黄土高原东部山区两种灌木群落的土壤碳通量研究

Soil CO_2 efflux from two shrub stands in a mountain area of the eastern Loess plateau

用动态气室法对2个灌木群落(样地1:黄刺玫+荆条Rosa xanthina+Vitex negundo;样地2:沙棘Periploca sepium)的土壤碳通量进行了连续4年(2005～2008年)的定位测定,研究灌木群落土壤碳通量的年内、年际变化与土壤温度和水分的关系.结果表明:土壤碳通量具有明显的季节变化特点,受土壤温度和水分的影响,土壤碳通量的最大值出现在土壤温度和水分均较高的月份,最低值出现在3、12月以及其它较为干旱的月份;2个灌木群落4年土壤碳通量的平均值分别为(5.20±4.02)、(3.21±2.38)、(3.48±2.48)、(2.73±1.46)μmol.m-.2s-1和(6.10±4.31)、(3.90±2.54)、(3.89±3.07)、(3.92±2.71)μmol.m-.2s-1(以CO2计),4年的总平均值为(4.02±3.08)μmol.m-.2s-1.2个灌木群落4年土壤碳通量的总平均值分别为1002和1169 g.m-.2a-1(以C计),样地间差异不显著(p=0.35).剔除土壤水分胁迫时测定的数据后,指数函数可以很好地拟合土壤碳通量与土壤温度的关系,方程的决定系数分别在0.47～0.85和0.78～0.85之间;样地1和样地2的土壤碳通量温度敏感性指数(Q10值)分别在1.88～2.73和3.61～4.28之间.土壤水分对土壤碳通量的影响在样地1较为明显.但是,对4年测定所有数据的分段分析表明,生长季土壤水分对土壤碳通量影响显著(p<0.05),非生长季土壤温度对土壤碳通量影响显著(p<0.05).用包含土壤水分和土壤温度的4个双变量关系方程可提高模型的预测能力,R2值在0.52～0.89之间.研究结果可为黄土高原东部山区灌丛土壤碳通量乃至生态系统碳平衡研究提供参考.

Soil CO2 efflux（Rs） was measured at two shrub stands（stand 1,Rosa xanthina ＋ Vitex negundo;stand 2,Periploca sepium） with a dynamic chamber（LICOR-6400-09） from 2005 to 2008.Soil temperature（Ts） at 10 cm depth and soil water content（Ws） in the top 10 cm were also measured simultaneously.The aim was to understand annual and interannual changes of Rs and its response to Ts and Ws.The results showed that the Rs exhibited pronounced seasonal variations that followed the seasonal changes in the Ts and the Ws,with higher values appearing in the months whenever both Ts and Ws were high,and lower values in March and December,or in the summer months when Ws was very low.The mean value of Rs was（5.20 ± 4.02）,（3.21 ± 2.38）,（3.48 ± 2.48）,（2.73 ± 1.46） μmol.m-.2s-1（CO2） in the shrub 1 and（6.10 ± 4.31）,（3.90 ± 2.54）,（3.89 ± 3.07）,（3.92 ± 2.71） μmol.m-.2s-1（CO2） in the shrub 2,respectively,for 2005,2006,2007 and 2008,with an overall average over the four years of（4.02 ± 3.08） μmol.m-.2s-1（CO2）.The annual CO2 efflux（Mar.to Dec.） for the two shrub stands,based on the daily-weighted monthly average of Rs,was 1002（stand 1） and 1169（stand 2） g.m-.2a-1（C）,which showed no significant difference（p = 0.35） in the CO2 efflux between the two shrub stands.An exponential function（Rs = a × ebTs） fit the data well,with the coefficients of determination ranging between 0.47 and 0.85 for stand 1,and between 0.78 and 0.85 for stand 2,when the data measured at low Ws were excluded from the analyses.The Q10 calculated from the fitted parameters rangedbetween 1.88 and 2.73 with a mean of 2.24 for stand 1,and between 3.61 and 4.28 with a mean of 4.07 for stand 2.The effect of Ws on Rs occurred only in stand 1.However,when we pooled the whole dataset from all four years together and separated it into two periods of growing and non-growing,the effect of Ws on Rs was significant in the growing season,and the effect of Ts on Rs was significant in the non-growing season.In order to describe relationships between Rs and both Ts and Ws,four two-variable linear（Rs = a ×（Ts × Ws） ＋ b and Rs = a ＋ b × Ts ＋ c × Ws） / non-linear（Rs = a × Ts b Ws c and Rs = a × ebt Ws c） models were developed.By using of those models it is possible to predict the Rs as function of both Ts and Ws,with R2 ranging from 0.52 to 0.89.Our research results bear important implications for the study of CO2 efflux as well as carbon balance in the brush mountain area of the eastern Loess plateau.