Analysis of the Characteristics and Influencing Factors of Carbon and Water Fluxes in the Restored Sites of Puding, Guizhou Province

Abstract

This study investigates the microclimate characteristics and temporal variations in carbon and water fluxes in the naturally restored plots in Puding County, Anshun City, Guizhou Province. Continuous monitoring and collection of environmental factors through the eddy covariance system reveals the dynamic change patterns of the microclimate environment in the early stages of returning farmland to forest and the dynamic characteristics of carbon and water fluxes in the ecosystem. The results show that the study area exhibits a pronounced carbon sink capacity at the annual scale. The net ecosystem exchange (NEE) gradually increased with the extension of the recovery time, and the carbon sink intensity reached its peak during the growing season (June–August). The water use efficiency (WUE) was significantly higher in the growing season compared to the non-growing season, indicating that the WUE and carbon sequestration capacity increased simultaneously. Structural equation model (SEM) analysis revealed that latent heat flux has the most significant regulatory effect on carbon sink function. Furthermore, temperature, relative humidity, and latent heat flux have a significant positive regulatory effect on NEE (p < 0.001), while rain and soil heat flux affect carbon sink function through an indirect path. By constructing a complex meteorological carbon flux relationship network, the main regulatory mechanism of the ecosystem carbon sink in the Puding area was revealed. These findings confirm that the implementation of returning farmland to forest has effectively enhanced the carbon sequestration capacity of karst ecosystems.

1. Introduction

The assessment report released by the Intergovernmental Panel on Climate Change (IPCC) shows that since the 1950s, greenhouse gas emissions caused by human activities have played a key role in the process of global warming, with a reliability of up to 95% [1]. In 2019, 86% of global carbon emissions came from the use of fossil fuels, 14% came from land use, and approximately 46% of CO₂ remained in the atmosphere, while marine and terrestrial ecosystems absorbed 23% and 31% of CO₂, respectively [2,3].

With rapid economic growth, global carbon dioxide emissions have gradually increased [4], among which China plays an important role in global emission reduction and climate change mitigation. Due to the fact that terrestrial ecosystems are often disturbed by various disasters, such as high temperatures and droughts, the intensification of droughts and the expansion of arid areas have led to the scarcity of terrestrial carbon, nitrogen, and phosphorus resources, which has a negative impact on the soil carbon pool [5]. Furthermore, climate change will greatly reduce the anti-interference resilience of ecosystems, impairing photosynthesis, inducing vegetation mortality, and ultimately resulting in a systematic decline in total primary productivity [6]. Through artificial intervention means, such as increasing timber and optimizing forest management, the carbon stock can be increased by 13.6 ± 1.5 Pg C from 2020 to 2100, while the impact of logging on forest age will postpone the carbon sink peak time by 10–30 years [7].

The karst area in Southwest China, as a key region of the global carbon cycle, has formed a unique biogeochemical carbon sink mechanism under the synergistic effect of monsoon climate and karst landforms [8,9,10]. Since the implementation of ecological projects such as “mountain closure and forest regeneration” in the 1990s [11], ecological conditions in this region have significantly improved, and at the same time, the carbon sequestration capacity of this region has been effectively enhanced [12,13]. Among them, returning farmland to forest can significantly increase the content of soil organic matter, improve soil structure, enhance soil fertility, and thereby improve the soil’s carbon sequestration capacity [14]. In addition, the vegetation in forest land and grassland can effectively absorb a large amount of carbon dioxide, reduce greenhouse gas emissions, and significantly enhance the carbon sink function of the ecosystem. Especially, the forest ecosystem still has a remarkable carbon sequestration capacity [15], while the middle and low mountain areas can store more organic carbon due to the optimization of water and heat conditions [16]. These findings provide key scientific evidence for optimizing carbon sink management in global karst areas and responding to climate change [17,18].

This article focuses on the ecological restoration plot in Puding County, Anshun City, Guizhou Province, which has typical karst landforms and rocky desertification phenomena. The rainfall in this area is abundant, providing favorable conditions for the growth of subtropical plants. Structural equation modeling (SEM), as a multivariate statistical analysis technique, by combining factor analysis and path analysis, can simultaneously quantify both direct and indirect effects. It is particularly suitable for studies involving the synergistic effects of multi-dimensional environmental factors such as the carbon and water cycles in ecosystems. This study utilized the high-quality datasets provided by the National Ecological Science Data Center to calculate ecological parameters, adopted SEM to evaluate the carbon–water cycle efficiency of the ecosystem, and revealed the complex relationships among multiple variables in the ecosystem.

2. Material and Methods

2.1. Site Description

The study area is located in Longga Village, Chengguan Town, Puding County, Anshun City, Guizhou Province. It is 5 km away from the county seat, with geographical coordinates of 26°36′ N and 105°79′ E, and an altitude of 1170 m (Figure 1). The area where Puding Railway Station is located has a subtropical monsoon humid climate with distinct monsoon alternations. The climate conditions throughout the year are relatively mild. The average temperature over many years is 15.96 °C, the annual rainfall reaches 1432 mm, and there is a frost-free period of more than 300 days. The karst landform in Puding County is very typical. It is widely developed and covers a wide variety of evolution forms and types. The phenomenon of rocky desertification in the local area is serious.

The natural restoration plots within the main station area of Puding Railway Station were returned to farmland in 2010. During the observation period, the plot was in the initial stage of natural recovery and the vegetation was significantly restored. Therefore, it can be regarded as an ecosystem of trees, shrubs, and grass.

2.2. Data Compilation

The data used in this paper are provided by the National Ecosystem Science Data Center, National Science & Technology Infrastructure of China (http://www.nesdc.org.cn/ (accessed on 5 August 2024)). The collected data are from the Puding Station, which joined the Chinese Ecological Research Network (CERN) in 2014. This paper selects the carbon and water flux observation dataset of the natural restoration plot in Puding, Guizhou Province, from 2015 to 2019 [19]. These data include flux data with a half-hour resolution and meteorological data with an hour resolution. It includes air temperature (TA, °C), relative humidity (RH, %), wind speed (WS, m·s⁻¹), rain (Rain, mm), net radiation (NR, W·m⁻²), soil heat flux (Ht, W·m⁻²), net exchange interpolation data of the ecosystem (NEE,μmol·m⁻²·s⁻¹), total primary productivity of the ecosystem (GPP, μmol·m⁻²·s⁻¹), latent heat interpolation data (LE, W·m⁻²), and sensible heat interpolation data (H, W·m⁻²).

Water use efficiency (WUE) is a key indicator used to describe the mutual influence relationship between plant drought tolerance and the carbon–water cycle in the ecosystem. It is usually defined as the amount of CO₂ fixed per unit of water consumption or the mass of dry matter produced. At the ecosystem scale, it is the ratio of gross primary productivity (GPP) to the evapotranspiration of vegetation (evapotranspiration, ET), closest to the concept used by WUE to analyze the distribution ratio of water fluxes in ecosystems [20]. ET is converted from the actual measured latent heat flux LE, and the original LE is preprocessed by EddyPro software version 1.1.3 (including coordinate rotation, WPL correction, and outlier elimination). Therefore, the calculation of WUE in this article adopts the following formula:

$$\mathrm{WUE}={\displaystyle \frac{GPP}{ET}}$$

(1)

2.3. Statistical Analysis

The meteorological data and flux data were integrated into daily–monthly–annual variation data using Microsoft Excel 2016, and their mean and total values were calculated. The daily–seasonal–annual interannual variation patterns of climate and flux were compared and analyzed. The influence mechanisms of environmental factors such as temperature, precipitation, and relative humidity on the carbon sink capacity of artificial ecological restoration sites were explored through the structural equation model (piecewise SEM) in R 4.4.1 and R studio 2024.04.2. The variation characteristics were plotted using Origin 2024b.

Among them, the structural equation model analysis is a multivariate statistical analysis technique. As ecological research often involves complex relationships among multiple interdependent variables, the structural equation model can be used to reveal and quantify the direct and indirect relationships among multiple variables in an ecosystem and their potential response mechanisms, and all causal relationships are displayed through the arrow directions in the diagram [21]. In recent years, the structural equation model incorporating composite variables has gradually been applied to ecological research [22]. The R package (piecewise SEM) was used for structural equation model analysis, and the chi-square test of Fisher’s C value could be performed to determine the overall fit of the model [23].

3. Results

3.1. Regional Climate Characteristics

As shown in Figure 2a, by comparing the temperature change curves from 0:00 to 23:00 on 1 August 2019 from 2015 to 2019, it can be seen that the temperature began to rise in the morning (represented by 08:00) in each year, reached the highest temperature from 2 p.m. to 4 p.m., and then showed a significant downward trend. The daily humidity dropped after the sunrise temperature rose at 8:00, reaching the lowest value of the day between 14:00 and 16:00, and then due to the cooling at sunset, it gradually rose above 90%. Wind speed showed seasonal variations from 2015 to 2019, reaching its peak in early April and then gradually declining. The variation in wind speed within each year was relatively small, while the seasonal variation was obvious. The average wind speed was higher in spring and autumn and lower in summer and winter, with the average minimum wind speed occurring in July. As shown in Figure 2d, the average net radiation from 2015 to 2019 was 67.4 W·m⁻². All extreme values occurred at 12:00 noon, indicating that the solar radiation intensity was the greatest at noon, and the impact on ecosystem respiration was also the most significant during this period. The seasonal variation characteristics of the average daily rainfall over many years from 2015 to 2019 were obvious, mainly concentrated during the meadow growth stage (April to October), and the rain during the non-growth period (November to March) was much lower than that during the growth period. And as shown in Figure 2f, the hourly variation characteristics of soil heat flux in the growing seasons over many years were similar. From 0:00 to 6:00, the soil heat flux rose slowly and steadily. Starting from 6:00, the soil heat flux increased significantly, reaching a peak at 12:00. After a brief decline, it rose again and reached a peak again at 14:00. Subsequently, the soil heat flux began to decrease significantly and reached the lowest value of the day at 19:00.

3.2. The Variation Characteristics of NEE and Its Responses to Environmental Factors

During the growing season of Puding (April to October), the daily net carbon dioxide exchange volume generally shows a “U”-shaped curve variation characteristic. The carbon flux intensity every 30 min from 2015 to August 1, 2019 is shown in Figure 3a, which presents a carbon sink and weak carbon source during the day and night, respectively. NEE shows a bimodal trend of increasing first and then decreasing with temperature changes and is generally manifested as a carbon sink during the vegetation growth stage.

The monthly CO₂ exchange situation of the ecological restoration plot all manifests a carbon sink. Its carbon sink capacity first increases and then decreases, reaching the peak from June to August, showing a seasonal variation pattern. Among them, the carbon sink capacity of this terrestrial ecosystem reaches the strongest value in August and the weakest value in January (Figure 3b). The annual total variation in the measured NEE in the ecological restoration plots from 2015 to 2019 is shown in Figure 3c (where 2015 is the 275 data items counted starting from April 1). It can be known that the net carbon dioxide absorption of the ecosystem in this area shows an increasing trend with the advancement of the ecological restoration project, and the carbon dioxide absorption capacity of the ecosystem gradually increases as a whole. It is indicated that the adopted restoration measures have a positive effect on the carbon sequestration capacity of the ecosystem.

The calculation results through the structural equation model (piecewiseSEM) show that the net ecosystem NEE in this ecological restoration area is mainly regulated by atmospheric temperature (TA), relative humidity (RH), latent heat flux (LE), and wind speed (WS). The direct action coefficients are 0.121, 0.074, −0.808, and 0.144, respectively (p < 0.001, Figure 4), among which the influence of latent heat flux on NEE reaches a very significant level. The results show that among multiple years, temperature, humidity, and latent heat flux are the main positive environmental factors affecting the NEE of the restored ecosystem.

3.3. The Variation Characteristics of WUE and Its Responses to Environmental Factors

The daily total value variation in WUE in ecological restoration plots from 2015 to 2019 is shown in Figure 5a. After removing invalid values, WUE continued to increase from January to July, reached the maximum value between July and August, and then showed a downward trend until December. The WUE mainly fluctuates between −10 and 10 g·kg⁻¹, and the WUE during the growing season is significantly higher than that throughout the year.

According to the analysis of the structural equation model (piecewiseSEM) in R studio, the regulatory relationship between WUE and environmental factors from 2015 to 2019 is shown in Figure 5b. Environmental factors such as heat, temperature, relative humidity, soil heat flux, and latent heat flux all have significant indirect effects on WUE (p < 0.001), and most of them are negative effects.

4. Discussion

Our results indicated that the ecosystem of the natural restoration plot in the Puding area is relatively sensitive to climate change. In the future, with the intensification of global warming, determining how to maintain and enhance the carbon sink functions of different regions will become an important research topic, and the effects of different environmental factors related to it on the carbon sink vary.

For ecosystems in different regions, the soil characteristics they possess, such as soil nutrient content, pH value, organic matter content, and texture, also affect the carbon exchange capacity of the ecosystem. By increasing temperature and humidity experiments at multiple levels above and below ground, the regulation of soil respiration by climate changes such as global warming can be estimated more accurately, so as to accurately estimate the variation mechanism of carbon flux when warming occurs between years [24]. Different land use types, such as forests, grasslands, farmlands and urban areas, exhibit distinct carbon absorption and release capabilities due to differences in vegetation composition, biomass accumulation, and soil characteristics. In the research on the alpine grassland of the Qinghai–Tibet Plateau, it was found that its carbon sequestration capacity has been continuously increasing, indicating that under the background of a warm and humid climate, the carbon sequestration capacity of these ecosystems is increasing [25], while most studies on soil carbon flux are based on areas with non-karst landforms [26,27,28]. Therefore, it is particularly important to conduct in-depth observation, research, and analysis of carbon storage and flow in karst systems.

Multiple climatic factors and human activities can both affect the carbon exchange process in ecosystems. Yang Yunfeng et al. found that climate warming intensifies the positive stimulating effect of soil in temperate grassland ecosystems, especially when climate warming occurs simultaneously with an increase in plant biomass, leading to an increase in CO₂ emissions from the soil into the atmosphere [29]. The enhancement in ecosystem respiration is mainly attributed to the combined increase in plant-related respiration and microbial respiration. The impact of climate warming on respiration is closely related to changes in soil conditions. Therefore, the response of ecosystems to climatic factors is particularly sensitive [30].

The analysis of the forest ecosystem in the early stage of ecological restoration in this article plays an important role in enhancing the regional carbon sink function of the ongoing ecological restoration projects, provides data support for optimizing ecological restoration measures, and emphasizes the importance of long-term monitoring in the research of the ecosystem carbon cycle. In the future, long-term monitoring and observation of natural restoration plots in the Puding area should continue to be strengthened, and more high-quality data resources should be accumulated to provide strong support for in-depth analysis of the characteristics and influencing factors of carbon flux changes. Meanwhile, the observation and research on the carbon cycle of ecosystems in other karst areas should also be strengthened to promote the in-depth development of global carbon cycle research.

5. Conclusions

This study conducts a systematic analysis of the carbon–water flux characteristics and influencing factors of the natural restoration plots in Puding area, Guizhou Province, and draws the following main conclusions:

Through long-term observations of the vortex-correlated system, the variation process of the carbon and water fluxes in the ecosystem of the natural restoration plot in the Puding area has been revealed. This research finds that the seasonal peak of WUE (July–August) is in harmony with the high-temperature and high-humidity environment, the sample plots exhibit significant carbon sink effects on an annual scale, and the dynamic variation characteristics of NEE indicate that the carbon sink function gradually enhances with the increase in recovery time. Especially during the growing season, the carbon sink effect is more obvious, which has positive significance for the global carbon cycle and regional climate change.

The main meteorological factors affecting the ecosystem carbon flux in the Puding area were estimated by using the piecewise SEM (structural equation model). The results show that factors such as air temperature, humidity, wind direction and speed, rain, net radiation, and soil heat flux have direct or indirect influences on the water evaporation capacity and carbon dioxide exchange capacity of the ecosystem. Among them, at the multi-year scale, the effect of latent heat flux on the carbon sink function is greatest, reaching an extremely significant level, which reveals the main regulatory mechanism of the ecosystem carbon sink in the Puding area. The carbon sink intensity significantly increased with the recovery time (NEE annual increase of 23%), and the peak during the growing season was synchronized with the increase in WUE during the high-temperature and high-humidity period. Vegetation restoration is the main controlling factor for the increase in carbon sink, which provides institutional evidence for the carbon sink effect of the “returning farmland to forest” project in global karst areas.