中文摘要：

土壤呼吸有着明显的时间变异，呈现小时、日、季节、年际的变化，对土壤呼吸制定最佳的监测频率非常必要．为选取土壤呼吸的最佳监测频率，以太岳山油松人工林为研究对象，于2011-2012年5—10月测定土壤呼吸速率的动态变化，以每周2次（天）的观测数据作为长期定位观测土壤呼吸速率的数据（CK），依此区划为不同监测频率：每周1次（SF2）、每两周1次（SF4）、每月1次（SF8），并分析不同监测频率的土壤呼吸速率与对照相比产生的偏差以及对生长季土壤呼吸速率值的估计概率．结果显示：油松人工林土壤呼吸速率季节变化明显，9：00．11：00土壤呼吸值与24h日平均土壤呼吸值相关性最大（R2=0．978）．降低监测频率使土壤呼吸速率与土壤温湿度的拟合效果（R2）以及土壤呼吸速率对温度的敏感性指数Q10的波动范围呈扩大趋势．随着监测频率的降低，所测的土壤呼吸速率与对照的离散程度呈增大趋势．在给定精度下，随着监测频率的降低，所测的土壤呼吸速率对生长季土壤呼吸速率值估计概率呈下降趋势，2011、2012、2011-2012年SF2在5％精度下估计概率均为100％，而SF4在5％的精度下估计概率分别为50％、75％、100％，SF8则在5％的精度下估计概率依次为50％、62．5％、87．5％．研究表明，整个生长季可使用9：00．10：00作为土壤呼吸速率测定代表性时段；对于单年观测期每周1次（天）的监测频率可以得到可靠的生长季土壤呼吸估算量（与对照相比偏差〈5％），而对于两年观测期每两周1次（天）的监测频率可以得到可靠的生长季土壤呼吸估算量（与对照相比偏差〈5％）．本研究结果为准确测定土壤呼吸并减少工作量提供了方法和依据．

英文摘要：

Soil respiration is the major path through which CO2 fixed by vegetation returns to the atmosphere, therefore plays an important role in the carbon cycle of terrestrial ecosystems. Soil respiration exhibit pronounced hourly, daily, seasonal and interannual variability, which is controlled by temperature, moisture and substrate supply. Optimizing a manual sampling frequency for estimating the growth season soil respiration （Rs）is an important but unresolved issue. This study measured the diurnal, seasonal and interannual dynamics of soil respiration rate in the plantation forest of P. tabulaeformis during the growth season in 2011-2012. A sampling frequency of twice per week was set as a long-term soil respiration observed data （CK）, then divided into different sampling frequencies： weekly （SF2）, biweekly （SF4）, and once every month （SFS）. The differences between different sampling frequencies in the deviation of annual soil respiration and the probabilities of obtaining estimates of soil respiration at a given precision were analyzed. The results showed a strong seasonal dynamic pattern in soil respiration rate in the plantation forest of P.tabulaeformis, with Rs during 9：00-11：00 a.m. significantly correlated with diurnal mean Rs （R2=0.978）. Decreasing measuring frequency expanded the fluctuation ranges of the fitting effect （Rs） of correlation model between soil respiration rate and soil temperature, humidity and the temperature sensitivity of soil respiration Q~0. The variances associated with the spread of % deviations observed increased with decreasing sampling frequency, but the probability of obtaining an estimate of annual Rs at a given precision showed an opposite trend. As the estimates rates of annual Rs within ＋5% of CK with a 100% probability, the probability estimate of SF2, SF4, SF8 were 100%, 50%, 50% in 2011, 100%, 75%, 62.5% in 2012, and 100%, 100%, 87.5% in 2011-2012, respectively and orderly. We therefore concluded that Rs during 9：00- 10：00 am could be