Global climate change in concert with other anthropogenic drivers is affecting biodiversity from regional to local scales by causing shifts in species’ distributional ranges and abundances. Continuing climate change will shape local plant community compositions, but the magnitude will depend on how individual species will cope with novel conditions. Both temperature and precipitation are expected to change globally, which will likely affect local plant communities differentially. We set up a four-year field experiment in semiarid, northern Mongolian steppe, which still supports pastoral nomads as it has for millennia. This region is predicted to experience above average increase in temperature, but it is not certain how the precipitation regime will change. We experimentally created three probable climate change scenarios across the two different slope locations, differing in ambient conditions, species composition and aspect. Using open-top passive warming chambers (OTC), we used canonical analysis of principal coordinates (CAP) to determine how plant community compositions were affected by elevated temperatures and the interactions with increased watering and reduced grazing.
Under all climates change scenarios, warming via OTCs affected species relative abundances but the magnitude varied across the two different slope locations. The warming effect was greatest at the more productive lower slope compared to the drier upper slope and was especially pronounced in the last year of the experiment. This implies that warming effects were site specific and initial conditions of the community or site can determine the magnitude of the species compositional changes. Specifically, at the drier upper slope we found that the warming effect was increased by watering addition as plots with water addition and increased temperature differed more from controls than did plots with increased temperature only. We found that the change on the upper slope was mostly explained by the response of generalist species, indicating that warming treatment might have created above optimal temperatures for generalist as common species are operating closer to their thermal optimum. The opposite was true for the lower slope where both generalist and specialist species responded differently to the grazing treatment rather than warming alone. The effect of grazing was less pronounced under higher temperatures, and the warming effect alone overshadowed the effect of grazing. The plant community response to warming depended both on precipitation and grazing regime, showing the importance of accounting climatic change in the context of land use.