Introduction

With poor soil, sparse vegetation and fragile ecological environments, the loess area of northern Shaanxi has become one of the most serious areas of soil erosion in the world under the influence of natural and irrational man-made factors1,2. Since 1999, when the Chinese government implemented ecological restoration projects such as returning farmland to forests (grasslands) and watershed closures, the vegetation coverage of this area has been greatly improved, and the ecological environment has been somewhat improved, and the ecological benefits have increased to achieve regional sustainable development and ecological civilization3,4. Vegetation, as an important component of terrestrial ecosystems, plays an important role in material circulation and energy flow in ecosystems5. Vegetation restoration plays an irreplaceable role in soil and water conservation and ecological environment construction by improving the physical and chemical properties of soil, effectively preventing soil erosion, and improving soil quality through its covering effect and root fixation6 and the high-quality soil can provide sufficient nutrients and water, which is beneficial for the healthy growth of vegetation and the maintenance of biodiversity7,8.

Soil quality is the concentration of the soil’s ability to maintain productive capacity and is a comprehensive indicator for evaluating soil fertility, environmental quality and health quality9. Soil quality depends on land use practices, ecosystem types, and the interaction of various features within the soil10. The quantitative evaluation of soil physical, chemical and biological indicators can objectively and directly reflect soil quality and ecosystem development11,12. Soil quality evaluation is an important for evaluating soil degradation and is the basis for evaluating the sustainable use of soil13. At present, various methods for evaluating soil quality, such as the composite index method14 gray correlation analysis15 principal component analysis16 and fuzzy comprehensive evaluation methods17 are available. Among these methods, the Soil Quality Index18 (SQI) method can be used to assess the impact of management measures on soil quality, is suitable for sustainable soil management, and remains the most widely used method.

Reasonable vegetation configuration patterns can effectively promote the coordination of soil water, fertilizer, and heat and thus improve soil quality, whereas unreasonable vegetation patterns may exacerbate soil erosion and degradation, affecting the recovery of regional ecosystems19. Different vegetation patterns have different effects on soil quality improvement20. At present, the local ecological and environmental problems of the Loess Plateau are still prominent, and in-depth research on the relationship between soil properties and vegetation in the process of vegetation restoration in different regions still lacks a systematic understanding, especially since the implementation of forestry ecological projects, such as returning farmland to forests (grasses), the relationship between the typical vegetation patterns formed and the soil is still unknown. Therefore, this study takes the soil quality that affects vegetation growth as an entry point, takes eight typical vegetation patterns in Wuqi County, Loess Area of northern Shaanxi as the research object, explores the differences in soil quality evaluation indexes among different vegetation patterns, clarifies the soil quality recovery under different vegetation patterns by calculating the soil quality index (SQI), and identifies the key driving factors that affect the soil quality under different vegetation patterns. This study provides a scientific basis and decision support for soil quality improvement, vegetation optimization design and inefficient forest renovation in the study area and similar areas.

Materials and methods

Overview of the study area

This study area is located in Wuqi County, Yan’an city, Shaanxi Province, with geographic coordinates of 36°33′33″−37°24′27″N, 107°38′37″−108°32′49″E, and the whole area belongs to the Loess Plateau Beam Hills and Gullies Region. Wuqi County has a semiarid continental climate in the middle temperate zone, and the soil is dominated by loess. The altitude ranges from 1233 ~ 1809 m, the average annual temperature is 7.8℃, the average annual rainfall is 483.4 mm, and it is concentrated from July-September, and the average annual frost-free period is 146 days. Before 1998, soil erosion in the county was extremely serious and the area experienced very high- intensity soil erosion. Since the implementation of the fallow forest (grass) project in 1998, the situation has improved, with the forest and grass cover reaching 49.6%. The main tree species in the region are Armeniaca sibirica, Robinia pseudoacacia, Pinus tabulaeformis, Populus hopeiensis, Populus simonii, lateral cypress Platycladus orientalis, Ailanthus altissima, Salix matsudana Koidz, etc. The dominant shrubs are Hippophae rhamnoides, Caragana intermedia, etc., and the main herbaceous plants include Lespedeza daurica, Medicago sativa, Ailanthus altissima, Salix matsudana, Agropyron cristatum, Bothriochloa ischaemum, Potentilla chinensis Se, etc. The plot information is shown in Fig. 1.

Fig. 1
figure 1

Sample plot information map. Map created using ArcGIS Desktop 10.8 (Esri), with base geographic data from Geospatial Data Cloud (www.gscloud.cn).

Full size image

Sample plot layout

After the collection of forestry data and field investigations, the group selected Wuqi County, Yan’an city, Shaanxi Province, a demonstration county for returning farmland to forests in China, was selected as the study area, eight typical vegetation patterns in the territory were used as the study objects, a fixed sample plot of 20 m*20 m was established in each sample plot, and soil collection and the determination of physicochemical indexes were carried out. A basic overview of each sample plot in the study area is shown in Table 1.

Table 1 Basic information of the sample site.
Full size table

Collection and determination of soil samples

In each standard sample plot, five 40 cm deep soil profiles were set up in an “S” shape, and the soil samples were collected via the ring knife method. The samples were vertically distributed from the surface downward into three layers, in the order of 0–10 cm, 10–20 cm, 20–40 cm. Three ring knives were set up in each soil layer of the same profile to determine the physical indexes of the soil. The soil layer of the same profile was set up with three ring knives to determine the physical index of the soil. The soil samples collected from each profile were put into plastic bags and brought back to the laboratory to dry naturally. The following methods were used: the ring knife method was used to determine the natural water content (SWC), soil bulk density (BD), maximum water holding capacity (MWHC), capillary water holding capacity (CWHC), and total soil porosity (TCP); the PHS-320 high-precision intelligent acidimeter was used to determine the soil pH value; the DDS-608 multifunctional meter was used to determine the electrical conductivity (EC); the soil organic carbon content (SOC) was determined via the external heating method; the soil alkaline nitrogen content (AN) via the alkaline diffusion method; the soil quick-acting phosphorus (AP) via the sodium bicarbonate method; the soil quick-acting potassium (AK) via the flame photometer method; the soil total nitrogen (TN) via the sulfuric acid decoction-sodium salicylate method; and the total phosphorus (TP) via the sodium hydroxide-molybdenum-antimony colorimetric method.

Soil quality evaluation methods

The soil quality index (SQI) method was used to evaluate the soil quality (SQ) of the study area21. The SQI is an indicator that can be used to quantify the overall health and fertility of the soil and is widely used by many scholars to assess the SQ22,23,24. The evaluation methods are as follows:

Selection of soil quality indicators

On the basis of the results of a previous25 study,16 soil physical and chemical indicators were selected as indicators for evaluating the soil quality of different vegetation patterns.

Calculation of evaluation indicator weights

To avoid errors caused by subjectivity, this study used principal component analysis (PCA) to calculate the weight value (Wi) of each evaluation indicator. The common factor variance obtained via PCA reflects the degree of contribution of an indicator to the overall variance. The greater the common factor variance is, the greater the contribution to the overall variance. The weights are determined by dividing the common factor variance of each evaluation indicator by the sum of the common factor variances of all indicators26.

Calculation of the affiliation of evaluation indicators

In fuzzy comprehensive evaluation, the affiliation value is determined by the affiliation function to which the evaluation index belongs. The affiliation functions generally include ascending and descending affiliation functions27.

The ascending order affiliation function is calculated as:

$$\:f\left(x\right)=\left\{\begin{array}{c}1.0,\:\:\:\:\:\:\:\:(x\ge\:b)\\\:0.9(x-a)/(b-a)+0.1\:\:\:\:\:(a\:<\:x\:<\:b)\\\:0.1,\:\:\:\:\:\:\:\:(x\le\:a)\end{array}\right.$$

The descending order affiliation function is calculated as:

$$\:f\left(x\right)=\left\{\begin{array}{c}1.0,\:\:\:\:\:\:\:\:(x\le\:a)\\\:0.9(b-x)/(b-a)+0.1\:\:\:\:\:(a\:<\:x\:<\:b)\\\:0.1,\:\:\:\:\:\:\:\:(x\ge\:b)\end{array}\right.$$

where f(x) is the degree of affiliation of each evaluation indicator, x is the average value of the indicator, a is the lowest value of the indicator and b is the highest value of the indicator.

SQI calculation

After the affiliation values and weights were calculated, the weighted evaluation method was applied to calculate the SQI. The higher the SQI is, the better the effect of soil quality of the vegetation. The formula for calculating the SQI is as follow:

$$\:\text{S}\text{Q}\text{I}={\sum\:}_{\text{i}=1}^{\text{n}}{\text{w}}_{\text{i}}\text{F}({\text{x}}_{\text{i}})$$

In the formula, SQI is the soil quality index, n is the number of evaluation indicators, is the weight value of each indicator, and F(xi) is the affiliation value of each indicator. The soil pH in the loess area of northern Shaanxi is alkaline, the pH needs to use a descending order affiliation function, and the rest of the indicators use an ascending order affiliation function28.

Software applications

In this study, Microsoft Excel (Version 2022, https://www.microsoft.com/en-us/microsoft-365/excel) software was used to preprocess the data and calculate the affiliation, weights and SQ of the soil indicators at a later stage. Principal component analysis with one-way ANOVA was performed via IBM SPSS Statistics (Version 27.0, https://www.ibm.com/products/spss-statistics) and correlation plotting was performed via Origin Pro (Version 2022, https://www.originlab.com SEM was performed via R4.4.2. The R package was constructed with the PLS-PM package29,30,31.

Results and analysis

Characterization of soil physicochemical indicators under different vegetation patterns

As shown in Table 2, there was no significant difference (P > 0.05) in the EC, AP, TN, TP, N/P, or C/P among the different vegetation patterns in the study area, and the remaining indicators were significantly different (P < 0.05). The coniferous forest MWHC, CWHC, TCP, SOC, AP, C/P, and C/N were significantly greater than those of the other vegetation patterns, whereas in the broadleaf forest, the MWHC, CWHC, SOC, TN, and TP were significantly lower than those of the other vegetation patterns at the MAF sample site. The coefficients of variation (CoVs) of the indicators ranged from 1.87–102.32%.The coefficients of variation (CoVs) of TCP (7.46%), EC (8.54%), and pH (1.87%) were lower, and according to the coefficient of variation criterion32 the degrees of variation of the indicators of TCP, EC, and pH were weakly variable; the coefficients of variation (CoVs) of TN (71.34%), N/P (63.73%), and AN (102.32%) and other indicators had high coefficients of variation, indicating moderate and strong variability.

Construction of a soil quality evaluation system based on fuzzy comprehensive evaluation

Characterization of the affiliation of soil indicators

As shown in Fig. 2, there were significant differences in the values of the soil indicators among the sample sites with different vegetation restoration patterns (P < 0.05). At the SF sample site, the SWC, BD, SOC and TP indicators were strongly related, whereas the other factors were identified as limiting factors. SWC, BD and AK exhibited high affinity in HPF, whereas other indicators were identified as limiting factors. BLF samples presented high affinities for SWC, BD and AP, and other indicators were identified as limiting factors. The SLP sample plot had a high affinity for the SWC, AP, and AK indicators, and the other indicators were identified as limiting factors. The PTF sample plots were highly correlated the MWHC, CWHC, TCP, SOC, AN, C/P, and C/N ratios, whereas the SWC, BD, EC, pH, AP, AK, TN, TP, and N/P ratio were identified as limiting factors. The MAF sample plot showed high correlations with BD, EC, and pH, except for the other indicators which were identified as limiting factors. The MF1 sample plot was highly correlated a high correlation with SWC and TN, while the remaining indicators were identified as limiting factors. The MF2 sample plots were highly correlated with EC, pH, and TP, and the remaining indicators were identified as limiting factors.

Table 2 Soil quality indicators for different vegetation patterns.
Full size table
Fig. 2
figure 2

Affiliation diagram of soil indicators for different vegetation types.

Full size image
Table 3 Principal component analysis and weights of the soil indicators.
Full size table
Table 4 SQI values for different vegetation patterns.
Full size table

Weighted values for the soil indicators

As shown in Table 3, the 16 indicators were subjected to principal component analysis. The results show that the eigenvalues of the first four principal components are all greater than 1, and the cumulative contribution rate reaches 77.482%, explaining most of the variables of the original information. The common factor variance of the 16 indicators ranges from 0.600 to 0.954, which are all greater than 0.5, and the common factor variance of the indicators is greater than 0.7, except for EC, AP, AK, TP, and C/N, which suggests that the common factor variance effectively expresses the SQ evaluation indicators that contain the information contained in the SQ evaluation indicators.

SQI analysis

As shown in Table 4, the SQI values of the different vegetation patterns in the study area ranged from 0.389 to 0.534. There were differences in soil quality among the different vegetation patterns, with the soil quality of coniferous forests significantly better than that of the other vegetation patterns; among the broad-leaved forests, the soil quality of the SLP sample site was the best, while the soil quality of the MAF sample site was significantly lower than that of the other vegetation patterns; and the differences in the soil quality of the two kinds of mixed forests were not significant. The order of soil quality size was as follows: PTF (0.534) > SLP (0.494) > SF (0.462) > MF1 (0.430) > HPF (0.423) > BLF (0.420) > MF2 (0.415) > MAF (0.389). Among them, the SQI of coniferous forests was improved by 15.58% greater than the SQI of shrub forests, and the SQI of oil pine forests was improved by 37.28% greater than the SQI of mountain almond forests, which was 37.28%.

Soil quality drivers for different vegetation patterns

Fig. 3
figure 3

Analysis of soil quality drivers for different vegetation types.

Full size image

SEM was used to analyze the relationships between soil quality and soil physical and chemical factors under different vegetation patterns, and the results are shown in Fig. 3. Each vegetation type was influenced by physical and chemical indicators. In general, the soil physical indicators PI (BD and MWHC) had a negative effect on soil quality with a throughput coefficient of −0.605, whereas soil organic carbon (SOC) had the most dominant positive effect with a throughput coefficient of 0.714. Among them, SOC had a significant positive effect on SF, PTF, MAF, MF1, and MF2 with throughput coefficients of 0.957, 0.996, 0.984, 0.854, 0.954, 0.996, 0.984, 0.954, and 0.996, respectively. The factors that had a significant positive effect on the HPF, BLF, and SLP were AN, PI (MWHC vs. TCP), and PI (BD vs. TCP), with through-travel coefficients of 0.9469, 0.774, and 0.833, respectively. The physical indicators (PIs) had a significant negative effect on the HPF, PTF, MF1, and MF2, with throughput coefficients of −0.7219, −0.253, −0.867, and − 0.301, respectively. In addition, AP had a negative effect on MF2. pH has a significant negative effect on BLF and SLP, with throughput coefficients of −0.196 and − 0.372, respectively, whereas in the HPF samples the throughput coefficient of pH was positively affected by 0.1920. In the SF and MAF samples, the negatively affecting indicators were the chemical indicators AK and AP, respectively (throughput coefficients of −0.337 and − 0.031).

Discussion

Analysis of differences in physical and chemical indicators between vegetation patterns

Different vegetation types influence soil health and community structure in direct or indirect ways. The results of this study revealed that the best soil quality with good soil porosity and water-holding capacity was found at the PTF sample site. Specifically, the maximum water holding capacity (MWHC), capillary water holding capacity (CWHC), and total porosity (TCP) were greater than those of the other vegetation patterns, while the soil bulk density (BD) was relatively low, and these characteristics indicated that the physical properties of the soil in the PTF plot were superior. Han et al.33 also found in their study of the effects of three types of coniferous forests on the soil and vegetation of semi-arid mountain ecosystems that the higher capillary porosity and total porosity and the lower soil bulk weight of the creosote bush forests provided a better basis for water infiltration and retention, which is favorable for water conservation, which is consistent with the findings of this study. In contrast, the soil bulk density (BD) and total soil porosity (TCP) of the MAF sample site were lower than those of the other sample sites, which may be due to the presence of a large number of anthropogenic trampling activities at the MAF sample site. Jia Xiaoyan et al.34 noted that the harvesting process in MAF resulted in an increase in soil bulk weight and a decrease in total porosity due to human trampling, which is consistent with the results of this study. The study revealed that the highest SOC content was found in the PTF, whereas the highest TN content was found in the MF1 and the highest TP content was found in the MF2. In contrast, the SOC, TN and TP contents of the MAF sample were significantly lower than the other samples. This may be because the fact that the PTF has better porosity and water-holding capacity, which improves soil moisture and provides a rich energy and material base for microbial activities, thus increasing the content of soil organic carbon and promoting the virtuous cycle of the soil and vegetation system. Zhou Zhiyong et al.35 reported in their study of the characteristics of soil quality and organic carbon stability evolution with forest age in the PTF of Taiyue Mountain, Shanxi, that the PTF has a high soil organic carbon content, and that the porosity and water-holding capacity of the forest soil improved and that the soil microbial biomass increased gradually over time. Zhu Chao et al.36 further confirmed that the N and P contents of mixed forests were significantly greater than those of pure forests, which indicated that different vegetation types had a significant effect on the rates of soil N and P element accumulation, which was consistent with the findings of this study. Studies have shown that C, N, and P, as the basic elements of living organisms, can reflect the way plants absorb nutrients37. The soil C/N ratio is an indicator of the soil nitrogen mineralization capacity, which is negatively correlated with the decomposition rate of soil organic matter, and the soil C/P ratio reflects the potential of microorganisms to mineralize soil organic matter to release phosphorus or to take up and fix phosphorus from the environment38. In this study, both the C/N and the C/P values were greater for the creosote bush, suggesting that the suggests that creosote bush has a positive effect on the accumulation aspect of organic carbon; however, the soil phosphorus effectiveness is low, and plant growth may be affected by phosphorus effectiveness39.

Analysis of soil quality under different vegetation patterns

Soil and vegetation are important factors in terrestrial ecosystems; soil provides the necessary water and nutrients for plant growth, while vegetation planting contributes to soil optimization and improvement40. Different vegetation configuration patterns are important for assessing soil quality status and maintaining ecosystem stability41. In this study, the soil quality under different vegetation restoration patterns was analyzed in depth via principal component analysis (PCA) and fuzzy comprehensive evaluation methods. The results revealed that among the eight different vegetation patterns, the soil quality of the PTF and SLP plots was significantly better than that of the other sample plots, which was mainly due to the high organic matter and nutrient contents in the soil of the PTF and SLP plots on the gently sloped slopes, which are important for improving the soil quality. A study by Lili Dong42 also revealed that soil organic matter is a key cementing substance for soil aggregate formation and stabilization, and that the soil structure of the four types of woodlands, including PTF, SLP, POF(Patycladus orientalis), and PAF(Prunus armeniaca), was significantly improved compared with that of agricultural land and apple orchards, which is consistent with the findings of this study. In addition, this study revealed that the soil quality of MF2 was lower than that of SF and BLF, which was attributed to the fact that BLF and SF, both of which are suitable species for the study area, grow faster and that mixing results in over-densification of the stand. Although higher stand density is favorable for the accumulation of stand biomass, it leads to deterioration of ecological conditions such as light, water, temperature and nutrients within the forest, and the consumption of soil nutrients by many plants cannot be replenished in a timely manner, which has a negative impact on the growth of understory vegetation and soil health, which corroborates the results of this study43,44. Therefore, when the return of farmland to forest (grass)projects in the loess hilly areas of northern Shaanxi is implemented, the rational selection of different vegetation patterns is highly important for improving soil quality and sustainable development, which can maximize the social, ecological and economic benefits45. In addition, considering that this study mainly focuses mainly on the physicochemical properties of the soil and does not involve the study of soil bioindicators, future research should strengthen the measurement of soil bioindicators and comprehensively consider more factors that have a significant impact on soil quality, with a view to further improve the soil quality assessment system in the region.

Analysis of soil drivers for different vegetation patterns

Significant differences exist in the effects of different environmental conditions on soil quality drivers46. The present study revealed significant differences in soil quality drivers among different vegetation patterns. Through a comprehensive assessment of the throughput coefficient and soil quality index (SQI), the chemical indicator SOC was found to be the key factor dominating the SQI of different vegetation patterns, which was mainly because different vegetation types accumulate large amounts of roots and apoplastic materials in the soil surface layer, and the organic matter produced by the decomposition of apoplastic materials returns to the soil through nutrient cycling, resulting in the accumulation of a large amount of organic carbon in the surface layer of the soil47. Soil organic carbon plays a key role in maintaining the stability of soil structure; enhancing soil fertility; maintaining the soil water balance; and mitigating climate change, which can affect soil quality by influencing soil chemical, physical, and biological indicators48. Moreover, this study revealed that the physical indicator MWHC is also an important factor dominating the SQI of different vegetation patterns, which is due mainly to the influence of soil moisture on the growth of surface vegetation and soil quality. Furthermore the maximum water holding capacity (MWHC) of forest soil is an important indicator of its water-holding capacity49, and changes in soil moisture directly affect the growth and development of vegetation, which in turn affects the carbon sequestration efficiency and water retention capacity of vegetation50. Effective water retention is an important factor in improving soil quality and promoting vegetation growth51. The present study also revealed that TN and AN had significant effect on the SQI of almost all the vegetation patterns. The contents of TN and AN, important nutrient indicators of soil, are highly important for improving soil quality52. Addressing water and carbon and nitrogen issues is particularly critical in the future when improving the soil quality of different vegetation patterns in a region. In addition, soil bulk density (BD) plays an important role in the SQI of different vegetation patterns. Relevant studies have noted that soil bulk weight is a key physical trait reflecting soil porosity, water retention and fertility53, and the smaller the soil bulk weight is, the greater degree of soil water retention54. Effective water retention is an important factor in improving soil quality and promoting vegetation growth, and if it is not reasonably maintained, the soil will be damaged in the event of heavy rainfall, leading to nutrient loss and a decline in soil quality. Therefore, water retention and soil fixation are important for future improvements in soil quality in this study area, and solving water, carbon and nitrogen problems is especially critical.

Conclusion

The results of this study revealed that there were significant differences in the physicochemical properties of the soil evaluation indicators across the different vegetation patterns. The PTF had relatively highly water holding capacity, porosity, organic carbon and phosphorus contents, which improved the soil quality. The soil organic carbon, total nitrogen, and total phosphorus contents of MAF were lower, and the soil capacity value was higher, with relatively poor soil quality. Different vegetation patterns can effectively improve soil quality to different degrees, among which, the PTF has the greatest effect on soil quality improvement, followed by the SLP and SF. Structural equation modeling analysis revealed that the overall SQI of different vegetation patterns was jointly influenced by physical and chemical indicators, in which SOC had a significant positive effect on the SQI of different vegetation patterns, whereas BD and MWHC had a significant negative effect on the SQI of different vegetation patterns. Therefore, the results of this study suggest that PTF, SLP (slow-slope afforestation), and SF are suitable for use in future forestry ecological project construction in this study area and similar areas.