Evaluating the Conservation Status and Effectiveness of Multi-Type Protected Areas for Carbon Sequestration in the Loess Plateau, China

Abstract

Evaluating the conservation effectiveness of multiple types of protected areas (PAs) on carbon sequestration services can enhance the role of PAs in mitigating global warming. Here, we evaluated the conservation status and effectiveness of national parks, nature reserves, forest parks, geo-parks, and scenic spots on carbon sequestration within the Loess Plateau throughout 2000–2020. The results show that all existing PA types have good representation and conservation effectiveness on carbon sequestration. Nature reserves are the most representative of carbon sequestration but are the least effective in protecting carbon sequestration and are the only ones that are weekly effective in protecting critical carbon sequestration. The main factors influencing these results are PA size, 2000 precipitation, slope, change rate of evapotranspiration, PA rank, and 2000 evapotranspiration. We suggest upgrading the critical carbon sequestration distribution areas in scenic spots, forest parks and geo-parks to national parks or nature reserves in the future and implementing appropriate protection and restoration measures in low carbon sequestration areas within grassland and wild plant nature reserves to help achieve the goal of carbon neutrality early.

1. Introduction

Human-induced climate change and biodiversity loss pose interconnected emergencies that threaten human well-being [1,2,3]. Policy frameworks such as the Sustainable Development Goals, United Nations Framework Convention on Climate Change, Convention on Biological Diversity, and others emphasise that ecological conservation and restoration should simultaneously contribute to biodiversity conservation and climate change mitigation [4,5,6,7]. Increasing concentrations of greenhouse gases like carbon dioxide (CO₂) are significant contributors to global warming [8,9], accelerated by increasing deforestation and vegetation degradation [10,11,12].

Establishing Protected areas (PAs) not only enhances biodiversity conservation [13,14,15] but also mitigates global warming by increasing vegetation’s carbon sequestration services [16,17]. In response to global climate change, most countries are still planning to reach net-zero carbon emissions by 2050–2070 [18], and the Chinese government has proposed achieving carbon neutrality by 2060 [19]. Strengthening vegetation’s carbon sequestration in PAs is a key strategy to achieve this goal [20,21]. China is building a PA system mainly composed of national parks to address issues of spatial overlap and multiple management of existing PA types [22]. Given global climate change and the opportunity to integrate and optimise PAs in China, it is necessary to scientifically assess the conservation effects of different PA types on carbon sequestration.

Current studies predominantly focus on analysing changes in total carbon sequestration within PAs. Melillo et al. [16] analysed the changes in carbon sequestration in over 150,000 PAs worldwide from 1700 to 2005, finding that global carbon sequestration in PAs was 0.5 PgC per year, accounting for about one-fifth of the annual carbon sequestration of all terrestrial ecosystems; Tian et al. [23] assessed changes in carbon sequestration in 8133 different types of terrestrial PAs in China from 1980 to 2020, showing that the amount of carbon sequestration in PAs over past 40 years had significantly increased, with nature reserves sequestering the enormous amount of carbon, and forest parks sequestering the highest amount of carbon per unit area, despite a clear downward trend.

Research on conservation effects is not uniform. Internationally, studies tend to analyse all PAs as a single type. For instance, Duncanson et al. [24] analysed the conservation effects of 260,000 PAs globally on carbon sequestration between 2000 and 2020 using the propensity score matching method (PSM), and the results showed that PAs are crucial for climate change mitigation globally, especially in areas with high carbon sequestration. In China, the focus is more on the conservation effects of nature reserves on carbon sequestration. For example, Cao [25] used PSM to analyse the conservation effect of 19 national nature reserves in the Qinling Mountains on carbon sequestration from 2010 to 2015, finding that carbon sequestration in national nature reserves increased more than that outside the reserves. However, Xu et al. [26] used the representativeness evaluation method to assess the protection status of carbon sequestration in 2412 nature reserves in China during 2010 and found that the representativeness of nature reserves on carbon sequestration was weak. More research on different PA types regarding conservation effectiveness assessment is needed, as various methods can yield different results. Exploring the conservation effectiveness of different PA types on carbon sequestration can help promote the effective management of carbon sequestration and thus mitigate climate change.

The Loess Plateau, with severe soil erosion and a fragile ecological environment, is one of the most concentrated regions in China in terms of population, resources, and environmental conflicts [27]. To curb soil erosion, the Chinese government launched the Grain for Green Project on the Loess Plateau in 1999, significantly changing the regional land use pattern [28]. This paper investigates the conservation status and effects of carbon sequestration in five significant PAs types (national parks, nature reserves, forest parks, geo-parks, and scenic spots) on the Loess Plateau from 2000 to 2020. The objectives of this study are (1) to clarify the conservation status and effects of different PA types on carbon sequestration, (2) to identify the critical factors affecting the conservation effects of carbon sequestration, and (3) to provide suggestions for subsequent conservation management, thereby offering scientific support for the integration and optimisation of China’s PAs and contributing to global goals of carbon neutrality and climate change mitigation.

2. Materials and Methods

2.1. Study Area

The Loess Plateau is a temperate semi-moist and semi-arid transition zone in China that spans approximately 649,000 km², accounting for 6.76% of China’s total land area [29]. The region’s average annual rainfall is between 20 and 700 mm, and the vegetation (Figure 1) from southeast to northwest is warm-temperate deciduous broad-leaved forest, forest grassland, typical grassland, and desert grassland [30]. In addition, the Loess Plateau is the world’s largest accumulation area of loess and an essential ecological function area in the middle reaches of the Yellow River [31]. In order to protect the regional ecological environment, more than 700 PAs of different types, accounting for 15.1% of the total area of the Loess Plateau, of which more than 160 are nature reserves, have been established.

2.2. Methods

In this study, we first selected PAs with carbon sequestration functions established within the Loess Plateau by 2020 and those established by 2000 and earlier. We then conducted a representative analysis and applied methods based on the strictness of PAs to evaluate their conservation status and effectiveness in carbon sequestration. Subsequently, we identified factors affecting conservation effectiveness based on the basic attributes of the PAs, such as natural geography, socio-economic, and demographic aspects. Finally, we provided targeted recommendations to enhance the effectiveness of these conservation efforts. Figure 2 shows the research flowchart.

2.2.1. Data Sources

We obtained data on the spatial distribution and basic information of PAs on the Loess Plateau from the State Forestry and Grassland Administration. To evaluate the conservation status and effectiveness, we obtained carbon sequestration data from the 2000 to 2020 China Ecosystem Services Dataset of the Ecological Environment Research Centre of the Chinese Academy of Sciences, with a resolution of 250 m. Data on influencing factors, including precipitation, temperature, evapotranspiration, population density, GDP, soil texture, DEM, and slope, were obtained from the Resource Environmental Science Data Registry and Publishing System at a spatial resolution of 1 km. We used a spatial resolution of 250 m to assess the conservation effects of cs, and a spatial resolution of 1 km was used to analyse the impact factors, both in WGS 1984 Albers coordinates and calculated in ArcGIS 10.8.

2.2.2. Selected Protected Areas

In the current conservation situation analysis, we choose PAs with an area larger than 5 km² and with more vegetation distribution to remove those without carbon sequestration or lower carbon sequestration functions. Selected five types of PAs, including national park pilot areas, forest parks, geo-parks, scenic spots, and nature reserves (dominated by forests, grasslands, wild animal, and wild plant types), to assess the 2020 conservation status of carbon sequestration in 2020. We only selected the PAs established in 2000 and before for the dynamic analyses. Because we could not obtain the boundaries of the PAs in each period, we assumed that the boundaries of these PAs had not changed.

2.2.3. Carbon Sequestration

We used the fixed amount of CO₂ to evaluate the physical quantity of carbon sequestration in natural ecosystems, which was calculated by the carbon sequestration rate method [32].

$${Q_t}_{{CO}_2}=M_{C{O}_2}/M_C\times \left(VCSR+SCSR\right)\times S$$

(1)

where, ${Q_t}_{{CO}_2}$ means the total carbon sequestration (t CO₂/a), $\frac{M_{C{O}_2}}{M_C}=44/12$, means the coefficient of conversion of C to CO₂, VCSR means the carbon sequestration rate of vegetation, SCSR means the carbon sequestration rate of soil, and S indicates the area of each ecosystem type.

The importance of ecosystem services indicates the significance of ecological protection for different ecological units. Based on the physical amount of carbon sequestration in the Loess Plateau region, we classified 50% of the area with the highest physical amount as critical [33].

2.2.4. Conservation Effectiveness

We used spatial overlay analysis and before-and-after comparisons to analyse the conservation effectiveness of carbon sequestration in PAs. For status quo conservation effectiveness, we only used the carbon sequestration data in 2020 by using representation analysis as follows [26]:

$$\frac{P{A}_{C{S}_{20}}}{L{P}_{C{S}_{20}}}\times 100\%\ge \frac{P{A}_{are{a}_{20}}}{L{P}_{area}}\times 100\%$$

(2)

For dynamic protection effectiveness assessments, we used the rate of change in the physical amount of carbon sequestration from 2000 to 2020 as follows:

$$\frac{P{A}_{C{S}_{20}}-P{A}_{C{S}_{00}}}{P{A}_{C{S}_{00}}}\times 100\%\ge W_i\times \frac{L{P}_{C{S}_{20}}-L{P}_{C{S}_{00}}}{L{P}_{C{S}_{00}}}\times 100\%$$

(3)

where $P{A}_{C{S}_{20/00}}$ means the physical amount (critical amount) of carbon sequestration in the PAs in 2020/2000; $L{P}_{C{S}_{20/00}}$ means the physical amount (critical amount) of carbon sequestration in the Loess Plateau in 2020/2000; $P{A}_{are{a}_{20}}$ means the total area of PAs in the Loess Plateau in 2020; $L{P}_{area}$ means the area of Loess Plateau.

$W_i$ means the weights of different PA types were set according to their strictness rating in China [34]. Prior to the establishment of national parks in China, nature reserves were set as the strictest type of PAs [35], subdivided into three function zones (e.g., core, buffer, and experimental), three major categories (e.g., natural ecosystem, wildlife, and natural relics), and nine types. Only a partial of the experimental zones can be opened to the public with permission. Natural ecosystem types nature reserves contained representativeness, typicality, and completeness ecosystems; wildlife types nature reserves focused on rare wildlife species and their habitats. Other PA types, such as forest parks, geo-parks, and scenic spots, were less subdivided and relatively more open to the public. Based on that, the weight of nature reserves is set to 2/3 (when considering nature reserve types alone, the weight of nature reserves of natural ecosystem types is set to 2/3, the weight of nature reserves of wildlife types is set to 1/3), the weight of geo-parks, forest parks, and scenic spots is set to 1/3.

2.2.5. Impact Factors

Using the rate of change in the physical volume of carbon sequestration within PAs as the dependent variable, Spearman correlation analyses were conducted using SPSS software to reveal the drivers, such as differences in PAs (e.g., time, area, and area), natural factors (e.g., dem, slope, temperature, precipitation, evapotranspiration, and soil texture), and threatening factors (e.g., population density, GDP, and cropland). We define a correlation coefficient of less than 0.3 as a weak correlation, 0.3 and 0.6 as a moderate correlation, and greater than 0.6 as a strong correlation.

3. Results

3.1. Conservation Status

Regarding physical quantity, the total amount of carbon sequestration in the Loess Plateau in 2020 was 277.41 Tg, of which 138.71 Tg is critical. Regarding spatial distribution (Figure 3), the amount of carbon sequestration shows a pattern that is higher in the southeast and lower in the northwest. The areas with high carbon sequestration are mainly distributed in the southern part of the Lvliang Mountains, the southern part of the Taihang Mountains, the Zhongtiao Mountains, the Ziwu Ridge, the northern part of the Qinling Mountains, the southern part of the Liupan Mountains, and the Daban Mountains; the areas with low carbon sequestration mainly concentrated in the northern desert area.

Based on the screening method, a total of 430 PAs (including national park pilot areas, forest parks, geo-parks, scenic spots, and nature reserves) have been selected, with a total area of about 70,500 km², accounting for 10.85% of the total area of the Loess Plateau in 2020. Regarding spatial distribution (Figure 4a), there are fewer PAs in the central part of the Loess Plateau, and most of them are distributed in the eastern part. In terms of the number of PA types (Figure 4b), the number of forest parks is higher, accounting for 51.86% of the total, followed by nature reserves, which account for 27.21%, scenic spots and national parks account for a smaller proportion of the total, with 6.05% and 0.23%, respectively. In terms of the area of the types (Figure 4b), the nature reserves accounted for the largest proportion of the area of the Loess Plateau at 5.70%, followed by forest parks at 2.79%, scenic spots, and national parks accounted for a relatively small proportion of the area at 0.51% and 0.37% respectively.

In 2020, the total amount of carbon sequestration in the 430 PAs was 104.52 Tg. The total amount of critical carbon sequestration was 69.44 Tg, which accounted for 37.68% and 50.07% of the total amount of the Loess Plateau, respectively; this is much more than the proportion of the total area of the Loess Plateau accounted for by the PAs (10.85%). As can be seen from Figure 5, the conservation effect of carbon sequestration in the five types of PAs was better, with the proportion of forest parks, geo-parks, and scenic spots exceeding 40% of the proportion of the area. Although nature reserves are more effective in protecting carbon sequestration, in terms of specific types, forest and wild animal types of nature reserves are more effective in safeguarding carbon sequestration. In contrast, grassland and wild plant types are relatively weaker.

In terms of specific quantities, about 100% of national parks, 88.46% of scenic spots, 75.21% of nature reserves, 72.65% of forest parks, and 60.32% of geo-parks are more effective in protecting the total amount of carbon sequestration. About 80.95% of forests, 77.27% of wild animals, 33.33% of wild plants, and 20% of grassland-type nature reserves are more effective in protecting the total carbon sequestration. About 100% of national parks, 91.30% of scenic spots, 84.62% of geo-parks, 80% of nature reserves, and 77.27% of forest parks protect critical carbon sequestration well; all wild animals and plants, and 75.38% of forest-type nature reserves protect critical carbon sequestration well.

3.2. Conservation Effectiveness

During the past 20 years, the total amount of carbon sequestration within the Loess Plateau increased from 152.48 Tg to 277.41 Tg, increasing about 81.93%. We selected 187 eligible PAs (Figure 6), with a total area of about 36,500 km², accounting for 5.62% of the total area of the Loess Plateau. Forest parks had the most significant number of sites, amounting to 93, followed by nature reserves with 53, scenic spots and geo-parks with 23 and 18 sites, respectively; nature reserves accounted for the largest share of the area, followed by forest parks, scenic spots, and geo-parks. In terms of the specific types of nature reserves, the number of forest and wild animal types of nature reserves is higher, the area of forest and wild plant types of nature reserves accounts for a higher proportion, and the area of grassland types of nature reserves accounts for a minor proportion.

Total carbon sequestration in the 187 PAs increased from 30.43 Tg to 46.99 Tg, increasing about 54.42%. In terms of specific types, the total amount of carbon sequestration in nature reserves increased the most during the 20 years, reaching 62.47%, followed by scenic spots and forest parks, with increases of 52.71% and 52.51%, respectively, and geo-parks with the smallest increase of 41.34%; despite the high increase in total carbon sequestration within nature reserves, in terms of specific types, forest and wild animal type nature reserves had the most considerable increase of 75.60% and 45.47%, respectively, followed by wild plant type nature reserves with a rise of 18.42%. However, grass-type nature reserves showed a certain degree of decline, with a decrease of 3.31%.

The total amount of critical carbon sequestration in the 187 PAs increased from 21.98 Tg to 31.95 Tg, an increase of about 45.36%. In terms of specific types, the total amount of critical carbon sequestration within forest parks increased the most during the 20 years, with 50.98%, followed by scenic spots and nature reserves, with 41.64% and 40.30%, respectively, and geo-parks with the smallest increase of 41.34%. In terms of specific types of nature reserves, the largest increase in the amount of critical carbon sequestration was in the forest-type nature reserves, followed by the wild animal-type nature reserves, and there was no distribution of critical carbon sequestration within the grass and wild plant type nature reserves.

The results show that during the past 20 years, all four types of PAs have a better protection effect on carbon sequestration, among which scenic spots and forest parks have a better protection effect, followed by geo-parks and nature reserves (Figure 7). Only forest and wild animal types of nature reserves had a better protection effect on carbon sequestration services. In contrast, grassland and wild plant types of nature reserves had a lesser protective effect. For critical carbon sequestration, forest parks, scenic spots, and geo-parks all have a better protection effect. In comparison, nature reserves have a weaker protection effect. Regarding the specific types of nature reserves, only wild animal-type nature reserves have a better protection effect.

In terms of the number of specific PAs, about 91.30% of scenic spots, 83.87% of forest parks, 77.78% of geo-parks, and 54.72% of nature reserves are more effective in carbon sequestration; about 75% of wild plants, 69.23% of wild animals, 56.25% of forests, and 25% of grassland-type nature reserves are more effective in carbon sequestration. About 69.57% of scenic spots, 56.99% of forest parks, 44.44% of geo-parks, and 15.09% of nature reserves are more effective in protecting critical carbon sequestration; about 53.85% of wild animals and 15.63% of forest-type nature reserves are more effective in safeguarding critical carbon sequestration.

3.3. Main Factors

Spearman’s correlation analysis showed (

Table 1. Correlation test results.

R	CS Change Rate	Critical CS Change Rate
PA establishment year	0.169 *	0.107
PA level	0.331 **	0.221 **
PA area	0.612 **	0.146 *
DEM	0.042	−0.264 **
Slope	0.380 **	0.345 **
Silt	−0.266 **	−0.453 **
Sand	0.106	0.246 **
Temperature in 2000	−0.03	0.278 **
Rate of temperature change	0.03	0.212 **
Precipitation in 2000	0.418 **	0.638 **
Rate of precipitation change	−0.14	−0.307 **
Evapotranspiration in 2000	−0.313 **	−0.369 **
Rate of evapotranspiration change	0.326 **	0.605 **
Population density in 2000	0.074	0.235 **
Rate of population density change	0.082	0.025
GDP in 2000	0.099	0.213 **
Rate of GDP change	0.047	0.069
Cropland in 2000	0.281 **	0.004
Rate of cropland change	−0.048	0.046

Notes: * indicates significance at the 0.05 level, ** indicates significance at the 0.01 level.

) that there were many factors affecting the rate of change of carbon sequestration, with highly significant strong positive correlation with the size of the PAs, highly significant moderate positive correlation with precipitation, slope, rate of change of evapotranspiration, and the grade of the PAs in 2000; highly significant moderate negative correlation with evapotranspiration in 2000; highly significant weak positive correlation with cropland area in 2000; and highly significant weak negative correlation with silt loam content.

The factors that were highly significant and strongly positively correlated with the rate of change of critical carbon sequestration were precipitation in 2000 and evapotranspiration change rate; the only factors that were highly significant and moderately positively correlated was slope, but they were highly significant and moderately negatively correlated with evapotranspiration in 2000, silt, and precipitation change rate; and highly significant and weakly positively correlated with the temperature in 2000, population density in 2000, the grade of PAs, GDP in 2000, and the temperature change rate, but highly significant and weakly negative with DEM; in addition, it was also highly significant and weakly positively correlated with the size of the PAs.

4. Discussion

This study analysed the conservation status of carbon sequestration in 430 PAs of five types on the Loess Plateau during 2020 using the representative evaluation method. We found that, without considering the spatial overlap of different PA types, directly adding up the areas of all PAs, the 430 PAs accounted for only 10.85% of the total area of the Loess Plateau but provided 37.68% and 50.07% of the total amount of carbon sequestration and critical carbon sequestration, which was much higher than the area share of the PAs; therefore, we believe that PAs on the Loess Plateau are better representative of the carbon sequestration and the critical sequestered carbon. However, the representativeness of different types of PAs for carbon sequestration services varies, with nature reserves having the best representativeness for carbon sequestration, followed by forest parks and geo-parks, which is consistent with the results of the national study [23], which may be because, nature reserves as the most stringent type of PAs in China, are established larger than other types to protect the integrity of the ecosystem [36]. In the case of the Loess Plateau, the area of nature reserves in the region exceeds that of forest parks by a factor of two and that of geo-parks by a factor of four. Larger areas generally provide more resources and habitats, thus supporting greater carbon sequestration [37]. Although nature reserves are better represented for carbon sequestration services, in terms of specific types, only forest and wild animal-type nature reserves are better represented for carbon sequestration. In contrast, grassland and wild plant-type nature reserves show lesser representativeness. The reason for this phenomenon is that the vegetation types of wild plant type nature reserves in the Loess Plateau region are dominated by grasslands, which in turn have a lower carbon sequestration capacity than that of forests [38,39], so overall grassland and wild plant type nature reserves are weakly represented for carbon sequestration services.

Based on the differences in the degree of protection stringency of different types of PAs [33], constructing the protection effect of multiple types of PAs on carbon sequestration services not only identifies the differences in the protection of each type of PAs but also proposes targeted protection advice which is conducive to the overall improvement of the protection effect. Until now, China has established more than 11,800 PAs [40], which can be classified into three types: strict protection, restricted use, and protected use [34,41]. According to the degree of strictness of protection, nature reserves are recognised as the strict type of reserve. They are divided into three categories and nine types [42]. While scenic spots, forest parks, and geo-parks can be classified as restricted use types. There are many nature reserves with complex classification systems, of which nearly 18% have spatial overlap and cross-cutting management authorities, resulting in ineffective management, especially in the Taihang Mountains region within the Loess Plateau [43]. Therefore, we constructed a method for assessing the conservation effect of multiple types of PAs on carbon sequestration services by setting reasonable weights, considering the variability of background conditions and the degree of protection stringency of PAs on the Loess Plateau. Our proposed method is, on the one hand, more scientific than determining the conservation effect or contribution based only on the increase or decrease of carbon sequestration in the before-and-after timeframe. On the other hand, it is more convenient to operate than the propensity score matching method. We have taken into account the strictness of the protection of different types of PAs, such as nature reserves being the most strict type; only the experimental area allows part of the scientific research activities, its management policies, measures, funds, personnel and other aspects of the investment are higher than other types of PAs, the primary purpose is to protect the main object from interference [44]. Forest parks, scenic spots, and geo-parks allow a certain degree of human activities, considering the functions of protection, landscape, recreation, and popular science education [45], so there is a difference in the setting of weights. In addition, the PSM method needs to consider more environmental variables [46]. It even needs to take relevant policy factors into account. Still, it is difficult to maintain the consistency of the policy factors implemented inside and outside the PAs, so not all the PAs can be found in the periphery of the environmental factors of the same matching samples (such as the PAs of small size, the environmental factors do not match) [47]. The workload is heavier and not very operable. The reasonable weights we set based on the difference in protection strictness can quickly and scientifically assess the protection effect of different types of PAs on carbon sequestration services.

Based on the proposed conservation effectiveness assessment method, we assessed the conservation effectiveness of carbon sequestration services in 187 PAs established in 2000 and before of four categories in 2000–2020. We found that the total carbon sequestration within the Loess Plateau increased by about 81.93% during the 20 years, while the total carbon sequestration services within the 187 PAs increased by about 54.42%, which was lower than the increase in total carbon sequestration within the Loess Plateau. This phenomenon may be due to the unique soil conditions of the Loess Plateau itself, where soil erosion occurs from time to time in areas not covered by PAs [48]. To combat this ecological problem, China has dramatically increased the vegetation cover in the region by implementing long-term natural forest protection projects, grain for green, soil and water conservation, and other conservation and restoration measures [28,49]. At the same time, PAs are constrained by relevant policies and laws and cannot carry out too much artificial restoration [48], so the total increase in carbon sequestration services within the Loess Plateau was more extensive than that in PAs, but this does not indicate that PAs are less effective in protecting carbon sequestration. Combining the strict degree of protection of PAs and the background condition of the Loess Plateau and by setting reasonable weights [34], we found that all 187 PAs of the four categories have a better protective effect on carbon sequestration services. Still, scenic spots and forest parks have a better protective effect on carbon sequestration than geo-parks and nature reserves, forest and wild animal type nature reserves have a better protective effect on carbon sequestration, and grassland and wild plant type nature reserves have a weaker protective effect on carbon sequestration, suggests a need for targeted interventions to enhance their conservation outcomes [50]. The main reason for this result is that scenic spots, forest parks, forest and wild animals type nature reserves are mainly distributed in the eastern and southern parts of the Loess Plateau. In contrast, geo-parks, large nature reserves, grassland, and wild plant nature reserves are distributed primarily in the northern part of the Loess Plateau. The vegetation type of the eastern and southern parts of the Loess Plateau is dominated by forests, while grasslands dominate that of the northern part. In contrast, precipitation and slope in the southeastern part of the Plateau are higher than in the north, and evapotranspiration is lower than in the north. Evapotranspiration is a key indicator of water availability and plant health, directly affecting carbon uptake [51]. Related studies have also shown that high precipitation and low evapotranspiration are essential influences on the carbon sequestration capacity of vegetation [52]. For critical carbon sequestration, forest parks, scenic spots, and geo-parks all have a better protection effect on it. Only nature reserves have a weaker effect on it, but wild animal-type nature reserves have a better effect. The main reason for this phenomenon is that there is a spatial mismatch between most of the nature reserves and critical carbon sequestration [26], which is concentrated in the southeastern part of the Loess Plateau, but the distribution of nature reserves is more expansive and not focused enough.

Our findings underscore the crucial role of considering PA types and background conditions when evaluating conservation effectiveness. The outstanding performance of scenic spots, forest parks, and geo-parks indicates that these PA types may benefit from management practices prioritising carbon sequestration. Conversely, the relatively weaker performance of grassland and wild plant nature reserves suggests a need for targeted interventions to enhance their conservation outcomes, such as carrying out artificial restoration of degraded grasslands with a multi-species configuration of native grasses with a variety of legumes [50,53]. Our results suggest that integrating and optimising the management of PAs should focus on upgrading the protection level of critical carbon sequestration areas within scenic spots, forest parks, and geo-parks [54]. Combined with the overlap** analysis of PAs [55], the area should be included in future national parks or nature reserves according to the actual situation. Implementing comprehensive protection and restoration measures in low-performing PAs, particularly grassland and wild plant nature reserves, can significantly enhance their carbon sequestration capacity. This approach aligns with China’s broader goal of achieving carbon neutrality and emphasises the dual benefits of biodiversity conservation and climate change mitigation.

The IUCN classifies PAs into six categories, with category Ia representing strict PAs [56]. Countries around the world have further subdivided PA types based on this. Despite these differences, all are classified into certain levels of strict protection. Therefore, this study provides a methodological reference for evaluating the effectiveness of PAs with varying protection levels in different regions and countries, supporting related international research, and then mitigating global climate change.

5. Conclusions

This study analysed the conservation status and effects of different types of PAs on carbon sequestration services within the Loess Plateau from 2000 to 2020. The results found that: (1) The existing PAs demonstrate good representativeness and conservation effectiveness for carbon sequestration services, though these vary by type. While stricter PAs generally show better representativeness, this does not always translate to higher conservation effectiveness. For example, nature reserves exhibit the best representativeness for carbon sequestration, yet their conservation effectiveness is lower compared to other types of PAs. (2) Factors that positively and significantly affect the conservation effect of carbon sequestration include the size of the PAs area, precipitation in 2000, slope, rate of change of evapotranspiration, and PAs class. (3) It is recommended to appropriately upgrade the protection level of areas with low protection stringency but high carbon sequestration and to implement necessary restoration measures in areas with high protection stringency but low carbon sequestration. This approach will not only serve as a reference for the integration and optimisation of PAs in China but also provide guidance for other regions worldwide with similar classifications of PAs so that we can comprehensively enhance the carbon sequestration capacity and help achieve the goal of carbon neutrality.