We indicated last time that the Dec. 4, 2008, issue of Nature has essays on five research programs aimed at improving the world food supply, and suggested that teachers, in particular, could use these for useful discussions in biology and agriculture classes.

Because we have significant ability these days to make drastic changes in crop plants, students are encouraged to let their minds range very widely. But today let's get acquainted with the five chosen by Nature.

Each essay focuses on one particular researcher, but it is important to remember that few scientists these days work as loners. Rather, they are members of teams, often multidisciplinary teams of several or even scores of colleagues.

The Green Revolution of the 1960s produced wheat and rice, in particular, that were both high yield and with certain types of disease resistance. (The genetic improvement of corn preceded that a bit.) But the traditional methods of plant breeding are too slow and too imprecise for modern needs. We must employ methods of genetic engineering (biotechnology, transgenics).

Stem rust in wheat is due to a potent fungus. The Green Revolution produced wheats with genetic resistance to the fungus. But all living things evolve, and in Uganda in 1999 a mutation occurred in the fungus so that it can now attack the previously resistant wheats. The evolved fungal strain, Ug99, has now moved out of Africa and into southwest Asia, from where it has ready access to the Old World's crops. The search is on among wheat's wild relatives to find new genetic material for resistance, which can then be bred into our wheat strains.

But in the meantime, Peter Dodds of Australia (cross reference Nature) is taking a more fundamental approach. He and his colleagues are trying to learn every step by which the fungus invades wheat plants, reproduces and reduces seed yield. Their plan is to design proteins that can block each step of the process. They will manufacture genes that can be inserted into wheat strains to manufacture proteins that will block one or two selected critical steps on fungal reproduction. If the fungus mutates a way around those blocks, the workers will be ready with genes to block other steps in the fungal life cycle.

This is a long-term program, but if successful it should provide a major new approach to pest control. Because the manufacture of resistance proteins does require energy, it is not wise to overload plants with multiple resistance genes. It is wiser to harness maximal energy for kernel production and employ a revolving battery of resistance genes as the fungus evolves.

Another agricultural problem is that most of our crops are annual plants, for which seeds must be sowed every year. Perennial plants, which live for multiple years and yield seed every year, would be far cheaper. Further, perennials put their roots down farther into the soil, getting water and nutrients that annuals cannot reach. They control erosion better, require less use of machinery that compacts the soil, and add increased levels of organic matter to improve the soils. But can we convert annuals into perennials? Jerry Glover at the Land Institute in Kansas, and colleagues worldwide, are making the attempt. Success would totally change the face of modern agriculture.

United States agriculture, in a basic sense, revolves around the idea that to make things grow and produce better, one applies plenty of fertilizer and water. But we all recognize that water supplies are limited and under increasing stress. Can plants be made to thrive, or even give increased yields, with less water? We'll explore that next time.