Growth-Climate Response and Drought Reconstruction from Tree-Ring of Mongolian Pine in Hulunbuir, Northeast China
Zhongjie Shi,Lihong Xu,Linshui Dong,Jixi Gao,Xiaohui Yang,Shihai Lu,Chaoyang Feng,Jianxun Shang,Aiyun Song,Hao Guo,Xiao Zhang
DOI: https://doi.org/10.1093/jpe/rtv029
2016-01-01
Journal of Plant Ecology
Abstract:AimsDrought affected by atmosphere-ocean cycle is a dominant factor influencing tree radial growth of sandy Mongolian pine (Pinus sylvestris var. mongolica) and regional vegetation dynamics in Hulunbuir, China. However, historical droughts and its correlations with tree radial growth and atmosphere-ocean cycle in this area have been little tested.MethodsWe developed tree-ring chronologies of Mongolian pine from Hulunbuir, Inner Mongolia, China and analyzed the correlations between tree-ring width index, the normalized difference vegetation index and Palmer drought severity index (PDSI), then developed a linear model to reconstruct the drought variability from 1829 to 2009. Long-term trends and its linkages with atmosphere-ocean cycle were performed by the power spectral, wavelet and teleconnection analysis.Important FindingsThe local moisture variations affected largely the regional vegetation dynamics and tree-ring growth of Mongolia pine in the forestgrassland transition. Using tree-ring width chronology of Mongolian pine, the reconstruction explains 49.2% of PDSI variance during their common data period (1951-2005). The reconstruction gives a broad-scale regional representation of PDSI in the Hulunbuir area, with drought occurrences in the 1850s, 1900s, 1920s, mid-1930s and at the turn of the 21st century. Comparisons with other treering drought reconstructions and historical records reveal some common drought periods and drying trends in recent decades at the northern margin zones of the East Asian summer monsoon (EASM). The drying trends in these zones occurred earlier than weakening of the EASM. A REDFIT spectral analysis shows significant peaks at 7.2, 3.9, 2.7-2.8, 2.4 and 2.2 years with a 0.05 significance level, and 36.9, 18.1 and 5.0 years with 0.1 significance level. Wavelet analysis also shows similar cycles. Drought variations in the study area significantly correlated with sea surface temperatures in the western tropical Pacific Ocean and middle and northern Indian Ocean, and the Pacific Decadal Oscillation and North Atlantic Oscillation. This suggests a possible linkage with the El Nino-Southern Oscillation, the EASM and the Westerlies.