Abstract: Analyzing the spatial distribution pattern of forest carbon density and exploring the influencing factors can provide a scientific basis for the management of forest carbon sink. Based on forest resources inventory data in 2019, the spatial distribution of carbon density of Cunninghamia lanceolata plantations in Ji’an City of Jiangxi Province was analyzed by geostatistical method, and the influence of site, stand and meteorological factors on carbon density was quantified by structural equation model. The results showed that carbon density of Cunninghamia lanceolata plantations had significant positive spatial autocorrelation. The linear model was the best fitting semivariogram model for carbon density, with the ratio of nugget to sill being 0.81. The carbon density was mainly concentrated in the range of 20.91-51.97 t·hm^-2. The spatial distribution of carbon density was generally higher in the northern and middle region and lower in the southern region, which was not consistent with the geomorphologic characteristics of the study area. The main factors affecting carbon density were average diameter at breast height (DBH), stand density, canopy density, stand age, mean tree height, slope gradient and annual average relative humidity, with their total impact coefficients being 0.650, 0.365, 0.110, 0.090, 0.080, -0.014 and -0.036, respectively. Among them, the average DBH had the greatest direct impact on carbon density, followed by stand density. Stand age, canopy density, annual average relative humidity, and slope gradient had indirect effects on carbon density mainly by affecting the mean height of trees. Carbon density of Cunninghamia lanceolata plantations was most affected by stand factors, followed by site factors, and least affected by the meteorological factors.

Key words: Cunninghamia lanceolata plantation, carbon density, spatial distribution, influencing factor, structural equation model