The Sudden oak death epidemic that has killed more than a million trees throughout coastal California started in two sites: Scotts Valley and on Mount Tamalpais in Marin County, a new genetic analysis reveals.

Through genetic detective wok, scientists found that pathogens at both locations - separated by 62 miles - share identical DNA footprints, indicating that they are related, probably through the nursery trade, said lead investigator Matteo Garbelotto of the University of California Berkeley.

"Our study reconstructs the Sudden oak death epidemic," Garbelotto said."Having multiple introductions explains why it is so extensive."

The discovery sheds new light on a horticultural murder mystery involving a pathogen found in Asia that is now devastating swaths of California forests. The finding does not lead to a cure for the epidemic. From a Scotts Valley rhododendron nursery in the Bean Creek area, the pathogen escaped into nearby forests, he said. People accelerated its spread by buying infected plants and planting them throughout the region. It now ravages parts of Monterey County's Pfeiffer Big Sur State Park, and has killed thousands of trees in Santa Clara, Santa Cruz, San Mateo and Alameda counties.

Although there is no nursery near the Mount Tamalpais site, there are several large homes with extensive landscaping. Once established, the pathogen - Phytophthora ramorum - spread north into Sonoma and Humboldt counties.

The research team does not know if the infected plants in Marin came from Scotts Valley - or whether both sites independently acquired their plants from the same infected shipment.

Similarly, researchers have not identified the source of the infected plants at the Scotts Valley business, which is a wholesale nursery.

"There is an incredibly complex pattern of plant trading between nurseries," Garbelotto said."It is not easily reconstructable."

Plants such as camellias and rhododendrons can spread the disease, which weakens but does not kill them. But it can be fatal to coast live oak, black oak and tanbark oak trees. Like humans, plants show wide variability in their vulnerability to disease; of the 22 species known to become infected, the disease kills just a handful.

It kills by girdling the tree, destroying its cell walls and disrupting the flow of nutrients. A federal quarantine affecting 10 California counties now bans the shipment of soil and plants from more than a dozen host species.

One of its relatives was responsible for the potato blight in Ireland in the 1840s, and another species nearly killed Scotland's soft-fruit industry in the 1920s. A strain of the same microbe introduced to Australia in the early 1900s has transformed forests into grasslands.

Rhododendrons had been sick at the Bean Creek nursery for several years before testing in 2001, according to Garbelotto.

Owner Peter Moerdyke of Bay Laurel Nursery - the only such business in Bean Creek - denies that his nursery was an original source of the epidemic. Because he propagates his own plants and does not import specimens, he contends that his nursery was infected by the surrounding forest - or that the four or five former nurseries in the area, now gone, brought it in.

He noticed some plants were sick for awhile but his repeated calls to the California Department of Forestry went unheeded for months, he said. It was only when a local agricultural extension agent notified scientists that the disease got attention, he said.

"I don't ship, like large commercial nurseries," he said."This was just the first place they studied it - because I was responsible and called their attention to it. By the time they got here, it was already so widespread."

Because symptoms take a long time to appear, scientists speculate that the microbe has been in the region for many years - perhaps as far back as the mid-1980s.

Scientists traced the microbe's ancestry by counting mutations in its genetic material. As it evolves over time, new infestations became genetically distinct from old infestations.

They analyzed genetic markers of nearly 300 P. ramorum samples taken from forests in five California counties. From the samples, they identified 35 unique strains of the pathogen. The Santa Cruz and Marin County locations had the most ancestral populations.

The study, published in the April edition of the journal Molecular Ecology, was presented today at the annual meeting of the California Oak Mortality Task Force in San Raphael.

To better understand how different areas ended up with matching strains, researchers analyzed how far the pathogen spreads naturally, using GPS coordinates to calculate the distance between plants with identical strains. They found that the vast majority of spread occurred within 200 to 300 yards.

However, the wind is able to blow spores as far as three miles, they found.

The inadvertent movement of infected plants by people explains why perfectly matched strains are found longer distances from one another.

Some current strategies, such as clear-cutting, are not a practical way to contain a pathogen that can travel three miles, he said.

A forest management plan that uses controlled burns, grazing and other tools could slow its spread, turning an epidemic into a less destructive endemic, said Garbelotto.

"There is no way we can get rid of it or stop it from moving around," he said.

"We need to do something that gives it less of an advantage."

Reach Lisa M. Krieger at lkrieger@mercurynews.com or 408-920-5565.
Sudden Oak Death

The fungus creates cankers under the bark of the trunk. But other organisms produce symptoms very similar to Sudden Oak Death. Many local arborists, foresters, and tree service companies have undergone training to diagnose the disease.

To reduce the risk of disease spread, it is recommended that dead trees be left on site. This material can be chipped and used as mulch. Composting can kill the pathogen, but the compost must reach a high temperature of 130 degrees for at least two weeks.

For more information click on the Sudden Oak Death Task Force Web site at www.suddenoakdeath.org.