Plant and Soil 252: 55–128, 2003. © 2003 FAO. Published by Kluwer Academic Publishers. Printed in the Netherlands. 55 Beans (Phaseolus spp.) – model food legumes W. J. Broughton 1,6,∗∗ , G. Hern´ andez 2 , M. Blair 3 , S. Beebe 3 , P. Gepts 4 & J. Vanderleyden 5 1 LBMPS, Universit´ e de Gen` eve, 1 ch. de l’Imp´ eratrice, 1292 Chamb´ esy, Gen` eve, Switzerland. 2 CIFN, UNAM, Cuernavaca, Mexico. 3 CIAT, Cali, Colombia. 4 University of California, Davis, USA. 5 CMPG, Katholieke Universiteit, Heverlee, Belgium. 6 Corresponding author ∗ Received 31 January 2002. Accepted in revised form 20 August 2002 Key words: expression analysis, expressed sequence tags, large-scale sequencing, molecular breeding, Phaseomics consortium, P. vulgaris, Rhizobium Abstract Globally, 800 million people are malnourished. Heavily subsidised farmers in rich countries produce sufficient surplusfood to feed the hungry, but not at a price the poor can afford. Even donating the rich world’s surplus to the poor would not solve the problem. Most poor people earn their living from agriculture, so a deluge of free food would destroy their livelihoods. Thus, the only answer to world hunger is to safeguard and improve the productivity of farmers in poor countries. Diets of subsistence level farmers in Africa and Latin America often contain sufficient carbohydrates (through cassava, corn/maize, rice, wheat, etc.), but are poor in proteins. Dietary proteins can take the form of scarce animal products (eggs, milk, meat, etc.), but are usually derived from legumes (plants of the bean and pea family). Legumes are vital in agriculture as they form associations with bacteria that ‘fix-nitrogen’ from the air. Effectively this amounts to internal fertilisation and is the main reason that legumes are richer in proteins than all other plants. Thousands of legume species exist but more common beans (Phaseolus vulgaris L.) are eaten than any other. In some countries such as Mexico and Brazil, beans are the primary source of protein in human diets. As half the grain legumes consumed worldwide are common beans, they represent the species of choice for the study of grain legume nutrition. Unfortunately, the yields of common beans are low even by the standards of legumes, and the quality of their seed proteins is sub-optimal. Most probably this results from millennia of selection for stable rather than high yield, and as such, is a problem that can be redressed by modern genetic techniques. We have formed an international consortium called ‘Phaseomics’ to establish the necessary framework of knowledge and materials that will result in disease-resistant, stress-tolerant, high-quality protein and high-yielding beans. Phaseomics will be instrumental in improving living conditions in deprived regions of Africa and the Americas. It will contribute to social equity and sustainable development and enhance inter- and intra-cultural understanding, knowledge and relationships. A major goal of Phaseomics is to generate new common bean varieties that are not only suitable for but also desired by the local farmer and consumer communities. Therefore, the socio-economic dimension of improved bean production and the analysis of factors influencing the acceptance of novel varieties will be an integral part of the proposed research (see Figure 1). Here, we give an overview of the economic and nutritional importance of common beans as a food crop. Priorities and targets of current breeding programmes are outlined, along with ongoing efforts in genomics. Recommendations for an international coordinated effort to join knowledge, facilities and expertise in a variety of scientific undertakings that will contribute to the overall goal of better beans are given. To be rapid and effective, plant breeding programmes (i.e., those that involve crossing two different ‘parents’) rely heavily on molecular ‘markers’. These genetic landmarks are used to position ∗ FAX No: +41-2-2906-1741. E-mail: william.broughton@bioveg.unige.ch ∗∗ With contributions from all members of the Phaseomics Consortium (see Appendix for details).