Bananas are different from most other cultivated plants in that almost all the varieties-including the Cavendish-lack seeds (that round, dark center in a banana slice is the vestige of what was once the fruit's reproductive core). Cultivated bananas never reproduce sexually on their own. Rather, new stems grow from the existing root, sometimes for many years. Forcing the pollen from one male flower to make its way to the female of another plant, however, is how traditional banana breeders like Aguilar's team develop new varieties. Most mornings, usually just as dawn is breaking, a team of hand pollinators pedal through FHIA's dirt-tracked fields on battered three-speed bicycles. They move from plant to plant, gathering the powdery pollen from the males and transferring it to receptive female flowers, keeping meticulous records of their activities (Aguilar calls the field "a giant spreadsheet"). The goal of all this is to get seeds, and to use them to grow Aguilar's experimental varieties, one of which, he hopes, will ultimately yield a tasty, market-friendly Cavendish replacement. What are the odds of an individual seed ultimately yielding a thriving hybrid? "About 1 in 10,000," Aguilar says.