Effects of different grazing intensities on plant species diversity at different spatial scales in a desert steppe in Inner Mongolia

Sample site overview and experimental design

The experimental site is located in Haden Hushu Gacha Town, Saihantala Town, Sunit Right Banner, Xilin Gol League, Inner Mongolia Autonomous Region (IMAR), China, at 41°55′∼43°39′N latitude, 111°08′∼114°16′E longitude, with an average altitude of 1,150.8 m, and a linear distance of 50.81 km from the banner set up by the government of China. This area has a typical arid temperate continental climate, with cold winters and hot summers. The highest yearly average temperature is 38.7 °C, while the lowest is −38.8 °C, with a large difference in the annual average temperature. The frost-free period is 130 days. From 2000 to 2020, the recorded average annual precipitation was 199.6 mm, while the average annual evaporation was 2,384 mm, and the average speed of prevailing westerly wind was 5.5 m/s. The study site is a semi-desert grassland region situated in the transition zone between steppe and desert, and has mainly brown soil rich in calcium.

The study area is dominated by herbaceous plants such as Stipa breviflora, Cleistogenes songorica, and Allium polyrhizum.

The fence grazing was adopted in this research, and three experimental blocks were randomly set up, with each block having five plots of different grazing intensities. There were five different treatments of grazing intensity (Fig. 1), including no grazing (control) (CK), light grazing (LG), moderate grazing (MG), heavy grazing (HG), and extremely heavy grazing (EG), with 0, 1.54, 1.92, 2.31, and 2.69 sheep units. hm⁻², respectively, grazing intensities were established with 0, 4, 5, 6 and 7 sheep per treatment. Each grazing plot had an area of about 2.6 hm² from the total area of approximately 39.2637 hm². The grazing treatments were applied from early May to late October (Xing et al., 2021). No permits were needed at any locations.

Sampling design

The sampling points were set up using random distribution. There was a total of 15 plots, with each plot being 1 m ×1 m with three samples. The number of occurrences of plants was recorded based on species. Based on the vegetation types (Stipa breviflora + Cleistogenes songorica +Allium polyrhizum) in the study area and the importance value (IV) data obtained from 45 sampled plots, plant species were categorized into three plant groups: dominant species, common species, and rare species (Pan, Liu & Niu, 2018; Yang et al., 2014). Stipa breviflora, Cleistogenes songorica, and Allium polyrhizum were the dominant species in the study area. There were, however, seven common species such as Convolvulus ammannii, Neopallasia pectinata, Setaria viridis, etc., and Asparagus cochinchinensis, Allium mongolicum, and Allium tenuissimum were among the seven rare species (Table 1).

Data processing and analysis

The plant species diversity of a desert steppe under different grazing intensities could be calculated using the following formulas: (1) α diversity index: $\mathrm{\alpha }=\frac{1}{3}{\sum }_{i=1}^3{k}_i,$where:k_i represents the sum of species in the ith 1 m  ×  1 m quadrat; (2) γ diversity index: γ = the total number of species recorded under each grazing intensity; (3) β diversity index: β = γ-α; (4) Importance Value (IV): IV = (relative height + relative cover + relative density)/3 × 100%.

Statistical analysis

All statistical analyses were performed using R-Studio software. The ggplot2 package was used to create the percent stacked bar charts to analyze the responses of different plant community-type frequencies to grazing intensities. The differences in the α diversity of dominant species, common species, and rare species among different grazing intensities were determined by the one-way analysis of variance (ANOVA), and bar graphs were plotted using the ggplot2 and ggsignif packages. Thereafter, the regression analysis was used to evaluate the relationships between five grazing intensities (CK, LG, MG, HG, and EG) taken as continuous variables and diversity indices by linear, quadratic, and exponential functions. The optimal model-fitting was implemented using R² and P values, and the ggplot2 and ggpmisc packages were used to draw the fitted curves. Finally, after averaging the obtained values for diversity indices, we employed the FactoMineR and CA packages in R to perform the corresponding analysis to explore the relationships among different grazing intensities and species diversity indices (α, β, γ, DA, CA, and RA).

Response of species frequency to grazing intensity

The dominant species, including S. breviflora and C. songorica, showed high grazing tolerance, which first increased and then stabilized with the increase in grazing intensity. In contrast, the grazing tolerance of A. polyrhizum is relatively weak, and it belongs to the plant species that are sensitive to grazing. Its frequency decreases as the grazing intensity increases (Fig. 2A). Common species generally exhibited neutral responses to grazing, e.g., changes in grazing intensity hardly resulted in the changing trends of the species frequency; the trends were, however, almost constant or fluctuating (Fig. 2B). Rare species were generally sensitive to grazing. That is, grazing can cause a sharp decline in their frequency (such as Allium mongolicum) or a trend of first increasing and then decreasing (such as Caragana stenophylla), but for Asparagus cochinchinensis, the changing patterns were not obvious, and thus, it can be classified as a grazing-insensitive species (Fig. 2C).

Response of alpha diversity of dominant species, common species and rare species to grazing intensity

The strongest response to different grazing intensities was achieved by the α diversity of rare species, followed by that of dominant species; the α diversity index of common species, however, was the least sensitive to grazing intensity (Fig. 3). The dominant species displayed the α diversity index ranging from 1.33 to 2.67, with a significant difference detected between the heavily grazed areas and the ungrazed (control) areas (P < 0.05). In addition, the α diversity of common species was between 3.67 and 5.33, but no significant difference was noted under different grazing intensities (P > 0.05). Rare species showed α diversity within the range of 0.33 to 2.33, and a changing trend of first decreasing and then stabilizing with the increase in grazing intensity, and different grazing intensities differed significantly (P < 0.05).

Response of α, β and γ diversity to grazing

With increased grazing intensity, all three α diversity, β diversity, and γ diversity underwent an initial decrease, followed by an increase. As shown in Fig. 4, following a downward trend of diversity indices in the CK and different grazing treatments, α and γ diversity indices started increasing slightly in the extremely heavy grazing treatment, while the β diversity index began increasing from the heavy grazing (HG) to EG. This indicates that the optimal grazing intensity was below the MG intensity. Compared with the control treatment, α, β, and γ diversity decreased by 13%, 66%, and 26% under light grazing, whereas under moderate grazing, decreases by 22%, 69%, and 34% were recorded, respectively. Moreover, when grazing was heavy, decreases of 28%, 62%, and 37% were obtained, and extremely heavy grazing caused decreases by 22%, 48%, and 29% in α, β, and γ diversity, respectively. Thus, β diversity was more sensitive to grazing, whereas α diversity was least responsive to it.

Correspondence between grazing intensity and species diversity

The first two principal components (PCs) could explain more than 97% of the variance in original data (Fig. 5), and the goodness-of-fit was very high, indicating that the correlations existed among grazing intensities and diversity indices. At the same time, CK, LG, and MG were found to be above the principal component horizontal axis (Dim1), while HG and EG were below it, which demonstrates that the effects of HG and EG treatments on the species diversity of the plant community were different from those of CK, LG, and MG, which were arranged from left to right along the horizontal axis. These results revealed that there was a linear variation in the species diversity of the plant community under grazing intensities below MG. The α diversity of dominant species (DA) and common species (CA), and the α and γ diversity of the plant community were insensitive to grazing intensity, and therefore, they were positioned in close proximity to LG, MG, HG, and EG. However, the α diversity of rare species (RA) and the β diversity of the plant community were sensitive to grazing, but they showed nonlinear relationships with different grazing intensities, and thus, they were far away from each grazing intensity point along the axis.

Effects of grazing on plant community species diversity

Grazing represents the most extensive land utilization mode worldwide. Grazing is one of the main factors affecting plant diversity (Fernando et al., 2022; Jinlan et al., 2022). In the study area located in a desert steppe, the α and γ diversity indices under the light, moderate, and heavy grazing decreased compared with those in the non-grazed areas (control), indicating that grazing would reduce plant diversity. Heavy grazing should be avoided to protect the diversity and stability of grassland ecosystems, which is different from the intermediate disturbance hypothesis (Michael, 2014). The research results obtained by Takehiro et al. (2007) also showed that the plant species diversity in a desert steppe could not support the intermediate disturbance hypothesis (Takehiro et al., 2007), but were consistent with the “dynamic equilibrium model”, which predicts that competition, predation, and productivity together contribute to controlling species diversity (Huston, 1979; Michael, 2014). The α and γ diversity indices increased under extremely heavy grazing compared to those in the heavily grazed site, which is due to the degradation of vegetation communities by grazing, especially for more biennial plants but fewer perennials (Yousif Mohamed et al., 2020; Zhen’an et al., 2016), and despite the increase in species diversity, these communities remained in a degraded state.

In the ungrazed, lightly grazed, and moderately grazed areas, the linear decrease in the β diversity index with the increase in grazing intensity was observed. Compared with the control treatment, the β, α, and γ diversity indices decreased by 69%, 21%, and 34%, respectively, indicating the highest sensitivity of the β diversity to the grazing intensity, followed by γ diversity and α diversity. The increase in grazing intensity caused decreases in herbivore selectivity (Edina et al., 2016; Péter et al., 2018), and the gradual disappearance of grazing-sensitive species, such as Kochia prostrata and Allium tenuissimum, occurs (Milchunas, Sala & Lauenroth, 1988), resulting in the decrease in the γ diversity index (Julia et al., 2021). The α diversity index, which is a measure of the average number of species within the community (Maria Teresa, Francesc & Enric, 2021), was less affected by grazing intensity. Therefore, the α and γ diversity together induced the sensitivity of the β diversity to grazing intensity.

Effects of plant taxa on species diversity under grazing conditions

At the level of community type, the α diversity of rare species was the most responsive index to changes in grazing intensity (Jiaojiao et al., 2024), followed by that of dominant species, while common species were the least sensitive in terms of alpha diversity. The possible reason is that most of the rare species are grazing-sensitive and highly palatable species (Shijie et al., 2024), such as Allium tenuissimum, Allium mongolicum, etc., that began to decrease or even disappear in light grazing areas (Shuxia et al., 2010). Furthermore, the α diversity of rare species tended to stabilize from moderate to extremely heavy grazing intensities, and there was a significant difference between MG and the control treatment (P < 0.05). The results revealed that although the rare species are inadequately represented in the community (Jennifer Nagel et al., 2022), they play a crucial role in the ecosystem and thus can be used as an important indicator for the assessment of the level of biodiversity (Zavaleta, 2004). The results of the correspondence analysis showed that the β diversity of the plant community and the α diversity of rare species responded strongly to different grazing intensities, which is consistent with the overall results of the present study.