Abstract
Gross primary productivity (GPP) has been proven to respond positively but asymmetrically to increases in precipitation, such that the productivity gains in wet periods may be unequal to the productivity declines in dry periods. Climate model projections indicate significant precipitation changes in terms of mean and variability over the coming decades; however, how the changes in precipitation regimes affect the asymmetric GPP responses remains an open question. In this study, the responses of GPP to precipitation changes were assessed by using model simulations under a range of experimental scenarios from 1995 to 2100. In particular, the asymmetry of GPP in response to precipitation was quantified by the asymmetry index (AI), where positive AI suggests larger increases in GPP during wet years and negative AI indicates greater decreases in GPP during dry years. The results showed that: (1) In the historical period of 1995–2014, positive GPP asymmetry occurred in 47.50% of the global vegetated area including central Australia (AI = 0.24), northern Russia (0.04), and southern western Africa (0.03), while negative asymmetry was observed in the Amazon (−0.14), central North America (−0.05), and eastern Europe (−0.03). (2) During the period of 2015–2100, the positive AI in central Australia (northern Russia) is projected to increase (decrease) at a rate of 0.007 decade −1 (−0.013 decade −1 ) under the high emission scenario. (3) In the perspective of different land cover types, clear positive GPP asymmetry is found in shrublands and grasslands during the historical period, and the positive sign is expected to be enhanced in the former but diminished in the latter by the end of the century. The changes in the asymmetric GPP responses to precipitation are suggestive of variations in terrestrial ecosystem carbon uptake and the need for effective ecosystem management strategies.