Bourland and plant physiologist Derrick Oosterhuis are working on yield stability - a big concern for cotton growers. In the 1990s, new cotton varieties provided higher yield potential and other benefits. But a bumper crop one year has often been followed by a terrible crop the next year.

Erratic weather, especially hotter than normal summer nights, is a factor, but genetics are most likely involved, Bourland says.

Hal Lewis, a cotton scientist and grower at Doddridge in the southwest corner of Arkansas with a doctorate in biochemistry, evaluated many years of data from his own fields and multiple locations of the National Cotton Variety Test. His findings may point to a genetic solution to the yield stability problem.

"Our evaluation of the 1999 cotton variety test agrees with Hal's findings," Bourland says.

Each cotton boll contains up to about 40 seeds. Cotton fabric is spun from fibers, or lint, that grows from the seed.

In his research, Lewis bases yield on two components - the amount of seed produced per acre and the amount of lint produced per seed.

"It takes more energy to produce seed than lint," Bourland says.

The theory is that, without realizing it, cotton breeders may have programmed plants to produce more seed, but not more lint per seed. Under favorable conditions, more seed has the desired effect. But when plants are stressed by weather or other factors, they may spend too much energy producing seed and not be able to adequately fill the bolls with lint.

Derrick Oosterhuis, in his plant physiology lab and greenhouses on the Fayetteville campus, is studying the response to stress on older varieties compared to newer varieties with regard to these yield components.

"We may be able to improve the genetics for lint per seed pretty quickly," Bourland says.

However, he adds, "Plant breeding is like trying to get a bunch of monkeys up a tree. As soon as you think you have them all up there, they start falling out." In other words, advances in one genetic trait may cause an interaction that produces undesired effects.

Nearly all cotton seed planted in Arkansas is produced by private companies, but the U of A Division of Agriculture breeding program and genetics research have a major impact on those varieties.

"We work on basic research issues like yield stability, and we develop improved breeding lines that are adapted to Arkansas conditions," Bourland says.

Genetic traits in commercial varieties that have been directly influenced by U of A breeders include early maturity, fiber quality, yield potential, disease resistance, resistance to insect damage, and the vigor of seeds and seedlings.

Howell Medders writes for the Arkansas Agricultural Experiment Station.