Response of Gross Primary Productivity to Drought Time-Scales Across China

Drought largely affects terrestrial gross primary productivity (GPP) by its duration (i.e., drought time-scales). However, the large-scale impact of drought time-scales on GPP is still not well understood, partly due to limited understanding of ecological processes and difficulties in quantifying drought time-scales. Here, we examined the responses of GPP to drought in China during the last four decades, using state-of-the-art GPP data estimates and a multiscalar drought metric. Most of the GPP data suggested that the time-scales at which drought has the largest influence on annual GPP (dominant drought time-scales) were well related to climate region and vegetation type. The GPP responded to drought more rapidly in arid regions than that in semiarid and subhumid regions, with a mean dominant drought time-scale of 15 and 20 months, respectively; as plants in arid regions can rapidly adapt to water shortage through physiological strategies, while, in semiarid and subhumid regions plants usually can withstand a degree of water deficits. In a humid region, GPP did not rapidly respond to drought, probably because the high soil water availability buffered the effect of drought. In addition, most of the GPP estimates suggested that grasslands and croplands tended to respond to drought at shorter time-scales than that of forests, shrublands, and sparse vegetation. The study highlights the importance of improving the effects of water availability on vegetation activity and further consideration of vegetation adaptability to drought in both remote sensing GPP algorithms and parameterizations of current terrestrial ecosystem models.