Agriculture

The Roman conquest of empire carried with it huge economic implications, not least in terms of the disposal of the acquired territory (Fulford 1992; D. J. Mattingly 1997: 18-19, 117-35). Land was both a key motor of military conquest and underpinned economic activity throughout the empire. In Egypt, for example, taxes on land and the products of farming are estimated to have contributed over 60 percent of state revenues (Duncan-Jones 1994: 53). The reallocation of land to subject peoples was a key stage in the transfer from military control to civil rule and was often accompanied by land survey. Where land was taken for the creation of colonies, these lands were normally surveyed and partitioned in detail on an imposed grid system (centuriation). The feats of the Roman land surveyors ( agrimensores) can be appreciated both from extant writings (Campbell 2000) and from the physical evidence of such systems in the landscape (Dilke 1971).

For the Mediterranean region, the basics of Roman agriculture were the ‘‘triad’’ of cereals (Spurr 1986), grape vine (Fleming 2001; Purcell 1985; Tchernia and Brun 1999), and olive (D. J. Mattingly 1996a). In most areas agriculture outweighed pastoralism, though even in Italy there were regions that were known for stock raising, such as the Apennine valleys (Whittaker 1988). The relatively arid Mediterranean climate, mainly simple technology, and variable soils imposed constraints on the productivity of Roman farming, but the overall extent of cultivated land during the Roman period was probably not again surpassed until recent centuries.

There was a well-developed tradition of writing agricultural manuals, though in truth these were more works on estate management than on the practical details of farming methods. The most influential surviving works were by the Roman aristocrats M. Porcius Cato and M. Terentius Varro and the Spaniard Columella (K. D. White 1977). These sources dealt unequally with the various components of farming: viticulture is generally given first place in terms of length of exposition (reflecting an age-old aristocratic interest in the production of wine), then cereals, then olives. In Varro’s account stock raising features prominently, as does market gardening, and in general terms the ‘‘ideal’’ Roman agricultural estate seems to have been conceived as a self-sufficient mixed-farming unit. All the writers also describe the management of slave labor (showing the Italian context of their experience - rural slaves were far less common in the provinces), and the construction of estate buildings (villas). The most spectacular archaeological example of one of these Italian aristocratic estates is the Settefinestre villa near Cosa (Carandini et al.1984).

From other Roman sources, we know that one result of Roman expansion was the creation of large numbers of smallholders, in part through land allocations to time-expired soldiers. However, they do not feature in the works of the agronomists and archaeological evidence suggests that in many areas there was a tendency towards the consolidation of larger estates (described as latifundia in Italy) out of the early smaller distributions. Many farmers ended up serving as tenants rather than as freeholders and, particularly outside Italy, the importance of this sort of dependent labor in supporting large estates cannot be overemphasized (Garnsey 1980; Kehoe, this volume).

The reality of Roman farming in the provinces was in many respects rather different from the picture that we derive from the literary sources of its practice in Italy. Although we can detect the formation oflarge estates in many regions, coupled with the emergence of a ‘ ‘villa economy,’’ there are several clear instances of crop specialization. Papyrological evidence from the Heroninos archive, relating to an estate in the Egyptian Fayum, reveals a complex infrastructure linking several production units. Although producing a variety of crops, the main cash crop was wine, which was produced and traded on quite a large scale (Rathbone 1991). Similar regional specializations in olive oil production can be detected in southern Spain and several regions in North Africa (D. J. Mattingly 1988a), and in wine in southern Gaul (Tchernia and Brun 1999). These economic success stories are also matched by regions such as Greece, where the early Roman period appears to have been one of contraction and underdevelopment, in contrast to renewed development in late Roman times (Alcock 1993).

Outside the Mediterranean zone in temperate Europe, the existence of heavier soils was offset by more reliable and abundant rainfall. Cereal cultivation became well developed, and viticulture gradually extended far to the north of its previous limits, even reaching Britain and Germany. The olive, on the other hand, remained restricted to the Mediterranean climatic zone, due to its vulnerability to cold. Stock raising for meat and secondary products was in general a more important element in northwestern Europe, exploiting the abundant pasture lands (A. King 2001). Beer, produced from barley malt, and the use of animal fats in cooking and for lighting remained key cultural markers of the north, despite the partial inroads by winemakers and the imports of olive oil.

In North Africa, two regions were crucial for the production of cereals needed to feed the city of Rome: the Nile delta in Egypt and northern Tunisia. The existence may be noted in both regions of large imperial estates and a complex infrastructure for gathering the annona grain from other producers (D. Crawford 1976; Kehoe 1988a; Rickman 1980). Less suitable cereal lands in Africa were strongly developed for other crops, notably the olive, making North Africa a far more dynamic region economically than its modern agriculture would suggest (D. J. Mattingly 1988a).

There is a large literature on the management and organization of Roman rural production (see Kehoe, this volume). The Roman provincial economies may have been built on the labor of lots of peasants, but they were dominated by the output of the bigger players, and major estates were a feature of most provinces of the Roman world. The key conclusion to be drawn from the archaeological and documentary evidence relating to these major estates is that they could be extremely large in scale, and rationally organized so that profitability and costs could be assessed adequately and tied into wider commercial networks for disposing of their often considerable surpluses. The Appianus estate in the Fayum, for example, comprised several units of production, based on villages each under the financial control of a phrontistes, with a very sophisticated system of monthly accounts that could take account of internal paper transfers due for the loan of animals and labor between different units (Rath-bone 1991). Archaeological evidence supports the view that such estates could have generated huge surpluses - as demonstrable in the case of wine production at sites in Italy and southern France, which possessed three-four presses and wine fermenting cellars capable of producing and storing several 100,000 liters (Amouretti and Brun 1993; Carandini et al. 1984). In Libya, the largest olive oil factory (oilery) yet discovered contained a massive 17 olive presses, capable of processing in a peak year well over 100,000 liters of olive oil (D. J. Mattingly 1988b).