Global inbreeding

The values of total (F _t), random (F _r) and non-random (F _n) inbreeding at the provincial level are shown in Fig. 2. Both ethno-social groups (Spanish and Mixed) exhibited low values of inbreeding and scarce preference by isonymic spouse, in spite of the selective behaviour traditionally attributed to the Spanish.

Fig. 2.

Province level: estimations of total (F _t), random (F _r) and non-random (F _n) inbreeding for the Spanish and Mixed groups.

Differences between the two groups were pronounced for F _n (the value for Spanish was almost four times greater than for Mixed) whereas F _r was only somewhat higher in the Spanish group. The relative loading of each component showed that while in the Spanish group the total inbreeding was primarily due to isonymic marriages, in the Mixed group the random component prevailed.

Inbreeding by parish and micro-differentiation rate

The estimations by parish of F _r, F _n and F _t for the Spanish and Mixed groups can be seen in Table 2 and Table 3, respectively.

Table 2.

Components of inbreeding by parish for the Spanish group

Table 3.

Components of inbreeding by parish for the Mixed group

In the Spanish group, the low values of F _r, and F _n obtained for Córdoba city suggest that it was a population without a preference for isonymic spouses. This city, during Colonial and early Independent times, has been described as the main receiving centre for intra-provincial migrations, as well as those of neighbouring provinces and Spain, in this latter case consisting mainly of men (Celton, Reference Celton1993; Tell, Reference Tell2008). Nevertheless, the attraction of the city for the feminine labour force of free condition (i.e. of Mixed women), coming from neighbouring regions, has also been mentioned (Celton, Reference Celton1994). Therefore, the addition of new surnames to the city may have played an important role, as reflected in the low F estimators.

In the Spanish group, the parish that showed the greatest preference for isonymic spouses was Tulumba, in the north-west, with a high value of F _n indicating a clear preference for spouses having the same surname. The already commented on social organization in this region, which had quite a solid social network, might have included as a preferential strategy marriage between relatives. Relatively high values of F _r and F _n were found for Tercero Arriba, a zone of more recent settlement, peopled by families from Córdoba city expanding their ranches to the south (Ghirardi et al., Reference Ghirardi, Colantonio and Celton2007). This may have resulted in a population with little variability of surnames, as expressed by the value of F _r (which was the highest and at least double the figures obtained for its neighbours). This familial peopling, linked with a high percentage of compound households and a dependent population (Ghirardi et al., Reference Ghirardi, Colantonio and Celton2007), seems to have been related to a dominant class in a good economic position that wished to maintain its status through marriage among relatives, among other strategies. The rest of the parishes, however, did not evidence inbreeding due to mating preference by surname, and even exhibited some degree of rejection of isonymic marriage.

In the Mixed group, none of the parishes revealed any significant inbreeding due to isonymy. The highest F _r, found for Tulumba, seemed to be related to its characteristic as an area from which people emigrated.

The parish that displayed the most similar behaviour between Spanish and Mixed groups was Córdoba city. The F _t values obtained from the census of 1832 for Córdoba city (Küffer & Colantonio, Reference Küffer and Colantonio2005) were 0.0054 for nobles and 0.0038 for pardos. These figures are of a low magnitude, and similar to those of the present study. This was so, in spite of the fact that 1795 was in the late Colonial period, whereas 1832 corresponds to the Independent period when the recently risen republic was totally free from Spain's monarchy. The results from other parishes showed fluctuating values. Nevertheless, both the Spanish and Mixed groups coincided in having no preference for isonymic marriage, except for the Spanish in the case of Tulumba.

The fact that isonymy tends to overestimate consanguinity, due to the polyphyletic origin of some surnames, gives more credibility that the values found in this work are revealing the scarce inbreeding in the population under study. A comparison with other studies can be seen in Table 4, where it can be observed that the estimations found here are in the range of those obtained for historical communities with similar characteristics. The populations of Córdoba province were more similar to those communities that could be described as ‘open’, that is to say, which displayed dynamism in the entrance and exit of people.

Table 4.

Estimations of consanguinity (F _r, F _n and F _t) found in this study and other studies of historical populations

The values separated by a bar denote the minimum and maximum; those in bold are communities of mountainous zones.

The micro-differentiation rate (F _st) among parishes was also relatively low, both in the Spanish (0.0034) and Mixed (0.0026) groups, compared with the values for historical populations of the province. In fact, the F _st obtained for smaller settlements, located near to each other in rural populations of Córdoba in the first half of the 19th century (Colantonio & López, Reference Colantonio and López1997), was 0.0042. These results indicate a more important genetic flow occurring among the major geographic units considered in the present study (parish) than among these local populations in the rural curatos of Córdoba.

The values of F _st found in this work are similar to those obtained for other historical populations of the same period in Spain, the United States and Finland (see comparative table in Peña et al., Reference Peña, Morales and Calderón1997). The lower value of F _st in the Mixed group suggests a greater geographic mobility of this group. It is very plausible that the displacements in search for new labour opportunities in the Mixed group had predominated over the Spanish group, mainly in the case of individuals migrating alone. This can be explained by considering that an easier access to property for the Spanish implied that they had less need to migrate than the Mixed group.

Repeated pairs

The estimates of RP, RP_r and (RP−RP_r)/RP_r by parish for the Spanish group are shown in Table 5. For eight of the ten parishes, RP was higher than RP_r, the exceptions being Córdoba city and Ischilín. That is to say, except for these two cases, the parishes exhibited a tendency towards subdivision within the population, as is shown by the (RP−RP_r)/RP_r indicator. According to the z-test, this tendency was statistically significant in six out of the eight parishes where such a trend was apparent.

Table 5.

Observed (RP) and expected (RP_r) repeated pairs, and(RP–RP_r)/RP_r, z-test and p-values by parish for the Spanish group

* p<0.05.

The estimations for the Mixed group appear in Table 6. For this group, repeated pairs of surnames were only found in two of the seven parishes studied (Tulumba and Río Segundo), and only one – Río Segundo – happened to be statistically significant. Therefore, in contrast to the Spanish group, preferences among family lines were not a prevailing feature.

Table 6.

Observed (RP) and expected (RP_r) repeated pairs, and (RP−RP_r)/RP_r, z-test and p-values by parish for the Mixed group

* p<0.05.

The major cause of these dissimilar trends could be related to the existing economic and social inequalities between the Spanish and Mixed groups. The former had a better standard of living, the majority of the productive land and political power being in their hands. Thus, it may be supposed that a common way to consolidate these privileges was the exchange of spouses between lineages, especially among the richest families.

The parishes that showed the (RP–RP_r)/RP_r minimum and maximum values, thus repeating the observed pattern in inbreeding for both groups, were Córdoba city and Tulumba, respectively. Córdoba city did not present a population subdivision, whereas in Tulumba the values of (RP–RP_r)/RP_r were the highest for both ethno-social groups, thereby showing the greatest tendency towards population subdivision. Thus, despite the discrepancies in spouse isonymy observed for the Spanish (preference) and Mixed (non-apparent preference) groups, they displayed a similar behaviour with respect to marriages among certain lineages in this parish. The findings for the city show a non-preferential behaviour of mate selection, at least regarding surnames, revealing a well-stratified society with a strong determination of the dominant class to preserve its privileges, at least if the whole population of the city rather than just some prominent families are considered. In contrast, results obtained for Tulumba suggest a clear selectivity, which could have been convenient, among other reasons, for the preservation of the familial patrimony in the Spanish group. In the Mixed group, however, this explanation is less appropriate because they had few opportunities to become land-owners. Maybe a combination of two factors took place: preserving and increasing the modest and hard-to-acquire properties, and following their employers' mating behaviour (marriages between both spouses' employees). Nevertheless, a third factor might have been even more important: the decrease in potential partners due to migration from Tulumba to other areas, either to Córdoba or other provinces.

An estimate, by taking the marriages for both the Spanish and Mixed groups together, could show whether the subdivision unveiled by surnames reflected the ethno-social stratification. If this were the case, then the values of (RP–RP_r)/RP_r obtained for the joined groups would be higher than those considering each one separately. Thus, it would be expected that both groups were sufficiently differentiated so as to be able to characterize them as two different (though not totally isolated) reproductive units. However, the figures of (RP–RP_r)/RP_r shown in Fig. 3, when compared with those of Tables 5 and 6, suggest that except for two of the seven parishes (Anejos and Punilla), an association between surnames and ethno-social groups was not evident. Even though populations where surnames were strongly associated to an ethnic group or geographic origin have been found (González-Martín et al., Reference González-Martín, García-Moro, Hernández and Moral2006), this was not the case in Córdoba's population. Surnames per se did not make it possible to identify the Spanish and Mixed groups as two different reproductive units, despite the well-documented social segregation (at least in reference to legal unions) that existed at that time in all of Spain's colonies in America. This is in agreement with the very close similarity in the geographical distribution of surnames and in relationships among parishes found for historical rural populations of Córdoba province in different ethno-social groups (Colantonio et al., Reference Colantonio, Fuster, Ferreyra and Lascano2006).

Fig. 3.

Indicator (RP−RP_r)/RP_r by parish, taking both groups together.

Summing up, a comprehensive examination of the results obtained by isonymy and repeated pairs at the parish level for the Spanish group shows that, while a preference for spouses of the same surname was not confirmed, there was a tendency towards internal population subdivision. This suggests that although marriage to a relative tended to be avoided, there was an exchange between certain lineages, which may constitute a strategy for maintaining the patrimony while avoiding unions between close relatives. This would also explain why, in most of the parishes, subdivision was verified for Spanish but found in only two for the Mixed group, since among the former were the richest families, who would have had more interest in contracting advisable unions.