1022-7954/05/4109- © 2005 Pleiades Publishing, Inc.
Russian Journal of Genetics, Vol. 41, No. 9, 2005, pp. 1046–1054. Translated from Genetika, Vol. 41, No. 9, 2005, pp. 1272–1282.
Original Russian Text Copyright © 2005 by Movsessian.
In spite of advances of molecular genetics in isolat-
ing mtDNA from skeletal remains, cranial morphology
remains to be the primary source of information on
polymorphism of ancient populations for biologists
studying human evolution. In view of this, of special
interest is the issue of constructing a genetically correct
phylogenetic tree of human races on the basis of mor-
The human skull offers a fairly large umber of dis-
crete variable, nonmetric traits: accessory bones, fon-
tanel ossicles, sutural bones, accessory and instable
foramina, processes, tubercles, etc.
As alternative variations of other morphological
traits, discrete cranial variations are typical phenes,
which permit to use methods and approaches of phe-
netic population studies in paleoanthropology. Phenes
are supposed to reﬂect the genetic constitution of an
individual, and their frequency, to describe genetic fea-
tures of the populations. The key principle of popula-
tion phenetics—implementing population genetic
methods and approaches to studying phene pools—is
very relevant in investigation of ancient human popula-
The phenetic approach to population characteriza-
tion even has some advantages over genetic analysis,
since several different genes can affect the phenotypic
expression of a trait, and thus “losing in precision of
marking, phenes win in covering the genotype” . It
is thought that each of discrete variations in the cranial
structure results from changes in about ten gene loci.
Consequently, 30 independent traits reﬂect 300 loci,
which constitute a considerable part of the genome .
Determination of genetic control of cranial varia-
tions is problematic, due to the fact that direct genetic
analysis of these traits on the fossil material is impossi-
ble. However, numerous studies have shown that genet-
ically distinct populations also clearly differ in frequen-
cies of nonmetric traits. For instance, matrices of inter-
population distances in nonmetric cranial trait
frequencies were found to correlate with the frequen-
cies of simple Mendelian characters .
Family analysis of metopic suture and variation in the
structure of squama occipitalis, based on X-ray studies,
has shown that these traits are controlled by a dominant
gene with varying penetrance [4, 5]. Genetic determina-
tion of the formation of sutural bones has also been
established . Later, the genetic nature of the palatine
torus and mandibular torus have been shown .
However, major studies in genetics of nonmetric
cranial traits were carried out on animals. The widest
popularity was gained by the theory on the threshold
type of inheritance of a number of traits, which has
been substantiated by the results of experiments on
mice. During several years Hans Gruneberg with col-
leagues studied skeletal variation in mice that were
morphologically analogous (and maybe even homolo-
gous) to their human counterparts [8, 9]. It has been
found that the trait frequencies in individual inbred
mouse lines were stable in several generations and
changed when lines were crossed, which suggested
genetic control of the traits. However, the interlinear
crosses did not produce Mendelian segregation in the
Phenetic Analysis in Paleoanthropology:
Phenogeography of Peoples of the World
A. A. Movsessian
Department of Anthropology, Biological Faculty, Moscow State University, Moscow, 119899 Russia;
Received December 27, 2004
—Phenetic diversity of peoples of the world in a system of nonmetric, discrete variable traits has been
studied. Sixty-two populations from North, Central, and Southeast Asia, Eastern and Western Europe, America,
East Africa, Australia, and Melanesia have been examined. The estimates of phenetic diversity within regions
) and the distances of the regions from the global means (
) proved to be comparable to the corresponding
estimates inferred from genetic data. This means that differentiation of populations in discrete variable traits is
related to the history of formation of their gene pools. A classiﬁcation tree of the world peoples constructed
using bootstrap implemented in the PHYLIP program package (Felsenstein, 1993) showed that the Australo–
Melanesian populations were close to the East African ones but separated from those of the Eurasian region.
The results of phylogenetic analysis of the reconstructed phene pools of the regional ancestral populations sup-
port the assumptions on the early colonization of Australia and Melanesia and on the later time of divergence
of the ancestors of modern Caucasoids and North Asian Mongoloids.