Welcome to La Tirimbina, a 740-acre research preserve tucked in the Costa Rican rainforest.   Although we're just 45 miles from the San Jose airport, the last bit of the trip is a bone-jolting ride best taken in sturdy four-wheel drive.

      Tropical biologist Allen Young points to a tree that would be like any other rainforest shrub except for one thing.  It's a cacao tree -- one of the few species with flowers and fruits that grow directly out of the trunk and lower branches of the tree.  Young hacks off a red-brown fruit about the shape and size of a kid's toy football, ridged and bumpy.  He slices it open with his machete and offers a taste of the flesh -- lemony and a little less dense than an apple.  A seed squirts out of his hand.

      "That's the strategy of this tree," he says.  "The tree seems to want birds and other animals to come by, eat this great-tasting fruit and distribute the seeds around the forest."   In fact wild cacao fruits don't ever get the chance to fall off the tree.  He slices open the slippery white seed.  The purplish bean inside is the raw ingredient of your favorite chocolate bar, but it still has a long way to go.

  "Here's a tree that produces prodigious numbers of flowers," says Allen Young, gazing at an ornate pink flower.  "You'll see waves of flowers on the trees at certain times of the year.  But only 2 to 3 percent of them are productive in terms of yielding harvestable-size pods."  A cherry orchard in Maryland or Wisconsin might produce seven or eight times as much fruit for that number of flowers.

      For 10 years he studied why so few blooms were pollinated.  As the entomology curator at the Milwaukee Public Museum, which owns the preserve, he focused, naturally, on the insects that frequented cacao, and found that instead of bees certain midge species -- a group of gnatlike flies -- were its chief pollinators.  Young scattered chopped-up bits of banana plants and leaf litter in stands of cacao and recorded the numbers of the beneficial midges.  He also experimented with other possible aids to breeding the flies, including artificial bromeliads constructed of plastic cups and decaying plant matter.

      He stayed on the lookout for clues -- the times of day and seasons most favorable for pollination; whether sun or shade, wet or dry areas fostered the largest population of midges; whether the midges might have a preference for certain kinds of rotting debris; whether more midges necessarily lead to larger numbers of healthy cacao pods.

      Every experiment pointed to the same conclusion:  The bigger a cacao plantation, the more it frustrates the midges in their efforts to pollinate individual cacao flowers.  The midges are happiest in the shady, humid rainforest full of damp leaf litter and epiphytes growing on the cacao trees.  They have little incentive to penetrate very far into the dryer, sunnier and well-tended plantations.

      Not only that, but he identified which of the 78 aromatic substances cacao flowers give off to attract the midges.  Young found that selective breeding of domesticated cacao trees inadvertently altered the flower's smell, making it less desirable to nearby midges.

  The chocolate tree (Theobroma cacao) probably evolved about 15,000 years ago as a wild tree in the Amazon basin growing in the understory of the tropical rainforest. It thrived in the shade of the forest floor and lived on the nutrients and water passed down from the plants above. By 1,000 B.C., Mesoamerican people knew that small plantings of cacao at the edge of the rainforest, mixed in with corn, vanilla, yucca and other food plants, would provide the highest yields of prized cacao seeds. But modern farmers ignored that wisdom.

  Silvino Villegas, a second-generation cacao farmer, started working on the cacao finca when he was 11, and stopped going to school four years later to work there full-time.  "It was all I did for 25 years. I took over management of the 35-acre plantation when my father retired."

      His traditional livelihood was destroyed by Monilia pod rot -- a fungus that decimated Costa Rica's chocolate exports.  "The Monilia hit this particular farm about 1980," he remembers. Symptoms included splotches of powdery white or yellow mold on the fruits.

      "We didn't know what it was, so we took the plants to the minister of agriculture.  He helped identify it.   We started cutting and cutting the bad fruits and branches, and that didn't work.  Then we cut some more and burned the branches on the spot.  By 1985 we pretty much stopped growing cacao here because of the Monilia."  Perversely all the cutting and burning may have helped spread the fungus even further.

      Worldwide production of raw cocoa went from 1.5 million metric tons in 1965 to 2.4 million metric tons in 1995 (an increase of more than 160 percent).  But Costa Rican production fell 73 percent, from 11,000 metric tons to just 3,000 metric tons.  Today the only known way to control the spread of pod rot is to hand-pick infected pods before the fungus releases its spores.

      Large, undifferentiated plantations of a single crop left Costa Rica's cacao trees vulnerable, says Allen Young, which is why Costa Rica's cacao industry was destroyed by Monilia pod rot in the 1980s.  "Monoculture invites problems, and makes you a sitting target for disease and pestilence.  But if you grow cacao as an integrated part of a forest ecosystem, what you're doing is softening the negative impact of these problems, because there are many more hosts around for those things to attack."

      Scientists at La Tirimbina are already working on a sustainable forestry experiment, and Young wants to include cacao in the plan.  "Rather than clear-cutting, researchers have been cutting down only selected trees, retaining the integrity of the forest, and carefully measuring the impact of this limited harvest," Young says.  The lessons of ecology and tradition suggest that cacao trees can be grown successfully in small stands in the forest understory where hardwoods have been removed.

      If Young's latest project goes as planned, a year from now as many as 1,000 new chocolate trees will be growing by the Sarapiqui River.  He hopes these plants will thrive and show small farmers in this corner of Costa Rica how they might once again make their living growing the magical stuff of brownies and birthday cakes.  If chocolate makes a comeback here, it could put many people back to work.  Young predicts that it will be safer and more economical, because insecticides and fungicides can be reduced when cacao trees are able to take advantage of protections provided by their natural habitat.