Monday, June 29, 1998

WRITER: Phil Williams, 706/542-8501, [email protected]
CONTACT: Jim Hamrick, 706/542-1826, [email protected]



ATHENS, Ga. -- Tropical forests have been disappearing at alarming rates for the past three decades. Farmers, ranchers and timber industries have cut millions of acres, and only in the past few years has the ecosystem damage become clear.

A new study by botanists at the University of Georgia now shows for the first time that trees left standing in pastures can actually dominate the reproduction in nearby remnant forests, creating a "genetic bottleneck." The research indicates that the survival of tropical forests could be far more complex than was known before and that new approaches to conservation strategies may be needed.

"The key is to understand how much genetic movement there is between fragments of forest," said Dr. James Hamrick. "When we lose fragments of forest, we lose genetic diversity. Gene exchange between fragments helps to maintain this diversity."

The study, by Hamrick and his graduate student Preston Aldrich, was published today in the journal Science.

Genetic diversity is vital in both plant and animal communities. Farmers have, for hundreds of years, bred crop plants and farm animals to maintain a healthy diversity of what were, before the 20th century, called traits. Now, with advanced techniques to determine the exact genetic makeup of individuals, scientists understand considerably more about how genes drift through populations.

Aldrich and Hamrick studied a tree species called Symphonia globulifera in a little-examined premontane rain forest area in southern Costa Rica. S. globulifera is a shade-tolerant canopy tree with bright red flowers that are pollinated primarily by hummingbirds. Bats disperse the seeds by eating fruits and then passing seeds on through guano at their resting sites. Like many areas in the tropics, the study area consisted of an area of fragmented forest with a number of large nearby members of the species standing alone in open pasture land. There were neither seedlings or saplings of S. globulifera in the pastures, suggesting poor habitat quality for germination and growth.

The question was simple: What trees are the parents of seedlings growing in the forest fragments? Finding the answer would have been nearly impossible even a decade ago until the invention of sensitive techniques that allow researchers to determine the exact genetic makeup of individual plants in an ecosystem. Even now, the problem is daunting, since there were more than 800 possible parent pairs.

"In trying to tell who the parents are, we had to use the same techniques used in forensic analysis to determine a child's parents," said Hamrick. "We were able to do this only because Preston was able to develop the techniques for our specific genetic analysis himself."

The breakthrough came in using segments of DNA called microsatellites as specific markers for S. globulifera. These markers allowed Aldrich and Hamrick to determine the pedigree for a number of seedlings and saplings in a one hectare forest fragment on their 38.5- hectare research area. (A hectare is a metric unit of area equal to 2.471 acres.) The scientists knew the genetic composition of all the adults, 42 individuals, in the study area.

The results of the genetic analysis were startling. Out of nearly 250 seedlings studied from a single forest fragment, some 68 percent were produced by adults in pastures ñ not from adults within the fragments themselves. Moreover, of the seedling produced by pasture trees, 77 percent came from only two trees. Adults left in the fragment produced less than 5 percent of the seedlings in their own patches. The importance of the discovery lies in the fact that the genetic diversity of seedlings in forest fragment may be relatively small indeed.

"If you looked at the number of seedlings superficially, you might say that this is a healthy rate of regrowth," said Hamrick. "But in truth, the effect is ecologically unhealthy due to the potential for inbreeding in subsequent generations." As humans know, inbreeding can expose deleterious genes.

The scientists have several theories why the pasture trees have such an overwhelming impact on gene flow. First, there is little competition for the pasture trees for sunlight and nutrients, giving them superior abilities to flower and fruit. Second, the abundance of flowers may attract more hummingbirds for pollination. Finally, bats have easy pickings of the fruit and take them from the pasture trees to the forest fragment, where they eat and then pass seeds back to soil in guano.

The study has important implications for conservation and forest restoration. And it shows that the impact of deforestation has been far more devastating that the simple removal of individual trees. At least in this species of tree, fragmentation has resulted in the possibility of a serious loss of genetic diversity in this test area.

Thus, areas that look healthy in terms of regrowth may not be healthy at all over the long term. They may be facing serious future problems due to problems with genetic drift and inbreeding. American farmers already know of the problems this can cause, since they barely avoided a near-disaster because of over-planting of corn with too little genetic diversity in the 1970s.

Still, scientists say it is beginning to become apparent that it will be difficult to make any kind of blanket statements about gene movement among populations of forest species. Indeed, botanists say we are now only beginning to understand what happens to genetic diversity in natural populations over time -- and why.

"One of the important things this study has shown us is that the superficial appearance of an area might not be telling you the whole story," said Hamrick. "Each situation has unique characteristics that make if very difficult to say that tropical trees in a certain situation will behave this way or that. Quite simply, our study area looked healthy, but it wasn't."


(Editors/Writers: James Hamrick will be out of the country until Monday, July 6. He may be reached at the above phone number at that time.)

Register for reporter access to contact details