Geophys. Res. Lett. 38, L19704 (2011)

The two recent major droughts in the Amazon Basin, in 2005 and 2010, provided an opportunity to study the effects of drought on tropical forests and improve predictive impact models for the region.

Michael Toomey from the Department of Geography at the University of California, Santa Barbara, USA, and his co-workers analysed remote-sensing data of land-surface temperatures in the Amazon to investigate the contribution of heat stress to the observed changes in aboveground living biomass during the droughts, such as increased tree death.

They found that heat stress was a better indicator of biomass variability during the droughts than water stress — models that only incorporated precipitation patterns were 17% less accurate in predicting the biomass changes than those that only used heat stress. Models that incorporated both heat and moisture stress were the best, able to account for around 65% of biomass variability. Heat stress played an important role in both droughts, the authors conclude, adding that models to predict drought impacts in tropical forests should take account of this.