The Response Of Tree-ring Radial Growth To Climate Variability In The Hexi Corridor

With their accurate dating,high resolution,eliability and ease of access,tree-rings have become increasingly valuable in disclosing the long-term dynamics of climate in different regions of the world.The Hexi Corridor,located in a typically arid and semiarid region of northwestern China,is a sensitive area of global climate change.Qinghai spruce(Picea crassifolia Kom.)as the dominant tree specie in the Hexi Corridor,which plays an important role in the ecosystem service.However,under the stress of high temperature and drought,there were a series of negative feedbacks at the Hexi Corridor.Therefore,under the background of global warming,it is great significance to explore the relationship between tree-ring radial growth and climate variability in this area.Based on the study of the Qinghai spruce in the Hexi corridor,we collected Qinghai spruce from five sampling sites along a hydrological gradient.Firstly,this paper established five chronologies from east to west,and carried out the correlation analysis,principal component analysis,the difference analysis of chronologies statistical characteristics and correlations analysis of tree-ring chronologies with mean temperature,monthly total precipitation,relative humidity and sc PDSI after first-order difference to study the radial growth differences of the Qinghai spruce(Sabina przewalskii Kom.)along a hydrological gradient.Secondly,this paper established the regional chronology,which was derived from the above five sites of Qinghai spruce,and analyzed the correlation of the chronology with mean temperature,monthly total precipitation,relative humidity and sc PDSI after first-order difference to clarify the response of Qinghai spruce tree-ring radial growth to climate variability in the Hexi Corridor.Finally,through FFT filtering analysis,ensemble empirical mode decomposition(EEMD)and wavelet analysis,this paper compared the responses of the Qinghai spruce and the Qilian juniper(1813-2001)to climate variability at different time scales and discussed the difference of radial growth between the Qinghai spruce and the Qilian juniper.Following conclusions were drawn in this study:(1)Qinghai spruce in the west was more sensitive to climate response than the east.The correlation between the five chronologies showed that with the increasing of the difference between dry and wet sample sites,the correlation coefficients between chronologies decreased gradually.The statistical parameters of the five tree-ring chronologies showed that the trees growth at the Dahuoshao,which at the westernmost sampling site,had the maximum values of interannual variation of standard deviation(SD),mean inter-series correlation of all series(Rbar)and mean sensitivity(MS),while the first-order autocorrelation(AC1)was lowest.At the same time,the values of interannual variation of SD,MS for the chronology at the Pingqiangkou,which at the easternmost sampling site,were clearly smaller than the other chronologies,while the AC1 is highest.It is proved that the chronology of the western region contained more abundant information than that of the eastern region and more sensitivity to climate.(2)Drought in the previous of the growing season and early growing season(March to June)was the main limiting factor for the growth of the Qinghai spruce.Correlations of the five Qinghai spruce tree-ring chronologies with climate data at the region after first-order difference showed that the control factors of Qinghai spruce tree growth in different dry and wet regions were similar,there was a significant positive correlation with precipitation,relative humidity and sc PDSI in the previous of the growing season and early growing season(0.18?r?0.6,P<0.05).The correlation between the region chronology and climate data after first-order difference also showed that there was a significant positive correlation with precipitation,relative humidity and sc PDSI in March to June of current year(0.26?r?0.51,P<0.05).(3)In terms of the responses in different time scales,there was a strong correlation between the interannual variation trend of the Qinghai spruce and the Qilian juniper,the interdecadal and above variation trends were weak.Five years high-pass filters,first order difference contrast and fifty years low-pass filters were used to compare the PC1 of the two species in the past 189 years.The results showed that there was a strong correlation between the interannual variation trend of the Qinghai spruce and the Qilian juniper(r=0.72,P<0.01),but the interdecadal and above variation trends were weak(r=0.11,P>0.01).EEMD analysis also showed that the correlation coefficients between the Qinghai spruce and Qilian juniper on the interdecadal cycle and above were low(r=0.26,P>0.01;r=-0.19,P>0.01),the interannual component cycle of the Qinghai spruce was similar to the interannual component cycle of the Qilian juniper(r=0.66,p<0.01).(4)At the 95% confidence level,there were obvious differences in the periodic changes of the Qinghai spruce and Qilian juniper.Qilian juniper showed a strong amplitude which associated with the 60-64 year periods from 1800 to 1940,while the short-lived Qinghai spruce not shown.Wavelet analysis showed that Qinghai spruce had a weaker amplitude response to the environment variations,mainly occured at 2?4years and 10?18 years periods.Qilian juniper showed a strong amplitude associated with the 60?64 years period from 1800 to 1940,while the Qinghai spruce not shown.In conclusion,drought in the previous of the growing season and early growing season(March to June)was the main limiting factor for the growth of the Qinghai spruce.The chronology of the western region contained more abundant information than that of the eastern region and more sensitive to climate.Besides,there was a strong correlation between the interannual variation trends of the Qinghai spruce and Qilian juniper,but the interdecadal and above variation trends were weak.No evidence showed that tree-ring width chronology of short-lived Qinghai spruce recorded the60?64 years period at the 95% confidence level,and it is also reminds us that we should be cautious when using short-lived trees such as spruce in response to multi-decadal changes in future studies.