Carbon Dynamics During Decomposition Of Coarse Woody Debris And Influencing Factors For Eleven Temperate Tree Species

Coarse woody debris (CWD) is a conspicuous structural and functional component of forest ecosystems. In spite of its large variability caused by forest types, successional stage, management history, etc., the CWD plays an important role in forest carbon (C) and nutrient cycling, and biodiversity maintenance. It is a major C pool in forest ecosystem and carbon source to the atmosphere because it accounts for the majority of forest floor and the C loss during CWD decomposition is mainly released as CO2. The forest in northeastern China is sensitive to climate change, and plays a crucial role in national and regional carbon balance. However, the contribution of CWD to forest carbon budgets and its responses to environmental change are poorly understood. In this study, an in situ long-term tracking experiment was conducted to comprehensively study dynamics of C, nitrogen (N), and structural compositions of CWD during its decomposition for 11 temperate tree species. The objective was to comparatively study the contribution of CWD decomposition and controlling mechanisms for the species. The tree species examined included Betula platyphylla, Populus davidiana, Tilia amurensis, Juglans mandshurica, Quercus mongolica, Acer mono, Ulmus japonica, Pinus koraiensis, Phellodendron amurense, Larix gmelinii, and Fraxinus mandshurica. The results were summarized as the following aspects:(1) Spatiotemporal dynamics of CO2 flux released from CWD decomposition (RCWD) and the relationship with environmental factors. The RCWD of the 11 tree species at four sites with various environmental conditions (Mongolian oak forest, hardwood forest, Korean pine plantation and open field) in northeastern China was measured with an infrared gas exchange analyzer (LI-6400 IRGA) in order to examine temporal dynamics in RCWD and response of RCWD to CWD temperature at 10 cm depth (TCWD) and water content (WCWD) from May to October 2008 and 2009. The results indicated that the diurnal variation of RCWD showed an overall bell-shaped curve during the growing season examined except for July and August, mainly driven by TCWD.The maximum RCWD occurred between 13:00 and 15:00, substantially delaying to the maximum daily air temperature (TA). In July and August, the diurnal changes in RCWD displayed non-peak or multi-peak patterns, less responding to the temperature change. The mean RCWD during the daytime was greater than that during the nighttime for all tree species. The RCWD was positively correlated with TCWD and TA (P<0.05). However, the RCWD was more significantly related with the TA at 2 hours before the measurement, rather than the instant TA, implying a hysteresis response of RCWD to the TA. The temperature coefficient of RCWD (Q10) changed with tree species and seasons. The Q10 values varied from 1.74 for B. platyphylla to 4.20 for Q. mongolica, tending to decrease with TCWD rising. The seasonality of RCWD showed an overall bell-shaped curve for all tree species with a maximum value in July and August, coinciding with the seasonal variation pattern of TCWD. The WCWD had less contribution to the seasonal changes in RCWD.The coefficient of variation (CV) of RCWD presented an opposite trend to that of TCWD, decreasing gradually with temperature increasing.The R15, defined as normalized RCWD to TCWD at 15℃, was significantly affected by tree species (P<0.001), of which the average value was the greatest for B. platyphylla and the least for P. amurense, and the former was 4.93 and 6.04 times as much as the latter in 2008 and 2009, respectivley. Overall, the R15 for the broad-leaved species was significantly higher than that for the coniferous species. The mass-based R15 (R15/g) was positively correlated with WCWD (r=0.76). However, the fitness of the correlation varied with species; and the softwood species with a less wood density had a better fitting than the hardwood and conifer species with a greater density. The responses of RCWD to TCWD were significantly different among the four sites (P<0.001). These responses were significantly greater in the Korean pine plantation and hardwood forest than in the Mongolian oak forest and open field.The integrated CO2 emission from CWD decomposition during the growing season (RG) fluctuated between 2.59 gC·m-2·growing-season-1 and 20.94 gC·m-2·growing-season-1, showing a similar variation trend to the R15. The estimates using different time-steps differed significantly. The daily time-step fitting between RCWD and TA relationship (R2=0.47) was better than that the growing season time-step fitting (R2=0.16). The RG estimated by the model with daily time step was 5.4%-69.9% greater than that by the model with the whole growing-season time-step. It is suggested that accurate estimation of CO2 emission from CWD should take the temporal and inter-specific variations in temperature sensitivity of RCWD and time steps of modeling into account.(2) Losses of C and N during CWD decomposition and influencing factors. The C concentration of CWD (Cc) did not change significantly for all species at the early stage of CWD decomposition (P>0.05). The CWD mass (M), C density (Cd), N concentration (Nc) and N density (Nd) decreased with the decomposition proceeding, whereas the C/N ratio increased. These parameters were significantly different among the species (P<0.001). The coniferous species had significantly lower decomposition rate than the broad-leaved species. There was a negative correlation between CWD size and decay rate. The mass loss, C and N release from CWD decomposition were positively correlated with the initial N content, but were negatively correlated with the initial C/N ratio. The changes of M, Cc, Cd, Nc and Nd were not significantly different among the sites (P>0.05), and showed similar trends as the decomposition proceeded. It is concluded that the CWD tended to be C and N sources during the first three years of the decomposition. (3) Changes of CWD structural compositions in relation with CO2 emission. The lignin and holocellulose contents of CWD were significantly different among the species (P< 0.001), which the softwood tree species had the minima. The ratio of lignin to N (Lc/N) ranked as: softwood tree species> hardwood tree species> coniferous species. Lignin concentration (Lc) increased during the first three years of decomposition for all the species (except for T. amurensis and U.japonica), but the changes were insignificant (P> 0.05). The holocellulose (Hd) and the lignin density (Ld) decreased to some degrees (P< 0.05) with the minimum loss in the conifer species and the maximum loss in the softwood species. The Lc/N increased during the decomposition for all species. After 3 years decomposition, R15 deceased for all the species except for L. gmelinii, A.mono and F. mandshurica. Overall, R15 of softwood species decreased 32.1%and that of coniferous species increased 23.3%. Additional, the temperature coefficient for coniferous species increased and that for broad-leaved species unchanged. R15 was positively correlated with Hc but negative correlation with LC and LC/N. The changes of R15 were no correlated with the changes of structural compositions and the structural compositions had little effect on Q10. The results indicated that the changes of structural compositions were not the major factors influencing the changes of RCWD.