Identifying critical climate periods for temporal dynamics of aboveground net primary productivity in the Eurasian steppe region

Abstract

Understanding correlations between aboveground net primary productivity (ANPP) and climate change has remained a long-standing goal in ecology. Annual ANPP experienced significant interannual variations in the Eurasian steppe region (EASR) over past three decades. However, critical climate periods for annual ANPP remain poorly understood. Here, we comprehensively examined ANPP–climate relationships in the EASR on a seasonal scale from 1982 to 2013. Partial correlation coefficients were calculated between annual ANPP and seasonal climatic variables. A multiple linear regression (MLR) model was implemented between annual ANPP and three climatic factors over four seasons. For each climatic variable during each season, the relative importance for the MLR model was determined to identify critical climate periods for annual ANPP. Results demonstrated that (1) seasonal differences could be observed in response of annual ANPP to climatic variables. The influence of precipitation (PRE) on annual ANPP gradually increased from previous-year winter to summer seasons, with a diminishing effect of solar shortwave radiation (SWD), and decreased from summer to autumn with an increasing effect of SWD. Summer TMP imposed the strongest effects, followed by previous-year winter TMP. The influence of TMP exhibited only slight changes between spring and autumn. (2) PRE, TMP and SWD were crucial climatic variables affecting annual ANPP in 41%, 31% and 28%, respectively, of the EASR, especially summer PRE, summer TMP and previous-year winter SWD. The critical climate periods for annual ANPP exhibited large spatial differences in the EASR. (3) Furthermore, the effects of PRE and SWD on annual ANPP increased from the first stage (1982–1995) to the second stage (1996–2007) with a decreasing influence of TMP, and then decreased from the second stage to the third stage (2008–2013) with an increasing influence of TMP. This study could be crucial for improving vegetation productivity predictions and understanding ecosystem-climate feedbacks.