Introduction
Grassland ecosystems cover a large proportion of the arid and semi-arid regions of the world, playing an important role in the global carbon cycle (Scurlock and Hall 1998; Schuman et al. 2002; Zhou et al. 2019; Lei et al. 2020). The degree to which terrestrial ecosystems serve as net carbon sinks or sources depends on the balance between the carbon fixed by plant photosynthesis and the carbon released into the atmosphere by plant and soil respiration (Peng et al . 2014; Li et al. 2017b; Jin et al. 2023). While numerous evidence has addressed the degree to which forested systems serve as net carbon sinks or sources (Martens et al. 2004; Rebane et al. 2020), the status of grassland ecosystems as sources or sinks can be highly variable (Dai et al. 2014; Smith 2014; Chang et al. 2021). Grasslands can be an important carbon sink in some places or times (Hafner et al. 2012; Sha et al. 2020), but in other places and times, grassland carbon fluxes can be in equilibrium (neither a source or sink) (Hao et al. 2017), a net carbon source (Kuzyakov and Gavrichkova 2010), or fluctuate between states (Daiet al. 2014; Zhang et al. 2020).
Livestock grazing is a major land-use category by which human activities can influence the structure and function of grassland ecosystems, profoundly altering the carbon cycle and stability of grassland productivity (Zhang et al. 2023). Grazing directly affects plant productivity and respiration because livestock foraging removes leaves while promoting compensatory growth, and their trampling and excretion redistributes soil organic matter and alters soil respiration (Caoet al. 2004; Chen et al. 2015; Barthelemy et al.2018; Veldhuis et al. 2018); Grazing also alters soil nitrogen content and other processes important to the carbon cycle, such as litter decomposition and photosynthate distribution (Xia and Wan 2008). As a result, grazing can moderate the net ecosystem exchange of grasslands and whether they serve as a net carbon sink or source. In some cases, light to moderate levels of grazing can facilitate grasslands being net carbon sinks (Derner et al. 2006; Shaet al. 2020; Chang et al. 2021), while high levels of grazing can accelerate the release of carbon and switch the ecosystem from a carbon sink to a carbon source (Liang et al. 2017; Tanget al. 2018); In other cases, grazing appears to have little influence on the carbon budget of grassland ecosystems (Fang et al. 2010; PiƱeiro et al. 2010).
Desert steppe is particularly vulnerable to degradation due to livestock grazing and to carbon sinks transitioning to carbon sources (Zhanget al. 2020). We assessed ecosystem carbon balances over a sustained 10-year period and explored the influencing factors. We concluded that precipitation patterns and grazing combine to cause changes in the carbon sink function of grasslands (Jin et al.2023; Wang et al. 2023), but that ecosystem carbon exchange is disturbed by a combination of environmental (soil, climate) and biological (grazing) factors. How environmental and biological factors influence net ecosystem carbon exchange depends on the relationship between carbon uptake via primary productivity and carbon release via plant and soil respiration. Furthermore, there is considerable variability and uncertainty regarding the factors influencing carbon exchange in grassland ecosystems (Liu et al. 2015; Sha et al. 2020). This is likely because the variability in grassland types is mediated by climate, vegetation and soil (Helfter et al. 2015; Hussain et al. 2015; Liang et al. 2020), as well as by grazing practices (Fang et al. 2010; Dai et al. 2014).
Thus, simply measuring net ecosystem exchange and aboveground biomass is not enough to fully understand the influence of biotic and abiotic factors on these rates (Li et al. 2017a; Bajgain et al.2018). It is necessary to more fully identify how carbon exchange and soil respiration are influenced by grazing and background environmental factors, in particular. This information will not only help us better understand the factors influencing the carbon dynamics of these important ecosystems, but it will also help inform the formulation of policies for the sustainable management and conservation of grassland resources. In this study, we measured ecosystem carbon fluxes and their associations in respond to a long-term (16-year) grazer manipulation experiment in a desert steppe grassland in Inner Mongolia, China. We specifically asked (1) how does grazing influence features of the plant community and soil conditions and (2) how those effects influence the parameters of net ecosystem carbon exchange, including gross ecosystem productivity and respiration. On the basis of our previous research, we further measured aboveground and belowground biomass, plant nutrients (carbon and nitrogen content of plant communities) and soil nutrient indexes to analyze the main drivers that influence the changes of CO2 fluxes in desert steppe ecosystems and their responses to grazing disturbances, and to provide theoretical basis for the adaptive management of desert steppe.