| Although the mid-and high latitude forests of the northern hemisphere have been repeatedly shown to be the main regions of current terrestrial carbon sink at the global scale.However,contributions of external factors still have great debate and uncertainty.One of the main reasons is that few existing carbon cycle models have ability to differentiate the effect between intrinsic factors(i.e.,stand age effect)and external factors(i.e.,atmospheric CO2 concentration,nitrogen(N)deposition,and climate change).This differentiation is important because stand age plays different roles at different stages of stand development,the effect of some external factors on stand trees growth may be disturbed by stand age.In addition,the change of forest productivity with stand age is different with various site conditions,and this difference will directly affect the accuracy of the simulation net primary productivity(NPP)results by using carbon cycle model.However,this difference has not been given sufficient attention or even overlooked in most current NPP-age relationship studies.Based on these debate and shortcomings,we selected Heilongjiang forest as the study region,the distribution of site class index(SCI)in 2010 was estimated by using the geographically weighted regression(GWR)model combined with stand factors and remote sensing factors.At the same time,the normalized burn ratio(NBR)and the differential normalized burn ratio(dNBR)were calculated to extract forest fires and classify forest fire severity during 2000-2010.According to the effects of different fire severity on stand age,the stand age was revised to make the distribution of stand age in 2010 become more accurate.On the other hand,the relationships between NPP and stand age at various SCI values for the 12 main forest types in Heilongjiang forest were established by using yield tables and forest inventory data,and we updated Integrated Terrestrial Ecosystem Carbon(InTEC)model based on these relationships and the effect of different fire severity on stand age and C pools,so as to improve the simulation mechanism of the InTEC model.Finally,the updated InTEC model was driven by stand intrinsic factors and external environmental factors to quantitative simulate and analyze the interannual variation and spatial distribution of C sources and sinks in Heilongjiang province,China from 1901 to 2010.In addition,we desighed a series of scenarios to test the relative individual and combined effects of intrinsic and external factors on the forest C sources and sinks.The main contents and conclusions are as follows:1.Quantifying the effects of different fire severities on forest age is conducive to improving the simulation method of forest fire disturbance in the C cycle model.Our results showed that,forest fires occurred mainly in Daxing'an Mountains and Xiaoxing'an Mountains areas in Heilongjiang Province during 2000-2010.The total burned area was close to 530000 hm2,in which low fire,moderate fire and high fire accounted for 60%,34%and 6%of the total burned area respectively.The burned area of Larix gmelinii and Quercus mongolica in the whole study area was larger than the other forest types,accounting for 29%and 47%of the total burned area respectively.The burned area of each forest type in different forest fire severities were:low fire>moderate fire>high fire.When low fire occurred,the age of forest types remained unchanged and the dominant tree species did not undergo succession.When middle fire occurred,only two forest types of Larix gmelinii and Betula platyphylla occured succession,and the forest age became 0.When a high fire occured,all forest types died and trees begin to grow again in the second year after fire.2.Accurate input data is the prerequisite to improve the simulation accuracy of the C cycle model.Our results indicated that,the GWR model could estimate the distribution of stand SCI well by combining with remote sensing factors and stand factors.The R2adj of estimated SCI was 0.61,and RMSE was 0.43m.The average stand age after disturbance was 49 years in 2010.3.Reasonable NPP-age relationships are the key to improve the simulation mechanism of the C cycle model.Our results showed that,these NPP-age relationships showed that forest NPP increased rapidly at the early development stage,reached a maximum around maturity and gradually declined in later ages.For the pure forest types,the NPP of coniferous forests was higher than broadleaved forests,but the age when NPP reached to the maximum was relatively later,generally between 20 and 30 years.However,the maximum value of NPP of broadleaved forest occurred earlier,generally around 10 years.For the mix forest types,the coniferous mixed forest had the strongest productivity among the three mixed forest types.In addition,a higher SCI leads to a higher NPP value,The NPP of the highest SCI could reach 2-3 times larger than that of the minimum SCI.4.The updated InTEC model was used to simulate the annual variation of forest NPP in Heilongjiang province,China from 1901 to 2010.The results showed that the NPP exhibited a significant positive trend with an increase of 31%over the study period.The highest was occurred in 1993 with a value of 464 g C m-2·a-1.After the early 1990s,there was a long-term downward trend in NPP,which continued until 2003.Since then,the NPP began gradually increase and reached 386 g C m-2·a-1 in 2010.The average NPP of the entire Heilongjiang forest area increased significantly during the 110 years,The growth was more obvious in the north region of the Xiaoxing'an Mountains and the Changbai Mountains than the other regions,meanwhile,the NPP of almost the entire Daxing'an Mountains region was higher than 400 g C·m-2·a-1.However,the NPP in the north region of the Xiaoxing'an Mountains was lower than that in the other regions.5.The fluctuation in average net biomass productivity(NBP)was relatively large,and Heilongjiang forest was a carbon source for many years before the 1930s.Since then,Heilongjiang forest was a large carbon sink for most of the year from 1930 to 2010.Similar to the trend of NPP,NBP also showed a long-term decline after 1990,with a decline rate of over 90%compared with the NPP in 1990.The NBP began to recover in recent years and reached 121 g C·m-2·a-1 in 2010.In 1901,The Daxing'an Mountains was the main carbon sinks areas,most of the Xiaoxing'an Mountains and the Changbai Mountains showed a carbon source trend.However,Heilongjiang forest is an obvious carbon sink in 2010,the Daxing'an Mountains and the south region of the Xiaoxing'an Mountains were the main carbon sinks areas,only a few area in the north region of the Xiaoxing'an Mountains showed a carbon source trend at this time.6.The total carbon stock and carbon density of forest ecosystem in Heilongjiang province showed an increasing trend during the 110 years,with an increasing range of 16%and 13%respectively.The region with the largest increasement was the south region of the Xiaoxing'an Mountains.However,the carbon pool of the ecosystem in the Daxing'an Mountains showed a decrease trend,the decrease range was 0-5 kg C-m-2.7.We designed a series of scenarios to quantitatively analysis the relative individual and combined effects of intrinsic and external factors on the NPP and NBP in Heilongjiang province,China over 1950-2010.Our analysis indicated that stand age was the dominant factor for accumulated NPP and NBP.Our results further revealed the importance of climate factors in controlling the variation of NPP and NBP.Such a pronounced impact of climate change on C dynamics,especially temperature illustrated that global warming has markedly affected mid-and high latitude forest ecosystems.The contribution of atmospheric CO2 and N deposition to forest productivity were relatively small,but it still showed that atmospheric component changes had a significant fertilization effect on stand trees growth. |
PDF Full Text Request