Effects of Long-Term Enclosing on Vertical Distributions of Soil Physical Properties and Nutrient Stocks in Grassland of Inner Mongolia

Abstract

Enclosing plays a crucial role in vegetation and soil quality in grassland. The biomass of green plants, litter, and vertical distributions of soil physical properties and nutrient stocks were evaluated at plot enclosed long term for 38 years inside a fence and a long-term grazing plot outside a fence in a semi-arid grassland of Inner Mongolia. The results showed that dry matter of green plants and litter during the 38-year enclosing treatment was higher than in the grazing treatment (p < 0.01). The soil silt (2–50 μm) in the 38-year enclosing treatment was 5.9% higher than in the grazing treatment (p < 0.01) in 0–10 cm soil, and the fine sand (100–250 μm) was 6.0% lower (p < 0.05). The 38-year enclosing treatment slightly decreased the bulk density and significantly decreased the electrical conductivity in each soil layer (0–100 cm). The 38-year enclosing treatment significantly increased the stocks of soil organic carbon (SOC), available phosphorus (AP), and available potassium (AK) on the surface soil, and obviously decreased the stocks of total nitrogen (TN), total phosphorus (TP), calcium (Ca), magnesium (Mg), sulfur (S), and available nitrogen (AN) in each soil layer (0–100 cm). In conclusion, long-term enclosing improved grassland production, but decreased most nutrient stocks in soil.

1. Introduction

Grazing is one of the most important ways for human beings to use grassland, and it plays crucial role in material cycles of ecosystems [1]. Grazing not only directly affects the community structure, primary productivity, biodiversity, and ecosystem stability [2], but also indirectly affects the physical and chemical properties of soil [3]. Livestock only absorb a small amount of the nutrients they ingest, and about 75–95% of nutrients return to soil through the form of urine and dung [4]. Moderate grazing can increase primary production of plants through compensatory growth. However, the loss of the balance of grassland ecosystems is often caused by unreasonable grazing [5].

Enclosing is usually deemed as a simple, effective, and economic method to rehabilitate and reconstruct degenerated grassland [6,7]. It is generally believed that grazing exclusion can improve vegetation production, increase the richness and biomass of species, and accumulate litter [8]. Enclosing also decreased the sand particle size composition and bulk density of soil, and improved soil fertility [9,10,11,12]. However, a negative effect mechanism on soil physical and chemical properties has been reported by some researchers [13,14,15]. Shan et al. (2011) indicated that long-term enclosing significantly decreased soil TN and TP contents in typical Xilin Gol grassland [16]. A large amount of vegetation in long-term grazing exclusion grassland needs more nutrients, and the nutrients might transfer from soil to plants. With the accumulation of litter in long-term enclosing grassland, the nutrients might exist in litter with a low decomposition rate.

The Ca, Mg, and S are the essential plant nutrients and AN, AP, and AK play vital roles in plant growth as well, but these mineral nutrients and available nutrients have been largely ignored in recent studies. The aims of present study were to: (1) analyze the differences in green plant biomass and litter biomass of long-term grazing and enclosing; (2) explore the changes in surface soil texture and the vertical changes in pH value, bulk density, and electrical conductivity of long-term grazing and enclosing; (3) clarify the vertical distributions (0–100 cm) of SOC, TN, TP, TK, Ca, Mg, S, AN, AP, and AK stocks of long-term grazing and enclosing. We hypothesized that: (1) long-term enclosing would increase the dry matter of green plants and litter, and improve soil physical properties; (2) the nutrient stocks in soil would be different between long-term grazing and long-term enclosing. This information will provide the basic guidelines for grassland management and sustainable development of grassland in Inner Mongolia.

2. Methodology

2.1. Study Site and Experimental Design

The study was carried out at the Inner Mongolia Grassland Ecosystem Research Station (IMGERS, 43°38′ N, 116°42′ E) of the Chinese Academy of Sciences, which belongs to the Xilin River Basin of Inner Mongolia, China. The climate in the region is a semi-arid steppe climate. The annual average temperature is 2.3 °C and the annual average precipitation is 330 mm (Figure 1) [17]. The soil is dark chestnut with a loamy sand texture. The original forage species were Stipa grandis and Leymus chinensis, and the mean percentage of Stipa grandis and Leymus chinensis was 57% and 21%, respectively. Strong winds usually occur from March to May, and the average monthly speed of wind is up to 4.9 m·s⁻¹. Wind erosion and dust storms often occur in this region [17]. The experimental site includes plot enclosed long term for 38 years inside a fence and a long-term grazing plot outside a fence. Livestock were excluded from the long-term enclosing plot from 1981, and the vegetation and soil have recovered well. The grazing intensity was about 5 sheep·hm⁻¹·year⁻¹ in the long-term grazing plot.

2.2. Sampling and Analysis

The total sampling area was 5000 m². A 50 m long transect was randomly placed in each plot on 9 August 2018. In the long-term grazing plot and long-term enclosing plot, five quadrats were established at 10 m intervals in each plot, and five replicates were performed. Plant and litter samples were collected within concentric quadrats of 1 m × 1 m, and then weighed after drying at 105 °C in the laboratory. From each soil sampling location, three soil cores were collected to make a composite sample for each soil layer (0–10 cm, 10–20 cm, 20–30 cm, 30–50 cm, 50–70 cm, and 70–100 cm depth). The soil samples were air-dried, and then coarsely ground through a 2 mm sieve after stones, roots, and other plant residues were removed.

Soil pH value and electrical conductivity (EC) were determined in a 1:5 (w/v) ratio (H₂O) using a pH meter and an EC meter, respectively [18]. Soil bulk density (BD) was measured by the metal-ring method. The H₂SO₄-K₂Cr₂O₇ method was used to determine soil organic carbon (SOC). The Kjeldahl method was used to determine soil total nitrogen (TN). A flame atomic absorption machine was used to measure soil calcium (Ca), magnesium (Mg), sulfur (S), total phosphorous (TP), and total potassium (TK). The diffusion method, UV spectrophotometer method, and flame photometer method were used to measure available nitrogen (AN), available phosphorous (AP), and available potassium (AK) [19]. The wet sieving method was used to determine soil particle size distributions [20].

2.3. Statistical Analysis

SPSS 10.0 was used to analysis all data. Group means were compared by an analysis of variance (ANOVA). Multiple comparisons among the means of soil parameters between two treatments in each soil layer were analyzed by Tukey’s honestly significant difference test. Significant differences between treatments were determined at 95% and 99% confidence levels.

Nutrient stock was calculated as follows:

Nutrient stock (g/m²) = nutrient content (g/kg) × bulk density (g/cm³) × soil depth (m) × area (1 m²) × 0.9 × 1000.

3. Results

3.1. Dry Matter of Green Plants and Litter

The 38-year enclosing treatment indeed increased the dry matter of green plants and litter compared to long-term grazing (p < 0.01) (Figure 2). The dry matter of green plants with the grazing treatment and 38-year enclosing treatment was 114.8 g/m² and 160.9 g/m², respectively. The dry matter of litter with the 38-year enclosing treatment reached 494.3 g/m², while there was almost no litter with the grazing treatment.

3.2. Soil Physical Properties

3.2.1. Soil Texture

Soil particle sizes were partitioned into five size fractions, and we designated <2 μm fraction as clay, 2–50 μm fraction as silt, 50–100 μm fraction as very fine sand, 100–250 μm fraction as fine sand, and 250–1000 μm fraction as coarse sand. The dominant soil particles in these two treatments were distributed in the size categories of silt (2–50 μm) and sand (50–250 μm), which on average accounted for 37.7% and 51.1% of the total mass, respectively (Figure 3). The size categories of silt (2–50 μm), very fine sand (50–100 μm), and fine sand (100–250 μm) in the grazing treatment accounted for 36.6%, 28.1%, and 24.1%, respectively, and those of the 38-year enclosing treatment accounted for 38.9%, 27.1%, and 22.77%, respectively. The mass percent of silt (2–50 μm) in the 38-year enclosing treatment was 5.9% higher than that of the long-term grazing plot (p < 0.01). The mass percent of fine sand (100–250 μm) in the 38-year enclosing treatment was 6.0% lower than that of the long-term grazing plot (p < 0.05).

3.2.2. Bulk Density, pH Value, and Electrical Conductivity

The soil bulk densities in these two plots slightly increased with increasing soil depth and there were no significant differences among soil layers (Figure 4A). The soil bulk density in 0–10 cm, 10–20 cm, 20–30 cm, 30–50 cm, 50–70 cm, and 70–100 cm soil in the 38-year enclosing treatment was 6.78%, 6.15%, 6.35%, 5.48%, 5.71%, and 4.93% lower than the long-term grazing plot, however, there was no significant difference. There were no obvious changes in pH values in these two plots with increasing soil depth (Figure 4B). The pH value of each soil layer in the 38-year enclosing treatment was slightly higher than the long-term grazing plot. The variation trends of electrical conductivity in these two plots quickly declined in the 0–30 cm soil depth and then quickly increased in the 30–100 cm soil depth (Figure 4C). The 38-year enclosing treatment significantly decreased the electrical conductivity of each soil layer compared to long-term grazing. The electrical conductivity in the 38-year enclosing treatment in the soil depths of 0–10 cm, 10–20 cm, and 20–30 cm was 6.4%, 39.6%, and 19.7% lower than the long-term grazing plot, respectively, and was 131.21%, 84.28%, and 81.17% lower in the soil depths of 30–50 cm, 50–70 cm, and 70–100 cm, respectively.

3.3. Soil Nutrient Stocks

3.3.1. Soil SOC, TN, TP, and TK Stocks

The SOC, TN, and TP stocks in these two plots decreased with increasing soil depth and were significantly higher (p < 0.01) in 0–10 cm soil than other soils depth (Figure 5A–C). However, the TK stocks in these two plots increased with increasing soil depth in the 0–50 cm soil depth (Figure 5D). The SOC stock in 0–10 cm soil in the 38-year enclosing treatment was 14.4% higher (p < 0.01) than in the grazing treatment. There were no significant differences in SOC stocks in 10–100 cm soil between these two plots. The 38-year enclosing treatment decreased the TN and TP stocks compared to the grazing treatment. The TN stock in the 38-year enclosing treatment in 0–10 cm, 10–20 cm, 20–30 cm, 30–50 cm, 50–70 cm, and 70–100 cm soil was 34.7%, 82.5%, 68.3%, 92.3%, 84.9%, and 106.8% lower than in the grazing treatment, respectively. The TP stock in the 38-year enclosing treatment in 0–10 cm, 10–20 cm, 20–30 cm, 30–50 cm, 50–70 cm, and 70–100 cm soil was 5.7%, 2.3%, 2.6%, 6.6%, 7.7%, and 4.5% lower than in the grazing treatment, respectively. There was no significant difference in TK stock in each soil layer between these two treatments.

3.3.2. Soil Ca, Mg, and S Stocks

Compared to the grazing treatment, the 38-year enclosing treatment decreased Ca, Mg, and S stocks in each soil layer (Figure 6A–C). Although there were no obvious changes in Ca and Mg stocks in these two plots with increasing soil depth, the Ca and Mg stocks in 0–10 cm soil were slightly higher than in the soil at 10–20 cm and 20–30 cm. The Ca stock in the 38-year enclosing treatment in the soil of 0–10 cm, 10–20 cm, 20–30 cm, 30–50 cm, 50–70 cm, and 70–100 cm was 23.5%, 18.0%, 13.8%, 10.3%, 19.2%, and 19.7% lower than in the grazing treatment, respectively. The Mg stock in the 38-year enclosing treatment in the soil of 0–10 cm, 10–20 cm, 20–30 cm, 30–50 cm, 50–70 cm, and 70–100 cm was 13.9%, 18.3%, 15.5%, 19.6%, 18.3%, and 17.1% higher than in the grazing treatment, respectively. The S stock in the grazing treatment quickly declined in 0–30 cm soil and was 39.1% and 86.8% higher in 0–10 cm soil than 10–20 cm soil and 20–30 cm soil, respectively. The S stock in the 38-year enclosing treatment in 0–20 cm soil was significantly higher than at other soil depths and was 60.9% higher in 0–10 cm soil than 10–20 cm soil. Compared with the grazing treatment, the S stock in the 38-year enclosing treatment in the soil of 0–10 cm, 10–20 cm, 20–30 cm, 30–50 cm, 50–70 cm, and 70–100 cm decreased by 15.8%, 33.9%, 11.3%, 29.0%, 59.1%, and 58.2%, respectively.

3.3.3. Soil AN, AP, and AK Stocks

The soil AN stocks in these two plots decreased with increasing soil depth and were higher (p < 0.01) in 0–50 cm soil than 50–100 cm soil (Figure 7A). The AN stock in the grazing treatment in 0–10 cm soil was 13.7%, 26.9%, and 37.0% higher than the soil of 10–20 cm, 20–30 cm, and 30–50 cm, respectively. The AN stock in the 38-year enclosing treatment in 0–10 cm soil was 8.8%, 18.3%, and 23.5% higher than the soil of 10–20 cm, 20–30 cm, and 30–50 cm, respectively. The 38-year enclosing treatment significantly decreased the AN stock in 0–50 cm soil compared to the grazing treatment and was 17.0%, 12.0%, 9.1%, and 5.4% lower in the soil of 0–10 cm, 10–20 cm, 20–30 cm, and 30–50 cm, respectively. The soil AP stocks in these two plots decreased with increasing soil depth (Figure 7B). The AP stock in the grazing treatment in 0–50 cm soil was significantly higher than in 50–100 cm soil and was 21.1%, 37.5%, and 56.4% higher in 0–10 cm soil than 10–20 cm, 20–30 cm, and 30-50 cm soil, respectively. The AP stock in the 38-year enclosing treatment in 0–20 cm soil was significantly higher than in other soil layers. The AP stocks in the 38-year enclosing treatment in the soil of 0–10 cm and 10–20 cm were 11.6% and 8.0% higher than in grazing treatment, respectively. However, AP stock was 27.1% and 26.0% lower in the 38-year enclosing treatment in the soil of 20–30 cm and 30–50 cm relative to the grazing treatment, respectively. The soil AK stock in the grazing treatment in 0–10 cm soil was significantly higher (p < 0.01) than other soil depths (Figure 7C). The soil AK stock at in the 38-year enclosing treatment in 0–20 cm soil was significantly higher (p < 0.01) than other soil depths and 86.2% higher (p < 0.01) in 0–10 cm soil relative to 10–20 cm soil. The soil AK stock in the 38-year enclosing treatment in the soil of 0–10 cm and 10–20 cm was 26.2% and 44.8% higher than in the grazing treatment, respectively.

3.4. Pearson Correlation among Soil Nutrients and Dry Matter of Green Plants and Litter

We carried out Pearson correlations among soil nutrients in 0–10 cm soil and dry matter of green plants and litter. As shown in

Table 1. Pearson correlation coefficient among soil nutrients and dry matter of green plants and litter.

	SOC	N	p	K	Ca	Mg	S	AN	AP	AK
Dry matter of green plants	−0.758	−0.783	−0.564	0.806	−0.517	−0.648	−0.347	−0.913 *	0.940 **	0.958 **
Dry matter of litter	−0.680	−0.925 **	−0.513	0.818 *	−0.465	−0.513	−0.261	−0.851 *	0.794	0.897 *

Note: * and ** show differences at p < 0.05 and p < 0.01, respectively.

, dry matter of green plants was negatively correlated with the soil AN (r = −0.913, p < 0.05), and positively correlated with the soil AP (r = 0.940, p < 0.01) and AK (r = 0.958, p < 0.01). Dry matter of litter was negatively correlated with the soil TN (r = −0.925, p < 0.01) and AN (r = −0.851, p < 0.05), and positively correlated with the soil TK (r = 0.818, p < 0.05) and AK (r = 0.897, p < 0.05).

4. Discussion

4.1. Dry Matter of Green Plants and Litter

The dry matter of green plants represents the productivity of above-ground vegetation. In this study, the dry matter of green plants in the grazing treatment and the 38-year enclosing treatment was 114.8 g/m² and 160.9 g/m², respectively. The results showed that long-term enclosing significantly increased the dry matter of green planta, which suggested that excluding grazing for 38 years was effective in vegetation restoration, and this result was consistent with some studies [21,22,23]. The significant improvement of vegetation after enclosing was directly related to the decreased consumption by grazing animals. Yan et al. (2008) studied the effects of different grazing and enclosing times on green plants in the same region (N 43°33′, E 116°40′) as our study, and showed that the dry matter of green plants in a plot grazed for 26 years and a plot where grazers were excluded for 26 years was 30.2 g/m² and 108.2 g/m², respectively [24]. Based on the result of Yan et al. (2008), we found that the dry matter of green plants in the 38-year grazing plot was higher than that of 26-year grazing plot, which might be attributed to the interannual changes in temperature and precipitation, and the difference in sampling time [24]. We found that there was 52.7 g/m² more dry matter of green plants in the 38-year enclosing plot relative to the 26-year enclosing plot [24], which indicated that the average of the dry matter of green plants only increased 4.4 g/m² per year with continuous enclosing for 12 years after enclosing for 26 years. In addition, Yan et al. (2008) showed that the dry matter of green plants when grazers were excluded for 6 years and 2 years was 113.1 g/m² and 96.9 g/m², respectively [24]. This further showed that there was small increase in the green plant biomass with increasing enclosing time, and the vegetation could be restored in a short time by enclosing. Litter refers to the vegetation that cannot conduct photosynthetic function, generally including above-ground standing dead plants and litter at ground level [25,26]. Litter plays critical roles in maintaining soil fertility and improving energy cycles in grassland [27]. Our study showed that there was almost no litter in the grazing treatment. Production and loss determined the amount of litter, hence, we inferred that the loss of litter in the grazing treatment was far greater than the production of litter, which might be ascribed to the fast decomposition of litter by the action of livestock. Yan et al. (2008) showed that the dry mass of litter after enclosing for 2 years, 6 years, and 26 years was about 68 g/m², 440 g/m², and 198 g/m², respectively [24]. Our study showed that the dry matter of litter after enclosing for 38 years was 494.3 g/m². This indicated that the litter could largely accumulate in short time or long term when enclosing.

4.2. Soil Physical Properties

The improvement of soil properties was good for the growth and production of plants. It is usually believed that grazing can change particle size composition, increase soil bulk density, and cause soil erosion [4]. Wind erosion and vegetation coverage play important roles in soil texture evolution [12]. The critical diameter of grain will become larger as vegetation coverage decreases and wind erosion intensity increases. Lower grazing pressure resulted in a large amount of vegetation which might have increased surface roughness and thus favored the deposition of fine soil particles carried by the wind [12,28]. This study showed that the grazing treatment decreased (p < 0.01) the amount of silt (2–50 μm) and increased (p < 0.05) the amount of fine sand (100–250 μm) of surface soil (0–10 cm) compared to long-term enclosing, which suggested that surface soil became coarser with long-term grazing in semi-arid grasslands of Inner Mongolia. The mass percent of clay and slit (<50 μm) of 0–10 cm soil was 39% in grassland where grazers were excluded for 11 years, and it was 5.57 times in continuously grazed grassland in Xilin Gol [29]. Tang et al. (2016) indicated that, generally, greater fine fractions and smaller coarse fractions are associated with reduced grazing pressure [30]. Yan et al. (2010) found that wind erosion over 24 years increased the mass percent of surface soil sand by 31.6% on overgrazed grassland in the typical steppe of Inner Mongolia [31]. In addition, Hoffmann et al. (2008) found that the deposition rate of sand in spring could reach 2.5 g/m²·d in grazing excluded grassland in Xilin Gol, and it was 85% higher than in grazing grassland [17]. Generally, the increase in soil bulk density is a critical indicator of soil degradation in grazing grassland. The higher soil bulk density in the grazing treatment relative to grazing exclusion was probably caused by animal trampling [32]. This study also found that long-term grazing increased the bulk density of each soil layer compared to long-term enclosing, which might be attributed to the hoof action of livestock [33,34]. The lower soil bulk density in the 38-year enclosing treatment might also be attributed to the growth of plant roots. The pH value could influence the mineralization and accumulation of nutrients in arid and semi-arid regions. In our study, long-term grazing decreased the pH value compared to long-term enclosing, which indicated that the soil mineralization in the grazing treatment was greater. In addition, we found that long-term grazing significantly increased the electrical conductivity of 30-100 cm soil, which might be attributed to the leaching of livestock excreta and the mineralization of plants and litter.

4.3. Soil Nutrient Stocks

Soil nutrients of grassland are mainly dependent on the inputs and outputs of nutrients, such as the amount of litter and dung, the decomposition rates of litter and dung, and livestock consumption [35,36,37]. Our study showed that long-term enclosing increased the SOC stock in surface soil, which was in agreement with the result of Ma et al. (2016) who indicated that SOC concentration in an enclosing plot was significantly higher than in a grazing plot in the Qinghai-Tibetan Plateau [38]. Long-term grazing could accelerate the decomposition of soil organic matter owing to the larger area of bare ground, ground temperature, and soil respiration for the reduction of green plants and litter [7,39,40]. Trampling of the vegetation by livestock hooves may cause a short-term increase in microbial mineralization of native soil carbon as a result of fresh carbon inputs [41]. At present, many studies have reported that a higher mass percent of fine sand and a higher bulk density could reduce SOC content. Yan et al. (2008) found that SOC content was significantly positively correlated with the mass percent of clay particles, and for every 1% increase in sand content in semi-arid grassland of Inner Mongolia, the SOC content would decrease 0.3345 g/kg [24]. Hence, the lower SOC stock in surface soil in the grazing treatment might also be attributed to the lower silt (2–50 μm) and higher fine sand (100–250 μm) levels. This study found that long-term enclosing obviously or slightly decreased the TN, TP, TK, Ca, Mg, and S stocks. Shan et al. (2011) indicated that long-term enclosing was beneficial for the restoration of physicochemical properties of soil, but significantly decreased soil TN and TP contents in typical Xilin Gol grassland [16]. Livestock only use a small proportion of the nutrients they ingest, and with about 75–95% return to the soil in the form of dung or urine which leads to a redistribution of soil nutrients in grazing grassland [4]. The TN, TP, TK, and Mg concentrations of herbage in grazing grassland were higher than enclosing grassland, which suggested that the mineral cycle of grazing grassland was greater. The C/N ratio of plants in grazing grassland was lower than that of enclosing grassland, therefore the decomposition rate of plant residue in grazing grassland was faster, which suggested that the grazing improved the nutrient cycling of the ecosystem. However, a large amount of litter with low decomposition rates accumulated in enclosing grassland might limit nutrient cycling. In addition, greater amount of vegetation in enclosing grassland needs a large amount of soil nutrients, which directly reduces soil nutrients. In this study, the grazing treatment slightly decreased the TK content in each soil layer, therefore, the higher TK stock in the grazing treatment was mainly caused by the higher soil bulk density. The P, Ca, Mg, and K in feed for cows were excreted into the grassland at 65%, 78%, 80%, and 11%, respectively, which might have contributed to the slight reduction in TK stock in grazing grassland [42]. In our study, we found that AN stocks in 0–50 cm soil in the grazing treatment were higher than that of the 38-year enclosing treatment, which might be attributed to greater mineralization of organic N and greater microbial activities in dung [6]. In addition, this study found that the grazing treatment significantly decreased the AP and AK stocks in surface soil, which was closely related to lower SOC stocks in surface soil.

In this study, we found that the stocks of most of nutrients in these two plots decreased with increasing soil depth, which was consistent with previous studies [43]. This tendency might be closely related to the effects of decomposition of plant residue, litter, and animal excreta in surface soil. In addition, the higher nutrient stocks in surface soil might also reflect atmospheric input and legume plant fixation of N.

5. Conclusions

Compared with grazing for 38 years, long-term enclosing was beneficial to restoring vegetation and improving soil physical properties, as well as significantly increasing the SOC, AK, and AP stocks in surface soil. However, long-term enclosing causes the accumulation of a large amount of litter and obviously decreased TN, TP, Ca, Mg, S, and AN stocks in the soil profile. Therefore, long-term enclosing could improve grassland production, but decrease most of the nutrient stocks in soil.