Consider Tribulus terrestrus, sun-lover, sidewalk-dweller and public nuisance.

An indomitable weed possessed of a sharp thorn, it grows up in the pavement cracks of the desert Southwest, basking in the summer heat and drought and serving no apparent purpose outside of pricking the unwary.

But to the scientists in one of biotechnology's boldest and most promising fields, once-ignored plants like Tribulus have suddenly become an object of great interest.

What if the same qualities that make sidewalk weeds or, say, cactus so durable could somehow be transferred to more useful crops? Would putting a few select Tribulus genes inside a corn plant give a farmer a better chance of surviving a drought?

Drought tolerance is by no means a new goal for plant breeders.

But with the new tools of bioengineering, which allow scientists to pop genes in and out of different plants like tape cassettes, the field has exploded, attracting new researchers and presenting an extraordinary range of possibilities.

Searches are underway for genes to make peanut roots grow deeper to find moist soil, genes that will grow long hairs on leaves to shade them against summer sun, genes like those of cactus that shift a plant's heat-sensitive metabolic processes to the cool of the evening, or like those of beans that go into hibernation during long dry spells.

At stake is a potentially enormous market among farmers around the globe, for whom the specter of drought has taken on a special urgency. Parts of the United States are enduring years-long droughts and, if global warming continues, new droughts may come to now-wet regions.

"Drought tolerance is the holy grail," said Peter Carlson, chief scientist at Crop Genetics International in Hanover, Md.

But the task isn't easy. To begin with, no one fully understands the ways a lack of water kills plants. That makes it difficult to know how to improve their drought tolerance and even more difficult to know what genes to look for in the plant species that thrive without much water.

"We have so little understanding of the basic physiological processes involved that genetic engineering is totally a shot in the dark," said Ted Hsiao, a researcher at the University of California at Davis.

Nor has anyone been able to figure out a way to make a plant drought resistant without seriously affecting its productivity during normal weather.

For example, Agriculture Department scientists recently found a gene, known as TR, that could make pearl millet, a forage grain grown in many parts of the United States, 25 percent more productive during drought conditions. The gene gave the millet leaves a thick, waxy skin that helped preserve moisture.

But those waxy leaves make the millet less digestible. Cattle don't like it as much. Threshing good seeds from it is not as easy.

"TR was interesting in that it was making changes that were significant," said Glenn Burton, a researcher at the USDA's Coastal Plain Experimental Station in Tifton, Georgia. "But when you finally added everything up, the disadvantages offset any economic advantages from surviving drought."

This kind of experience is typical. In most cases, changes simply cannot be made to most crop varieties, which are finely tuned by years of breeding and selection to produce under optimal conditions, without affecting yield.

For example, one logical way of protecting a plant from drought is to narrow the tiny breathing holes on each leaf -- called stomata (singular: stoma) -- through which plants lose much moisture. But during good weather a plant needs wide-open stomata to take in the maximum amount of carbon dioxide from the air, which it uses to carry out photosynthesis. Preparing for dry weather, in other words, means paying a price during wet weather.

Or take the idea of giving a plant a deeper root system so that it can find water from deep soil during times of drought. This is how many plants survive in the desert. But the effort put into an extensive root system leaves the plant with less energy to grow above ground.

There are also practical implications. A corn plant's roots normally cluster tightly around the base of a stem. If they ran out to the side in all directions, they could capture more water by expanding the territory tapped for moisture. But to make that work, a farmer would have to plant his corn farther apart, cutting yields.

This doesn't mean that finding an economical way to introduce drought resistance is impossible. But the complexity of the task has forced scientists to look for more sophisticated means.

One possibility is to introduce new genetic traits to help out during a drought, but to fine-tune them so that they have a small effect on plant performance in normal times. Stomata, for example, could be narrowed, but another gene could be added that would make sure they narrowed only in extended dry periods.

Another possibility is to accept the yield loss but to introduce new genes that minimize its effect. Suppose, for example, drought tolerance abilities took up so much energy in a corn plant that it grew six inches shorter than a normal stalk. Another gene might be found to take that energy from the stem and not from the ear.

"It is unlikely that we are going to make plants grow as large and weigh as much under dry conditions as they do in normal weather," said John Boyer, a biochemist at the University of Delaware. "But by manipulating the partitioning of that weight in various parts of the plant, we can encourage the production of the harvestable parts."

Boyer has shown that corn ears don't need much water to survive, although the rest of the plant does. In the laboratory he has deprived corn plants of so much water that the stems and leaves virtually dried up. But by injecting a special mix of concentrated nutrients, with relatively little water, he has kept the ears alive and growing.

Boyer thinks this means corn might be endowed with a backup metabolic system that would kick in during times of drought. If some way can be found to enhance the ability of a corn plant to produce the nutrients in his special mix, then in a severe drought the plant could shut down most of its anatomy and only muster enough nutrient to keep the ear alive. In good weather, theoretically, that system should have no downside: It should simply mean that the ears get even bigger.

But Boyer is a long way from finding how to get plants to produce those key nutrients. And in most cases researchers say that even when those mechanisms are identified, transferring them to plants is going to be a more complicated genetic engineering process than has ever been attempted.

"We're at the Tinkertoy stage," said Randy Ryan, a researcher at the Unversity of Arizona. "We're building very simple models."