If true, this conclusion does not bode well for the biosphere; and, hence, it is important that it be carefully scrutinized.  Hence, we here review some items that raise significant questions about the study's data and their interpretation.

In a news story in the 25 April 2003 issue of Science that is devoted to the Clark et al. paper, Kaiser (2003) reports that the U.S. Forest Service's Ariel Lugo of San Juan, Puerto Rico, describes Clark et al.'s attempt to link a small portion of their single patch of forest (that part that comprises but six of the more than 300 tree species present) to atmospheric CO2 levels across the entire tropics as "very daring."  In addition, Kaiser notes that Richard Condit of the Smithsonian Tropical Research Institute in Panama questions Clark et al.'s assumption that tropical forests worldwide grow more slowly in warm years, saying "it isn't what we see in Panama."

It's not what scientists see in other parts of the world either.  Phillips and Gentry (1994), for example -- in a study of forty tropical forests from all around the world -- determined that the turnover rates of these forests have been rising ever higher since at least 1960, with an apparent pantropical acceleration beginning about 1980.  Over this entire period, as may be verified by visiting the World Temperatures section of our website, surface air temperature records depict sustained and significant warming, which (if true) suggests that tropical forests exhibit just the opposite growth response to rising temperature as that proclaimed by Clark et al.  Pimm and Sugden (1994) also note another possible source of the worldwide increasing tropical forest growth rates, stating that the consistency and simultaneity of the forest growth trends on several continents observed by Phillips and Gentry led them to conclude that "enhanced productivity induced by increased CO2 is the most plausible candidate for the cause of the increased turnover."

A few years later, Phillips et al. (1998) analyzed forest growth rate data for the period 1958 to 1996 for several hundred plots of mature tropical trees scattered around the world, finding that tropical forest biomass, as a whole, increased substantially over the period of record.  In fact, the increase in the Neotropics was equivalent to approximately 40% of the missing terrestrial carbon sink of the entire globe.  Hence, and in total contradiction of the conclusions of Clark et al., Phillips et al. concluded that tropical forests "may be helping to buffer the rate of increase in atmospheric CO2, thereby reducing the impacts of global climate change [our italics]."  And, again, they identified the aerial fertilization effect of the ongoing rise in the air's CO2 content as one of the primary factors that may be responsible for this phenomenon.

But what about the second aspect of the Clark et al. paper, i.e., the inversion tracer-transport model results that imply large net fluxes of CO2 from the terrestrial tropics to the atmosphere when the La Selva, Costa Rica, data depict reduced forest growth in warmer years?  These model calculations are anything but simple; and the results employed by Clark et al. have been seriously questioned because they are so different from those that have been calculated by other scientists.  And because of these facts, according to the Clarks (as reported by Kaiser), "the paper ran into problems in review at Nature and Science," being rejected by those journals before finding acceptance by the Proceedings of the National Academy of Sciences.

Another red flag about Clark et al.'s statement that the phenomenon they claim to have discovered "would accelerate global warming" is raised by their statement that, although the calculated tropical terrestrial net CO2 flux increased from a 0.9 Pg C per-year source in the 1980s to a 2.6 Pg C per-year source in the 1990s, the estimated global land flux of CO2 went from a 0.2 Pg C per-year sink in the 1980s to a much greater 0.8 Pg C per-year sink in the 1990s, due to concomitant increased carbon uptake in temperate and boreal zones.  Hence, even if Clark et al. were correct about tropical forest growth response to increasing air temperature (which they appear not to be), by their own acknowledgement the terrestrial world, as a whole, appears to remove more CO2 from the air when temperatures rise, which would tend to do just the opposite of what they suggest and decelerate global warming.

In conclusion, there are sufficient inconsistencies in both the empirical and modeling efforts of Clark et al. that little credence should be given to the alarm they attempt to raise about the claimed inability of the biosphere to temper the ongoing rise in the air's CO2 content, not the least of which is the strange behavior of the trees they studied.  As the senior author is quoted by Kaiser as saying with respect to the tiny piece of Costa Rican forest in which she and her husband live and work, "it seems to me we're getting all kinds of weirdness."

We couldn't agree more.

References
Clark, D.A.  2002.  Are tropical forests an important carbon sink?  Reanalysis of the long-term plot data.  Ecological Applications 12: 3-7.

Clark, D.A., Piper, S.C., Keeling, C.D. and Clark, D.B.  2003.  Tropical rain forest tree growth and atmospheric carbon dynamics linked to interannual temperature variation during 1984-2000.  Proceedings of the National Academy of Sciences, USA 100: 10.1073/pnas.0935903100.

Kaiser, J.  2003.  An intimate knowledge of trees.  Science 300: 566-567.

Phillips, O.L. and Gentry, A.H.  1994.  Increasing turnover through time in tropical forests.  Science 263: 954-958.

Phillips, O.L., Malhi, Y., Higuchi, N., Laurance, W.F., Nunez, P.V., Vasquez, R.M., Laurance, S.G., Ferreira, L.V., Stern, M., Brown, S. and Grace, J.  1998.  Changes in the carbon balance of tropical forests: Evidence from long-term plots.  Science 282: 439-442.

Pimm, S.L. and Sugden, A.M.  1994.  Tropical diversity and global change.  Science 263: 933-934.