A price tag on the planet's ecosystems.

It's like trying to audit the books of someone missing a lot of receipts and bank statements. Yet a group of 13 researchers has pulled together an emerging body of studies on the value of ecosystems and come up with a rough figure for the annual worth of Earth's natural goods and services: $33 trillion.

That includes an estimated value for just about everything under the sun--from recreational beaches to forest lumber to hidden services like the ocean's regulation of atmospheric carbon dioxide and grasslands' provisioning for pollinators. A few natural resources were omitted, such as nonrenewable fuels and minerals. In comparison to the ecosystems' worth, the researchers report in the May 15 Nature, the world's annual gross national products total about $18 trillion.

One goal of the exercise, says Robert Costanza of the Institute for Ecological Economics at the University of Maryland in Solomons, is to answer a lingering question in economics: "Are environmental services big potatoes or small potatoes? ... We're saying they're very big potatoes."

They're probably even bigger than $33 trillion, the researchers and other observers say. That figure, the average of calculations ranging from $16 trillion to $54 trillion, comes from converting and tallying a range of ecosystem values from more than 100 studies. For example, says Costanza, data relating the size of shrimp harvests in Louisiana to the extent of local wetlands were included in calculations of the average value of a hectare of wetland, which was then applied to the global extent of that habitat. Recreational values came from reports of people's willingness to pay for access to a coral reef, lake, or other natural area.

The overall average is conservative because it omits ecosystem services in several biomes, such as desert and tundra, that haven't been studied in terms of ecosystem values. Ecologist Stuart L. Pimm of the University of Tennessee in Knoxville points out that It also ignores the services in urban areas, despite the substantial value of green spaces like New York's Central Park.

The most valuable ecosystems per hectare turn out to be estuaries and wetlands--in part, because these areas have been the object of the most study, says Costanza. Other ecosystems will probably increase in value as they are examined more extensively.

The field of ecological economics underlying this analysis is only about a decade old, and its practitioners acknowledge its limitations--the researchers list a dozen caveats to their study. "We're putting this forth as a starting point," says Costanza.

It's a reasonable one, says environmental economist Lawrence H. Goulder of Stanford University. Pimm says the power of the analysis comes from the per hectare value of habitats and "the way in which it will inform local decisions" when it comes to planning.