Clues to the origins of flowering plants.

Meticulous microscopic examinations of the primitive seeds of a desert shrub commonly known as "Mormon tea" have yielded new evidence concerning the evolutionary link between flowering and nonflowering plants.

The research suggests that a now-extinct ancestor of the lanky shrub -- whose scientific name is Ephedra -- first evolved a seed structure called the endosperm to nourish its developing plant embryos. William E. Friedman, a botanist at the University of Georgia in Athens, speculates that this innovation set the stage for the evolution of flowering plants by providing a more efficient means for plants to nurture successive generations.

In the Jan. 17 SCIENCE, Friedman reports that the starchy endosperm that makes up the bulk of most plants' seeds is really the fraternal twin of the embryo it feeds. As such, he says, "it is essentially a deviant plant embryo" that develops from a second fertilized egg within the plant ovary.

Paleobotanists surmise that before the advent of endosperm, all growing plant embryos fed on cells from their embryo sacs -- the maternal tissue that gives rise to unfertilized eggs. But this strategy had a serious limitation, Friedman explains. Plants had to develop plump, nutrient-rich embryo sacs even before their offspring were conceived -- investments of time and energy that didn't pay off if the eggs never got fertilized by a sperm cell or if the young embryos died on the vine.

Because endosperm is spawned at the same time as each embryo -- and always grows just a little faster than its twin baby plant -- it ensures that plants don't make such needless investments, Friedman says. Spending less time and energy on reproduction could have allowed early plants to evolve more sophisticated life cycles that included flowering, he asserts.

Endosperm is also a richer source of food for developing plants than embryo sac cells, Friedman says. Because it contains genes from both the maternal and paternal plants, it is less likely to have a defect caused by one faulty gene. And because it bears multiple copes of both parent plants' genes, it has extra sets of blueprints for building up greater stores of food faster, Friedman says.

He found traces of endosperm's evolution within 34 fertilized eggs collected from Ephedra plants growing wild outside Tucson, Ariz. Using a microscope, he observed that in all instance, each of the Ephedra sperm cells' two nuclei had fused with an egg. Even though the extra fertilized egg later dies off -- and the nonflowering Ephedra's embryo ultimately feeds on its embryo sac cells -- Friedman says this is evidence of how endosperm once formed.

Lloyd Mogensen, a plant embryologist at Northern Arizona University in Flagstaff, agrees that Friedman has "nailed down" the origin of endosperm. "This is the first example . . . of how endosperm came into being genertically," Mogensen says.

Two years ago, researchers described a 110-million-year-old fossil containing an herb-like plant bearing the oldest known flowers (SN: 2/10/90, p.85). The buds are likely descendants of Ephedra's flowering ancestral cousin.