Tropical forest temperature thresholds for gross primary productivity

Abstract Tropical forests are hyper‐diverse and perform critical functions that regulate global climate, yet they are also threatened by rising temperatures. Canopy temperatures depart considerably from air temperatures, sometimes by as much as air temperatures are projected to increase by the end of this century; however, canopy temperatures are rarely measured or considered in climate change analyses. Our results from near‐continuous thermal imaging of a well‐studied tropical forest show that canopy temperatures reached a maximum of ~34°C, and exceeded maximum air temperatures by as much as 7°C. Comparing different canopy surfaces reveals that bark was the warmest, followed by a deciduous canopy, flowers, and coolest was an evergreen canopy. Differences among canopy surfaces were largest during afternoon hours, when the evergreen canopy cooled more rapidly than other canopy surfaces, presumably due to transpiration. Gross primary productivity ( GPP ), estimated from eddy covariance measurements, was more strongly associated with canopy temperatures than air temperatures or vapor pressure deficit. The rate of GPP increase with canopy temperatures slowed above ~28–29°C, but GPP continued to increase until ~31–32°C. Although future warming is projected to be greater in high‐latitude regions, we show that tropical forest productivity is highly sensitive to small changes in temperature. Important biophysical and physiological characteristics captured by canopy temperatures allow more accurate predictions of GPP compared to commonly used air temperatures. Results suggest that as air temperatures continue to warm with climate change, canopy temperatures will increase at a ~40% higher rate, with uncertain but potentially large impacts on tropical forest productivity.