Study On Individual Tree Growth Model Of Dominant Trees In Chinese Fir Plantations Based On Climate Change

With the aggravation of global climate change,exploring the response of forest growth to climate change can provide technical support for sustainable forest management.The growth status of dominant wood is the embodiment of stand production potential,and it is generally believed that the growth of dominant trees is less affected by stand density and thinning strength.Therefore,taking dominant tree as the research object can better analyze the relationship between the growth of single wood and climate.This study took dominant trees of Cunninghamia lanceolate plantation in Jiangxi province as the research object,three models(random forest,the boosted regression trees and Cubist)and date of dominant trees and 10 climate factors extracted from Climate AP were used to construct the growth models of DBH,height and volume of individual tree with climate factors.In this study,the prediction of the growth of indivitual trees in the Chinese fir plantations under three future climate scenarios(RCP 2.6,RCP 4.5 and RCP 8.5)from 2010 to 2060provided a reference for the sustainable management and management of Chinese fir plantation under climate change.The main results and conclusions were as follows:(1)The determination coefficient(R~2)of the growth model of DBH and height constructed by three machine learning algorithms reached above 0.80,and the R~2 of the volume growth model all reached above 0.55.The fitting accuracy and prediction effect of the growth models of DBH,height and volume were shown as:the boosted regression trees>Cubist>random forest.The boosted regression tree models of optimal parameter combination were:n.trees=10000,shrinkage=0.01 and interactive.depth=9.Based on cross validation,the optimal growth model of DBH was shown as follows:the root mean square error(RMSE)=1.3960 cm,average absolute error(MAE)=1.0846 cm;the optimal growth model of height was shown as follows:RMSE=1.1884 m,MAE=0.8973 m;the optimal growth model of volume was shown as follows:RMSE=0.0384 m~3 and MAE=0.0215 m~3.The application of the boosted regression tree model could better predict the growth law of dominant trees in Chinese fir plantation.(2)The calculation results of relative contribution rate of independent variables of the three models all showed that age contributed the most to the growth of Chinese fir,and temperature had a greater influence on the growth of Chinese fir than precipitation.For the optimal growth model of DBH,the boosted regression tree model could be obtained as follows:Age>GSAT>GSMT>MAT>GSXP>GSMP>TD>GSAP>GSXT>AHM>MAP;for the optimal growth model of height,the boosted regression tree model could be obtained as follows:Age>MAT>TD>GSMT>GSXP>GSAP>GSAT>GSXT>GSMP>MAP>AHM;for the optimal growth model of volume,the boosted regression tree model could be obtained as follows:Age>TD>GSMT>MAT>GSXP>GSAP>GSAT>GSXT>GSMP>MAP>AHM.(3)At the same age,compared with the current climatic conditions,the growth of Chinese fir’s DBH,height and volume all increased in the future climate.In the climate scenario of RCP 8.5,the growth performance of DBH and height were the best,followed by RCP 4.5,while RCP 2.6 was the worst.In the climate scenario of RCP 4.5,the growth performance of volume was the best,followed by RCP 2.6,and RCP 8.5 was the worst.In Chinese fir 0-10 a,climate change had little effect on the growth of Chinese fir.After 10 a,climate change had a promoting effect on the growth.Chinese fir grew rapidly and began to slow down after 30-35 a.At 30 a,the total growth of DBH was 1.30 times of the current total growth under the climate scenario of RCP 8.5,and then the growth was slow.At 35 a,the total growth of height was 1.50 times of the current total growth amount under the climate scenario of RCP 8.5.At 30 a,the total growth of volume was 2.37 times of the current total growth under the climate scenario of RCP 4.5,and 40 a reached the maturity age of volume.