采用Biolog-ECO 微平板技术、皮尔逊(Pearson)相关性分析及典范对应分析(CCA)等方法对福建省三明市小湖村的格氏栲(Castanopsis kawakamii Hay.)天然林和人工林土壤的生态化学计量特征、微生物群落特征及两者之间的关联性进行了比较研究。结果表明:天然林土壤的有机碳含量、全氮含量、C/ P 比、C/ K 比和N/ K 比以及土壤微生物群落的Shannon-Wiener 指数和McIntosh 指数均显著高于人工林;而2 个林分间土壤的全磷含量、全钾含量、容重、含水量、分形维数、C/ N 比、N/ P 比和P/ K 比以及土壤微生物群落的Simpson 指数差异不显著。用31 个碳源或不同类型碳源进行培养,随培养时间延长2 个林分土壤微生物群落的平均颜色变化率(AWCD)均逐渐升高;在31 个碳源及碳水化合物、羧酸、多聚物和酚酸4 类碳源中,天然林土壤微生物群落的AWCD 值均高于人工林;总体上,天然林土壤微生物群落对碳源的利用率高于人工林。相关性分析结果表明:格氏栲林土壤微生物群落对碳水化合物的利用率与土壤的N/ P 比和N/ K 比呈极显著正相关,与土壤的C/ P 比和C/ K 比呈显著正相关,与土壤分形维数呈显著负相关;对羧酸的利用率与土壤的N/ P 比和N/ K 比呈显著正相关;对多聚物的利用率与土壤的C/ P 比和C/ K 比呈极显著正相关。CCA 分析结果表明:格氏栲林土壤微生物群落对碳水化合物的利用率与土壤的C/ P比、C/ K 比、N/ P 比和N/ K 比的相关性均较强,对羧酸的利用率与土壤的N/ P 比和N/ K 比的相关性也较强;此外,而5 个天然林样地主要分布在CCA 排序图的第3 和第4 象限,而5 个人工林样地则主要分布在第1、第2 和第4 象限,表明天然林样地的土壤养分比例协调,而人工林样地的土壤肥力偏差、土壤结构不稳定。研究结果显示:格氏栲天然林的土壤生态系统优于人工林,建议适当减少对人工林的人为干扰,以促进林下土壤养分和微生物群落的良性发展。

Using methods of Biolog-ECO micro-plate technique, Pearson correlation analysis and canonical correspondence analysis ( CCA), etc, ecological stoichiometric characteristics, microbial community characteristics and their relevance in soil of natural and artificial forests of Castanopsis kawakamii Hay. at Xiaohu Village in Sanming City at Fujian Province were studied. The results show that organic carbon and total nitrogen contents, C/ P, C/ K and N/ K ratios in soil and Shannon-Wiener and McIntosh indexes of soil microbial community of natural forest all are significantly higher than those of artificial forest, while differences in total phosphorus and total potassium contents, bulk density, water content, fractal dimension and C/ N, N/ P and P/ K ratios of soil and Simpson index of soil microbial community between two forest types are not significant. Using 31 carbon sources or different types of carbon sources for culturing, average well color development (AWCD) of soil microbial community of two forest types both increases gradually with prolonging of culture time. In 31 carbon sources and four types of carbon sources including carbohydrates, carboxylic acids, polymers and phenolic acids, AWCD values of soil microbial community of natural forest all are higher than those of artificial forest. Overall,utilization of soil microbial community of natural forest to carbon source is higher than that of artificial forest. The correlation analysis result shows that utilization of soil microbial community of C. kawakamii forest to carbohydrates appears an extremely significantly positive correlation with N/ P and N/ K ratios of soil, a significantly positive correlation with C/ P and C/ K ratios of soil, a significantly negative correlation with soil fractal dimension. Utilization to carboxylic acids appears a significantly positive correlation with N/ P and N/ K ratios of soil. Utilization to polymers appears an extremely significantly positive correlation with C/ P and C/ K ratios of soil. CCA analysis result shows that correlations of utilization of soil microbial community of C. kawakamii forest to carbohydrates with C/ P, C/ K, N/ P and N/ K ratios of soil all are stronger, those of its utilization to carboxylic acids with N/ P and N/ K ratios of soil are also stronger. Besides, five natural forest plots mainly distribute in the third and the fourth quadrants of CCA ordination diagram, while five artificial forest plots mainly distribute in the first, the second and the fourth quadrants, meaning that soil nutrient proportion of C. kawakamii natural forest is coordinate, while soil fertility of artificial forest is deviate and its soil structure is instability. The research results indicate that soil ecosystem of C. kawakamii natural forest is better than that of artificial forest, it is suggested that it is necessary to appropriately reduce the artificial interference to artificial forest, in order to promote better development of nutrients and microbial community in forest soil.