Responses of seven wetlands carbon sources and sinks to permafrost degradation in Northeast China

Purpose Warming-induced permafrost degradation is anticipated to change the global carbon cycle. We attempted to determine the effect of permafrost degradation on carbon emissions and carbon sequestration of seven wetlands in three zones of Northeast China, aiming to investigate the responses of carbon sources/sinks to permafrost degradation. Methods Three zones (permafrost zone, PZ; discontinuous permafrost zone, DPZ; and permafrost degradation zone, PDZ) were selected to represent permafrost degradation stages. In each zone, we selected seven wetlands along the moisture gradient, namely, marsh (M), thicket swamp (TS), forested swamps (alder swamp, FAS; birch swamp, FBS; and larch swamp, FLS), forested fen (larch fen, FLF), and forested bog (larch bog, FLB). We determined the annual carbon emissions of soil heterotrophic respiration from seven wetlands and the annual net carbon sequestration of vegetation, evaluated the net carbon balance by calculating the difference between annual net carbon sequestration and annual carbon emissions, and then determined the magnitude and direction of carbon-climate feedback. Results and discussion With permafrost degradation, most forested wetlands (excluding FAS in PDZ) still acted as carbon sinks in DPZ (0.30 - 1.88 t ha(-1) year(-1)) and PDZ (0.31 - 1.76 t ha(-1) year(-1)) in comparison to PZ (0.46 - 2.43 t ha(-1) year(-1)). In contrast, M and TS acted as carbon sources in DPZ (-1.72 and -0.82 t ha(-1) year(-1)) and PDZ (-2.66 and -0.98 t ha(-1) year(-1)) in comparison to PZ (-0.86 and 0.03 t ha(-1) year(-1)), this result could be attributed to the increased CO2 emissions (promoted by warmer soil temperatures) and CH4 emissions (promoted by warmer soil temperatures, higher water tables and greater thaw depths), the two significantly increased the annual carbon emissions (increased by 8.8 - 14.4% in DPZ and by 35.0 - 46.0% in PDZ), and the annual carbon emissions > the annual net carbon sequestration. Furthermore, in terms of net radiative forcing, five forested wetlands still showed negative net radiative forcing in DPZ (-6.90 to -1.10 t CO2-eq ha(-1) year(-1)) in comparison to PZ (-8.91 to -1.62 t CO2-eq ha(-1) year(-1)). In contrast, in PDZ, only FLB showed negative net radiative forcing (-6.29 t CO2-eq ha(-1) year(-1)) and significantly increased by 288.3% compared to PZ (P < 0.05), indicating an ever-increasing net cooling impact, while the other four forested wetlands all turned into positive net radiative forcing (0.84 - 53.56 t CO2-eq ha(-1) year(-1)) because of higher CH4 (CO2-eq) emissions, indicating net warming impacts. Conclusions Our results indicated that permafrost degradation affected the carbon sources/sinks of seven wetlands via different mechanisms. M and TS acted as carbon sources in both DPZ and PDZ, while permafrost degradation did not change the overall direction of the net carbon balance of five forested wetlands. Most forested wetlands (excluding FAS in PDZ) still acted as carbon sinks in both DPZ and PDZ, although there were fluctuations in carbon sink values. Moreover, despite being carbon sinks, most forested wetlands (excluding FLB) in PDZ showed positive net radiative forcing compared to DPZ and PZ (negative net radiative forcing) when using the methodology of CO2 equivalent, indicating climatic warming impacts, while FLB showed negative net radiative forcing, indicating a climatic cooling impact. Therefore, FLB should be protected as a priority in the subsequent carbon sink management practices in permafrost zones.