A high stocking rate can intensify wind erosion in grasslands, and strong wind can carry away soil surface particles and their nutrients, which leads to soil barrenness. In this research, the dust flux, total carbon (TC) and total nitrogen (TN) contents and fluxes of aeolian sediment were computed in the non-growing season (mid-October to mid-April of the following year) and growing season (mid-April to mid-October) from 2018 to 2020 at a long-term grazing gradient experiment platform and wind-erosion monitoring experiment in a desert steppe in Inner Mongolia, China. The results were as follows, 1) with the increasing of stocking rate, the fluxes of aeolian sediment at 10 cm (H10), 30 cm (H30), and 100 cm (H100) were greatly increased ( P<0.05). Aeolian sediment fluxes followed the order control (CK) < light stocking rate (LG) < moderate stocking rate (MG) < heavy stocking rate (HG). 2) TC and TN contents of aeolian sediment reflected stocking rates and followed the order CK>LG>MG>HG, and also reflected aeolian sediment collection height, as H10P>0.05). TC and TN contents in HG at different heights reduced by 31.1% and 25.9% on average in comparison to the CK in the non-growing season, and by 30.1% and 25.0% in the growing season. TC loss was higher than TN loss. Overall, as stocking rate increased, wind erosion increased, leading to the loss of soil nutrients and significant loss of the carbon pool in this desert steppe.

Abstract Qinghai-Tibetan Plateau is an important ecological security barrier in China, and alpine ecosystem presents a trend of overall improvement under the influence of climate change and human activities, yet there are localized deteriorations. In order to improve the ecological function of the Tibetan Plateau, large-scale ecological restoration projects (ERPs) have been carried out in Tibetan Plateau over the past 30 years. From following the footstep of national ecological constructions to the implementation of projects specific to the plateau characteristics, the ERP of the plateau can be divided into three stages. Major ERPs focusing on four types of projects, i.e. forest protection and construction, grassland protection and construction, water and soil erosion control and desertification land management, with a total area of 850,000km 2 across the Qinghai-Tibetan Plateau. The positive effect at sampling quadrat scale has been widely verified, and the responses of productivity and species diversity were inconsistent. The positive effect at the regional scale gradually emerges, yet the spatial difference is significant, and the quantification of driving forces is an important prerequisite. The positive and negative effects of existing ERP on biodiversity conservation are revealed, for instance, there is a significant increase of rare wild animals in natural reserves and in artificially planted areas it failed to effectively curb the loss of animal and plant diversity. The long-term effects of various ERPs on the biodiversity should be taken into account so as to optimize the ecological measures and build a sustainable management model after project restoration. By summarizing the achievements and problems of ERP across the Qinghai-Tibetan Plateau in the past 30 years, we proposed to set stepwise ecological restoration and management scheme by different restoration goals and reference modes in three stages, i.e. “Environmental treatment—Ecological restoration projects—Adaptive management”.

Permafrost regions are under particular pressure from climate change resulting in widespread landscape changes, which impact also freshwater chemistry. We investigated a snapshot of hydrochemistry in various freshwater environments in the lower Kolyma river basin (North-East Siberia, continuous permafrost zone) to explore the mobility of metals, metalloids and non-metals resulting from permafrost thaw. Particular attention was focused on heavy metals as contaminants potentially released from the secondary source in the permafrozen Yedoma complex. Permafrost creeks represented the Mg-Ca-Na-HCO 3-Cl-SO 4 ionic water type (with mineralisation in the range 600-800 mg/L), while permafrost ice and thermokarst lake waters were the HCO 3-Ca-Mg type. Multiple heavy metals (As, Cu, Co, Mn and Ni) showed much higher dissolved phase concentrations in permafrost creeks and ice than in Kolyma and its tributaries, and only in the permafrost samples and one Kolyma tributary have we detected dissolved Ti or Hg. In thermokarst lakes, several metal and metalloid dissolved concentrations increased with water depth (Fe, Mn, Ni and Zn - in both lakes; Al, Cu, K, Sb, Sr and Pb in either lake), reaching 1370 µg/L Cu, 4610 µg/L Mn, and 687 µg/L Zn in the bottom water layers. Permafrost-related waters were also enriched in dissolved phosphorus (up to 512 µg/L in Yedoma-fed creeks). The impact of permafrost thaw on river and lake water chemistry is a complex problem which needs to be considered both in the context of legacy permafrost shrinkage and the interference of the deepening active layer with newly deposited anthropogenic contaminants.

：Vegetation blanket restoration techniques of different materials in mine drainage field provide different environments for plant growth, but effects of vegetation blanket cover on soil microbial community structure and their diversity characteristics were not well known， especially in arid areas. In this study, high-throughput sequencing was used on the study site, located at Dafeng Mine, the Helan Mountains, Ningxia, China. Soil microbial communities were analyzed under four different treatments: Straw, Straw-Coir, Coir vegetation blanket types, and a bare soil control. The results showed that the contents of soil total nitrogen, available potassium, urease, and catalase were significantly increased in different types of vegetation blankets. High-throughput sequencing showed that the straw vegetation blanket increased bacterial diversity while the coir vegetation blanket increased fungal diversity. The dominant bacterial phyla were Proteobacteria, Bacteroidetes, Actinobacteria, and Chloroflexi, whose main influencing factors were total nitrogen, catalase, urease, and protease. The dominant fungal phylum was Ascomycota, whose main influencing factors were alkaline phosphatase, soil organic carbon, total nitrogen, and electrical conductance. These results indicate that vegetation blanket cover can improve the physicochemical properties of soil, increase the diversity and richness of soil microorganisms, and improve the structural composition of the community, thus improving the soil environment in the mining areas in arid regions, while laying a good foundation for further restoration measures.

On 19 September 2021, a new monogenetic volcano (Tajogaite) erupted on the island of La Palma (Canary Islands, Spain). After 86 days of strombolian eruption, with emissions of volcanic material, a pyroclastic cone 200 m high and 800 m in diameter was formed. Successive lava flows descended the western slopes and reached the sea on 29 September. On descending the coastal cliffs and entering the sea, the lava flows formed two lava deltas of 75 and 5 ha, on the submarine island shelf, backed by fossilized coastal cliffs. This work presents an approach, using comparative and numerical methods, to estimate the geoheritage value and support the conservation of a new volcanic landscape in an island territory with high anthropic pressure on land uses. In a first phase, a cartographic inventory was made of all the volcanic formations similar to the new volcano in the geological domain of the Canary Islands. In a second phase, their representativeness (A), rarity (R), diversity (D), integrity (I) and observability (O) was quantitatively measured by means of dimensional estimators. The results obtained show that the new volcano presents a geological value of the first order in the context of the Canary Islands, which is one of the most prominent hot-spot archipelago worldwide. Its value is based above all on its high conservation state with respect to the similar volcanoes in the Canary Islands. The high mismatch found between the intrinsic geological value and the environmental protection of this area, justifies the development and application of urgent basic guidelines for its protection, as well as the promotion of geotourism as opposed to alternative land uses.

Global food demand is expected to increase in the next decades pushing agricultural expansion and deforestation. However, food production in agricultural lands is just one dimension of food security, to which forest goods and services also contribute. In this paper, we aimed to explore the relationship between forest cover and food security. We hypothesised that food security is improved by both human-made and green infrastructure combined. To test this relationship, we explore the relationships between forest cover and a multidimensional index of food security that included both socioeconomic and natural variables taken from Brazilian official databases for 1,141 municipalities of the Brazilian Caatinga (a seasonally tropical dry forest). The index was formed by 12 principal components axes (12 PCs) and we found that financial poverty (PC 1) and economic inequality (PC 2) were the main determinants of food insecurity in Caatinga. We found that lowest food security values were found in two contrasting contexts: one is represented by poor and unequal municipalities with high forest cover while the other refers to poor and less unequal municipalities but with little forest cover. Municipalities with intermediate levels of forest cover had slightly higher food security, a consistent pattern across time (2006 and 2017). Win-win scenarios where both forest cover and food security increased with time were almost as common as lose-lose situations (25% and 22% respectively). This suggests a sort of balance between forests and human-made land uses and reinforces that natural capital contributes to food security. Zero-hunger is a main issue for sustainable development goals and our results adds to the notion that both sustainable use of forests and socioeconomic improvements must coexist rather than being treated as antagonistic policies.

Soil phosphorus (P) is an essential nutrient that controls wetland productivity and ecological functions. However, the effects of soil P forms on P availability during wetland degradation are relatively unknown. Soil samples from differently degraded marshes, including relatively pristine marsh (RPM), lightly degraded marsh (LDM), moderately degraded marsh (MDM), and heavily degraded marsh (HDM), were collected to investigate the changes in soil P forms and its regulation on P availability in the Zoige Plateau, China. We observed that compared with RPM, the main changes in total P concentration were a significant increase of 31.6%–44.2% in the 0–30 cm soil layers of LDM and MDM, and the available P concentration increased in LDM and MDM but decreased in HDM with a lower P activation coefficient. Marsh degradation increased the concentration and proportion of dicalcium phosphates, P occluded in iron hydroxides, and organic P but decreased those of iron oxide surfaces adsorbed P and apatite P. Soil available P was mainly related to organic P and P non-occluded in iron oxide minerals that might also be non-negligible direct source of available P. The transformation from apatite P to organic P was an important regulation mechanism of P availability in soils during marsh degradation. This study revealed the risk of P limitation in heavily degraded marsh soils and established the mechanism by which marsh degradation significantly influences soil P availability. Therefore, some measure of on improving P availability should be implement for the ecological restoration of heavily degraded marsh in the future, such as grazing exclusion and the application of organic fertiliser.

Snow avalanches are a major component of the mountain cryosphere that frequently create a menace for the road network. Deposit characteristics determine the extent of damage and disruption to communication networks, but the factors controlling snow-deposit volumes remain largely unknown. This study investigates the influence of meteorological and snowpack conditions on snow-avalanche deposits and road-network vulnerability based on 1986 deposit volumes from 182 paths located in two regions of the French Alps between 2003 and 2017: Guil and Haute-Maurienne valleys. During the period, 195 avalanches impacted the road network in these areas, leading to major disruptions. In Haute-Maurienne, correlations between deposit volumes and meteorological and snowpack conditions are high in winter. However, the relationships differ with path elevation and orientation. Results do not show any significant relationship between volumes and meteorological or snowpack conditions for the spring season. Focusing on deposits that disturbed the road network, winter and spring reveal a distinct influence of meteorological and snow variables compared to the overall data set, with snowfall intensity as the predominant control variable of deposit volumes leading to road cuts. When the same analysis is conducted by considering Guil valley separately or by aggregating Haute-Maurienne with Guil valley area data, results do not show any significant relationship, highlighting the specific regional nature of relations between deposit volumes and meteorological and snowpack conditions.

One of the core challenges to achieve Land Degradation Neutrality (LDN) is to spatially identify, and strategically prioritise, the areas to implement actions to avoid, reduce and reverse land degradation. To achieve this, a tool for a participatory and data-driven assessment considering both the biophysical, and socio-economic dimensions of land degradation across scales was developed for Ecuador. In this paper, we present the methodology and results obtained, including the spatially explicit interactive tool developed to integrate indicators that support the scaling-up of sustainable land management (SLM). The process involved specialists from various national and international institutions, as well as decision makers from the public sector and other relevant stakeholders. Cloud computing allowed the integration of five main sources of data: (1) the results of a participatory land degradation assessment based on an expert knowledge questionnaire following the Land Degradation Assessment in Drylands (LADA) and World Overview of Conservation Approaches and Technologies (WOCAT) methodology; (2) the Hand-in-Hand Initiative Ecuador typology maps based on poverty maps and estimated agricultural potential and efficiency scores using household surveys on agricultural production; (3) National data sets on land cover and land use, soil properties, and hydro climatic indicators; (4) global satellite-derived LDN indicators, such as Land Productivity Dynamics; and (5) Documented SLM practices from WOCAT Global SLM Database. The tool is based on a Google Earth Engine application and allows decision makers to easily compare results and obtain statistics at different spatial scales and landscapes, including 647 Land Use Systems delimited by experts. It also includes a multi-criteria module to identify areas with specific characteristics to prioritise different types of interventions to achieve the country´s LDN targets. Convergence of local and global evidence allowed the identification of hotspots of degradation as well as areas of false positives/negatives - if only global or remote sensing indicators were considered. The participatory process contributed to strengthening multi-sector cooperation mechanisms and to guaranteeing ownership of the tool and the results. The system will support Ecuador’s efforts to monitor and report progress towards LDN to the United Nations Convention to Combat Desertification. The system´s code is shared as a repository at Earth Engine and can be adapted to and used by other countries and regions.

科学合理布设耕地质量（CLQ）监测点，可以及时准确地提供耕地质量现状和变化信息，对保障国家粮食安全具有重要意义。传统的 CLQ 监测点选择方法是基于土地利用斑块的 CLQ。由于大补丁可能有不同等级，被选为监控点会降低监控CLQ的可靠性。此外，传统的监测点部署方法主要只考虑CLQ，而忽略了道路可达性和地形等因素，导致部分监测点无法进入。因此，为了提高 CLQ 监测的可靠性，本研究提出了一种部署 CLQ 监测点的新方法。第一的，像素尺度的 CLQ 是使用遗传算法 - 反向传播神经网络 (GA-BPNN) 模型基于具有 30 m 空间分辨率的 Landstat8 数据估计的。其次，采用分层抽样模型确定最佳样本点。最后，应用改进的空间模拟退火算法（ISSA），同时考虑坡度和道路可达性，以优化监测点的位置。本研究在中国广东省广州市从化区进行。结果突出表明 (1) 与测量的 CLQ 的准确度相比，准确度 (R 应用改进的空间模拟退火算法（ISSA），同时考虑坡度和道路可达性，以优化监测点的位置。本研究在中国广东省广州市从化区进行。结果突出表明 (1) 与测量的 CLQ 的准确度相比，准确度 (R 应用改进的空间模拟退火算法（ISSA），同时考虑坡度和道路可达性，以优化监测点的位置。本研究在中国广东省广州市从化区进行。结果突出表明 (1) 与测量的 CLQ 的准确度相比，准确度 (R2  =   0.63, RMSE = 79.32, NRMSE = 13.77%) 用遥感技术估计的 CLQ 是可靠的，不同等级的像素级 CLQ 数据比斑块级 CLQ 数据更合理。(2)基于分层抽样模型，最终在研究区确定了132个监测点。（3）与空间模拟退火算法（SSA）和标准网格法相比，本研究提出的方法总分更高（F=94.61）。此外，获得的样本点主要位于道路和平坦地形附近。这样可以有效避开人迹罕至的地方。因此，基于本研究提出的新方法的结果为获得最佳CLQ监测点提供了科学依据和技术支持。

The change in the structure of the polar landscape since the termination of the Little Ice Age (ca. 1900) is expressed by widespread retreat of glaciers, progressive exposure of glacial landforms at ice margins and opening ice marginal zones to increasing paraglacial and periglacial processes operating synchronously in adjacent areas. This study provides insights into the rate of post-LIA deglaciation and associated paraglacial transformation in foreland of Gåsbreen, a glacier situated in north-western Sørkapp Land, region characterized by one of fastest deglaciation rates in the entire Svalbard Archipelago. During the investigated period, 1938-2020, Gåsbreen was in a recession that accelerated after 1990 and as a result the area of its marginal zone almost tripled from 2.2 km² to 5.8 km². This process had a significant impact on the development of the relief in glacier foreland. The dynamics of landscape transformation in the glacier marginal zone manifested in degradation in the surface of ice-cored moraines and the forms that are underlain by dead-ice. Mass movements and debris flow on ice-cored moraines and fluvioglacial processes had a great influence on this transformation. Larger volume of proglacial waters intensified denudation, transport and accumulation of sediments, which resulted, in: an increase in the surface of sandurs and proglacial riverbeds, an increase in the area of lakes, extending and changing of the course of rivers in marginal zone.

Aeolian desertification is a severe ecological and environmental problem in arid regions. Research on its temporal and spatial distribution, development model, and driving force is necessary to prevent the development of aeolian desertification. In this study, the Moltsog dune field in Mongolia and the Ujimqin dune field in China were selected as the study areas, as both contain dunes with similar natural conditions. Using Landsat data from 1988, 1995, 2002, 2009, 2016, and 2020, the spatial-temporal distribution and degree of development of aeolian desertification in the two dune fields over the past 30 years were compared. Two periods of high-resolution images were then used to compare the surface morphological changes induced by aeolian desertification in the dune fields. Climatic and socio-economic data of the same period were used to compare and analyze the causes of changes in aeolian desertification in these regions. The results show that: (1) Over 30 years, the degree and development rate of aeolian desertification in the Ujimqin dune field were generally higher than those in the Moltsog dune field, and the former had a high degree of fragmented aeolian desertification patches with an expanding range. (2) The main form of aeolian desertification is the reactivation of fixed dunes, which includes the development of blowouts on the flat grassland under the influence of human activities in the Ujimqin dune field. (3) The desertification in Moltsog is mainly affected by climatic factors, while that in Ujimqin is mainly affected by human activities. The latter is specifically affected by the high grazing intensity before 2000 and increased mining activities after 2000. These findings provide a reference for comparing the aeolian desertification process and meaningful information for preventing and managing aeolian desertification and enabling the sustainable development of dune fields in arid regions.

Planted vegetation is considered one of the most important soil-forming factors in the reclamation of degraded post-mining lands for forestry. The objective of this study was to compare the effect of N-fixing tree species and non-N-fixing species on the organic phosphorus (P org) forms in technosols developing from various substrates. Samples were taken from the uppermost layer (depth 0 – 5 cm) of technosols afforested with black locust ( Robinia pseudoaccaccia), black alder ( Alnus glutinosa), silver birch ( Betula pendula) and Scots pine ( Pinus sylvestris). Samples of the tree foliage and the O horizons were taken as well. The studied technosols developed from Quaternary sands (Sands), fly ashes after lignite combustion (Ashes) and Miocene clays (Clays). The soil samples were measured for the contents of labile (P Olab) and moderately labile organic phosphorus (P Omod), phosphorus contained in fulvic and humic acids (P fulv and P hum, respectively) and residual organic phosphorus (P Ores). The foliage and O horizon samples were measured for the concentrations of C, N and P. The N-fixing trees had higher P concentration in the foliage than non-N-fixing species. However, in the O horizon the highest P concentration was determined under birch and not under N-fixers. The effect of tree species on the organic P (P org) concentrations in the mineral soil was limited with significantly lower P org concentrations under pine. Soil under this species contained less P fulv, P hum and P Ores. However, the percentages of P Olab, P Omod, P fulv, P hum and P Ores in soil P org were nearly the same under all tree species. The largest effect on the P org concentration and on the contents of particular P org fractions was from the substrate type. Sands contained much less P org than the Clays and Ashes but their percentage of P Olab in P org was much higher than in two other substrates. We conclude that N-fixing trees do not affect the concentration of labile organic P and the major factor controlling this P org fraction is the soil substrate quality.

： The problem of grassland degradation resulting from livestock production at the expense of local grasses has become a serious environmental problem worldwide. To maintain livestock production and sustainable grassland development, a better grasp of the extent and intensity of grassland utilization needs to be realized. We developed a model to monitor the intensity of grazing in a typical grazing area (Selinco basin) of the Qinghai-Tibet Plateau. Based on combining the number of livestock at the township scale with their matching livestock habitat area and location, the biomass consumed by livestock was then assigned to the productivity supply map (NPP) by a mathematical iterative algorithm from the perspective of different foraging habits of different livestock. The objective is to more accurately measure the overall grazing pressure on the pasture as well as to estimate livestock pasture utilization. The model has confirmed that there are obvious spatial differences in the intensity of grassland utilization in the Selinco basin, which is significantly lower except in the southwestern region. Overall grazing area derived from the statistics on the pixel is 150, thousands km 2. Among them, the area of moderate grazing area is 130, thousands km 2, accounting for 87% of the total grazing area, and the area of overgrazing area is 20, thousands km 2, accounting for 13% of the total grazing area. Our model can quantify human activities spatially and provide a reliable and accurate scientific basis for livestock production development and ecological environment management.

The effects of different straw returning and nitrogen addition levels on soil quality are important for proper coastal saline soil remediation. Two maize/wheat straw returning levels (1.0 × 10 4 kg ha -1 (2S) and 5.0 × 10 3 kg ha -1 (S)) and three inorganic nitrogen addition levels (300 kg ha -1 (N2), 150 kg ha -1 (N) and 75 kg ha -1 (N1/2))—were studied, with 150 kg ha -1 inorganic nitrogen and without straw addition treatment as the control (CK), to elucidate the response of soil physical and chemical properties to the two factors. Dry-sieving technique was applied to fractionate the soils into silt-plus-clay particles (< 0.053 mm, CS), microaggregates (0.053–0.25 mm, MI), small macroaggregates (0.25–2.0 mm, SM), and large macroaggregates (> 2 mm, LM). After four consecutive wheat-maize cycles, different straw and N fertilizer treatments obviously decreased the salinity contents, increased the total nutrient contents, and optimized the soil structure of the saline soil. The saline soil reclamation effects showed significant distinctions among the different straw and N fertilizer treatments. The 2SN2 treatment displayed the greatest effects in regard to decreasing salinity, increasing the total soil nutrient contents and optimizing the soil structure, which resulted in the best remediation effect. Straw returning play a major role in decreasing soil salinity and enhancing saline soil aggregate formation. N fertilizer addition supplies rich nutrients for straw decomposition, and promotes soil microbial growth and reproduction, which brought about C sequestration in coastal saline soil. During the coastal saline soil remediation process in the Yellow River Delta, it is suggested to prioritize straw returning and moderate N fertilizer addition, and live together with moderate P fertilizer application.

Agricultural Policy has been a major driver of Land Use Changes in Portugal, especially the south-eastern Alentejo: the commons division in the early 20 th century, autarkic policies from the fascist regime, and subsidies for forestry, cereal, and cattle production since joining the common market. This study focuses on the Serra de Mértola, an area with poor soils historically subject to land degradation. Data from the Vale Formoso Erosion Centre (1961-present) was used to compute erosion rates under different land uses representative of the study area. Higher values were found in vertical Fallow (964kg/ha/yr), Wheat (90kg/ha/yr) and horizontal Fallow (66kg/ha/yr), with lower values associated with Spontaneous Vegetation (3kg/ha/yr), Quercus (4kg/ha/yr), and Pines (7kg/ha/yr). Analysis of trends in Landsat NDVI values allowed to discriminate between land cover change and natural oscillations in weather and vegetation growth: in dry years NDVI values decrease for winter and spring, maintaining lower values throughout the year, while natural vegetation recovery shows a slow but steady increase until it stabilizes or breaks due to human action. Breaks in NDVI values were correlated with occasional ploughing in pastureland and spontaneous vegetation as well as the different moments in the agricultural year for cereal production. Three tendencies were observed: in Cultivated areas, ploughing provides higher sediment availability in valley floors, resulting in aggradation; Pastureland shows rock fragments on slopes and lower sediment availability, leading to incision on valley floors; Spontaneous Vegetation/Abandoned Land leads to a sharp decrease of sedimentary yield, with incision until its natural growth stabilizes hillslope processes.

The sulfur cycle is one of the geochemical element cycles in which microorganisms play a key driving role. The microbial function of soil S cycling in response to desert degradation, however, remains largely unknown. We used metagenomics to analyze the characteristics of microbial communities and their functional genes involved in the S cycles under natural water gradients with three typical halophytes shrubs in the Ebinur Lake Basin Desert, China. Our results showed that the rhizosphere effect, water gradient, and halophyte type played a major role in shaping the S cycle. On the whole, in the rhizosphere type and low water environment, the functional genes involved in the S cycle had high abundance, and the SOX system in Alhagi sparsifolia had a high expression level. In the S cycle network structure, as the soil water content decreased, the complexity in S gene networks increased, showing the characteristics of clustering and high connectivity. Indicates the strengthening mode in microbial interactions with the water content. Interestingly, the negative correlation of the network changed with the water content, and there was more competition among communities under the low water gradient and more cooperation under the high water gradient. Through the correlation between environmental factors and the network, nitrate (NO 3 −) and soil available S (AS) constrained most S gene ecology networks. The key species involved in the S cycle were halophilic microorganisms. These results can enhance the understanding of soil S biogeochemical processes and contribute to the mitigation of desertification by improving soil conservation.

Vegetation restoration in ecologically fragile areas has a significant carbon sequestration effect, and the driving factors affecting it are complex. In particular, it is usually difficult to achieve quantitative assessment at the regional scale for this part of human activity intervention in the ecological environment. The Chinese government’s ex-situ poverty alleviation and relocation project has relocated approximately 10 million people from areas with a fragile ecological environment to urban centralized resettlement, which is a typical case of weakened environmental intervention by human activities. Guizhou Province, an ecologically fragile region in southwest China, was selected as the study area, with a relocated population of 1.92 million, accounting for approximately 20% of the total relocated population in China. The carbon sequestration (CS) model of vegetation photosynthesis and spatial analysis of geographic information were used to quantify the contribution of human activities to the natural restoration of vegetation carbon sequestration at the regional scale caused by the weakening of environmental interventions, based on the data of net primary productivity (NPP) of vegetation from 2000 to 2020. The results show that the implementation of the ex-situ poverty alleviation and relocation project acts as an external force to drive vegetation restoration and carbon sequestration, which increases the slope of the carbon density change trend (from k=30.9 to k=57.41), resulting in an overall carbon density increase of 26.51 tCkm -2. The results of the regional spatial analysis showed that the correlation coefficients between carbon density and relocation intensity in the 5-year and 10-year change intervals before and after relocation were r=0.976 (p<0.01) and r=0.949 (p<0.05), respectively, indicating a significant positive correlation between vegetation carbon sequestration and relocation intensity. Based on this, the carbon sequestration contribution of vegetation in 84 counties (districts) in Guizhou Province that implemented ex-situ poverty alleviation and relocation projects was calculated, showing that 79 counties (districts) contributed positively, accounting for 94%. The average contribution of carbon sequestration by vegetation restoration in each county (district) was 0.0556 Tg, and the CO 2 emissions during the emission reduction period were 0.2059 Tg. The other five counties (districts) with a negative contribution to carbon sequestration were distributed in regions with relatively stable ecosystems and mature forests. This shows that human intervention in the environment changes significantly in fragile ecological areas.