1 Introduction

Soil nutrients play essential roles in soil quality, plant growth, and nutrient cycling processes in ecosystems. Carbon (C), nitrogen (N), phosphorus (P), and potassium (K) are the competitive macro-elements and the dominant nutrients in soil for plants and microorganisms (Craine and Dybzinski 2013; Wang et al. 2014b). The stoichiometry of these macro-elements in soil would also indicate the nutrient constraints for the organisms and for ecosystem function and is considered one of the dominant drivers for nutrient cycling (Delgado-Baquerizo et al. 2017; Li et al. 2016). Redfield (1958) reported that the atomic ratio of C/N/P in planktonic biomass was similar to that of marine water at 106:16:1, resulting in insight into the biochemical processes of marine ecosystems. However, unlike marine ecosystems, terrestrial ecosystems are characterized by high spatial heterogeneity, exceptional biological diversity, and structural complexity due to varying conditions (e.g., ecosystem type, vegetation cover, land-use type, topography and geomorphology, anthropogenic interference, climate factors, etc.), resulting in significant spatial heterogeneity of soil nutrients and their associated elemental stoichiometry (Bing et al. 2016; Cleveland and Liptzin 2007; Zhang et al. 2014). Previous studies on soil nutrients and elemental stoichiometry in various ecosystems have mainly discussed C, N, and P contents, but the role of K has often been neglected when estimating nutrient levels and stoichiometric relationships in terrestrial nutrient cycling (Lawniczak et al. 2009; Sardans and Peñuelas 2014; Sardans and Peñuelas 2015). The nutrient K and its stoichiometry with other elements are more sensitive in response to environmental changes than N or P, for its close relationship with soil water availability and plant uptake (Sardans et al. 2012a; Rivas-Ubach et al. 2012; Sardans and Peñuelas 2014; Yavitt et al. 2004). Therefore, it is still necessary to investigate soil nutrients and elemental stoichiometry in terrestrial ecosystems, especially K and its associative elemental stoichiometry with other nutrients.

In addition, there are many influencing factors which can strongly alter soil nutrient levels and elemental stoichiometry in terrestrial ecosystems including natural degenerative processes (e.g., overgrazing and species invasion) and anthropogenic interferences (e.g., weathering, fertilization, and crop harvesting) (Wang et al. 2014b; Zhou et al. 2018). Vegetation types, land use, and the landscape are also reported to have significant impacts on soil nutrients and elemental stoichiometry (Cleveland and Liptzin 2007; Li et al. 2012; Sardans et al. 2012c; Xu et al. 2013). Many studies have been conducted in unique landscapes or ecosystems, such as alpine ecosystems (Bing et al. 2016; Tian et al. 2018; Wang et al. 2017), loess hilly regions (Cao and Chen 2017; Li et al. 2016), and wetlands (Liu et al. 2017b; Lu et al. 2018). Previous studies have also reported influences of climate change drivers (mainly warming and altered precipitation) on soil nutrient levels and elemental stoichiometry (Sardans et al. 2012b; Tian et al. 2010; Yu et al. 2018; Yue et al. 2017). Changes in temperature and precipitation can lead to the decoupling of soil elemental stoichiometry and the variation in soil nutrients through the alteration of ground temperature, soil moisture, and microbial activity (Mooshammer et al. 2017; Penuelas et al. 2013). A negative correlation between the soil nutrient content and the mean average temperature (MAT) was observed by Liu et al. (2017b), while Yu et al. (2018) found that total carbon (TC), total nitrogen (TN), and total phosphorus (TP) content decreased with an increase in mean annual precipitation (MAP) in subtropical China. However, little information is available regarding the influence of other climate factors (e.g., extreme temperature, air humidity, and wind speed) on soil nutrients and elemental stoichiometry. Furthermore, in addition to the factors mentioned above, previous studies have shown that soil nutrients and elemental stoichiometry are also impacted by soil properties such as soil pH (Gao et al. 2014; Liu et al. 2017b), bulk density (BD) (Wang et al. 2014a; Wu et al. 2017), cation exchange capacity (CEC) (Gao et al. 2014), soil moisture (Wu et al. 2016), and soil texture (such as sand, silt, and clay content) (Serrano et al. 2016; Tian et al. 2018). However, the key factors among the above-mentioned variables influencing the distributions of soil nutrients and elemental stoichiometry required further study.

Therefore, the objectives of this study were (1) to investigate soil nutrients and elemental stoichiometry under different land uses in terrestrial ecosystems, especially K and its associative elemental stoichiometry with other nutrients; (2) to explain the impacts of climate factors, soil texture, and soil physicochemical properties; and (3) to identify the key factors on soil elemental stoichiometry and nutrient contents. The findings are favorable for incorporating nutrient cycling into ecosystem models and predicting soil nutrients and ecosystem changes under future global changes.

2 Materials and methods

2.1 Study area

The Huang–Huai–Hai River Basin, located between 30°57′ N–42°43′ N and 95°53′ E–122°42′ E, is composed of three first-level basins (i.e., the Yellow River Basin, the Hai River Basin, and the Huai River Basin). The total area of the basin is 14.45 × 105 km2, accounting for approximately 15% of China’s total area, and the terrain can be divided from west to the east into four levels in decreasing order of elevation: the Qinghai–Tibet Plateau, the Inner Mongolian Plateau, the Loess Plateau, and the Huang–Huai–Hai Plain. Due to its vastness, the climate areas within the basin are significant, consisting mainly of subtropical monsoon, temperate monsoon, temperate continental, and plateau mountain climates. The distribution of precipitation in the region is uneven; the maximum precipitation can be greater than 2000 mm in the southeastern coastal area, while the minimum precipitation can be less than 400 mm in the northeastern arid region (Wang and Zhang 2015). Furthermore, the Huang–Huai–Hai River Basin is composed of various soil types and textures and has abundant natural and artificial vegetation, including grasslands, shrublands, woodlands, croplands, etc. (Fig. 1) (Yan et al. 2017).

Fig. 1
figure 1

Land-use types and sampling sites across the Huang–Huai–Hai River Basin (1979–1985)

Full size image

2.2 Data sources

2.2.1 Soil data

Soil data (including TC, soil organic matter (SOM), TN, available nitrogen (AN), TP, available phosphorus (AP), total potassium (TK), available potassium (AK), pH, BD, CEC, volumetric water content (VMC), clay content, silt content, and sand content) were collected from the China Dataset of Soil Properties for Land Surface Modeling (Shangguan et al. 2013) and the China Soil Characteristics Dataset (Shangguan et al. 2012). Both datasets were made available by WESTDC (Environmental and Ecological Science Data Center for West China, National Natural Science Foundation of China, http://westdc.westgis.ac.cn). The original soil data were produced based on the Second National Soil Survey of China (1979–1985). Land-use types from 1979 to 1985 were collected from the Resource and Environment Data Cloud Platform (http://www.resdc.cn/Default.aspx) managed by the Institute of Geographic Sciences and Natural Resources Research, CAS.

All 282 sites (Fig. 1) were evenly distributed throughout the Huang–Huai–Hai River Basin: 158 were in the Yellow River Basin, 62 were in the Huai River Basin, and 62 were in the Hai River Basin. We excluded 12 out of the 282 sites in this study, which were sites located in water bodies or on bare soil or for which there were no available data. The sites included in this study had typical land-use types: 29 were in woodlands, 92 were in croplands, 99 were in grasslands, and 50 were in shrublands. The sites represented the eastern, central, and western sections of the region as well as all typical land-use types.

2.2.2 Climate data

The climate data used in this study included MAT, mean minimum temperature (Tmin), mean maximum temperature (Tmax), MAP, mean relative humidity in air (RHU), mean wind speed (WIN), and mean evapotranspiration (EVP). All these data for 1979–1985 were obtained at a resolution of 0.25° × 0.25° and derived from the daily observational data from meteorological stations maintained by the China Meteorological Administration (Xu et al. 2009).

3 Statistical analyses

The soil elemental stoichiometry used the mole ratio of the studied elements. The least significant difference (LSD) test and one-way analysis of variance (one-way ANOVA) were used to estimate whether the soil elemental stoichiometry and nutrient contents differed significantly among the land-use types. The ordinary least squares (OLS) method was applied to investigate the correlations between soil elemental soil stoichiometry or nutrient contents and climate factors or soil physicochemical properties. A general linear model was applied to estimate the integrative effects of climate variables and soil physicochemical properties on soil elemental stoichiometry and nutrient contents.

4 Results and discussion

4.1 Soil nutrient content and the associated stoichiometry in different land uses

The soil nutrient content varied significantly across different land uses (Fig. 2). Woodlands had the highest SOM (3.80%), followed by grasslands (2.21%), croplands (1.76%), and shrublands (0.85%). Croplands and grasslands had higher TC (2.20% and 2.15%, respectively) than woodlands (1.74%) and shrublands (1.70%), while woodlands showed the highest TN (0.19%), followed by grasslands and croplands. The shrubland TN (0.05%) was lowest overall. The grassland TP (0.08%) was significantly higher than the TP in the other three land-use types (p < 0.05), while no significant difference was found in TK (1.92–2.00%) among all four land-use types. Regarding available N, P, and K, woodlands possessed the highest AN (140.76 mg/kg), followed by grasslands and croplands. Shrublands had the lowest AN (41.51 mg/kg) overall. Furthermore, shrublands had significantly lower AP (3.84 mg/kg) than the other three land-use types (5.18–5.63 mg/kg) (p < 0.05), while the AK of woodlands and grasslands was significantly higher than the corresponding values in croplands and shrublands (p < 0.05). The results from the present study agreed with the results of Liu et al. (2017a), which showed that cropland SOC, TN, and AN content was lower compared to the corresponding values in grasslands and woodlands. The reasons for this might be the harvest of crop biomass (soil C and N losses) (McLauchlan 2006) and agricultural practices (destructing soil aggregates, decreasing SOM stabilization, stimulating the emission of nitrous oxide and nitric oxide gases, leading to SOM decomposition and N losses) (Blanco-Moure et al. 2012; Yao et al. 2009). Although the application of N fertilizers increases soil N in croplands over the short term, the fertilizers are easily absorbed by crops, transferred by nitrification and denitrification, or lost by leaching (Zhu and Chen 2002). Soil TP and TK are also necessary nutrients for plant growth. We found that TP did not differ significantly among croplands, woodlands, and shrublands, and TK did not vary significantly among land-use types, indicating that these land uses had less effect on TP and TK than on TC and TN; this result was consistent with results from Liu et al. (2017a). The soil P patterns observed among the different land-use types could have resulted from soluble crystal lattices, which could potentially retain most soil TP (Withers and Hodgkinson 2009). Soil TK was mainly controlled by the weathering rates of the parent material (Augusto et al. 2017).

Fig. 2
figure 2

Soil nutrients (SOM (a), TC (b), TN (c), TP (d), TK (e), AN (f), AP (g), AK (h)) in the topsoil (the top 10 cm soil layer) under various land-use types in the Huang–Huai–Hai River Basin. Different letters indicate significant differences among the land-use types (p < 0.05). Error bars indicate the standard errors of the mean values. SOM: soil organic matter; TC: total carbon; TN: total nitrogen; TP: total phosphorus; TK: total potassium; AN: available nitrogen; AP: available phosphorus; AK: available potassium

Full size image

The soil elemental stoichiometry also varied significantly among land uses in the Huang–Huai–Hai River Basin (Fig. 3). Shrublands had a higher C/N ratio (52.7) than grasslands (32.2), croplands (30.8), and woodlands (19.8) (p < 0.05). There was no significant difference in the C/N ratio between grasslands and croplands. The soil C/N ratio can be used as an indicator of the N mineralization capacity as well as an index of soil quality. Low C/N ratios increase microbial activity and accelerate organic N decomposition, while high C/N ratios conversely decelerate these processes (Marty et al. 2017). The C/N ratios showed significant differences among the different land uses, which could be an effect of the significant influence of C and N content and their transformation (Bai et al. 2012). C/P and C/K were highest in croplands (99.71 and 3.83, respectively), followed by grasslands, woodlands, and shrublands. The C/P ratio is considered an indicator of the organic P mineralization capacity; a lower C/P ratio is conducive to nutrient release during organic matter decomposition, causing an increase in AP that favors the growth of plants (Lu et al. 2018). Both the N/P and N/K ratios showed the same order: woodlands (6.38, 0.27) > grasslands (3.94, 0.17) > croplands (3.85, 0.15) > shrublands (1.97, 0.08). Gao et al. (2014) also reported a lower cropland N/P ratio than the corresponding woodland ratio, which may be due to the application of N and P fertilizers. Shrublands had the lowest P/K ratio (0.040), while grasslands had the highest (0.048), and the woodland and cropland P/K ratios were not significantly different from the corresponding grassland and shrubland ratios (p < 0.05). Our results were similar to those from Liu et al. (2017a), although different from results obtained on a global scale: the soil C/N ratios (19.8–52.7) from different land uses in this study were higher than those on a global scale (14.3), and the C/P (75.7–99.7) and N/P (2.0–6.4) ratios in our study were much lower than the global ratios of C/P (186.0) and N/P (13.1) (Cleveland and Liptzin 2007). C and N contents in China soils are lower than the global levels due to soil erosion and nutrient losses, which are especially severe in the Loess Plateau in Huang–Huai–Hai River Basin (Tang et al. 2019; Tian et al. 2010). The results and differences confirmed that soil nutrients and their stoichiometry existed large spatial heterogeneity on regional and global scales due to the effects of various environmental factors in terrestrial ecosystems. It is notable that changes in the C/K, N/K, and P/K ratios were mainly due to changes in C, N, and P content, respectively, but not in the K content, because K remained more or less constant throughout. Stoichiometric changes including N (N/P, N/K, and C/N) were mainly due to changes in N content because N exhibited the most obvious differences among the different land uses.

Fig. 3
figure 3

Soil nutrient stoichiometry (C/N (a), C/P (b), C/K (c), N/P (d), N/K (e), P/K (f)) in the topsoil (the top 10 cm soil layer) under various land-use types in the Huang–Huai–Hai River Basin. Different letters indicate significant differences among the different land-use types (p < 0.05). Error bars indicate the standard errors of the mean values

Full size image

4.2 Effects of soil properties on soil elemental stoichiometry and nutrient contents

It has been well established that soil nutrients are closely correlated to the physicochemical properties of soil (Bai et al. 2012; Feng et al. 2017b). Relationships among soil nutrients (including TC, SOM, TN, AN, TP, AP, TK, and AK) and soil physicochemical properties (including BD, pH, CEC, VMC, clay, silt, and sand content) were investigated in this study (Fig. 4). SOM, TN, AN, and AP were found to be significantly negatively correlated with pH, and TP and TK were significantly positively correlated with soil pH (p < 0.05). The correlations between SOM, TN, AN contents, and pH were strong (|R| > 0.5). The effect of soil pH on nutrient contents is related to the restrictive effects of pH on C and N decomposition due to its effects on microbial activity at higher pH levels (Lu et al. 2018). Soil pH did not have a significant effect on AK, in accordance with results from a previous study conducted in the eastern region of the Qinghai–Tibet Plateau (Tian et al. 2017). Soil pH also affects the rates of SOM accumulation or mineralization by regulating the growth of vegetation (Wu et al. 2017). SOM, TN, AN, AP, and AK had positive correlations with CEC (p < 0.05), and the correlations between SOM, TN, AN contents, and pH were strong (|R| > 0.5). Parfitt et al. (1995) reported that soils with high SOM typically also have a high CEC, while Jun et al. (2013) found that a higher CEC could increase P adsorption, and soil with a high CEC could be high in P content; however, we found no significant differences between TP and CEC in our study (p > 0.05). SOM, TN, TP, AN, AP, and AK were significantly negatively correlated to BD (p < 0.05), which corresponds to results from Tian et al. (2018), Lu et al. (2018), and Wang et al. (2014b). However, only the correlation between AK and BD was moderate (0.3 < |R| < 0.5, p < 0.05); other correlations were weak or not significant. Our results reported that SOM, AN, TC, TN, and TK were significantly positively correlated with VMC (p < 0.05) (Fig. 4). High soil moisture favors the accumulation of chemical elements (such as C and N), especially under anaerobic conditions (Loomis and Craft 2010), and soil moisture has been reported to enrich plant production and consequently enhance nutrient content (Yang et al. 2008). For soil texture, we found a significant positive correlation between AP, AK, and TC content and soil clay content (p < 0.01). TK was significantly negatively correlated with silt content (p < 0.01), and TN, AP, and AK were significantly positively correlated with soil silt content (p < 0.05). Moreover, AP, AK, and TC were significantly negatively correlated, and TK showed a significantly positive correlation with the soil sand content (p < 0.05) (Fig. 4). In soil, silt and clay are finer particles and have both a high specific surface capacity and a high adsorption capacity (Xiao et al. 2012); thus, their diffusion rates are generally slower, while their available nutrients are higher and move more slowly than those in more coarsely textured soil (Tian et al. 2019). Therefore, this could result in the accumulation of available nutrients, particularly in moist conditions.

Fig. 4
figure 4

Relationships between soil nutrients and soil properties across the Huang–Huai–Hai River Basin. SOM: soil organic matter; TC: total carbon; TN: total nitrogen; TP: total phosphorus; TK: total potassium; AN: available nitrogen; AP: available phosphorus; AK: available potassium; CEC: cation exchange capacity; BD: bulk density; VMC: volumetric water content

Full size image

Figure 5 shows the relationship between soil properties (e.g., pH, BD, CEC, VMC, clay, silt, and sand content) and elemental stoichiometry. A significant negative correlation between pH and the C/P, C/K, N/P, and N/K ratios and a significant positive correlation between pH and C/N were found in the present study (p < 0.05) (Fig. 5). However, Feng et al. (2017a) and Wang et al. (2014b) found that the ratios of C/N decreased linearly with pH, and Tian et al. (2018) found no significant correlation between the C/N ratio and soil pH. We found a significant positive correlation between CEC and the N/P and N/K ratios, and a significant negative correlation between CEC and the C/N and P/K ratios (p < 0.05). The correlations between C/N and pH or CEC were moderate (0.3 < |R| < 0.5), and between N/P, N/K with pH or CEC were strong (|R| > 0.5). The C/N and C/P ratios were significantly positively correlated with BD (p < 0.05), while the C/K, N/P, N/K, and P/K ratios were significantly negatively correlated with BD (p < 0.05). However, the correlations were all weak with BD (|R| < 0.3). The ratios of C/P and C/K were moderately positively correlated with VMC (0.3 < |R| < 0.5, p < 0.05), for increasing moisture will hinder decomposition and result in C accumulation and will also yield much organic phosphorus which flow out from soil and resulting in a P deficiency and an increase in C/P ratios (Zhang et al. 2017). The correlations of other ratios with VMC were weak or not significant. For soil texture, C/N ratios were significantly negatively correlated with silt content and positively correlated with sand content (p < 0.05), and the correlations were moderate (0.3 < |R| < 0.5). Several previous studies have shown that a coarse soil texture increased the soil C/N ratio (Homann et al. 2007), and a positive relationship between sand content and the C/N ratio was discovered by Marty et al. (2017). C/K and N/P ratios were significantly positively correlated with clay and silt content, while negatively correlated to sand content (p < 0.05), but the correlations were weak (|R| < 0.3).

Fig. 5
figure 5

Relationships between soil nutrient stoichiometry and soil properties across the Huang–Huai–Hai River Basin. CEC: cation exchange capacity; BD: bulk density; VMC: volumetric water content

Full size image

4.3 Effects of climate factors on soil elemental stoichiometry and nutrient contents

Climate regulates soil development and soil biota, impacting soil nutrient contents, availability, and cycling (Dijkstra et al. 2012; Oleksyn et al. 2003; Tian et al. 2010). Figure 6 shows the relationship between soil nutrients and climate factors. We found no significant correlation between TC, TP, and AP content with MAT, Tmax, and Tmin (p > 0.05), while SOM, TN, TK, AN, and AK content were significantly negatively correlated with MAT, Tmin, and Tmax (p < 0.05), where correlations between SOM, TN, and AN with MAT, Tmin, and Tmax were moderate (0.3 < |R| < 0.5) and others were weak (|R| < 0.3). Liu et al. (2017b) and Zhang et al. (2017) also found the similar nutrient and temperature patterns, and Bai et al. (2012) reported higher SOM accumulation in a cold climate. For precipitation, AN and AP were significantly positively correlated with MAP (p < 0.05), and the reason for this might be that microbial activities and nitrification, denitrification processes could be accelerated by increased MAP (Allison and Treseder 2008; Chen et al. 2016). AK and TK were significantly negatively correlated with MAP (p < 0.05), but the correlations were weak (|R| < 0.3). Furthermore, K is reported to be controlled by parent material weathering (Augusto et al. 2017), which decreases with elevated rainfall, so an increase in precipitation and temperature leads to a higher K leaching rate and a decrease in TK and AK content in soils. TC, AN, and AP had a significant positive correlation with RHU (p < 0.05), and TK had a significant negative correlation with RHU (p < 0.05); the correlations were weak (|R| < 0.3). RHU was closely correlated with MAP, and the relationship between soil nutrient contents and RHU was similar to that between soil nutrient contents and MAP. Nutrient content (except for TK) had a significant negative correlation with soil EVP (p < 0.05), and the correlations between SOM, TN, and AN with EVP were moderate (0.3 < |R| < 0.5). Yin et al. (2019) reported that a lower MAP or a higher EVP might irregularly reduce soil nutrient availability, leading to lower nutrient requirements and uptake by plants. Additionally, SOM and TN had a significant positive correlation with WIN (p < 0.05), but Wang et al. (2018) conducted wind tunnel experiments and found that no significant correlations were obtained between WIN and removal of SOM and TN. We found that TP had a significant negative correlation with WIN (p < 0.05), and the correlation was all weak (|R| < 0.3). The reason for this might be that P is enriched within the fine particle fraction, and as wind speed increases, more TP was removed from the soil surface (Wang et al. 2018).

Fig. 6
figure 6

Relationships between soil nutrients and climate factors across the Huang–Huai–Hai River Basin. SOM: soil organic matter; TC: total carbon; TN: total nitrogen; TP: total phosphorus; TK: total potassium; AN: available nitrogen; AP: available phosphorus; AK: available potassium; MAP: mean annual precipitation; RHU: mean relative humidity in air; EVP: mean evapotranspiration; MAT: mean annual temperature; Tmax: mean maximum temperature; Tmin: mean minimum temperature; WIN: mean wind speed

Full size image

The C/N and C/P ratios were significantly positively correlated with MAT, Tmax, and Tmin (p < 0.05); the C/K ratio was significantly positively correlated with MAT and Tmin (p < 0.05), and the correlations were weak (Fig. 7, |R| < 0.3); the N/P and N/K ratios were moderately negatively correlated with MAT, Tmax, and Tmin (0.3 < |R| < 0.5, p < 0.05). The main reason for this was the significantly negative correlation between TN and temperature (p < 0.05) (Fig. 4). A significant negative correlation was found between MAP and the C/N ratio (p < 0.05), and a significant positive correlation was found between MAP and the C/P, C/K, N/P, and N/K ratios (p < 0.05), and the correlations were weak (|R| < 0.3). The relationship between elemental stoichiometry and RHU was similar to that between elemental stoichiometry and MAP. The relationship between elemental stoichiometry and EVP exhibited a similar pattern to that with temperature for EVP was closely positively correlated to MAT (Zheng et al. 2016). For wind speed, N/K and N/P were significantly positively correlated with WIN (p < 0.05) (Fig. 7), and previous studies also reported that stronger soil weathering resulted in higher N/P ratios (Tian et al. 2010).

Fig. 7
figure 7

Relationships between soil nutrient stoichiometry and climate factors across the Huang–Huai–Hai River Basin. MAP: mean annual precipitation; RHU: mean relative humidity; EVP: mean evapotranspiration; MAT: mean annual temperature; Tmax: mean maximum temperature; Tmin: mean minimum temperature; WIN: mean wind speed

Full size image

4.4 Factors that affect soil nutrient availability dynamics

Soil physicochemical properties and climate factors interact to impact soil elemental stoichiometry and nutrient contents in the Huang–Huai–Hai River Basin (Figs. 4, 5, 6, and 7; Table 1). Analyses applying a general linear model revealed that all climate factors and soil physicochemical properties in this study explained greater than 50% of SOM, AN, TN, N/P, and N/K, and climate factors contributed 5.64–28.16% to soil nutrient levels and elemental stoichiometry (except for AP and TK) (Table 1). All AP was attributed to soil properties in this study, indicating that AP was minimally affected by climate factors. Among the climate factors, temperature (MAT, Tmax, and Tmin) explained most of the soil nutrient levels (except for AK). Soil elemental stoichiometry and nutrient contents could be governed by different environmental controlling factors. The results indicated that climatic variables only weakly affected soil nutrients, but they could indirectly influence soil nutrient dynamics by altering the weathering rates of parent materials as well as by influencing plant growth and productivity (Tian et al. 2019). Climate factors explained 68.35% of TK. The results again proved once again that K was primarily derived from rock weathering (Delgado-Baquerizo et al. 2013; Tian et al. 2018; Whitehead 2000). SOM, AN, TN, N/P, and N/K were attributed mostly to soil pH, which explained 34.2–44.5% (Table 1). Soil pH could directly affect the existing forms, bioavailability, and migration of nutrients. For TC, given that VMC was the primary explanatory factor, VMC was also the main explanatory factor for the ratios of C/P and C/K.

Table 1 Effects (expressed by proportion of variances explained) of environmental factors on soil nutrients and their stoichiometry across the Huang–Huai–Hai River Basin using the general linear model
Full size table

4.5 Relationships among soil elemental stoichiometry and nutrient contents

There was a strong linear relationship between SOM and TN (R2 = 0.941). The reason for this might be that SOM generally contains fixed proportions of N in soil (Pribyl 2010), reflecting the abilities of microorganisms to meet their own nutrient requirements or those of plants, through nutrient immobilization and mineralization (Delgado-Baquerizo et al. 2017; Tian et al. 2010). The soil C/N, C/P, and C/K ratios were more significantly correlated with TC than with TN, TP, and TK, respectively (Fig. 8). The N/K and N/P ratios were more significantly correlated with TN than with TP and TK, respectively. The P/K ratio was correlated with TP but more weakly correlated with TK. The results showed that variations in elemental stoichiometry (not including the C/N ratio) were dependent on TC and TN dynamics, so the elemental stoichiometry and soil nutrient contents varied according to changes in environmental variables (Figs. 4, 5, 6, and 7).

Fig. 8
figure 8

Stoichiometric associations among soil nutrients in the topsoil (the top 10 cm of the soil layer) across the Huang–Huai–Hai River Basin. TC: total carbon; TN: total nitrogen; TP: total phosphorus; TK: total potassium

Full size image

The C/N ratio was more related to TN than to TC in our study. However, Tian et al. (2018) reported that the C/N ratio was not significantly correlated with either TC or TN. In our study, the correlation between the C/N ratio and temperature and wind speed was weak, and C/N had weak responses to elevated CO2 concentration, warming, and N deposition (Yue et al. 2017). These results indicated that the soil had high stoichiometric homeostasis, where TC and TN remained closely coupled under changing environmental conditions. N/P was closely correlated with TN but not with TP (Fig. 8). TN decreased significantly with MAT, Tmax, and Tmin, while TP was not significantly correlated with temperature or MAP (Fig. 6). It is reasonable to infer that with increasing temperature, precipitation, and N deposition, the nutrient limitation would shift from P limitation to N or N-P limitation in the Huang–Huai–Hai River Basin.

5 Conclusions

Soil elemental stoichiometry and nutrient contents varied significantly among the different land uses in the Huang–Huai–Hai River Basin. Woodlands had the highest SOM, TN, and AN, followed by grasslands, croplands, and shrublands. SOM, TN, AN, N/P, and N/K were strongly negatively correlated to soil pH, moderately negatively correlated to EVP, MAT, Tmax, and Tmin (p < 0.05), and were strongly positively correlated to soil CEC (p < 0.05). Overall, soil pH explained most of the total variation in soil nutrients, and climate factors explained 5.64–28.16% of soil nutrient contents and elemental stoichiometry (except for AP (0.0%) and TK (68.35%)). Among the climate factors, temperature explained most of the soil nutrient levels and elemental stoichiometry. Moreover, C and N as well as N and P were closely coupled, which could result in a shift from P-limitation to N-limitation under conditions of climate change. The results suggest that soil properties and climate factors both affect soil nutrients and elemental stoichiometry, and soil properties generally contribute more than climate factors to soil nutrients and elemental stoichiometry. The findings will help to improve our knowledge of nutrient flux to climate change while also improving earth system models about soil response to future climate.