Effect Of Thinning Intensity On Carbon Storage Of Larix Gmelinii-betula Platyphylla Natural Secondary Forest In Greater Khingan Range

The forest area of Heilongjiang Province is the third in the country,and the proportion of natural secondary forests is more than half,which is due to the blind logging of early human beings.Secondary forest has seriously affected its important role as forest carbon sink due to its poor stand environment and low productivity.Natural secondary forest can improve stand quality and enhance its carbon sequestration capacity through reasonable thinning.Therefore,it is very important to carry out the research on thinning management measures of secondary forest in Heilongjiang Province.In this study,based on the measured carbon content of tree species,the effect of thinning intensity on carbon storage of Larix gmelinii-Betula platyphylla natural secondary forest in Greater Khingan Range was studied,and the prediction model of carbon content of tree species was established by near infrared spectroscopy.On the one hand,this study is helpful to better understand the relationship between the level and state of secondary forest management and carbon storage,and provide specific theoretical guidance for forest development with the purpose of carbon sink.On the other hand,near infrared technology can be used to quickly detect the carbon content of tree species and soil,and provide new ideas for forest carbon storage prediction.In this study,Larix gmelinii-Betula platyphylla natural secondary forest in Greater Khingan Range after 12 years of thinning was selected as the research object,and six thinning groups were selected,including the blank control group(CK,0%)and five thinning groups with thinning intensities of 16.7%,25.5%,34.4%,49.6% and 59.9%,combined with the measured carbon content of various organs of tree species.AHP-Entropy was used to determine the combination weight of indicators,and principal component analysis was used to comprehensively evaluate the influence of tending and thinning intensity on the total carbon storage of natural secondary forest.Finally,in the research,near infrared spectroscopy combined with Partial least squares(PLS)and Back propagation neural network(BP)were used to construct a near infrared prediction model of carbon content of tree species and a prediction model of soil carbon content,which provided a new idea for large-scale detection of carbon content of secondary forest.The main conclusions of this study are as follows:(1)By comparing diameter at breast height(DBH),tree height,crown width,individual volume and volume of larch and birch,combined with single factor analysis of variance,the effect of thinning on tree growth was analyzed.The results showed that the average DBH,tree height,crown width and average individual volume of larch and birch increased simultaneously in the 49.6% thinning intensity group.Among them,the proportion of large diameter class of larch and birch in 49.6% thinning intensity group was more than 50%.Moreover,the Shapiro-Wilk normality test of stand diameter class showed that the DBH distribution of larch,birch and stand scale of 49.6% thinning group met normal distribution.In conclusion,49.6% thinning is beneficial to better growth of trees.(2)This study analyzed the effects of thinning on carbon storage of different tree species(larch and birch)in the tree layer.The results showed that 49.6% thinning intensity significantly increased the carbon storage of individual trees and organs of larch and birch compared with CK.Moreover,thinning intensity had an effect on the distribution of carbon storage in various organs,and the effect on birch was greater than that on larch.In addition,tending thinning significantly reduced the total carbon storage of larch in the tree layer,increased the total carbon storage of birch,and generally reduced the total carbon storage of tree layer.Among them,when the tending thinning intensity was 16.7%,34.4% and 49.6%,the carbon storage of tree layer had no significant difference with that of CK group,and the carbon storage of 49.6% tending thinning intensity group was the closest to that of CK group.(3)In this study,soil physical and chemical properties and carbon storage were compared and correlation analysis,revealing the influence mechanism of thinning on soil carbon storage.The results showed that thinning intensity changed soil carbon storage(positively correlated)by affecting soil moisture content,organic carbon content,p H value,soil maximum water holding capacity,total P and total K.After 12 years of thinning,34.4% thinning intensity significantly increased soil water content and soil organic carbon content,resulting in increased soil carbon storage in 0～10 cm and 0～20 cm.(4)Seventy-eight indexes of forest growth,birch carbon storage,larch carbon storage and soil carbon storage were selected,the combination weight of indexes was determined by AHPEntropy,and the influence of thinning intensity on the total carbon storage of natural secondary forest was comprehensively evaluated by principal component analysis.AHP-Entropy method determines the first-level index combination weight,the weight of forest growth is relatively large,and the weight of soil carbon storage is relatively small.Comprehensive evaluation shows that only 49.6% thinning intensity is most suitable for carbon storage accumulation of natural secondary forest.(5)In this study,the prediction models of carbon content of tree species and soil were established by near infrared spectroscopy.Taking Larch-Betula platyphylla tree species as the research object,the near infrared model for predicting the carbon content of tree species,which was established by linear regression of partial least squares(PLS)after the whole spectrum was pretreated by the first derivative(1D),has the best performance,and the carbon storage of secondary forest tree layer calculated by using the predicted carbon content is only 1.4% lower than the actual carbon storage.Taking soil as the research object,after the whole spectrum is preprocessed by orthogonal signal correction(OSC),the near-infrared model for predicting soil carbon content established by nonlinear regression of BP Neural Network has high correlation coefficient,and the prediction accuracy can meet the determination of soil carbon content.Near-infrared spectroscopy prediction model can be used as a reliable and rapid detection method for large-area tree species and soil carbon content in secondary forest.