4.2 Effect of grassland managements on carbon and nitrogen stocks
The grazing regime alters ecosystem carbon and nitrogen cycles through livestock feeding, trampling, and manure return in grassland ecosystems (Tserang Donko Mipam et al., 2021; G. Zhou et al., 2017). We found that GE significantly increased soil organic carbon storage in this study, which is consistent with other studies in alpine meadows (Xiong et al., 2014), desert steppe (Niu et al., 2011), and arid and semiarid grasslands (Yu et al., 2021). A meta-analysis concluded that GE mostly increased the soil C pool from 78 study sites in the Tibetan alpine grassland (Yu, Chen, Sun, & Huang, 2019) and 164 sites across grasslands worldwide (Abdalla et al., 2018). This positive effect may result from the following mechanisms. First, we found that GE significantly increased the aboveground and root biomass (Fig. 1). A study (Xiong et al., 2014) indicated that the soil organic carbon and nitrogen stocks of fencing grasslands were positively correlated with plant root biomass. Positive feedback may exist between soil organic carbon and plant biomass (Dai, Guo, et al., 2021), because most soil organic carbon comes from root exudates and litter decomposition (Kaiser, 2000; Xiong et al., 2014). Fencing-induced increases in root, litter, and aboveground biomass subsequently increase organic carbon input and promote the contribution of root-derived C to SOC by increasing root carbon content (Su & Xu, 2021; Yang, Wang, & An, 2021), whereas, grazing decreases the organic matter in the soil by consumption of the plant. This was supported by a significant positive correlation between plant biomass and SOC stocks (Fig. 5). Second, soil stoichiometry plays a critical role in the regulation of grassland nutrient cycles (Chen, Wang, & Baoyin, 2021). A previous study indicated that GE promoted nitrogen release from roots and litter, thereby increasing organic carbon (Yu et al., 2019). We found that the C/N ratio was significantly higher after years of GE than that observed after different grazing methods (Table 1), suggesting that the grassland may have nitrogen restrictions without exogenous N input after years of fencing, while a lower C/N ratio of the grazing methods can promote the decomposition rate of soil microorganisms and thus reduce SOC (Yang et al., 2021). Additionally, GE could facilitate soil moisture and temperature and decrease soil degradation by increasing plant cover and preventing livestock trampling, respectively, which in turn stimulates soil microbial biomass C and plant growth and reduces soil carbon leaching (Feyisa et al., 2017; Niu et al., 2011; Xiong et al., 2014). Furthermore, grazing causes a decrease in palatable species, which have faster litter decomposition and nutrient release than unpalatable species, leading to a lower SOC content than GE (Dai, Fu, et al., 2021). However, few studies have found that GE decreases SOC accumulation and storage in subtropical (Wilson, Strickland, Hutchings, Bianchi, & Flory, 2018), semiarid (Chen et al., 2021), alpine grasslands (Wu et al., 2021) and causes no variation in SOC of alpine meadows (Yuan & Jiang, 2021). These different conclusions may be due to variations in belowground allocation, such as root biomass, fine root exudates, and microbial biomass (Wu et al., 2021).
Contrary to SOC stocks, GE significantly decreased soil TN stocks compared to grazing methods in the present study (Fig. 2b), indicating that grazing improved the soil N content in this region. This result is in agreement with other studies (Contosta et al., 2021; Wu et al., 2021; Y. Zhou et al., 2020) and a meta-analysis showing that grazing led to a decrease in SOC but an increase in TN stocks at a global scale (Abdalla et al., 2018). Grazing results in higher levels of plant N, efficient use of N uptake, and accelerated N release from roots and plankton, leading to a significant increase in TN during grazing (Dai, Fu, et al., 2021; Yu et al., 2019). A study by Zhu, Liu, Wang, Sun, & Han (2021) found that grazing disturbances could change the activities of nitrogen assimilation-related enzymes which are beneficial for nitrogen assimilation by grassland plants. On the contrary, grazing could increase soil N content by returning excreta from livestock into the soil, as urine and faeces contain a large amount of available N for uptake (Oenema, Oudendag, & Velthof, 2007). Meanwhile, the grazing methods were implemented using 60 kg·ha-1 urea and 300 kg·ha-1 of superphosphate in late June and mid-to-late October, respectively; this suggests that the grazing grassland contains sufficient nitrogen but lacks C due to plant intake when compared to grassland subjected to GE, which resulted in a lower C/N ratio and in turn inhibited the decomposition of soil organic matter for nitrogen utilisation (Kumar, Kundu, Ghorai, Mitra, & Singh, 2018). We found that soil TN stock was significantly negatively correlated with the C/N ratio (Fig. 4). Furthermore, soil N content may depend on stocking rates because soil N sequestration potential tends to decrease with increasing stocking rates, which is beneficial for soil N sequestration; however, heavy grazing leads to lower N sequestration potential and N loss (Y. Li et al., 2012). The stocking rate in this study was 10 sheep units / hm2. However, some previous studies have reported that GE increased SOC stocks and TN stocks in alpine meadows (W. Li et al., 2011; Xiong et al., 2014). The meta-analysis of 32 studies reported that grazing significantly decreased both SOC (-20%) and TN (-15%) in alpine grasslands (Yan et al., 2020). As very few meta-analyses have been conducted on the responses of soil SOC and TN stocks to grazing at regional and global scales, the effects and related mechanisms have not been consistently determined due to the complicated effects of grazing on soil-plant systems depending on the regional climate (precipitation gradient, mean annual temperature), altitude, soil depth, stocking rate, fencing years, and preceding grazer type to some degree (Abdalla et al., 2018; Su & Xu, 2021; Yan et al., 2020; G. Zhou et al., 2017). For instance, grassland SOC was enhanced by mean annual temperature, but stocking rates had little effect, regardless of the mean annual temperature (Throop, Munson, Hornslein, & McClaran, 2022).