Characteristics of Soil Organic Carbon Distribution in Different Economic Forests in Gangu County, Gansu Province, China

In this study, five types of economic forest plots were selected as sample plots (Malus pumila, Juglans regia, Zanthoxylum bungeanum, Prunus persica, and Prunus armeniaca) in Gangu County, Gansu Province, and wasteland was used as a control to investigate the changes of soil organic carbon content in 0–100 cm of different economic forests in Gangu County. The results showed that the soil organic carbon content of different economic forests in Gangu County ranged from 12.65 ± 0.09…7.20 ± 0.13 g/kg in layer 0–10 cm. The soil organic carbon content at 0–100 cm depth ranged from high to low: apple (Malus pumila), pepper (Zanthoxylum bungeanum), walnut (Juglans regia), apricot (Prunus armeniaca), peach (Prunus persica) and wasteland. The SOC content of the five forest types and the wasteland showed apparent surface aggregation, with the highest SOC content of 17.20 g/kg for Malus pumila and the lowest SOC content of 12.65 g/kg for the wasteland in the 0–10 cm surface layer; in the 80–100 cm deep layer, the highest SOC content of 14.90 g/kg for Juglans regia and the lowest SOC content of 9.17 g/kg for the wasteland. SOC content was positively correlated (p < 0.05, n = 126) with soil water content, soil enzyme activity and soil microbial population, and significantly negatively correlated (p < 0.05, n = 126) with pH and soil bulk density. Overall, the mean value of SOC content in 0–100 cm soils of Malus pumila and Zanthoxylum bungeanum is higher among the five economic forest species in Gangu County. The mean SOC content of Malus pumila and Zanthoxylum bungeanum in 100 cm soil is relatively high among the five economic forest species, therefore Malus pumila is the preferred choice for economic forest planting in Gangu County, and in the future economic development and ecological restoration practice of Gangu County, it is recommended to plant a mixed forest planting pattern of Malus pumila and Zanthoxylum bungeanum, which is conducive to improving the potential carbon sink function of economic forests in the area.


INTRODUCTION
The soil carbon pool is the largest reservoir in terrestrial ecosystems, about 3.3 times larger than the atmospheric carbon pool and 4.5 times larger than the biogenic carbon pool.Among these, soil organic carbon pools, an important indicator of soil fertility and one of the significant carbon pools in terrestrial ecosystems, amount to 1220-1576 Pg above 1 m globally [34].Soil organic carbon is essential for providing soil nutrients and reducing erosion [41].To cope with climate change and improve fragile habitats, since the late 1970s, China has implemented national projects such as the "Three Norths" afforestation project  and the "Return of Cropland to Forests (Grass)" project (1999-2020) [4,12].Planted forests have increased ecosystem productivity and soil carbon stocks in fragile habitats and have slowed down soil erosion and soil erosion to some extent [33].Assessing different vegetation types' soil organic carbon (SOC) content contributes to carbon sequestration and ecological improvement [21].
SOC content varies depending on the type of soil or land use [2].Differences in vegetation type also affect SOC content, as plant carbon sequestration varies due to different vegetation productivity.In contrast, the quality and quantity of apoplastic matter vary between vegetation, and the fraction returned to the soil is affected by microbial influences that further affect SOC.Wang [19] studied the carbon storage capacity of artificial lateral cypress and acacia growing on the Loess Plateau and concluded that the carbon sink capacity of acacia was higher than that of lateral cypress in the same geographical area.Fan [36] showed that the SOC content of evergreen broadleaved forests was higher than that of pure fir plantations of the same age.Wan [27] found that the SOC content of typical vegetation types in the upper reaches

SOIL CHEMISTRY
of the Qilianshan Shiyang River Basin was alpine meadow > Picea crassifolia > Populus simonii > subalpine shrub.Luo [13] used a typical vegetation community in a semi-arid loess area as a study object and found that SOC content was highest in Populus simonii, followed by barren grassland, and smallest in Caragana korshinskii.
Gangu County is a central agricultural county in northwest China and one of the base counties of Gansu Province's substantial plateau summer dish "Western Cuisine to East Transfer" [41].At this stage, understanding the content level and distribution of SOC in Gangu County has important practical significance for improving soil carbon sequestration potential and guiding agricultural production.At present, relevant research mainly focuses on the change of soil organic carbon content in different land use modes (such as grassland, forest, wetland, etc.) [29], but the effect of carbon sequestration and sink enhancement of different economic forest soils, and its influencing factors are still unclear.Therefore, based on the distribution characteristics of SOC (0-100 cm) under different economic forest types (Malus pumila, Juglans regia, Zanthoxylum bungeanum, Prunus persica, Prunus armeniaca, wasteland), this study explored the difference of SOC content of different economic forest vegetation types and its relationship with other environmental factors, providing a scientific basis for the optimal selection, rational management and soil carbon sequestration and sink enhancement of economic fruit forests in Gangu County.

Study area.
Gangu County (34°32′-35°03′ N, 105°02′-105°31′ E) in Tianshui City, Gansu Province, is located in the southeast of Gansu Province, upstream of the Wei River, and belongs to the semiarid region of the Loess Plateau.The county has an average altitude of 1972 m and a semi-humid continental monsoon climate, with an average annual temperature of 11.5°C, including the highest temperature in July, with an average of 25°C and the lowest temperature in January, with an average of -1°C.The rainfall is low in spring and high in summer, with an annual rainfall of less than 500 mm [9]; the annual sunshine is about 2100 h; the soil in the territory is The DEM (Digital Elevation Model, DEM) is an irregular triangular network derived from contour lines spaced 1 m apart using ArcGIS 10.2 (Fig. 1).The soil types in Gangu County are complex and diverse, with widespread loess, deep cover, vertical joints and fissures.Soil types include: Eutric Cambisols, Eutric Luvisols, Chernozems, Calcic Cambisols and so on (Reports on World Soil Resources, no.106.FAO, Rome).Among them, Eutric Cambisols accounts for 1.79% of the total area, Eutric Luvisols accounts for 8.79%, Chernozems accounts for 12.08%, and Calcic Cambisols accounts for 60.69%.Soil sampling.On the basis of field research, we followed the principles of science, reasonableness and representativeness to ensure that the sampling sites could reflect the land cover of different economic forest vegetation.Finally, we selected Malus pumila, Juglans regia, Zanthoxylum bungeanum, Prunus persica, Prunus armeniaca and wasteland as the research sample sites.Wasteland refers to land that has been cultivated previously but has been abandoned.The study sample sites are all in the Loess Plateau area, the soil types are all Calcic Cambisols, the mineral composition is dominated by quartz and feldspar with little variation among the layers, and the study sample sites have the same climatic conditions, water system conditions and land use history.Seven 20 × 20 m sample plots were randomly selected for each economic forest species and wasteland, and the soil samples were sampled in October 2020, with the last local rainfall on October 13, and the sampling started after one week of dry weather.According to the actual area and topography of each sample plot, three small sample squares of 5 × 5 m were selected along the diagonal (ends and midpoints), avoiding fertilizer ditches (pits) and irrigated wetlands, and soil was collected according to the depths of 0-10, 10-20, 20-40, 40-60, 60-80 cm, and 80-100 cm, respectively (Fig. 2).Three parallel samples were collected for each soil layer.After removing plant roots and large stones, the soil samples were placed in self-sealing bags and numbered.While sampling, information such as latitude, longitude and altitude of the sampling site was recorded with a handheld Global Positioning System (GPS), soil temperature at the sampling site was recorded with the aid of a soil thermometer, and the basic conditions of the sample site were recorded.
Laboratory analyses.A soil depth of 100 cm is likely to reflect the main root distribution of the forest [18].Soil samples were taken for a profile depth range of 0-10, 10-20, 20-40, 40-60, 60-80, and 80-100 cm, using a cutting ring whose volume was equivalent to 1 × 10 -4 m 3 [26].After labeling, soil samples for soil bulk density (SBD) and soil water content (SWC) were transported to the laboratory and oven-dried to constant weight at 105°C.Soils were air-dried and stored in a dark place.After removing impurities such as root material, soils were processed using a 2 mm mesh sieve.SOC was determined using a potassium dichromate oxidation ferrous sulfate titration method [14].Electrical conductivity (EC 1:5 ) and pH were analyzed in a 1 : 5 soil:water extract using a conductivity meter (DDS-307, Leici, China) and a pH meter (PHS-3C, Leici, China), respectively.The soil microorganisms were cultured by the dilution pouring coating plate method [23].Bacteria were counted using beef paste peptone medium in an inverted incubator at 37°C for 48 h (DNP-9162, Ningbo le electrical instrument manufacturing company, Ningbo).Three dilutions of 10 -4 , 10 -5 and 10 -6 were selected.The soil fungi were incubated in inversion for 3-4 d at 28°C using Mar-tin's Bengal Red medium, and three dilution concentrations of 10 -1 , 10 -2 and 10 -3 were selected.Soil actinomycetes were incubated in inversion for 5-7 d using modified Koch 1 medium (to which a 3% solution of potassium dichromate was added to inhibit bacterial growth) and dilution concentrations of 10 -2 , 10 -3 , and 10 -4 were selected.The activity of soil urease (S-UE) was determined by phenol sodium hypochlorite colorimetry, the activity of soil catalase (S-CAT) was determined by potassium permanganate titration, and the activity of soil sucrase (S-SC) was determined by 3,5-dinitro salicylic acid colorimetry.The SOC stock was calculated by multiplying the SOC concentration (g kg -1 ) with SBD (g cm −3 ) and depth (D, cm) [35]: Data analysis.Data were analyzed using SPSS 22.0 statistical software (SPSS Inc.Chicago, USA).Linear regression analysis was conducted on the relationship between SOC and soil depth by LSD (least significant difference) test (p < 0.05).The relatedness between SOC and other soil properties (such as SWC, pH, SBD, EC 1:5 ), soil enzyme activities, and soil microor- ontent SBD D. t ganism was expressed by correlations, the hypothesized pathways of their direct or indirect effects were verified using structural equation modeling (SEM) [5].SEM is a common tool for causal reasoning in ecological studies, it can isolate multiple pathways of influence and is an effective method for studying ecologically complex relationships.Also, SEM can test the plausibility of hypothetical models based on a priori information about the relationships between specific variables.The Origin Pro 9.0 and ChiPlot software were then adopted to visualize the data through appropriate visual and statistical diagnostic plots.

SOC content and SOC pools in different economic
forests. Figure 3 shows that the SOC content of the five economic forest vegetation and wasteland ranged from 9.17 ± 0.15…17.20 ± 0.13 g/kg, with Malus pumila, Juglans regia, Zanthoxylum bungeanum, Prunus persica, Prunus armeniaca and wasteland having a SOC content of 13.14 ± 0.14 … 17.20 ± 0.13, 14.90 ± 0.12 … 16.81 ± 0.23, 13.26 ± 0.06 … 16.86 ± 0.19, 10.25 ± 0.18 … 15.20 ± 0.12, 12.80 ± 0.06 … 15.70 ± 0.19, 9.17 ± 0.15 … 12.65 ± 0.09 g/kg.In the same soil layer, the SOC content of different economic forest plantations varied.In the top layer of 0-10 cm, the SOC content of Malus pumila was the highest at 17.20 g/kg, while the SOC content of wasteland was the lowest at 12.65 g/kg.The SOC content of the same vegetation type varied at different soil depths.The maximum SOC stock was observed under Malus pumila (197.80 t ha -1 ), followed by Zanthoxylum bungeanum (185.30t ha -1 ) (Fig. 4).The vertical distribution of SOC in the six vegetation types can be well described by a quadratic function (R 2 in the range of 0.713 to 0.921) (Fig. 5).
Content in SWC, pH, EC 1:5 and SBD in different economic forests.Figure 6 shows that the SWC in the study area is between 11.33 and 20.31%.Significant differences (p < 0.05) were detected between different vegetation types in SWC in the 0-100 cm soil layer.The SWC of 6 vegetation types decreased with the increase in soil depth.Among the six vegetation types, the SWC in the 0-20 cm soil layer is markedly higher than in other soil layers.The SBD studied is between 0.88-2.20 g/cm 3 .The SBD varied significantly (p < 0.05) concerning vegetation type and soil depth.There was an increasing trend in the SBD with soil depth in all vegetation types.For all vegetation types, the SBD of the 80-100 cm soil layer was significantly higher than that of the other soil layers.The pH of the area studied is 7.95-8.52,and the soil is alkaline.Vegetation type and soil depth significantly affected the pH; the pH value increases with the increase of soil depth.The EC 1:5 in the area studied is between 197.60 and 273.67 μs/cm.The linear trend of EC 1:5 change with soil depth is not apparent, the maximum value appears
Juglans regia had the highest S-UE and S-CAT of 0.93 mg g -1 d -1 and 11.47 mL g -1 , respectively.Malus pumila had the highest S-SC at 16.8 mg g -1 d -1 .S-UE, S-CAT and S-SC were all lowest in the wasteland.S-UE, S-CAT and S-SC of the five economic forest vegetation species and the wasteland all showed a decreasing trend as the soil profile deepened.Three soil enzymes were more active in the 0-20 cm soil layer, declining rapidly below the 60 cm soil layer, then leveling off.

Correlation analysis of soil organic carbon content of different economic forests with other environmental factors, soil microorganisms, and soil enzyme activities.
Figure 9 shows a positive correlation between SOC content, soil enzyme activity, and soil microbial population in all five economic forest species.The correlation coefficient between SOC content and S-UE for Juglans regia was 0.865 (p < 0.01), and that between SOC content and S-SC for Zanthoxylum bungeanum was 0.853 (p < 0.01), both of which were highly statistically significant.The SOC content of Malus pumila showed a significant correlation with the number of bacteria; the SOC content of Prunus armeniaca showed a significant correlation with the number of fungi.Soil conductivity of the five economic forest vegetation species showed a weak positive correlation with SOC content.The correlation between SWC and SOC content was positive for all five economic forest vegetation species, with the correlation between SWC and SOC content of Prunus persica and Prunus armeniaca being higher at 0.899 and 0.934, respectively (p < 0.05).There was also a significant negative correlation between SBD and SOC content.As SBD increased, the SOC content of the economic forest vegetation tended to decrease, with the SOC content of the wasteland being most affected by SBD.The soils in the study area are predominantly primary and calcareous, with the pH of the calcareous soils ranging from 8.18 to 8.53.The SOC content of the economic forest vegetation decreases as the pH increases.

Structural equation modelling.
Structural equation modeling (SEM) quantified the relationship between soil factors and soil microbial population and soil enzyme activity (Fig. 10).The path coefficients of SEM showed that SOC was influenced by pH and EC 1:5 with path coefficients of -0.700 and 0.013, respectively.SOC significantly affected soil microorganism (0.852) and soil enzyme activity (0.525), and soil enzyme activity also had a significant positive effect on soil microorganism (0.852).The increase in SOC content increased the SWC and had a negative effect on SBD (-0.852).

DISCUSSION
Effect of vegetation type on SOC content.Different vegetation types differ in their growth characteristics and root secretions, resulting in differences in the amount of surface invertebrates and root microbial activity, which can lead to changes in litter C input and SOC output [40].Gloria [17] showed that vegetation mainly uses its return of invertebrates, root secretions, and soil microbial decomposition capacity to influence SOC and storage.The results of this study showed that the highest SOC content was found in the 0-10 cm surface soil of Malus pumila, followed by Zanthoxylum bungeanum, which was richer in SOC, while the SOC content of wasteland was the lowest.This may be because Malus pumila in Gangu County have been planted for a relatively long time, that the undergrowth of Malus pumila is rich in biodiversity such as grass cover and soil fauna, that the surface layer of dead litter is thick, that humus is also thick, and that the soil microbial population is rich in soil microbial taxa and root secretions, and therefore SOC is richer.As there are relatively few anthropogenic activities under the Zanthoxylum bungeanum, the dam-  Soil depth, cm 0-10 age to its surface is relatively small, which is conducive to the preservation and input of plant residues.A study by Deng [25] proved that the apoplastic material produced by trees could improve the carbon sequestration capacity of the soil.At the same time, the canopy structure of trees can play an influential role in shading, reducing the soil temperature in the understory and slowing down the rate of organic carbon mineralization, which directly affects the exchange of material and energy between the plant community and the environment [22].The five economic forest vegetation species studied in this paper are deciduous trees whose apoplastic material provides a source of surface SOC, and the canopy provides a boundary to maintain microclimatic zones that influence soil carbon release [3].
Since the study area is located in the Loess Plateau area in northwestern China, this area is a unique distribution area of Calcic Cambisols in the world.Calcic Cambisols is mainly powdered sand particles, loose soil, easily disintegrated by water, and easily eroded and transported by running water, and the seasonal distribution of precipitation in the area is uneven, with concentrated and heavy rainfall in July, August and September.Therefore, compared with the soil covered by economic forests, the soil organic carbon content of the wasteland is the lowest.
Vertical distribution of soil organic carbon with soil depth.The SOC content of all five types of economic forest vegetation and wasteland in Gangu County decreased with the deepening of the soil layer, and the SOC content under each cover type showed a negative correlation with soil depth (Fig. 5).In the 0-10 cm topsoil, different plants produce different amounts and quality of apoplastic material, which increases the input of exogenous organic matter.The decaying plant residues in the soil effectively change the physicochemical properties of the soil, affecting the decomposition and transformation of microorganisms in the soil through soil respiration and other means and helping the soil to accumulate organic carbon.The decomposition products of vegetation withers and arthropod exoskeletons continue to migrate downwards from the surface layer under the action of soil leaching.However, with the increase in soil depth, the Vegetation characteristics determine root characteristics that influence soil nutrient inputs [37], differences in rooting depth of different tree species also influence the vertical distribution of SOC in the soil profile [12], and deeper plant roots reduce the loss of SOC due to rainfall [39].Plant root secretions also influence the distribution pattern of SOC [32].Juglans regia is a deep-rooted tree species, and in terms of vertical distribution, the 0-70 cm soil layer is the central area of vertical root distribution, with root biomass accounting for 85.45% of the total vertical distribution [6].The root system of Malus pumila can reach a depth of 80-100 cm in the soil layer, and the root mass in the 0-60 cm soil layer accounts for about 80% of the total root system [38].Zanthoxylum bungeanum is a shallow-rooted tree with a shallow vertical root system and a wide horizontal distribution.Prunus persica is a shallow-rooted fruit tree with absorbing roots usually within 40 cm of the soil layer, with 10-30 cm being the zone of vigorous growth and development.This is consistent with the ranking of the size of the soil organic carbon content of each tree species in this study, with Juglans regia and Malus pumila having higher SOC content in the deeper soil layers of 60-80 and 80-100 cm, while the SOC content of Zanthoxylum bungeanum decreases significantly from the 0-60 cm soil layer, and the SOC of Prunus persica has the lowest organic carbon content in the 60-80 and 80-100 cm soils.

Relationship between soil organic carbon content and other environmental factors, soil microorganisms, and soil enzyme activity in different economic forests.
The synergistic effects of plant communities and soil physicochemical properties significantly impact SOC, mainly sourced from above-ground invertebrates, soil microorganisms, roots, and their secretions, and it is also influenced by soil moisture and pH [20].In this study, it was found that the soil physicochemical properties of different economic forest types showed different trends with increasing soil depth (Fig. 6).The number of soil microorganisms and soil enzyme activities showed a gradual decrease with deeper soil depth (Figs. 7, 8).In all five economic forests, SOC content was positively correlated (p < 0.05, n = 126) with SWC, soil enzyme activity and soil microbial population, and S-CAT, mL/g significantly negatively correlated (p < 0.05, n = 126) with pH and SBD (Fig. 9).Previous studies have shown that SOC contains sufficient substrates to promote enzyme synthesis [8] and that the increased organic matter input provides a large amount of raw material for microbial growth and enzyme synthesis [15].SOC shows a positive correlation with SWC, this may be related to the properties of SOC, which increases water holding capacity regardless of soil texture [10].SOC is able to constantly improve the water holding capacity of soil by rendering surfaces hydrophobic [7] and thus protecting aggregates from mineralization [14].SBD was negatively and significantly correlated with SOC content, the reason for this inverse relationship may be that the development of the root network and the accumulation of leaf litter after afforestation allows more soil organic matter to collect in the soil, leading to a decrease in SBD, since the density of humus is   -3 , while the density of minerals such as quartz is 2.65 g cm -3 [24].This is consistent with the findings of Don [9], who showed that an increase in SOC content at 0-30 cm soil depth resulted in a negative change in SBD after afforestation of agricultural fields and grasslands.pH was significantly and negatively correlated with SOC content, it can be found through previous studies that elevated soil pH is unfavorable for the absorption of nutrients by plants, limiting soil carbon accumulation [31], which is supported by Yin's [42] findings, namely the significant negative correlation between exchangeable Ca or Mg and pH.The effect of pH on SOC content was also investigated by Zhang [16].pH in the study area ranged from 7.97 to 9.02, and their analysis showed that smaller SOC content corresponded to higher pH, and greater SOC desorption was found at higher pH, and strong SOC desorption was partly responsible for the low SOC content in high pH soils.
Structural equation model.In the structural equation model (Fig. 10), soil factors have a direct effect on SOC content and are counteracted by SOC.pH had a direct effect on SOC content, enzyme activity, respectively.Zhang [28] found that pH can alter enzyme activity by affecting soil enzyme stability, and in subtropical ecosystems, C-degrading enzyme activity is driven by soil pH following land use change.SOC content has a direct effect on soil microbial indicators and soil enzyme activity, and soil enzyme activity can also have an indirect effect on SOC content through its effect on soil microbial population.This is a good illustration of the complex and diverse effects of soil factors and soil microorganisms on soil biochemistry, where various soil factors such as pH, temperature, SWC and SOC, as well as the diversity and abundance of soil microbial communities, interact to shape the characteristics and properties of the soil environment.Therefore, a comprehensive consideration of various indicators such as soil factors, soil microorganisms and soil enzyme activities can provide a better understanding of the biological, chemical and physical processes of the soil and develop more scientific and reasonable soil management measures to promote the improvement of the soil environment and the sustainable development of agriculture [30].

CONCLUSIONS
The SOC content of different economic forests ranged from 10.17-17.20 g/kg, and the SOC content at 0-100 cm soil depth was in descending order: Malus pumila, Zanthoxylum bungeanum, Juglans regia, Prunus armeniaca, Prunus persica, and wasteland, which indicated that Malus pumila was the preferred choice for economic forest planting in Gangu County and planting Malus pumila could improve the soil carbon sequestration level in the area to a certain extent.Soil factors have a direct effect on SOC content and are counteracted by SOC.Therefore, agricultural cultiva-tion needs to improve and adjust the growing environment of crops according to the actual situation.

FUNDING
We sincerely appreciate the National Natural Science Foundation of China (31760135), the Gansu Provincial Natural Science Foundation (20JR10RA089), and the Gansu Provincial Forestry and Grassland Science and Technology Innovation (KJCX2021005).

Fig. 1 .Fig. 2 .
Fig. 1.Geographical position and digital elevation model of the study area.

Fig. 6 .
Fig. 6.Soil property parameters of 0-100 cm soil layer of different economic forest vegetation.

Fig. 7 .
Fig. 7. Number of soil microorganisms in different economic forest vegetation under 0-100 cm soil depth.