Introduction
Grasslands are one of the largest terrestrial ecosystems on Earth (Suttie et al. 2005). Aboveground biomass production is one of the important functions in grasslands, providing many essential services to humanity such as feed for livestock, carbon storage, and climate mitigation (Bengtsson et al. 2019). However, increasing nitrogen deposition (Galloway et al. 2004) and accelerating herbivore extirpation (Ripple et al. 2015; Atwood et al. 2020) are altering aboveground biomass and stable provision (i.e. stability) of aboveground biomass in grasslands (Hautier et al. 2015, 2020; Blüthgen et al. 2016; Qin et al. 2019; Zhang et al.2019; Borer et al. 2020). Stability is a multi-dimensional concept (Donohue et al. 2013; Arnoldi et al. 2019); here we examine temporal invariability, calculated as the mean of aboveground biomass through time divided by its standard deviation in local communities (alpha stability) and larger spatial scales (aggregated local communities; gamma stability). Metacommunity theory clarifies that alpha stability and asynchronous dynamics among local communities (spatial asynchrony) determine gamma stability (Wang & Loreau 2016; Wang et al. 2019).
In grassland ecosystems, nutrient addition often decreases alpha stability (Hautier et al. 2015; Koerner et al. 2016; Zhanget al. 2016; Liu et al. 2019), and these effects may propagate to gamma stability (Zhang et al. 2019; Hautier et al. 2020). In contrast, herbivore exclusion has been found to have positive, neutral, or negative effects on grassland stability, depending on the herbivore species excluded and spatial scales studied (Halpern et al. 2005; Hautier et al. 2015; Blüthgen et al. 2016; Ren et al. 2018; Ganjurjav et al. 2019; Qin et al. 2019; Saruul et al. 2019; Liu et al. 2021). Effects of herbivores on stability may be more apparent at the larger spatial scale because herbivores usually promote vegetation heterogeneity in the landscape due to selective grazing, trampling, and localized deposition of urine and dung (Glenn et al. 1992; Howison et al. 2017). However, to our knowledge, no study has investigated how nutrients and herbivores jointly regulate grassland stability across multiple spatial scales.
In contrast to stability, a range of studies have examined the joint effects of nutrients and herbivores on grassland species richness (Proulx & Mazumder 1998; Worm et al. 2002; Bakker et al.2006; Hillebrand et al. 2007; Alberti et al. 2010, 2011; Yang et al. 2013; Borer et al. 2014b; Beck et al.2015; Koerner et al. 2018), community evenness (Hillebrandet al. 2007), community composition (Milchunas & Lauenroth 1993; Chase et al. 2000; Grellmann 2002; Hartley & Mitchell 2005; Alberti et al. 2017; Hodapp et al. 2018), and aboveground biomass (Milchunas & Lauenroth 1993; Chase et al. 2000; Moran & Scheidler 2002; Alberti et al. 2010, 2011; Borer et al.2020). The majority of these studies found strong interactive effects between nutrients and herbivores. For instance, herbivores typically decrease species richness at low nutrients or productivity, while increasing it at higher nutrients or productivity (Proulx & Mazumder 1998; Bakker et al. 2006; Hillebrand et al. 2007; Boreret al. 2014b). Herbivores also can consume extra aboveground biomass stimulated by nutrient addition particularly when their abundance is high (Borer et al. 2020).
As plant diversity and biomass can impact stability indirectly and directly, it is likely that nutrients and herbivores also jointly impact grassland stability. For instance, nutrients or herbivores can indirectly impact gamma stability by regulating species richness at the local and larger spatial scales (alpha and beta diversity, respectively) (Hautier et al. 2015; Zhang et al. 2019; Liang et al. 2020). Although less often tested, nutrients or herbivores can also impact alpha and gamma stability through community evenness (Grman et al. 2010; Liang et al. 2020) and community dissimilarity across time and space (Koerner et al. 2016; Zhang et al. 2019). Our understanding of plant diversity (including alpha and beta diversity, evenness, community dissimilarity across time and space) on stability are limited because different studies usually focus on different plant diversity metrics (Grime 1998; Tilman et al. 2006; Polleyet al. 2007; Grman et al. 2010; Hautier et al.2015; Koerner et al. 2016). Therefore, it remains unclear which plant diversity metrics are the major factors mediating the effects of nutrient and herbivores on alpha and gamma stability. Assessing the relative contribution of different facets of plant diversity to alpha and gamma stability can deepen our understanding of the underlying mechanisms for stability across spatial scales and help prioritize conservation efforts.
Here, we used a globally coordinated grassland experiment, Nutrient Network (NutNet) (Borer et al. 2014a) to assess (1) the joint effects of nutrient addition and herbivore exclusion on the temporal stability of aboveground biomass at the local and larger spatial scales (i.e. alpha and gamma stability); (2) the relative contribution of different facets of plant diversity in mediating the responses of alpha and gamma stability to nutrients and herbivores. We hypothesized that (1) nutrient addition decreases alpha and gamma stability, while herbivore exclusion worsens these decreases particularly at sites with high grazing intensity; (2) Nutrient addition decreases gamma stability via reducing alpha stability, which is regulated by all facets of plant diversity. Herbivore exclusion decreases gamma stability not only via alpha stability but also via spatial asynchrony, because herbivores often increase vegetation heterogeneity in the landscape (i.e. spatial community dissimilarity).