Ecological stoichiometric characteristics and influencing factors of carbon, nitrogen, and phosphorus in the leaves of Sophora alopecuroides L. in the Yili River Valley, Xinjiang

Site description

The study area was located in the Yili River Valley of the Xinjiang Uygur Autonomous Region (80°09′–84°56′E, 42°14′–44°50′N). High mountains form the north, east, and south sides of the Yili River Valley. The terrain ranges from high in the east to low in the west, in the shape of a trumpet, so the natural landform can be described as “three mountains and two valleys,” and is considered a “wet island.” The elevation ranges from 530 m to 1,000 m. The valley spans 360 km from east to west and 275 km from north to south, covering an area of 56,400 km². The Yili River Valley is the wettest area in Xinjiang, with a warm and humid temperate continental climate. The average annual temperature is 10.4 °C. Annual sunshine ranges between 2700 and 3000 h, and the average annual precipitation is 417.6 mm, of which approximately 60∼70% occurs during the spring and summer. The Yili River Valley has abundant natural resources, high species diversity, various mineral deposits, and unique wetland landscapes. Vegetation types in the valley are mainly grassland, meadow, and forest.

Study site and sample collection

Field investigation and observation revealed that S. alopecuroides, having strong invasion ability, often colonizes four landscapes—forest, roadside, farmland, and desert—in which it becomes widely distributed. Sampling sites were established in Qapqal County, Yili River Valley, covering those four habitat types (Fig. 1). Soil samples were collected in September 2018, and plant samples were collected in July 2019.

In the four invasive habitats of S. alopecuroides, 20 m ×20 m plots were selected respectively, and 3 quadrats were randomly selected from each plot, with an area of 5 m ×5 m. In each plot, 20 complete S. alopecuroides leaves were randomly selected, and taken back to the laboratory for cryopreservation. All collected leaves were dried at 70 °C for 24 h. The leaves were then crushed into powder using a pulverizer, then sealed for storage. The contents of total C, total N, and total P in the leaves of S. alopecuroides leaves were determined.

According to the same method, 20 m ×20 m plots were set in four invasive habitats of S. alopecuroides, and three quadrats (5 m ×5 m) were randomly set in each plot. Soil samples of 0∼10 cm, 10∼20 cm, and 20∼30 cm depth were collected from each quadrats. 36 soil samples were collected from the four habitats. The collected soil samples were placed in sealed plastic bags. After transportation to the laboratory, the soil samples were dried, ground with a mortar and pestle, then screened through a 0.5 mm mesh for soil physical-chemical properties o analyzed.

Analysis of soil and plant properties

Soil physicochemical properties

The total potassium (K) content in the soil was measured using atomic absorption spectrophotometry (Bao, 2000). To determine the content of ammonium nitrogen (NH₄⁺-N) and nitrate nitrogen (NO₃⁻-N) in the soil, firstly, 10.00 g of the soil sample was weighed into a plastic bottle, calcium chloride (CaCl₂) extractant was added, and the mixture was shaken for 30 min between 20 °C ∼25 °C, then measured the content of NH₄⁺-N and NO₃⁻-N by colorimetry after filtration (Bao, 2000). The content of soil organic matter (OM) was measured by a K₂CrO₇-H₂SO₄ oxidation procedure (Bao, 2000). Soil pH was measured using a pH meter; To determine soil available phosphorus (AP), 2.50 g of each soil sample was weighed into a plastic bottle, then NaHCO₃ extract and 1 g of phosphorus-free activated carbon was added; The sample was shaken for 30 min between 20 °C ∼25 °C, then filtered, and the content of AP in the soil sample was measured by colorimetry (Bao, 2000). The content of available potassium (AK) was determined by the flame photometric method (Bao, 2000).

Statistical analysis

Excel 2010 (Microsoft) and SPSS 19.0 statistics software were used to analyze the data after integration. One-way ANOVA method was used to analysis the differences in C, N, and P contents and their stoichiometric ratios in the leaves of S. alopecuroides across the four different habitats. And significance analysis was performed using Duncun multiple comparison. It should be noted that the factors significantly related to the ecological stoichiometric characteristics of C, N, and P need to be selected by Monte Carlo analysis before the redundancy analysis. According to the detrended correspondence analysis (DCA) of leaf C, N, and P contents, it can be seen that the length of gradient (LGA) of the sorting axis was less than 3, indicating the data had suitable linearity between the leaf nutrient levels and the soil environmental factors, which is suitable for the redundancy analysis (RDA).

In the sorting diagram, the quadrant in which the arrow is located represents a positive and negative correlation between the factors and the sorting axis, while the hollow arrow represents the ecological stoichiometric characteristics of the leaves. The solid arrow represents the physicochemical factors of the soil. The length of the line represents the relationship between the ecological stoichiometric characteristics of the leaves of S. alopecuroides and the soil chemical factors. The angle between the two arrows represents the correlation between the ecological stoichiometric characteristics of the leaves and soil chemical factors. The smaller the angle, the greater is the correlation. The solid line represents the environmental factors that were significantly related to the stoichiometric characteristics of the leaves (p < 0.05).

Content and stoichiometric ratios of C, N, and P in S. alopecuroides leaves

As indicated by Table 1, the average values of C, N, and P contents in the leaves of S. alopecuroides in the Yili River Valley were 470.09 g/kg, 32.71 g/kg, and 1.43 g/kg, respectively. The coefficients of variation of C, N, and P were 10.96%, 30.41%, and 30.86%, respectively. The average values of C/N, C/P, and N/P were 16.88, 364.67, and 23.20, respectively, and the variation coefficients were 57.04%, 38.42%, and 24.00%, respectively. The coefficients of variation of C, N, P, and the stoichiometric ratios of leaves were generally large, with the largest being the coefficient of variation of the C/N ratio, indicating that the C and N contents of the leaves had the highest degree of variation and the strongest variability. As shown in Figs. 2A and 2B, there was no significant correlation between leaf C content and leaf N and P content (p > 0.05), while there is a strong positive correlation (p < 0.01) between leaf N and P contents (Fig. 2C).The regression equation (y = 0.0009x² −0.0095x + 0.7532) clearly reflects the increasing trend of P content in leaves with increasing N content.

Contents and stoichiometric ratios of C, N, and P in S. alopecuroides leaves in different habitats

There were some differences in the C and P contents of S. alopecuroides leaves in different habitats (Table 2). The C content of leaves in the habitats showed an increasing trend: the C content of leaves from the desert habitat was much higher than those from the forest. The coefficients of variation of C content in the leaves were 11.58%, 13.28%, 14.93%, and 5.73% for the forest, roadside, farmland, and desert habitats, respectively. There was no significant difference in N content among the four habitats. The variation coefficients of N content in the forest, roadside, farmland, and desert habitats were 7.52%, 63.30%, 3.37%, and 31.24%, respectively. The P content of the leaves sampled from the four habitats in order from highest to lowest was farmland >forest >roadside >desert, and the P content of the leaves from the desert was significantly lower than that of the farmland. The variation coefficients of P content in the leaves of the forest, roadside, farmland, and desert were 13.62%, 40.12%, 6.96%, and 44.23%, respectively.

There was a significant difference in the stoichiometric ratio of C/P in the leaves of different habitats, but there were no significant differences in C/N and N/P (Table 3). The C/N ratio of the forest leaves was slightly lower than those of the desert, roadside, and farmland areas. The coefficients of variation of C/N in the forest, roadside, farmland, and desert were 7.59%, 75.31%, 12.19%, and 22.65%, respectively. The N/P ratio of leaves from the farmland habitat was less than that of the forest, roadside, and desert areas. The coefficients of variation of N/P in the forest, roadside, farmland, and desert habitats were 13.31%, 41.23%, 5.09%, and 19.31%, respectively. The leaf C/P of the four habitats, from largest to smallest, was desert >roadside >forest >farmland, and the leaf C/P of the farmland habitat was significantly lower than that of the desert. The coefficients of variation of C/P in the forest, roadside, farmland, and desert habitats were 8.69%, 32.98%, 14.29%, and 38.67%, respectively.

Correlation between ecological stoichiometric characteristics of S. alopecuroides leaves and soil physicochemical factors

RDA was used to study the correlation between the ecological stoichiometric characteristics of S. alopecuroides leaves and the soil physicochemical factors (AK, pH, NO₃⁻-N, NH₄⁺-N, AP, OM, soil C content, soil N content, soil P content, and soil K content). It can be seen from Table 4 that the interpretation amounts of the first and second sorting axes were 54.8% and 26.9%, respectively. The first two axes jointly explained 81.7% of the change in the leaf ecological stoichiometric characteristics. The cumulative explanations of the leaf ecological stoichiometric characteristics and soil physicochemical factors of S. alopecuroides leaves reached 86.8%, indicating that the first two axes could reflect the large difference between the soil physicochemical factors and the leaf stoichiometric characteristics, and were mainly determined by the first sorting axis.

According to the RDA (Fig. 3), the arrow line between AK and pH was the longest, which is consistent with the importance ranking results in Table 5. Together, AK and pH had the greatest impact on the ecological stoichiometric characteristics of S. alopecuroides leaves. AP and pH were positively correlated with leaf C, N, P, and N/P, and negatively correlated with leaf C/P and C/N. The direction of the arrow line of AK and leaf P content was almost the same with a small angle, indicating that AK was significantly positively related to leaf P content, and AK may be an important factor affecting leaf P content in S. alopecuroides leaves in the Yili River Valley.

Different soil physicochemical factors were found to exhibit significant differences in the ecological stoichiometric characteristics of S. alopecuroides leaves (Table 5). The effects of soil physicochemical factors on stoichiometric characteristics were as follows: AK >pH >NO₃⁻-N >soil P content >NH₄⁺-N >soil N content >soil K content >soil OM >soil C content >AP. AK and pH had a significant effect on the stoichiometric characteristics of the leaves (p ≤ 0.05). AK had the most significant effect on the stoichiometric characteristics of leaves, accounting for 19.9% of the total interpretation (4.487, p = 0.05). NO₃⁻-N, soil P content, NH₄⁺-N, soil N content, soil K content, OM, soil C content, and AP had no significant effect on the stoichiometric characteristics of the leaves (p > 0.05).

Ecological stoichiometric characteristics of C, N, and P in the leaves of S. alopecuroides in different habitats

The average C content (470.09 mg/g) of S. alopecuroides leaves in the Yili Valley was slightly higher than the average of 492 global terrestrial plants leaves (464 mg/g; Elser et al., 2000a), indicating that the organic C content of S. alopecuroides in the Yili River valley was higher, therefore the ability to store C storage was stronger. There was a certain difference in leaf C content among the four habitats, which indicated that leaf carbon accumulation of S. alopecuroides in the four habitats was different.

The higher C content from the desert leaves may be because carbon usually exists in plants in the form of OM. Under conditions of low soil moisture and high salt content, a habitat will likely be high in stress and low in competitive interference. S. alopecuroides readily stores carbon, reduces its reproductive and competitive ability, maintains normal growth, and achieves balanced resource allocation (Zhang, 2000). It may also be the case that in an environment with sufficient resources, it is easy to reach environmental capacity saturation, which would aggravate interspecific competition and reduces the available natural resources.

Yili Valley lies within the arid and semi-arid region of Xinjiang. The average N content in the leaves of S. alopecuroides (32.71 mg/g) is not only much higher than the average of Chinese plant leaves (20.2 mg/g; Han et al., 2005), but also higher than average N content in plant leaves at a global scale (20.6 mg/g; Elser et al., 2000a). This seems consistent with previous observations that the N content of leaves from arid desert environments was relatively high (Li et al., 2010); however, the N content in S. alopecuroides leaves showed no significant difference across the four habitats. S. alopecuroides is a leguminous plant with strong nodulation, and therefore nitrogen-fixation ability; this would explain the high N content in the leaves, as well as the high internal stability of N levels.

The average content of P in the leaves of S. alopecuroides (1.43 mg/g) was lower than the average of plants in China (1.5 mg/g; Han et al., 2005) and of that of global plants (1.99 mg/g; Elser et al., 2000a). The abundant precipitation of the Yili River Valley’s temperate continental climate enhances the leaching of available P in the soil—which is not conducive to binding and accumulating P—resulting in less P absorption by plant leaves (Wang et al., 2018). Previous studies have shown that human disturbance, such as the application of fertilizer, affects the plant chemometrics. Cultivation, fertilization, irrigation, and other activities improve the local soil’s nutrient content and quality and improve the effective P content (Luo & Gong, 2016), which would provide a favorable environment for S. alopecuroides growth. This explains the higher P content in the S. alopecuroides leaves from the farmland habitat compared to the other habitats.

C/N and C/P represent the ability to assimilate C when plants absorb nutrient elements. To some extent, it can reflect the utilization efficiency of nutrient elements in plants, and the ratios are closely related to the growth rates of organisms (Davis, Childers & Noe, 2006). In this study, the C/N ratio (16.88) of S. alopecuroides leaves was lower than the global plant average (22.5; Elser et al., 2000b), and C/P (364.67) was much higher in the S. alopecuroides leaves than the global plant average (232; Elser et al., 2000b), indicating that the nutrient utilization efficiency of S. alopecuroides L. was low. As shown in Table 3, the C/N and C/P ratios of S. alopecuroides leaves grown in the desert and roadside habitats were higher than those of the farmland and forest habitats. This finding may be because, in an environment with sufficient nutrition, the growth rate of plants is high, the synthesis of organic carbon is high, and the C/N and C/P ratios are low. In nutrient-deficient environments, plant growth is relatively slow, the utilization efficiency of nutrient elements is high, and the C/N and C/P ratios are high (Ng et al., 2014).

The ratio of N to P in plant leaves reflects the dynamic balance between soil nutrient supply and plant nutrient demand; the N/P ratio can be used to determine the limiting growth factors of plant nutrients (Duan et al., 2004). Studies have shown that when leaf N/P <14, plant growth is mainly restricted by N; when leaf N/P >16, plant growth is mainly restricted by P; and when leaf NP is more than 14, but less than 16, plant growth is restricted by both N and P (Aerts & Chapin, 2000). S. alopecuroides is a nitrogen-fixing plant species, so it is generally N-limited, due to its higher leaf P content (mean = 1.53 mg/g), indicating that it may not be P-limited. Therefore, the rule-of-thumb for judging restrictive elements—that is, an N/P of 14 or 16—may not greatly apply to nitrogen-fixing plants. Further studies are needed to determine which nutrients limit plant growth.

Factors affecting the ecological stoichiometric characteristics of C, N, and P in S. alopecuroides leaves

Plants need to absorb nutrients from the soil to supplement the nutrients needed for the growth and development of leaves, so soil physicochemical factors have a greater impact on the C, N, and P ecological stoichiometric characteristics of S. alopecuroides leaves. According to the observed RDA ranking, AK and pH were the main factors affecting the C, N, and P stoichiometric characteristics. Soil pH can change soil nutrient content and distribution area, thus affecting plant growth and developmental processes (Zhan, Yu & He, 2013). In this study, pH was positively correlated with C, N, and P contents and N/P in S. alopecuroides leaves and was negatively correlated with C/P and C/N, indicating that pH is closely related to leaf growth, which is similar to the finding that pH affects the growth and development of plants by affecting the physical, chemical, and biological characteristics of soil (Xu et al., 2015). In this study, it was observed that with an increase in the content of AK, the C, N, P, the N/P ratio in the leaves also increased, but C/P and C/N decreased. The content of P in the leaves was positively correlated with the content of AK, indicating that AK was the main factor affecting the content of P in the leaves. This may be because the absorption efficiency of soil nutrients to S. alopecuroides leaves is different in arid and semi-arid areas compared to other environments. The absorption efficiency of AK is higher in arid and semi-arid areas, which benefits the growth of S. alopecuroides leaves.

The internal stability of grassland ecosystems in the arid and semi-arid areas of Xinjiang was explored by analyzing the influence of soil physicochemical factors on the ecological stoichiometric characteristics of S. alopecuroides leaves. The RDA indicated that the stoichiometric characteristics of C, N, and P in the sampled S. alopecuroides leaves were not significantly affected by the OM, nitrate nitrogen, ammonium nitrogen, AK, or the soil C, N, P, and K levels. However, the independent analysis of the effect of soil physicochemical factors on the stoichiometric characteristics showed some deficiencies. First, the effects of soil physicochemical factors on the C, N, and P ecological stoichiometric characteristics of S. alopecuroides leaves were not independent. Second, soil physicochemical factors had mutual influence and restrictions. Therefore, based on the independent analysis, it is necessary to further analyze the double or even multiple synergistic effects of soil physicochemical factors on the ecological stoichiometric characteristics of C, N, and P in S. alopecuroides leaves, to make the results more accurate (Li et al., 2019).

As the dominant species degrading grassland in the Yili River Valley, the growth, development, and distribution of S. alopecuroides seriously affects the ecosystem in the valley. Studying the relationships between the C, N, and P ecological stoichiometric characteristics of leaves and environmental factors is of great significance in revealing the ecological mechanisms of the successful diffusion of S. alopecuroides plants in the Yili River Valley.