Mountains and genes: Population history of the Pyrenees

The Pyrenees Mountains rise where the Iberian Peninsula and mainland Europe meet. The range is 430 km long from the Atlantic to the Mediterranean shore. Its highest peak, Aneto (3404 m), lies in the central section, which is also the widest section (160 km). The western end smoothly joins the lower Cantabrian range, whereas to the east the Pyrenees fall abruptly to the Mediterranean Sea. Both ends are crossed by the main roads between France and Spain; there are few passes in the center, and all are above 1600 m. Most Pyrenean valleys lie in a north-south direction, although a few, such as the Cerdanya, have the same east-west orientation as the range. The Spanish-French border runs from east to west, following the crest line and leaving room for the tiny principality of Andorra.

The Pyrenees have been continuously populated since Paleolithic times; they have often acted as a refuge when the surrounding plains were not safe. Four extant languages were born in the Pyrenees: Basque, a non-Indo-European relict language, whose relationship to South Caucasian (Kartvelian) languages is widely disputed (Lafon 1933; Ruhlen 1991); and Gascon, Aragonese, and Catalan, which are Romanic dialects. Gascon and Aragonese are on the verge of complete replacement by French and Spanish, respectively; Catalan, with more than six million speakers, enjoys official status. It should be noted that Gascon is a distinct dialect of Occitan, certain features of which point to the influence of a Bascoid substrate (Allieres 1986). Languedocien, another Occitan dialect, is mainly spoken in the Languedoc plain, but it also reaches the Pyrenees in the Foix valley.

There is a long tradition of anthropological studies in the Pyrenees. In 1889 Aranzadi described the peculiar cranial morphology of the Basques and suggested that they were the remnants of the most ancient peoples of Europe. When the first Rh blood group frequencies were available for the Basques (Etcheverry 1945), Mourant (1947) argued in favor of Aranzadi's hypothesis on the origin of the Basque genetic distinctiveness. Later, Alcobe (1948, 1974), Fuste (1955), and Ruffie (1958), among others, devoted their work to blood groups, dermatoglyphics, and cranial morphology of ancient and extant Pyrenean populations.

Since the discovery and diffusion of the techniques for the detection of a large number of polymorphic genetic markers, Pyrenean populations have been studied repeatedly. A team at the Centre de Recherches sur le Polymorphisme Genetique des Populations Humaines (Toulouse, France) studied 13 polymorphisms in 11 areas on the French side of the Pyrenees (Vergnes et al. 1980, 1981). Subsequently, two Pyrenean populations (Bearn and Catalogne) were included in a comprehensive survey of 14 French regions (Ohayon and Cambon-Thomsen 1986). On the southern side the Basques have been thoroughly studied [for a summary, see Agui et al. (1991); in Catalonia several Pyrenean valleys have also been sampled (Aluja 1987; Moral 1988; among others)]. Moral (1988) used 7 markers (20 alleles) to compute genetic distances between 6 Pyrenean populations, which is only a small fraction of the information now available. Hazout et al. (1991) applied novel computational techniques in an attempt to relate genetic differentiation to geographic location in 11 populations of the French Pyrenees; they used one polymorphism, GM. Therefore a remarkable body of knowledge on the genetic polymorphisms in Pyrenean populations has grown in the last decade; roughly one-third of the genetic data available for the Iberian Peninsula and France (Calafell and Bertranpetit 1994) concerns Pyrenean populations. Nevertheless, there is still much work to be done to understand the genetic makeup of the populations dwelling between France and Spain.

In a previous work (Calafell and Bertranpetit 1994), the Pyrenees were identified as the region where spatial genetic change is steepest within France and the Iberian Peninsula. This clue by itself justifies a close study of Pyrenean populations with a specific purpose: the interpretation of the spatial genetic structure as the result of prehistoric demographic processes. Population genetics provides a basis for the interpretation, in terms of population history, of the observed differences between the populations. Evidence from other disciplines, such as paleodemography, paleoanthropology, archeology, and historical linguistics, is also needed to build plausible hypotheses from the genetic landscapes that can be described using sophisticated statistical tools (Cavalli-Sforza et al. 1993, 1994; Barbujani 1991). A few of the numerous unsolved issues in the population dynamics of the Pyrenees are discussed here from a genetic point of view, for instance, the extent of the Basque or related peoples along the range and the imprint left by the Indo-European-speaking peoples that crossed the Pyrenees.

MATERIALS AND METHODS

After a thorough search of the literature, we compiled a database containing all the available information on the populations in the Pyrenees. The genetic polymorphisms included blood groups, red cell enzymes, serum proteins, immunoglobulins, and HLA antigens. Samples come from a large number of villages and valleys, with an uneven distribution of genes studied. First, data were pooled in eight groups, according to linguistic and geographic considerations but also taking into account the availability of genetic information. All the populations consist of autochthonous samples geographically defined as follows (Figure 1): (Figure 1 omitted)