AB
Fig.3 Contribution of CO2 flux at deep layers(80-200cm)to the total CO2 flux of soil-atmosphere interface in different use-land types (crop land, CH10a, CH20a,CH30a,CH40a). Violin plot represents proportion of CO2 flux at deep profile (80-200cm) to the total flux at soil-atmosphere interface in growing seasons and non-growing seasons (A ) . Box plot represents proportion of CO2flux at deep profile (80-200cm) to the total flux at soil-atmosphere interface in whole season (B ). Columnar plot represents the accumulated variance contribution of soil CO2 flux at deep profile (80-200cm) to the total flux at soil-atmosphere interface in whole season of different land- use type (B ).
Fig.4 The content of soil organic carbon(g kg-1) at different soil profiles depth(0-5cm,5-20cm,20-80cm,80-140cm,140-200cm) in crop land andRobinia pseudoacacia of four kinds of stand age(CH10a, CH20a, CH30a, CH40a). Values ​​are means of three replicates (n=3) and error bars are standard errors of means. The letters in a column represent a significant difference(p<0.01) between means (n=3) of different soil profiles depth within a land-use type.
Fig.5 Regression parameters and statistics equations for index model between soil CO2 flux (F) and soil temperature (T) in different profiles.The parameters (mean) are defined in Eqs. (F=a×ebT). Q10, the proportional increase in soil CO2 flux with a 10℃ increase in temperature,Q10=e10b, where, b is parameter in Eqs. (F=a×ebT); R2, coefficient of determination; P-value, parameter testing regression significant.** Significant at the 0.01 probability level.
Fig.6 Regression parameters and statistics equations for quadratic Objective function model between soil CO2 flux (F) and soil moisture (M) in different profiles.The parameters (mean) are defined in Eqs. ((F=aM2+bM+c)). R2, coefficient of determination; P-value, parameter testing regression significant. * Significant at the 0.05 probability level,** Significant at the 0.01 probability level.
Fig.7 Correlation analysis between soil CO2flux, temperature, moisture, and soil organic carbon in Shallow layer (5-20cm) (Heat map: A ). Correlation analysis between soil CO2 flux, temperature, moisture, and soil organic carbon in layer(80-200cm) (Heat map: B ). P-value, parameter testing correlation significant.
* Significant at the 0.05 probability level.
** Significant at the 0.01 probability level.
Fig.8 The CO2 flux at different soil temperature, moisture, and organic carbon . 3D-scatter diagram(a) presents the CO2 flux in shallow layers (5-20cm). 3D-scatter diagram (b) presents the CO2 flux in deep layers(80-200cm).
Fig.9 Observed and modeled CO2 flux in different ecosystems with the T&M&C-model. (a) , (b) ,(c) , (d) , (e) and (f) represent temperate forests,subtropical forests,neotropical rain forests,mediterranean arable land,fir plantation,coniferous and broad-leaved forests, respectively; statistics are shown inAppendix A .
Fig.10 observed and modeled CO2 flux in different ecosystems with the T&M-model. (a) , (b) ,(c) , (d) , (e) and (f) represent temperate forests,subtropical forests,neotropical rain forests,mediterranean arable land,fir plantation,coniferous and broad-leaved forests, respectively; statistics are shown inAppendix A .
Fig.11 Contribution rates of T&M&C model to the CO2 flux estimation improvement compared with T&M model.
Table 1 Description of sample plot used in the study.