In terms of greenhouse-gas accounting, forests land in the credit column because they absorb carbon dioxide (CO2) from the atmosphere. As a result, these carbon sinks, as they are called, can counterbalance increasing man-made CO2 emissions. But quantifying their effects is extremely difficult. Further complicating matters, new findings published today in the journal Science suggest that some of the assumptions currently being employed to make such estimates may be incorrect.

The roots of a tree provide the connection between it and the surrounding soil. Until recently researchers hypothesized that roots less than two millimeters in diameter, which are responsible for most below-ground nutrient cycling, live for about a year. Roser Matamala of Argonne National Laboratory and her colleagues have now completed a five-year-long experiment that suggests otherwise. The team exposed two types of trees--sweetgums and loblolly pines--to CO2 labeled with carbon 13 and then tracked the heavy carbon as it appeared in the soil. According to the report, the roots lasted between 1.2 and nine years, with the pine tree roots lasting longer on average than those of the sweetgum. "These long turnover times suggest that root production and turnover in forests have been overestimated," the authors write, "and that sequestration of anthropogenic atmospheric carbon in forest soils may be lower than currently estimated."

Unlike balancing a checkbook, reconciling the greenhouse gas budget is extraordinarily complex. Tiny tree roots may seem like a small thing to be focusing on, but Susan E. Trumbore of the University of California at Irvine and Julia B. Gaudinski of the University of California at Santa Cruz note in an accompanying commentary that "unless we recognize that root behavior is as complex as that of its counterparts above ground, the rules governing allocation of carbon to roots and the role of roots in soil carbon cycling will remain well-kept secrets."