Factors Controlling Soil Organic Carbon Sequestration of Highland Agricultural Areas in the Mae Chaem Basin, Northern Thailand

: Understanding the effect of the environment, crop types, and land management practices on the organic carbon sequestration of top soil is crucial for adopting management strategies in highland agricultural areas. The objectives of this study are: (1) to estimate top soil organic carbon density (SOCD) of different crop types and (2) to analyze the factors controlling top SOCD in highland agricultural areas. The top soil layers from 0 to 30 cm depths were collected from the Mae Chaem basin, Northern Thailand. The results showed that the highest top SOCD was found soil used for growing upland rice, which contained an average of 58.71 Mg C ha − 1 . A significant difference between the top SOCD was detected between areas where minimum tillage and conventional tillage of various crops, with average of values 59.17 and 41.33 Mg C ha − 1 , respectively, for areas growing strawberries; 61.14 and 37.58 Mg C ha − 1 , respectively, for cabbage, and 71.15 and 39.55 Mg C ha − 1 , respectively, for maize. At higher elevation, the top SOCD was high, which may be due to high clay content and low temperature. Increased use of chemical fertilizers lead to increases in top SOCD, resulting in increased crop yields. Elevation, bulk density, N and K 2 O fertilizers were the main factors controlling the top SOCD at all sites. gooseberry, an ha , respectively. Higher elevation and application rates of N, 2 , K 2 O fertilizers could lead to increases in top SOCD. A positive relationship between all crop yields and top SOCD was detected, indicating the increasing of top SOCD can increase crop yields. We conformed that minimum tillage methods can preserve top SOCD and enhance crop yields. Multiple regression analysis indicated that elevation, bulk density, N and K 2 O fertilizers the main factors influencing changes in top SOCD. Practicing minimum tillage, applying N and K 2 O fertilizers could be beneficial for SOC sequestration, which is technically applicable for highland agricultural areas.


Introduction
Land degradation problems are mainly related to land use changes [1], which result in reduced crop yields and declining soil quality [2]. Soil organic carbon (SOC) can play a crucial role in increasing crop yields, enhancing soil fertility [3][4][5] and mitigating greenhouse gas emissions [6] by serving as potential carbon sinks, and providing improvements in soil structure and soil water storage [7]. Low levels of SOC is caused by having a poor ability to store nutrients in the soil and may result in decreased crop yields, significantly affecting soil fertility and causing nutrient loss [8]. Several studies have examined the potential of SOC on decelerating climate change and encouraging plantation crops. For example, a report claimed that an increase of SOC in the upper soil layer (0-20 cm) may result in increased grain yields of 430 kg per hectare and 3.5% reductions in crop instability [9]. However, other studies argued that the potential to increase SOC is often variable, depending on climatic conditions and soil types [10][11][12]. As a result, it is necessary to identify the relationships between carbon input, SOC accumulation, and crop yields.
In highland areas, geographical factors play a crucial role in determining the distribution of heat, water, clay minerals, and nutrients, all of which affect both the decomposition and accumulation of

Study Area
The location of the Mae Chaem basin studied is 18°6′0′′ and 19°10′0′′ N between 98°4′0′′ and 98°34′0′′ E, covering an area of 3,853 km 2 . This area is in the Chiang Mai Province in the Northern part of Thailand. The highest point is the Doi Inthanon (Mount Inthanon), 2,535 m above sea level (a.s.l.), which is the highest peak of Thailand. The lowest elevation considered is 282 m a.s.l., which will allow researchers to compare various climates of different elevations [32]. These geographic parameters result in a wide spatial distribution of precipitation [33]; the annual precipitation was 1,145 mm (wet season from about May to September) and the average temperature was 27.5 °C in 2010 [34].
3 Vegetation types are varied following different elevations. Hill evergreen forest occurs above 2,500 m a.s.l. Tropical mixed pine forest ranges 900 to 1,500 m a.s.l., while dry dipterocarp and mixed deciduous forests are found below 1,000 m a.s.l. [35].
Based on World Reference Base for Soil Resources (WRB), the dominant soil is Acrisols [36]. Soil from the area is reddish-brown lateritic, and the soil texture being mostly sandy loam and clay. A high clay content is at the deeper soil layer. The soil bulk density range is 1. 16-1.41 Mg m −3 and the maximum water holding capacity is 41.42%-56.78%. The soil is generally acidic, with pH ranging between 5.0 and 5.8. The organic matter range is 2.60-3.86% [37].

Soil Sampling and Analysis
Soil from a total of 138 sites were collected from the Mae Chaem basin within the study area at elevations of 400-1,500 m a.s.l.. The elevation was measured in each site because it was used as a parameter for controlling temperature and precipitation. At each site, soil samples from 0-to 30-cm depth were identified the color and texture. Then, three soil samples of each site were collected by using soil auger. Soil samples from nine crop types; cabbage (18 sites), upland rice (13 sites), lowland rice (31 sites), tomato (7 sites), strawberry (7 sites), potato (7 sites), Japanese pumpkin (6 sites), cape gooseberry (5 sites), and maize (44 sites) were collected during January 2017-June 2019 ( Figure 1) and compared. However, the number of each crop sites were different because some crop types were grown by a small number of farmers. A soil moisture meter was used to measure soil moisture in the field. The physical and chemical properties of whole soils were determined using the procedures described by the United States Department of Agriculture (USDA), National Soil Survey Center [38]. Soil bulk density was determined as the dry weight per unit volume of the soil core after a 24-hr drying period in an oven at 105°C [39]. Soil texture was measured by a hydrometer, and soil pH was determined in a 1:2.5 soil 5 to water mixture using a pH meter. Organic carbon was determined by the method of Walkley and Black [40].

Estimating Soil Organic Carbon Density (SOCD)
SOCD was calculated using the following equation [41]: where SOCD is the total amount of SOC at depth h per unit area (Mg C ha −1 ); is the number of layers considered; represents the i th layer; is thickness (cm); is SOC content (g kg −1 ); is bulk density (Mg m −3 ); and is the proportion (%) of coarse (> 2 mm) fragments in the i th layer

Management Practices
Management practice data were obtained from the farm owners at each sampling site. At each site, the farmer was asked about crop type, tillage methods, rates and types of application of fertilizers, manure, crop residues management, and crop yields.

Statistical Analysis
SPSS Version 20.0 (Chicago, IL, USA) was employed to analyze statistics including mean and standard deviation, while the Kolmogorov-Smirnov (K-S) test was run to determine whether the distribution of quantitative data was normal. The t-test was conducted to identify differences in top SOCD across all tillage methods. The variance inflation factor (VIF) and tolerance (TOL) indices were used for testing the collinearity diagnostics between independent variables. A clear signal of multicollinearity was considered if VIF < 10 and TOL > 0.10 [42]. Pearson's correlation was used to identify the relationship between two variables by fitting a linear equation. If the correlation value of any variable > 0.8, it was not considered due to avoiding the multicollinearity among variables.
Stepwise multiple regression analysis was conducted to investigate the relationship between top SOCD and environment (elevation), fertilization, and crop yields.

Soil Properties, Management Practices, and Top SOCD in Agricultural Highland Areas
In general, the bulk density range was 1.10-1.55 Mg m −3 ; the soil moisture range was 6.23-30.21%; sand accounts for 40.42-58.0% of soil content; silt accounts for 15.54-20.80% of soil content. Clay accounts for 21.48-39.51% of soil content, so the majority of the soil content studied here was sandy clay loam. However, the soil texture of Japanese pumpkin and maize areas were clay loam, while the upland rice area was predominantly sandy clay. Fine particles (clay plus silt) accounts for 45.57-62.39% of the soil content; and the pH range was 5.40-5.97 (Table 1). Elevation, bulk density, and clay content were significantly positively correlated with top SOCD (r = 0.522**, 0.210* and 0.367**, respectively), whereas sand content was significantly negatively correlated with top SOCD (r = −0.351**). Elevation was significantly negatively correlated with bulk density and sand content (r = −0.176*, and −0.213*, respectively) and a significantly positively correlated with clay content (r = 0.203*) ( Table 2). Three types of soil tillage used for land preparation were identified in our study: 1) conventional tillage ( Figure 2 (a)); the practice in which farmers use machines such as the tractor and discs to turn over and loosen the soil before planting cabbage, lowland rice, strawberry, potato, and maize (Table  3); 2) minimum tillage ( Figure 2 (b)); the practice that minimizes the disruption of soil by using special equipment (e.g., hoe) to plant cabbage, tomato, strawberry, and maize (Table 3); and 3) no-till ( Figure  2 (c)); the practice that causes even less disturbance to the soil in which as little soil as possible is disturbed in the process of sowing seeds. The no-till method was used for planting upland rice, tomato, Japanese pumpkin, and cape gooseberry. The strawberry plantation applied the highest average amount of N, P2O5, and K2O fertilizers, which were 567.7, 363.27, and 957.04 kg ha −1 , respectively. In contrast, the lowest average levels of N (15.12 kg ha −1 ), P2O5 (7.66 kg ha −1 ), and K2O (4.34 kg ha −1 ) fertilizers were used in upland rice plantations. Maize residue was burnt after harvesting, while other crop areas were left in the fields (Table 3).    The highest top SOCD was found in upland rice area with the average of 58.71 Mg C ha −1 . This might be due to no-tillage practices combined with the fact that the practice of burning of rice residue was not employed, these characteristics, taken together, cause minimal disturbance to the soil and crop residues were not removed from the field. The amount of top SOCD in areas used to grow maize were the second, after upland rice, followed by strawberry, cabbage, tomato, lowland rice, Japanese  (Table 3).

Relationships Between Top SOCD and Elevation
The results showed that the top SOCD had a positive relationship with elevation at all sites (R 2 = 0.19; p < 0.01; Figure 3), explaining the observation that at high elevations, levels of top SOCD were also high. However, altitude is not a variable directly influencing top SOCD, but a variety of climatic functions associated with different aspects of altitude do affect SOC budget.

Relationships Between Top SOCD and Chemical Fertilizers
Application rates of N, P2O5, and K2O fertilizers had a positive relationship with top SOCD (Figure 4), indicating that increasing of chemical fertilizer application can lead to increases in top SOCD. N fertilizer use here had the highest positive relationship with top SOCD in the strawberry, cabbage, lowland rice, and upland rice areas, with the R 2 values of 0.93 (Figure 4(e)), 0.62 ( Figure  4(a)), 0.59 (Figure 4(c)), and 0.52 (Figure 4(b)) (p < 0.01), respectively. The P2O5 fertilizer could explain 44% and 35% of the changes of top SOCD observed in the areas growing cabbage and upland rice, respectively (Figure 4(a) and (b)), while the 39% and 38% could be explained by the amount of K2O fertilizer used to grow in areas planted with lowland rice (Figure 4(c)), and cabbage (Figure 4(a)), respectively.
We found that only areas planted with upland rice, tomato and potato applied cattle manure, with average levels applied being 0.98, 0.87, and 1.02 Mg ha −1 , respectively ( Table 3). As for manure application, however, this study did not detect any significant relationships between manure and any of the tested soil or geographic attributes. This may be because fewer farmers applied the cattle manure than other fertilizers here.

Relationships Between Top SOCD and Crop Yields
As shown in Figure 5, crop yields were positively correlated with increasing levels of top SOCD. This finding could explain the observation that cabbage yield increases 53% ( Figure 5(a)), followed lowland rice, upland rice, and maize yields, at 52% ( Figure 5(c)), 51% ( Figure 5(b)), and 50% ( Figure  5(i)), respectively. These results indicate that increasing of top SOCD can increase crop yields.

Top SOCD and Crop Yields Under Different Tillage Methods
Based on our survey, conventional tillage and minimum tillage methods were used to maintain cabbage, strawberry, and maize plantations for land preparation. Results here revealed that there was significant difference between top SOCD maintained by conventional tillage and minimum tillage methods (p < 0.01) in areas growing cabbage and maize ( Figure 6). The top SCOD was 37.58 and 61.14 Mg C ha −1 under conventional tillage and minimum tillage methods in areas growing cabbage, respectively, while levels were 39.55 and 71.15 Mg C ha −1 were found using conventional tillage and minimum tillage methods, respectively, in areas growing maize. The top SCOD in strawberry areas, moreover, contained 41.33 and 59.17 Mg C ha −1 under conventional tillage and minimum tillage methods, respectively, which also produced a significant difference when compared to conventional tillage and minimum tillage methods (p < 0.1) ( Figure 6).

14
Differences in cabbage and strawberry yields, when compare to the use of either conventional tillage or minimum tillage methods, were highly significantly different (p < 0.05) (Figure 7). The average cabbage yield was 17.38 and 18.71 Mg ha −1 , while strawberry yield was 18.20 and 19.58 Mg ha −1 under conventional tillage and minimum tillage methods, respectively. However, no significant differences were observed when comparing conventional tillage and minimum tillage methods used to grow maize, with yield values of 4.32 and 4.58 Mg ha −1 , respectively (Figure 7). 15 Figure 7. Comparison of yield between conventional tillage and minimum tillage of cabbage, strawberry, and maize areas. *, **, *** Significant difference at 10%, 5%, and 1% level, respectively

Factors Controlling Top SOCD and Crop Yield in Highland Agricultural Areas
The TOL and VIF values were detected, with ranges of 0.556 to 0.905 and 1.074 to 1.798, respectively, when SOCD was assigned as the dependent variable. Ranges of 0.630 to 0.980 and 1.030 to 1.587 were found for TOL and VIF values, respectively, when yield was assigned as the dependent variable. These indicated that there was no signal of multicollinearity in this study.

17
In areas where cabbage is grown, chemical fertilizers (N, P2O5, and K2O fertilizers) largely governed level of top SOCD, while top SOCD was the main factor influencing cabbage yield. Top SOCD in upland rice was controlled by bulk density, clay content, soil moisture and sand content, while top SOCD and bulk density were two major factors controlling upland rice yield. For lowland rice, N fertilizer, soil moisture and K2O fertilizer played important roles in explaining top SOCD, while elevation and SOCD control lowland rice yield. In addition, elevation and P2O5 fertilizer were two key factors controlling top SOCD in areas used to grow tomatoes, but there was no factor that could predict changes in tomato yield.
Changes in top SOCD of areas growing strawberries were largely related to N and P2O5 fertilizers, while soil moisture, clay content and P2O5 fertilizer were largely responsible for changes in strawberry yield. In areas where potatoes are grown, top SOCD and yield were controlled by elevation and N fertilizer, respectively. Silt content was main factor controlling top SOCD in Japanese pumpkin agricultural areas, while soil moisture, clay plus silt content, and P2O5 fertilizer levels were the main factors influencing Japanese pumpkin yield. Changing of top SOCD in areas growing maize was mainly controlled by elevation and N fertilizer, while top SOCD, elevation, N fertilizer, and clay plus silt governed maize yield. However, no factors were detected influencing top SOCD and yield changes for cape gooseberry in this study (Table 4). This might be due to the relatively low number of cape gooseberry areas growing that were sampled.

Environmental Factors Affecting Top SOCD in Highland Agricultural Areas
SOC stocks can differ significantly due to the influence of local factors such as topography, soil properties, or land use and management [43][44][45]. Previous studies reported that elevation is significantly correlated with SOC content [46][47][48][49], which is consistent with the results of the present study in which top SOCD increased with increasing altitude (Figure 3). This is because the accumulation of SOC at a higher altitude may potentially result from decreases in temperature and increasing levels of soil moisture [50][51][52]. Pabst et al. [53] confirmed that elevation can be used as a parameter to control temperature and precipitation, which are probably the major factors influencing SOC stocks along the investigated elevation gradient. Increasing soil temperature accelerates the rate of SOC decomposition [54,55], improve the level of soil respiration [56], and lead to a significant decrease in SOC content [57][58][59]. On the other hand, SOC content is positively correlated with annual mean precipitation, which increases soil moisture [58,60]. Moreover, Florinsky [61] reported that topography influenced SOC content more than climate and parent material in the A horizon because of gravity-driven lateral migration, the accumulation of water and differences in the temperature of different slopes. However, carbon loss through erosion is mainly related to topography [62]. With increasing slope, infiltration decreases because increasing slope area and the velocity of water flow leads to decreasing soil moisture content in the deeper layers through increasing runoff and evaporation rate [61]. However, our results showed a low positive relationship between top SOCD and elevation at all sites (R 2 = 0.19; p < 0.01) (Figure 3), indicating that top SOCD in agricultural areas may be influenced by management measures (tillage methods and chemical fertilizers) than elevation did. Unlike forests that have less human interference, agricultural land is often disturbed by land management. Interestingly, we found that the elevation was included in the regression models of tomato, potato, and maize areas, accounting for 42% of the total sampling sites of this study (Table  4). This can be explained by the high values of standard deviation of elevation as shown in Table 3, which indicated the soil samples of these areas were investigated from the wide range of elevation (892 to 1603 m a.s.l. for tomato areas, 701 to 1178 m a.s.l. for potato areas, and 498 to 1436 m a.s.l. for maize areas). The results found that at the high elevation, the value of SOCD was high in tomato, potato, and maize areas. Therefore, our results demonstrated that the elevation acted as the main factor controlling SOCD in these areas even the same crop type is grown, which is in line with Tsui et al. [63] who concluded that elevation is a simple and effective predictor of SOC stock.

Effect of Burning Crop Residues on Top SOCD
Fire has been used as a tool to fertilize soils and control plant growth, but it can accelerate soil erosion and cause desertification [64]. This is similar to the Mae Chaem basin, where farmers usually choose to burn maize residue and leave the land fallow for 2-3 months before preparing the land for the next crop. However, the average values of top SOCD in areas growing maize was the second highest after upland rice (52.48 Mg C ha −1 ) (Table 3), compared to the other areas in this study, which is due to highest clay content (39.51%) was detected in maize plantation (Table 1). Moreover, it may be due to elevation (climate factors), and burning maize residue (Table 3) may increase the top SOCD in this area. The possible reasons to explain why burning maize residue increases the top SOCD. Firstly, fire can increase SOC [65,66] by transformation the labile compounds into recalcitrant organic forms [67], which are components of long-term carbon sequestration [68]. Secondly, fires increased the microbial metabolism after burning [69], increased degradation of soil organic matter and mortality of microbes, resulting in increased carbon mineralization, ammonium nitrogen (NH4-N), available phosphorus (P), and exchangeable potassium (K) and calcium (Ca) into the soil [70]. Notwithstanding, frequent burning of crop residues in highland can cause soil degradation [71], resulting in the loss of top SOCD due to surface runoff, water erosion [72], and sedimentation processes [73] as well as fires increase CO2, oxides of nitrogen, and black carbon into the atmosphere [74].

Effects of Physical and Chemical Soil Properties on Top SOCD in Highland Agricultural Areas
Top SOCD had a positive correlations with elevation, bulk density, and clay content (Table 2), which is in line with other studies [75][76][77] that reported the positive effects associated with higher proportions of clay on top SOCD levels, which are likely due to the ability of clay particle to absorb organic matter and protect it from microbial decomposition [78]. Moreover, clay content and iron (Fe) oxides affect carbon turnover in soil [79,80], which is dependent on regional climate and varies with topographic positions [80], thus may affect soil properties, plant biomass, and SOC stocks [81]. Therefore, soil texture must be considered for preservation and sequestration of SOC in cropland.
We also found higher soil bulk densities at lower altitude sites due to the negative correlation detected between bulk density and elevation ( Table 2), indicating that the soils with high SOCD contain high organic matter accumulation and high percent pore space, resulting in lower soil bulk density. The study of Kumar et al. [82] supported our finding that at high elevations with low bulk density indicate high coarse structure of organic matter in which increase the spaces for SOC accumulation. Also, at high elevation, clay content was high, but sand content decreased (Table 2), which may be due to the availability and movement of water that moved the sand particles to the lower areas, resulting in the larger proportion of clay particles that remained. These factors contribute to increasing top SOCD with increasing altitude.

Effects of Fertilizers on Top SOCD
As shown in Figure 4, we found greater top SOCD when increasing amounts of chemical fertilizers were applied, particularly with respect to increasing the use of N and K2O fertilizers, confirming findings of previous studies which showed that not only chemical fertilizer, but also manure, could increase soil organic matter [83]. Unfortunately, a few areas in this study were identified in which famers applied manure, but we were able to observe that applying cattle manure could lead to high top SOCD, especially in areas where upland rice is grown (Table 3). Although chemical fertilizers alone may accelerate microbial activity and decomposition of soil organic matter, manure plus chemical fertilizers had greater effects on soil organic matter [84].
However, chemical fertilizers did not directly increase organic matter in the soil. They only enhanced carbon input by increasing root and stubble biomass retention (enhanced net primary productivity, above ground biomass and underground roots) [85]; as a result, SOC increased when crop residues remained in the fields.

Effects of Tillage on Top SOCD
Tillage is practiced to aerate and incorporate the soil, crop residue, manure, and fertilizer into the root zone [86]. It increases aeration and breaks up aggregates, exposing organic matter to microbial attacks that accelerate soil organic matter decomposition rates of the top soil [87]. In our study, we found the top SOCD under minimum tillage conditions was higher than top SOCD when conventional tillage was used in areas growing cabbage, strawberry, and maize ( Figure 6), which is consistent with the previous studies reporting that minimum tillage/no-tillage can greatly increase SOC stock, particularly in the top soil layer (0-30 cm) compared with conventional tillage because it disturbs the soil physical aggregate structure less [88][89][90]. When conventional tillage in highland agricultural areas is used, soil erosion and nutrient losses are major problems due to low vegetation cover during heavy rainfall [91,92]. Biddoccu et al. [93] found that conventional tillage and reduced tillage caused the soil losses about 111.5 and 207.7 Mg ha −1 , respectively, while only 25.6 Mg ha −1 was observed in controlled grass cover plot in a hilly vineyard (Piedmont, NW Italy) that can reduce runoff and soil losses by 68% and 61%, respectively [94].

Effects of Top SOCD and Crop Yields
Crop yield stability is an essential indicator of agricultural sustainability [12], where increasing SOC stocks can reduce yield variation [9,95]. In this study, all crop yields had a positive relationship with top SOCD ( Figure 5). This is also consistent with other studies who reported that increased SOC can increase crop yields through inputs of organic materials (manures, crop residues) [96][97][98][99], and reducing the practice of intensive tillage and crop residue burning [100]. Top SOCD and crop yield may also vary depending on other factors such as altitude, rainfall, and agricultural practices, which is in agreement with a study by Lal [101] who reported that increasing top SOCD 1 Mg C ha −1 in the root zone can increase annual food production by 30-50 million Mg in developing countries. Qiu et al. [102] predicted increases in grain yield of 454 kg ha −1 in Huantai country because of increases of 1 g C kg -1 SOC.
Differences in crop types and crop production affect SOC storage due to differences in the amount of crop residues that are left in the fields, and belowground biomass (roots) that remains in the soil. Eghball et al. [103] found that greater SOC storage could be observed under continuous corn than corn-soybean rotations because corn left behind more residues in the production process.
Upland rice areas showed the highest amounts of top SOCD compared with other crops evaluated (Table 3). This is because rice residues left in the fields after harvest can cause carbon and nitrogen increased in the soil, thus enhancing the entire ecosystem [104,105]. Studies have claimed that SOC stocks in the upper 20 cm of soil have globally increased by 0.24-0.46 Mg C ha −1 year −1 due to the levels of compost applied in the last 10 years [106]. Another report added that in the southern part of China, SOC stocks in the upper soil layer (0-20 cm) have increased by 3.8 Mg C ha −1 with a 22year-long application of compost, in comparison with soil that only received chemical fertilizers [98]. In addition, it was estimated that SOC stocks will increased by 3.8 Tg C year −1 in soil that is incorporated with post-harvest rice straw [104]. Another feature of the Mae Chaem basin is that upland rice is often grown only once per year, leaving farmers in no rush to clear their rice fields for the next crop by burning or removing rice residues, thus allowing organic matter is allowed to accumulate in the soil and make soil carbon densities higher in upland rice farms than in farms of other crops. This finding agrees with the study of Stan et al. [107], which shows that crop residues that are not burned or removed from the field will contribute to the maintenance of SOC, enhance soil fertility and biological activity.

Conclusion
The highest top SOCD was found in areas growing upland rice with the average of levels reaching 58.71 Mg C ha −1 . Areas growing maize were the second highest levels of top SOCD, after upland rice, followed by strawberry, cabbage, tomato, lowland rice, Japanese pumpkin, potato, and 20 cape gooseberry, with an average top SOCD level of 52.48, 49.0, 48.05, 42.14, 35.21, 34.0, 29.45, and 19.56 Mg C ha −1 , respectively. Higher elevation and application rates of N, P2O5, and K2O fertilizers could lead to increases in top SOCD. A positive relationship between all crop yields and top SOCD was detected, indicating the increasing of top SOCD can increase crop yields. We conformed that minimum tillage methods can preserve top SOCD and enhance crop yields. Multiple regression analysis indicated that elevation, bulk density, N and K2O fertilizers the main factors influencing changes in top SOCD. Practicing minimum tillage, applying N and K2O fertilizers could be beneficial for SOC sequestration, which is technically applicable for highland agricultural areas.