INTRODUCTION Arable and pastoral farming was central to the economy of Roman Britain. Throughout the Roman world, the production of agricultural capital, alongside the availability and exploitation of land and labour, were key factors in the logistics of farming practices, while technological innovations were crucial for instigating changes in the scale of crop and livestock production (Erdkamp 2005, 12–54). However, in comparison with other Roman provinces, particularly around the Mediterranean (e.g. Bowman 2013), written evidence that provides details of land-ownership, agricultural strategies, and tax systems is almost completely absent from Britain (Fulford 2004). In contrast, British archaeology, particularly since the establishment of developer-funded archaeology, has produced a wealth of environmental evidence to draw upon. The Rural Settlement of Roman Britain project has assembled one of the largest zooarchaeological and archaeobotanical datasets available for the Roman period, drawing upon the project database. In the previous two chapters, these data were examined in isolation to reveal broad regional and chronological patterns of arable and pastoral farming. In this chapter, the two datasets are combined in order to consider technological innovation, diversification and scale in farming practice within five case study areas in England and Wales, before a detailed comparison of the various agricultural strategies is explored. The possible motivations for these patterns are then considered in light of social, economic and environmental factors, particularly with regards to the demands of the urban and military populations that became established after A.D. 43. An important consideration for understanding the scale and functioning of the agrarian economy of Roman Britain is the evidence for the movement of animal and plant resources. It is clear, from the presence of specialist processing structures such as corndryers, mills and granaries, that agricultural surpluses were being produced in quantities that were beyond the basic needs of the rural population (Halstead 1989; Van der Veen and O’Connor 1998). Expanding rural and urban populations, alongside the establishment and maintenance of the Roman army, placed substantial demands on food supplies and probably resulted in the development of more complex economic systems, perhaps incorporating elements of free market exchange with the demands of the state. While a certain amount of agricultural surplus may have been traded through markets unrestricted, it is uncertain how much exchange was driven or stimulated by political concerns. Taxation would have been the principal means of the state for extracting a proportion of this surplus in order to supply the army, and support the cursus publicus and the provincial administration (Mattingly 2006, 494–6). Extensive written evidence for the supply of garrisons stationed at Vindolanda at the turn of the second century A.D. provides unique details of requisitions for a wide range of consumer products (Bowman 2003). Other potential evidence for an imperial command economy largely derives from literary records such as Diocletian’s price edict, which brought about financial reforms at the beginning of the fourth century A.D. (Salway 1993, 243–5), though it is unclear how successful these reforms were in practice, particularly in provinces such as Britain. Nonetheless, it is abundantly evident that coinage circulated widely in the agricultural heartlands of Britain from the second century A.D. onwards, probably reflecting the growing use of coins for exchange and tax purposes (see Ch. 6). Approaches to the study of Romano-British agriculture have advanced considerably since the development of environmental archaeology in the 1970s (Fulford and Brindle 2016). One of the first attempts at conceptualising an agrarian history of Roman Britain was attempted by Applebaum (1958; 1972), whose interpretation of the available evidence, made with little use of environmental remains, was largely conjectural. Several efforts since have focused on how individual farming settlements might have functioned, with Bignor villa, West Sussex (Applebaum 1958, 69–70; 1975; Frere 1987, 265) and Barton Court Farm, Oxfordshire (Miles 1986; 1989), forming two stand-out examples. By the early 1990s, however, soon after the implementation of PPG16, the potential contribution of animal bones and plant remains to knowledge of the Romano-British economy began to be highlighted. Overviews devoted to zooarchaeological (King 1991) and archaeobotanical (M. Jones 1991) evidence were 142 CHAPTER 4 AGRICULTURAL STRATEGIES IN ROMAN BRITAIN By Martyn Allen and Lisa Lodwick