Seasonal and interannual variations of ecosystem photosynthetic characteristics in a semi-arid grassland of Northern China

Abstract The ecosystem apparent quantum yield (α), maximum rate of gross CO2 assimilation (Pmax) and daytime ecosystem respiration rate (Rd), reflecting the physiological functioning of ecosystem, are vital photosynthetic parameters for the estimation of ecosystem carbon budget. Climatic drivers may affect photosynthetic parameters both directly and indirectly by altering the response of vegetation. However, the relative contribution and regulation pathway of environmental and physiological controls remain unclear, especially in semi-arid grasslands. We analyzed seasonal and interannual variations of photosynthetic parameters derived from eddy-covariance observation in a typical semi-arid grassland in Inner Mongolia, Northern China, over 12 years from 2006 to 2017. Regression analyses and a structural equation model (SEM) were adopted to separate the contributions of environmental and physiological effects. The photosynthetic parameters showed unimodal seasonal patterns and significantly interannual variations. Variations of air temperature (Ta) and soil water content (SWC) drove the seasonal patterns of photosynthetic parameters, while SWC predominated their interannual variations. Moreover, contrasting with the predominant roles of Ta on α and Rd, SWC explained more variance of Pmax than Ta. Results of SEM revealed that environmental factors impacted photosynthetic parameters both directly and indirectly through regulating physiological responses reflected by stomatal conductance at the canopy level. Moreover, leaf area index (LAI) directly affected α, Pmax and Rd and dominated the variation of Pmax. On the other hand, SWC influenced photosynthetic parameters indirectly through LAI and canopy surface conductance (gc). Our findings highlight the importance of physiological regulation on the photosynthetic parameters and carbon assimilation capacity, especially in water-limited grassland ecosystems.