Selection Of Strip Harvesting Width And Fertilization Measures For Moso Bamboo Based On Carbon Stock

Moso bamboo（Phyllostachys pubescens）is an important non-timb er forest resource in southern China.Due to its typical uneven-aged stand structure,artificial selective cutting is usually used for bamboo harvesting and stand structure adjustment.This method has high costs and low effici ency,limiting the further development of the Moso bamboo industry.In th is context,strip cutting has attracted widespread attention for its low cost,high efficiency,and easy mechanization.This study took the Moso bambo o forests in Shangping Township,Yong’an City,Fujian Province as the res earch object and set up five standard plots with the same fertilization treat ment and different cutting widths（3m,6m,9m,12m,15m）,and 27 standa rd plots with the same width（9m）and different fertilization treatments.Fe rtilization tests were carried out with three ratios of N:P₂O₅:K₂O（2:1:1.5,3:1:2,5:1:2.5,A1,A2,A3）and three fertilization rates（30,40,50 kg/mu,B1,B2,B3）.Using unmanned aerial vehicle（UAV）remote sensing techn ology,the above-ground carbon storage and soil organic carbon storage of Moso bamboo forests were obtained.The strip cutting width and fertilizat ion measures were selected based on the above-ground carbon storage,soi l organic carbon storage,and total carbon storage two years after strip cutt ing.The main research conclusions are as follows:（1）The study uses power function,exponential function,and polyno mial function to construct models for the biomass of individual Moso bam boo plants and their various organs.Among them,the power function has the best accuracy.The expressions for each organ model are:W_individual=0.201×D^1.9524,W_pole=0.113×D^2.0971,W_canopy=0.1172×D^1.5,W_branch=-0.0045x²+0.4374D-1.2885,W_leaf=0.0206（DH2）^0.5543.Their R² values are 0.86,0.84,0.71,0.59,and 0.48;and the overall accuracy（TA）values are 0.81,0.78,0.79,0.68,and 0.57.（2）Referring to Zhou Guomo’s research,using 0.50 as the conversio n factor for aboveground biomass to aboveground carbon storage,the abo veground carbon storage of the standard plots is obtained.Combining dro ne remote sensing data,multiple linear regression models,random forest,BP neural network,and Light GBM are constructed as three machine learn ing models.After model testing,the multiple linear regression model has t he highest accuracy,with R²,RMSEr,and overall accuracy（TA）of 0.76,0.29,and 0.75,respectively.（3）The exposure of the forest land during the recovery period of strip-cut Moso bamboo forests provides a basis for obtaining direct spectral information of the forest land and constructing a soil organic carbon inversion model.Based on this,the study uses the spectral features of bare land points and the measured soil organic carbon storage to construct multiple linear regression models and random forest,BP neural network,and Light GBM as three machine learning models.After model testing,the BP neural network regression model has the highest accuracy,with R²,RMSEr,and overall accuracy（TA）of 0.43,0.29,and 0.75,respectively.（4）Analysis of aboveground carbon stock and soil organic carbon stock under the same fertilization treatment with different logging widths.After a two-year recovery period,the aboveground carbon stock varied between 284.79 kg/mu and 748.97 kg/mu for different logging widths,with the specific order being 9m>6m>3m>CG>12m>15m.Following strip logging,the soil organic carbon stock showed a trend of initial decline followed by an increase.Compared to pre-logging levels,the soil carbon stock recovery rate in the second year after logging ranged from 84.82%to94.52%,with the specific order being 12m>6m>9m>15m>6m.Considering both aboveground carbon stock and soil organic carbon stock,the net carbon accumulation ranged from 295.43 kg/mu to 531.24 kg/mu,with the specific order being 9m>6m>12m>CG>15m>3m.（5）Analysis of aboveground carbon stock and soil organic carbon sto ck under the same logging width but different fertilization treatments.Aft er a two-year recovery period,the aboveground carbon stock varied betwe en 693.04 kg/mu and 843.17 kg/mu for different fertilization treatments,w ith the specific order being F7>F5>F8>F9>F6>CG>F3>F4>F1>F2.Compared to pre-logging levels,the soil carbon stock recovery rate in the second year after logging ranged from 89.24%to 119.82%,with th e specific order being F9>F6>F8>F5>F1>F2>F4>F7>F3.Consi dering both aboveground carbon stock and soil organic carbon stock,the n et carbon accumulation ranged from 275.08 kg/mu to 1240.96 kg/mu,wit h the specific order being F5>F9>F8>F6>CG>F2>F3>F7>F1>F4.In conclusion,after two years of strip logging,adopting a manageme nt strategy with a logging width of 9 meters,an N:P₂O5:K₂O ratio of 3:1:2,and a fertilization rate of 40 kg/mu can promote an increase in abovegr ound carbon stock and soil carbon stock.