马尾松人工林土壤有机碳特征及其 与凋落物质量的关系

以马尾松人工林为研究对象,通过标准地调查和样品测定获取数据,利用主成分提取反映凋落物质量特征的变量并将其与土壤有机碳建立回归模型,探讨土壤有机碳与凋落物质量的相关关系。结果表明:① 林分0～100 cm土壤有机碳密度为39.76～85.85 t/hm2,随林龄的增加而增加,表现为近熟林、成熟林显著大于幼龄林、中龄林(P<0.05)。随着土壤深度增加,不同龄组土壤有机碳含量及各层10 cm厚度平均有机碳密度均逐渐减小; ② 林分凋落物总现存量(包括半分解层与未分解层)为1.54～2.25 t/hm2,随着林龄的增加而逐渐增加,但凋落物中各元素含量均无明显的变化规律。各龄组不同分解层凋落物现存量及N、P含量均表现为半分解层大于未分解层,C含量及C/N值则表现为相反的规律,未分解层平均C含量、C/N值显著大于半分解层(P<0.05),而N含量则显著低于半分解层(P<0.05),不同分解层平均P含量差异不显著(P>0.05); ③ 多元线性回归分析表明,土壤有机碳密度与反映凋落物未分解层质量特征的主成分相关关系达显著水平(P<0.05),其未分解层C含量及C/N值与土壤有机碳密度呈负相关,即凋落物未分解层C含量及C/N值越小,分解速率越快时,土壤有机碳密度增加也越快。因此,通过补种阔叶树种或施用氮肥等经营措施能促进未分解层凋落物的迅速分解,对增加土壤碳汇具有重要意义。
In order to explore the relationship between the soil organic carbon(SOC)and litter quality, we extracted the variables which reflected the litter quality characteristics by principal component analysis and set up a regression model between the litter quality and the SOC based on the standard investigation and sample measurement data of Pinus massoniana plantation. The results showed that: ① The SOC density of stand at 0-100 cm soil depth was 39.76-85.85 t/hm2, which increased with the stand age, it mean that the SOC of near-mature forest and mature forest were significantly higher than the young forest and middle-aged forest(P<0.05). The SOC content and average density of 10 cm thickness at each soil layer among different age groups decreased gradually with the increase of soil depth. ② The total existing amount of litter(including semi-decomposed and undecomposed layer)was 1.54-2.25 t/hm2, which increased with the stand age. However,element contents of litter showed no obvious variation. The existing amount of litter and the N, P content of each age group at different decomposed layers all showed that the semi-decomposed layer were higher than the undecomposed, the C content and C/N ratio showed the opposite rule. The average C content and C/N ratio of undecomposed layer were significantly higher than that in the semi-decomposed(P<0.05), and the average N content in undecomposed layer was significantly lower than that in semi-decomposed(P<0.05), the average P content of different decomposed layers had no significant difference(P>0.05). ③ The analysis of multiple linear regression between SOC and litter quality showed that the principal component variables which reflected the litter quality of undecomposed layer had a significant correlation with the soil organic carbon density(P<0.05), C content and C/N ratio were negatively correlated with organic carbon density(P<0.05), which displayed that the soil organic carbon density