extension agent to farmer. New paradigms in agricultural extension programs have recognized that local people con­duct research on their own farms; it is even argued that their experiential knowledge is derived from their skills as experi­menters (Stanley and Rice, 2003).
     Although the absence of formal and rigorous method­ology employed by formal institutions such as robust sta­tistical design make analysis and interpretation of farmer experimentation difficult, such paradigms are increasingly becoming acceptable to the scientific community. An exam­ple is the agroforestry project of the International Institute of Rural Reconstruction (IIRR) in the Philippines, in which scientists worked with village farmers after a failed nursery operation that relied on exotic species. Local people were then asked to identify locally growing species (indigenous and introduced) according to criteria considered by the community as important—such as hardiness, fire resistance, general utility and seed availability. The exercise resulted in the formulation of community action plans for reforestation (IIRR, 1996).
     Achieving sustainable  agriculture in the region will require the integration of agricultural knowledge systems, practices and technologies through the provision of finan­cial and infrastructural support to facilitate research, dis­semination and constraint-specific utilization of all available technologies. Specifically, this will include but not be limited to:
•   Increasing support for and applying traditional and low-input systems that function productively and in a socially inclusive manner, particularly in low productiv­ity areas throughout the ESAP region.
•   Augmenting indigenous knowledge with appropriate modern practices that can enhance the system, such as microbial inoculations, appropriate scale mechaniza­tion and small-scale technology—such as gravity-fed technology for sprinkler and drip irrigation for vegeta­ble and fruit cultivation.
•   Integrating elements of traditional and organic practices (such as rotational, trap and intercropping and agrofor­estry) in modern agricultural settings to maximize sys­tem productivity and natural resource sustainability.
•   The application of emerging technologies such as bio and nanotechnology that have been rigorously evaluated in comparison with existing or developing technologies not only on the basis of potential gains in productivity, but on their ability to maintain ecosystem integrity, hu­man and animal health and social and economic well-being in adopting communities and countries.

It is critical that unbiased science precedes, rather than fol­lows the commercialization of new agricultural technolo­gies such as genetic engineering and nanotechnology ap­plications. Each new transgenic product has potential for unintended impact; hence each new technology should be evaluated on a case by case basis (NAS, 2000), acknowl­edging farmer needs and conditions. A stringent biosafety framework, enforcement, and rigorous site-specific scien­tific and social monitoring protocol are needed. ESAP coun­tries intending to implement new technologies will need to ensure that their infrastructure is sufficient to support the safe development, transfer and application of the tech-

nologies with special attention paid to developing relevant policies, information systems and training in biotechnology risk assessment and biosafety procedures. Gene flow and resistance management issues in particular, among others, still warrant caution and long-term monitoring in the field, along with a holistic assessment that includes analysis of the social and economic consequence of biotechnology adop­tion by farming communities.
     Methods and results of environmental risk assessments could be shared between countries that have similar agri­cultural environments, thus reducing the burden of proof for any one country (Nuffield Council on Bioethics, 2003). Adoption of such a model in the ESAP region may include data sharing to satisfy regulatory needs and similar exten­sions of prior findings, all of which could substantially reduce unwarranted restrictions and improve the benefits of these technologies for resource-poor farmers, as long as their socioeconomic and cultural environments are also considered alongside the agricultural. While there may be reason to be optimistic about the potential for a variety of bio and nanotechnologies to be beneficial in increasing ag­ricultural productivity while reducing some inputs, there is also an imperative to exert the highest scientific, regulatory and policy standards to ensure negligible long-term ecologi­cal and human impact prior to their deployment.
     Facilitating the access of rural communities to informa­tion through a range of ICTs and knowledge bases can lead to an increased understanding of the consequences of vari­ous management practices and the adoption of appropriate and sustainable agricultural and aquaculture management practices. The use of ICTs can improve the reliability of cli­mate forecasts and the prediction of extreme weather events as well as their likely effects on agricultural ecosystems and rural livelihoods—both of which are invaluable in devis­ing coping strategies. The adoption of common regional standards for sharing information will facilitate technology transfer and proliferate best management practices across the ESAP region for a relatively low R&D investment on the part of any given country.
     The ESAP region accounts  for about two-thirds of the world's rural poor, primarily concentrated in sparsely populated arid or marginal lands and forests of South Asia (CGIAR, 1999; Fan and Kang, 2004). Data on poverty dis­tribution by land type for India and China indicates that in the 1990s, over 80% and 60% of the rural poor lived in low potential or rainfed areas of India and China, respectively (Fan and Hazell, 1999; Fan et al., 2002). Despite high rates of out-migration from these areas, populations continue to grow, often with increasing poverty and degradation of nat­ural resources (Hazell et al., 2000). Government and donor investment in the past favored areas of high yield potential, leading to better infrastructure, schools, health facilities, credit programs, production pricing policies and access to agricultural technologies and services in these areas (Fan and Kang, 2004). Funding for public agricultural research and development grew at an annual rate of about 8.7% in the 1970s; this slowed to 6.2% annually in the 1980s and has likely been decreasing since, as have investments in new irrigation infrastructure (Gruhn et al., 2000).
     Improving  agricultural productivity  and halting the degradation of the region's natural resource base will re-