Agricultural innovation to protect the environment.

In a world of 9.5 billion people, global demand for food, fiber, and biofuels has to be met with minimal possible increases in land, water, fossil fuels, and the minerals used to produce fertilizers (1–4). The problem is debated at three levels: first, that agriculture will not be able to produce enough because it will come up against both biophysical and environmental limits that restrict yields (3, 5, 6); second, that the need to expand and intensify agriculture will destroy the broader environmental values of forests, wetlands, marine systems, and their associated biodiversity (7–9); and third, that there are institutional obstacles to the diffusion and adoption of the innovations that could solve these problems. Although there is debate on these issues, there is also strong consensus that we are witnessing unprecedented changes in our major agricultural systems (6). Major shifts are occurring in the way food and other agricultural commodities are produced, in the scale at which this happens, in the geographical locations of agriculture, and perhaps most notable, the agencies and actors driving these processes (10–14). Growth in demand for agricultural products will mainly occur in markets of emerging economies, particularly in the most populous countries of Asia and Sub-Saharan Africa. Therefore, the ways in which China, India, Indonesia, Bangladesh, Nigeria, Ethiopia, and South Africa respond to growing food demand will be major determinants of environmental change at a global scale (3, 6, 11). The papers in this special feature of PNAS highlight innovations in agriculture that could contribute to producing more food without increasing environmental pressures. The papers are based on some of the more exciting ideas that emerged from a forum in Beijing in October 2011 that brought together agricultural and environmental scientists from China with their peers from the rest of the world (12, 13). The papers collectively consider how agricultural science is responding to environmental challenges. Agricultural land is now required to deliver multiple environmental and production services (9, 14, 15). The issues are often beset by “wicked problems” (16, 17) where different communities of scientists and practitioners are unable to agree on the framing of questions and therefore advocate divergent solutions (18, 19). The papers explore implications of different combinations of technologies, institutional arrangements, and policies on the agriculture– environment nexus (20, 21) and attempt to link the global resource management discourse with the realities faced by poor farmers in developing countries (3). They endorse four strategic objectives: ensuring production of adequate food, alleviating poverty, achieving better health and nutrition for a growing population, and conserving the natural resource base upon which all of this depends (22–24). Agricultural innovation is essential to address environmental problems in a world that must soon support more than 9 billion humans. Poverty and food insecurity go hand in hand (1). For the 2 billion malnourished poor in developing countries, shortterm food security is inevitably a higher priority than long-term environmental sustainability. A large proportion of rural poor in the tropics live in regions with marginal land and climate for agriculture (25) or in areas with more favorable climate that lie at the interface between agriculture and remaining carbon-rich and biodiverse natural ecosystems such as rainforests, wetlands, grasslands, and savannas (26). Feeding 9 billion people and lifting rural poor out of poverty is a prerequisite for maintaining the planet’s environment. Many people are leaving rural areas and seeking employment in manufacturing and services in cities. However, this opportunity is not open to all. Large numbers of poor farmers continue to practice extensive agriculture. Inevitably they will continue to encroach on hitherto uncultivated lands unless they can adopt innovative systems that allow for agricultural intensification and development of agricultural equipment industries, farm inputs, and food processing capacities. To this end, much agricultural research continues to focus on how to increase productivity on this existing farm land. Improved efficiency in the use of land and agricultural inputs is already contributing to environmental goals. Quantifying food production capacity of currently farmed land has focused on estimating “yield gaps” (i.e., the difference between current farm yields and the potential that can be achieved with good crop and soil management). Yield gap analysis allows the identification of regions with the greatest potential for higher yields (27–29). Need for more precise and geospatially explicit yield gap estimates are the target of the Global Yield Gap Atlas (www.yieldgap.org). However, increasing productivity is necessary but not sufficient to ensure food security, reduce poverty, improve nutrition, and maintain the natural resource base for sustainable development (6). Innovations across a broader spectrum of policies and technologies are needed to confront the complex array of challenges at the agriculture–environment nexus (1, 21). Many practicing agricultural scientists are working to solve immediate problems of poor farmers. A marked shift is occurring in the way agricultural research is conducted. In particular, there has been a move from single-factor, mainly on-station research toward active engagement with farmers and farm communities to encourage experimentation and innovation. A recurring theme is the use of concepts such as Integrated Agricultural Research for Development (IAR4D) (30). This “systems science” approach (31) and a number of similar concepts share much with the underlying principles of Sustainability Science. IAR4D attempts to harness science to address complex multifunctional agricultural objectives and to engage farmers and their communities in the