Grazing and climate change have site-dependent interactive effects on vegetation in Asian montane rangelands

Climate over Asian montane rangelands is changing faster than the global average, posing serious threats to the future of the region's livestock-based economies and cultures. Effects of climate change on rangeland vegetation likely depend on grazing by herbivores but the potential responses of vegetation to such changes in climate and grazing regimes remain unclear. 2. We examined vegetation responses to experimentally simulated climate change (warming, drought and increased rainfall) and grazing (clipping vegetation) between 2015 and 2018 at two mountain rangeland sites: Spiti valley, in the Indian Trans-Himalaya and Tost, in the Gobi-Altai Mountains in Mongolia. 3. Clipping and climate change manipulations interactively reduced vegetation cover and biomass but did not affect species richness. Treatment effects and their interactions varied between sites. In ungrazed plots, vegetation cover and biomass declined sharply in response to warming (18%–35%) and drought (20%–50%) at the two sites, and, surprisingly also declined slightly in response to increased rainfall (20%) at Tost. While the effects of climate treatments were largely similar in the grazed and ungrazed plots in Tost, they were larger in the ungrazed plots in Spiti. The decline in vegetation cover was driven by a decline in the cover of both forbs and grasses. 4. In combination, grazing and warming (Tost) or drought (Spiti) had sub-additive effects, that is, the decrease in vegetation cover in response to grazing and warming/drought was less than the sum of their independent effects but greater than the effect of either manipulation alone. Of the two, warming had a greater effect than drought at the more arid site (Tost), whereas drought had a larger effect at the more mesic site (Spiti). 5. Synthesis and applications. Our findings show that future changes in climate, including just over 1°C of warming, could undermine the sustainability of pastoral economies and the persistence of wildlife across Asian montane rangelands. 2 | Journal of Applied Ecology KOHLI et al. 1 | INTRODUCTION Rangelands occupy nearly 45% of the Earth's terrestrial surface (Reid et al., 2008). They are vital for global food security and livelihoods of millions of pastoralists, and provide ecosystem services such as soil erosion control, biodiversity maintenance and carbon sequestration. Because rangeland vegetation dynamics and functioning are jointly shaped by climate and grazing by herbivores (Frank et al., 1998; Liang & Gornish, 2019; Milchunas et al., 1988), changes in climate and grazing (e.g. due to loss or replacement of native with domestic herbivores, or changes in the size or composition of livestock) could interactively impact the ability of rangelands to provide food, livelihoods and other ecosystem services (Dangal et al., 2016; Eldridge et al., 2016). Asian montane rangelands are a part of the largest contiguous grassland system in the world, and despite being historically important for pastoralism are still relatively understudied. This rangeland system stretches over central Asia (Kyrgyzstan, Turkmenistan, Uzbekistan, Tajikistan and Kazakhstan), trans-Himalaya (Pakistan, India, Nepal and Bhutan), Tibetan Plateau (China) and steppe and Gobi regions in Mongolia, occupying between 40% and 65% of land area over this region (Angerer et al., 2008; Gintzburger et al., 2005). The bulk of this region's livestock-based economies and cultures, and a vast majority of this region's mostly rural population, are dependent on these rangelands for nomadic and agro-pastoralism. Furthermore, these rangelands are home to several endangered species of wildlife (e.g. Saiga Saiga tatarica, Tibetan antelope Pantholops hodgsonii and snow leopard Panthera uncial; Berger et al., 2013), and comprise a globally important carbon reservoir (e.g. Genxu et al., 2002). Understandably then, there is a concern about the degradation of these rangelands in recent times. This degradation has been linked with rapid changes in livestock grazing regimes arising from ongoing socio-economic changes (Angerer et al., 2008; Berger et al., 2013; Harris, 2010; Miller, 1990). For example, livestock numbers have increased nearly 9-fold in over a 50-year period in Inner Mongolia (Angerer et al., 2008) and similar rapid increases have been reported from the Gobi region in Mongolia (Berger et al., 2013), Indian trans-Himalayas (Namgail et al., 2007), Qinghai-Tibetan Plateau (Lu et al., 2017) and central Asia (Gintzburger et al., 2005). This intensification of livestock holding and a shift towards smaller bodied livestock has been accompanied by a decline in native herbivores and a large decline in rangeland productivity across the Asian steppes (Angerer et al., 2008; Berger et al., 2013). In addition to changes in patterns of grazing, these rangelands are experiencing some of the most rapid climatic changes globally (Christensen et al., 2007). For example, average winter temperature over Qinghai-Tibet Plateau (QTP) has increased by more than 1.5°C in over two decades (Du & Ma, 2004), and projections suggest a further increase of 2–5°C in the coming decades (Christensen et al., 2007). Similarly, both summer and winter precipitation patterns are changing, although changes in precipitation are harder to project and vary greatly over this region, especially due to local orographic features in mountainous regions (Christensen et al., 2007; Xu et al., 2008). Together with changes in livestock production systems, these climatic changes are likely to influence rangeland vegetation and various aspects of ecosystem functioning, including carbon cycling and storage, hydrological cycles and forage production. Furthermore, the effects of grazing are likely to be influenced by changes in climate and vice versa (Klein et al., 2004), making it necessary to study these factors together to better predict future rangeland functioning. Thus far, most experimental research in these Asian montane rangelands has focused on the effects of warming and has been predominantly conducted on the Tibetan plateau (Ganjurjav et al., 2015; Klein et al., 2004; Lu et al., 2017). However, the effects of changing climatic regimes and grazing systems on rangeland vegetation in other regions of Asia have been less well studied experimentally. Furthermore, most experimental studies manipulate a single climate variable (by imposing either warming or drought) at a single site, making comparisons across different drivers and across sites with different environmental conditions difficult. To address this gap, we experimentally manipulated growingseason temperature and precipitation, and simulated changes in grazing over 3 years at two sites (Spiti valley, India and Tost, Mongolia) in semi-arid Asian montane rangelands that have a long history of livestock grazing. We examined changes in the cover and composition of the vegetation community in response to these treatments. Previous research from the region suggests that site-level precipitation regimes and micro-site differences in soil moisture influence the direction of vegetation response to warming (Ganjurjav et al., 2016; Klein et al., 2004; Liancourt et al., 2012). As our study sites are dry (see Section 2: average rainfall in the growing season is <200 mm), and vegetation is typically moisture limited (Bagchi & Ritchie, 2011; Liancourt et al., 2013), we expected warming to lower soil moisture and thus lower overall vegetation cover. Furthermore, we expected that vegetation cover would decline with reduced rainfall and increase Furthermore, grazing by herbivores will play an important role in mediating rangeland responses to climate change; thus, pasture management in concert with local pastoralists will be crucial in mitigating the adverse effects of climate change on rangelands, pastoral livelihoods and wildlife populations.