Nina G. Jablonskiand George ChaplinDepartment of Anthropology,California Academy ofSciences, Golden Gate Park,San Francisco, CA94118-4599, U.S.A. E-mails:njablonski@calacademy.org;gchaplin@calacademy.org

Received 25 March 1999Revision received 10 August1999 and accepted18 February 2000

Keywords: ultravioletradiation, UVB, melanin,integument, folate,photoprotection, neural tubedefects, vitamin D,depigmentation,thermoregulation, raceconcepts.

The evolution of human skin coloration

Skin color is one of the most conspicuous ways in which humans varyand has been widely used to define human races. Here we presentnew evidence indicating that variations in skin color are adaptive, andare related to the regulation of ultraviolet (UV) radiation penetrationin the integument and its direct and indirect effects on fitness. Usingremotely sensed data on UV radiation levels, hypotheses concerningthe distribution of the skin colors of indigenous peoples relative to UVlevels were tested quantitatively in this study for the first time.

The major results of this study are: (1) skin reflectance is stronglycorrelated with absolute latitude and UV radiation levels. The highestcorrelation between skin reflectance and UV levels was observed at545 nm, near the absorption maximum for oxyhemoglobin, sug-gesting that the main role of melanin pigmentation in humans isregulation of the effects of UV radiation on the contents of cutaneousblood vessels located in the dermis. (2) Predicted skin reflectancesdeviated little from observed values. (3) In all populations for whichskin reflectance data were available for males and females, femaleswere found to be lighter skinned than males. (4) The clinal gradationof skin coloration observed among indigenous peoples is correlatedwith UV radiation levels and represents a compromise solution tothe conflicting physiological requirements of photoprotection andvitamin D synthesis.

The earliest members of the hominid lineage probably had a mostlyunpigmented or lightly pigmented integument covered with darkblack hair, similar to that of the modern chimpanzee. The evolutionof a naked, darkly pigmented integument occurred early in theevolution of the genus Homo. A dark epidermis protected sweatglands from UV-induced injury, thus insuring the integrity of somaticthermoregulation. Of greater significance to individual reproductivesuccess was that highly melanized skin protected against UV-inducedphotolysis of folate (Branda & Eaton, 1978, Science 201, 625–626;Jablonski, 1992, Proc. Australas. Soc. Hum. Biol. 5, 455–462, 1999,Med. Hypotheses 52, 581–582), a metabolite essential for normaldevelopment of the embryonic neural tube (Bower & Stanley, 1989,The Medical Journal of Australia 150, 613–619; Medical ResearchCouncil Vitamin Research Group, 1991, The Lancet 338, 31–37) andspermatogenesis (Cosentino et al., 1990, Proc. Natn. Acad. Sci.U.S.A. 87, 1431–1435; Mathur et al., 1977, Fertility Sterility 28,1356–1360).

As hominids migrated outside of the tropics, varying degrees ofdepigmentation evolved in order to permit UVB-induced synthesis ofprevitamin D3. The lighter color of female skin may be required topermit synthesis of the relatively higher amounts of vitamin D3

necessary during pregnancy and lactation.Skin coloration in humans is adaptive and labile. Skin pigmentation

levels have changed more than once in human evolution. Because ofthis, skin coloration is of no value in determining phylogeneticrelationships among modern human groups.

� 2000 Academic Press

Journal of Human Evolution (2000) 39, 57–106doi:10.1006/jhev.2000.0403Available online at http://www.idealibrary.com on

0047–2484/00/070057+50$35.00/0 � 2000 Academic Press

58 . . .

Introduction

Melanin accounts for most of the variationin the visual appearance of human skin.Variation in cutaneous melanin pigmen-tation in humans has been attributed tomany factors, with most authorities agreeingthat the observed variations reflect biologicaladaptations to some aspect of the environ-ment. Many hypotheses have centered onthe role of melanin pigmentation in theregulation of the penetration of sunlightand, more specifically, ultraviolet (UV)radiation, into the skin. The more highlymelanized skins of indigenous tropicalpeoples have been said to afford greaterprotection against the deleterious effects ofUV radiation, such as sunburn, skin cancer(Fitzpatrick, 1965) and nutrient photolysis(Branda & Eaton, 1978). The more lightlypigmented skins of peoples inhabitinglatitudes nearer the Arctic have beenexplained as adaptations to the lower UVradiation regimes of those regions and theimportance of maintaining UV-induced bio-synthesis of vitamin D3 in the skin (Murray,1934; Loomis, 1967). Other adaptationisthypotheses have emphasized the role ofskin pigmentation in regulating sensitivity tofrostbite (Post et al., 1975a), in diseaseprevention (Wassermann, 1965, 1974),thermoregulation (Roberts & Kahlon, 1976;Roberts, 1977), concealment (Cowles,1959), or combinations of these factors(Roberts & Kahlon, 1976; Roberts, 1977).

The concept of integumentary pigmen-tation as an environmental adaptation hasnot been universally accepted, however.Some authorities have argued that becauseUV-radiation-induced sunburn or skincancers rarely affect reproductive success,melanin pigmentation must be consideredonly slightly adaptive or nonadaptive (Blum,1961; Diamond, 1991). It has also beenargued that because the relationshipbetween skin color and sunlight appearsimperfect, skin color should not be viewed

only as an adaptation to the environment(Diamond, 1991). Integumentary pigmen-tation has also been regarded as a pleiotropicbyproduct of selection acting on first func-tions of pigmentation genes, such as theregulation of metabolic pathways (Deol,1975), or as a characteristic controlled bysexual selection for the enhancement ofattractiveness to the opposite sex (Diamond,1988, 1991).

In this paper we demonstrate that thepreviously observed relationship betweenskin pigmentation in indigenous humanpopulations and latitude is traceable to thestrong correlation between skin color andUV radiation. We also present evidence sup-porting the theory that variations in melaninpigmentation of human skin are adaptiveand that they represent adaptations for theregulation of the effects of UV radiation ondeep strata of the integument. Our presen-tation combines analysis and visualization ofdata on the skin coloration of indigenouspeoples coupled with remote sensing dataon UV radiation levels at the Earth’s surfaceand physiological evidence concerning theeffects of UV radiation on levels of essentialvitamins and metabolites.

What skin color was primitive for thehominid lineage?

Before questions about changes in integu-mentary pigmentation in modern humanevolution can be addressed, considerationmust be given to the probable primitivecondition of the integument in the earliestmembers of the human lineage. It is likelythat the integument of the earliest proto-hominids was similar to that of our closestliving relative, the chimpanzee, being whiteor lightly pigmented and covered with darkhair (Post et al., 1975b). In the chimpanzee,exposed areas of skin vary considerably intheir coloration depending on the speciesand subspecies under consideration, but inall groups facial pigmentation increases with

59

UV radiation at the surface of theEarth

The sun emits electromagnetic radiationfrom short-wavelength X-rays to long-wavelength radio waves. Because X-rays andthe shortest UV waves (UVC, <280 nm) donot penetrate the atmosphere, middle(UVB, 280–320 nm) and long (UVA 320–400 nm) wavelength UV radiation, alongwith the visible wavelengths and infrared,are of greatest biological significance.

In order for solar radiation to produce aphotochemical effect, it must be absorbed.For an effective photon of radiation to beabsorbed, it must first travel from the sun tothe body’s surface through the atmosphere.Atmospheric absorption and scattering oflight are functions of the air mass throughwhich light must travel (Daniels, 1959). Theair mass, in turn, is a function of the angle ofthe sun to the observer, which is determinedby the time of day, season, latitude andaltitude. The air mass has a great influenceon the transmission of UV radiationbecause, in general, scattering is greatest inthe UV and shorter visible wavelengths(Daniels, 1959). The transmission of UVand visible radiation to the Earth’s surfaceare also determined by absorption in theozone layer, clouds, dust, haze and variousorganic compounds (Daniels, 1959, 1964).

Prior to the utilization of remote sensingtechnology, the intensity of UV radiation atthe Earth’s surface and the erythemalresponse of human skin to UV radiationwere estimated using mathematical models(Paltridge & Barton, 1978). Although thesemodels were parameterized to account forthe effects of solar elevation, the amounts ofozone and aerosols in the atmosphere,and the surface albedo on estimated UV

age and exposure to UV radiation (Postet al., 1975b). Except for the face, eyelids,lips, pinnae, friction surfaces, and anogenitalareas, the epidermis of most nonhumanprimates is unpigmented due to an absenceof active melanocytes (Montagna &Machida, 1966; Montagna et al., 1966a,b),suggesting that this is the primitive condi-tion for primates in general. The hairlessareas listed above are pigmented to greateror lesser extents in all primate species(Montagna & Machida, 1966; Montagnaet al., 1966a,b), suggesting that the potentialfor induction of melanogenesis (Erickson &Montagna, 1975) in exposed skin is alsoprimitive for the group.

Physiological models have demonstratedthat the evolution of hairlessness and anessentially modern sweating mechanismwere coordinated with the higher activitylevels associated with the modern limbproportions and striding bipedalism(Montagna, 1981; Schwartz & Rosenblum,1981; Wheeler, 1984, 1996; Chaplin et al.,1994). Throughout this transitional period,the critical function of the integument inthermoregulation was maintained throughevolution of an increased number of sweatglands, particularly on the face (Cabanac &Caputa, 1979; Falk, 1990), that increasedthe maximum rate of evaporative coolingavailable at any one time (Wheeler, 1996;see also Mahoney, 1980). The brain isextremely heat sensitive, and its temperatureclosely follows arterial temperature (Nelson& Nunneley, 1998). Evolution of a whole-body cooling mechanism capable of finelyregulating arterial temperature was, there-fore, a prerequisite for brain expansionand increased activity levels. Naked skinitself affords a thermoregulatory advantagebecause it makes for a reduced total ther-mal load requiring evaporative dissipation(Wheeler, 1996). As the density of body hairdecreased and the density of sweat glandsincreased, the need for protection of sub-epidermal tissues against the destructive

effects of UV radiation, particularly UVB,also increased. This protection was accom-plished by an increase in melanization of theskin.

60 . . .

radiation, they were merely models, notdirect measurements. With the applicationof remote sensing to this problem, however,it has been possible to directly measure theUV radiation reaching the Earth’s surface,taking ozone concentration and scene reflec-tivities (cloud conditions, and snow and icecover) into account (Herman & Celarier,1996). The existence of directly measureddata means that it is possible to know exactlyhow much UV radiation made it to theEarth’s surface at a specific day and time; itis no longer necessary to rely on calculatedestimates. The present study is the first inwhich direct measurements of UV radiationat the Earth’s surface have been used to testhypotheses concerning the evolution ofhuman skin pigmentation.

Relethford (1997) recently suggested thathemispheric differences in human skin colorwere due to hemispheric differences in UVradiation. In response to this claim, we havedemonstrated, using direct measurements ofUV radiation, that the annual erythemalmeans for UV radiation in the northern andsouthern hemispheres were not significantlydifferent, but that significant differencesbetween the hemispheres are found at thesummer and winter solstices (Chaplin &Jablonski, 1998). The magnitude of thesedifferences can be predicted from today’sperihelion effect (i.e., that the Earth isclosest to the sun on 3 January, during theAustral Summer). We have also shown(Chaplin & Jablonski, 1998) that thehemispheric differences in skin color thatRelethford demonstrated are due in largepart to the fact that, in the southern hemi-sphere, the areas receiving high annual UVradiation constitute a much larger percent-age of the total habitable land area than inthe northern hemisphere. In other words, alarger proportion of the habitable land ofthe southern hemisphere lies closer to theequator than it does in the northern hemi-sphere. This area receives high annual dosesof UV radiation. In contrast, the vast

majority of habitable land in the northernhemisphere lies north of the Tropic ofCancer and receives low annual doses ofUV radiation. The hemispheric bias in thelatitudinal distribution of land masses hashad profound consequences for the evol-ution of skin pigmentation for the humansinhabiting the two hemispheres (Chaplin &Jablonski, 1998).

Melanin and the effects of UVradiation on human skin

In the skin, melanin acts as an optical andchemical photoprotective filter, whichreduces the penetration of all wavelengthsof light into subepidermal tissues (Daniels,1959; Kaidbey et al., 1979). Electro-magnetic radiation impinging on the humanskin can undergo a number of interactions:it can be reflected by the surface of thestratum corneum; it can enter the stratumcorneum after a slight change in its directionof travel; it can interact with melanin dust inthe stratum corneum, resulting in partial ortotal absorption; it can traverse deeper layersand experience a possible additional scatter-ing; or it can enter the viable epidermiswhere it encounters melanin packagedwithin melanosomes (Kollias et al., 1991).Epidermal melanin has different forms atdifferent sites within the skin and theseinteract differently with radiation (Kolliaset al., 1991). At all of these sites, melaninworks an optical filter to attenuate radiationby scattering (Kollias et al., 1991). It alsoacts as a chemical filter through its functionas a stable free radical that can absorbcompounds produced by photochemicalaction which would be toxic or carcinogenic(Daniels, 1959; Kollias et al., 1991). Themelanin dust of the stratum corneumappears to be a degradation product ofthe melanosomes, the organelles in whichthe melanin pigment resides. This serves thehighly desirable function of attenuatingradiation close to the skin’s surface, thus

61

sparing the deeper, viable layers. Most epi-dermal melanin is packaged in intactmelanosomes that are located deeper in theskin, however. These organelles are dis-tributed to keratinocytes throughout theMalpighian layer of the epidermis by thedendritic processes of melanocytes. Thus,the superior photoprotection of highly mela-nized skin is accomplished by absorptionand scattering, which are influenced by thedensity and distribution of melanosomeswithin keratinocytes in the basal and para-basal layers of the epidermis (Kaidbey et al.,1979) and by the presence of specks ofmelanin dust in the stratum corneum(Daniels, 1964). The heavily pigmentedmelanocytes of darker skins also have theability to resume proliferation afterirradiation with UVB radiation, indicatingthat they can recover more quickly from thegrowth-inhibitory effects of UVB exposure(Barker et al., 1995).

Most studies of the effects of UV radiationon human skin have utilized as a standardthe minimum-erythemal dose (UVMED),which is the quantity of UV radiationrequired to produce a barely perceptiblereddening of lightly-pigmented skin.

Apart from its beneficial role in vitamin Dsynthesis, the effects of UVB radiation onthe skin are universally harmful. Suppres-sion of sweating and subsequent disruptionof thermoregulation due to sunburn-induced damage to sweat glands are themost serious immediate effects of excessiveUVB exposure (Daniels, 1964; Pandolfet al., 1992). Other short-term harmfuleffects include discomfort, lowering of thepain threshold, vesiculation and possiblesecondary infection following severe sun-burn, desquamation and nutrient photolysis(Branda & Eaton, 1978; Daniels, 1964).Degenerative changes in the dermis andepidermis due to UVB exposure are realizedover longer time courses, some eventuallyresulting in skin cancers. The question froman evolutionary point of view is do any of

these effects, singly or in combination, havedirect effects on reproductive success? Basalcell and squamous cell carcinomas, forinstance, while a significant cause of mor-bidity in lightly pigmented peoples, rarelyinfluence fitness because they generallyaffect individuals after reproductive age andare rarely fatal (Blum, 1961; Roberts, 1977;Robins, 1991). Malignant melanomas,though rare (representing 4% of skin cancerdiagnoses) are more often fatal than otherskin cancers, but their effect on reproductivesuccess is still limited because of their lateonset, well beyond the age of first reproduc-tion (Johnson et al., 1998).

We argue here that protection againstnutrient photolysis, and specifically photo-lysis of folate, was a prime selective agentwhich brought about the evolution of deeplypigmented skins among people living underregimes of high UVB radiation throughoutmost of the year because of the directconnection between folate and individualreproductive success. The importance toindividual fitness of protection of sweatglands and maintenance of thermoregu-latory capability is also seen as contributingto increased melanization.

Folate and human reproductivesuccess

Folic acid is an essential nutrient, which isrequired for nucleotide and, therefore, DNAbiosynthesis. Folic acid is converted in thebody to its conjugated form, folate. The lackof folate has long been known to bring abouta macrocystic megaloblastic anemia becausefolate is required for the maturation ofbone marrow and the development of redblood cells. Folate deficiency in nonhumanmammals has also been shown to producemultiple fetal anomalies including malfor-mations of the eye, central nervous system,palate, lip, gastrointestinal system, aorta,kidney and skeleton (Omaye, 1993, andreferences cited therein). All of these

62 . . .

problems can ultimately be traced to folate’sroles in purine and pyrimidine biosynthesis(Omaye, 1993; Fleming & Copp, 1998).

It has recently been confirmed that thereis a connection between defective folatemetabolism and neural tube defects (NTDs)in humans (Fleming & Copp, 1998; Bower& Stanley, 1989; Medical Research CouncilVitamin Research Group, 1991). Neuraltube defects comprise a family of congenitalmalformations that result from incompleteneurulation and that are expressed asdeformities of varying severity. In cranio-rachischisis totalis, the entire neural tube failsto close; embryos fail early in their develop-ment and are spontaneously aborted.Failure of the cranial neuropore to fuseresults in anencephalus or craniorachischisis,in which the brain is represented by anexposed dorsal mass of undifferentiatedtissue. Anencephalic embryos often surviveinto late fetal life or term, but invariably diesoon after birth. When failure of neural tubeclosure disrupts the induction of the over-lying vertebral arches, the resulting con-dition of an open neural canal is called spinabifida. Cases of spina bifida vary in theirseriousness. The most disabling involve theprotrusion of neural tissue and meningesthrough an open neural canal (meningo-myelocele). The least serious involve failureof a single vertebral arch to fuse with noherniation of underlying neural tissue (spinabifida occulta).

Since records have been kept in developedcountries, anencephalus and spina bifida havebeen found to be particularly common inlight-skinned populations. These defectsaccounted for as much as 15% of all peri-natal and 10% of all postperinatal mortalityin the worst-affected populations prior tothe introduction of preventative nutritionalsupplementation (Elwood & Elwood, 1980).Since the advent of prenatal diagnosis, theprevalence of NTDs in relation to all con-ceptions has been shown to be significantlyhigher than birth prevalence (Velie & Shaw,

1996; Forrester et al., 1998), partly becauseof the high rate of early miscarriage in thecase of the most serious defects (C. Bower,personal communication).

Folic acid prevents 70% of NTDs inhumans and is thought to act by directnormalization of neurulation through regu-lation of pyrimidine biosynthesis neces-sary for DNA production (Minns, 1996;Fleming & Copp, 1998). In the body, folateis most sensitive to two major environmentalagents, ethanol and UV radiation. In thecase of ethanol, folate deficiency is a well-documented problem of chronic alcoholics(Tamura & Halsted, 1983). The photo-sensitivity of folate is significant, and thenutrient can be readily degraded bynatural sunlight or UV light (Kaunitz &Lindenbaum, 1977). Photolysis of folate inhumans has been demonstrated by a signifi-cant decline in folate levels when serum(in vitro) or light-skinned subjects (in vivo)were exposed to simulated natural sunlight(Branda & Eaton, 1978). A causal relation-ship between in vivo folate photolysis inhumans and NTDs, first suggested byJablonski (1992, 1999), has recently beensupported by the report of NTDs in threeunrelated subjects whose mothers had beenexposed to UV light on the sun beds oftanning salons during the first weeks oftheir pregnancies (Lapunzina, 1996). Folatedeficiencies brought about by UV photolysishave also been implicated in the etiology ofNTDs in some amphibian populations(Jablonski, 1998) because such defects havebeen seen in amphibian embryos exposedto UV radiation (Higgins & Sheard, 1926;Licht & Grant, 1997).

In addition to its important role in ensur-ing embryonic survival through properneurulation, folate has been shown to becritical to another important process centralto reproduction, spermatogenesis. In bothmice (Cosentino et al., 1990) and rats(Mathur et al., 1977), chemically inducedfolate deficiency resulted in spermatogenic

63

arrest and male infertility, findings whichprompted investigations of antifolate agentsas male contraceptives in humans.

Thus, through folate’s roles in the survivalof embryos through normalization ofneurulation and maintenance of malefertility, and its involvement in a range ofother physiological processes dependent onnucleotide biosynthesis, regulation of folatelevels appears to be critical to individualreproductive success. Folate levels inhumans are influenced by dietary intake offolic acid and by destructive, exogenousfactors such as UV radiation. Therefore, thesolution to the evolutionary problem ofmaintaining adequate folate levels in areasof high UV radiation involved the ingestionof adequate amounts of folic acid in the dietand protection against UV radiation-induced folate photolysis. The latter wasaccomplished by increasing the concen-tration of the natural sunscreen, melanin,in the skin. The low prevalences of severefolate deficiency (Lawrence, 1983;Lamparelli et al., 1988) and NTDs (Carter,1970; Elwood & Elwood, 1980; Buccimazzaet al., 1994; Wiswell et al., 1990; Shaw et al.,1994) observed among native Africans andAfrican Americans, even among individualsof marginal nutritional status, are probablydue to a highly melanized integument,which protects against folate photolysis.

Skin pigmentation and vitamin Dsynthesis

Vitamin D3 is essential for normal growth,calcium absorption and skeletal develop-ment. Deficiency of the vitamin can causedeath, immobilization, or pelvic deformitieswhich prevent normal childbirth (Neer,1975). Requirements for vitamin D3 areelevated in females during pregnancy andlactation because of the need for enhancedmaternal absorption of calcium in order tobuild the fetal and neonatal skeletal system(Whitehead et al., 1981; Kohlmeier &

Marcus, 1995; Brunvand et al., 1996). Inmost humans, casual exposure to sunlightleads to conversion of 7-dehydrocholesterolin the skin to previtamin D3 by UV photonsand subsequent isomerization of the latter tovitamin D3 at body temperature (Loomis,1967; Webb et al., 1988). Increasing themelanin in human skin increases the lengthof exposure to UV light that is neededto maximize synthesis of previtamin D3

(Holick et al., 1981). Deeply melanizedskins become nonadaptive under conditionswhere the concentration of melanin is toohigh to permit sufficient amounts of vitaminD3 precursor to be synthesized in the skinunder conditions of available UV radiation(Loomis, 1967; Holick et al., 1981; Clemenset al., 1982). If the duration of UV exposureis not sufficient to catalyze previtamin D3

synthesis, individuals are at much higher riskof vitamin D3 deficiency and its manifes-tations (rickets, osteomalacia, and osteo-porosis), as has been demonstrated byrecent migrants from Ethiopia to Israel(Fogelman et al., 1995) or from the Indiansubcontinent to urban centers in the U.K.(Henderson et al., 1987).

Following Murray (1934), Loomis (1967)argued that depigmentation of the skin was anecessary adaptation for humans attemptingto inhabit regions outside the tropics,especially those north of 40�N, whichreceive low average amounts of UV radi-ation throughout the year. In addition, headvocated the position that deeply pig-mented skin was also an adaptation forphysiological regulation of vitamin D3

levels. To Loomis, a highly melanizedintegument prevented UV-radiation-induced vitamin D3 toxicity caused byover-synthesis of previtamin D3. Both ofLoomis’s claims about integumentary pig-mentation in humans and vitamin D3 levelshave been challenged.

The hypothesis that deeply pigmentedskin protects against hypervitamosis Dunder conditions of high UV radiation has

64 . . .

been conclusively disproven. Vitamin Dtoxicity is prevented by a ceiling effect onprevitamin D3 synthesis combined within vivo photolysis of vitamin D3 itself(Holick et al., 1981; Holick, 1987). VitaminD3 intoxication can occur as the result ofindividuals overdosing on vitamin supple-ments, but no cases of naturally occurringvitamin D3 intoxication have ever beenreported (Robins, 1991).

Loomis’s central hypothesis concerningdepigmentation of the integument ofhumans living at high latitudes as a require-ment for adequate vitamin D synthesis hasbeen most strongly challenged by Robins(1991). Robins’ main points are that earlyHomo at high latitudes was exposed to thefull impact of the natural environment. Evenwhen clad with animal skins and furs, heclaims that enough of the body’s surfacewould be exposed to permit synthesis ofadequate amounts of previtamin D3. Whilehe admits that winters in Europe duringglacials would have been particularly coldand dim, he indicates that the late springand summer would have afforded goodopportunities for hominids to expose theirskin to UV radiation. According to Robins(1991), the long, dull winters would nothave triggered widespread hypovitamosis Dand rickets because vitamin D can be storedin fat and muscle (Rosenstreich et al., 1971;Mawer et al., 1972). He suggests thatvitamin D deficiencies are a product of‘‘industrialization, urbanization and over-population’’ (Robins, 1991:207) and sup-ports his claim by pointing out that themajority of modern human populations thatsuffer from such problems are urban dwell-ers, deprived of natural sunlight and theopportunity to synthesize previtamin D3.According to Robins, dark-skinned humanpopulations living under conditions of lowannual UV radiation do not suffer fromvitamin D deficiencies that can be attributedto lack of sunlight. The depigmentation con-spicuous in peoples inhabiting the northern

hemisphere above 40�N cannot be traced,according to Robins, to the need to synthe-size vitamin D in their skin. Rather, peoplewith deeply pigmented skin can survive wellunder conditions of low annual UV radi-ation provided that they spend sufficienttime outdoors in the spring and summer tobuild up physiological stores of vitamin Dfor the winter (Robins, 1991). As is dis-cussed below, the findings of the currentstudy and the weight of current clinicalevidence do not support Robins’ counter-arguments against the vitamin D hypothesisof depigmentation.

Testing the relationship between skincolor and levels of UV radiation

Previous studies of the relationship betweenenvironmental variables and the skin colorof indigenous populations using skin re-flectance spectrophotometry demonstratedhighest associations with latitude (Roberts &Kahlon, 1976; Roberts, 1977; Tasa et al.,1985). These were interpreted as reflectingthe crucial role of UV radiation in determin-ing skin color. Until this time, accurate testsof this relationship were not possiblebecause accurate data on UVMED levelsat the Earth’s surface were not available.Because these data are now available(Herman & Celarier, 1996), we were able toapply them to two critical analyses: (1) inthe determination of the geographic distri-bution of the potential for previtamin D3

synthesis; and (2) in assessment of the re-lationship between surface UV radiationlevels and skin color reflectances for indig-enous populations.

Methods

Skin reflectance dataA database of skin reflectance data wascompiled from the literature. The databasecomprises samples designated as male,female, and both sexes, for reflectance at

65

425 nm (blue filter), 545 nm (green filter)and 685 nm (red filter), for the upper innerarm site. In many primary reports, inter-mediate wavelength reflectance readingswere also provided, but these were not used.The complete skin reflectance data set withbibliographic sources is presented in theAppendix. The vast majority of these datahad been collected with the E.E.L. (EvansElectric Limited) reflectometer. Readingsfor Colombian natives were the only onestaken using the Photovolt abridged spectro-photometer. These measurements were con-verted to E.E.L.-comparable values by usingthe regression equation for Belize Creolesdeveloped by Lees & Byard (1978) desig-nated in the Appendix. As is obvious froman inspection of the collated data, the skinreflectance data set is far from complete andis strongly skewed toward representation ofindigenous human populations from the OldWorld. The inherent deficiencies of thisdatabase limited the scope of analyses thatcould be undertaken.

Indigenous populations were defined forpurposes of this study as those which hadexisted in their current locations for a longtime prior to European colonization. Thenames by which areas or groups wereidentified were taken from the originalcitations. Populations known to have highlevels of admixture or to have recentlymigrated to their current locations wereexcluded. Some populations were repre-sented by males only, others by males andfemales, and others where the values for thetwo sexes had been averaged by the originalworkers. In some cases, the sex of the sub-jects was not stated in the original report.Because more skin reflectance data exist formales than for females, analyses of skinreflectance in which data for the sexes werecombined are slightly biased toward malereflectance values. In cases where more thanone skin reflectance was measured for aspecific population, either for each sex or bymultiple workers, the means for each sex

were averaged, except in the cases notedbelow where the reflectance values for thesexes were analyzed separately.

Reflectance data were specified in orig-inal reports at varying levels of geographicaccuracy. Those specified to a small geo-graphic area such as township or countycould be matched to a small number ofaverage UVMED cell values. For thosespecified at the regional or country level, asingle skin reflectance value was associatedwith multiple UVMED cell values, depend-ing on the size of the region. These variedfrom one cell (e.g., Germany, Mainz;India, Goa) to 453 cells (Zaire). To over-come the problem of cell replication andartificial exaggeration of variance, theUVMED data for each specified area wasaveraged for that area to produce a singlevalue. This data reduction exercise made itpossible for a single skin reflectance valueto be evaluated relative to a single averageUVMED value.

Because Relethford (1997) has suggestedthat a different relationship between UVradiation and skin reflectance exists betweenthe northern and southern hemispheres,separate analyses were performed for bothhemispheres together and for each hemi-sphere separately. (The only analysis inwhich this was not done was that comparingthe significance of differences in skin reflect-ance between the sexes. See below.)Although the annual UVMED is equal forboth hemispheres, the seasonal distributionof UV radiation is different (Chaplin &Jablonski, 1998). The hemispheres also dif-fer in their respective land surface area(Chaplin & Jablonski, 1998) and degree ofland mass connectivity. It is impossible, forinstance, for non-seafarers to move aroundthe globe in the southern hemisphere.Lastly, the proportion of different vegetationzones between the hemispheres is different,with the northern hemisphere exhibitingmore extensive areas of forest, desert andmountains.

66 . . .

Calculation of annual average UVMEDThe annual average values for UVMEDwere derived from readings taken from theNASA Total Ozone Mapping Spectrometer(TOMS) which was flown aboard theNimbus-7 satellite between 1978 and 1993(Herman & Celarier, 1996). The data rep-resent the relative daily areal exposure of UVradiation effective in causing skin irritation,computed at each 1·25 degree longitude byone degree latitude pixel, between 65�S and65�N; the solar flux was measured at noon(Herman & Celarier, 1996). These readingswere computed to account for the totalozone column and scene reflectivities (cloudand snow cover) in the same latitude–longitude pixel. These measurements werethen combined with the results of radiativetransfer calculations, terrain height, solarzenith angle and the model action spectrumof erythemal sensitivity for caucasoid skin(Herman & Celarier, 1996). The wave-length range of 280–400 nm sampled bythe TOMS ensured that all significant con-tributions to the erythemal exposure wereincluded in the UVMED. Because theaction spectrum is defined up to an arbitrarymultiplicative constant, the basic UVMEDdata are of arbitrary dimensions.

The original dataset was very large (over190,000 data points representing 37,400readings taken each day from 1979–1992).Therefore, an abridged dataset (the annualaverage UVMED) was produced by takingan average for all years of the average of the21st to the 23rd days of each month.

All data analyses in this study were under-taken using S-Plus� 2000 and ArcView�,reprojected in Mercator projection.

Analysis and visualization of the potential forvitamin D3 synthesisThe rate of previtamin D3 synthesis inlightly pigmented human skin exposed tonatural sunlight has been experimentallydetermined (Webb et al., 1988). Working inBoston, Massachusetts in 1986, Webb et al.

(1988) found that the earliest occurrence inthe year of previtamin D3 synthesis in neo-natal caucasoid foreskin samples was on 17March. Because the actual UVMED for thatplace and time was available from the NASATOMS dataset, values representing thepotential for vitamin D3 synthesis relative toannual average UVMED could be calcu-lated for all locations included in the NASATOMS dataset. It was then possible todefine distinct geographic zones represent-ing the differing potentials for vitamin D3

synthesis for light-skinned individuals. Theannual average UVMED and average skinreflectance measurement for each zone werethen calculated and compared. The skinreflectance data used for this analysis com-prised data for indigenous people (males,females, both sexes and unspecified sex),averaged by country and then by zone.

We also sought to demonstrate the poten-tial for vitamin D3 synthesis in the skin ofdark-skinned human populations. Thelength of time required for endogenousvitamin D3 biosynthesis increases withincreasing melanin concentration in the skin(Clemens et al., 1982; Holick et al., 1981).Thus, the data showing the time of exposureto UV radiation necessary to maximize pre-vitamin D3 formation in different humangroups were used to calculate the geographi-cal zones in which annual UVMED was notsufficient, averaged over the year, to catalyzeprevitamin D3 synthesis in moderately andhighly melanized skin.

Sexual differences in skin reflectanceIn order to test the hypothesis that femalesand males of the same populations differ intheir average skin reflectance, we selecteddata from specific population where skinreflectance measurements had been takenand reported separately for males andfemales. A standard two-sample t-test wasused to determine the significance of thedifferences between the sexes at the samelocality. Separate analyses for the northern

67

and southern hemispheres were not under-taken because the small sizes of the data-sets that met the above criteria wouldhave rendered hemisphere-specific analysesmeaningless. Further, our objective here wasnot to test the hypothesis that there is adifference between the hemispheres in thedegree of differentiation between the twosexes in skin reflectance.

The relationship of annual average UVMEDto skin reflectanceA correlation matrix was used to test thestrength of the relationships between annualaverage UVMED, latitude and skin re-flectance. This analysis was undertakenspecifically to determine the strength of thecorrelation of skin reflectance to UVMEDrelative to its correlation with latitude. Theskin reflectance data used in this analysiscomprised those from indigenous popu-lations, all-sex samples combined, averagedby country. For this analysis, latitude wastransformed to absolute latitude, i.e., theabsolute value of the latitude, whichexpresses the angle of a location from theEquator rather than relative north andsouth. It is unclear from previous studies ofhuman skin pigmentation as to whetherworkers utilized conventional latitude orabsolute latitude for purposes of correlationor regression analyses. Transformation oflatitude to absolute latitude is necessarybecause of the nonlinear relationshipbetween solar insolation and latitude. Thetransformation to absolute latitude rendersthis relationship more linear, and thus yieldsmuch higher correlations.

The relationship between UVMED andskin color reflectance was explored ingreater detail using a least squares regres-sion. For this analysis, the skin reflectancedata used were the same as those employedin the correlation matrix analysis describedabove. Separate regressions were developedfor data from each hemisphere, for each

widely used skin reflectance filter (425, 545and 685 nm).

Predicted vs. observed skin reflectancesIn order to compare predicted vs. observedvalues for the skin reflectances of indigenouspeoples, the largest available dataset forobserved skin reflectance (at 685 nm) wasused. It is important to note again that somepopulations were represented by males only,as both sexes combined, or as unspecifiedsex, as discussed. Although these variationsin sexual classification of the raw data addedto the expected variance of the pooled data-set, this problem could not be avoided.Significant outliers from the regression linewere then identified.

In order to construct a map of predictedskin colors for modern terrestrial environ-ments, a regression was computed betweenannual average UVMED and the observedskin reflectance. The observed reflectancesfor indigenous populations were based on allavailable data for a particular area or group.A regression equation derived from a largerdata set including values for males, females,both sexes and samples of unknown sex wasused:

Predicted skin color=annual average UVMED

(��0·1088)+72·7483.

Results

Analysis and visualization of the potential forvitamin D3 synthesisThree zones representing different poten-tials for UV-induced vitamin D3 synthesisin light-skinned humans were identified(Figure 1). The annual average UVMEDand average skin reflectance measurementfor each zone are presented in Table 1.Zone 1 was delimited as the area in whichthe average daily UVMED was sufficient tocatalyze previtamin D3 synthesis throughoutthe year. This zone comprises the area from

68 . . .

Fig

ure

1.T

hepo

tent

ial

for

synt

hesi

sof

prev

itam

inD

3in

light

lypi

gmen

ted

hum

ansk

inco

mpu

ted

from

annu

alav

erag

eU

VM

ED

.T

hehi

ghes

tan

nual

valu

esfo

rU

VM

ED

are

show

nin

light

viol

et,w

ith

incr

emen

tally

low

erva

lues

inda

rkvi

olet

,the

nin

light

toda

rksh

ades

ofbl

ue,o

rang

e,gr

een

and

gray

(64

clas

ses)

.W

hite

deno

tes

area

sfo

rw

hich

noU

VM

ED

data

exis

t.M

erca

tor

proj

ecti

on.

Inth

etr

opic

s,th

ezo

neof

adeq

uate

UV

radi

atio

nth

roug

hout

the

year

(Zon

e1)

isde

limit

edby

bold

blac

klin

es.L

ight

stip

plin

gin

dica

tes

Zon

e2,

inw

hich

ther

eis

nots

uffici

entU

Vra

diat

ion

duri

ngat

leas

tone

mon

thof

the

year

topr

oduc

epr

evit

amin

D3

inhu

man

skin

.Z

one

3,in

whi

chth

ere

isno

tsu

ffici

ent

UV

radi

atio

nfo

rpr

evit

amin

D3

synt

hesi

son

aver

age

for

the

who

leye

ar,

isin

dica

ted

byhe

avy

stip

plin

g.

69

approximately 5� north of the Tropic ofCancer to approximately 5� south of theTropic of Capricorn. Here, previtamin D3

could be synthesized in lightly pigmentedskin as a result of casual exposure to UVBthroughout all months of the year. Zone 2was defined as the area in which the aver-age daily UVMED for at least one monthwas not sufficient to produce previtaminD3 in lightly pigmented skin. (This zone isindicated by light stippling in Figure 1.)Large areas of the northern hemisphere fallwithin this zone. Zone 3 was defined as thearea in which the daily UVMED, averagedover the whole year, was not sufficient tocatalyze previtamin D3 synthesis in lightlypigmented skin. That is, it was the zone inwhich previtamin D3 synthesis resultingfrom UVB exposure, as averaged over thecourse of a year, was not sufficient tomeet minimum physiological requirements.(This zone is indicated by heavy stipplingin Figure 1.) Although there were somemonths around the summer solstice duringwhich there was sufficient UV radiationto bring about cutaneous previtamin D3

synthesis in Zone 3 (and the attendantpotential for vitamin D storage), the aver-age daily dose, when averaged over thecourse of a year, was not sufficient tocatalyze the reaction. In other words, in-sufficient ambient UV radiation is availableon a year by year (or long-term) basis to

permit adequate cutaneous biosynthesis ofprevitamin D3.

It is important to note that the zonesdefined in Figure 1 were calculated on thebasis of the potential for previtamin D3

synthesis in light skin. No comparable datawere available for dark skin. Data concern-ing the times of exposure necessary to maxi-mize previtamin D3 synthesis in skinsamples with different melanin concen-trations are known, however. These wereused to calculate the geographical zones inwhich annual UVMED was not sufficient,averaged over the year, to catalyze pre-vitamin D3 synthesis in moderately andhighly melanized skin (Table 2; Figure 2).Figure 2 is a comparison of the estimatedpositions of Zone 3 from Figure 1 for humanpopulations with lightly, moderately anddeeply melanized skin. The fact that for-mation of previtamin D3 takes more thanfive times as long in highly melanized (TypeVI) skin as it does in lightly melanized (TypeIII) skin means that the continental areasthat are ‘‘vitamin D safe’’ for darker skinsare considerably smaller than they are forlighter skins (Figure 2).

Sexual differences in skin reflectanceFemales were found to be significantlylighter (i.e., displaying higher reflectancevalues) than males regardless of the filterused to measure the reflectance (Table 3).

Table 1 Annual average UVMED and averageobserved skin reflectance measurements at685 nm (red filter) for each of the zones depictedin Figure 1 and described in the text

Zone

Annualaverage

UVMED

Average observedskin reflectance

at 685 nm

1 313·2 37·22 158·3 55·03 68·1 68·9

UVMED is measured in units of arbitrary dimension(Herman & Celarier, 1996).

The relationship of annual average UVMEDto skin reflectanceWhen the relationship between annual aver-age UVMED and values of skin reflectancefor the combined all-sexes dataset wasexamined using a correlation matrix it wasfound that, for skin reflectance at all wave-lengths, the correlation with annual averageUVMED and absolute latitude were bothhigh (Table 4). The correlation betweenskin reflectance and annual averageUVMED was stronger than the correlationbetween skin reflectance and absolute

70 . . .

latitude in three instances: for the 545 nm(green) filter reflectance for the separatehemispheres and for the 685 nm (red) filterreflectance for the southern hemisphere.The correlation between skin reflectanceand annual average UVMED was onlyslightly weaker than the correlation betweenskin reflectance and absolute latitude in afurther four instances: for the 425 nm (blue)filter reflectance for the southern hemi-sphere and both hemispheres combined, forthe 545 nm (green) filter reflectance for bothhemispheres combined, and for the 685 nm(red) filter reflectance for both hemispherescombined. In only two cases [425 nm (blue)for the northern hemisphere and 685 nm(red) for the northern hemisphere] was thecorrelation between skin reflectance andabsolute latitude much higher than thatbetween skin reflectance and annual averageUVMED. The correlation between annualaverage UVMED and absolute latitudebased on data from both hemispheres com-bined was found to be �0·935, muchhigher than the correlation with untrans-formed latitude, �0·741.

When the relationship between annualaverage UVMED and values of skin reflect-ance was examined using a least squaresregression model (Table 5), the highest r2

values were associated with green (545 nm)filter reflectance.

Predicted vs. observed skin reflectancesThe differences between observed and pre-dicted skin reflectance values at 685 nm forindigenous populations were generally small(Table 6; Figures 3 and 4), and the differ-ence between the average observed and pre-dicted values was not significant (P=0·9983;two-sample t-test), and did not detract fromthe validity of the predictive model.

Discussion

The results presented here strongly supportthe theory that the degree of melanin pig-mentation in human skin is an adaptationfor the regulation of penetration of UVradiation into the epidermis.

Table 2 The effect of melanin concentration on the estimated annual average UVMED values that,averaged for the year, are not sufficient to catalyze previtamin D3 synthesis in human skin

Skin type(humanpopulation)

Time of exposure tomaximize pre D3

formation (h)(range and mean)

Multiplier foramount of time

relative to skin TypeIIIa, calculated from

mean time of exposure

Estimated annualaverage UVMED

necessary tocatalyze pre D3

formation

Type IIIa (European) 0·50–0·75 (0·625) 1 68·1Type V (Asian) 0·75–1·50 (1·125) 1·8 122·6Type VI (African American) 3·00–3·50 (3·25) 5·2 354

The annual average UVMED value for Zone 3 in Table 1 (for lightly pigmented skin) was multiplied by a factorrepresenting the increased number of hours necessary to catalyze previtamin D3 synthesis in more heavilypigmented skin types, based on the results of Holick and colleagues (Holick et al., 1981). This yielded an estimateof the minimum annual average UVMED necessary to catalyze previtamin D3 synthesis in darker skin types. Thedesignations of skin type and human population are from the original source.

The geographic distribution of the potential forvitamin D3 synthesisSkin pigmentation is determined as much bythe requirements of previtamin D3 synthesisas by photoprotection. Within Zone 1 asdefined in this study, a clinal distribution ofskin reflectance relative to annual averageUVMED can be detected from inspectionof the values presented in Table 5. A cline

71

Fig

ure

2.A

com

pari

son

ofth

ees

tim

ated

area

sin

whi

chan

nual

UV

ME

Dis

not

suffi

cien

t,av

erag

edov

erth

eye

ar,

toca

taly

zepr

evit

amin

D3

synt

hesi

sin

light

ly,

mod

erat

ely

and

high

lym

elan

ized

skin

.A

llzo

nes

wer

ede

fined

byth

eva

lues

for

prev

itam

inD

3sy

nthe

sis

pote

ntia

lpr

esen

ted

inT

able

2.W

idel

ysp

aced

obliq

ueha

chur

eco

vers

the

nort

hern

mos

tre

gion

ofth

eN

orth

ern

Hem

isph

ere

inw

hich

ther

eis

not

suffi

cien

tU

Vra

diat

ion,

aver

aged

over

the

enti

reye

ar,

toca

taly

zeth

efo

rmat

ion

ofpr

evit

amin

D3

inlig

htly

pigm

ente

d(T

ype

IIIa

)hu

man

skin

(Zon

e3

from

Fig

ure

1).

Nar

row

lysp

aced

obliq

ueha

chur

ede

note

sth

ear

ea,i

nad

ditio

nto

that

show

nby

wid

ely

spac

edob

lique

hach

ure,

inw

hich

ther

eis

not

suffi

cien

tU

Vra

diat

ion

toca

taly

zeth

efo

rmat

ion

ofpr

evit

amin

D3

inm

oder

atel

ym

elan

ized

(Typ

eV

)sk

in.T

hela

rge

circ

um-E

quat

oria

lare

ade

note

dby

stip

plin

gco

vers

the

area

,in

addi

tion

toth

epr

evio

ustw

o,in

whi

chth

ere

isno

tsu

ffici

ent

UV

radi

atio

nav

erag

edov

erth

een

tire

year

toca

taly

zeth

efo

rmat

ion

ofpr

evit

amin

D3

inhi

ghly

mel

aniz

ed(T

ype

VI)

skin

.

72 . . .

Table 3 Difference between the means of femaleand male skin reflectances for a combined dataset of indigenous human populations, based onsamples from individual populations for whichseparate data for males and females were avail-able

Filter nm

Femalemean

reflectance

Malemean

reflectance

Two-samplet-test

P-value

425 (blue) 19·20 16·88 0545 (green) 23·93 22·78 0·0089685 (red) 47·20 45·09 0

For reflectances at all wavelengths, females werefound to be lighter (exhibiting higher reflectancevalues) than males.

Table 4 Results of the correlation matrix per-formed to examine the relationships betweenskin reflectance (at three wavelengths), absolutelatitude and annual average UVMED

Filter nm by hemisphere(NH, SH or both)

Absolutelatitude

Annualaverage

UVMED

425 (blue), NH 0·974 �0·714425 (blue), SH 0·694 �0·668425 (blue), both 0·966 �0·954545 (green), NH 0·908 �0·917545 (green), SH 0·620 �0·967545 (green), both 0·964 �0·957685 (red), NH 0·844 �0·632685 (red), SH 0·423 �0·498685 (red), both 0·827 �0·809

Table 5 The results of a least squares regressionmodel in which the relationship between skinreflectance values for indigenous humanpopulations and average annual UVMED wasexamined, by hemisphere

Filter nm, byhemisphere (NH,SH or both) Intercept Slope r2

425 (blue), NH 34·032 �0·059 0·4572425 (blue), SH 25·517 �0·045 0·3402545 (green), NH 45·046 �0·083 0·7747545 (green), SH 44·193 �0·092 0·7013685 (red), NH 65·575 �0·073 0·3995685 (red), SH 66·908 �0·103 0·2485685 (red), both 72·797 �0·109 0·6551

For the 685 nm (red) filter reflectance, a value is alsopresented for combined hemispheric skin reflectanceand UVMED data; this value was used to generate thepredicted skin color values in Table 6 and Figure 3, andto compare with those published in previous studies(see Discussion).

of depigmentation is more noticeable inZone 2, where increasing depigmentationhas helped to ensure maintenance ofadequate vitamin D3 production in the skinthroughout most months of the year.Although individuals in this zone do notreceive enough UVB during at least onemonth each year to catalyze previtamin D3

synthesis, intake of vitamin D3-containingfoods and the ability of the body to storevitamin D3 mitigates this short lapse inprevitamin D3 biosynthesis. For humansliving in this zone, however, the adoptionof more housebound lifestyles with lowUV radiation exposure combined with

inadequate vitamin D3 intake in recent yearshas resulted in a high prevalence of hypo-vitaminosis D (Thomas et al., 1998). Thisproblem is exacerbated if the human inhab-itants of Zone 2 are relatively dark-skinned.The high prevalence of hypovitaminosisD, rickets, osteomalacia and osteoporosisamong populations recently migrated fromthe Indian subcontinent to the U.K. ishigher than that for the light-skinned generalpopulation, and shows a marked north–south gradient consistent with UV-radiationdosage dependence (Hodgkin et al., 1973;Henderson et al., 1987).

Humans inhabiting Zone 3 are at highestrisk of severe vitamin D3 deficiency. Suc-cessful, long-term human habitation of thiszone has depended upon two key factors:the evolution of a depigmented integumentcapable of permitting maximum cutaneousprevitamin D3 synthesis under conditions ofavailable UV radiation and the consumptionof foodstuffs naturally high in vitamin D3

(such as fish and marine mammals). Recentmigrants to high latitude regions, such asGreenlanders, appear to be only very slowly

73

undergoing depigmentation because of theirvitamin D3-rich diet.

Robins (1991) has challenged the theorythat depigmentation in high latitude humanpopulations is due to the requirements ofendogenous vitamin D3 synthesis. He hasconcluded that, for early Homo sapiens onthe Eurasian or North American landscapesof the Pleistocene, ‘‘there is not a scintilla ofpaleontological or experimental evidencethat rickets (or osteomalacia) ever did orwould have manifested, regardless of skincolour [sic]’’ (Robins, 1991:208). Thepaleoenvironmental and clinical evidencenow at hand militates against thisinterpretation.

Populations of Homo sapiens (sensu stricto)have been established above the 40th paral-lel, in areas of low annual UV radiation, for,approximately, only the last 50,000 years.(It is important to note in this connectionthat there is no equivalent land mass in thesouthern hemisphere available for humanhabitation.) The period from approximately50,000 to 10,000 years ago comprises theLast Glaciation, and includes the LastGlacial Maximum (LGM) at 18,000 B.P.During this period, human habitation northof the Tropic of Cancer and, particularly,north of 40�N was limited by low tempera-tures and the proximity of ice sheets.Climatic oscillations were frequent andsudden, and had profound impacts on thedistributions of humans and the mammalsthey depended upon (Williams et al., 1993).The minimum distance of habitation sitesfrom the southern margin of the ice sheetswas approximately 400–600 km (Madeyska,1992). During the LGM, the areas ofEurope habitable by humans were charac-terized by extreme cold, with annual meanair temperatures 8�C colder than presentday, widespread decreases in precipitation,and fierce dust storms (Frenzel, 1992a,b).Accurate reconstruction of patterns ofhuman activity north of the 40th parallelfrom the period of 50,000 to 10,000 B.P. is

not possible. It can be safely said, however,that human populations—especially aroundthe LGM—here would have worn moreprotective clothing and would have neededto seek shelter from the elements more thanthey do today. The latter would have beenparticularly true of females and youngchildren.

Abundant clinical evidence attests to thefact that moderately to deeply pigmentedhumans suffer from various manifestationsof hypovitaminosis D3 (including rickets,osteomalacia and osteoporosis) when theiropportunities for endogenous vitamin D3

synthesis are restricted as a result of changesin lifestyle or geographical relocation.Manifestations of vitamin D3 deficiency insuch populations can be attributed to life-style changes such as more indoor living inurban environments and the wearing of con-cealing garments outdoors (Bachrach et al.,1979; Haworth & Dilling, 1986; Hendersonet al., 1987; Gullu et al., 1998; Wauters &Soesbergen, 1999). The phenomenon isequally true of populations who haverecently relocated from an area of highannual UV radiation to one of lower UV,and who have otherwise maintained allaspects of their original lifestyle (Hodgkinet al., 1973; Fogelman et al., 1995).Pregnant or lactating women and smallchildren undergoing rapid bone growth aremost susceptible to changes of UV radiationregime (Bachrach et al., 1979; Fogelmanet al., 1995; Gessner et al., 1997; Namgunget al., 1998; Waiters et al., 1999). VitaminD3 deficiencies are also common among theelderly, where they render individuals moresusceptible to osteoporosis and its sequelae(Davies et al., 1986; Thomas et al., 1998).The key point here is that overwhelmingclinical evidence demonstrates that even arelatively minor decrease in endogenoussynthesis of vitamin D3 as a result ofreduced annual UV radiation exposure cantrigger vitamin D3 deficiencies in moder-ately to deeply pigmented people. These

74 . . .

Observed vs. predicted values for human skin reflectance at 685 nm based onthe results of a linear regression model between observed reflectance for allavailable groups of indigenous humans and country averages of annualaverage UVMED

Country and population or area andhemisphere designation (NH or SH)

Observedreflectanceat 685 nm

Predictedreflectanceat 685 nm

Afghanistan/Iran (NH) 55·70 45·55Algeria (Aures) (NH) 58·05 47·91Australia (Darwin) (SH) 19·30 36·24Belgium (NH) 63·14 65·66Botswana (San) (SH) 42·40 39·45Brazil (Caingan) (SH) 49·40 48·53Brazil (Guarani) (SH) 47·20 45·29Burkina Faso (Kurumba) (NH) 28·60 34·23Cambodia (NH) 54·00 38·99Cameroon (Fali) (NH) 21·50 34·37Chad (Sara) (NH) 24·60 34·77China (Southern) (NH) 59·17 50·49China (Tibet) (NH) 54·70 41·78Ethiopia (NH) 31·70 32·70Ethiopia (Highland) (NH) 33·55 31·35Germany (Mainz) (NH) 66·90 65·21Greenland (Southern) (NH) 55·70 70·31India (NH) 44·60 48·85India (Bengal) (NH) 49·73 44·33India (Goa) (NH) 46·40 38·93India (Nagpur) (NH) 41·30 41·53India (Northern) (NH) 53·26 44·23India (Orissa) (NH) 32·05 41·52India (Punjab) (NH) 54·24 47·89India (Rajasthan) (NH) 52·00 42·19India (Southern) (NH) 46·70 37·60Iraq/Syria (Kurds) (NH) 61·12 51·50Ireland (Ballinlough) (NH) 65·20 67·11Ireland (Carnew) (NH) 64·50 65·84Ireland (Longford) (NH) 65·00 66·99Ireland (Rossmore) (NH) 64·75 66·73Israel (NH) 58·20 48·67Japan (Central) (NH) 55·42 58·51Japan (Hidakka) (NH) 59·10 63·58Japan (Northern) (NH) 54·90 61·34Japan (Southwest) (NH) 53·55 56·68Jordan (NH) 53·00 45·36Kenya (NH) 32·40 34·21Lebanon (NH) 58·20 50·74Liberia (NH) 29·40 40·52Libya (Cyrenaica) (NH) 53·50 44·19Libya (Fezzan) (NH) 44·00 41·31Libya (Tripoli) (NH) 54·40 48·83Malawi (SH) 27·00 38·67Mali (Dogon) (NH) 34·10 34·54Morocco (NH) 54·85 49·09Mozambique (Chopi) (SH) 19·45 43·84

Observed vs. predicted values were not found to be significantly different(P=0·9983), based on a standard two-sample t-test between the combined means forthe observed (46·449) and predicted (46·445) 685 nm (red) filter skin reflectancevalues.

Table 6

75

(Continued)

Country and population or area andhemisphere designation (NH or SH)

Observedreflectanceat 685 nm

Predictedreflectanceat 685 nm

Namibia (SH) 25·55 38·29Namibia (Okavango) (SH) 22·92 38·63Namibia (Rehoboth Baster) (SH) 32·90 36·49Nepal (Eastern) (NH) 50·42 46·31Netherlands (NH) 67·37 65·94Nigeria (Ebo) (NH) 28·20 41·86Nigeria (Yoruba) (NH) 27·40 39·62Pakistan (NH) 52·30 44·15Papua New Guinea (SH) 35·30 37·26Papua New Guinea (Goroka) (SH) 33·30 34·20Papua New Guinea (Karker) (SH) 32·00 37·25Papua New Guinea (Lufa) (SH) 31·20 36·88Papua New Guinea (Mt Hagan) (SH) 35·35 31·56Papua New Guinea (Port Moresby) (SH) 41·00 35·45Peru (Maranon) (SH) 43·05 42·28Peru (Nunoa) (SH) 47·70 34·89Philippines (Manila) (NH) 54·10 41·53Russia (Chechen) (NH) 53·45 59·04Saudi Arabia (NH) 52·50 38·65South Africa (SH) 42·50 45·67South Africa (Cape) (SH) 50·96 50·71South Africa (Hottentot) (SH) 46·80 43·91South Africa (San Central) (SH) 43·75 41·14Spain (Basques) (NH) 65·70 62·38Spain (Leon) (NH) 64·66 60·80Sudan (NH) 35·50 33·45Swaziland (SH) 35·60 44·62Tanzania (Nyatura) (SH) 25·80 34·12Tanzania (Sandewe) (SH) 28·90 32·13Tunisia (NH) 56·30 52·03Turkey (NH) 59·15 55·56United Kingdom (Cumberland) (NH) 66·75 66·99United Kingdom (London) (NH) 62·30 65·84United Kingdom (Northern) (NH) 66·10 67·49United Kingdom (Wales) (NH) 65·00 66·15Vietnam (NH) 55·90 43·59Zaire (NH) 33·20 37·46Zaire (Konda) (SH) 29·40 39·43Average 46·18 46·52

Table 6

deficiencies are more marked in repro-ductive females, growing children and theelderly. Because moderately to deeply pig-mented people require from two to six timesas much UV radiation as lightly pigmentedindividuals to catalyze the synthesis of anequivalent amount of previtamin D3 (Table2; Figure 2) (Holick et al., 1981; Clemenset al., 1982), they are highly susceptible to

vitamin D3 deficiencies caused by a changeof UV radiation regime. We therefore con-clude, contrary to Robins, that moderatelyto deeply pigmented modern human popu-lations are optimally tuned for endogenoussynthesis of vitamin D3 under the specificUV radiation regimes under which theyevolved. This fine-tuning is easily disruptedby changes of lifestyle or locale. This

76 . . .

Fig

ure

3.P

redi

cted

shad

ing

ofsk

inco

lors

for

indi

geno

ushu

man

sba

sed

onth

ere

sult

sof

alin

ear

regr

essi

onm

odel

inw

hich

skin

refle

ctan

ce(a

t68

5nm

)fo

rin

dige

nous

peop

les

inbo

thhe

mis

pher

esw

asal

low

edto

resp

ond

toan

nual

aver

age

UV

ME

Dfo

rbo

thhe

mis

pher

es.

The

pred

icte

dsk

inre

flect

ance

valu

esw

ere

first

divi

ded

into

50eq

ual

inte

rval

san

dth

engr

aphi

cally

repr

esen

ted

ingr

aysh

ades

rang

ing

from

dark

est

gray

(gre

ates

tm

elan

izat

ion)

tolig

htes

tgr

ay(l

east

mel

aniz

atio

n).D

arke

rsh

ades

ofgr

ayre

pres

ent

ahi

gher

degr

eeof

skin

mel

aniz

atio

nan

ddo

not

repr

esen

tac

tual

pred

icte

dsk

inco

lors

.M

erca

tor

proj

ecti

on.

77

Fig

ure

4.G

rada

tion

ofsk

inco

lors

for

know

nin

dige

nous

hum

anpo

pula

tion

s,re

pres

ente

dby

shad

ing

from

dark

est

tolig

htes

tgr

ay(g

reat

est

tole

ast

mel

aniz

atio

n,as

inF

igur

e3)

,ba

sed

onob

serv

edsk

inre

flect

ance

sat

685

nmre

port

edin

Tab

le6.

Mer

cato

rpr

ojec

tion

.

78 . . .

Sexual differences in skin reflectanceComparison of skin reflectances betweenthe sexes confirmed previous observationsthat human females are consistently lighterthan males (Byard, 1981; Robins, 1991).Although female skin coloration darkensthrough adolescence and adulthood andonly lightens after 45 (Wiskemann & Wisser,1956; Byard, 1981), females are still signifi-cantly lighter (that is, show a higher valuefor skin reflectance) than males for popu-lations in which males and females arerepresented from the same location. Thissuggests that the lighter skin pigmentation offemales is needed to permit relatively greaterUV light penetration of the integument forprevitamin D3 synthesis. The extra calciumneeds of females during pregnancy andlactation are met by increasing plasma con-centrations of 1.25-dihydroxyvitamin D,

which in turn enhances calcium absorp-tion in the intestine (Wilson et al., 1990;Whitehead et al., 1981). Skin pigmentationin human females thus represents a complexcompromise between the exigencies of photo-protection and previtamin D3 synthesis.

The documentation of lighter skin infemales than in males for all populationsraises a serious question concerning thevalidity of the hypothesis that human skincoloration is in large part determined bysexual selection (Diamond, 1988, 1991).Were this so, one would have to postulatethat there was a universal preference ofmales for females slightly lighter thanthemselves, even in populations purportedto prefer dark skin (e.g., Tasmanians;Diamond, 1988, 1991). Avoidance of thesun by females is practiced in some culturesand is probably maintained as a custom inthose cultures in part by mate choice. It isunlikely, however, that this acquired customwould have any effect on constitutive pig-mentation through time unless inherentlymore lightly pigmented females who alsoavoided the sun were more reproductivelysuccessful than more darkly pigmentedfemales who followed the same practice. Wesuggest that lighter pigmentation in humanfemales began as a trait directly tied toincreased fitness and was subsequentlyreinforced and enhanced in many humanpopulations by sexual selection.

Annual average UVMED and skin reflectancePrevious studies of the relationship betweenskin reflectance and environmental variables(Roberts & Kahlon, 1976; Tasa et al.,1985), undertaken before remote sensingdata on UV radiation were available, foundhighest correlations between 685 nm filterskin reflectance and latitude (r=0·835 byRoberts & Kahlon, 1976, and r=0·82 byTasa et al., 1985). Relative to these values,calculations based on the larger skin reflect-ance dataset amassed for this study showeda very similar correlation (r=0·827). The

situation applies equally to populations ofearly Homo sapiens that undertook migra-tions from eastern Africa into the circum-Mediterranean region and Europe. Theiroriginal levels of pigmentation would haveprecipitated vitamin D3 crises in their newenvironments, especially among females andinfants, and these crises would have beenmore severe the farther north the popu-lations ventured. Clinical evidence indicatesthat the ability of humans to store vitaminD3 and its metabolites for long periods oftime in fat and skeletal muscle is dependenton their pre-existing stores of the vitamin(Mawer et al., 1972). Individuals receivinglarge doses of vitamin D3 by injection for thefirst time showed rapid elimination of vita-min D3 metabolites within 7–10 days; thosewho had previously received them showed amuch more protracted schedule of elimi-nation (Mawer et al., 1972). Thus, thepotential for vitamin D3 storage is great ifthe intake or the endogenous synthesis ishigh; otherwise it is not. Robins’ argumentsagainst the depigmentation hypothesis are,therefore, not supported.

79

generally very high correlations between skinreflectance at all wavelengths and annualaverage UVMED speaks of the close re-lationship of UV radiation to skin color(contra Diamond, 1991).

The generally small differences foundbetween observed and expected values forskin coloration appear to reflect differencesbetween populations in duration of habi-tation in their respective areas. Populationsbelieved to have inhabited their current areaof distribution for 10–20,000 years (e.g.,Spanish Basques) conform most closely topredicted values for skin reflectance. Thosewhich are thought to have migrated intotheir current locations more recently (e.g.,Aboriginal Australians from Darwin who aremigrants from the Central Desert) conformless closely to predicted values.

Cultural practices have had significanteffects on rates of change of integumentarypigmentation in human populations thathave undergone migrations, especially in thelast 20,000 or so years. The first modernhuman inhabitants of Tibet, for instance,were clad, not naked. The skin of modernTibetans is lighter than would be predictedfrom the annual average UVMED aloneprobably because of the relative recency oftheir migration and because the obligatewearing of clothes has meant that the areasof skin available for previtamin D3 syn-thesis (such as the face) had to remainrelatively unpigmented. By the same token,immigrants to the New World who inhabittropical forests are also relatively lightly pig-mented. This is probably again because ofthe relative recency of their immigration, butalso because of their habitual shade-seekingbehavior (Hames, 1992), which mitigatesthe photolytic effects of UV radiation.

The specific role of melanin in human skinMelanin has general absorption from theinfrared to the ultraviolet, but the ab-sorption maxima of other integumentarypigments are more specific. Although it is

subject to the effects of melanin, the absorp-tion from oxyhemoglobin can be sampledwith a green (545 nm) filter. UV absorptionat this value is greater in lightly pigmentedindividuals and produces a greater ery-themal response (Daniels & Imbrie, 1958;Daniels, 1964). The strongest statisticalcorrelations between skin reflectance andUVMED are observed for the 545 nm filter(Tables 4 and 5), suggesting that the mainrole of melanin pigmentation in humans isregulation of the effects of UV radiation onthe contents of cutaneous blood vesselslocated in the dermis. This finding supportsthe proposition that skin reflectance, par-ticularly in the wavelength approximatingthe absorption maximum of oxyhemoglobin(545 nm), has been optimized by naturalselection to balance the conflicting require-ments of prevention of folate photolysis(photoprotection) and previtamin D3 syn-thesis (requiring UV penetration). Wewould predict that the relationship betweenannual average UVMED and green filterreflectance would be even stronger infemales than in males because of their needto optimize the endogenous synthesis ofprevitamin D3. Unfortunately, becauseinsufficient sex-specific reflectance data existfor each hemisphere, this hypothesis couldnot be conclusively tested.

Conclusions

The results presented here demonstrate thatskin coloration in humans is highly adaptiveand has evolved to accommodate thephysiological needs of humans as they havedispersed to regions of widely varying annualUVMED. The dual selective pressures ofphotoprotection and vitamin D3 synthesishave created two clines of skin pigmentation.The first cline, from the equator to the poles,is defined by the significantly greater needfor photoprotection at the equator in par-ticular and within the tropics in general.Deeply melanized skin protects againstfolate photolysis and helps to prevent

80 . . .

UV-induced injury to sweat glands (and sub-sequent disruption of thermoregulation). Thesecond cline, from approximately 30�N to theNorth Pole, is defined by the greater need inhigh latitudes to accommodate as much pre-vitamin D3 synthesis as possible in areas oflow annual UVMED. Humans inhabiting re-gions at the intersection of these clines dem-onstrate a potential for developing varyingdegrees of facultative pigmentation (tanning)(Quevedo et al., 1975). Moderately melan-ized skin would appear to be at risk of vitaminD3 deficiency and rickets under conditionswhere UV radiation is restricted as a result oflatitude, cultural practices or both.

The results of this study suggest that skinpigmentation is relatively labile, and thatadaptations to local UVMED conditionscan occur over relatively short periods ofgeological time. Thus, it is likely that somehuman lineages through time may have gonethrough alternating periods of depigmen-tation and pigmentation (or vice versa) asthey moved from one UVMED regime toanother. As the pace of human migrationshas quickened in recent centuries, more andmore populations are finding themselves liv-ing under UV irradiation regimes to whichthey are inherently poorly adapted (e.g.,the English who settled in Australia in thenineteenth and twentieth centuries, and theIndians and Pakistanis who have moved tonorthern England in recent decades), withmajor public health consequences (Kaidbeyet al., 1979; Henderson et al., 1987). Cul-tural practices such as sun-bathing and pur-dah have in some cases exacerbated theseconditions and mitigated others. Because ofits high degree of responsiveness to environ-mental conditions, skin pigmentation is ofno value in assessing the phylogeneticrelationships between human groups.

Acknowledgements

We express our deepest thanks to CarolBower, whose intellectual input and moral

support helped to move this research for-ward in its earliest stages. We also wishto express our gratitude to James Rohr,Fiona Stanley and Tom Poiker for usefuldiscussions, to Elizabeth Weatherhead forproviding access to the NASA TOMSUVMED data, to Kenneth Beals for provid-ing a digital copy of the unpublished data-base collated at Oregon State University,and to Charles Convis of the ConservationSupport Program of Environmental SystemsResearch Institute for donations of GIS soft-ware. Patty Shea-Diner is thanked for herconsiderable help in gathering bibliographicresources. Comments from C. Loring Braceand three anonymous reviewers led tosignificant improvements in the manuscript.

References

Bachrach, S., Fisher, J. & Parkes, J. S. (1979).An outbreak of vitamin D deficiency rickets in asusceptible population. Pediatrics 64, 871–877.

Banerjee, S. (1984). The inheritance of constitutiveand facultative skin colour. Clin. Genet. 35,256–258.

Barker, D., Dixon, K., Medrano, E. E., Smalara, D.,Im, S., Mitchell, D., Babcock, G. & Abdel-Malek,Z. A. (1995). Comparison of the responses of humanmelanocytes with different melanin contents to ultra-violet B irradiation. Cancer Research 55, 4041–4046.

Barnicot, N. A. (1958). Reflectometry of the skin insouthern Nigerians and in some mulattoes. Hum.Biol. 30, 150–160.

Blum, H. F. (1961). Does the melanin pigment ofhuman skin have adaptive value? Quart. Rev. Biol.36, 50–63.

Bower, C. & Stanley, F. J. (1989). Dietary folate as arisk factor for neural-tube defects: evidence from acase-control study in Western Australia. The MedicalJournal of Australia 150, 613–619.

Branda, R. F. & Eaton, J. W. (1978). Skin colorand nutrient photolysis: An evolutionary hypothesis.Science 201, 625–626.

Brunvand, L., Quigstad, E., Urdal, P. & Haug, E.(1996). Vitamin D deficiency and fetal growth. EarlyHum. Dev. 45, 27–33.

Buccimazza, S. S., Molteno, C. D., Dunne, T. T. &Viljoen, D. L. (1994). Prevalence of neural tubedefects in Cape Town, South Africa. Teratology 50,194–199.

Buchi, E. C. (1957). Eine spektrophotometrischeUntersuchung der Hautgarbe von Angehorigen ver-schiedener Kasten in Bengalen. Bull. Scheizr. Ges.Anthropol. Ethnol. 34, 7–8.

81

Byard, P. J. (1981). Quantitative genetics of humanskin color. Yearb. phys. Anthrop. 24, 123–137.

Byard, P. J. & Lees, F. C. (1982). Skin colorimetry inBelize. II. Inter- and intra-population variation. Am.J. phys. Anthrop. 58, 215–219.

Cabanac, M. & Caputa, M. (1979). Natural selec-tive cooling of the human brain: evidence of itsoccurrence and magnitude. J. Physiol. 286, 255–264.

Carbonnel, J. P. & Olivier, G. (1966). Documentsreflectrometriques sur la couleur de la peau dans lesud-est asiatique. Bull. Mem. Soc. Anthropol. Paris 9,137–144.

Caro, L. (1980). La reflectancia de la piel en Espanolesdel Noroeste. Bol. Soc. Espanola Anthropol. Biol. 1,24–31.

Carter, C. (1970). Ethnic origin and neural tubemalformations. Developm. Med. Child Neurol. 12,372–373.

Cartwright, R. A. (1975). Skin reflectance results fromHoly Island, Northumberland. Ann. Hum. Biol. 2,347–354.

Chamla, M.-C. & Demoulin, F. (1978). Reflectancede la peau, pigmentation des cheveux et des yeuxdes Chaouias de Buizina (Aures, Algerie).L’Anthropologie 82, 61–94.

Chaplin, G. & Jablonski, N. G. (1998). Hemisphericdifference in human skin color. Am. J. phys. Anthrop.107, 221–224.

Chaplin, G., Jablonski, N. G. & Cable, N. T. (1994).Physiology, thermoregulation and bipedalism.J. hum. Evol. 27, 497–510.

Clemens, T. L., Henderson, S. L., Adams, J. S. &Holick, M. F. (1982). Increased skin pigmentreduces the capacity of skin to synthesize vitamin D3.The Lancet 1982, 74–76.

Conway, D. L. & Baker, P. T. (1972). Skin reflectanceof Quecha Indians: the effect of genetic admixture,sex and age. Am. J. phys. Anthrop. 36, 267–282.

Cosentino, M. J., Pakyz, R. E. & Fried, J. (1990).Pyrimethamine: an approach to the development of amale contraceptive. Proc. Natn. Acad. Sci. (U.S.A.)87, 1431–1435.

Cowles, R. B. (1959). Some ecological factors bearingon the origin and evolution of pigment in the humanskin. Am. Nat. 93, 283–293.

Daniels, F. Jr (1959). The physiological effects ofsunlight. J. Invest. Dermatol. 32, 147–155.

Daniels, F. Jr (1964). Man and radiant energy: solarradiation. In (D. B. Dill & E. F. Adolph, Eds)Adaptation to the Environment. Handbook of Physi-ology, pp. 969–985. Washington, D.C.: AmericanPhysiological Society.

Daniels, F. Jr & Imbrie, J. D. (1958). Comparisonbetween visual grading and reflectance measure-ments of erythema produced by sunlight. J. Invest.Dermatol. 30, 295–304.

Das, S. R. & Mukherjee, D. P. (1963). A spectro-photometric colour survey among four Indian castesand tribes. Zeit. Morphol. Anthropol. 54, 190–200.

Davies, M., Mawer, E. B., Hann, J. T. & Taylor, J. L.(1986). Seasonal changes in the biochemical indicesof vitamin D deficiency in the elderly: a comparison

of people in residential homes, long-stay wards andattending a day hospital. Ageing 15, 77–83.

Deol, M. S. (1975). Racial differences in pigmen-tation and natural selection. Ann. Hum. Genet. 38,501–503.

Diamond, J. (1988). Survival of the sexiest. Discover 9,74–81.

Diamond, J. (1991). The Rise and Fall of the ThirdChimpanzee. London: Radius.

Diaz Ungria, A. G. (1965). La pigmentacion de la pielen los indigenas Guahibos. Homenaje a Juan Comasen su 65 Anniversario, vol. 11. Antropologica Fisica.Mexico City: Editorial Libros de Mexico.

Ducros, A., Ducros, J. & Robbe, P. (1975). Pig-mentation et brunissement compares d’Eskimo(Ammassalimut, Groenland de l’Est). L’Anthropologie79, 299–316.

Elwood, J. M. & Elwood, J. H. (1980). Epidemiologyof Anencephalus and Spina Bifida. Oxford: OxfordUniversity Press.

Erickson, K. L. & Montagna, W. (1975). The induc-tion of melanogenesis by ultraviolet light in thepigmentary system of rhesus monkeys. J. Invest.Dermatol. 65, 279–284.

Falk, D. (1990). Brain evolution in Homo: The‘‘radiator’’ theory. Brain, Behav. & Evol. 13,333–381.

Fitzpatrick, T. B. (1965). Introductory lecture. In(E. J. Bower, Ed.) Recent Progress in Photobiology,pp. 365–373. New York: Academic Press.

Fleming, A. & Copp, A. J. (1998). Embryonic folatemetabolism and mouse neural tube defects. Science280, 2107–2109.

Fogelman, Y., Rakover, Y. & Luboshitsky, R. (1995).High prevalence of vitamin D deficiency amongEthiopian women immigrants to Israel: exacerbationduring pregnancy and lactation. Isr. J. Med. Sci. 31,221–224.

Forrester, M. B., Merz, R. D. & Yoon, P. W. (1998).Impact of prenatal diagnosis and elective terminationon the prevalence of selected birth defects in Hawaii.Am. J. Epidemiol. 148, 1206–1211.

Frenzel, B. (1992a). Maximum cooling of the LastGlaciation (about 20,000 to 18,000 yr B.P.). In (B.Frenzel, M. Pecsi & A. A. Velichko, Eds) Atlas ofPaleoclimates and Paleoenvironments of the NorthernHemisphere, pp. 97–99. Budapest: GeographicalResearch Institute, Hungarian Academy of Sciences.

Frenzel, B. (1992b). The Summer surface albedo atabout 18,000 B.P. In (B. Frenzel, M. Pecsi & A. A.Velichko, Eds) Atlas of Paleoclimates and Paleoenviron-ments of the Northern Hemisphere, p. 100. Budapest:Geographical Research Institute, HungarianAcademy of Sciences.

Frisancho, A. R., Wainwright, R. & Way, A. (1981).Heritability and components of phenotypic expres-sion in skin reflectance of Mestizos from the Peruvianlowlands. Am. J. phys. Anthrop. 55, 203–208.

Gessner, B. D., deSchweinitz, E., Petersen, K. M. &Leandowski, C. (1997). Nutritional rickets amongbreast-fed black and Alaska Native children. AlaskaMed. 39, 72–74.

82 . . .

Gullu, S., Erdogan, M. F., Uysal, A. R., Baskal, N.,Kamel, A. N. & Erdogan, G. (1998). A potential riskfor osteomalacia due to sociocultural lifestyle inTurkish women. Endocr. J. 45, 675–678.

Hames, R. (1992). Time allocation. In (E. A. Smith &B. Winterhalder, Eds) Evolutionary Ecology andHuman Behavior, pp. 203–237. Hawthorne, NewYork: Aldine de Gruyter.

Harrison, G. A. (1973). Differences in human pigmen-tation: measurement, geographic variation, andcauses. J. Invest. Dermatol. 60, 418–426.

Harrison, G. A. & Owen, J. J. T. (1964). Studies on theinheritance of human skin colour. Ann. Hum. Genet.28, 27–37.

Harrison, G. A. & Salzano, F. M. (1966). The skincolour of the Caingang and Guarini Indians of Brazil.Hum. Biol. 38, 104–111.

Harrison, G. A., Kuchemann, C. F., Moore, M. A. S.,Boyce, A. J., Baju, T., Mourant, A. E., Godber,M. J., Glasgow, B. G., Kopec, A. C., Tills, D. &Clegg, E. J. (1969). The effects of altitudinal vari-ation in Ethiopian populations. Phil. Trans. R. Soc.Lond. Ser. B. 256, 147–182.

Harvey, R. G. & Lord, J. M. (1978). Skin colour of theAinu of Hidaka, Hokkaido, northern Japan. Ann.Hum. Biol. 5, 459–467.

Haworth, J. C. & Dilling, L. A. (1986). Vitamin-D-deficient rickets in Manitoba, 1972–84. CMAJ 134,237–241.

Henderson, J. B., Dunnigan, M. G., McIntosh, W. B.,Abdul-Motaal, A. A., Gettinby, G. & Glekin, B. M.(1987). The importance of limited exposure to ultra-violet radiation and dietary factors in the aetiology ofAsian rickets: a risk factor model. Q. J. Med. 63,413–425.

Herman, J. & Celarier, E. (1996). TOMS Version 7UV-Erythemal Exposure: 1978–1993. (Data devel-oped by NASA Goddard Space Flight Center OzoneProcessing Team.)

Hiernaux, J. (1972). La reflectance de la peau dans unecommunaute de Sara Madjingay (Republique duTchad). L’Anthropologie 76, 279–300.

Hiernaux, J. (1977). Long-term biological effects ofhuman migration from the equatorial forest: a casestudy of human adaptation to a hot and wet climate.In (G. A. Harrison, Ed.) Population Structure andHuman Variation. International Biological Program,pp. 187–217. Cambridge: Cambridge UniversityPress.

Higgins, G. M. & Sheard, C. J. (1926). Effects ofultraviolet radiation on the early larval developmentof Rana pipiens. J. Exp. Zool. 46, 333–343.

Hodgkin, P., Kay, G. H., Hine, P. M., Lumb, G. A. &Stanbury, S. W. (1973). Vitamin-D deficiency inAsians at home and in Britain. The Lancet 1973,167–172.

Holick, M. F. (1987). Photosynthesis of vitamin D inthe skin: effect of environmental and life-stylevariables. Fed. Proc. 46, 1876–1882.

Holick, M. F., MacLaughlin, J. A. & Doppelt, S. H.(1981). Regulation of cutaneous previtamin D3

photosynthesis in man: skin pigment is not anessential regulator. Science 211, 590–593.

Huizinga, J. (1968). Human biological observations onsome African populations of the thorn savanna belt: Iand II. Proc. Kon. Ned. Akad. Weten. Ser. C 71,356–390.

Hulse, F. S. (1967). Selection for skin color among theJapanese. Am. J. phys. Anthrop. 27, 143–156.

Hulse, F. S. (1969). Skin color among the YemeniteJews of the isolate from Habban. Proc. 8th Cong.Anthropol. Ethnol. Sci., pp. 226–228. Tokyo.

Hulse, F. S. (1973). Skin colour in Northumberland.In (D. F. Roberts & E. Sunderland, Eds) GeneticVariation in Britain, vol. 12. Symposia of the Society forthe Study of Human Biology, pp. 245–257. London:Taylor & Francis.

Jablonski, N. G. (1992). Sun, skin colour and spinabifida: an exploration of the relationship betweenultraviolet light and neural tube defects. Proc.Australas. Soc. Hum. Biol. 5, 455–462.

Jablonski, N. G. (1998). Ultraviolet light-inducedneural tube defects in amphibian larvae and theirimplications for the evolution of melanized pigmen-tation and declines in amphibian populations.J. Herpetol. 32, 455–457.

Jablonski, N. G. (1999). A possible link between neuraltube defects and ultraviolet light exposure. Med.Hypotheses 52, 581–582.

Jaswal, I. J. (1979). Skin colour in northern Indianpopulations. J. hum. Evol. 8, 361–366.

Johnson, T. M., Hamilton, T. & Lowe, L. (1998).Multiple primary melanomas. J. Am. Acad. Dermatol.39, 422–427.

Kahlon, D. P. (1973). Skin colour in the Sikh commu-nity in Britain. In (D. F. Roberts & E. Sunderland,Eds) Genetic Variation in Britain, vol. 12. Symposiaof the Society for the Study of Human Biology,pp. 265–276. London: Taylor & Francis.

Kaidbey, K. H., Poh Agin, P., Sayre, R. M. & Kligman,A. M. (1979). Photoprotection by melanin—a com-parison of black and Caucasian skin. J. Am. Acad.Dermatol. 1, 249–260.

Kalla, A. K. (1969). Affinities in skin pigmentationof some Indian populations. Hum. Hered. 19, 499–505.

Kalla, A. K. (1972). Parent–child relationship and sexdifferences in skin tanning potential in man. HumanGenetik 15, 39–43.

Kalla, A. K. (1973). Ageing and sex differences inhuman skin pigmentation. Z. Morph. Anthrop. 65,29–33.

Kalla, A. K. & Tiwari, S. C. (1970). Sex difference inskin colour in man. Acta Genet. Med. Gemmelogicae19, 472–476.

Kaunitz, J. & Lindenbaum, J. (1977). Bioavailability offolic acid added to wine. Ann. Int. Med. 87, 542–545.

Kohlmeier, L. & Marcus, R. (1995). Calcium disordersof pregnancy. Endocrinol. Metab. Clin. North Am. 24,15–39.

Kollias, N., Sayre, R. M., Zeise, L. & Chedekel, M. R.(1991). Photoprotection by melanin. J. Photochem.Photobiol. B: Biol. 9, 135–160.

83

Lamparelli, R. D., Bothwell, T. H., MacPhail,A. P., van der Westhuyzen, J., Baynes, R. D. &MacFarlane, B. J. (1988). Nutritional anaemia inpregnant coloured women in Johannesburg. S. Afr.Med. J. 73, 477–481.

Lapunzina, P. (1996). Ultraviolet light-related neuraltube defects? Am. J. Med. Gen. 67, 106.

Lasker, G. (1954). Photoelectric measurement of skincolor in a Mexican Mestizo population. Am. J. phys.Anthrop. 12, 115–122.

Lawrence, V. A. (1983). Demographic analysis ofserum folate and folate-binding capacity in hospital-ized patients. Acta Haematol. 69, 289–293.

Lees, F. & Byard, P. (1978). Skin colorimetry in Belize.1. Conversion formulae. Am. J. phys. Anthrop. 48,515–522.

Leguebe, A. (1961). Contribution a l’etude de la pig-mentation chez l’homme. Bull. Inst. Roy. Sci. Nat.Belg. 37, 1–29.

Licht, L. E. & Grant, K. P. (1997). The effects ofultraviolet radiation on the biology of amphibians.Am. Zool. 37, 137–145.

Loomis, W. F. (1967). Skin-pigment regulationof vitamin-D biosynthesis in man. Science 157,501–506.

Lourie, J. A. (1973). III. Physical characteristics ofYemenite and Kurdish Jews in Israel. Phil. Trans. R.Soc. Lond., Ser. B 266, 101–112.

Madeyska, T. (1992). Human occupation of the OldWorld during the last Glaciation. In (B. Frenzel, M.Pecsi & A. A. Velichko, Eds) Atlas of Paleoclimatesand Paleoenvironments of the Northern Hemisphere,pp. 130–131. Budapest: Geographical ResearchInstitute, Hungarian Academy of Sciences.

Mahoney, S. A. (1980). Cost of locomotion and heatbalance during rest and running from 0 to 55�C in apatas monkey. J. Appl. Physiol. 49, 789–800.

Mathur, U., Datta, S. L. & Mathur, B. B. (1977). Theeffect of aminopterin-induced folic acid deficiency onspermatogenesis. Fertility Sterility 28, 1356–1360.

Mawer, E. G., Backhouse, J., Holman, C. A., Lumb,G. A. & Stanbury, S. W. (1972). The distributionand storage of vitamin D and its metabolites inhuman tissues. Clin. Sci. 43, 413–431.

Medical Research Council Vitamin Research Group.(1991). Prevention of neural tube defects: results ofthe Medical Research Council vitamin study. TheLancet 338, 131–137.

Mehrai, H. & Sunderland, E. (1990). Skin colour datafrom Nowshahr City, northern Iran. Ann. Hum. Biol.17, 115–120.

Minns, R. (1996). Folic acid and neural tube defects.Spinal Cord 34, 460–465.

Montagna, W. (1981). The consequences of having anaked skin. Birth Defects Original Article Series 17, 1–7.

Montagna, W. & Machida, H. (1966). The skin ofprimates. XXXII. The Philippine tarsier (Tarsiussyrichta). Am. J. phys. Anthrop. 25, 71–84.

Montagna, W., Machida, H. & Perkins, E. (1966a).The skin of primates. XXVII. The stump-tailmacaque (Macaca speciosa). Am. J. phys. Anthrop. 24,71–86.

Montagna, W., Machida, H. & Perkins, E. M. (1966b).The skin of primates. XXXII. The skin of theangwantibo (Arctocebus calabarensis). Am. J. phys.Anthrop. 25, 277–290.

Murray, F. G. (1934). Pigmentation, sunlight andnutritional disease. Am. Anthrop. 36, 438–445.

Namgung, R., Tsang, R. C., Lee, C., Han, D. G., Ho,M. L. & Sierra, R. I. (1998). Low total body bonemineral content and high bone resorption in Koreanwinter-born versus summer-born newborn infants.J. Pediatr. 132, 421–425.

Neer, R. M. (1975). The evolutionary significance ofvitamin D, skin pigment, and ultraviolet light. Am. J.phys. Anthrop. 43, 409–416.

Nelson, D. A. & Nunneley, S. A. (1998). Brain tem-perature and limits on transcranial cooling inhumans: quantitative modeling results. Eur. J. Appl.Physiol. 78, 353–359.

Ojikutu, R. O. (1965). Die Rolle von Hautpigmentund Schewissdrusen in der Klimaanpassung desMenschen. Homo 16, 77–95.

Omaye, S. T. (1993). Nutrient deficiencies and preg-nancy outcome. In (R. P. Sharma, Ed.) DietaryFactors and Birth Defects, pp. 12–41. San Francisco:Pacific Division AAAS.

Paltridge, G. W. & Barton, I. J. (1978). Erythemalultraviolet radiation distribution over Australia—thecalculations, detailed results and input data includingfrequency analysis of observed Australia cloud cover.C.S.I.R.O. Division of Atmospheric Physics TechnicalPaper No. 33, 1–48.

Pandolf, K. B., Gange, R. W., Latzka, W. A., Blank,I. H., Kraning, K. K. d. & Gonzalez, R. R. (1992).Human thermoregulatory responses during heatexposure after artificially induced sunburn. Am. J.Physiol. 262, R610–R616.

Post, P. W., Daniels, F. Jr & Binford, R. T. (1975a).Cold injury and the evolution of ‘‘white’’ skin. Hum.Biol. 47, 65–80.

Post, P. W., Szabo, G. & Keeling, M. E. (1975b). Aquantitative and morphological study of the pigmen-tary system of the chimpanzee with light and electronmicroscope. Am. J. phys. Anthrop. 43, 435–444.

Quevedo, W. C. Jr, Fitzpatrick, T. B., Pathak, M. A. &Jimbow, K. (1975). Role of light in human skin colorvariation. Am. J. phys. Anthrop. 43, 393–408.

Rebato, E. (1987). Skin colour in the Basque popu-lation. Anthrop. Anz. 45, 49–55.

Relethford, J. & Lees, F. (1981). Admixture and skincolor in the transplanted Tlaxcatecan population ofSaltillo, Mexico. Am. J. phys. Anthrop. 56, 259–267.

Relethford, J. H. (1997). Hemispheric differencesin human skin color. Am. J. phys. Anthrop. 104,449–458.

Ritgers-Aris, C. A. (1973a). A reflectometric study ofthe skin in Dutch families. J. hum. Evol. 2, 123–136.

Ritgers-Aris, C. A. (1973b). Reflectometrie cutanee desFali (Cameroun). Proc. Kon. Ned. Akad. Weten. Ser.C 76, 500–511.

Roberts, D. F. (1977). Human pigmentation: itsgeographical and racial distribution and biologicalsignificance. J. Soc. Cosmetic Chem. 28, 329–342.

84 . . .

Roberts, D. F. & Kahlon, D. P. (1976). Environmentalcorrelations of skin colour. Ann. Hum. Biol. 3,11–22.

Roberts, D. F., Kromberg, J. G. & Jenkins, T. (1986).Differentiation of heterozygotes in recessive albinism.J. Med. Genet. 23, 323–327.

Robins, A. H. (1991). Biological Perspectives on HumanPigmentation. Cambridge: Cambridge UniversityPress.

Rosenstreich, S. J., Rich, C. & Volwiler, W. (1971).Deposition in and release of vitamin D3 from bodyfat: evidence for a storage site in the rat. J. Clin.Invest. 50, 679–687.

Schwartz, G. G. & Rosenblum, L. A. (1981).Allometry of primate hair density and the evolutionof human hairlessness. Am. J. phys. Anthrop. 55,9–12.

Shaw, G. M., Jensvold, N. G., Wasserman, C. R. &Lammer, E. J. (1994). Epidemiologic characteristicsof phenotypically distinct neural tube defects among0·7 million California births, 1983–1987. Teratology49, 143–149.

Smith, J. & Mitchell, R. J. (1973). Skin colour studiesin South Wales, the Isle of Man and Cumbria. In(D. F. Roberts & E. Sunderland, Eds) Genetic Vari-ation in Britain, vol. 12. Symposia of the Society forthe Study of Human Biology, pp. 259–264. London:Taylor & Francis.

Sunderland, E. (1979). Skin color variability in theMiddle East and Asia. Physiological and MorphologicalAdaptation and Evolution. World Anthropology,pp. 1–18. The Hague: Mouton.

Sunderland, E. & Coope, E. (1973). XII. Geneticstudies in Jordan. Phil. Trans. R. Soc. Lond., Ser. B266, 207–220.

Sunderland, E. & Woolley, V. (1982). A study ofskin pigmentation in the population of the formercounty of Pembrokeshire, Wales. Hum. Biol. 54,387–401.

Sunderland, E., Tills, D., Bouloux, C. & Doyl, J.(1973). Genetic studies in Ireland. In (D. F. Roberts& E. Sunderland, Eds) Genetic Variation in Britain,vol. 11. Symposia of the Society for the Study ofHuman Biology, pp. 141–159. London: Taylor &Francis.

Tamura, T. & Halsted, C. H. (1983). Folate turnoverin chronically alcoholic monkeys. J. Lab. Clin. Med.101, 623–628.

Tasa, G. L., Murray, C. J. & Boughton, J. M. (1985).Reflectometer reports on human pigmentation. Curr.Anthrop. 26, 511–512.

Thomas, M. K., Lloyd-Jones, D. M., Thadhani, R. I.,Shaw, A. C., Deraska, D. J., Kitch, B. T., Vamvakas,E. C., Dick, I. M., Prince, R. L. & Finkelstein, J. S.(1998). Hypovitaminosis D in medical inpatients.New Engl. J. Med. 338, 777–783.

Tiwari, S. C. (1963). Studies of crossing betweenIndians and Europeans. Ann. Hum. Genet. 26,219–227.

Tobias, P. V. (1963). Studies on skin reflectance inBushman–European hybrids. In Second International

Congress of Human Genetics, pp. 461–471. Rome:Instituto G. Mendel.

Tobias, P. V. (1974). The biology of the South AfricanNegro. In (W. D. Hammond-Tooke, Ed.) TheBantu-Speaking Peoples of South Africa, pp. 3–45.London: Routledge & Kegan Paul.

Van Rijn-Tournel, J. (1966). Pigmentation de la peaude Belges et d’Africains. Bull. Soc. roy. Belged’Anthropol. Prehist. 76, 79–96.

Velie, E. M. & Shaw, G. M. (1996). Impact of prenataldiagnosis and elective termination on prevalence andrisk estimates of neural tube defects in California,1989–1991. Am. J. Epidemiol. 144, 473–479.

Waiters, B., Godel, J. C. & Basu, T. K. (1999).Perinatal vitamin D and calcium status of northernCanadian mothers and their newborn infants. J. Am.Coll. Nutr. 18, 122–126.

Walsh, R. J. (1963). Variations of melanin pigmen-tation of the skin in some Asian and Pacific peoples.Man 93, 126–132.

Wasserman, H. P. (1965). Human pigmentation and en-vironmental adaptation. Arch. Env. Health 11, 691–694.

Wassermann, H. P. (1974). Ethnic Pigmentation:Historical Physiological and Clinical Aspects.Amsterdam: Excerpta Medica.

Wasserman, H. P. & Heyl, T. (1968). Quantitative dataon skin pigmentation in South African races. S. Afr.Med. J. 42, 98–101.

Wauters, I. M. & Soesbergen, R. M. (1999). Diseasecaused by lack of sunlight: rickets and osteomalacia.Ned. Tijdschr. Geneeskd. 143, 593–597.

Webb, A. B., Kline, L. & Holick, M. F. (1988).Influence of season and latitude on the cutaneoussynthesis of vitamin D3: synthesis in human skin.J. Clin. Endocrinol. Metabol. 67, 373–378.

Weiner, J. S., Sebag-Montefiore, N. C. & Peterson, J.N. (1963). A note on the skin colour of the AguaranaIndians of Peru. Hum. Biol. 35, 470–473.

Weiner, J. S., Harrison, G. A., Singer, R., Harris, R. &Jopp, W. (1964). Skin colour in southern Africa.Hum. Biol. 36, 294–307.

Weninger, M. (1969). SpektrophotometrischeUntersuchungen der Haut an einem Bantu-Stamm(Chope) aus Mocambique. Anthropologie 7, 53–58.

Wheeler, P. E. (1984). The evolution of bipedality andloss of functional body hair in hominids. J. hum.Evol. 13, 91–98.

Wheeler, P. E. (1996). The environmental context offunctional body hair loss in hominids (a reply toAmaral, 1996). J. hum. Evol. 30, 367–371.

Whitehead, M., Lane, G., Young, O., Campbell, S.,Abeyasekera, G., Hillyard, C. J., MacIntyre, I.,Phang, K. G. & Stevenson, J. C. (1981). Interrela-tions of calcium-regulating hormones during normalpregnancy. Brit. Med. J. (Clinical Research Edition)283, 10–12.

Williams, M. A. J., Dunkerley, D. L., De Deckker, P.,Kershaw, A. P. & Stokes, T. (1993). QuaternaryEnvironments. London: Edward Arnold.

Williams-Blangero, S. & Blangero, J. (1991). Skin colorvariation in eastern Nepal. Am. J. phys. Anthrop. 85,281–291.

85

Wilson, S. G., Retallack, R. W., Kent, J. C., Worth,G. K. & Gutteridge, D. H. (1990). Serum free1,25-dihydroxyvitamin D and the free 1,25-dihydroxyvitamin D index during a longitudinalstudy of human pregnancy and lactation. Clin.Endocrinol. 32, 613–622.

Wiskemann, A. & Wisser, H. (1956). Die Beziehungder directen Pigmentierung zur Konstitution.Strahlentherapie 99, 594–599.

Wiswell, T. E., Tuttle, D. J., Northam, R. S. & Simonds,G. R. (1990). Major congenital neurologic malfor-mations. A 17-year survey. Am. J. Dis. Child 144, 61–67.

86 . . .

Ap

pen

dix

Ori

gina

lsk

inre

flect

ance

data

for

indi

geno

ushu

man

popu

lati

ons

asde

fined

inth

ete

xt. P

opul

atio

nde

sign

atio

n(a

sin

Tab

le5)

Cou

ntry

orre

gion

Sub

regi

onor

city

Pop

ulat

ion

nam

egi

ven

inst

udy

Sex

Afg

hani

stan

/Ira

nA

fgha

nist

anA

fgha

ni/I

rani

anM

Alg

eria

Aur

esA

lger

iaA

ures

Cha

ouia

sfr

omB

ouzi

naM

Alg

eria

Aur

esA

lger

iaA

ures

Cha

ouia

sfr

omB

ouzi

naF

Aus

tral

iaD

arw

inA

ustr

alia

Dar

win

Abo

rigi

nes

MB

elgi

umB

elgi

umB

elgi

ans

3B

elgi

ans

BB

elgi

umB

elgi

umB

russ

els

Bel

gian

sM

Bel

gium

Bel

gium

Bru

ssel

sB

elgi

ans

FB

elgi

umB

elgi

umB

russ

els

Bel

gian

s(e

xcep

t2

Gre

eks,

1Ir

aqi)

MB

elgi

umB

elgi

umB

russ

els

Bel

gian

s(e

xcep

t2

Gre

eks,

1Ir

aqi)

FB

otsw

ana

San

Bot

swan

aK

alah

ari

Des

ert

Cen

tral

Bus

hmen

MB

otsw

ana

San

Bot

swan

aK

alah

ari

Des

ert

Cen

tral

Bus

hmen

FB

otsw

ana

San

Bot

swan

aK

alah

ari

Des

ert

Yel

low

Bus

hmen

atL

one

Tre

e,C

entr

alS

anM

Bot

swan

aS

anB

otsw

ana

Kal

ahar

iD

eser

tY

ello

wB

ushm

enat

Lon

eT

ree,

Cen

tral

San

FB

otsw

ana

San

Bot

swan

aK

alah

ari

Des

ert

Yel

low

Bus

hmen

atT

akas

hwan

i,C

entr

alS

anM

Bot

swan

aS

anB

otsw

ana

Kal

ahar

iD

eser

tY

ello

wB

ushm

enat

Tak

ashw

ani,

Cen

tral

San

FB

otsw

ana

San

Bot

swan

aK

alah

ari

Des

ert

Yel

low

Bus

hmen

atG

hanz

i,C

entr

alS

anM

Bot

swan

aS

anB

otsw

ana

Kal

ahar

iD

eser

tY

ello

wB

ushm

enat

Gha

nzi,

Cen

tral

San

FB

razi

lC

aing

anB

razi

lP

aran

aC

aing

ang

Indi

ans

MB

razi

lC

aing

anB

razi

lP

aran

aC

aing

ang

Indi

ans

FB

razi

lG

uara

niB

razi

lP

aran

aG

uari

niIn

dian

sF

Bra

zil

Gua

rani

Bra

zil

Par

ana

Gua

rini

Indi

ans

MB

urki

naF

aso

Kur

umba

Bur

kina

Fas

oF

orm

erly

Upp

erV

olta

Kur

umba

from

Roa

nga

FB

urki

naF

aso

Kur

umba

Bur

kina

Fas

oF

orm

erly

Upp

erV

olta

Kur

umba

from

Roa

nga

MC

ambo

dia

Cam

bodi

aK

hmer

sM

Cam

eroo

nF

ali

Cam

eroo

n(N

orth

ern)

Fal

iT

ingu

elin

MC

amer

oon

Fal

iC

amer

oon

(Nor

ther

n)F

ali

Tin

guel

inF

Cam

eroo

nF

ali

Cam

eroo

n(N

orth

ern)

Fal

iK

ango

uM

Sex

desi

gnat

ions

:F=

fem

ale;

M=

mal

e;B

=bo

th.A

llda

taar

efr

ompu

blis

hed

sour

ces

exce

ptfo

rth

ose

indi

cate

dto

befr

om‘‘

OS

U-B

eal’s

Dat

abas

e’’,

whi

chw

ere

kind

lypr

ovid

edby

Dr

Ken

neth

Bea

ls.S

kin

refle

ctan

ces

for

filte

rsof

diff

eren

twav

elen

gth

are

iden

tifie

dby

the

wav

elen

gth

only

rath

erth

anth

eir

filte

rnu

mbe

r.(F

orin

stan

ce,

the

609

filte

rw

hich

mea

sure

ssk

inre

flect

ance

at68

5nm

isid

enti

fied

only

as68

5nm

.)T

heon

lyor

igin

alm

easu

rem

ents

conv

erte

dfr

omP

hoto

volt

toE

.E.L

.va

lues

wer

eth

ose

from

Col

ombi

a,w

hich

wer

eco

nver

ted

usin

gth

efo

rmul

afo

rB

eliz

eC

reol

esof

Lee

s&

Bya

rd,

1978

.

87

A

pp

end

ix(C

onti

nu

ed)

Pop

ulat

ion

desi

gnat

ion

(as

inT

able

5)C

ount

ryor

regi

onS

ubre

gion

orci

tyP

opul

atio

nna

me

give

nin

stud

yS

ex

Cam

eroo

nF

ali

Cam

eroo

n(N

orth

ern)

Fal

iK

ango

uF

Cha

dS

ara

Cha

dN

dila

Sar

aM

adjin

gay

MC

had

Sar

aC

had

Ndi

laS

ara

Mad

jinga

yF

Chi

naS

outh

ern

Chi

naM

Chi

naS

outh

ern

Chi

naH

ong

Kon

gC

hine

seM

Chi

naT

ibet

Tib

etIn

dia

Mus

soor

ieT

ibet

ans

MC

hina

Tib

etT

ibet

Indi

aM

usso

orie

Tib

etan

sF

Col

ombi

aA

mer

icas

Reg

ion

De

Pla

nos

Gua

hibo

sM

Eth

iopi

aE

thio

pia

Res

iden

tsof

Adi

-Ark

ai(1

500

mal

titu

de)

ME

thio

pia

Eth

iopi

aR

esid

ents

ofA

di-A

rkai

(150

0m

alti

tude

)F

Eth

iopi

aH

ighl

and

Eth

iopi

aH

ighl

ands

Res

iden

tsof

Deb

arec

h(3

000

mal

titu

de)

ME

thio

pia

Hig

hlan

dE

thio

pia

Hig

hlan

dsR

esid

ents

ofD

ebar

ech

(300

0m

alti

tude

)F

Ger

man

yM

ainz

Ger

man

yM

ainz

Ger

man

and

Am

eric

anW

hite

sM

Gre

enla

ndS

outh

ern

Gre

enla

ndG

reen

land

Eas

tE

skim

oA

mm

assa

limiu

tM

Indi

aIn

dia

Ang

ami

Nag

asB

Indi

aB

enga

lIn

dia

Ben

gal

Low

Cas

teM

Indi

aB

enga

lIn

dia

Ben

gal

Kay

asth

aM

Indi

aB

enga

lIn

dia

Ben

gal

Bra

hman

MIn

dia

Ben

gal

Indi

aB

enga

lV

aidy

aM

Indi

aB

enga

lIn

dia

Cal

cutt

aR

arhi

Bra

hman

MIn

dia

Goa

Indi

aG

oaG

oans

MIn

dia

Nag

pur

Indi

aN

agpu

ran

dK

ampt

eeM

ahar

MIn

dia

Nor

ther

nIn

dia

Nor

thIn

dia

Ban

iya

MIn

dia

Nor

ther

nIn

dia

Nor

thIn

dia

Jat

Sik

hsM

Indi

aN

orth

ern

Indi

aN

orth

Indi

aH

arya

naJa

tsM

Indi

aN

orth

ern

Indi

aN

orth

Indi

aK

hatr

isM

Indi

aN

orth

ern

Indi

aN

orth

Indi

aB

rahm

ans

MIn

dia

Nor

ther

nIn

dia

Nor

thIn

dia

Aro

ras

MIn

dia

Nor

ther

nIn

dia

Nor

thIn

dia

Aro

ras

FIn

dia

Nor

ther

nIn

dia

Nor

thIn

dia

Kha

tris

FIn

dia

Nor

ther

nIn

dia

Del

hiS

axen

aK

ayas

tha

(hus

band

–wif

e–ch

ildre

n)F

Indi

aN

orth

ern

Indi

aD

elhi

Sax

ena

Kay

asth

a(h

usba

nd–w

ife–

child

ren)

MIn

dia

Nor

ther

nIn

dia

Del

hiS

axen

aK

ayas

tha

(hus

band

–wif

e–ch

ildre

n)B

Indi

aN

orth

ern

Indi

aD

elhi

Agg

arw

alM

Indi

aN

orth

ern

Indi

aD

elhi

Agg

arw

alF

Indi

aN

orth

ern

Indi

aE

ngla

ndN

orth

Indi

ans

(Hig

her

Cas

tes)

M

88 . . .

Ap

pen

dix

(Con

tin

ued

)

Pop

ulat

ion

desi

gnat

ion

(as

inT

able

5)C

ount

ryor

regi

onS

ubre

gion

orci

tyP

opul

atio

nna

me

give

nin

stud

yS

ex

Indi

aO

riss

aIn

dia

Kor

aput

Tow

n,O

riss

aB

aren

gP

aroj

aM

Indi

aO

riss

aIn

dia

Kor

aput

Tow

n,O

riss

aB

ado

Gad

aba

MIn

dia

Pun

jab

Indi

aD

elhi

Pun

jabi

Indi

ans

(mot

her)

FIn

dia

Pun

jab

Indi

aD

elhi

Pun

jabi

Indi

ans

(fat

her)

MIn

dia

Pun

jab

Indi

aD

elhi

Pun

jabi

Indi

ans

(son

)M

Indi

aP

unja

bIn

dia

Del

hiP

unja

biIn

dian

s(d

augh

ter)

FIn

dia

Pun

jab

Indi

aE

ngla

ndS

ikhs

(hus

band

–wif

e–ch

ildre

n)F

Indi

aP

unja

bIn

dia

Eng

land

Sik

hs(h

usba

nd–w

ife–

child

ren)

MIn

dia

Pun

jab

Indi

aE

ngla

ndS

ikhs

(hus

band

–wif

e–ch

ildre

n)F

Indi

aP

unja

bIn

dia

Eng

land

Sik

hs(h

usba

nd–w

ife–

child

ren)

MIn

dia

Pun

jab

Indi

aP

unja

bK

hatr

isP

unja

bM

Indi

aP

unja

bIn

dia

Sik

hs2

Sik

hsB

Indi

aR

ajas

than

Indi

aN

orth

Indi

aR

ajpu

tsM

Indi

aS

outh

ern

Indi

aS

outh

Indi

aM

Iran

Now

shah

rIr

anIr

an(N

orth

ern)

Now

shah

rC

ity

MIr

anN

owsh

ahr

Iran

Iran

(Nor

ther

n)N

owsh

ahr

Cit

yF

Iraq

Syr

iaK

urds

Syr

iaM

ainz

Syr

ians

and

Iran

ians

MIr

aqS

yria

Kur

dsIr

aqIr

aqi/S

yria

nM

Iraq

Syr

iaK

urds

Iraq

Iraq

i/Syr

ian

FIr

elan

dB

alin

loug

hIr

elan

d(E

ire)

Bal

linlo

ugh

MIr

elan

dB

alin

loug

hIr

elan

d(E

ire)

Bal

linlo

ugh

FIr

elan

dC

arne

wIr

elan

d(E

ire)

Car

new

FIr

elan

dC

arne

wIr

elan

d(E

ire)

Car

new

MIr

elan

dL

ongf

ord

Irel

and

(Eir

e)C

ount

yL

ongf

ord

BIr

elan

dL

ongf

ord

Irel

and

(Eir

e)C

ount

yL

ongf

ord

MIr

elan

dL

ongf

ord

Irel

and

(Eir

e)C

ount

yL

ongf

ord

FIr

elan

dR

ossm

ore

Irel

and

(Eir

e)R

ossm

ore

FIr

elan

dR

ossm

ore

Irel

and

(Eir

e)R

ossm

ore

MIs

rael

Pal

esti

neP

ales

tine

MIs

rael

Pal

esti

neP

ales

tine

FJa

pan

Cen

tral

Japa

nJa

pan

(Cen

tral

)Ja

pane

seM

Japa

nC

entr

alJa

pan

Japa

n(C

entr

al)

Japa

nese

FJa

pan

Cen

tral

Japa

nC

entr

alJa

pane

seC

entr

alJa

pane

seB

Japa

nC

entr

alJa

pan

Tok

yoJa

pane

seM

Japa

nH

idak

aH

okka

ido

Hid

aka

Ain

uM

Japa

nH

idak

aH

okka

ido

Hid

aka

Ain

uF

89

A

pp

end

ix(C

onti

nu

ed)

Pop

ulat

ion

desi

gnat

ion

(as

inT

able

5)C

ount

ryor

regi

onS

ubre

gion

orci

tyP

opul

atio

nna

me

give

nin

stud

yS

ex

Japa

nN

orth

ern

Japa

nJa

pan

(Nor

th)

Japa

nese

MJa

pan

Nor

ther

nJa

pan

Japa

n(N

orth

)Ja

pane

seF

Japa

nS

outh

wes

tJa

pan

Sou

thw

est

Sou

thw

est

MJa

pan

Sou

thw

est

Japa

nS

outh

wes

tS

outh

wes

tF

Jord

anJo

rdan

Jord

anN

on-v

illag

eA

rabs

MJo

rdan

Jord

anJo

rdan

All

Ara

bsM

Jord

anJo

rdan

Jord

anM

Jord

anA

zzar

qaJo

rdan

Jord

anA

zraq

depr

essi

onA

rab

villa

ger,

child

FJo

rdan

Azz

arqa

Jord

anJo

rdan

Azr

aqde

pres

sion

Ara

bvi

llage

rsM

Ken

yaK

enya

Ken

yaM

Kur

dish

Jew

sIr

aqN

eB

aghd

adK

urdi

shJe

ws

MK

urdi

shJe

ws

Iraq

Ne

Bag

hdad

Kur

dish

Jew

sM

Leb

anon

Leb

anon

Leb

anes

eM

Leb

anon

Leb

anon

Leb

anes

eF

Lib

eria

Lib

eria

Ger

man

yM

ainz

Afr

ican

sfr

omG

hana

and

Lib

eria

ML

ibya

Cyr

enai

caL

ibyi

aC

yren

aica

ML

ibya

Cyr

enai

caL

ibyi

aC

yren

aica

FL

ibya

Fez

zan

Lib

yia

Fez

zan

ML

ibya

Tri

poli

Lib

yia

Tri

polit

ania

ML

ibya

Tri

poli

Lib

yia

Tri

polit

ania

FM

alaw

iM

alaw

iM

ainl

yC

ewa

MM

ali

Dog

onM

ali

San

gaan

dB

oni

Dog

onM

Mal

iD

ogon

Mal

iS

anga

and

Bon

iD

ogon

FM

oroc

coM

oroc

coM

orac

cans

Mor

ocan

sB

Mor

occo

Mor

occo

Mor

occo

MM

ozam

biqu

eC

hopi

Moz

ambi

que

Sof

ala

Cho

piM

Moz

ambi

que

Cho

piM

ozam

biqu

eS

ofal

aC

hopi

FN

amib

iaN

amib

iaK

urun

gkur

uK

raal

Kur

ungk

uru

Kra

alB

Nam

ibia

Nam

ibia

Ton

doro

Ton

doro

BN

amib

iaO

kava

ngo

Nam

ibia

Bor

deri

ngon

Ang

ola

Oka

vang

oB

antu

,K

uang

ali

MN

amib

iaO

kava

ngo

Nam

ibia

Bor

deri

ngon

Ang

ola

Oka

vang

oB

antu

,K

uang

ali

FN

amib

iaO

kava

ngo

Nam

ibia

Bor

deri

ngon

Ang

ola

Oka

vang

oB

antu

,K

uang

ali

MN

amib

iaO

kava

ngo

Nam

ibia

Bor

deri

ngon

Ang

ola

Oka

vang

oB

antu

,K

uang

ali

FN

amib

iaO

kava

ngo

Nam

ibia

Bor

deri

ngon

Ang

ola

Oka

vang

oB

antu

,M

’buk

ushu

atB

agan

iF

Nam

ibia

Oka

vang

oN

amib

iaB

orde

ring

onA

ngol

aO

kava

ngo

Ban

tu,

M’b

ukus

huat

Bag

ani

MN

amib

iaO

kava

ngo

Nam

ibia

Bor

deri

ngon

Ang

ola

Oka

vang

oB

antu

,M

’buk

ushu

atB

agan

iM

Nam

ibia

Oka

vang

oN

amib

iaB

orde

ring

onA

ngol

aO

kava

ngo

Ban

tu,

M’b

ukus

huat

Bag

ani

F

90 . . .

Ap

pen

dix

(Con

tin

ued

)

Pop

ulat

ion

desi

gnat

ion

(as

inT

able

5)C

ount

ryor

regi

onS

ubre

gion

orci

tyP

opul

atio

nna

me

give

nin

stud

yS

ex

Nam

ibia

Reh

obot

hB

aste

rN

amib

iaB

orde

ring

onA

ngol

aB

lack

Bus

hmen

atB

agan

iM

Nam

ibia

Reh

obot

hB

aste

rN

amib

iaB

orde

ring

onA

ngol

aB

lack

Bus

hmen

atB

agan

iF

Nam

ibia

Reh

obot

hB

aste

rN

amib

iaB

orde

ring

onA

ngol

aB

lack

Bus

hmen

atB

agan

iM

Nam

ibia

Reh

obot

hB

aste

rN

amib

iaB

orde

ring

onA

ngol

aB

lack

Bus

hmen

atB

agan

iF

Nam

ibia

Reh

obot

hB

aste

rN

amib

iaR

ehob

oth

Bas

ters

MN

amib

iaR

ehob

oth

Bas

ter

Nam

ibia

Reh

obot

hB

aste

rsF

Nep

alE

aste

rnN

epal

Eas

tern

Jire

lB

Nep

alE

aste

rnN

epal

Eas

tern

Sun

war

BN

epal

Eas

tern

Nep

alE

aste

rnS

herp

aB

Nep

alE

aste

rnN

epal

Eas

tern

Tam

ang

BN

epal

Eas

tern

Nep

alE

aste

rnB

rahm

anB

Nep

alE

aste

rnN

epal

Eas

tern

Che

tri

BN

epal

Eas

tern

Nep

alE

aste

rnA

vera

geB

Nep

alE

aste

rnN

epal

Eas

tern

Jire

lM

Nep

alE

aste

rnN

epal

Eas

tern

Jire

lF

Net

herl

ands

Net

herl

ands

Dut

chB

Net

herl

ands

Net

herl

ands

Dut

ch(m

ainl

yre

side

ntin

Utr

echt

)M

Net

herl

ands

Net

herl

ands

Dut

ch(m

ainl

yre

side

ntin

Utr

echt

)F

Nig

eria

Ibo

Nig

eria

Sou

ther

n(I

bada

n)Ib

oM

Nig

eria

Yor

uba

Nig

eria

Sou

ther

n(I

bada

n)Y

orub

aM

Nig

eria

Yor

uba

Nig

eria

Sou

ther

n(I

bada

n)Y

orub

aF

Nig

eria

Yor

uba

Nig

eria

Lag

osA

fric

ans

MP

akis

tan

Pak

ista

nP

akis

tan

MP

akis

tan

Pak

ista

nP

akis

tan

FP

akis

tan

Pak

ista

nM

Pap

uaK

arke

rP

apua

-New

Gui

nea

Kar

kar

Isla

ndK

arke

rIs

land

ers

FP

apua

Luf

aP

apua

-New

Gui

nea

Wes

tern

Hig

hlan

dsL

ufa

villa

gers

MP

apua

Luf

aP

apua

-New

Gui

nea

Wes

tern

Hig

hlan

dsL

ufa

villa

gers

FP

apua

Mt

Hag

anP

apua

-New

Gui

nea

Mou

ntH

agan

Wes

tern

Hig

hlan

dsM

Pap

uaM

tH

agan

Pap

ua-N

ewG

uine

aM

ount

Hag

anW

este

rnH

ighl

ands

FP

apua

New

Gui

nea

Pap

ua-N

ewG

uine

aP

apua

nsP

apua

nsB

Pap

uaP

ort

Mor

esby

Pap

ua-N

ewG

uine

aP

ort

Mor

esby

Han

uaba

daB

Per

uM

aran

onP

eru

Mar

anon

Val

ley

Agu

aran

aIn

dian

sM

Per

uM

aran

onP

eru

Mar

anon

Val

ley

Agu

aran

aIn

dian

sF

Per

uN

unoa

Per

uN

unoa

Az

BP

hilip

pine

sM

anila

Phi

lippi

nes

Man

ilaF

ilipi

noM

91

A

pp

end

ix(C

onti

nu

ed)

Pop

ulat

ion

desi

gnat

ion

(as

inT

able

5)C

ount

ryor

regi

onS

ubre

gion

orci

tyP

opul

atio

nna

me

give

nin

stud

yS

ex

Rus

sia

Che

chen

Rus

sia

Che

chni

aC

hech

enF

Rus

sia

Che

chen

Rus

sia

Che

chni

aC

hech

enM

Sau

diA

rabi

aS

audi

Sau

diM

Sou

thA

fric

aS

outh

Afr

ica

Joha

nnes

burg

S.

A.

Neg

roes

(73%

Tsw

ana

and

Xho

sa)

MS

outh

Afr

ica

Sou

thA

fric

aJo

hann

esbu

rgS

.A

.N

egro

es(7

3%T

swan

aan

dX

hosa

)F

Sou

thA

fric

aS

outh

Afr

ica

Ban

tu(9

6%X

hosa

)M

Sou

thA

fric

aS

outh

Afr

ica

Ban

tu(9

6%X

hosa

)F

Sou

thA

fric

aC

ape

Sou

thA

fric

aC

ape

Pro

vinc

eC

ape

Col

oure

dsM

Sou

thA

fric

aC

ape

Sou

thA

fric

aC

ape

Pro

vinc

eC

ape

Col

oure

dsF

Sou

thA

fric

aC

ape

Sou

thA

fric

aC

ape

Pro

vinc

eC

ape

Col

oure

dsM

Sou

thA

fric

aC

ape

Sou

thA

fric

aC

ape

Pro

vinc

eC

ape

Col

oure

dsF

Sou

thA

fric

aH

otte

ntot

Sou

thA

fric

aN

amaq

uala

ndH

otte

ntot

MS

outh

Afr

ica

Hot

tent

otS

outh

Afr

ica

Nam

aqua

land

Hot

tent

otF

Sou

thA

fric

aS

anC

entr

alN

amib

iaW

arm

bath

Hot

tent

otM

Sou

thA

fric

aS

anC

entr

alN

amib

iaW

arm

bath

Hot

tent

otF

Spa

inB

asqu

esS

pain

Gui

puzc

oaB

asqu

esF

Spa

inB

asqu

esS

pain

Gui

puzc

oaB

asqu

esM

Spa

inB

asqu

esS

pain

Viz

caya

Bas

ques

FS

pain

Bas

ques

Spa

inV

izca

yaB

asqu

esM

Spa

inB

asqu

esS

pain

Bis

cay

Bas

que

and

non-

Bas

ques

MS

pain

Leo

nS

pain

Leo

nM

eset

aM

Spa

inL

eon

Spa

inL

eon

Cab

rera

MS

pain

Leo

nS

pain

Leo

nB

ierz

oM

Spa

inL

eon

Spa

inL

eon

Mon

tana

MS

pain

Leo

nS

pain

Leo

nM

arag

ater

iaM

Sud

anS

udan

Sud

anM

Sw

azila

ndS

outh

Afr

ica

Joha

nnes

burg

and

Sw

azila

ndS

waz

isM

Sw

azila

ndS

outh

Afr

ica

Joha

nnes

burg

and

Sw

azila

ndS

waz

isF

Syr

iaD

ruze

Jord

anJo

rdan

Azr

aqD

epre

ssio

nD

ruze

MT

anza

nia

Nya

tura

Tan

zani

aN

yatu

raM

Tan

zani

aN

yatu

raT

anza

nia

Nya

tura

FT

anza

nia

San

dew

eT

anza

nia

San

daw

eM

Tan

zani

aS

ande

we

Tan

zani

aS

anda

we

FT

unis

iaT

unis

iaT

unis

ians

Tun

isia

nsB

Tur

key

Tur

key

Tur

key

MT

urke

yT

urke

yT

urke

yF

92 . . .

Ap

pen

dix

(Con

tin

ued

)

Pop

ulat

ion

desi

gnat

ion

(as

inT

able

5)C

ount

ryor

regi

onS

ubre

gion

orci

tyP

opul

atio

nna

me

give

nin

stud

yS

ex

Uni

ted

Kin

gdom

Cum

berl

and

Uni

ted

Kin

gdom

Cum

berl

and

FU

nite

dK

ingd

omC

umbe

rlan

dU

nite

dK

ingd

omC

umbe

rlan

dM

Uni

ted

Kin

gdom

Lon

don

Uni

ted

Kin

gdom

Lon

don

Eur

opea

nsF

Uni

ted

Kin

gdom

Lon

don

Uni

ted

Kin

gdom

Lon

don

Eur

opea

nsM

Uni

ted

Kin

gdom

Nor

ther

nU

nite

dK

ingd

omN

orth

umbe

rlan

dH

oly

Isla

ndM

Uni

ted

Kin

gdom

Nor

ther

nU

nite

dK

ingd

omN

orth

umbe

rlan

dH

oly

Isla

ndF

Uni

ted

Kin

gdom

Nor

ther

nU

nite

dK

ingd

omL

iver

pool

Eur

opea

nsB

Uni

ted

Kin

gdom

Nor

ther

nU

nite

dK

ingd

omN

orth

umbe

rlan

dN

orth

Nor

thum

berl

and

FU

nite

dK

ingd

omN

orth

ern

Uni

ted

Kin

gdom

Nor

thum

berl

and

Nor

thN

orth

umbe

rlan

dM

Uni

ted

Kin

gdom

Nor

ther

nU

nite

dK

ingd

omN

orth

umbe

rlan

dS

outh

-eas

tN

orth

umbe

rlan

dF

Uni

ted

Kin

gdom

Nor

ther

nU

nite

dK

ingd

omN

orth

umbe

rlan

dS

outh

-eas

tN

orth

umbe

rlan

dM

Uni

ted

Kin

gdom

Wal

esU

nite

dK

ingd

omB

riti

shIs

les

Isle

ofM

anF

Uni

ted

Kin

gdom

Wal

esU

nite

dK

ingd

omB

riti

shIs

les

Isle

ofM

anM

Uni

ted

Kin

gdom

Wal

esU

nite

dK

ingd

omW

ales

Mer

thyr

Tyd

filF

Uni

ted

Kin

gdom

Wal

esU

nite

dK

ingd

omW

ales

Mer

thyr

Tyd

filM

Uni

ted

Kin

gdom

Wal

esU

nite

dK

ingd

omW

ales

(Sou

thw

est)

Nor

thP

embr

okes

hire

FU

nite

dK

ingd

omW

ales

Uni

ted

Kin

gdom

Wal

es(S

outh

wes

t)N

orth

Pem

brok

eshi

reM

Uni

ted

Kin

gdom

Wal

esU

nite

dK

ingd

omW

ales

(Sou

thw

est)

Sou

thP

embr

okes

hire

FU

nite

dK

ingd

omW

ales

Uni

ted

Kin

gdom

Wal

es(S

outh

wes

t)S

outh

Pem

brok

eshi

reM

Vie

tnam

Vie

tnam

Vie

tnam

ese

MV

ietn

amV

ietn

amV

eitn

ames

eB

Zai

reZ

aire

Zai

re2

Zai

reB

Zai

reZ

aire

Con

gole

seex

cept

3C

amer

oon

fem

ales

MZ

aire

Zai

reC

ongo

lese

exce

pt3

Cam

eroo

nfe

mal

esF

Zai

reK

onda

Zai

reK

onda

0·9–

18·2

Oto

MZ

aire

Kon

daZ

aire

Kon

da0·

9–18

·2T

wa

M

93

A

pp

end

ix

Pop

ulat

ion

desi

gnat

ion

(as

inT

able

5)E

EL

(E)

orph

otov

olt

(P)

420

nm42

5nm

450

nm46

5nm

485

nm51

5nm

525

nm54

5nm

550

nm

Afg

hani

stan

/Ira

nE

24·2

31·7

Alg

eria

Aur

esE

Alg

eria

Aur

esE

Aus

tral

iaD

arw

inE

Bel

gium

EB

elgi

umE

36·5

37·7

33·5

44·7

46·7

46·7

38·6

44·8

42·7

Bel

gium

E35

·736

·532

·743

·545

·345

·738

·244

·642

·3B

elgi

umE

37·1

39·1

33·8

45·5

46·6

46·1

38·6

43·8

42·5

Bel

gium

E36

·338

·232

·843

·744

·544

·838

·143

·241

·9B

otsw

ana

San

E12

·712

·79·

315

·516

·217

·213

·818

·119

·1B

otsw

ana

San

E13

·213

·79·

716

·617

·418

·214

·419

·020

·0B

otsw

ana

San

E14

·4B

otsw

ana

San

E15

·2B

otsw

ana

San

E16

·2B

otsw

ana

San

E17

·2B

otsw

ana

San

E15

·6B

otsw

ana

San

E16

·4B

razi

lC

aing

anE

16·5

22·9

Bra

zil

Cai

ngan

E18

·426

·1B

razi

lG

uara

niE

17·2

23·4

Bra

zil

Gua

rani

E16

·823

·6B

urki

naF

aso

Kur

umba

E8·

610

·86·

312

·011

·011

·08·

411

·612

·2B

urki

naF

aso

Kur

umba

E8·

811

·46·

412

·511

·611

·48·

612

·012

·5C

ambo

dia

EC

amer

oon

Fal

iE

6·5

7·3

5·0

7·4

7·4

7·6

5·7

7·9

7·7

Cam

eroo

nF

ali

E6·

77·

75·

17·

67·

67·

95·

98·

28·

1C

amer

oon

Fal

iE

6·7

7·4

5·2

7·7

7·7

8·0

6·0

8·4

8·1

Cam

eroo

nF

ali

E7·

28·

15·

38·

08·

08·

56·

38·

79·

0C

had

Sar

aE

Cha

dS

ara

EC

hina

Sou

ther

nE

Chi

naS

outh

ern

EC

hina

Tib

etE

Chi

naT

ibet

EC

olom

bia

PE

thio

pia

E8·

311

·0E

thio

pia

E9·

813

·1

94 . . .

Ap

pen

dix

(Con

tin

ued

)

Pop

ulat

ion

desi

gnat

ion

(as

inT

able

5)E

EL

(E)

orph

otov

olt

(P)

420

nm42

5nm

450

nm46

5nm

485

nm51

5nm

525

nm54

5nm

550

nm

Eth

iopi

aH

ighl

and

E9·

512

·7E

thio

pia

Hig

hlan

dE

10·3

14·3

Ger

man

yM

ainz

E37

·540

·334

·645

·547

·447

·139

·845

·243

·6G

reen

land

Sou

ther

nE

22·4

29·9

Indi

aE

Indi

aB

enga

lE

13·6

13·9

10·1

17·0

18·2

19·9

15·1

20·4

20·2

Indi

aB

enga

lE

16·4

16·3

12·9

19·7

21·1

22·8

19·0

24·3

23·3

Indi

aB

enga

lE

17·1

17·2

13·4

20·4

21·7

24·0

19·6

24·3

24·1

Indi

aB

enga

lE

17·6

18·4

14·1

21·3

24·0

25·4

20·8

25·3

25·2

Indi

aB

enga

lE

17·1

18·3

13·4

20·3

21·9

23·2

19·8

24·5

24·2

Indi

aG

oaE

Indi

aN

agpu

rE

12·3

13·3

9·2

14·8

15·7

16·9

13·4

18·2

18·7

Indi

aN

orth

ern

E14

·322

·0In

dia

Nor

ther

nE

15·4

23·6

Indi

aN

orth

ern

E15

·724

·0In

dia

Nor

ther

nE

16·5

25·1

Indi

aN

orth

ern

E17

·226

·0In

dia

Nor

ther

nE

20·7

Indi

aN

orth

ern

E21

·1In

dia

Nor

ther

nE

21·7

Indi

aN

orth

ern

E14

·913

·911

·617

·118

·919

·617

·321

·521

·2In

dia

Nor

ther

nE

17·0

16·8

13·4

20·3

22·8

23·5

20·1

24·8

24·4

Indi

aN

orth

ern

E15

·614

·512

·118

·019

·920

·918

·222

·322

·4In

dia

Nor

ther

nE

Indi

aN

orth

ern

EIn

dia

Nor

ther

nE

18·6

21·4

14·9

25·2

26·5

28·0

22·2

27·9

26·6

Indi

aO

riss

aE

9·6

9·8

7·0

10·3

10·7

11·5

9·6

12·4

13·7

Indi

aO

riss

aE

9·7

10·3

7·1

10·7

11·2

11·8

12·7

14·1

Indi

aP

unja

bE

18·2

17·9

14·4

22·0

24·4

25·9

21·6

26·2

26·4

Indi

aP

unja

bE

18·6

18·8

14·8

23·3

25·5

27·0

22·2

27·0

27·0

Indi

aP

unja

bE

18·2

17·8

14·4

22·4

24·5

26·0

21·7

26·2

26·4

Indi

aP

unja

bE

18·7

18·6

14·9

23·0

25·5

27·0

22·4

27·2

27·3

Indi

aP

unja

bE

20·1

23·4

16·6

27·5

28·3

30·5

24·4

30·9

29·6

Indi

aP

unja

bE

19·5

23·1

16·0

27·0

27·7

29·5

23·6

29·7

28·5

Indi

aP

unja

bE

19·3

21·8

15·8

26·0

26·7

28·7

23·2

29·3

28·2

Indi

aP

unja

bE

18·4

20·5

15·0

24·2

25·0

27·0

22·0

27·4

26·8

95

A

pp

end

ix(C

onti

nu

ed)

Pop

ulat

ion

desi

gnat

ion

(as

inT

able

5)E

EL

(E)

orph

otov

olt

(P)

420

nm42

5nm

450

nm46

5nm

485

nm51

5nm

525

nm54

5nm

550

nm

Indi

aP

unja

bE

21·3

Indi

aP

unja

bE

Indi

aR

ajas

than

E16

·224

·7In

dia

Sou

ther

nE

Iran

Now

shah

rE

17·9

18·2

14·6

21·6

24·4

26·6

21·1

27·2

25·8

Iran

Now

shah

rE

22·8

28·0

19·4

32·9

35·6

37·3

28·5

37·6

33·6

Iraq

Syr

iaK

urds

E35

·738

·932

·441

·242

·743

·338

·143

·041

·6Ir

aqS

yria

Kur

dsE

27·4

34·1

Iraq

Syr

iaK

urds

E31

·239

·1Ir

elan

dB

alin

loug

hE

35·4

40·9

Irel

and

Bal

inlo

ugh

E36

·241

·9Ir

elan

dC

arne

wE

37·2

42·1

Irel

and

Car

new

E34

·939

·4Ir

elan

dL

ongf

ord

E35

·036

·732

·241

·044

·844

·637

·041

·440

·8Ir

elan

dL

ongf

ord

EIr

elan

dL

ongf

ord

EIr

elan

dR

ossm

ore

E35

·741

·8Ir

elan

dR

ossm

ore

E34

·640

·7Is

rael

E26

·433

·6Is

rael

E29

·637

·3Ja

pan

Cen

tral

E18

·815

·124

·426

·327

·022

·627

·727

·3Ja

pan

Cen

tral

E20

·717

·128

·630

·432

·125

·232

·930

·2Ja

pan

Cen

tral

E26

·730

·5Ja

pan

Cen

tral

EJa

pan

Hid

aka

E20

·616

·729

·230

·831

·925

·231

·729

·9Ja

pan

Hid

aka

E22

·919

·033

·335

·136

·828

·337

·631

·6Ja

pan

Nor

ther

nE

19·4

15·7

26·1

27·8

28·8

23·4

29·5

28·2

Japa

nN

orth

ern

E21

·017

·529

·731

·233

·125

·633

·630

·6Ja

pan

Sou

thw

est

E18

·114

·523

·024

·325

·721

·426

·926

·3Ja

pan

Sou

thw

est

E20

·617

·129

·731

·431

·025

·633

·730

·3Jo

rdan

E22

·430

·4Jo

rdan

E21

·629

·3Jo

rdan

E23

·731

·1Jo

rdan

Azz

arqa

E16

·724

·2Jo

rdan

Azz

arqa

E20

·327

·6K

enya

E11

·914

·0

96 . . .

Ap

pen

dix

(Con

tin

ued

)

Pop

ulat

ion

desi

gnat

ion

(as

inT

able

5)E

EL

(E)

orph

otov

olt

(P)

420

nm42

5nm

450

nm46

5nm

485

nm51

5nm

525

nm54

5nm

550

nm

Kur

dish

Jew

sE

21·1

29·2

Kur

dish

Jew

sE

24·2

33·7

Leb

anon

E27

·034

·4L

eban

onE

28·7

36·6

Lib

eria

E9·

013

·67·

212

·112

·213

·38·

813

·713

·1L

ibya

Cyr

enai

caE

19·0

26·6

Lib

yaC

yren

aica

E24

·332

·3L

ibya

Fez

zan

E14

·920

·7L

ibya

Tri

poli

E19

·827

·4L

ibya

Tri

poli

E24

·733

·6M

alaw

iE

9·0

1·2

10·0

Mal

iD

ogon

E11

·511

·011

·412

·911

·914

·0M

ali

Dog

onE

11·8

11·2

11·5

13·1

11·9

14·4

Mor

occo

EM

oroc

coE

21·2

28·1

Moz

ambi

que

Cho

piE

7·3

Moz

ambi

que

Cho

piE

7·7

Nam

ibia

EN

amib

iaE

Nam

ibia

Oka

vang

oE

Nam

ibia

Oka

vang

oE

Nam

ibia

Oka

vang

oE

Nam

ibia

Oka

vang

oE

Nam

ibia

Oka

vang

oE

Nam

ibia

Oka

vang

oE

Nam

ibia

Oka

vang

oE

Nam

ibia

Oka

vang

oE

Nam

ibia

Reh

obot

hB

aste

rE

Nam

ibia

Reh

obot

hB

aste

rE

Nam

ibia

Reh

obot

hB

aste

rE

Nam

ibia

Reh

obot

hB

aste

rE

Nam

ibia

Reh

obot

hB

aste

rE

17·2

Nam

ibia

Reh

obot

hB

aste

rE

20·6

Nep

alE

aste

rnE

16·5

25·2

Nep

alE

aste

rnE

17·7

26·3

Nep

alE

aste

rnE

19·7

29·5

97

A

pp

end

ix(C

onti

nu

ed)

Pop

ulat

ion

desi

gnat

ion

(as

inT

able

5)E

EL

(E)

orph

otov

olt

(P)

420

nm42

5nm

450

nm46

5nm

485

nm51

5nm

525

nm54

5nm

550

nm

Nep

alE

aste

rnE

15·9

24·4

Nep

alE

aste

rnE

19·0

28·4

Nep

alE

aste

rnE

17·7

26·1

Nep

alE

aste

rnE

17·4

26·1

Nep

alE

aste

rnE

16·2

24·8

Nep

alE

aste

rnE

19·5

23·9

Net

herl

ands

EN

ethe

rlan

dsE

37·8

39·7

34·7

46·8

48·3

47·8

39·3

44·7

43·3

Net

herl

ands

E36

·437

·733

·144

·645

·845

·838

·043

·242

·0N

iger

iaIb

oE

8·7

8·7

6·5

9·2

9·3

10·7

8·6

11·2

11·7

Nig

eria

Yor

uba

E7·

78·

05·

98·

38·

19·

77·

410

·19·

7N

iger

iaY

orub

aE

8·4

8·5

6·3

8·9

8·9

10·5

8·3

11·1

11·0

Nig

eria

Yor

uba

E9·

912

·87·

712

·613

·114

·110

·114

·914

·4P

akis

tan

E20

·526

·9P

akis

tan

E21

·928

·4P

akis

tan

EP

apua

Kar

ker

E10

·115

·0P

apua

Luf

aE

9·1

13·3

Pap

uaL

ufa

E9·

313

·8P

apua

Mt

Hag

anE

Pap

uaM

tH

agan

EP

apua

New

Gui

nea

EP

apua

Por

tM

ores

byE

Per

uM

aran

onE

24·5

Per

uM

aran

onE

24·7

Per

uN

unoa

EP

hilip

pine

sM

anila

ER

ussi

aC

hech

enE

21·2

29·1

Rus

sia

Che

chen

E27

·433

·6S

audi

Ara

bia

E21

·028

·0S

outh

Afr

ica

E12

·415

·39·

217

·917

·018

·113

·118

·218

·8S

outh

Afr

ica

E13

·316

·510

·019

·318

·720

·014

·520

·620

·5S

outh

Afr

ica

E10

·110

·67·

612

·312

·412

·810

·513

·714

·7S

outh

Afr

ica

E11

·612

·68·

614

·514

·815

·512

·316

·817

·4S

outh

Afr

ica

Cap

eE

16·1

15·8

12·5

19·4

20·5

21·5

18·7

22·7

23·1

Sou

thA

fric

aC

ape

E17

·918

·514

·322

·424

·325

·421

·327

·026

·0

98 . . .

Ap

pen

dix

(Con

tin

ued

)

Pop

ulat

ion

desi

gnat

ion

(as

inT

able

5)E

EL

(E)

orph

otov

olt

(P)

420

nm42

5nm

450

nm46

5nm

485

nm51

5nm

525

nm54

5nm

550

nm

Sou

thA

fric

aC

ape

ES

outh

Afr

ica

Cap

eE

Sou

thA

fric

aH

otte

ntot

ES

outh

Afr

ica

Hot

tent

otE

Sou

thA

fric

aS

anC

entr

alE

Sou

thA

fric

aS

anC

entr

alE

Spa

inB

asqu

esE

31·0

29·2

28·5

35·9

40·9

40·9

34·4

40·9

38·7

Spa

inB

asqu

esE

31·6

30·0

28·7

37·0

41·2

40·9

34·1

39·6

38·3

Spa

inB

asqu

esE

30·8

29·0

28·1

35·5

40·1

40·4

34·0

40·0

38·3

Spa

inB

asqu

esE

31·5

30·0

28·7

36·6

41·1

40·7

34·2

39·6

38·4

Spa

inB

asqu

esE

32·6

34·1

29·0

36·2

38·3

39·8

35·1

39·5

38·6

Spa

inL

eon

E31

·129

·014

·336

·439

·024

·436

·8S

pain

Leo

nE

29·3

26·3

33·4

36·5

36·1

Spa

inL

eon

E30

·027

·434

·437

·135

·5S

pain

Leo

nE

30·8

28·5

36·0

37·8

35·1

Spa

inL

eon

E28

·425

·112

·731

·834

·833

·4S

udan

E12

·214

·7S

waz

iland

E9·

89·

77·

211

·010

·313

·310

·113

·913

·1S

waz

iland

E11

·210

·58·

212

·212

·314

·912

·016

·515

·9S

yria

Dru

zeE

22·5

30·0

Tan

zani

aN

yatu

raE

7·8

10·2

Tan

zani

aN

yatu

raE

8·6

10·6

Tan

zani

aS

ande

we

E8·

811

·1T

anza

nia

San

dew

eE

9·1

11·7

Tun

isia

ET

urke

yE

28·8

36·6

Tur

key

E30

·038

·0U

nite

dK

ingd

omC

umbe

rlan

dE

36·9

42·4

Uni

ted

Kin

gdom

Cum

berl

and

E35

·841

·8U

nite

dK

ingd

omL

ondo

nE

34·4

34·3

30·8

41·4

43·1

43·9

35·6

40·5

39·7

Uni

ted

Kin

gdom

Lon

don

E33

·332

·829

·539

·741

·441

·634

·037

·938

·1U

nite

dK

ingd

omN

orth

ern

E34

·139

·8U

nite

dK

ingd

omN

orth

ern

E34

·139

·6U

nite

dK

ingd

omN

orth

ern

E35

·036

·131

·441

·643

·043

·736

·541

·040

·4U

nite

dK

ingd

omN

orth

ern

E35

·336

·532

·842

·145

·846

·438

·544

·842

·5U

nite

dK

ingd

omN

orth

ern

E33

·934

·231

·340

·043

·845

·337

·343

·841

·4

99

Ap

pen

dix

(Con

tin

ued

)

Pop

ulat

ion

desi

gnat

ion

(as

inT

able

5)E

EL

(E)

orph

otov

olt

(P)

420

nm42

5nm

450

nm46

5nm

485

nm51

5nm

525

nm54

5nm

550

nm

Uni

ted

Kin

gdom

Nor

ther

nE

34·7

35·6

32·1

41·1

44·6

45·5

37·8

43·9

41·8

Uni

ted

Kin

gdom

Nor

ther

nE

33·2

33·0

30·4

38·9

42·7

44·0

36·2

42·1

40·4

Uni

ted

Kin

gdom

Wal

esE

36·8

41·8

Uni

ted

Kin

gdom

Wal

esE

36·6

41·9

Uni

ted

Kin

gdom

Wal

esE

33·2

38·7

Uni

ted

Kin

gdom

Wal

esE

32·8

38·7

Uni

ted

Kin

gdom

Wal

esE

35·6

40·9

Uni

ted

Kin

gdom

Wal

esE

34·2

16·4

40·8

Uni

ted

Kin

gdom

Wal

esE

34·1

40·5

Uni

ted

Kin

gdom

Wal

esE

33·7

39·6

Vie

tnam

EV

ietn

amE

Zai

reE

Zai

reE

9·0

7·4

6·6

8·6

9·0

9·7

8·9

10·4

12·3

Zai

reE

10·5

9·0

7·6

10·7

11·5

12·5

10·6

12·8

14·8

Zai

reK

onda

E2·

3Z

aire

Kon

daE

100 . . .

Ap

pen

dix

Pop

ulat

ion

desi

gnat

ion

(as

inT

able

5)57

5nm

595

nm60

0nm

655

nm67

0nm

685

nmO

rigi

nal

refe

renc

eS

ourc

eof

data

Afg

hani

stan

/Ira

n55

·7(S

unde

rlan

d,19

79)

Ori

gina

lA

lger

iaA

ures

57·4

(Cha

mla

&D

emou

lin,

1978

)(R

obin

s,19

91)

Alg

eria

Aur

es58

·7(C

ham

la&

Dem

oulin

,19

78)

(Rob

ins,

1991

)A

ustr

alia

Dar

win

12·9

19·3

(Wal

sh,

1963

);re

des

tim

ated

Ori

gina

lB

elgi

um60

·1F

OS

73fr

omO

SU

-Bea

lss

data

base

OS

U-B

eals

sda

taba

seB

elgi

um48

·959

·545

·665

·766

·567

·3(L

egue

be,

1961

)(R

obin

s,19

91)

Bel

gium

48·8

58·5

45·1

64·3

65·4

65·9

(Leg

uebe

,19

61)

(Rob

ins,

1991

)B

elgi

um47

·557

·445

·062

·965

·164

·5(V

anR

ijn-T

ourn

el,

1966

)(R

obin

s,19

91)

Bel

gium

47·2

57·3

44·5

62·6

64·2

63·7

(Van

Rijn

-Tou

rnel

,19

66)

(Rob

ins,

1991

)B

otsw

ana

San

21·1

28·4

20·5

39·4

32·0

42·5

(Tob

ias,

1963

)(R

obin

s,19

91)

Bot

swan

aS

an22

·530

·221

·340

·833

·043

·6(T

obia

s,19

63)

(Rob

ins,

1991

)B

otsw

ana

San

40·5

(Wei

ner

etal

.,19

64)

(Rob

ins,

1991

)B

otsw

ana

San

43·1

(Wei

ner

etal

.,19

64)

(Rob

ins,

1991

)B

otsw

ana

San

43·0

(Wei

ner

etal

.,19

64)

(Rob

ins,

1991

)B

otsw

ana

San

43·6

(Wei

ner

etal

.,19

64)

(Rob

ins,

1991

)B

otsw

ana

San

43·0

(Wei

ner

etal

.,19

64)

(Rob

ins,

1991

)B

otsw

ana

San

44·6

(Wei

ner

etal

.,19

64)

(Rob

ins,

1991

)B

razi

lCai

ngan

48·1

(Har

riso

n&

Sal

zano

,19

66)

(Rob

ins,

1991

)B

razi

lCai

ngan

50·7

(Har

riso

n&

Sal

zano

,19

66)

(Rob

ins,

1991

)B

razi

lGua

rani

46·5

(Har

riso

n&

Sal

zano

,19

66)

(Rob

ins,

1991

)B

razi

lGua

rani

47·9

(Har

riso

n&

Sal

zano

,19

66)

(Rob

ins,

1991

)B

urki

naF

aso

Kur

umba

13·0

17·1

12·2

27·1

21·7

27·9

(Hui

zing

a,19

68)

(Rob

ins,

1991

)B

urki

naF

aso

Kur

umba

13·5

17·4

12·5

28·0

22·4

29·3

(Hui

zing

a,19

68)

(Rob

ins,

1991

)C

ambo

dia

50·2

54·0

(Car

bonn

el&

Oliv

er,

1966

);re

des

tim

ated

(Har

riso

n,19

73)

Cam

eroo

nF

ali

8·5

11·3

8·3

16·1

15·9

20·0

(Rit

gers

-Ari

s,19

73b)

(Rob

ins,

1991

)C

amer

oon

Fal

i8·

812

·08·

216

·916

·320

·9(R

itge

rs-A

ris,

1973

b)(R

obin

s,19

91)

Cam

eroo

nF

ali

8·9

11·9

8·3

16·8

16·5

21·1

(Rit

gers

-Ari

s,19

73b)

(Rob

ins,

1991

)C

amer

oon

Fal

i9·

312

·69·

018

·618

·425

·5(R

itge

rs-A

ris,

1973

b)(R

obin

s,19

91)

Cha

dS

ara

23·2

(Hie

rnau

x,19

72)

(Rob

ins,

1991

)C

had

Sar

a26

·0(H

iern

aux,

1972

)(R

obin

s,19

91)

Chi

naS

outh

ern

56·6

59·9

(Car

bonn

el&

Oliv

er,

1966

);re

des

tim

ated

(Har

riso

n,19

73)

Chi

naS

outh

ern

54·3

57·7

(Wal

sh,

1963

);re

des

tim

ated

Ori

gina

lC

hina

Tib

et54

·5(K

alla

&T

irw

ari,

1970

)(R

obin

s,19

91)

Chi

naT

ibet

54·9

(Kal

la&

Tir

war

i,19

70)

(Rob

ins,

1991

)C

olom

bia

24·6

(Dia

zU

ngri

a,19

65)

(Har

riso

n,19

73)

101

A

pp

end

ix(C

onti

nu

ed)

Pop

ulat

ion

desi

gnat

ion

(as

inT

able

5)57

5nm

595

nm60

0nm

655

nm67

0nm

685

nmO

rigi

nal

refe

renc

eS

ourc

eof

data

Eth

iopi

a30

·2(H

arri

son

etal

.,19

69)

(Rob

ins,

1991

)E

thio

pia

33·2

(Har

riso

net

al.,

1969

)(R

obin

s,19

91)

Eth

iopi

aH

ighl

and

31·4

(Har

riso

net

al.,

1969

)(R

obin

s,19

91)

Eth

iopi

aH

ighl

and

35·7

(Har

riso

net

al.,

1969

)(R

obin

s,19

91)

Ger

man

yM

ainz

50·1

60·6

46·3

66·7

66·9

66·9

(Ojik

utu,

1965

)(R

obin

s,19

91)

Gre

enla

ndS

outh

ern

55·7

(Duc

ros

etal

.,19

75)

(Rob

ins,

1991

)In

dia

44·6

OS

U-B

eals

sda

taba

seO

SU

-Bea

lss

data

base

Indi

aB

enga

l24

·129

·922

·141

·534

·244

·8(B

uchi

,19

57)

(Rob

ins,

1991

)In

dia

Ben

gal

27·4

35·6

28·1

45·3

43·7

48·6

(Buc

hi,

1957

)(R

obin

s,19

91)

Indi

aB

enga

l28

·536

·329

·347

·545

·149

·9(B

uchi

,19

57)

(Rob

ins,

1991

)In

dia

Ben

gal

30·2

37·9

30·0

47·2

45·9

50·7

(Buc

hi,

1957

)(R

obin

s,19

91)

Indi

aB

enga

l29

·136

·929

·547

·745

·249

·7(D

as&

Muk

herj

ee,

1963

)(R

obin

s,19

91)

Indi

aG

oa42

·046

·4(W

alsh

,19

63);

red

esti

mat

edO

rigi

nal

Indi

aN

agpu

r21

·628

·419

·637

·631

·341

·3(D

as&

Muk

herj

ee,

1963

)(R

obin

s,19

91)

Indi

aN

orth

ern

48·6

(Jas

wal

,19

79)

(Rob

ins,

1991

)In

dia

Nor

ther

n50

·6(J

asw

al,

1979

)(R

obin

s,19

91)

Indi

aN

orth

ern

51·5

(Jas

wal

,19

79)

(Rob

ins,

1991

)In

dia

Nor

ther

n52

·7(J

asw

al,

1979

)(R

obin

s,19

91)

Indi

aN

orth

ern

52·8

(Jas

wal

,19

79)

(Rob

ins,

1991

)In

dia

Nor

ther

n55

·8(K

alla

,19

69)

(Rob

ins,

1991

)In

dia

Nor

ther

n56

·1(K

alla

,19

69)

(Rob

ins,

1991

)In

dia

Nor

ther

n56

·4(K

alla

,19

69)

(Rob

ins,

1991

)In

dia

Nor

ther

n24

·832

·825

·741

·540

·845

·8(K

alla

,19

72)

(Rob

ins,

1991

)In

dia

Nor

ther

n29

·037

·629

·346

·545

·250

·4(K

alla

,19

72)

(Rob

ins,

1991

)In

dia

Nor

ther

n26

·034

·427

·043

·742

·648

·0(K

alla

,19

72)

(Rob

ins,

1991

)In

dia

Nor

ther

n49

·5(K

alla

,19

73)

(Rob

ins,

1991

)In

dia

Nor

ther

n50

·0(K

alla

,19

73)

(Rob

ins,

1991

)In

dia

Nor

ther

n31

·540

·531

·449

·447

·652

·5(T

iwar

i,19

63)

(Rob

ins,

1991

)In

dia

Ori

ssa

14·6

20·0

14·3

28·8

24·3

31·8

(Das

&M

ukhe

rjee

,19

63)

(Rob

ins,

1991

)In

dia

Ori

ssa

15·2

20·7

14·6

29·7

24·8

32·3

(Das

&M

ukhe

rjee

,19

63)

(Rob

ins,

1991

)In

dia

Pun

jab

31·2

40·5

31·5

49·7

48·0

53·6

(Ban

erje

e,19

84)

Ori

gina

lIn

dia

Pun

jab

31·8

41·4

32·2

50·8

48·8

54·5

(Ban

erje

e,19

84)

Ori

gina

lIn

dia

Pun

jab

30·9

40·6

31·6

50·0

48·3

54·3

(Ban

erje

e,19

84)

Ori

gina

lIn

dia

Pun

jab

32·3

41·8

32·5

51·0

49·2

54·9

(Ban

erje

e,19

84)

Ori

gina

lIn

dia

Pun

jab

36·5

45·3

34·6

52·0

51·1

55·6

(Kah

lon,

1973

)(R

obin

s,19

91)

Indi

aP

unja

b34

·843

·533

·450

·649

·854

·5(K

ahlo

n,19

73)

(Rob

ins,

1991

)

102 . . .

Ap

pen

dix

(Con

tin

ued

)

Pop

ulat

ion

desi

gnat

ion

(as

inT

able

5)57

5nm

595

nm60

0nm

655

nm67

0nm

685

nmO

rigi

nal

refe

renc

eS

ourc

eof

data

Indi

aP

unja

b34

·643

·133

·250

·449

·554

·1(K

ahlo

n,19

73)

(Rob

ins,

1991

)In

dia

Pun

jab

32·6

41·0

31·6

48·4

47·7

52·4

(Kah

lon,

1973

)(R

obin

s,19

91)

Indi

aP

unja

b56

·1(K

alla

,19

69)

(Rob

ins,

1991

)In

dia

Pun

jab

54·0

RO

B72

from

OS

U-B

eals

sda

taba

seO

SU

-Bea

lss

data

base

Indi

aR

ajas

than

52·0

(Jas

wal

,19

79)

(Rob

ins,

1991

)In

dia

Sou

ther

n42

·446

·7(W

alsh

,19

63);

red

esti

mat

edO

rigi

nal

Iran

Now

shah

r31

·239

·730

·246

·545

·850

·0(M

ehra

i&

Sun

derl

and,

1990

)(R

obin

s,19

91)

Iran

Now

shah

r41

·950

·938

·656

·655

·759

·7(M

ehra

i&

Sun

derl

and,

1990

)(R

obin

s,19

91)

Iraq

Syr

iaK

urds

46·7

57·7

44·2

62·3

63·6

64·0

(Ojik

utu,

1965

)(R

obin

s,19

91)

Iraq

Syr

iaK

urds

57·4

(Sun

derl

and,

1979

)O

rigi

nal

Iraq

Syr

iaK

urds

60·2

(Sun

derl

and,

1979

)O

rigi

nal

Irel

and

Bal

inlo

ugh

65·3

(Sun

derl

and

etal

.,19

73)

(Rob

ins,

1991

)Ir

elan

dB

alin

loug

h65

·1(S

unde

rlan

det

al.,

1973

)(R

obin

s,19

91)

Irel

and

Car

new

64·6

(Sun

derl

and

etal

.,19

73)

(Rob

ins,

1991

)Ir

elan

dC

arne

w64

·4(S

unde

rlan

det

al.,

1973

)(R

obin

s,19

91)

Irel

and

Lon

gfor

d46

·258

·243

·763

·164

·764

·9(L

ees

etal

.,19

79)

Ori

gina

lIr

elan

dL

ongf

ord

64·9

(Lee

set

al.,

1979

)O

rigi

nal

Irel

and

Lon

gfor

d65

·2(L

ees

etal

.,19

79)

Ori

gina

lIr

elan

dR

ossm

ore

64·8

(Sun

derl

and

etal

.,19

73)

(Rob

ins,

1991

)Ir

elan

dR

ossm

ore

64·7

(Sun

derl

and

etal

.,19

73)

(Rob

ins,

1991

)Is

rael

57·6

(Sun

derl

and,

1979

)O

rigi

nal

Isra

el58

·8(S

unde

rlan

d,19

79)

Ori

gina

lJa

pan

Cen

tral

33·0

42·8

32·4

50·5

48·7

53·3

(Hul

se,

1967

)(R

obin

s,19

91)

Japa

nC

entr

al37

·146

·435

·153

·251

·755

·7(H

ulse

,19

67)

(Rob

ins,

1991

)Ja

pan

Cen

tral

54·6

(Har

vey

&L

ord,

1978

;H

ulse

,19

67)

Ori

gina

lJa

pan

Cen

tral

54·3

57·8

(Wal

sh,

1963

);re

des

tim

ated

Ori

gina

lJa

pan

Hid

aka

35·6

46·0

35·0

54·3

52·1

57·4

(Har

vey

&L

ord,

1978

)O

rigi

nal

Japa

nH

idak

a41

·051

·136

·158

·259

·460

·8(H

arve

y&

Lor

d,19

78)

Ori

gina

lJa

pan

Nor

ther

n34

·044

·133

·251

·449

·554

·1(H

ulse

,19

67)

(Rob

ins,

1991

)Ja

pan

Nor

ther

n37

·946

·835

·553

·552

·055

·7(H

ulse

,19

67)

(Rob

ins,

1991

)Ja

pan

Sou

thw

est

31·7

41·6

31·2

49·0

47·2

51·6

(Hul

se,

1967

)(R

obin

s,19

91)

Japa

nS

outh

wes

t37

·947

·035

·553

·451

·955

·5(H

ulse

,19

67)

(Rob

ins,

1991

)Jo

rdan

52·7

(Sun

derl

and

&C

oope

,19

73)

Ori

gina

lJo

rdan

52·2

(Sun

derl

and

&C

oope

,19

73)

Ori

gina

lJo

rdan

54·1

(Sun

derl

and,

1979

)O

rigi

nal

103

A

pp

end

ix(C

onti

nu

ed)

Pop

ulat

ion

desi

gnat

ion

(as

inT

able

5)57

5nm

595

nm60

0nm

655

nm67

0nm

685

nmO

rigi

nal

refe

renc

eS

ourc

eof

data

Jord

anA

zzar

qa48

·0(S

unde

rlan

d&

Coo

pe,

1973

)O

rigi

nal

Jord

anA

zzar

qa51

·5(S

unde

rlan

d&

Coo

pe,

1973

)O

rigi

nal

Ken

ya32

·4(S

unde

rlan

d,19

79)

Ori

gina

lK

urdi

shJe

ws

54·9

(Lou

rie,

1973

)O

rigi

nal

Kur

dish

Jew

s60

·0(L

ouri

e,19

73)

Ori

gina

lL

eban

on58

·0(S

unde

rlan

d,19

79)

Ori

gina

lL

eban

on58

·4(S

unde

rlan

d,19

79)

Ori

gina

lL

iber

ia15

·621

·212

·525

·422

·329

·4(O

jikut

u,19

65)

(Rob

ins,

1991

)L

ibya

Cyr

enai

ca53

·0(S

unde

rlan

d,19

79)

Ori

gina

lL

ibya

Cyr

enai

ca54

·0(S

unde

rlan

d,19

79)

Ori

gina

lL

ibya

Fez

zan

44·0

(Sun

derl

and,

1979

)O

rigi

nal

Lib

yaT

ripo

li53

·6(S

unde

rlan

d,19

79)

Ori

gina

lL

ibya

Tri

poli

55·2

(Sun

derl

and,

1979

)O

rigi

nal

Mal

awi

27·0

(Tob

ias,

1974

)(R

obin

s,19

91)

Mal

iD

ogon

13·7

18·2

19·4

32·4

27·8

33·7

(Hui

zing

a,19

68)

(Rob

ins,

1991

)M

ali

Dog

on14

·018

·519

·933

·528

·434

·5(H

uizi

nga,

1968

)(R

obin

s,19

91)

Mor

occo

56·4

FO

S73

from

OS

U-B

eals

sda

taba

seO

SU

-Bea

lss

data

base

Mor

occo

53·3

(Sun

derl

and,

1979

)O

rigi

nal

Moz

ambi

que

Cho

pi17

·8(W

enin

ger,

1969

)(R

obin

s,19

91)

Moz

ambi

que

Cho

pi21

·1(W

enin

ger,

1969

)(R

obin

s,19

91)

Nam

ibia

26·6

(Wei

ner

etal

.,19

64)

OS

U-B

eals

sda

taba

seN

amib

ia24

·5(W

eine

ret

al.,

1964

)O

SU

-Bea

lss

data

base

Nam

ibia

Oka

vang

o21

·6(W

eine

ret

al.,

1964

)(R

obin

s,19

91)

Nam

ibia

Oka

vang

o22

·1(W

eine

ret

al.,

1964

)(R

obin

s,19

91)

Nam

ibia

Oka

vang

o24

·2(W

eine

ret

al.,

1964

)(R

obin

s,19

91)

Nam

ibia

Oka

vang

o25

·6(W

eine

ret

al.,

1964

)(R

obin

s,19

91)

Nam

ibia

Oka

vang

o22

·2(W

eine

ret

al.,

1964

)(R

obin

s,19

91)

Nam

ibia

Oka

vang

o22

·6(W

eine

ret

al.,

1964

)(R

obin

s,19

91)

Nam

ibia

Oka

vang

o22

·4(W

eine

ret

al.,

1964

)(R

obin

s,19

91)

Nam

ibia

Oka

vang

o25

·5(W

eine

ret

al.,

1964

)(R

obin

s,19

91)

Nam

ibia

Reh

obot

hB

aste

r28

·2(W

eine

ret

al.,

1964

)(R

obin

s,19

91)

Nam

ibia

Reh

obot

hB

aste

r29

·4(W

eine

ret

al.,

1964

)(R

obin

s,19

91)

Nam

ibia

Reh

obot

hB

aste

r28

·0(W

eine

ret

al.,

1964

)(R

obin

s,19

91)

Nam

ibia

Reh

obot

hB

aste

r32

·4(W

eine

ret

al.,

1964

)(R

obin

s,19

91)

Nam

ibia

Reh

obot

hB

aste

r47

·9(W

eine

ret

al.,

1964

)(R

obin

s,19

91)

Nam

ibia

Reh

obot

hB

aste

r51

·9(W

eine

ret

al.,

1964

)(R

obin

s,19

91)

104 . . .

Ap

pen

dix

(Con

tin

ued

)

Pop

ulat

ion

desi

gnat

ion

(as

inT

able

5)57

5nm

595

nm60

0nm

655

nm67

0nm

685

nmO

rigi

nal

refe

renc

eS

ourc

eof

data

Nep

alE

aste

rn48

·9(W

illia

ms-

Bla

nger

o&

Bla

nger

o,19

91)

Ori

gina

lN

epal

Eas

tern

51·5

(Will

iam

s-B

lang

ero

&B

lang

ero,

1991

)O

rigi

nal

Nep

alE

aste

rn53

·6(W

illia

ms-

Bla

nger

o&

Bla

nger

o,19

91)

Ori

gina

lN

epal

Eas

tern

48·7

(Will

iam

s-B

lang

ero

&B

lang

ero,

1991

)O

rigi

nal

Nep

alE

aste

rn52

·1(W

illia

ms-

Bla

nger

o&

Bla

nger

o,19

91)

Ori

gina

lN

epal

Eas

tern

50·0

(Will

iam

s-B

lang

ero

&B

lang

ero,

1991

)O

rigi

nal

Nep

alE

aste

rn50

·1(W

illia

ms-

Bla

nger

o&

Bla

nger

o,19

91)

Ori

gina

lN

epal

Eas

tern

49·5

(Will

iam

s-B

lang

ero

&B

lang

ero,

1991

)O

rigi

nal

Nep

alE

aste

rn55

·7(W

illia

ms-

Bla

nger

o&

Bla

nger

o,19

91)

Ori

gina

lN

ethe

rlan

ds66

·5H

RM

64fr

omO

SU

-Bea

lss

data

base

OS

U-B

eals

sda

taba

seN

ethe

rlan

ds49

·060

·946

·167

·567

·968

·9(R

itge

r-A

ris,

1973

)(R

obin

s,19

91)

Net

herl

ands

47·4

58·7

44·8

65·3

66·0

66·7

(Rit

ger-

Ari

s,19

73)

(Rob

ins,

1991

)N

iger

iaIb

o12

·617

·012

·424

·621

·728

·2(B

arni

cot,

1958

)(R

obin

s,19

91)

Nig

eria

Yor

uba

10·9

14·3

10·4

20·6

18·8

23·6

(Bar

nico

t,19

58)

(Rob

ins,

1991

)N

iger

iaY

orub

a12

·216

·211

·823

·120

·626

·1(B

arni

cot,

1958

)(R

obin

s,19

91)

Nig

eria

Yor

uba

16·9

22·5

14·4

28·3

24·7

32·5

(Ojik

utu,

1965

)(R

obin

s,19

91)

Pak

ista

n52

·0(S

unde

rlan

d,19

79)

Ori

gina

lP

akis

tan

51·3

(Sun

derl

and,

1979

)O

rigi

nal

Pak

ista

n49

·853

·6(W

alsh

,19

63);

red

esti

mat

edO

rigi

nal

Pap

uaK

arke

r33

·2(H

arve

y&

Lor

d,19

78)

(Rob

ins,

1991

)P

apua

Luf

a30

·8(H

arve

y&

Lor

d,19

78)

(Rob

ins,

1991

)P

apua

Luf

a31

·6(H

arve

y&

Lor

d,19

78)

(Rob

ins,

1991

)P

apua

Mt

Hag

an29

·521

·934

·8(W

alsh

,19

63);

red

esti

mat

edO

rigi

nal

Pap

uaM

tH

agan

30·7

22·5

35·9

(Wal

sh,

1963

);re

des

tim

ated

Ori

gina

lP

apua

New

Gui

nea

35·3

HA

R71

from

OS

U-B

eals

sda

taba

seO

SU

-Bea

lss

data

base

Pap

uaP

ort

Mor

esby

36·2

25·3

41·0

(Wal

sh,

1963

);re

des

tim

ated

Ori

gina

lP

eru

Mar

anon

42·8

(Wei

ner

etal

.,19

63)

(Rob

ins,

1991

)P

eru

Mar

anon

43·3

(Wei

ner

etal

.,19

63)

(Rob

ins,

1991

)P

eru

Nun

oa47

·7C

ON

72fr

omO

SU

-Bea

lss

data

base

OS

U-B

eals

sda

taba

seP

hilip

pine

sM

anila

50·3

54·1

(Wal

sh,

1963

);re

des

tim

ated

Ori

gina

lR

ussi

aC

hech

en51

·9(S

unde

rlan

d&

Coo

pe,

1973

)O

rigi

nal

Rus

sia

Che

chen

55·0

(Sun

derl

and

&C

oope

,19

73)

Ori

gina

lS

audi

Ara

bia

52·5

(Sun

derl

and,

1979

)O

rigi

nal

Sou

thA

fric

a20

·727

·819

·339

·030

·841

·7(R

obin

s,19

91)

(Rob

ins,

1991

)S

outh

Afr

ica

23·4

31·2

21·1

41·5

33·1

44·3

(Rob

ins,

1991

)(R

obin

s,19

91)

Sou

thA

fric

a16

·422

·515

·430

·125

·232

·1(W

aser

man

n&

Hey

l,19

68)

(Rob

ins,

1991

)S

outh

Afr

ica

19·7

26·5

18·1

35·0

29·4

38·9

(Was

erm

ann

&H

eyl,

1968

)(R

obin

s,19

91)

105

A

pp

end

ix(C

onti

nu

ed)

Pop

ulat

ion

desi

gnat

ion

(as

inT

able

5)57

5nm

595

nm60

0nm

655

nm67

0nm

685

nmO

rigi

nal

refe

renc

eS

ourc

eof

data

Sou

thA

fric

aC

ape

26·8

36·1

27·9

45·2

43·7

49·2

(Was

erm

ann

&H

eyl,

1968

)(R

obin

s,19

91)

Sou

thA

fric

aC

ape

31·0

40·3

30·9

48·6

47·1

52·1

(Was

erm

ann

&H

eyl,

1968

)(R

obin

s,19

91)

Sou

thA

fric

aC

ape

50·1

(Wei

ner

etal

.,19

64)

(Rob

ins,

1991

)S

outh

Afr

ica

Cap

e51

·3(W

eine

ret

al.,

1964

)(R

obin

s,19

91)

Sou

thA

fric

aH

otte

ntot

45·5

(Wei

ner

etal

.,19

64)

(Rob

ins,

1991

)S

outh

Afr

ica

Hot

tent

ot48

·1(W

eine

ret

al.,

1964

)(R

obin

s,19

91)

Sou

thA

fric

aS

anC

entr

al41

·9(W

eine

ret

al.,

1964

)(R

obin

s,19

91)

Sou

thA

fric

aS

anC

entr

al45

·6(W

eine

ret

al.,

1964

)(R

obin

s,19

91)

Spa

inB

asqu

es45

·656

·042

·364

·564

·266

·4(R

ebat

o,19

87)

(Rob

ins,

1991

)S

pain

Bas

ques

44·1

55·0

41·7

63·4

63·8

65·5

(Reb

ato,

1987

)(R

obin

s,19

91)

Spa

inB

asqu

es44

·955

·241

·863

·963

·665

·8(R

ebat

o,19

87)

(Rob

ins,

1991

)S

pain

Bas

ques

44·5

55·0

41·7

63·3

63·7

65·3

(Reb

ato,

1987

)(R

obin

s,19

91)

Spa

inB

asqu

es42

·952

·641

·661

·161

·662

·3(R

ebat

o,19

87)

Ori

gina

lS

pain

Leo

n62

·866

·1(C

aro,

1980

)(R

obin

s,19

91)

Spa

inL

eon

61·0

64·6

(Car

o,19

80)

(Rob

ins,

1991

)S

pain

Leo

n60

·664

·3(C

aro,

1980

)(R

obin

s,19

91)

Spa

inL

eon

60·4

64·0

(Car

o,19

80)

(Rob

ins,

1991

)S

pain

Leo

n59

·562

·7(C

aro,

1980

)(R

obin

s,19

91)

Sud

an35

·5(S

unde

rlan

d,19

79)

Ori

gina

lS

waz

iland

14·5

18·4

14·3

27·6

24·3

32·4

(Rob

erts

etal

.,19

86)

(Rob

ins,

1991

)S

waz

iland

18·1

22·7

17·3

32·8

28·8

38·8

(Rob

erts

etal

.,19

86)

(Rob

ins,

1991

)S

yria

Dru

ze52

·8(S

unde

rlan

d&

Coo

pe,

1973

)O

rigi

nal

Tan

zani

aN

yatu

ra25

·3W

eine

r,19

69ci

ted

in(R

itge

rs-A

ris,

1973

b)(R

obin

s,19

91)

Tan

zani

aN

yatu

ra26

·3W

eine

r,19

69ci

ted

in(R

itge

rs-A

ris,

1973

b)(R

obin

s,19

91)

Tan

zani

aS

ande

we

28·1

Wei

ner,

1969

cite

din

(Rit

gers

-Ari

s,19

73b)

(Rob

ins,

1991

)T

anza

nia

San

dew

e29

·7W

eine

r,19

69ci

ted

in(R

itge

rs-A

ris,

1973

b)(R

obin

s,19

91)

Tun

isia

56·3

FO

S73

from

Ore

gon

data

OS

U-B

eals

sda

taba

seT

urke

y59

·3(S

unde

rlan

d,19

79)

Ori

gina

lT

urke

y59

·0(S

unde

rlan

d,19

79)

Ori

gina

lU

nite

dK

ingd

omC

umbe

rlan

d67

·0(S

mit

h&

Mit

chel

l,19

73)

(Rob

ins,

1991

)U

nite

dK

ingd

omC

umbe

rlan

d66

·5(S

mit

h&

Mit

chel

l,19

73)

(Rob

ins,

1991

)U

nite

dK

ingd

omL

ondo

n43

·954

·342

·661

·563

·363

·1(B

arni

cot,

1958

)(R

obin

s,19

91)

Uni

ted

Kin

gdom

Lon

don

41·4

52·4

40·9

59·9

62·0

61·5

(Bar

nico

t,19

58)

(Rob

ins,

1991

)U

nite

dK

ingd

omN

orth

ern

63·4

(Car

twri

ght,

1975

)(R

obin

s,19

91)

Uni

ted

Kin

gdom

Nor

ther

n62

·3(C

artw

righ

t,19

75)

(Rob

ins,

1991

)U

nite

dK

ingd

omN

orth

ern

45·2

54·8

43·1

61·7

63·2

62·3

(Har

riso

n&

Ow

en,

1964

)(R

obin

s,19

91)

106 . . .

Ap

pen

dix

(Con

tin

ued

)

Pop

ulat

ion

desi

gnat

ion

(as

inT

able

5)57

5nm

595

nm60

0nm

655

nm67

0nm

685

nmO

rigi

nal

refe

renc

eS

ourc

eof

data

Uni

ted

Kin

gdom

Nor

ther

n50

·059

·645

·766

·766

·968

·9(H

ulse

,19

73)

(Rob

ins,

1991

)U

nite

dK

ingd

omN

orth

ern

48·9

58·7

44·8

66·2

66·0

68·6

(Hul

se,

1973

)(R

obin

s,19

91)

Uni

ted

Kin

gdom

Nor

ther

n48

·959

·145

·065

·966

·168

·3(H

ulse

,19

73)

(Rob

ins,

1991

)U

nite

dK

ingd

omN

orth

ern

47·4

56·8

43·7

64·4

64·8

66·8

(Hul

se,

1973

)(R

obin

s,19

91)

Uni

ted

Kin

gdom

Wal

es67

·0(S

mit

h&

Mit

chel

l,19

73)

(Rob

ins,

1991

)U

nite

dK

ingd

omW

ales

65·9

(Sm

ith

&M

itch

ell,

1973

)(R

obin

s,19

91)

Uni

ted

Kin

gdom

Wal

es63

·5(S

mit

h&

Mit

chel

l,19

73)

(Rob

ins,

1991

)U

nite

dK

ingd

omW

ales

62·8

(Sm

ith

&M

itch

ell,

1973

)(R

obin

s,19

91)

Uni

ted

Kin

gdom

Wal

es63

·1(S

unde

rlan

d&

Woo

lley,

1982

)(R

obin

s,19

91)

Uni

ted

Kin

gdom

Wal

es62

·9(S

unde

rlan

d&

Woo

lley,

1982

)(R

obin

s,19

91)

Uni

ted

Kin

gdom

Wal

es62

·7(S

unde

rlan

d&

Woo

lley,

1982

)(R

obin

s,19

91)

Uni

ted

Kin

gdom

Wal

es62

·4(S

unde

rlan

d&

Woo

lley,

1982

)(R

obin

s,19

91)

Vie

tnam

51·3

55·0

(Car

bonn

el&

Oliv

ier,

1966

);re

des

tim

ated

(Har

riso

n,19

73)

Vie

tnam

56·8

FO

S73

from

OS

U-B

eals

sda

taba

seO

SU

-Bea

lss

data

base

Zai

re33

·9F

OS

73fr

omO

SU

-Bea

lss

data

base

OS

U-B

eals

sda

taba

seZ

aire

13·3

18·2

13·3

25·6

23·0

29·6

(Van

Rijn

-Tou

rnel

,19

66)

(Rob

ins,

1991

)Z

aire

15·4

21·7

16·0

30·1

26·6

36·1

(Van

Rijn

-Tou

rnel

,19

66)

(Rob

ins,

1991

)Z

aire

Kon

da28

·8(H

iern

aux,

1977

)O

rigi

nal

Zai

reK

onda

30·0

(Hie

rnau

x,19

77)

Ori

gina

l