Explaining inter-annual variability of gross primary productivity from plant phenology and physiology

Climate variability influences both plant phenology and physiology, resulting in inter-annual variation in terrestrial gross primary productivity (GPP). However, it is still difficult to explain the inter-annual variability of GPP. In this study, we propose a Statistical Model of Integrated Phenology and Physiology (SMIPP) to explain the contributions of maximum daily GPP (GPP_max), and start and end of the growing season (GS_start and GS_end) to the inter-annual variability of GPP observed at 27 sites across North America and Europe. Strong relationships are found between the anomalies of GS_start and spring GPP (r = 0.82 ± 0.10), GPP_max and summer GPP (r = 0.90 ± 0.14), and GS_end and autumn GPP (r = 0.75 ± 0.18) within each site. Partial correlation analysis further supports strong correlations of annual GPP with GS_start (partial r value being 0.72 ± 0.20), GPP_max (0.87 ± 0.15), and GS_end (0.59 ± 0.26), respectively. In addition, the three indicators are found independent from each other to influence annual GPP at most of the 27 sites. Overall, the site-calibrated SMIPP explains 90 ± 11% of the annual GPP variability among the 27 sites. In general, GPP_max contributes to annual GPP variation more than the two phenological indicators. These results indicate that the inter-annual variability of GPP can be effectively estimated using the three indicators. Investigating plant physiology, and spring and autumn phenology to environmental changes can improve the prediction of the annual GPP trajectory under future climate change.