their babies (Wharton & Bishop 2003). The dele-
terious effects of vitamin D3 deficiency encompass a suite of problems affecting
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
608 JABLONSKI
evolutionary fitness, including those involving the formation and maintenance of
the skeleton, control of normal cell growth, inhibition of cancerous cell growth,
and maintenance of normal immune system function (Grant 2002; Holick 1991,
2001; Wharton & Bishop 2003). An important, but little reported consequence of
vitamin D deficiency in laboratory mice and rats is a marked reduction in female
fertility and female reproductive failure apparently due to failure of vitamin D to
interact normally with its receptor on the ovary (Jones et al. 1998).
An abundance of clinical and epidemiological evidence now supports the ar-
gument that depigmentation of the skin evolved in humans living outside of the
tropics because of the importance of maintaining adequate vitamin D3 production
in the skin for as long as possible throughout the year. Alterations in the function
of the vitamin D endocrine system in darkly pigmented people as a consequence of
diminished exposure to sunlight result in vitamin D3 insufficiency and deficiency,
as recently reviewed elsewhere (Mitra & Bell 1997, Wharton & Bishop 2003).
These problems potentially afflict dark-skinned people who have migrated to or
who inhabit UVB-poor regions (e.g., northern Europe, the northern United States,
or Canada) or darkly pigmented people living in sunny regions who habitually stay
indoors or consistently wear concealing clothing when outdoors (Atiq et al. 1998,
Bereket 2003, Brunvand & Haug 1993, Fogelman et al. 1995, Fonseca et al. 1984,
Gessner et al. 1997, Hodgkin et al. 1973, Holick 1995, Wharton & Bishop 2003).
In these populations, vitamin D3 deficiency is exacerbated by breast feeding be-
cause of the low concentration of vitamin D3 in human breast milk (Gessner et al.
1997, Shaw 2003, Wharton & Bishop 2003).
Vitamin D3 insufficiency and deficiency also afflict lightly pigmented people
who are not exposed to sufficient sunlight because of occupation, advanced age, or
hospitalization, or people who consistently wear protective clothing or sunscreen
when outdoors (Holick 1995, 1997, 2001; Thomas et al. 1998). Rickets (known
to many as the English disease) was, in fact, first recognized as a disease of light-
skinned children living in dark, multistoried structures devoid of sunlight (Holick
1991, 1995).
Brace (1963) argues that depigmentation of human skin occurred not as the
result of active selection for lighter pigmentation, but because of the relief of
selective pressure on pigmentary systems as humans populated increasingly high
latitudes where dark pigmentation was no longer required as a shield against
UVR. This structural reduction hypothesis is based on the "probable mutation
effect" whereby mutations in the genes controlling melanin pigmentation would
accumulate, leading to reduction of or failure to produce melanin (Brace 1963).
A recent variation on this argument states that where natural selection for dark
skin is sufficiently weak, a sexual preference for lighter skin could have driven the
evolution of light skin (Aoki 2002, Ihara & Aoki 1999) (see below also). These
arguments find limited support today with respect to the evolution of human skin
coloration in light of the impressive body of recent clinical evidence cited above
that attests to the many and highly significant functions of vitamin D3 in humans,
which directly impact human health and reproductive competence.
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
SKIN AND SKIN COLOR 609
Strong natural selection for vitamin D3 production in human skin was likely a
powerful factor influencing the evolution of skin pigmentation in human popula-
tions at high latitudes. Preliminary study of the distribution of paleontological and
archaeological sites for the genus Homo in relationship to the vitamin D3 synthe-
sis zones described above indicates that year-round hominin habitation of Zone
3, i.e., latitudes generally higher than 50
, occurred only after human populations
had developed the technological competence to harvest fish, marine mammals, or
other sources of food [such as reindeer lichen, or reindeer meat, organs, or milk
(Bjorn & Wang 2000)] rich in vitamin D3 (N. Jablonski, G. Chaplin & D. Tyler,
manuscript in preparation). This capability is associated almost primarily with
Upper Paleolithic peoples, living approximately 15,000­10,000 years ago, who
are known to have made extensive use of fish hooks, fish traps, nets, harpoons, and
other implements for the harvesting of marine animals.
SEXUAL DIMORPHISM IN HUMAN SKIN COLOR The observation that females exhi-
bit lighter skin pigmentation than do males in all populations examined (Jablonski
& Chaplin 2000, van den Berghe & Frost 1986) has invited speculation that the
phenomenon may be due to infantile mimicry, sexual selection, or a combination
of both factors (Aoki 2002, Frost 1988, Ihara & Aoki 1999, van den Berghe &
Frost 1986). These hypotheses are based on the observations that the attraction of
human infants and human females is partly due to their lighter pigmentation, and
that lighter-colored adult females are perceived as more feminine than are darker
females, and therefore are preferred as partners (Frost 1988). Jablonski & Chaplin
(2000) have advanced the idea that sexual dimorphism in skin pigmentation is
primarily due to natural selection, on the basis of the need of females to maximize
cutaneous vitamin D3 production in order to meet their absolutely higher calcium
requirements of pregnancy and lactation. Also, darker pigmentation may have been
the object of natural selection in males because of the importance of maintaining
optimal levels of folate in order to safeguard sperm production, a process depen-
dent on folate for DNA synthesis (G. Chaplin, personal communication). Sexual
selection is thus considered to have played a role in increasing the disparity in
skin color between the sexes in some societies through preference for more lightly
pigmented females, but this was not its ultimate cause (Jablonski & Chaplin 2000).
TANNING AND BLEACHING Thetemporarydevelopmentofincreasedmelaninpig-
mentation through exposure to UVR is called facultative pigmentation or tanning.
Individuals with very light constitutive pigmentation (skin phototypes I and II)
never tan or tan minimally, whereas those with moderate to dark constitutive pig-
mentation (phototypes V and VI) tan profusely (Taylor 2002). Considerable vari-
ation in tanning potential exists even between people with ostensibly very similar
levels of constitutive pigmentation (Lee & Lasker 1959). Tanning develops in two
stages (immediate and delayed) over the course of several hours or days, depending
on the wavelength and duration of UVR exposure (Ortonne 1990). Exposure to
UVA causes tanning to develop quickly (Ortonne 1990), possibly as an adaptation
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
610 JABLONSKI
to protect against photodegradation of essential biomolecules. Facultative pig-
mentation is probably most important in areas such as the circum-Mediterranean
that receive low levels of UVB but receive moderate levels of UVA that cause
photodegradation of folate, DNA, and vitamin D3.
The practice of recreational tanning has been eschewed by health care workers
in the past 20 years because of the explosion in skin cancer rates due to increased
UVR exposure. A tanned skin is still viewed by many as fashionable or as a sign
of well-being, however, and this positive image has spurred the development of a
simulated tanning industry in Europe, the Americas, and Australia (Brown 2001,
Randle 1997).
In many countries, however, tanned or dark skin does not connote member-
ship in a fashionable class, and the possession of light skin--especially among
women--was and still is viewed as highly desirable and indicative of higher social
standing. In many Asian countries, most women practice sun avoidance diligently.
In other countries where constitutive pigmentation is darker, skin-bleaching agents
(including potent topical corticosteroids and hydroquinone formulations) have be-
come popular (Taylor 2002).
THE MULTIFACTORIAL DETERMINATION OF SKIN PIGMENTATION IN MODERN HU-
MANS The evolution of skin pigmentation in humans has been determined by
many factors (Figure 6). By far the most important of these is the UVR regime
of the environment because intensity of UVR has been the main selective fac-
tor influencing the evolution of melanin pigmentation in the skin. Through time,
the number of factors influencing the evolution of human skin pigmentation has
increased, and culture clearly has reduced the scope for the action of natural se-
lection on human skin. Cultural behaviors such as the wearing of clothes and the
utilization of shelter have become more common through time and have affected
the evolution of skin pigmentation in some populations because of their effects of
reducing an individual's UVR exposure. Related to this phenomenon is the length
of time that a population has inhabited an area with a particular UVR regime and
the latitudinal distance traversed from the ancestral to the new homeland. There
is certainly a considerable lag time between the time of settlement of an area and
time that a population reaches its "optimum" skin color for the UVR conditions
of the area. The length of that lag period for any population is not known but
would depend on the intensity of natural selection exerted on the population by
environmental influences. In early prehistory, humans possessed a simpler material
culture, spent considerable time accumulating food, and had fewer cultural trap-
pings to buffer themselves against the environment. Under these conditions, natural
selectionwouldhavepromotedmainlybiologicaladaptationstotheenvironment--
including changes in skin coloration, body proportions, and regulation of thermal
cooling. With increasing cultural competence over time, cultural solutions to the
environmental challenges of sun, heat, and cold became preeminent. The oft-cited
example of the skin colors of the native populations of equatorial South America is
worth revisiting in this connection. These populations have long been recognized
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
SKIN AND SKIN COLOR 611
Figure 6 The factors influencing human skin pigmentation, through evo-
lutionary time and during the course of a human lifetime.
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
612 JABLONSKI
as being more lightly pigmented than are their counterparts at similar latitudes
and altitudes in the Old World (Frisancho 1981, Jablonski & Chaplin 2000). This
fact is almost certainly due to the recency of populations' migration into South
America from Asia (within the past 10,000­15,000 years) and the fact that the im-
migrant populations into South America possessed many cultural behaviors and
accoutrements that protected them from high UVR exposure (Jablonski & Chaplin
2000).
Diet has also played a part in the evolution of human skin pigmentation in
very recent human history, as is well illustrated by the Eskimo-Aleut peoples
of the northeast Asian and North American Arctic. Eskimo-Aleuts exhibit skin
pigmentation darker than would be predicted on the basis of the UVMED in their
habitats (Jablonski & Chaplin 2000). Several factors have likely contributed to
this phenomenon, including the relative recency of their migration to the far north
from a lower-latitude Asian homeland and its implication that their skin color
has not caught up with their current location. This is almost certainly not the
entire story, however. The UVR regime of the latitudes in which Eskimo-Aleuts
reside comprises almost exclusively UVA throughout the year, with virtually no
vitamin D­inducing UVB except for extremely small doses in the summer months
(Chaplin 2001, Johnson et al. 1976). Habitation of this latitude (Figure 5, Zone 3)
by humans would be impossible without reliance on a highly vitamin D­rich diet.
The major components of the aboriginal Eskimo-Aleut diet--marine mammals,
fish, and caribou--provide vitamin D3 in abundance. Much of the dietary vitamin
D3 is stored in body fat (Mawer et al. 1972), denoting a possible evolutionary
connection between the development of generous subcutaneous fat stores and
vitamin D3 storage in these populations. With selection pressure on depigmentation
apparently relaxed because of diet, Eskimo-Aleuts have evolved darker skin to
protect themselves from high levels of UVA as a result of direct solar irradiation
and reflection from snow and ice. This scenario is supported by epidemiological
studies showing that departure from traditional diets in Eskimo-Aleut populations
has resulted in a high prevalence of vitamin D3­deficiency diseases, especially
rickets (Gessner et al. 1997, Haworth & Dilling 1986, Moffatt 1995).
THE GENETICS OF HUMAN SKIN COLORATION The study of genetics of human
skin pigmentation has lagged considerably behind the study of the diversity and
causation of diverse human skin color phenotypes. This situation is now changing
rapidly as comparative genomics, especially detailed studies of the genes regu-
lating coat color pigmentation in mice (Barsh 1996, Sturm et al. 2001), begin to
permit identification of the genes responsible for the pigmentation of human hair,
skin, and eyes. Sixty of the 127 currently recognized pigmentation genes in the
mouse appear to have human orthologs (Bennett 2003).
Human skin pigmentation has long been considered a polygenic trait that fol-
lows a quasi-Mendelian pattern of inheritance (Brues 1975, Byard 1981, Byard
& Lees 1981), with a few major genes of dramatic effect and additional modi-
fier genes (Sturm et al. 2001). Because pigmentation is a trait determined by the
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
SKIN AND SKIN COLOR 613
synchronized interaction of various genes with the environment (John et al. 2003),
determination of the relative roles of variant genes and varying environments has
proven extremely challenging (Sturm et al. 1998). Classical genetic studies of in-
heritance of human skin coloration have shed little light on the molecular basis
of skin color variation, beyond showing that interbreeding between light and dark
skin color phenotypes produces offspring of intermediate pigmentation (Robins
1991; Sturm et al. 1998, 2001). As a result of recent advances in the understand-
ing of the chemistry and enzymology of the biosynthesis of melanins, the genetic
regulation of the many steps in melanin production is now beginning to be under-
stood. Among the numerous mutations affecting melanocyte function in human
populations are the P-gene and members of the TYRP and SILV gene families,
which direct the assembly and maturation of melansomes within melanocytes
(Sturm et al. 2001). Investigation of the influence of these genes on skin pigmen-
tation phenotypes is just beginning (Akey et al. 2001), however, and it remains
to be demonstrated whether polymorphism in these gene systems correlates with
pigmentary differences between populations.
To date, the greatest scholarly attention has been focused on the melanocortin-1
receptor (MC1R) gene, which is the human homologue of the Agouti locus that
in mice regulates the production of the eumelanin and pheomelanin pigments of
the coat (Barsh 1996, Rana et al. 1999). In humans, the synthesis of eumelanin is
stimulated by the binding of -melanotropin (-melanocyte-stimulating hormone)
to the functional MC1R expressed on melanocytes (Scott et al. 2002). The MC1R
appears to be one of the major genes involved in the determination of human hair
and skin pigmentation, with MC1R polymorphisms in northern European popula-
tions associated with red hair and fair skin, reduced tanning ability, and high risk
of melanoma and nonmelanoma skin cancer (Healy et al. 2001, Scott et al. 2002,
Smith et al. 1998). The MC1R locus is characterized by high levels of polymor-
phism in light-skinned individuals outside of Africa and lower levels of variation
in dark-skinned individuals within Africa (John et al. 2003, Rana et al. 1999).
This is opposite the pattern observed in most other loci, where Africans are most
polymorphic (Shriver et al. 1997). The observed pattern of variation in the MC1R
suggests that different selective pressures among individuals with dark and light
skin have shaped the genetic variation at this locus, with functional constraints
operating to limit variation in African populations (John et al. 2003). The numer-
ous MC1R polymorphisms in light-skinned individuals were originally thought
to denote relaxation of selection for production of eumelanin outside of tropical
latitudes (Harding et al. 2000). A reinterpretation of these data indicates, however,
that adaptive evolution for sun-resistant MC1R alleles began when humans first
became hairless in tropical Africa, and that human movement into the less sunny
climes of Eurasia favored any mutant MC1R allele that did not produce dark skin
(Rogers et al. 2004). Recent study of the MC1R promoter function casts doubt on
the relaxation hypothesis and suggests instead the possible action of purifying or
diversifying selection on some MC1R variants in Asian and Europeans (Makova
et al. 2001). A study comparing populations in southern Africa of Bantu-language
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
614 JABLONSKI
speakers and San people showed some variation in MC1R sequences, but investi-
gators concluded that although some MC1R mutants are tolerated in Africa, this
gene has been the object of purifying selection and has played an important role
in the maintenance of dark pigmentation in Africans (John et al. 2003). The pres-
ence of higher levels of MC1R variation in dark-skinned populations subjected to
lower levels of UVR in southern Africa (as compared to equatorial Africa) sup-
ports the notion that number and kinds of MC1R variants are strongly influenced
by purifying selection (John et al. 2003). Further genetic studies of more African
populations are needed to determine if the great diversity of skin color observed
in populations in sub-Saharan Africa (Relethford 2000) can be related to specific
patterns of MC1R or other polymorphisms that evolved in response to the region's
considerable heterogeneity of UVR and precipitation regimes (Chaplin 2001).
The study of the genetics of human skin pigmentation is still in infancy, and
much remains to be learned about the levels, effects, and interactions of polymor-
phisms in the loci influencing skin color phenotype. The production of eumelanin
is under strong functional constraint as a result of natural selection in regions of
the world with high levels of UVR (Sturm et al. 2001), and there is increasing
evidence that at least MC1R variation is an adaptive response to selection for dif-
ferent alleles in different environments (Makova et al. 2001, Sturm et al. 2001).
From what is known of the timing and nature of movements of groups of early
Homo species and of Homo sapiens in prehistory, it appears that populations of
humans have moved in and out of regions with different UVR regimes over the
course of thousands of years. This finding would suggest that natural selection
would have favored the evolution of dark and light skin pigmentation in disparate
places at different times, resulting in the independent evolution of dark and light
skin phenotypes and possibly involving recurrent episodes of repigmentation and
depigmentation (Jablonski & Chaplin 2000). This phenomenon would have been
pronounced in the early history of the genus Homo (including the early history of
Homo sapiens) when cultural buffers against the environment were less effective
and sophisticated.
SKIN COLOR AND RACE Skin color is the most obvious visible attribute of the
human body. It has been the primary characteristic used to classify people into
purportedly genetically distinct geographic groups or "races." The biological ba-
sis of skin pigmentation in humans, however, strongly argues against its use as a
diagnostic classificatory trait. Critical examination of the distribution of skin color
phenotypes in humans leads to the conclusion that skin pigmentation is adaptive,
and its evolution in specific populations has been strongly influenced by the envi-
ronmental conditions (the UVR regimes, in particular) of specific places. Highly
adaptive phenotypic characteristics of organisms are of little use in classification
because they are subject to homoplasy (parallelism or convergent evolution) and
are extremely labile. Emerging genetic evidence indicates that the evolution of
pigmentation genes has been driven by purifying and diversifying selection work-
ing to produce adaptive responses in different environments (Makova et al. 2001,
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
SKIN AND SKIN COLOR 615
Rogers et al. 2004, Sturm et al. 2001). This evidence indicates that similar skin
colors have evolved independently in human populations inhabiting similar en-
vironments. Darkly or lightly pigmented skin, therefore, provides evidence only
about the nature of the past environments in which people have lived, rendering
skin pigmentation useless as a marker for membership in a unique group or "race."
The continued social importance of skin color in human affairs reflects a high
degree of sensitivity to skin color, brought about by historic and complex cultural
attitudes toward skin colors (Ehrlich & Feldman 1969, Lewontin 1995, Parra et al.
2003). The apparent existence of a difference between so-called human races and
subgroups is predicated on an exaggerated perception and heightened sensitivity
to a visually obvious attribute of human appearance. The enormity of this bias is
revealed when the small amounts of actual genetic variation within purported racial
groups are revealed (Lewontin 1995, Marks 2002). Overall, human populations are
remarkably similar to one another, with the greatest fraction of human variation
being accounted for by differences between individuals (Lewontin 1972, 1995;
Marks 2002). This collective evidence militates that the concept of biological
race be abandoned and publicly disavowed (Lewontin 1995, Marks 2002, Muir
1993). Race thus emerges as a cultural construct devoid of explanatory power and
destructive of human and social relations (Lewontin 1972, 1995; Muir 1993).
PROSPECTUS
The past decade has witnessed a tremendous advance in the understanding of the
evolution of human skin and especially skin color, largely as a result of two phe-
nomena. First is availability of remotely sensed environmental data that permit
hypotheses about the adaptive value of properties of skin to be thoroughly tested.
Second is the proliferation of studies of the molecular genetics of the skin color
that are permitting new insight into the origins of skin color phenotypes and the
mechanisms by which they have evolved. Growth is anticipated in both of these
areas, and great potential exists for their interaction, in particular for the testing of
hypotheses of adaptation through the simultaneous and detailed study of patterns
of phenotypic, genotypic, and environmental variation, such as has been done in
the study of butterfly pigmentation (Watt et al. 2003). The study of the evolution of
human skin and skin color will also be advanced by the documentation of differen-
tial survival of well-defined phenotypes and genotypes in different environmental
regimes through the use of epidemiological data, as has been undertaken recently
in the study of geographic patterns of melanoma (Garland et al. 2003).
Continued study of the evolution of human skin and skin color is important
not only to our realization of a more complete picture of human evolution, but
also it is important because the skin is involved in so many aspects of human well-
being. Many humans now live in regions far distant from their ancestral homelands,
but they retain a covering of skin adapted to remote Pleistocene conditions. As
is evidenced by modern rates of skin cancer and vitamin D deficiencies, human
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
616 JABLONSKI
behavior and culture are not perfect buffers against the effects of these major
translocations.Anappreciationofthemanyrolesofskinwillimprovehumanhealth
and attitudes toward diversity and will promote the fundamental understanding of
why it is that people look the way they do.
ACKNOWLEDGMENTS
Conversations and discussions with many colleagues in the months and years
leading up to the writing of this review greatly enriched the content of this paper.
These include Walter Alvarez, Carol Boggs, Carol Bower, C. Loring Brace, Jim
Cleaver, Paul Ehrlich, Roberto Frisancho, Cedric Garland, Maciej Henneberg, the
late Gabriel Lasker, Charles Oxnard, Bill Nye, Lynn Rothschild, Donald Tyler, and
Ward Watt. I thank Rachel Wolf for helping to assemble references. Figures 1, 2,
3, and 6 were produced by Jennifer Kane, whom I thank for her skillful renderings
and patient revisions. I thank George Chaplin for producing Figures 4 and 5, and for
more than ten years worth of challenging and inspiring discussions on the evolution
of human skin and skin coloration. The spatial analyses and maps summarized here
were made possible by generous donations of geographical information systems
software to the California Academy of Sciences from Charles Convis and the
Conservation Program of the Environmental Systems Research Institute (ESRI).
The Annual Review of Anthropology is online at http://anthro.annualreviews.org
LITERATURE CITED
Akey JM, Wang H, Xiong M, Wu H, Liu
W, et al. 2001. Interaction between the
melanocortin-1 receptor and P genes con-
tributes to inter-individual variation in skin
pigmentation phenotypes in a Tibetan popu-
lation. Hum. Genet. 108:516­20
Aoki K. 2002. Sexual selection as a cause of hu-
man skin colour variation: Darwin's hypoth-
esis revisited. Ann. Hum. Biol. 29:589­608
AtiqM,SuriaA,QamaruddinNizamiS,Ahmed
I. 1998. Maternal vitamin-D deficiency
in Pakistan. Acta Obstet. Gynecol. Scand.
77:970­73
Baker PT. 1958. Racial differences in heat tol-
erance. Am. J. Phys. Anthropol. 16:287­305
Barker D, Dixon K, Medrano EE, Smalara D,
Im S, et al. 1995. Comparison of the re-
sponses of human melanocytes with different
melanin contents to ultraviolet B irradiation.
Cancer Res. 55:4041­46
Barsh G. 1996. The genetics of pigmentation:
from fancy genes to complex traits. Trends
Genet. 12:299­305
Barsh GS. 2003. What controls variation in hu-
man skin color? Public Libr. Sci. Biol. Oct.
1(1):E27
Bennett DC. 2003. IL-12. The colours of mice
and men--100 genes and beyond? Pigment
Cell Res. 16:576­77
Bereket A. 2003. Rickets in developing coun-
tries. In Vitamin D and Rickets, ed. Z
Hochberg, pp. 220­32. Basel: Karger
Bjorn LO, Wang T. 2000. Vitamin D in an eco-
logical context. Int. J. Circumpolar. Health
59:26­32
Blum HF. 1961. Does the melanin pigment of
human skin have adaptive value? Q. Rev.
Biol. 36:50­63
Bower C, Stanley FJ. 1989. Dietary folate as a
risk factor for neural-tube defects: evidence
from a case-control study in Western Aus-
tralia. Med. J. Aust. 150:613­19
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
SKIN AND SKIN COLOR 617
Brace CL. 1963. Structural reduction in evolu-
tion. Am. Nat. 97:39­49
Branda RF, Eaton JW. 1978. Skin color and nu-
trient photolysis: an evolutionary hypothesis.
Science 201:625­26
BrownDA.2001.Skinpigmentationenhancers.
See Goldsmith 1991, pp. 637­75
Brues AM. 1975. Rethinking human pigmenta-
tion. Am. J. Phys. Anthropol. 43:387­92
Brunvand L, Haug E. 1993. Vitamin D defi-
ciency amongst Pakistani women in Oslo.
Acta Obstet. Gynecol. Scand. 72:264­68
Byard PJ. 1981. Quantitative genetics of human
skin color. Yearb. Phys. Anthropol. 24:123­
37
Byard PJ, Lees FC. 1981. Estimating the num-
beroflocideterminingskincolourinahybrid
population. Ann. Hum. Biol. 8:49­58
Cabanac M, Caputa M. 1979. Natural selec-
tive cooling of the human brain: evidence
of its occurrence and magnitude. J. Physiol.
286:255­64
Caldwell MM, Bjorn LO, Bornman JF, Flint
SD, Kulandaivelu G, et al. 1998. Effects of
increasedsolarultravioletradiationonterres-
trial ecosystems. J. Photochem. Photobiol. B.
46:40­52
Chaplin G. 2001. The geographic distribution
of environmental factors influencing human
skin colouration. M.Sc. thesis, Manchester
Metropolitan Univ.
Chaplin G. 2004. Geographical distribution
of environmental factors influencing skin
coloration. Am. J. Phys. Anthropol. 123:In
press
Chaplin G, Jablonski NG. 1998. Hemispheric
difference in human skin color. Am. J. Phys.
Anthropol. 107:221­24
Chaplin G, Jablonski NG, Cable NT. 1994.
Physiology, thermoregulation and bipedal-
ism. J. Hum. Evol. 27:497­510
Chu DH, Haake AR, Holbrook K, Loomis CA.
2003. The structure and development of skin.
See Freedberg et al. 2003, pp. 58­88
Chung JH. 2001. The effects of sunlight on the
skin of Asians. See Goldsmith 1991, pp. 69­
90
Cleaver JE, Crowley E. 2002. UV damage,
DNA repair and skin carcinogenesis. Front.
Biosci. 7:1024­43
Clemens TL, Henderson SL, Adams JS, Holick
MF. 1982. Increased skin pigment reduces
the capacity of skin to synthesise vitamin D3.
Lancet 1:74­76
Cornish DA, Maluleke V, Mhlanga T. 2000.
An investigation into a possible relationship
between vitamin D, parathyroid hormone,
calcium and magnesium in a normally pig-
mented and an albino rural black population
in the Northern Province of South Africa.
BioFactors 11:35­38
Cosentino MJ, Pakyz RE, Fried J. 1990.
Pyrimethamine: an approach to the devel-
opment of a male contraceptive. Proc. Natl.
Acad. Sci. USA 87:1431­35
Cowles RB. 1959. Some ecological factors
bearing on the origin and evolution of pig-
ment in the human skin. Am. Nat. 93:283­
93
Daniels F. 1964. Man and radiant energy: solar
radiation. In Adaptation to the Environment,
ed. DB Dill, EF Adolph, CG Wilber, pp. 969­
85. Washington, DC: Am. Physiol. Soc.
Daniels F, Post PW, Johnson BE. 1972. The-
ories of the role of pigment in the evolution
of human races. In Pigmentation: Its Genesis
and Biologic Control, ed. V Riley, pp. 13­22.
New York: Appleton-Century-Crofts
Deol MS. 1975. Racial differences in pigmenta-
tion and natural selection. Ann. Hum. Genet.
London 38:501­3
Desruelle AV, Candas V. 2000. Thermoregula-
tory effects of three different types of head
cooling in humans during a mild hyperther-
mia. Eur. J. Appl. Physiol. 81:33­39
Dmi'el R, Prevulotzky A, Shkolnik A. 1980. Is
a black coat in the desert a means of saving
metabolic energy? Nature 283:761­62
Domb LG, Pagel M. 2001. Sexual swellings ad-
vertise female quality in wild baboons. Na-
ture 410:204­6
Edwards EA, Duntley Q. 1939. The pigments
and color of living human skin. Am. J. Anat.
65:1­33
EhrlichPR,FeldmanSS.1969.TheRaceBomb:
Skin Color, Prejudice, and Intelligence. New
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
618 JABLONSKI
York: Quadrangle/N.Y. Times Book Co. 207
pp.
Elias PM, Feingold KR, Fluhr JW. 2003. Skin
as an organ of protection. See Freedberg et al.
2003, pp. 107­18
Ellis RA, Montagna W. 1962. The skin of
primates. VI. The skin of the gorilla (Go-
rilla gorilla). Am. J. Phys. Anthropol. 20:79­
93
Falk D. 1990. Brain evolution in Homo: the "ra-
diator" theory. Behav. Brain Sci. 13:333­81
Feldman D, Glorieux FH, Pike JW, eds. 1997.
Vitamin D. San Diego: Academic
Fisher GJ, Kang S, Varani J, Bata-Csorgo Z,
Yinsheng W, et al. 2002. Mechanisms of pho-
toaging and chronological skin aging. Arch.
Dermatol. 138:1462­70
Fitzpatrick TB. 1988. The validity and practi-
cality of sun reactive skin type I through VI.
Arch. Dermatol. 124:869­71
Fitzpatrick TB, Becker SW Jr, Lerner AB,
Montgomery H. 1950. Tyrosinase in human
skin: demonstration of its presence and of
its role in human melanin formation. Science
112:223­25
Fitzpatrick TB, Seiji M, McGugan AD. 1961.
Melanin pigmentation. New Engl. J. Med.
265:328­32
Fitzpatrick TR, Ortonne J-P. 2003. Normal skin
color and general considerations of pigmen-
tary disorders. See Freedberg et al. 2003, pp.
819­25
Fleming A, Copp AJ. 1998. Embryonic folate
metabolism and mouse neural tube defects.
Science 280:2107­9
Fogelman Y, Rakover Y, Luboshitsky R. 1995.
High prevalence of vitamin D deficiency
among Ethiopian women immigrants to Is-
rael: exacerbation during pregnancy and lac-
tation. Isr. J. Med. Sci. 31:221­24
Folk GEJ, Semken HAJ. 1991. The evolution of
sweat glands. Int. J. Biometeor. 35:180­86
Fonseca V, Tongia R, El-Hazmi M, Abu-Aisha
H. 1984. Exposure to sunlight and vitamin
D deficiency in Saudi Arabian women. Post-
grad. Med. J. 60:589­91
Freedberg IM, Eisen AZ, Wolff K, Austen FK,
Goldsmith LA, Katz SI, eds. 2003. Fitz-
patrick's Dermatology in General Medicine.
New York: McGraw-Hill
Frisancho AR. 1981. Human Adaptation: A
Functional Interpretation. Ann Arbor: Univ.
Mich. Press. 209 pp.
Frost P. 1988. Human skin color: a possible
relationship between its sexual dimorphism
anditssocialperception.Perspect.Biol.Med.
32:38­59
Gambichler T, Sauermann K, Bader A, Alt-
meyer P, Hoffmann K. 2001. Serum folate
levels after UVA exposure: a two-group par-
allel randomised controlled trial. BMC Der-
matol. 1(1):8
Garland CF, Garland FC, Gorham ED. 2003.
Epidemiologic evidence for different roles of
ultraviolet A and B radiation in melanoma
mortality rates. Ann. Epidemiol. 13:395­404
Gessner BD, deSchweinitz E, Petersen KM,
Lewandowski C. 1997. Nutritional rickets
among breast-fed black and Alaska native
children. Alaska Med. 39:72­87
Giacomoni PU, ed. 2001. Sun Protection in
Man. Amsterdam: Elsevier
Goldsmith LA, ed. 1991. Physiology, Biochem-
istry, and Molecular Biology of the Skin. New
York: Oxford Univ. Press
Goldsmith LA. 2003. Biology of eccrine and
apocrine sweat glands. See Freedberg et al.
2003, pp. 99­106
Grant WB. 2002. An estimate of premature can-
cer mortality in the U.S. due to inadequate
doses of solar ultraviolet-B radiation. Can-
cer 94:1867­75
Halaban R, Hebert DN, Fisher DE. 2003. Bi-
ology of melanocytes. See Freedberg et al.
2003, pp. 127­47
Harding RM, Healy E, Ray AJ, Ellis NS, Flana-
gan N, et al. 2000. Evidence for variable
selective pressures at MC1R. Am. J. Hum.
Genet. 66:1351­61
Haworth JC, Dilling LA. 1986. Vitamin-D-
deficient rickets in Manitoba, 1972­84. Can.
Med. Assoc. J. 134:237­41
Healy E, Jordan SA, Budd PS, Suffolk R, Rees
JL, Jackson IJ. 2001. Functional variation of
MCR1 alleles from red-haired individuals.
Hum. Mol. Genet. 10:2397­402
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
SKIN AND SKIN COLOR 619
Herman J, Celarier E. 1996. TOMS Version
7 UV-erythemal exposure: 1978­1993. Data
developed by NASA Goddard Space Flight
Cent. Ozone Process. Team
Hirakawa K, Suzuki H, Oikawa S, Kawanishi
S. 2002. Sequence-specific DNA damage in-
duced by ultraviolet A-irradiated folic acid
via its photolysis product. Arch. Biochem.
Biophys. 410:261­28
Hitchcock RT. 2001. Ultraviolet Radiation.
Fairfax, VA: Am. Ind. Hyg. Assoc. 49 pp.
Hodgkin P, Kay GH, Hine PM, Lumb GA,
Stanbury SW. 1973. Vitamin-D deficiency in
Asians at home and in Britain. Lancet 2:167­
72
Holick MF. 1991. Photobiology, physiology,
and clinical applications for vitamin D. See
Goldsmith 1991, pp. 928­56
HolickMF.1995.Environmentalfactorsthatin-
fluence the cutaneous production of vitamin
D. Am. J. Clin. Nutr. 61:638S­45
Holick MF. 1997. Photobiology of vitamin D.
See Feldman et al. 1997, pp. 33­39
Holick MF. 2001. A perspective on the benefi-
cial effects of moderate exposure to sunlight:
bone health, cancer prevention, mental health
and well being. See Goldsmith 1991, pp. 11­
37
Holick MF. 2003. Vitamin D: a millenium per-
spective. J. Cell Biochem. 88:296­307
Holick MF, MacLaughlin JA, Doppelt SH.
1981. Regulation of cutaneous previtamin
D3 photosynthesis in man: Skin pigment is
not an essential regulator. Science 211:590­
93
Ihara Y, Aoki K. 1999. Sexual selection by male
choice in monogamous and polygynous hu-
man populations. Theor. Popul. Biol. 55:77­
93
Ito S. 2003. A chemist's view of melanogenesis.
Pigment Cell Res. 16:230­36
Jablonski NG, Chaplin G. 2000. The evolution
of skin coloration. J. Hum. Evol. 39:57­106
Jimbow K, Fitzpatrick TB, Wick MM. 1991.
Biochemistryandphysiologyofmelaninpig-
mentation.SeeGoldsmith1991,pp. 873­909
Jimbow K, Quevedo WC Jr, Fitzpatrick TB, Sz-
abo G. 1976. Some aspects of melanin biol-
ogy: 1950­1975. J. Invest. Dermatol. 67:72­
89
John PR, Makova K, Li WH, Jenkins T, Ramsay
M. 2003. DNA polymorphism and selection
at the melanocortin-1 receptor gene in nor-
mally pigmented southern African individu-
als. Ann. N.Y. Acad. Sci. 994:299­306
Johnson FS, Mo T, Green AES. 1976. Average
latitudinal variation in ultraviolet radiation at
the Earth's surface. Photochem. Photobiol.
23:179­88
Jones G, Strungnell SA, DeLuca HF. 1998. Cur-
rent understanding of the molecular actions
of vitamin D. Physiol. Rev. 78:1193­231
Kaidbey KH, Agin PP, Sayre RM, Kligman
AM. 1979. Photoprotection by melanin--a
comparison of black and Caucasian skin. Am.
Acad. Dermatol. 1:249­60
Kesavan V, Pote MS, Batra V, Viswanathan G.
2003. Increased folate catabolism following
total body y-irradiation in mice. J. Radiat.
Res. 44:141­44
Kielbassa C, Roza L, Epe B. 1997. Wavelength
dependence of oxidative DNA damage in-
duced by UV and visible light. Carcinogen-
esis 18:811­16
Knip AS. 1977. Ethnic studies on sweat gland
counts. In Physiological Variation and its
Genetic Basis, ed. JS Weiner, pp. 113­23.
London: Taylor & Francis
Kollias N. 1995a. The physical basis of skin
color and its evaluation. Clin. Dermatol.
13:361­67
Kollias N. 1995b. The spectroscopy of human
melanin in pigmentation. In Melanin: Its
Role in Human Photoprotection, ed. L Zeise,
MR Chedekel, TB Fitzpatrick, pp. 11­22.
Overland Park: Valdenmar
Kollias N, Sayre RM, Zeise L, Chedekel MR.
1991. Photoprotection by melanin. J. Pho-
tochem. Photobiol. B. 9:135­60
Kraning KK. 1991. Temperature regulation and
the skin. See Goldsmith 1991, pp. 1085­95
Kripke ML, Applegate LA. 1991. Alterations in
the immune response by ultraviolet radiation.
See Goldsmith 1991, pp. 1222­39
Lasker GW. 1954. Seasonal changes in skin
color. Am. J. Phys. Anthropol. 12:553­38
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
620 JABLONSKI
Lavker RM, Bertolino A, Freedberg IM, Sun T-
T. 2003. Biology of hair follicles. See Freed-
berg et al. 2003, pp. 148­58
Lee MMC, Lasker GW. 1959. The sun-tanning
potential of human skin. Hum. Biol. 31:252­
60
Lees FC, Byard PJ. 1978. Skin colorimetry in
Belize. I. Conversion formulae. Am. J. Phys.
Anthropol. 48:515­22
Lewontin RC. 1972. The apportionment of hu-
man diversity. Evol. Biol. 6:381­98
Lewontin RC. 1995. Human Diversity:
Freeman
Lock-Anderson J, Knudstorp ND, Wulf HC.
1998. Faculative skin pigmentation in cau-
casians: an objective biological indicator of
lifetime exposure to ultraviolet radiation? Br.
J. Dermatol. 138:826­32
Loomis WF. 1967. Skin-pigment regulation
of vitamin-D biosynthesis in man. Science
157:501­6
Lucock M, Yates Z, Glanville T, Leeming R,
Simpson N, Daskalakis I. 2003. A critical
role for B-vitamin nutrition in human devel-
opment and evolutionary biology. Nutr. Res.
23:1463­75
Lynn B. 1991. Cutaneous sensation. See Gold-
smith 1991, pp. 779­815
MacKintosh JA. 2001. The antimicrobial prop-
erties of melanocytes, melanosomes and
melanin and the evolution of black skin. J.
Theor. Biol. 211:101­13
MacLaughlin JA, Anderson RR, Holick MF.
1982. Spectral character of sunlight modu-
lates photosynthesis of previtamin D3 and
its photoisomers in human skin. Science
216:1001­3
Madronich S, McKenzie RL, Bjorn LO, Cald-
well MM. 1998. Changes in biologically ac-
tive ultraviolet radiation reaching the Earth's
surface. J. Photochem. Photobiol. B. 46:5­19
Makova K, Ramsay M, Jenkins T, Li W-H.
2001. Human DNA sequence variation in a
6.6-kb region containing the melanocortin 1
receptor promoter. Genetics 158:1253­68
Marks J. 2002. What It Means to be 98%
Chimpanzee: Apes, People, and Their Genes.
Berkeley: Univ. Calif. Press. 312 pp.
Marks R. 1991. Mechanical properties of the
skin. See Goldsmith 1991
Martin RD. 1990. Primate Origins and Evolu-
tion: A Phylogenetic Perspective. Princeton,
NJ: Princeton Univ. Press. 804 pp.
Mathur U, Datta SL, Mathur BB. 1977. The
effect of aminopterin-induced folic acid de-
ficiency on spermatogenesis. Fertil. Steril.
28:1356­60
Matsuoka LY, Wortsman J, Dannenberg MJ,
Hollis BW, Lu Z, Holick MF. 1992. Clothing
prevents ultraviolet-B radiation-dependent
photosynthesis of vitamin D3. J. Clin. En-
docrinol. Metabol. 75:1099­103
Mawer EB, Backhouse J, Holman CA, Lumb
GA,StanburySW.1972.Thedistributionand
storage of vitamin D and its metabolites in
human tissues. Clin. Sci. 43:413­31
McHenry HM, Berger LR. 1998. Body pro-
portions in Australopithecus afarensis and A.
africanus and the origin of the genus Homo.
J. Hum. Evol. 35:1­22
Minns RA. 1996. Folic acid and neural tube de-
fects. Spinal Cord 34:460­65
Mitra D, Bell NH. 1997. Racial, geographic,
genetic, and body habitus effects on vitamin
D metabolism. See Feldman et al. 1997, pp.
521­32
Moffatt MEK. 1995. Current status of nutri-
tional deficiencies in Canadian Aboriginal
people. Can. J. Physiol. Pharmacol. 73:754­
58
Montagna W. 1967. Comparative anatomy and
physiology of the skin. Arch. Dermatol.
96:357­63
Montagna W. 1971. Cutaneous comparative bi-
ology. Arch. Dermatol. 104:577­91
Montagna W. 1981. The consequences of hav-
ing a naked skin. Birth Defects: Orig. Artic.
Ser. 17:1­7
Morison WL. 1985. What is the function
of melanin? Arch. Dermatol. 121:1160­
63
MuirDE.1993.Race:themythicrootofracism.
Sociol. Inq. 63:339­50
Murray FG. 1934. Pigmentation, sunlight, and
nutritional disease. Am. Anthropol. 36:438­
45
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
SKIN AND SKIN COLOR 621
Neer RM. 1975. The evolutionary significance
of vitamin D, skin pigment, and ultraviolet
light. Am. J. Phys. Anthropol. 43:409­16
Neer RM. 1985. Environmental light: effects of
vitamin D synthesis and calcium metabolism
in humans. Ann. N.Y. Acad. Sci. 453:14­20
Nelson DA, Nunneley SA. 1998. Brain tem-
perature and limits on transcranial cooling in
humans: quantitative modeling results. Eur.
J. Appl. Physiol. 78:353­59
Odland GF. 1991. Structure of the skin. See
Goldsmith 1991, pp. 3­62
Olivier G. 1960. Pratique Anthropolgique.
Paris: Vigot Fr´eres
Online Mendelian Inheritance in Man. 2003.

139200 group-specific component; GC.
In Online Mendelian Inheritance in Man.
Bethesda, MD: Natl. Cent. Biotechnol. Inf.
Ortonne J-P. 1990. Pigmentary changes of the
ageing skin. Br. J. Dermatol. 122:21­28
Ortonne J-P. 2002. Photoprotective properties
of skin melanin. Br. J. Dermatol. 146:7­10
Pagel M, Bodmer W. 2003. A naked ape would
have fewer parasites. Proc. R. Soc. London B
270:S117­19
Pandolf KG, Gange RW, Latzka WA, Blank IH,
Kraning KK, Gonzalez RR. 1992. Human
thermoregulatory responses during heat ex-
posure after artificially induced sunburn. Am.
J. Physiol. 262:R610­16
Parker F. 1981. The biology of pigmentation.
Birth Defects 17:79­91
Parra FC, Amado RC, Lambertucci JR, Rocha
J, Antunes CM, Pena SDJ. 2003. Color and
genomic ancestry in Brazilians. Proc. Natl.
Acad. Sci. USA 100:177­82
Post PW, Daniels F Jr, Binford RT Jr. 1975.
Cold injury and the evolution of "white" skin.
Hum. Biol. 47:65­80
Prota G. 1992a. Melanin-producing cells. See
Prota 1992b, pp. 14­33
Prota G, ed. 1992b. Melanins and Melanogen-
esis. San Diego: Academic
Prota G. 1992c. Photobiology and photochem-
istry of melanogenesis. See Prota 1992b, pp.
208­24
Quevedo WC Jr, Fitzpatrick TB, Pathak MA.
1975. Role of light in human skin color
variation. Am. J. Phys. Anthropol. 43:393­
408
Rana BK, Hewett-Emmett D, Jin L, Chang BH,
Sambuughin N, et al. 1999. High polymor-
phism at the human melanocortin 1 receptor
locus. Genetics 151:1547­57
Randle HW. 1997. Suntanning: differences in
perceptions throughout history. Mayo Clin.
Proc. 72:461­66
Rawles ME. 1948. Origin of melanophores
and their role in development of color pat-
terns in vertebrates. Physiol. Res. 28:383­
408
Relethford JH. 1997. Hemispheric difference
in human skin color. Am. J. Phys. Anthropol.
104:449­57
Relethford JH. 2000. Human skin color diver-
sity is highest in sub-Saharan African popu-
lations. Hum. Biol. 72:771­80
Roberts DF. 1977. Human pigmentation: its ge-
ographical and racial distribution and biolog-
ical significance. J. Soc. Cosmet. Chem. 28:
329­42
Roberts DF, Kahlon DPS. 1976. Environmental
correlations of skin colour. Ann. Hum. Biol.
3:11­22
Robins AH. 1991. Biological Perspectives
on Human Pigmentation. Cambridge: Cam-
bridge Univ. Press. 253 pp.
Rogers AR, Iltis D, Wooding S. 2004. Genetic
variation at the MC1R locus and the time
since loss of human body hair. Curr. Anthro-
pol. 45:105­7
Rozema J, van Geel B, Bjorn LO, Lean J,
Madronich S. 2002. Toward solving the UV
puzzle. Science 296:1621­22
Ruff CB, Trinkaus E, Holliday TW. 1997.
Body mass and encephalization in Pleis-
tocene Homo. Nature 387:173­76
Ruff CB, Trinkaus E, Walker A, Larsen CS.
1993. Postcranial robusticity in Homo. I:
Temporal trends and mechanical interpreta-
tion. Am. J. Anthropol. 91:21­53
Ryan TJ. 1991. Cutaneous circulation. See
Goldsmith 1991, pp. 1019­84
Sarna T, Swartz HM. 1998. The physical prop-
erties of melanins. In The Pigmentary Sys-
tem: Physiology and Pathophysiology, ed. JJ
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
622 JABLONSKI
Nordlund, RE Boissey, VJ Hearing, pp. 333­
57. New York: Oxford Univ. Press
Schwartz GG, Rosenblum LA. 1981. Allometry
of primate hair density and the evolution of
human hairlessness. Am. J. Phys. Anthropol.
55:9­12
Scott MC, Suzuki I, Abdel-Malek ZA. 2002.
Regulation of the human melanocortin 1 re-
ceptor expression in epidermal melanocytes
by paracrine and endocrine factors and by
ultraviolet radiation. Pigment Cell Res. 15:
433­39
Shaw NJ. 2003. Vitamin D deficiency rickets.
In Vitamin D and Rickets, ed. Z Hochberg,
pp. 93­104. Basel: Karger
Shea CR, Parrish JA. 1991. Nonionizing radi-
ation and the skin. See Goldsmith 1991, pp.
910­27
ShonoS,ImuraM,OtaM,OnoS,TodaK.1985.
The relationship of skin color, UVB-induced
erythema, and melanogenesis. J. Invest. Der-
matol. 84:265­67
Shriver MD, Jin L, Ferrell RE, Deka R. 1997.
Microsatellite data support an early popu-
lation expansion in Africa. Genome Res. 7:
586­89
Smith RM, Healy E, Siddiqui S. 1998.
Melanocortin 1 receptor variants in an Irish
population. J. Invest. Dermatol. 111:119­22
Steegmann AT Jr. 1967. Frostbite of the human
face as a selective force. Hum. Biol. 39:131­
44
Sturm RA. 2002. Skin colour and skin cancer--
MC1R, the genetic link. Melanoma Res.
12:405­16
Sturm RA, Fox NF, Ramsay M. 1998. Human
pigmentation genetics: the difference is only
skin deep. BioEssays 20:712­21
Sturm RA, Teasdale RD, Fox NF. 2001. Human
pigmentation genes: identification, structure
and consequences of polymorphic variation.
Gene 277:49­62
Suh JR, Herbig AK, Stover PJ. 2001. New per-
spectives on folate catabolsim. Annu. Rev.
Nutr. 21:255­82
Sulaimon SS, Kitchell BE. 2003. The biology
of melanocytes. Vet. Dermatol. 14:57­65
Szabo G, Gerald AB, Pathak MA, Fitzpatrick
TB. 1969. Racial differences in the fate of
melanosomes in human epidermis. Nature
222:1081­82
Taylor SC. 2002. Skin of color: biology, struc-
ture, function, and implications for derma-
tologic disease. J. Am. Acad. Dermatol.
46:S41­62
Thody AJ, Higgins EM, Wakamatsu K, Ito S,
Burchill SA, Marks JM. 1991. Pheomelanin
as well as eumelanin is present in human epi-
dermis. J. Invest. Dermatol. 97:340­44
Thody AJ, Smith AG. 1977. Hormones and skin
pigmentation in the mammal. Int. J. Derma-
tol. 16:657­64
Thomas MK, Lloyd-Jones DM, Thadhani RI,
Shaw AC, Deraska DJ, et al. 1998. Hypovi-
taminosis D in medical inpatients. New Engl.
J. Med. 338:777­83
Urbach F. 2001. The negative effects of solar
radiation: a clinical overview. See Goldsmith
1991, pp. 39­67
van den Berghe PL, Frost P. 1986. Skin color
preference, sexual dimorphism and sexual
selection:acaseofgene-culturecoevolution?
Ethnic Rac. Stud. 9:87­113
Vermeer M, Schmieder GJ, Yoshikawa T, van
den Berg J-W, Metzman MS, et al. 1991.
EffectsofultravioletBlightoncutaneousim-
mune responses of humans with deeply pig-
mented skin. J. Invest. Dermatol. 97:729­34
von Luschan F. 1897. Beitrage zur Volk-
erkunde der Deutschen Schutzgebieten.
Berlin: Deutsche Buchgemeinschaft
Wagner JK, Parra EJ, Norton HL, Jovel C,
Shriver MD. 2002. Skin responses to ultravi-
olet radiation: effects of constitutive pigmen-
tation, sex, and ancestry. Pigment Cell Res.
15:385­90
Walsberg GE. 1988. Consequences of skin
color and fur properties for solar heat gain
and ultraviolet irradiance in two mammals.
J. Comp. Physiol. B 158:213­21
Walter H. 1958. Der zusammenhang von haut-
farbenverteilung und intensitat der ultravio-
letten strahlung. Homo 9:1­13
Walter H. 1971. Remarks on the environmen-
tal adaptation of man. Humangenetik 13:85­
97
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
SKIN AND SKIN COLOR 623
Wassermann GD. 1981. On the nature of the
theory of evolution. Philos. Sci. 48:416­37
Wassermann HP. 1965a. The circulation of
melanin--its clinical and physiological sig-
nificance. S. Afr. Med. J. 39:711­16
Wassermann HP. 1965b. Human pigmentation
and environmental adaptation. Arch. Envi-
ron. Health 11:691­94
Wassermann HP. 1974. Ethnic Pigmentation.
New York: Blackwell Sci. 284 pp.
Watt WB, Wheat CW, Meyer EH, Martin J-F.
2003. Adaptation at specific loci. VII. Nat-
ural selection, dispersal and the diversity
of molecular-functional variation patterns
among butterfly species complexes (Colias:
Lepidoptera, Pieridae). Mol. Ecol. 12:1265­
75
Webb AR, Kline L, Holick MF. 1988. Influ-
ence of season and latitude on the cutaneous
synthesis of vitamin D3: Exposure to winter
sunlight in Boston and Edmonton will not
promote vitamin D3 synthesis in human skin.
J. Clin. Endocrinol. Metabol. 67:373­78
Weiner JS. 1951. A spectrophotometer for mea-
surement of skin colour. Man 51:152­53
WeinerJS.1977.Variationinsweating.InPhys-
iological Variation and its Genetic Basis, ed.
JS Weiner, pp. 125­37. London: Taylor &
Francis
Wenger BC. 2003. Thermoregulation. See
Freedberg et al. 2003, pp. 119­26
Westerhof W, Estevez-Uscanga O, Meens J,
Kammeyer A, Durocq M, Cario I. 1990. The
relation between constitutional skin color
and photosensitivity estimated from UV-
induced erythema and pigmentation dose-
responsecurves.J.Invest.Dermatol.94:812­
16
Wharton B, Bishop N. 2003. Rickets. Lancet
362:1389­400
Wheeler PE. 1984. The evolution of bipedality
and loss of functional body hair in hominids.
J. Hum. Evol. 13:91­98
Wheeler PE. 1985. The loss of functional body
hair in man: the influence of thermal envi-
ronment, body form and bipedality. J. Hum.
Evol. 14:23­28
Wheeler PE. 1991a. The influence of bipedal-
ism on the energy and water budgets of early
hominids. J. Hum. Evol. 21:117­36
Wheeler PE. 1991b. The thermoregulatory ad-
vantages of hominid bipedalism in open
equatorial environments: the contribution of
increased convective heat loss and cutaneous
evaporative cooling. J. Hum. Evol. 21:107­
15
Wheeler PE. 1992. The thermoregulatory ad-
vantages of large body size for hominids
foraging in savannah environments. J. Hum.
Evol. 23:351­62
Wheeler PE. 1996. The environmental context
of functional body hair loss in hominids (a re-
ply to Amaral, 1996). J. Hum. Evol. 30:367­
71
Young AR, Sheehan J. 2001. UV-induced pig-
mentation in human skin. See Goldsmith
1991, pp. 357­75
Zihlman AL, Cohn BA. 1988. The adaptive re-
sponse of human skin to the Savanna. 3:397­
409
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
SKIN AND SKIN COLOR C-1
Figure 3 Human skin coloration as predicted from multiple regression formulae. See text
for discussion.
Figure 5 The potential for synthesizing vitamin D3 in the skin relative to levels of annual
average UVMED. The highest UVMED levels are indicated in deep violet, with incremen-
tally lower levels indicated in shades of red, orange, yellow, green, and gray. Zone 1 (area
without hachure enclosing the tropics) represents the region with adequate UVR throughout
the year to catalyze vitamin D3 synthesis. Zone 2 (vertical hachure) represents the area in
which there is insufficient UVR during at least one month of the year to produce vitamin D3.
Zone 3 (cross-hatched area) represents the region in which there is insufficient UVR aver-
aged over the entire year to photosynthesize vitamin D3. See text for further description.
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
Annual Review of Anthropology
Volume 33, 2004
CONTENTS
Frontispiece--Marilyn Strathern xiv
PREFATORY CHAPTER
The Whole Person and Its Artifacts, Marilyn Strathern 1
ARCHAEOLOGY
The Archaeology of Ancient State Economies, Michael E. Smith 73
Political Economic Mosaics: Archaeology of the Last Two Millennia in
Tropical Sub-Saharan Africa, Ann Brower Stahl 145
Primary State Formation in Mesoamerica, Charles S. Spencer
and Elsa M. Redmond 173
The Archaeology of Communication Technologies, Stephen D. Houston 223
Origins and Development of Urbanism: Archaeological Perspectives,
George L. Cowgil 525
BIOLOGICAL ANTHROPOLOGY
Early Dispersals of Homo from Africa, Susan C. Ant´on
and Carl C. Swisher, III 271
Social Status and Health in Humans and Other Animals,
Robert M. Sapolsky 393
The Peopling of the New World: Perspectives from Molecular
Anthropology, Theodore G. Schurr 551
The Evolution of Human Skin and Skin Color, Nina G. Jablonski 585
LINGUISTICS AND COMMUNICATIVE PRACTICES
Language Revitalization and New Technologies: Cultures of Electronic
Mediation and the Refiguring of Communities, Patrick Eisenlohr 21
New Technologies and Language Change: Toward an Anthropology of
Linguistic Frontiers, Susan E. Cook 103
Language Birth and Death, Salikoko S. Mufwene 201
Talk and Interaction Among Children and the Co-Construction of Peer
Groups and Peer Culture, Amy Kyratzis 625
vii
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
viii CONTENTS
INTERNATIONAL ANTHROPOLOGY AND REGIONAL STUDIES
Christianity in Africa: From African Independent to
Pentecostal-Charismatic Churches, Birgit Meyer 447
Anthropology in Area Studies, Jane I. Guyer 499
SOCIOCULTURAL ANTHROPOLOGY
Music and the Global Order, Martin Stokes 47
The Globalization of Pentecostal and Charismatic Christianity,
Joel Robbins 117
Hang on to Your Self: Of Bodies, Embodiment, and Selves,
Steven Van Wolputte 251
The Body Beautiful: Symbolism and Agency in the Social World,
Erica Reischer and Kathryn S. Koo 297
Inscribing the Body, Enid Schildkrout 319
Culture, Globalization, Mediation, William Mazzarella 345
The World in Dress: Anthropological Perspectives on Clothing, Fashion,
and Culture, Karen Tranberg Hansen 369
Anthropology and Circumcision, Eric K. Silverman 419
Thinking About Cannibalism, Shirley Lindenbaum 475
THEME I: THE BODY AS A PUBLIC SURFACE
Hang on to Your Self: Of Bodies, Embodiment, and Selves,
Steven Van Wolputte 251
The Body Beautiful: Symbolism and Agency in the Social World,
Erica Reischer and Kathryn S. Koo 297
Inscribing the Body, Enid Schildkrout 319
The World in Dress: Anthropological Perspectives on Clothing, Fashion,
and Culture, Karen Tranberg Hansen 369
Anthropology and Circumcision, Eric K. Silverman 419
The Evolution of Human Skin and Skin Color, Nina G. Jablonski 585
THEME II: NEW TECHNOLOGIES OF COMMUNICATION
Language Revitalization and New Technologies: Cultures of Electronic
Mediation and the Refiguring of Communities, Patrick Eisenlohr 21
Music and the Global Order, Martin Stokes 47
New Technologies and Language Change: Toward an Anthropology of
Linguistic Frontiers, Susan E. Cook 103
The Archaeology of Communication Technologies, Stephen D. Houston 223
Culture, Globalization, Mediation, William Mazzarella 345
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.
CONTENTS ix
INDEXES
Subject Index 651
Cumulative Index of Contributing Authors, Volumes 25­33 663
Cumulative Index of Chapter Titles, Volume 25­33 666
ERRATA
An online log of corrections to Annual Review of Anthropology
chapters may be found at http://anthro.annualreviews.org/errata.shtml
Annu.Rev.Anthropol.2004.33:585-623.Downloadedfromarjournals.annualreviews.org
byAppalachianStateUniversityon08/13/08.Forpersonaluseonly.