Patterns of ancestral human diversity: an analysis of Alu-insertion and restriction-site polymorphisms.

We have analyzed 35 widely distributed, polymorphic Alu loci in 715 individuals from 31 world populations. The average frequency of Alu insertions (the derived state) is lowest in Africa (.42) but is higher and similar in India (.55), Europe (.56), and Asia (.57). A comparison with 30 restriction-site polymorphisms (RSPs) for which the ancestral state has been determined shows that the frequency of derived RSP alleles is also lower in Africa (.35) than it is in Asia (.45) and in Europe (.46). Neighbor-joining networks based on Alu insertions or RSPs are rooted in Africa and show African populations as separate from other populations, with high statistical support. Correlations between genetic distances based on Alu and nuclear RSPs, short tandem-repeat polymorphisms, and mtDNA, in the same individuals, are high and significant. For the 35 loci, Alu gene diversity and the diversity attributable to population subdivision is highest in Africa but is lower and similar in Europe and Asia. The distribution of ancestral alleles is consistent with an origin of early modern human populations in sub-Saharan Africa, the isolation and preservation of ancestral alleles within Africa, and an expansion out of Africa into Eurasia. This expansion is characterized by increasing frequencies of Alu inserts and by derived RSP alleles with reduced genetic diversity in non-African populations.

Population structure and history in East Asia.

Archaeological, anatomical, linguistic, and genetic data have suggested that there is an old and significant boundary between the populations of north and south China. We use three human genetic marker systems and one human-carried virus to examine the north/south distinction. We find no support for a major north/south division in these markers; rather, the marker patterns suggest simple isolation by distance.

Genetic traces of ancient demography.

Patterns of gene differences among humans contain information about the demographic history of our species. Haploid loci like mitochondrial DNA and the nonrecombining part of the Y chromosome show a pattern indicating expansion from a population of only several thousand during the late middle or early upper Pleistocene. Nuclear short tandem repeat loci also show evidence of this expansion. Both mitochondrial DNA and the Y chromosome coalesce within the last several hundred thousand years, and they cannot provide information about the population before their coalescence. Several nuclear loci are informative about our ancestral population size during nearly the whole Pleistocene. They indicate a small effective size, on the order of 10,000 breeding individuals, throughout this time period. This genetic evidence denies any version of the multiregional model of modern human origins. It implies instead that our ancestors were effectively a separate species for most of the Pleistocene.

Effect of sample bias on paleodemographic fertility estimates.

Paleodemographers must work to understand how representative any archaeologically recovered skeletal series is and the potential effects of series bias on their demographic reconstructions. We examine two forms of bias: 1) infant underenumeration caused by differential preservation or incomplete archaeological recovery and 2) the underenumeration of individuals over age 45 related to methodological bias. We generated 60 simulated skeletal series of 250 individuals each based on the Brass ([1971] Biological Aspects of Demography (London: Taylor and Francis), pp. 69-110) logit models. In the first test, age bias was introduced deterministically for all individuals with age at death over 40 years using the Lovejoy et al. ([1985] Am. J. Phys. Anthropol. 68:1-14) bias estimates. In the second test, 50% of all individuals under 5 years old were removed from each simulated distribution. The simulated series were analyzed using the model life table fitting procedure developed by the authors (Milner et al. [1989] Am. J. Phys. Anthropol. 80:49-58; Paine [1989] Am. J. Phys. Anthropol. 79:51-62). Forms of adult age estimation bias described by Lovejoy and coworkers inflate estimates by 10-20% of the true crude birth rate (CBR) (the number of births per year per 1,000 population). Overestimation of fertility and birth rates increases both absolutely and as a percentage of the true rate as population growth increases. This bias is very consistent. Because Lovejoy and colleagues have estimated the methodological bias itself, its effects can be estimated. Infant underenumeration is a more serious obstacle. It is not presently possible to estimate infant underenumeration reliably without prior knowledge of fertility rates. This reduces fertility reconstructions based on infant-biased samples to minimum fertility estimates.

Alu evolution in human populations: using the coalescent to estimate effective population size.

There are estimated to be approximately 1000 members of the Ya5 Alu subfamily of retroposons in humans. This subfamily has a distribution restricted to humans, with a few copies in gorillas and chimpanzees. Fifty-seven Ya5 elements were previously cloned from a HeLa-derived randomly sheared total genomic library, sequenced, and screened for polymorphism in a panel of 120 unrelated humans. Forty-four of the 57 cloned Alu repeats were monomorphic in the sample and 13 Alu repeats were dimorphic for insertion presence/absence. The observed distribution of sample frequencies of the 13 dimorphic elements is consistent with the theoretical expectation for elements ascertained in a single diploid cell line. Coalescence theory is used to compute expected total pedigree branch lengths for monomorphic and dimorphic elements, leading to an estimate of human effective population size of approximately 18,000 during the last one to two million years.

Microsatellite diversity and the demographic history of modern humans.

We have examined differences in diversity at 60 microsatellite loci among human population samples from three major continental groups to evaluate the hypothesis of greater African diversity in this rapidly evolving class of loci. Application of a statistical test that assumes equal mutation rates at all loci fails to demonstrate differences in microsatellite diversity, while a randomization test that does not make this assumption finds that Africans have significantly greater microsatellite diversity (P < 10(-8)) than do Asians and Europeans. Greater African diversity is most apparent at loci with smaller overall variance in allele size, suggesting that the record of population history has been erased at repeat loci with higher mutation rates. A power analysis shows that only 35-40 microsatellites are needed to establish this difference statistically, demonstrating the considerable evolutionary information contained in these systems. On average, African populations have approximately 20% greater microsatellite diversity than do Asian and European populations. A comparison of continental diversity differences in microsatellites and mtDNA sequences suggests earlier demographic expansion of the ancestors of Africans.

Assessing the reliability of paleodemographic fertility estimators using simulated skeletal distributions.

The reliability of published paleodemographic fertility reconstruction methods was assessed using simulated age-at-death distributions and a published cemetery series from a population with known birth rates. In the first test, the Brass ([1971] Biological Aspects of Demography, pp. 69-110) LOGIT models were used to generate 180 simulated skeletal samples of various sizes (N = 50, 100, 250) from hypothetical populations with known demographic rates. The base populations were expanding (r = 0.01), stationary, or declining (r = -0.01), yet all had the same life expectancy. Growth differences resulted from different fertility rates. The simulated skeletal series were then analyzed using the model life table fitting procedure outlined by Paine ([1989a] Am. J. Phys. Anthropol. 79:51-62), three commonly employed age ratio tests (Bocquet-Appel and Masset [1892] J. Hum. Evol. 11:321-333; Buikstra et al. [1986] Am. Antiquity 51:528-546), and one age-at-death ratio not previously published. In the second test the model life table fitting procedure was used to estimate fertility for a historical population, the Newton Plantation, Barbados (Corruccini et al. [1989] Am. Antiquity 54:609-614), with known demographic characteristics.

Craniometric variation, genetic theory, and modern human origins.

Recent controversies surrounding models of modern human origins have focused on among-group variation, particularly the reconstruction of phylogenetic trees from mitochondrial DNA (mtDNA) and the dating of population divergence. Problems in tree estimation have been seen as weakening the case for a replacement model and favoring a multiregional evolution model. There has been less discussion of patterns of within-group variation, although the mtDNA evidence has consistently shown the greatest diversity within African populations. Problems of interpretation abound given the numerous factors that can influence within-group variation, including the possibility of earlier divergence, differences in population size, patterns of population expansion, and variation in migration rates. We present a model of within-group phenotypic variation and apply it to a large set of craniometric data representing major Old World geographic regions (57 measurements for 1,159 cases in four regions: Europe, Sub-Saharan Africa, Australasia, and the Far East). The model predicts a linear relationship between variation within populations (the average within-group variance) and variation between populations (the genetic distance of populations to pooled phenotypic means). On a global level this relationship should hold if the long-term effective population sizes of each region are correctly specified. Other potential effects on within-group variation are accounted for by the model. Comparison of observed and expected variances under the assumption of equal effective sizes for four regions indicates significantly greater within-group variation in Africa and significantly less within-group variation in Europe. These results suggest that the long-term effective population size was greatest in Africa. Closer examination of the model suggests that the long-term African effective size was roughly three times that of any other geographic region. Using these estimates of relative population size, we present a method for analyzing ancient population structure, which provides estimates of ancient migration. This method allows us to reconstruct migration history between geographic regions after adjustment for the effect of genetic drift on interpopulational distances. Our results show a clear isolation of Africa from other regions. We then present a method that allows direct estimation of the ancient migration matrix, thus providing us with information on the actual extent of interregional migration. These methods also provide estimates of time frames necessary to reach genetic equilibrium. The ultimate goal is extracting as much information from present-day patterns of human variation relevant to issues of human origins.(ABSTRACT TRUNCATED AT 400 WORDS)

Signature of ancient population growth in a low-resolution mitochondrial DNA mismatch distribution.

A mismatch distribution is a tabulation of the number of pairwise differences among all DNA sequences in a sample. In a population that has been stationary for a long time these distributions from nonrecombinant DNA sequences become ragged and erratic, whereas a population that has been growing generates mismatch distributions that are smooth and have a peak. The position of the peak reflects the time of the population growth. The signature of an ancient population expansion is apparent even in the low-resolution mtDNA typings described by Merriwether et al. (1991). The smoothness of the mismatch distribution, an indicator of population expansion, is hardly affected by population structure, whereas mean sequence divergence increases in a pooled sample from highly isolated subpopulations.

Age structure and sex-biased mortality among Herero pastoralists.

Details of the population pyramid of living Herero and Mbanderu of Botswana suggest that infant and childhood mortality of males has been substantially greater than that of females. Direct tests from reproductive histories show that the hazard ratio is approximately 3 to 1 in favor of female survival in infancy and 2 to 1 in childhood. This biased mortality began in about 1960 in concert with recovery from infertility. A model of asymmetric fitness benefits between siblings is weakly supported. Logistic regression shows that heterogeneity among mothers explains much of the mortality of children under 2 years of age. A direct test for heterogeneity provides strong support for the hypothesis. Field methods appropriate for anthropologists are contrasted with those that are standard in demography. Finally, a contrast between these data and those from south Asia suggests that strong sex preferences may exist without being culturally articulated. Cultural norms do not necessarily coincide with behavior.

PIP: Ethnographer collected reproductive histories from about 600 women basically among Hereros in western Ngamiland, Botswana. They constructed a population pyramid which revealed a generation time of 19-24 years (shorter than that of most human populations). The Herero have had pathologic sterility, which may account for the shortened generation time. Fertility returned around 1960. Further, the pyramid had a missing bulge of males in the 1960s. Applying direct tests to the reproductive histories, they ascertained that, beginning around 1960, female infants were almost 3 times as likely to survive then males (p=.000001) and female children were 2 times as likely to survive than males (p.01). Yet the risk of infant mortality for males in Botswana was only 1.5 times that of females. The logistic regression showed that male death was more random than female death and that the risk rose with previous deceased siblings but fell with the number of previous living siblings of both sexes. These results indicated that heterogeneity among mothers in their ability to keep their children alive existed and that characteristics of the mothers and not those of the child were risk factors for infant and early childhood mortality. Next they used a likelihood ratio to test for heterogeneity among mothers which revealed that 72% of the mothers never had a daughter die and the risk for the remaining mothers was .20 (p.001), but the risk for having a son die was .03 for 55% of the mothers and .3 for 45% (p.0001). This test also indicated very significant heterogeneity among women. Still the women always denied any sex preference or differential care of their children. Yet they did admit the higher death rate for male infants. Therefore stated cultural beliefs are not necessarily guides to behavior. In conclusion, public health workers in Botswana should concentrate on mothers' characteristics.

Pattern matching of age-at-death distributions in paleodemographic analysis.

Model age-at-death distributions are generated from fertility and mortality rates derived from two present-day, traditional human societies with widely differing cultural systems: the !Kung hunters-and-gatherers and Yanomamo horticulturalists. Visual examination of these models demonstrates that fertility has more of an effect than mortality on the overall configuration of the age-at-death distributions of stable populations. Comparisons with a late prehistoric Oneota skeletal sample from the American Midwest illustrate how reference age-at-death schedules can be used 1) to identify whether a given skeletal sample approximates an age-at-death distribution expected of an extant human population and 2) to provide a basis for developing further testable hypotheses about the demographic and cultural characteristics of past populations.

Population structure and quantitative characters.

A migration matrix model is used to investigate the behavior of neutral polygenic characters in subdivided populations. It is shown that gametic disequilibrium has a large effect on the variance among groups but none at all on its expectation. The variance of among-group variance is substantial and does not depend on the number of loci contributing to variance in the character. It is just as large for polygenic characters as for single loci with the same additive variance. This implies that one polygenic character contains exactly as much information about population relationships as one single-locus marker. The theory is compared with observed differentiation of dermatoglyphic and anthropometric characters among Bougainville islanders.

Gammaglobulin groups of the Khoisan peoples of Southern Africa: evidence for polymorphism for a Gm1,5,13,14,21 haplotype among the San.

The Gm and Inv types were determined for eight San (Bushman) populations, two Khoikhoi (Hottentot) populations, one Coloured population, 112 San families in which the genotypes of the parents could be unambiguously determined, and for 65 San families in which the genotype of one or both parents could not be determined with certainty. The population and family data establish that the haplotype array of the San is composed of Gm1,21, Gm1,13, Gm1,5,13,14, and Gm1,5,13,14,21; Gm1,5,6 and Gm1,5,6,14 are also present but may have been acquired through admixture with Negroes. The Gm1,5,13,14,21 haplotype has not been found to be polymorphic in any other population. The haplotype array of the Khoikhoi is composed of Gm1,2,21, Gm1,13, and Gm1,5,13,14; Gm1,5,6 and Gm1,5,6,14 are also present but, as in the case of the San, may be due to admixture. The San and Khoikhoi differ from each other in that the former have the Gm1,21 and Gm1,5,13,14,21 haplotypes not present in the latter, and the Khoikhoi have the Gm1,2,21 haplotype not present in the San. These three haplotypes and Gm1,13 serve to distinguish the Khoisan people from other African peoples.