2005). Major US private universities that historically have trained large numbers of agricultural policy analysts have closed key academic units.  The Land-Grant universities tended to focus on state agricultural interests rather than international agricultural R&D. Also, several states have made funding foreign graduate students more difficult (Fal­con and Naylor, 2005).
     In addition to the environmental concerns and the de­velopment of the concept of multifunctional agriculture, market-based  economic   liberalization   and  globalization were dominant drivers from 1986 until the early 2000s. These market forces contributed to large-scale agricultural industrialization. The main consequences were a shift from producing commodities to manufacturing products, empha­sis on the entire food chain with increasing specialization, re-alignment and increasing power of retail and flexible sys­tem adjustment to changes in consumer demand, economic conditions and technological improvements (Van Keulen, 2008). Further, information technology was increasingly utilized to enhance the value chain's competitive ability. De­velopment of new products was aided through new tech­nologies: improved logistics brought about by integration of transport and storage systems, improved preservation sys­tems, the communication "revolution" (through electronic data exchange as well as investments on efficient consumer response), biotechnology, active packaging, precision farm­ing and an increased use of integrated pest management (Van Keulen, 2008).
     These trends in AKST approaches after the Second World War were more prominent in research, extension and training than in higher education. In higher education the general trends were similar but changes proceeded more slowly and met with more resistance.

4.4 Changes in the Integration of Perspectives within AKST
Integration of perspectives within AKST has several dimen­sions, integration among scientific disciplines and actors representing multiple interrelated goals (e.g., different di­mensions of sustainability, different policy goals), system levels (e.g., loops in the food chain, rural development), as well as spatial (local, national, global) and temporal scales (short- and long-term dimension of sustainability). Integra­tion within AKST refers also to integration among educa­tion, research and extension. Integration between AKST and KST is also of interest. Integration and disintegration may take place in terms of approaches, methods and conventions of science and innovation, as well as through development of organizational structures.

4.4.1 Evolution
In the past AKST was well integrated, if informally, with practical agriculture and beneficiaries as well as among the emerging disciplines. This changed at the time when the disciplinary basis of universities and research institutes was established. Distancing occurred both in relation to the practitioners and among emerging disciplines (i.e., vertical and horizontal disintegration). This distancing was more extensive in Europe than in the US as the higher education, agricultural research and extension systems of the latter

were established in a more integrated way. More recently AKST has moved towards re-integration.
     The integration of the early days was biased towards (1) farmers and rural populations at the cost of consumers and other interest groups and (2) soil, crop and animal sci­ences as well as farm economics at the cost of human nutri­tion, ecological and social sciences. The re-integration has mainly proceeded in the form of specific integrative research approaches without this earlier bias. The latter were often first adopted in developing countries, simultaneously with still continuing disciplinary fragmentation. Thus, in most places, integration has been a functional rather than a struc­tural, organizational phenomenon. In Europe, the strongest formal incentive to integration has been provided by recent EU Framework Programs, conceived to respond to the ma­jor socio-economic challenges facing Europe (Buhler et al., 2002).
     Up until the middle of the 19th century, training of ag­ricultural scientists did not advance rapidly. Advancement required the introduction of a new science system for agri­culture, which occurred largely between 1860 and 1920. To establish this system, research methods were borrowed from the general sciences (e.g., chemistry, botany, physics) (Huff­man, 1998ab). Even though the historical ideals of unity and synthesis of knowledge in natural sciences served as the first models for agricultural sciences, a fragmentary ten­dency dominated the infrastructure of science until the mid 20th century. This tendency was characterized by the split­ting of disciplines into new subspecialties (Klein, 1990) and by focusing on separate topics, increasingly ignoring their interrelations. Thus agricultural science structures—both in education and research—rewarded a narrow orientation as a sign of a truly scientific approach. However, science and technology  developed  bidirectionally,  facilitated  by the agricultural roots of most agricultural scientists (Huff­man, 1998ab). Additional methodologies were developed to meet the special circumstances associated with agricul­ture (Huffman, 1998a) and much applied research became multidisciplinary. While the earliest documented use of the term "interdisciplinary" in research appeared in general education and in the social sciences in the 1920s, the first problem-oriented interdisciplinary research was conducted in the 1940s in agriculture and defense (Bruun et al., 2005). In many comparative studies agriculture has turned out to be one of the most interdisciplinary science fields (Clayton, 1985; Qin et al., 1997; Song, 2003). However, these studies often used the term "interdisciplinarity" meaning multidis-ciplinarity with no requirement of interaction of sciences. Also combinations of closely related fields were much more common than interactions between natural and social sci­ences (Bruun et al., 2005).
     As described above (and in more detail in Chapter 3) the narrow focus in AKST and agriculture after the Second World War on productivity of labor and land caused nega­tive externalities which gradually become more apparent. These unintended consequences raised concern about frag­mentation and overspecialization in agricultural and food sciences (Carson, 1962; White, 1967). The recognition that ecological, economic and social dimensions needed to be taken into account simultaneously led to the introduction