1 INTRODUCTION
Climate change may amplify inter-annual precipitation fluctuation in terrestrial ecosystems (Alan K. Knapp, Avolio, et al., 2017; Sloat et al., 2018; W. Zhang & Zhou, 2019), and interactions between precipitation fluctuation and anthropogenic activities (e.g. overgrazing, land use change and fertilisation) could have significant impacts on grassland ecosystem structure and functioning (Ahlborn et al., 2021; Brambila, Chesnut, Prugh, & Hallett, 2020; Cui et al., 2020a; Han et al., 2021; L. Huang et al., 2019). Arid and semi-arid grassland ecosystems generally experience high inter-annual precipitation variation (Carmona, Mason, Azcarate, & Peco, 2015; Alan K. Knapp, Ciais, & Smith, 2017; A. K. Knapp & Smith, 2001; Kong et al., 2013) and anthropogenic activity–induced soil nutrient enrichment (Ma et al., 2020; Yahdjian, Gherardi, & Sala, 2011). In grassland ecosystems, arbuscular mycorrhizal (AM) fungi can colonize over 80% of terrestrial plants, which can improves host plant elements absorption (Smith & Read, 2008). However, much of our current understanding of plant community performances in desert steppe in relation to inter-annual precipitation fluctuation is based on the changes of plant productivity and diversity (Ahlborn et al., 2021; Han et al., 2021; Alan K. Knapp, Avolio, et al., 2017) and assessed independently of the effects of added nutrients and AM fungi (J. Huang et al., 2021; J. Huang et al., 2016; Qiao et al., 2019). As such, it is critical that we clarify plant community responses to the potential interactive impacts of multiple factors such as inter-annual fluctuation in precipitation, soil nutrient availability and AM fungi. Additionally, a reliable benchmark is needed to develop a robust understanding of arid grassland sensitivity using terrestrial ecosystem models (Püschel, Bitterlich, Rydlová, & Jansa, 2021; Xin Yang, Shen, Badgery, Guo, & Zhang, 2018).
Precipitation in growing season is a vital driver affecting variation in ANPP and plant community composition in grasslands (Bai, Han, Wu, Chen, & Li, 2004; Irisarri et al., 2016). However, plant communities can vary drastically in response to precipitation fluctuation (Hsu, Powell, & Adler, 2012; Alan K. Knapp, Avolio, et al., 2017; Koerner, Collins, Blair, Knapp, & Smith, 2014; Reichmann et al., 2018). For example, annual primary production is more sensitive to inter-annual fluctuation in precipitation in arid grassland than in humid grassland (Maurer, Hallmark, Brown, Sala, & Collins, 2020). Furthermore, evidence from recent field studies shows that environmental factors (e.g. soil nutrient availability and soil microorganisms) may affect plant communities in response to inter-annual precipitation fluctuation (Felton, Slette, Smith, & Knapp, 2020; J. Huang et al., 2016). Thus, our current konwledge of the responses of plant communities to inter-annual precipitation fluctuation may be impeded by a limited understanding of multiple factors.
Anthropogenic input of nitrogen (N) and phosphorous (P) can synergistically promote aboveground net primary productivity (ANPP), reduce plant species richness and change the dominance of plant species in nutrient-limited grasslands (Suding et al., 2005). The impacts of nutrients input on ANPP and plant diversity rely on the abundance and nutrient uptake strategy of plant species; N input favours the dominance of perennial grasses rather than other plant species (Avolio et al., 2014; Ma et al., 2020; Stevens et al., 2020). Nevertheless, P input alone promotes high abundance and shoot biomass in legumes because they have higher P demands than other plant species (Benner & Vitousek, 2007). Thus, it is necessary to consider the nutrient responses of various plant species to effectively interpret how the addition of nutrients changes plant communities.
Nutrient supply to plants can also alter the structure and functioning of AM fungi via changing the exchange of carbon (C) and P (Johnson, Wilson, Wilson, Miller, & Bowker, 2015; Kiers et al., 2011). Field studies found that N and/or P input could weaken the impacts of AM fungi on ANPP and plant species richness (G. W. Yang et al., 2014; T. Zhang & Feng, 2021). In general, AM fungi promote ANPP when soil available P is limited and vice versa (Collins & Foster, 2009; Johnson et al., 2015; G. W. Yang et al., 2014). In contrast, AM fungi have a neutral impact on ANPP when N is limited but reduce it when N is enriched (Jiang et al., 2018). Relatively high soil nutrient availability, especially high P, can decrease belowground C allocation in plants, thus directly reducing C supply for AM fungi (Zheng et al., 2018) and changing this symbiotic relationship from mutualism to parasitism (Jiang et al., 2018; Johnson, 2010).
Desert steppe generally has low ANPP and plant diversity because of soil water and nutrients (N and P) are strangely limited (Alan K. Knapp, Ciais, et al., 2017). Precipitation in the growing season in the desert steppe in China varies drastically between years, ranging from 111.4 to 382.4 mm (Zhao et al., 2019). AM fungi can still benefit plant N and P nutrition in this water- and nutrient-limited conditions (Qiao et al., 2019). Thus, we conducted a two-year field study to evaluate the inter-annual response of ANPP and plant species composition to nutrient addition (N, P and N+P) and AM fungi suppression in desert steppe. We examined the following three hypotheses: (1) inter-annual precipitation fluctuation alters ANPP and plant species composition response to N and P addition and AM fungal suppression; (2) the addition of nutrients enhances ANPP by changing the plant growth response in combination with AM fungi; and (3) inter-annual precipitation fluctuation changes the dominant species in a community regardless of nutrient addition and mycorrhizal suppression.