Global consistency in response of terrestrial ecosystem respiration to temperature

Many studies have been carried out to quantify the trend of terrestrial ecosystem respiration (Re) in a warming world, but a convincing explanation has not yet been provided, because numerous contradictory findings have been reported across different spatial and temporal scales. In this study, we applied nine empirical models to a 1,387 site-years FLUXNET dataset adopting a grouping regression method and found a dramatically consistent result. The sigmoid function outcompeted other statistical models in almost all datasets, and on average, achieved a staggering R2 value of 0.92. The results reconcile previous findings on "hot-acclimatization", "cold-acclimatization" and inconsistent reported Q10 values. Even though Re of all biomes followed a sigmoid function, the shape of the S-curve varied strongly across sites, which empirically confirms the hypothesis that each ecosystem adapts to its physical environment. The comparatively small year-to-year variation of the sigmoid pattern (a minor fraction of site-to-site variation) implies that our results feature good potentials in estimating global Re under a warming climate. Similar to others, we found that the simulated respiration value under mean annual temperature (MAT) showed an excellent match with mean annual respiration (R2=0.90). In contrast, we did not find consistent patterns for either the site-to-site or year-to-year variation of the sigmoid function. We concluded that year-to-year covariation between MAT and mean annual Re (i.e., respiration under global warming) could not be simply inferred from studies at smaller scales, but instead, detailed warming regime and year-to-year ecosystem acclimatization studies should be conducted.