【摘要】 以太原天龙山自然保护区的2个灌木群落（样地1和样地2）为研究对象，通过同步观测土壤呼吸（R_s）及土壤温度(T₁₀)、水分（W_r）、植被指数（VI），分析R_s的季节变化及其与环境因子季节变化的关系。结果表明：2个灌木群落R_s的季节变化趋势一致，最高值出现在夏季，最低值绝大多数出现在冬季，偶尔也出现在W_r较低的夏季。在季节尺度上，R_s与单一环境因子的关系为：与T₁₀为极显著；剔除水分低于1/3田间持水量的数据后，R_s与T₁₀的相关性提高；R_s与W_r的关系仅在样地1显著；R_s与VI的关系均为极显著。R_s与T₁₀和W_r的复合模型能够提高土壤呼吸的解释能力。土壤呼吸温度模型的残差与W_r的关系表明，样地1和2受干旱胁迫的W_r下限阈值分别为田间持水量的0.35和0.42。在R_s与T₁₀和W_r的复合模型中增加VI后，拟合模型的决定系数增加，赤池信息准则（AIC）值降低，模型的拟合效果得到改善。从AIC值看，3个3因素模型中温度-水分-植被指数的指数-双幂模型相对较好；3个植被指数中，比值植被指数模型相对较好。该研究结果可以为利用水热和植被因子精确估算灌丛土壤碳通量提供参考。更多还原

【Abstract】 Based on the measured soil respiration（R_s）,soil temperature(T₁₀),relative water content（W_r）and the vegetation index（VI）in two shrub communities（plot 1 and plot 2）in Tianlong mountain Nature Reserve,the relationships between R_s and the environmental factors were analyzed.The results showed that the seasonal variation in R_s in the two plots was of consistent trend,and that the highest R_s value appeared in summer,and the lowest value mostly appeared in winter,and occasionally did in summer whenever moisture was very low.On a seasonal scale,the relationship of R_s with T₁₀ was highly significant in both plots,and when the data from the measurement of less than 1/3 of water holding capacity were deleted from the analysis,their correlation improved;the relationship between R_s and W_r was significant only in the plot 1.The relationships between R_s and VI were highly significant in the both plots.The explanatory power of the synthetic relations of R_s with both T₁₀ and W_r increased.The relationship between the residual of the temperature model of R_s and W_r indicated that lower threshold value of W_r in the plot 1 and 2 was 0.35 and 0.42 of the water holding capacity.Adding a VI variable into the synthetic model of R_s with T₁₀ and W_r,the determination coefficient of the fitted model increased,and akaike information criterion（AIC）decreased to some extent.Among the three models of 3 variables,the exponential-double power function of T₁₀,W_r and VI was better than the other two models.And among the three vegetation indices,the ratio vegetation index model was better than the other two.The results can give some implications for accuracte estimation of local carbon efflux by using soil moisture,temperature and vegetation factors.更多还原