Spain
Spain has only recently started have faunal remains reported on regularly for Roman sites, and so the sample is not large (Fig. 5). It is possible to detect a trend of increasing pig percentages in the early imperial period, rising to 40% or more, before declining again in the late Roman period. However, it is important to note that we do not see the classic 'Roman' pattern being established, since cattle percentages are generally higher, for instance, and there are other smaller variations in the dietary pattern. It seems best to suggest that a local pattern persists that becomes modified as a result of the introduction of 'Roman' ideas of diet, but which is never entirely supplanted.
The rich agricultural region of Baetica has yielded more osteological data than other regions of Spain, so that it can be used as a case-study of the changes that occurred during the Roman period. Most of the pre-Roman sites date to the period of Phoenician and Tartessian influence, 8th-6th century BC, e.g. Huelva (Belen et al. 1977; Cereijo & Paton 1989; 1990), Setefilla (Estévez 1983), Castulo (Molero 1985), Castillo de Doña Blanca (Morales et al. 1994), El Carambolo (Martin Roldán 1959) and Toscanos (Uerpmann & Uerpmann 1973). The data from these sites have recently been summarized and tabulated numerically by Morales et al. (Roselló and Morales 1994, Tab. 10.3) and will not be repeated here. For most of them, there is a pattern with relatively low pig percentages, typically 20% or less, with a couple of exceptions; and higher sheep/goat percentages than those of ox. Huelva, for instance, yields figures for one of the assemblages of 36.9% ox, 44.6% sheep/goat and 18.5% pig (n = 502; Belen et al. 1977). This would imply a meat dietary pattern dominated by beef, followed by mutton/goat-meat and pork. In agricultural terms, a mixed pattern is implied, with the presence of pasture for the cattle, probably woodland for the pigs, and probably a generalized grain and animal-rearing economy. There is, however, a good degree of variability in the basic pattern, which suggests that local topography had significant influence on the economies of some sites. A good example of this is Castillo de Doña Blanca, adjacent to the coast near Cadiz. Ox and pig are both very low here, with sheep/goat percentages consequently being c. 80% of the total. Conditions were presumably not favourable for the former, and in fact, it is clear from the large quantities of fish and marine mollusc remains that much of the meat diet was derived from the sea.
It is unfortunate that the succeeding Iberic and Turdetanian period has yielded very few bone assemblages. Those earlier sites with phases dated to the 5th cent. BC and later (e.g. Setefilla, Castillo de Doña Blanca) appear to continue their own established patterns, and it may well be the case that despite the overall diminution in settlement activity, agricultural patterns showed a continuity with previous practices. One site, Cerro Macareno (Sevilla), is interesting in showing generally increasing percentages of pig through time: Phoenician phase 13.5%, early Iberic 11.9%, Iberic 22.4%, Iberic/Roman 26.1% . There are also ox percentages at c. 46-63% and sheep/goat percentages gradually declining from 38.9% to 20.7% (Amberger 1985). This hints at changes to the agricultural economy, with less reliance on sheep/goat and more on pig, while cattle were dominant throughout. This change may reflect intensification of production, similar to that seen in Peñaflor's bone assemblage at a slightly later date (see below).
The early phases at Peñaflor are of this period (King forthcoming a), but are too insignificant in osteological terms to allow us to establish any dietary or agricultural patterns. The final pre-forum phase hints at a pattern with relatively few ox bones and roughly equivalent numbers of sheep/goat and pig, which may be a reflection of local conditions before Roman cultural influence became very strong at the site. Since two of the thirteen pig bones are probably of wild boar, domestic pig is marginally less well-represented than sheep/goat, but the numbers overall are too low to be able to comment meaningfully on this detail. The period 5th-1st century BC is obviously important for the major cultural transitions that took place, and it would be most welcome to be able to fill in the archaeozoological picture for this from future excavations in the region.
With the construction of the forum at Peñaflor in the early 1st century AD, the influence of Roman tastes on the diet can be seen. The percentage of sheep/goat is lowered, in favour of ox and pig in fairly equal proportions (resulting in an absolute predominance of beef in the meat diet). This is possibly a weak reflection of the sort of diet seen in western central Italy at this time (see Italy), but differs in having a high percentage of ox, which would be unusual in the Italian diet. The decline in sheep/goat may be significant in agricultural terms, in that it could correspond with an intensification of cropping, perhaps cash-cropping of vines and olives, that pushed sheep grazing onto more distant territory, with a lesser number, predominantly goats , kept nearby for meat and milk.
Intensification was a process at Peñaflor that did not apparently go as far as reducing the amount of cattle pasture as well, which would have been the case with further intensification of arid-zone agriculture, since cattle compete with humans for water resources and occupy valuable land that could be used for crops. It is interesting to note that 2nd-century levels at nearly Munigua show very similar proportions (Boessneck & Driesch 1980), including a fairly high proport ion of goats, and may reflect a similar agricultural regime.
By the late 2nd century AD, the processes of Romanisation of diet and intensification of agriculture had probably reached their apogee. Pig bones account for nearly half of the main domesticate assemblage, while ox numbers are down. Such a pattern would not look out of place in an Italian context, and it may be the case that pigs were being raised on a commercial scale, as has been recorded in Italy at this time, at Settefinestre (King 1985). This would be a form of cash-cropping that would fit with intensive cultivation of olives, etc. A form of pig husbandry could have existed that utilized the shady conditions afforded by the cultivated trees as well as actual woodland, similar perhaps to the system used until recently in Extremadura (Parsons 1962).
In the late Roman period at Peñaflor, sheep/goat at 51.6% forms the most common domesticate, with pig and ox lower but roughly equal to each other in numerical terms. Sheep are now more common than g oats. This is seen elsewhere in Baetica at Munigua, at other urban sites in Spain (Tarragona, Zaragoza, Tiermes: Appendix, Table B) and indeed in many other parts of the western Mediterranean as well. There seems to have been a widespread change in the late Roman period, probably towards a less intensive form of agriculture. Cash-cropping may have declined, to be replaced by ranching, pasturing, and probably a return to cultivation of grain crops that could be rotated in sympathy with animal hu sbandry regimes.
Not all Spanish sites reflect this change, however. Two villas in the north of the country, Vilauba in Catalonia (King 1988b) and Arellano in Navarra (Mariezkurrena & Altuna 1994) display different patterns - the former having relatively high pig and cattle percentages throughout, and the latter high sheep/goat and ox. Both are presumably production sites in areas where the topography must have influenced local possibilities in animal rearing, and thus have affected the resulting bone asse mblages. In general, Catalonian sites are sheep/goat dominated (55-63%) in the pre-Roman period (Estévez 1987, 211), which is similar to the pattern in Provence (see below). During Roman times, pig is higher (54-58%) than sheep/goat (23-32%) at Ampurias in Catalonia (Estévez 1987, 212), reflecting a possibly anomalous 'Roman' pattern at this early colony, not unlike that at Fréjus in Provence. It is unfortunately not clear in our present state of knowledge what the usual Roman-period pattern was in Catalonia . Was it a continuation of the pre-Roman pattern, a change to a more 'Romanized' diet or the establishment of a distinctive local pattern along the lines of the Vilauba percentages? Undoubtedly many regional patterns remain to be established as further Hispano-Roman sites are examined from an archaeozoological point-of-view.