Recently, forest carbon (C) budgets have been significantly affected by climate variability, nitrogen (N) deposition, an increasing global atmospheric CO2 concentration, and disturbances (
Forests are the primary component of terrestrial ecosystems and store 85% of terrestrial biomass. Forest ecosystems play a vital role in the global carbon (C) cycle. Large differences in the net C exchange between forests and the atmosphere can result from small shifts in the balance between photosynthesis and ecosystem respiration (
Previous studies show that two aspects strongly influence regional C budgets. First, CO2 fertilization, climate variability, and nitrogen (N) deposition affect C budgets because these factors determine the length of the growing season and the rates of photosynthesis and heterotrophic respiration. Second, disturbances (
Biogeochemical models are an effective method to understand the response mechanisms of C cycling processes involving different environmental variables. For example,
The Integrated Terrestrial Ecosystem Carbon (InTEC) model is a process-based biogeochemical model produced by
According to many studies, a massive sink of atmospheric CO2 is unaccounted for in the budgets between intermediate and high latitudes in the Northern hemisphere (
Based on these uncertainties and inadequacies, the goal of this study was to integrate the effects of disturbance and nondisturbance factors using the InTEC model to simulate annual C budgets from 1901 to 2013 in the forest regions of Heilongjiang Province, which is one of the areas with the most abundant forest resources in China. Furthermore, we analyzed changes in spatiotemporal patterns from 1901 to 2013, and then discussed possible reasons and mechanisms to explain the results. The results of this study will provide a new perspective on the net changes in C and the underlying drivers for the changes, which will reduce current uncertainties concerning terrestrial C cycle processes and improve future forest management strategies.
The InTEC model simulated an annual NPP value for each pixel from the initial year (1901) combining the effects of disturbance and nondisturbance factors. Disturbances are particularly important for a C balance because they typically affect processes that release C into the atmosphere. Changes in forest stand age structure, or the normalized forest productivity, were used to determine the effect of disturbances on NPP. In the InTEC model, three assumptions concerned disturbance (
For the nondisturbance factors, the effect of climate on NPP was expressed through modifications in growing season length and direct effects on the photosynthesis rate. As the source of intercellular CO2 content, atmospheric CO2 concentration affects photosynthesis. The link is direct between N and the mechanisms of C decomposition, and the proportion of C and N in foliage and fine root C pools has a negative influence on C decomposition. In this study, soil texture was used to determine the parameters of soil moisture content and soil temperature. Forest stand age was used to determine the timing of the last stand-replacing disturbance.
The NPP for each pixel of a region in any year was calculated from the initial value of NPP (
The
The C dynamics of the entire forest system were divided into living biomass and soil (nonliving biomass) C pools. Living biomass C pools included four individual components (
where
For the initial values of living biomass and soil C pools, an assumption of “equilibrium age” was set, which assumed that C and N exchanges between terrestrial ecosystems and the atmosphere were in equilibrium for stands at equilibrium age under the mean conditions of climate, CO2 concentration and N deposition in the preindustrialization period such that (
where
NPP is equal to the difference between gross primary production (GPP) and autotrophic respiration (
NEP was used to represent carbon dynamics or balances to identify carbon sinks or sources in the forests of Heilongjiang, with positive NEP values indicating C sinks and negative values indicating sources of C to the atmosphere.
The InTEC model has been validated and applied widely. For instance,
Because of the lack of historical data on fires, harvests and insect disasters, these disturbance factors were not differentiated for the entire simulation period, and therefore all the disturbance factors were treated as fire.
To drive the InTEC model, a series of data sets were produced in this study. All spatial data were employed in the UTM projection and WGS-84 coordinate system and interpolated to 1 km resolution (0.0136° × 0.0089°).
For Heilongjiang, China, from 1901 to 2013, the 0.5° global data set interpolated by the UK Climate Research Unit (http://www.cru.uea.ac.uk/cru/data/) provided monthly mean temperature, relative humidity, and total precipitation. The data set was produced by measurements at available stations of the National Meteorological Administration in China. Monthly solar irradiance data for the period before 1948 were produced using the Bristow-Campbell model derived from historical temperature, humidity, and precipitation data. For the period from 1948 to 2013, monthly solar irradiance data from the T62 Gaussian reanalysis data of the US National Center for Atmospheric Research (NCAR) were used (http://www.esrl.noaa.gov).
The annual atmospheric CO2 concentrations from 1958 to 2013 were from the data set obtained from the Mauna Loa Observatory (20° N, 156° W - http://cdiac.esd.ornl.gov/ftp/trends/co2/maunaloa.co2). The pre-1958 CO2 concentration was estimated based on the CGCM2 (
Spatial N deposition data in 1993 were obtained from a data set that was simulated based on a chemical transport model (TM3). This data set included the predicted value of N deposition in 1860, 1993, and 2050, and the spatial resolution was 5° in longitude and 3.75° in latitude. Spatial N deposition from 1901 to 2013 (except 1993) was calculated using the equation based on historical greenhouse gas emissions and the average N deposition data in 1993 (
where
A map of forest stand age in 2005 was produced from the 7th National Forest Inventory data recorded from 2004 to 2008 in Heilongjiang Province (
The forest-type map of Heilongjiang in 2006 at a 1 km resolution was obtained from the Northeast Institute of Geography and Agroecology, Chinese Academy of Sciences, published in 2007 (
Physical properties of the soil included the field capacity of soil water, depth of soil, wilt point, and fractions of clay, silt and sand. These parameters were used to estimate soil moisture content and soil temperature, which were further used to estimate soil heterotrophic respiration. The field capacity of soil water and wilt point were derived from the International Geosphere-biosphere Program, Global Gridded Surfaces of Selected Soil Characteristics. Soil depth was derived from the global soil texture data set from the Oak Ridge National Laboratory Distributed Active Archive Center, TN, USA. The fractions of clay, silt, and sand were obtained from the Harmonized World Soil Database (HWSD) constructed by the Food and Agriculture Organization of the United Nations (FAO) and the International Institute for Applied Systems Analysis (IIASA).
The reference-year NPP of 2003 in Heilongjiang forests was obtained using the BEPS based on land cover, LAI maps derived from MOD15A2 products at 8-day intervals, soil texture, and daily meteorological data. The reference-year NPP data were used to calibrate the initial value of NPP.
A total of 11 NPP-age curves were used in the InTEC model. The relationships between NPP and age were established using yield tables for Heilongjiang (
where the parameters
The average NPP (
Heilongjiang was divided into three ecoregions to analyze the distribution of average NPP. The ecoregions were the Daxing’an Mountains, Xiaoxing’an Mountains, and Changbai Mountains within Heilongjiang. The highest average NPP was in the Daxing’an Mountains, and the increment was 37% from 1901 to 2013. The average NPP of the Xiaoxing’an Mountains was the lowest; however, this ecoregion had the largest increment of 45% from 1901 to 2013. The increment of average NPP in the Changbai Mountains was 39% from 1901 to 2013.
In this paper, we primarily analyzed the spatial distribution of the average NPP of the three ecoregions in 1901, 1950, and 2013 (
The range of average NEP in Heilongjiang was relatively large (
Similar to NPP, we analyzed the spatial distribution of the average NEP of three ecoregions in 1901, 1950, and 2013 (
This study showed that Heilongjiang was a large C sink after 2000, which is in agreement with
Because historical eddy covariance data were not available, the simulated NEP was not comparable with annual NEE measurements at the tower flux station (45° 24.215′ N, 127° 39.651′ E -
In recent years, an increasing number of studies have attempted to quantify C sinks and sources for all terrestrial ecosystems in northeastern China.
The average NPP was low in the early 20th century and slowly increased during the following 30 years, which may have been caused by low temperatures, precipitation, atmospheric CO2 concentrations, and N deposition. Thereafter, NPP began to increase with increases in climate variability and atmospheric CO2 concentration. However, NPP began to decrease in the early 1960s because China suffered heavily from natural disasters and the Cultural Revolution in the 1960s and 1970s, when large forest areas were harvested and destroyed and large-scale plantations were uncommon (
In this study, we analyzed the temporal variation in NEP in Heilongjiang. Heilongjiang was a carbon source in the 7 years before 1930, but was a huge carbon sink in 2013. During 1931-1949, although China suffered from the Second World War and the Chinese Civil War, Heilongjiang was not the primary battlefield, and therefore the forests in Heilongjiang were not seriously damaged. During the 113 years, two long-term reductions were apparent: one was in the late 1970s, and the other was in the early 21st century. According to the records for forest fires from 1953 to 2012 (
Point-by-point comparisons between simulated NPP by BEPS and MODIS NPP in 2003 suggested that the model generally captured the magnitude of NPP in the 433 points (
We also compared InTEC results for NPP from 2000 to 2013 with MODIS NPP (MOD17A3 -
The National Forest Inventory data generally included age, height, diameter at breast height, stand density, species composition, and area. Based on these parameters, we calculated the biomass for different ages in sample plots using the individual biomass models produced by Ding et al. (
where
A total of 100 sample plots in the inventory data set for 2005 were used to calculate NPP and were validated by the InTEC model. InTEC slightly overestimated NPP compared with the inventory data (
We also compared simulated biomass C with previous studies from different periods in Heilongjiang. For example,
Similarly, we compared the soil C stock with that in other studies. However, because the definitions of forest area, cover type, and soil depth were different between the InTEC model and those in other studies, comparisons of the results for soil C storage were difficult. Comparisons of the estimates of soil C density between our study and those in others indicated that the InTEC model simulated C changes in soil reasonably well (
In this paper, we used the InTEC model to estimate NPP and NEP in Heilongjiang from 1901 to 2013 and then analyzed the dynamic changes in the spatiotemporal C balance over the 113 years.
Our analysis indicated that NPP was low in the early 20th century; thereafter, NPP began to increase steadily and reached peak values in the early 1990s. Recently, NPP has been maintained at a relatively high level between 390 g C m-2 a-1 and 410 g C m-2 a-1. We concluded that NPP was closely related to climate change and the length of the growing season. However, greater efforts toward forest protection and a more reasonable stand age structure were the decisive factors for the increase in NPP in Heilongjiang.
The average NEP of Heilongjiang varied greatly, but we found that most forest disturbances occurred in the identical periods as long-term decreases in NEP. Disturbances periodically released large amounts of C to the atmosphere and further changed stand age structures. Our study showed that NEP in Heilongjiang forests was greatly influenced by disturbance factors, which is consistent with previous studies (
According to the spatial distribution of NPP in Heilongjiang in 1901, 1950, and 2013, NPP in the Xiaoxing’an Mountains showed a tendency to increase from north to south, which is consistent with the distribution of forest coverage. The distribution of NPP in the Changbai Mountains was more uniform than that of the Xiaoxing’an Mountains; however, the lowest average NPP was in these mountains. The spatial distribution of NEP was similar to that of NPP in Heilongjiang. We believe that this result was due to low forest coverage; therefore, large-scale forest plantations in these two regions will improve both NPP and NEP.
In this study, a method was used that estimated the distribution of the C balance in Heilongjiang and quantitatively analyzed the changes in spatiotemporal C budgets by offering a reference to assess the C balance in the past and project the variation into the future. Some aspects of this study were inadequate; for example, effects of disturbances were determined by changes in forest stand age structure rather than separating the disturbances into harvests, insect attacks, and fires. Furthermore, the 1 km resolution was too coarse to adequately explore the C dynamics in each pixel, because numerous pixels were mixed with nonforest plants. Additionally, the lack of available historical eddy covariance data resulted in insufficient validation of NEP. Hence, these aspects of the estimates of the C budgets of Heilongjiang will be addressed in future studies.
Research grants from the National Natural Science Foundation of China (31470640 and 31300420), the Natural Science Foundation of Jiangsu (BK20130987) and the Fundamental Research Funds for the Central Universities (2572016AA30) supported this research. Models in support of this article were from the research collaboration with Prof. Jingming Chen and Fangmin Zhang at the University of Toronto; please contact the authors for more details.
Average net primary production (NPP) from 1901 to 2013.
Spatial distribution of net primary production (NPP) in the study region of Heilongjiang. (a) 1901; (b) 1950; (c) 2013.
Average net ecosystem production (NEP) from 1901 to 2013.
Spatial distribution of net ecosystem production (NEP) in the study region of Heilongjiang. (a) 1901; (b) 1950; (c) 2013.
Comparison of net primary production (NPP) from Integrated Terrestrial Ecosystem Carbon (InTEC) model and MODIS product from 2000 to 2013.
Comparison of calculated net primary production (NPP) from Integrated Terrestrial Ecosystem Carbon (InTEC) model and inventory plots.
Integrated Terrestrial Ecosystem Carbon (InTEC) model parameters.
Symbol | Units | Description | Coniferous forests | Broadleaved forests | Mixed forests | Unique value | Reference |
---|---|---|---|---|---|---|---|
|
- | Maximum value of carbon per unit LAI | 70 | 31.5 | 53.25 | - |
|
μmol m-2 s-1 | Maximum carboxylation rate | 33 | 60 | 40 | - |
|
|
- | Sensitivity of N fixation rate to temperature | - | - | - | 2.3 |
|
|
- | Sensitivity of electron transport to temperature | - | - | - | 1.75 |
|
|
- | Sensitivity of Rubisco activity to temperature | - | - | - | 2.4 |
|
|
g N m-2 | Actual leaf nitrogen content | - | - | - | 1.2 |
|
Foliage turnover rates (
Tree species | e | Reference | |
---|---|---|---|
0.385 | 1.03 |
|
|
|
0.4 | 1.4 |
|
|
1 | 1.4 |
|
|
1 | 1.2 |
|
|
1 | 1.26 |
|
|
1 | 1.2 |
|
|
1 | 1.2 |
|
Comparison of carbon stock in biomass from previous forest inventories with estimates in this study.
Period |
|
InTEC model | ||||
---|---|---|---|---|---|---|
Area(×1011 m2) | C Pool(×1011 kg) | C density(kg m-2) | Area(×1011 m2) | C Pool(×1011 kg) | C density(kg m-2) | |
1973-1976 | 2.51 | 7.96 | 3.17 | 1.88 | 6.92 | 3.70 |
1977-1981 | 1.53 | 5.41 | 3.55 | 1.88 | 7.29 | 3.90 |
1984-1988 | 1.56 | 5.66 | 3.64 | 1.88 | 8.23 | 4.40 |
1989-1993 | 1.61 | 5.88 | 3.65 | 1.88 | 8.60 | 4.60 |
1994-1998 | 1.76 | 6.22 | 3.54 | 1.88 | 9.35 | 5.00 |
1999-2003 | 1.80 | 6.01 | 3.34 | 1.88 | 9.72 | 5.20 |
Comparison of carbon stock in soil from previous forest inventories with the estimates obtained in this study.
Value inthis study(kg m-2) | Estimate(kg m-2) | Method | Period | Region | Reference |
---|---|---|---|---|---|
13.27 | 13.57 | Based on the Second National Soil Survey | 1980s | Heilongjiang |
|
13.30 | 21.78 | FORCCHN model | 1981-2002 | Northeast China |
|
13.33 | 14.70 | Based on multi-purpose regional geochemical survey | 2006 | Heilongjiang |
|
17.38 | 17.26 | Based on the date plots measured | 2008-2011 | Daxing’an Mountains of Heilongjiang |
|
12.67 | 16.79 | Based on the date plots measured | 2008-2011 | Xiaoxing’an Mountains of Heilongjiang |
|