This week, the country's leading thinkers on biodiesel are in San Francisco for the 2009 National Biodiesel Conference and Expo (see Biofuel Maker to Switch From Deep-Fat Fryers to Jatropha). Here are some of the highlights so far:

One of the more cost effective places to grow biodiesel in the country could be the side of the road, according to Dallas Hanks from Utah State University. He heads up a project called Freeways to Fuel to study the feasibility of planting crops in the 30-foot wide shoulder next to roads.

The key is that the land would essentially be free. In the U.S. there are approximately 10 million acres lining 4 million miles of road that could be planted with oil seed crops, he said. There's another 1 million acres alongside the 140,000 miles of railroad right-of-ways.

"Right now, we mow it," he said. Utah alone spends about $300 a year per mile to keep lawns and open space around its roads clean, he said. Growing an oil crop would at least generate revenue at the same time.

The nation's 20,000 airports can probably spare about 2 million acres and the military-which has 29 million acres under its control-could probably spare 8 million acres.

"The state owns the land. The military owns its land and they pay people to take care of it," he said. "We can probably produce $2 a gallon biodiesel on this type of system."

The problem? Soils on the side of the road tend to get quite compact: Furrows that get cut in roadside soil tend to stay there, as if carved in Play-Doh. Transportation engineers also worry about harvesting and planting ruining the structural integrity of their roads. In a limited experiment in Utah in 2007 and 2008, the plants yielded very little oil. Utah, however, is one of the driest states in the nation. Universities in the Midwest and Northeast are going to conduct similar experiments where yields could be higher.

Looking for a novel feedstock? Castor could be the answer, according to Dick Auld from Texas Tech. The castor plant is drought tolerant, salt tolerant, grows on marginal land, probably amenable to genetic modification and is quite oily. The plant could yield 63 to 210 gallons of oil an acre-that's low compared to some crops but it would grown on marginal lands, thereby dropping the cost of production. Auld speculated that the Texas Panhandle, the dry 80,000 square miles toward the top of the state, could turn into Castor country.

"We believe we could be the most economical source of biological feedstocks," he said.

It's also not a food crop. The plant, originally from the tropics, produces the highly toxic ricin.

But if there's a feedstock on everyone's lips, it's the Chinese Tallow tree, according to Courtney McColgan, an associate at Draper, Fisher Jurvetson.

Ben Franklin is credited with bringing the tree to the continent when the U.S. was still a set of colonies owned by Britain. Since then, it's become a pesky, invasive species in the South. Some experts say it could produce several hundred gallons of feedstock per acre.

Algae was criticized by some attendees for requiring too much water, she said. An expert at the show said algae could still be four to five years away.

And one more feedstock to think about: Gumweed, an unattractive desert plant currently under study at the University of Nevada at Reno.