Introduction

The hilly and gully region in the northern part of the Loess Plateau has suffered from ecological degradation and severe soil erosion due to its transitional geographical location, climate, complex terrain, and intense human activities1. Vegetation restoration provides an effective way to control soil erosion and improve soil properties, leading to the diversification of vegetation types and environmental improvement2. Since 1999, the “Grain for Green” project has been implemented in the Loess Plateau to combat soil erosion and restore and improve the ecological environment3. However, excessive vegetation cover has resulted in a shortage of farmland, creating a contradiction between agricultural development and ecological protection. To address this contradiction, maintain agricultural development achievements, balance short-term and long-term interests, solve food shortages, increase farmer benefits, and enhance the sustainable development capacity of the northwest region, Yan’an City, Shaanxi Province, implemented the “Gully Land Reclamation” project in 2013. This project utilizes mechanized slope cutting and gully filling to create new farmland in low-lying gully areas, rapidly forming new land in the gullies, thereby increasing cultivated area and improving grain production. The primary method involves removing soil from surrounding hills to create flat and large-scale farmland in the gullies. However, during the process of expanding the area of newly cultivated farmland, mechanized operations and some engineering measures such as deep plowing, covering with infertile soil, using heavy machinery, and land leveling employed in land preparation have damaged the physical and chemical properties of the soil4. This has compressed the loose soil and disrupted the soil structure. Consequently, the newly cultivated farmland has low nutrient levels, which is unfavorable for crop growth5. This has led to a lack of soil moisture and nutrients in the newly cultivated land, rendering the soil infertile, reducing ecosystem productivity, and lagging agricultural development.

Soil nutrients are an important indicator of soil fertility, capable of supplying and coordinating the nutrition and environment for plant growth6. Vegetation and soil form an interactive and influential organic whole. Plant growth helps reduce soil erosion, enhance soil nutrient, renew land productivity, improve soil physicochemical properties, and provide opportunities for the reuse of farmland7. Wu et al.8 conducted a biological improvement experiment of planting castor bean on coastal saline soil in Yancheng, northern Jiangsu, and found that after two growing seasons of castor bean planting, the soil salinity was 0.92%, lower than the control (2.86%). Castor bean treatment significantly improved soil bulk density and soil nutrient status. Yuan et al.9 investigated the soil improvement effects of different leguminous plants (white clover > siberian peashrub > sainfoin > alfalfa > sweet clover) on Loess Plateau spoil heaps during initial planting. Their one-year study demonstrated these species effectively enhanced soil moisture and nutrient conditions. Jiang Min et al.10 researched the effects of different vegetation such as Mongolian scots pine, poplar, and sand willow on soil organic matter after ecological restoration in the Shendong Mining Area and found that the soil organic matter content in the Mongolian scots pine plot was significantly increased compared to the control collapsed quicksand area. Ren Pu11 screened alfalfa, waterweed, and ryegrass in salt-affected soil in degraded inland wetlands in the middle reaches of the Heihe River and conducted a planting experiment to improve the salt-affected soil. By comparing their soil improvement effects, the results showed that the soil improvement and fertilization effect of the grass-planting measures was significant compared with the non-grass-planting control treatment. Yang Lu et al.12 conducted an experiment on improving soil nutrients by planting grasses in the rows of one-year-old apple saplings in an orchard base in Tai’an City, Shandong Province. Using grasses such as grama grass, coarse oatgrass, and little bluestem as materials and clean tillage as a control, the research showed that different grass treatments significantly increased the soil nutrient levels in the orchard, with bayleaf bentgrass having the greatest impact on soil organic matter. Based on a comprehensive evaluation of the above research findings, the study’s time scale is relatively short, making it difficult to assess the persistence of the improvement effects. Additionally, the mechanistic research is not sufficiently in-depth, with most work remaining at the level of conventional physicochemical indicators. In future research, long-term dynamic monitoring should be conducted, and methods such as microbial metagenomics and isotope tracing techniques should be flexibly applied to achieve a more comprehensive and systematic analysis.

Alfalfa (Lotus corniculatus L.), a perennial leguminous herb, has high herbage yield and fast growth rate. As an important nitrogen-fixing plant, planting alfalfa can effectively increase soil nitrogen content and improve the soil13. Oatgrass (Avena satival L.), an annual grass herb belonging to the Gramineae family, is cold-tolerant and resistant to poverty, with strong adaptability. Planting oatgrass has the effect of regulating soil physicochemical properties and is commonly used for vegetation restoration in ecologically fragile areas such as the Loess Plateau, Tibetan Plateau, and Inner Mongolia14. Both oatgrass and alfalfa are widely planted in northwest China. However, there is currently insufficient research on their effects in improving soil nutrients in newly cultivated lands of northern Shaanxi. This paper takes newly cultivated land planted with alfalfa and oatgrass in the northern Shaanxi Loess Plateau as the experimental object to study the soil nutrient improvement effects of alfalfa and oatgrass on newly cultivated land. Combined with principal component analysis, the most suitable grass for soil improvement on newly cultivated land in northern Shaanxi was determined, providing scientific and theoretical support for effectively reducing soil infertility.

Materials and methods

Overview of the study area

The research area is located in Yanhe Bay Town, Ansai District, Yan’an City, Shaanxi Province (36°45’N, 109°22’E), with a total area of 210.64 km2. It is situated in the central part of Ansai District and is a typical hilly and gully region. The area has a temperate continental semi-arid monsoon climate, with distinct seasons of varying lengths and clear dry and wet periods. The average annual temperature is 9.4 °C, with an average annual precipitation of 501.70 mm. The average frost-free period is about 157 days, the average annual Sunshine duration is 2395 h, and the annual average total solar radiation is 118.9-132.2 kcal·cm−2. The soil type of the newly cultivated land experimental plot is yellow loess, which is alkaline and has been newly cultivated for two years. The soil is nutrient-deficient and classified as a sixth-grade poor soil. The basic properties of the experimental plot soil are shown in Table 1.

Table 1 Basic physical and chemical properties of newly created soil in Ansai District.
Full size table

Experimental design

In April 2022, oatgrass and alfalfa were planted. The oatgrass and alfalfa used in this study were sourced from the College of Grassland Science at Inner Mongolia Agricultural University. The experimental plot was set up with three treatments: oatgrass, alfalfa, and a control treatment. Each treatment was randomly arranged with three repetitions. Soil samples were collected in October 2023. Three sampling points were designed in an “S” shape in each plot where oatgrass, alfalfa, and the control treatment were planted. Stratified extraction was conducted at depths of 0–20 cm, 20–40 cm, and 40–60 cm. Soil samples from the same depth were mixed, plant residues and other impurities were removed, and the samples were placed in self-sealing bags, transported back to the laboratory for drying, grinding, and sieving before use. In the oatgrass and alfalfa plots, three 2 m2 plots were randomly selected to collect the aboveground and belowground biomass, which were taken back to the laboratory along with the soil samples. After drying, the nutrient indicators of the three soil types and the plant biomass within the plots were measured.

Methods for determination of soil nutrients

Soil organic matter (SOM), total nitrogen (TN), total phosphorus (TP), total potassium (TK), alkali-hydrolyzable nitrogen (AN), available phosphorus (AP) and available potassium (AK) were determined. SOM was determined by external heating of potassium dichromate15, TN was determined by semi-micro Kjeldahl nitrogen determination16, TP was determined by concentrated sulfuric acid-perchloric acid-deboiling molybdenum-antimonic resistance colorimetric spectrophotometry17, and the content of AN was determined by alkaline hydrolysis diffusion method18. The content of AP was determined by sodium bicarbonate extraction and molybdenum-antimony resistance colorimetric spectrophotometry19, and the content of AK was determined by ammonium acetate extraction and flame spectrophotometry20.

Data processing

Excel 2019 and Origin 2021 were used for data analysis and mapping processing, SPSS22 was used for data correlation analysis and difference significance analysis, and principal component analysis was carried out for soil nutrient indexes under two types of grass, and a mathematical model for comprehensive evaluation of soil nutrient of vegetation types was established.

Results

Soil nutrients under different grasses

The newly cultivated land planted with different vegetation types showed variations in soil organic matter (SOM). As indicated in Table 2; Fig. 1, within the 0–60 cm soil layer, the average mass fraction of SOM followed the order: oatgrass > control > alfalfa, with both alfalfa and oatgrass showing increasing SOM content as soil depth increased. As shown in Table 3, compared to the blank control group in the 0–60 cm soil layer, alfalfa reduced the average Mass fraction of SOM by 0.15%, while oatgrass increased it by 3.09%.

Table 2 Effects of different vegetation types on soil nutrients in 0–60 cm newly cultivated land in loess region.
Full size table
Fig. 1
figure 1

Soil nutrient content of newly cultivated land under different plant types. Upper case letters indicate significant (P < 0.05) differences in indicators between vegetation types at the same soil depth; lower case letters indicate significant (P < 0.05) differences in indicators between soil layers of the same vegetation type.

Full size image
Table 3 Soil nutrient increase ratio of different vegetation types compared with blank control area.
Full size table

The newly cultivated land planted with different vegetation types exhibited variations in total nitrogen (TN) content. As presented in Table 2; Fig. 1, within the 0–60 cm soil layer, the average mass fraction of TN followed the order: oatgrass > alfalfa > control, with both alfalfa and oatgrass demonstrating increasing TN concentrations with greater soil depth. According to Table 3, compared to the blank control group in the 0–60 cm soil layer, alfalfa and oatgrass increased the average Mass fraction of TN by 5.83% and 23.41%, respectively.

The newly cultivated land with different vegetation types showed distinct variations in total phosphorus (TP) distribution. As evidenced by Table 2; Fig. 1, the average Mass fraction of TP in the 0–60 cm soil profile displayed the following order: alfalfa > control > oatgrass, with both alfalfa and oatgrass exhibiting increasing TP concentrations along the soil depth gradient. Comparative analysis in Table 3 revealed that relative to the blank control, alfalfa cultivation increased the average TP Mass fraction by 3.22% in the 0–60 cm layer, whereas oatgrass treatment resulted in a substantial 45.25% reduction of TP content.

The newly cultivated lands with different vegetation types exhibited variations in total potassium (TK) distribution. As shown in Table 2; Fig. 1, within the 0–60 cm soil profile, the average mass fraction of TK followed the order: alfalfa > oatgrass > control, with neither alfalfa nor oatgrass showing significant changes in TK content with increasing soil depth. Statistical analysis revealed no significant differences (p > 0.05) in TK content among the three soil layers (0–20 cm, 20–40 cm, and 40–60 cm) for both alfalfa and oatgrass treatments. Comparative results in Table 3 demonstrated that, relative to the blank control group, alfalfa and oatgrass increased the average TK Mass fraction by 0.85% and 0.70%, respectively, in the 0–60 cm soil layer.

The newly cultivated land planted with different vegetation types showed significant variations in available nitrogen (AN) content. As demonstrated in Table 2; Fig. 1, within the 0–60 cm soil profile, the average mass fraction of AN followed the sequence: alfalfa > control > oatgrass. Notably, alfalfa exhibited increasing AN concentrations with soil depth, while oatgrass maintained stable AN levels throughout the soil profile. Comparative analysis in Table 3 revealed that, relative to the blank control group, alfalfa increased the average AN Mass fraction by 18.51% in the 0–60 cm layer, whereas oatgrass treatment resulted in a 5.76% reduction of AN content.

The newly reclaimed lands with different vegetation types exhibited distinct variations in available potassium (AK) distribution. As evidenced by Table 2; Fig. 1, the average Mass fraction of AK in the 0–60 cm soil profile displayed the following order: control > alfalfa > oatgrass, with neither alfalfa nor oatgrass showing significant vertical variation in AK content throughout the soil layers. Comparative analysis in Table 3 indicated that relative to the blank control, both vegetation types reduced AK levels in the 0–60 cm layer, with alfalfa and oatgrass decreasing the average AK Mass fraction by 1.83% and 4.03%, respectively.

The newly cultivated lands with different vegetation types showed significant differences in available phosphorus (AP) distribution. As presented in Table 2; Fig. 1, within the 0–60 cm soil profile, the average mass fraction of AP followed the order: oatgrass > control > alfalfa, with both alfalfa and oatgrass maintaining stable AP concentrations across different soil depths. Comparative results in Table 3 demonstrated that, relative to the blank control group, alfalfa decreased the average AP Mass fraction by 1.12% in the 0–60 cm soil layer, while oatgrass treatment led to a substantial 45.37% increase in AP content.

Soil nutrient profile characteristics of newly reclaimed land planted with different vegetation

As illustrated in Fig. 2, compared to the original background values of the newly cultivated land without any treatment, the contents of SOM, TP, TK, AN, and AK in the 0–20 cm soil layer significantly decreased in the control treatment, with TN and AP contents also decreasing. After planting alfalfa, the contents of SOM, TN, TP, AN, and AK significantly decreased in the 0–20 cm soil layer, while TK and AP contents decreased slightly. Similarly, after planting oatgrass, the contents of SOM, TN, TP, AN, and AK significantly decreased in the 0–20 cm soil layer, with TK and AP contents also showing a decrease. In the 20–40 cm soil layer of the control treatment, the contents of SOM, TN, TP, and AN significantly decreased, while there were no significant changes in TK and AK contents, but AP content significantly increased. After planting alfalfa, the contents of SOM, TP, and AN significantly decreased in the 20–40 cm soil layer, with no significant changes in TN, TK, AK, and AP contents. After planting oatgrass, the contents of SOM, TP, and AN significantly decreased in the 20–40 cm soil layer, with no significant changes in TN, TK, and AK contents, but AP content significantly increased. In the 40–60 cm soil layer of the control treatment, the contents of TP and TK significantly decreased, with no significant changes in SOM, AN, AK, and AP contents, but TN content significantly increased. After planting alfalfa, the contents of TP and TK significantly decreased in the 40–60 cm soil layer, with no significant change in SOM content, but TN, AN, AK, and AP contents significantly increased. After planting oatgrass in the 40–60 cm soil layer, TP content significantly decreased, with TK and AN contents also decreasing, while AK content showed no significant change. However, SOM, TN, and AP contents significantly increased.

Fig. 2
figure 2

Soil nutrient profile characteristics of different vegetation types.

Full size image

Correlation analysis of basic soil properties

As shown in Table 4, SOM exhibited a highly significant positive correlation with TP (0.914) (P < 0.01) and significant positive correlations with TK (0.781), AN (0.726), and AK (0.775) (P < 0.05). The results demonstrate that organic-rich soils can enhance the accumulation and availability of phosphorus and potassium, likely through organic matter decomposition that releases nutrients and improves soil nutrient retention capacity. However, the correlations between SOM and TN (−0.197), AP (−0.355), aboveground biomass (0.458), and belowground biomass (0.572) were not significant (P > 0.05). This may reflect the competitive relationship between organic matter and nitrogen. TN showed a highly significant positive correlation with AP (0.806) (P < 0.01) and significant negative correlations with AN (−0.787) and belowground biomass (−0.832) (P < 0.05). No significant correlations were observed between TN and TP (−0.511), TK (−0.548), AK (−0.598), or aboveground biomass (0.781) (P > 0.05).This is associated with nitrogen-phosphorus synergistic effects. TP had highly significant positive correlations with TK (0.800), AN (0.913), and AK (0.928) (P < 0.01) and a significant negative correlation with aboveground biomass (−0.818) (P < 0.05). The correlations between TP and AP (−0.596) or belowground biomass (0.745) were not significant (P > 0.05). The results suggest that phosphorus and potassium availability may be co-regulated through shared mechanisms. TK displayed a highly significant positive correlation with AN (0.820) (P < 0.01) and a significant positive correlation with AK (0.697) (P < 0.05). No significant correlations were found between TK and AP (−0.400), aboveground biomass (0.025), or belowground biomass (−0.058) (P > 0.05). AN had a highly significant positive correlation with AK (0.858) (P < 0.01) and a significant negative correlation with AP (0.713) (P < 0.05). The correlations between AN and aboveground biomass (−0.711) or belowground biomass (0.774) were not significant (P > 0.05). AK did not show significant correlations with AP (−0.643), aboveground biomass (−0.370), or belowground biomass (0.366) (P > 0.05). AP exhibited a highly significant positive correlation with aboveground biomass (0.993) (P < 0.01) and a highly significant negative correlation with belowground biomass (−0.948) (P < 0.01). Aboveground biomass had a highly significant negative correlation with belowground biomass (−0.953) (P < 0.01). Phosphorus availability directly promotes shoot growth, leading plants to reduce root investment and shift resources toward aboveground biomass production under sufficient phosphorus conditions.

Table 4 Correlation of soil basic properties.
Full size table

Principal component analysis

Variance contribution rate of principal component analysis for soil nutrients

In principal component analysis (PCA), the magnitude of variance indicates the degree of dispersion of measured indicators along the principal component direction. A larger variance suggests a greater role of the principal component in analyzing sample data. PCA was conducted on soil nutrient indicators from two vegetation types. Based on the selection criterion for the number of principal components (eigenvalue λ > 1), two principal components were obtained (Table 5). As shown in Table 5, the eigenvalue corresponding to the first principal component is 5.161, with a variance contribution rate of 73.732%. The eigenvalue corresponding to the second principal component is 1.158, with a variance contribution rate of 16.537%. The cumulative variance contribution rate of the first and second principal components reaches 90.269%, which can reflect 90.269% of the variation information in soil nutrients.

Table 5 Variance contribution from principal component analysis of soil nutrients.
Full size table

Principal component scores for soil nutrients of different vegetation types

Table 6 Principal component load, factor score coefficient and comprehensive score coefficient matrix of soil nutrients of the same vegetation type.
Full size table

Based on the principal component loading matrix presented in Table 6, it can be deduced that the correlation coefficients between Principal Component 1 and the contents of SOM, TP, TK, AN, and AP are all greater than 0.8, indicating that it effectively represents the information of these indicators. Principal Component 2 has a strong correlation with TN content. According to the score coefficients of each indicator in the first two principal components (Table 6), the functional expressions for the first two principal components of soil nutrients can be summarized as:

$$F_{1} = 0.355918X_{1} - 0.32001X_{2} + 0.420595X_{3} + 0.373614X_{4} + 0.428154X_{5} + 0.407126X_{6} - 0.3246X_{7}$$
(1)
$$F_{2} = 0.514528X_{1} - 0.595832X_{2} + 0.224768X_{3} + 0.226056X_{4} - 0.07598X_{5} + 0.071336X_{6} - 0.517457X_{7}$$
(2)
$$F = 0.260978X_{1} - 0.10319X_{2} + 0.260804X_{3} + 0.23495X_{4} + 0.227639X_{5} + 0.234291X_{6} - 0.11547X_{7}$$
(3)

In the equations: X represents the soil nutrient measurement indicators, with X1-X7 denoting SOM, TN, TP, TK, AN, AK, and AP respectively. F1-F2 represent the score values of the first and second principal components of soil nutrients. Using the variance contribution rates of the two principal components of soil nutrients in Table 5 as weights, a weighted normalization summation is performed to establish a comprehensive evaluation mathematical model for soil nutrients of different vegetation types. With the use of standard data, a comprehensive score F can be obtained (Table 7). The overall effect of plants on soil nutrient improvement is ranked as follows: alfalfa > oatgrass.

Table 7 Results of comprehensive evaluation of soil nutrients of different vegetation types.
Full size table

Discussion

The content of soil organic matter constitutes a very small proportion of soil composition, yet it plays a crucial role in vegetation growth21. In this study, within the 0–60 cm soil layer, the newly cultivated land planted with oatgrass exhibited the highest SOM content. The SOM content in both alfalfa and oatgrass soils increased with soil depth, and the effect of these two plants on soil organic Matter was Mainly concentrated in the 40–60 cm deep soil layer. As a deep-rooted forage crop, alfalfa possesses a large and developed root system with strong rhizobium nitrogen fixation capabilities. Rhizobia and numerous fibrous roots can increase SOM22. Luo et al.22 showed that, in the semi-arid region of the central Gansu Loess Plateau, planting alfalfa increased soil organic carbon content and changed its proportion, significantly improving soil carbon sequestration, which is similar to the findings of this study. Liu et al.23 investigated the effects of different vegetation restoration types on soil nutrients in a desertified area of Gansu Province and found that the SOM content of Populus hopeiensis, Caragana korshinskii, and Juniperus sabina decreased with soil depth. Jiang et al.10 demonstrated that during the ecological restoration of soil in the Shendong Mining Area, vegetation such as Pinus sylvestris var. mongolica, Populus, Salix psammophila, and Hippophae rhamnoides had the highest SOM content in the 0–20 cm soil layer, and SOM content decreased with soil depth, which is contrary to the results of this study. This discrepancy may be due to the concentration of some plant roots in the surface soil layer, where soil animals and microorganisms involved in SOM decomposition are also primarily present. Surface litter decomposition, facilitated by water or other media, percolates downward and enters the surface soil first. The combined influence of these factors results in a higher SOM content in the surface soil24. As a deep-rooted plant, alfalfa can directly increase organic matter in deep soil layers through root activity and exudates. In contrast, although oatgrass is a gramineous species, it may develop relatively deep fibrous roots. In regions with abundant rainfall, surface organic matter may migrate downward with water infiltration. However, in arid and semi-arid areas like the Loess Plateau, such leaching effects are likely to be weaker.

The total soil nitrogen content reflects the soil’s potential to supply nutrients to vegetation25. In this study, within the 0–60 cm soil layer, the newly cultivated land planted with oatgrass had the highest TN content. Both alfalfa and oatgrass exhibited an increase in TN content with soil depth, and their primary influence on soil TN was observed in the 40–60 cm deep soil layer. The main sources of soil nitrogen in alfalfa fields are the nitrogen inherently contained in the soil and the nitrogen fixed by alfalfa during its growth period26. As a leguminous plant, alfalfa’s nitrogen fixation by rhizobia primarily occurs in deeper soil layers (40–60 cm), which is associated with the deep root penetration in newly cultivated land. The nitrogen input from deep soil fixation may compensate for the leaching losses of nitrogen, resulting in an increase in total nitrogen content with depth. Wang et al.27 studied the effects of alfalfa on the physicochemical properties of saline-alkali soil in abandoned farmland in the alpine desert of the Qaidam Basin and found that the increase in soil nitrogen content after alfalfa planting was more significant in deep soil than in surface soil, which is consistent with the results of this study. Wang et al.28 investigated the distribution of soil nitrogen in the vegetation buffer zone of the Liaohe River riparian in Jilin Province and found that vegetation type Mainly affects soil TN content within the 0–40 cm depth range, with its influence gradually decreasing as soil depth increases. Liao et al.29 studied the variation characteristics of soil TN in alpine meadows in Xinjiang mountainous regions and found that soil TN content decreased with increasing soil depth, which differs from the results of this study. Soil air permeability decreases with increasing soil depth, leading to a reduction in convertible nitrogen sources. Additionally, litter, an important source of soil nitrogen, is mainly distributed in surface soil, resulting in a gradual decrease in soil TN content from top to bottom30. The variation in soil TN content and its trends may also differ under different conditions due to factors such as soil conditions, climate, and moisture. Oatgrass, commonly used as a cover crop for conditioning and improving bare, barren soils, competes with weeds for space, water, and nutrients by increasing biomass production and canopy density, thereby suppressing weed growth. Its extensive root system can absorb and temporarily immobilize NO3⁻ from deep soil layers, reducing leaching losses and consequently enhancing nitrogen content in deeper soils31.

Soil phosphorus is one of the essential nutrients for plants, directly influencing plant growth and development32. In this study, within the 0–60 cm soil layer, newly cultivated land planted with alfalfa had the highest TP content, while soil planted with oatgrass had the highest AP content. Both TP content in alfalfa and oatgrass fields increased with soil depth, whereas AP content in both did not change with soil depth. TP exhibits low mobility in soils, primarily existing in mineral or adsorbed forms. It tends to accumulate in deeper layers during soil formation processes, particularly in loess-derived parent materials like those on the Loess Plateau, where clay layers or caliche horizons strongly fix phosphorus in subsurface strata. AP mainly depends on organic matter mineralization and microbial activity, processes predominantly concentrated in surface layers. Although subsoil contains higher TP, low organic matter content and limited microbial activity restrict phosphorus availability in these deeper horizons. The intertwining roots of alfalfa and oatgrass, the decomposition of surface litter, and other microbial activities can improve soil structure, promote the formation of soil aggregates, and increase the clay content of soil layers. High clay content enhances soil’s ability to adsorb P, leading to the retention and storage of more nutrients within the soil28. Wang et al.27 planted alfalfa on abandoned and degraded land in the alpine desert region of the Qaidam Basin and found that soil AP content decreased from the top to the bottom soil layers, primarily due to the large P demand during alfalfa growth, resulting in a sharp decline in AP content. He et al.33 studied the effects of switchgrass planting on soil physicochemical properties in a typical saline-alkali region of Yinbei, Ningxia, and found that the AP content in 0–60 cm saline-alkali soil was significantly higher after switchgrass planting compared to saline-alkali wasteland. Additionally, the longer the planting duration, the higher the soil AP content, which is similar to the results of this study. An increase in plant species diversity leads to an increase in litter quantity with the increase in vegetation types, resulting in an increase in soil organic matter content, soil microbial carbon sources, and enhanced soil P mineralization, which in turn increases soil TP content. Under the action of root exudates and microorganisms, soil TP is transformed into different forms of P, allowing more AP to persist in the soil34. Sun et al.35 showed that there was no significant difference (p > 0.05) in soil TP content among four typical natural wetlands, including reed wetlands, tamarisk wetlands, Suaeda salsa wetlands, and bare tidal flats in the Yellow River Delta. This differs from the results of this study, potentially because the main source of soil P in the Yellow River Delta wetlands is input from the Yellow River, which deposits along with sediment and remains within the sediments. The physical and chemical properties of soil P in this region are relatively stable and less affected by vegetation.

Soil potassium, an essential nutrient for plants, enhances plant metabolism and photosynthesis and increases plant stress resistance36. In this study, within the 0–60 cm soil layer, newly cultivated land planted with alfalfa had the highest TK content. Neither TK nor AK content in the soils under alfalfa and oatgrass varied with soil depth, and there was no significant difference in AK content between the two vegetation types. This may be because newly cultivated land in the loess region is susceptible to erosion by rainstorm runoff, and under the influence of soil water infiltration and groundwater leaching, a significant amount of AK is lost through leaching. Soil AK, existing in an ionic state and highly soluble in water, is easily absorbed by plants and influenced by hydrological conditions37. Soil moisture affects the separation and release between litter and organic matter, the absorption and fixation of AK by roots, and root absorption and fixation of soil AK by altering litter, soil redox conditions, and microbial activity. Wu et al.38 studied the effect of green manure ryegrass on soil nutrient in kiwi orchards in Guizhou by planting ryegrass and found that incorporating ryegrass increased soil TK content, similar to the results of this study. Zhang et al.37 analyzed the impact of vegetation types such as bermudagrass, wild rice, and miscanthus on soil AK in the Mengjin District Yellow River Wetland Nature Reserve in Luoyang City and found that different vegetation types significantly affected soil AK content, with bermudagrass + miscanthus soil having significantly higher AK content than other vegetation types. Wei et al.39 studied the characteristics of soil nutrient changes under eight widely distributed plant species across four typical artificial vegetation types in the Yuzhong Basin of Gansu Province and found significant differences in soil AK content within the same soil layer under different plant samples. Specifically, the 0–10 cm soil layer under pear tree samples had the highest AK content, and AK content decreased as soil depth increased. The different environments in which root-associated microorganisms of various plant communities reside result in varying amounts of litter residue and decomposition rates, as well as differences in the depth and scope of root activities, leading to differences in soil AK content among different vegetation types. The Loess Plateau features homogeneous parent material with loose texture. Potassium (especially TK) primarily exists in mineral forms such as feldspar and mica, exhibiting minimal vertical distribution variation due to limited leaching effects. Although AK is plant-accessible and water-mobile, the region’s limited precipitation is insufficient to create significant stratigraphic differentiation.

Compared to the original nutrient background values of newly cultivated land without any treatment, the overall soil nutrient content gradually decreased over time. The nutrient loss was most severe in the topsoil layer (0–20 cm), followed by the subsoil layer (20–40 cm), while the nutrient content in the deeper soil layer (40–60 cm) showed little or no significant change. The study found a significant negative correlation between belowground biomass and total nitrogen (−0.832, P < 0.05), as well as extremely significant negative correlations between belowground biomass and available phosphorus (−0.948, p < 0.01) and aboveground biomass (−0.953, p < 0.01). Aboveground biomass showed a significant negative correlation with total phosphorus (−0.818, p < 0.05) and an extremely significant positive correlation with available phosphorus (0.993, p < 0.01). In this study, soil nutrient content was absorbed by plants. Considering the correlation between nutrient uptake during plant growth and the absence of plant incorporation back into the field, the planting of oatgrass and alfalfa led to an increase, decrease, or no significant change in various soil nutrient contents in the short term compared to the control group. After planting alfalfa, the humus left in the soil by its rhizobia and numerous fibrous roots increased soil organic matter and improved soil aggregate structure, exhibiting positive significance in biological carbon sequestration, enhancing soil nutrient, and improving soil quality. Correlations reflect static relationships, whereas long-term soil improvement requires continuous monitoring of soil nutrient dynamics and plant responses. The growth performance of alfalfa and oatgrass can serve as effective indicators for evaluating soil amelioration effects. Organic matter shows positive correlations with most nutrients, highlighting its pivotal role in enhancing overall soil nutrient status through organic matter enrichment. However, single-factor amendments may prove ineffective due to nutrient antagonism, necessitating integrated nutrient management strategies.

Conclusion

In the initial stages of planting alfalfa and oatgrass on newly cultivated land, the soil in the 0–60 cm layer exhibited the highest contents of TP, TK, and AN after planting alfalfa in the loess region. Conversely, the soil after planting oatgrass showed the highest contents of SOM, TN, and AP. Both alfalfa and oatgrass demonstrated an increasing trend in SOM, TN, TP, and AN contents with soil depth. However, soil TK, AK, and AP did not show significant variations with soil depth for both vegetation types. Principal component analysis was conducted on soil nutrient indicators for alfalfa and oatgrass, and a mathematical model for comprehensive evaluation of soil nutrient improvement by vegetation types was established. The comprehensive effect of plants on soil nutrient improvement was ranked as alfalfa > oatgrass.