Ancestral Traits, Parental Populations, and Hybrids ' WILLIAM S. POLLITZER Department of Anatomy, Department of Anthropology, Genetics Training Committee, and Carolina Population Center, University of N o r t h Carolina, Chapel Hill,N o r t h Carolina The distance traveled over the centuries by any human population is ABSTRACT a rocky road; the fusion of the paths of two meandering populations often leads to a n even rougher one. Tracing such a wayward child of history back to its long-lost parents presents complex problems for the physical anthropologist. For those parents may share unequally in their legacy to their offspring a t the start of the journey. Many genes may disappear into the gutters of genctic drift; some byways may stop at the barricades of selection; fresh migrants may join by side roads to swell the stream; and inbreeding may narrow the roadbed. Searching for the genes that the hybrid population holds in common with alleged parents provides one clue to the relative role of the ancestors. Comparison of measurements of the living hybrid with the nearest known relatives of the presumed parents yields another clue. Resuscitation of ancestors from their bones and teeth supplies yet another link. Distinctive features of such skeletal remains are most valuable where their inheritance is understood. Bones may yield another clue from substances left within them of known genetic control. And finally all the evidence of history, every glimpse caught of the population along the route, gives knowledge of the contribution of the forebears to the living people. The ideal estimate will utilize a wide variety of information about the known groups. While the most primitive of human societies have recognized the vast distinction between themselves and their neighbors, it is only in recent times that sophisticated scientists have sought to measure the precise degree of difference between two or more populations and the probable contribution of parental peoples to hybrids. Increasing technological skill adds to the procedures available to the anthropologist who indulges in this endeavor. Materials available from extinct people are bones and teeth, which can be measured in considerable detail for comparison with similar remains of other populations. Distinctive and unusual traits such as Carabelli's cusps noted among them provide clues to affinity. Where inherited disease, such as sickle cell anemia, leaves its tell-tale stamp on a skull table, x-ray may aid the eye in suggesting classification in the past. Where bones from a larger number of persons, identifiable as individuals, are available, ABO typing by inhibition methods should at least provide supplementary data. First performed by Candela ('39) and by Boyd ('39), and rightly questioned by others (Thieme and Otten, '57), these tests for blood group substances in ancient bones were greatly advanced by the AM. J. PHYS. ANTHROP.,30: 41-20. work of Gray ('58) who controlled bacterial contamination. We have been able to show that the bacteria cultured from excavated ancient bones are not for the most part the ones known to have a significant amount of A, B, and 0 substance. The further chemical purification of the material from cancellous bones holds promise for future identification. In our study of the Catawba Indians (Pollitzer, '67) we were able to relate the present remnant of these Siouan people to similar populations of the Carolinas some 400 years earlier, and even to the Shell Mound people of Indian Knoll of 3000 to 5000 B.C., using primarily cranial measurements. Krogman ('62) has contributed methods for correcting from skeletal measurements to measurements on the living. The typing of remains of 16th century people of the Yadkin River as 24 Group 0 and 2 Group A, while in line with living Indians of the area, is far too small a sample for significance. As bones of the past and their interpretation will be skillfully examined in subsequent talks on this symposium, I may be forgiven this mere skeleton outline of the 1 Dedicated to m y mother in memory of m y father, Dr. R. M. Pollitzer (1883-1969). 415 416 WILLIAM S. POLLITZER topic as I hasten to flesh out a consideration of the living. Since the days of Karl Pearson, methods have been sought to estimate “racial likeness” from physical traits. Guha advanced such methods through his study of the peoples of India (’31). Mahalanobis and his colleagues (’49) contributed D2 as a measure of “distance,” i.e., dissimilarity between populations. Employing the correlation between traits as well as the difference between traits in the populations, he analyzed the people of India for their relative “distance.” One inherent shortcoming of the Da procedure is the increase of distance with each trait added. We are beset, too, by the choice of traits: Is total morphological face height a preferable parameter to upper face height? Is bizygomatic better than bigonial width? We must be careful to study only those who have reached maturity for growth alters these relationships, and not necessarily at the same rate in all people. We should calculate a different distance for males and for females, a procedure which can result in estimating that the men of tribe A are closely related to the men of tribe B, while their women are (perhaps understandably) more “distant.” Basically such measures as Dzrest upon likeness, whether of genetic Rature or not. Since morphologic traits are governed by several genes, it is possible for the same phenotype to be expressed by quite different gene combinations. And the same genotype in different environments may be phenotypically different. If we choose those traits most unlike in the populations to be compared, we have a greater chance of finding homozygosity at the loci involved. Under an additive genetic model similar phenotypes for a given trait in a hybrid between the two will indicate the same proportion of genes for that trait even if the underlying genotype varies. Genetic drift is less likely to alter these morphologic traits due to the coadaptation of the genes underlying them. Where great passage of time and distance are involved selection may also operate to obscure the fundamental kinship between populations. While the use of such monogenic traits as the blood types do not suffer from the same defects they have their problems, too. By the use of ABO genes alone one can prove that the Eskimos are “blood brothers” to the Bantu tribes of Africa. By the use of Rh positive and negative alone, a hybrid population of an equal mixture of Chinese and Englishmen appear identical to American Negroes. These single gene traits are subject to genetic drift, especially where the bottlenecks of history have reduced a flourishing river of people to a thin stream. And the alterations due to selection, such as disease associations, are only now coming to be considered if their cause is not as yet defined and evaluated. But with the large number of loci made available through modern technology, it is now possible not only to calculate the similarity between two related populations, but to compare the answer to this problem given by each genetic system. The contribution of one population to another to form a hybrid provides a relatively simple model of gene flow, successi d l y used by Glass and Li ( ’ 5 3 ) to calculate admixture of Africans and Europeans in the formation of the American Negro and even to calculate the rate of admixture. Roberts and Hiorns ( ’ 6 2 ) devised a least squares method of greater general applicability in the dynamics of racial intermixture, providing for more than two parental populations. They eliminate one allele from each system, and the estimate obtained depends on which is eliminated. Krieger et al. (’65) proposed a maximum likelihood solution to the problem using a modcl that actual data do not fit. Elston in as yet unpublished work has devised an improved least squares estimate and suggested changes in the maximum likelihood estimate, utilizing all alleles of each genetic system. In all calculations of the relative contribution of parental to hybrid populations, mere similarity does not prove descent, and the supportive evidence of history is needed to prove derivation, while language and culture may provide additional clues. With the population as with the individual i t is a wise child who knows his own father. Let us now turn to certain populations of the United States Southeast and explore their composition. The Melungeons are a people of uncertain origin, distinctive features, and tawny ANCESTORS AND HYBRIDS coloration who have long inhabited the crests of the Appalachian mountains in western Virginia and eastern Tennessee. The source of the name is in dispute; it has been attributed to French “melange” for “mixed,” to Portuguese “melungo” for “shipmate,” to Greek “melan” for “colored,” and to a possible corruption of two surnames common in the isolate. The earliest writings viewed the Melungeons as a whiteIndian-Negro mixture, and Dromgoole (1891) believed there were Cherokee Indian, African Negro, English, and Portuguese elements in the population, each of which was indicated by distinctive surnames. The people themselves have long iavored the idea of Portuguese ancestry. In his thorough demographic study Price (’51) noted an aggregation of about 1,000 in Hancock County, Tennessee, 500 in adjacent Lee County, Virginia, and smaller pockets in the neighboring highlands. From careful study of the census he concluded that the nucleus of the Melungeons were “free persons of color” who had migrated from North Carolina in the last decade of the eighteenth century. Today the isolate has contracted through out-migration, and new elements have joined it through marriage. In our survey of 177 persons studied in 1965-1966, 128 were adults and 49 were children. Of birthplace recorded for 125 adults, 118 were born within Hancock and Lee counties. Blood samples were obtained on young and old, and physical measurements on the adults. Skin color determinations by reflectometry bore out the wellknown tendency of women to average slightly lighter than men, and older people were lighter than younger. In general morphology, hair form and pigmentation, they resemble whites more closely than Indian or Negro groups. Striking among blood type findings is the 50% incidence of Group A, half of which have the A, gene. The M gene is almost twice as coinmon as N. Phenotypes with Rh, comprise over half the sample; and Rho, like Hgb. S, is absent. Haptoglobin 2-2 patterns are more than twice as frequent as 1-1. Using ABQ, M-N, Rh, Hgb., and Hp. loci it is possible to test several hypotheses of probable parental populations and their contribution to the hybrid. If we consider English, Portuguese, 417 Negro, and Indian gene frequencies we find that all methods agree that Portuguese is the more likely of the two European ancestors, contributing the majority of the genes with the remainder contributed more by Indian than Negro. The data are best fitted by a 90% Portuguese component, with some Indian and Negro admixture. If we assume Portuguese are the only European component and there is no Negro component, the Portuguese contribute 94% of the genes, and the Indians account for the remaining 6 % . If we assume that English was the only European component, they account for 94% of the genes of the hybrid and the remainder can be explained by Negro admixture or by Negro and Indian admixture. Although these computations agree that European ancestry predominates, and that Portuguese is one likely origin of the European genes, they suggest even more forcibly that with several possible parental populations, different answers are equally likely. When we reflect that only a few individuals from any of these populations entered into the isolate initially our estimates become more suspect. (While we say “triracial” no three individuals mate; in the language of the statistician we not only lack the “means” but such a practice would constitute more than a “standard deviation.”) Physical appearance, history, and census reports combine with genetic studies to suggest that the Melungeons are probably derived primarily from a European people, possibly Portuguese but equally other ethnic groups as well, plus some Indian and (perhaps) Negro strains. Let us next turn our attention to the Seminole Indians of Florida. Numbering a thousand, they cluster in three reservations: Dania, just north of Miami; Big Cypress, on the edge of the Everglades; and Brighton, on the northwestern rim of Lake Okeechobee. Disease, warfare and slaving raids virtually destroyed the native Indian population of Florida by 1715, a void slowly filled primarily by Creeks from Georgia. One group of Muskogean speakers, derived from the Lower Creek tribe, the Oconee, settled at Alachua, near present Gainesville, in north central Florida in 1750; they formed the nucleus of the Cow Creek Indians, who later settled in the 418 WILLIAM S. POLLITZER Brighton area and speak what is now known as the Seminole language. A second group, speaking the Hitchiti dialect, settled at Appalachee around Lake Miccosuke near Tallahassee, in northwestern Florida; and subsequently peopled the Big Cypress community, where they still speak the distinctive but related Mikasuki language, The population of the Seminoles of Florida has varied widely from a peak of 5,000 in the 1820's, 500 in 1853, to 208 in 1884, and back up to 590 in 1930 and a thousand today, MacCauley in his report of 1884 describes the men as dark, coppercolored, tall, with square faces; the women as under average height and shapely; he reports with evidently equal favor that their health was good, that education was stoutly resisted, and that Christian missionaries had had little influence on their religion. We were able to obtain blood samples and pedigree information on 416 Seminoles six years and over on their three reservations, and measurements on the adults. In shape and appearance they resemble other living Indian tribes. They are somewhat shorter, more round-headed, roundfaced and wider-nosed than the full-blood Sioux of the plains. In two-thirds, hair is straight and in three-fourths it is deep black. Females are shorter, lighter, and more long-headed than males, and show a curious tendency toward marked obesity in the lower half of the body beginning in their early 30's. In blood factors the Seminoles are very similar to the Cherokee Indian full-bloods. In their general morphology, the Indians of Big Cypress are distinctive from those of Brighton, in keeping with their history. While the sickle cell trait and Diego occur only in Big Cypress the differentiation of the peoples of this reservation from those of Brighton and Dania is not as clear cut for monogenic traits as it is for the polygenic ones. This probable operation of genetic drift is also indicated when blood factors in those under 2 1 are compared with those 2 1 and over. The younger people show no Group B, 5 cases of Ns, and 3 of rh negative, while the older people show three instances of Group B and none of Ns or rh negative. Computations based on genetic loci in Seminoles and parental populations show the very high Indian component, but they also reflect variations on the three reserv a tions. These studies remind us of the pitfalls as well as the possibilities in estimating the genetic contribution of parental groups to hybrid population. Due to the fundamental theoretical difference between a morphological estimate like Da and one based on gene frequencies, any combination of the two remains difficult and arbitrary. When we know more about the role of selection on both kinds of traits, when we have information on the genic foundations of the morphology, and when we use a large battery of measurements on a large number of people, we shall be able to make the wisest estimates. Even so, we would do well to check all available historical and archeological data for the continuity of the populations over the span OP space and time. In conclusion, I want to thank all those who made the migration here today. While I may not have produced any mutation in your thinking, I hope I have made such a wise selection of materials that all of you have been able to follow my genetic drift. ACKNOWLEDGMENTS I am deeply indebted to Dr. D. F. Roberts at Newcastle-on-Tyne, and to Dr. Robert Elston at the University of North Carolina for their development of methods and their calculations of parental contribution to hybrid populations. I am also grateful for the help in blood typing given by Dr. Jerry Niswander and Mr. Chad Leyshon at the National Institutes of Health, and for the determination of serum proteins and red cell enzymes made by Dr. Don Rucknagel and Dr. Richard Tashian and Dr. Lowell Weitkamp at the University of Michigan. I am particularly appreciative of the assistance given by colleagues at Chapel Hill: Mr. W. H. Brown, Mrs. Lois Tillman and Mrs. Mikki Austin for collecting blood samples; Dr. David Phelps (now at the University of Florida) and Miss Margot Hall for the work on the blood typing of bone: Dr. Tom Donnelly and Mrs. K. Namboodiri for analysis of data; Miss Cassi Parrish and Mrs. R. M. Bock (now at the Univer- ANCESTORS AND HYBRIDS sity of Chicago) for construction of pedigrees. I am happy to thank Dr. W. F. Hill in Florida and Dr. Truett Pierce in Tennessee for obtaining the cooperation of the subjects, and, finally, to thank the subjects themselves. These studies have indeed been “group research,” and I am well aware of the criticism that group research is founded on the theory that if one makes nine women pregnant he can produce a baby in one month. I can only reply that it has been fun trying. These studies have been supported by grant GM-10601 from NIH and by funds from the United Medic21 Foundation of North Carolina. LITERATURE CITED Boyd, L. T., and W. C. Boyd 1939 Blood group ing reactions of preserved bone and muscle. Am. J. Phys. Anthrop., 25: 421-434. Candela, P. B. 1939 Blood tests on stains, mummified tissue and cancellous bone. Am. J. Phys. Anthrop., 25: 187-214. Dromgoole, W. A. 1891 The Malungeons. The Arena, 3: 470-479. Elston, R. C. Personal Communication. Glass, 13. B., and C. C. Li 1953 The dynamics of racial intermixture - a n analysis based on the American Negro. Am. J. Hum. Genet., 5 : 1-20. 419 Gray, M. 1958 A method for reducing nonspecific reactions i n the typing of human skeletal material. Am. J. Phys. Anthrop., 16: 135139 and Personal Communication. Guha, B. S. 1931 Census of India, V. 1, part I11 A. Krieger, H., N. E. Morton, M. P. Mi, Eliane Azevedo, A. Freire-Maia and N. Yasuda 1965 Racial admixture i n Northeastern Brazil. Annals Human Genetics, 29: 113-125. Krogman, W. M. 1962 The human skeleton in forensic medicine. Charles C Thomas, Springfield. IXacCauley, Clay 1884 The Seminole Indians of Florida. Bureau of American Ethnology. Annual Report. Mahalanobis, P. C., D. N. Majumdar and C. R. Rao 1949 Anthropometric survey of the United Provinces, 1941: A statistical study. Sankya, 9: 89-324. Pearson, K. 1926 On the co-efficient of racial likeness. Biometrika, XVIII: 106-117. Pollitzer, W. S., D. S. Phelps, R. E. Waggoner and W. C. Leyshon 1967 Catawba Indians: morphology, genetics, and history. Am. J. Phys. Anthrop., 26: 5-14. Price, E. T. 1951 The Melungeons: a mixedblood strain of the southern Appalachians. Geographical Review, 41 : 256-271. Roberts, D. F., and R. W. Hiorns 1962 The dynamics of racial intermixture. Am. J. Hum. Genet., 14. 261-277. Thieme, F. P., and C. M. Otten 1957 The unreliability of typing aged bonc. Am. J. Phys. Anthrop., 15: 387-398.