Quick! How many cotton varieties can you name that were among the most frequently planted in 1995? Here's a clue — most of them only had two or three numbers, and they didn't have the letters B or R in their names.

Cotton growers have seen a lot of changes in varieties in the last 10 years, but the changes have not always been for the better when it comes to fiber quality, according to a plant breeder with the University of Arkansas.

Since 1995, Arkansas producers have switched from 100 percent conventional varieties to less than 1 percent conventional or nontransgenic varieties in 2005, said Fred Bourland, director of the Northeast Research and Extension Center in Keiser. But fiber characteristics have not shown much improvement.

“The transgenic varieties, for the most part, have gone the wrong direction on quality,” said Bourland, who participated in a panel on Georgia's cotton fiber quality at the 2005 Engineered Fiber Selection or EFS Conference in Memphis, Tenn.

An update on the Georgia situation was included in this year's conference after mill buyers said they were having problems spinning cotton from that state in a presentation at last year's EFS meeting. Similar complaints were lodged at this year's conference, but Georgia Extension specialists said growers are still finding markets for their cotton.

Transgenic varieties have produced higher lint yields — a 12-pound advantage — in the 1996-2004 University of Arkansas Cotton Variety Tests, Bourland noted. (Although test data from Arkansas was used in the presentation, the trends are similar for most U.S. cotton production regions, he says.)

“In four of the first six years, conventional entries yielded more than transgenic entries,” he said. “However, lint yields of transgenic entries relative to conventional entries have steadily increased over the past three years.

No yield drag

“This trend suggests that transgenic varieties have not imposed a drag on lint yield in these tests and that relative yields of transgenic entries are improving. An alternative interpretation is that the relative yields of conventional varieties have not improved.”

Among the transgenic variety types, the Liberty Link and BXN transgenics appear to have the most positive effect on yield, he said, although with only one year of testing, the conclusion may be premature for Liberty Link. “BXN entries consistently provided improved yields over conventional entries, but BXN cotton is no longer on the market.”

Over all the years, Bt and BR (or stacked gene) entries had higher lint yields than conventional varieties. “In these tests, any entry possessing the Bt gene (either alone or stacked with Roundup Ready) has the advantage of Bt toxin control of sub-threshold injury associated with certain insect pests,” said Bourland.

Roundup Ready varieties (without the stacked gene) have yielded less than the conventional entries. “The first Roundup Ready varieties yielded very well compared to conventional, but most of those were removed from the market due to their susceptibility to bronze wilt,” he notes.

“Since 1996, Roundup Ready entries have consistently yielded less than conventional entries until this trend was reversed in 2004. Additional testing of new Roundup Ready entries is needed to determine if this reversal reflects actual improvement in Roundup Ready entries.”

Among the fiber properties, Bourland says, the largest variations are occurring between strength, micronaire and length.

“Although the differences were not great, transgenic means for each of the fiber parameters (except elongation) were in the wrong direction with respect to improved fiber quality. Overall, fiber of conventional varieties was finer (lower micronaire), longer, more uniform and stronger than fiber of transgenic varieties.”

He said the trend of increasing micronaire in transgenic varieties appeared to be reversed in 2001 through 2003, but was present again in 2004. Other than in 1996, strength of transgenic entries was less than strength of the conventional entries in all 10 years. “This suggests that improvement of fiber quality has not been a major goal in the development of transgenic varieties.”

Bourland acknowledges that means or averages of entries in variety tests do not always reflect performance of varieties planted by producers. From 1995 through 2004, seed companies entered an average of 36 varieties in the Arkansas Cotton Variety Tests. But the three most planted varieties in each of those years occupied two-thirds of the cotton acreage in north Arkansas.

Inbreeding effects

Variety performance may also be affected by a certain degree of inbreeding, he notes.

“The top varieties in the tests represent a relatively narrow genetic base, particularly considering that 1) Deltapine 20, Deltapine 50 and Deltapine 51 were derived from the same parents; 2) Stoneville 474 is the recurrent parent for BXN47, ST4793RR and ST4892BR; and 3) Stoneville 132 was directly derived from one of two parents of the conventional line used to develop Paymaster 1218BR.”

The yield, micronaire, length and strength of the top three planted varieties in the Arkansas Cotton Variety Tests were consistent over the years 1995-2004.

“The principal varieties used by producers tended to have higher yields, but coarser, shorter and weaker fibers than the average variety in the tests,” said Bourland. “These trends suggest that producers tend to choose varieties on yielding ability rather than fiber quality.

“Although these data represent only the north Arkansas region, the dominance of a few varieties, the narrow genetic base of the few varieties and the priority on yield over fiber quality when choosing a variety appear to be common trends in most U.S. cotton growing regions.”

Trying to breed new varieties for higher yields, strength, length, length uniformity, premium micronaire and other quality factors is not an easy task, Bourland concedes.

“The late Dr.0 Bob Bridge liked to compare cotton breeding to trying to get a bunch of monkeys up a tree,” he said. “About the time you think they are there, one of them always falls out.”

In 2003, nine of the 33 entries in the Arkansas Cotton Variety Test expressed average or better micronaire, length and strength values than the test mean. Those included ST5599BR, five FiberMax varieties, an Arkansas breeding line and two Deltapine varieties. Only four of those produced yields that exceeded the test mean.

“As indicated by a few notable examples, high yields and excellent fiber quality can be combined in the same variety,” says Bourland. “The scarcity of notable examples is indicative of how difficult it is to find this combination. With changes that have been and are occurring in marketing of cotton, it is essential that breeders successfully improve both lint yield and fiber quality.”

e-mail: flaws@primediabusiness.com