Northern Mongolia is expected to experience some of the greatest temperature increases due to global climate change. We use passive warming chambers to elevate temperatures and explore its ecological impacts in heavily grazed steppe vegetation near the Russian border. Experimental replicates are set up in two south-facing locations within a single river valley, across a relatively flat area near the river and at a higher (+100m) elevation with a much steeper slope. We used both open-top chambers that warm during the day and open-sided chambers that reduce rainfall throughput and warm at night. Warming treatments are crossed with a grazing treatment, supplemental watering, and vegetation removal. The latter is relevant to understanding how plant-plant interactions are altered by warming and its abiotic consequences. We are measuring a number of ecologically relevant variables over the four-year experiment, e.g. soil and air temperature, soil respiration, total soil carbon and nitrogen, available nitrogen, soil microbial communities, litter decomposition, plant community composition and productivity, and weekly flower production. Here we focus on flowering and soil moisture in 2009, our first full field season.

Results/Conclusions

Warming impacted flowering and the effect of vegetation on soil moisture differentially at the two elevations. At the higher elevation, warming increased total flower production of all species combined in the second half of the season, but there was no such effect of warming at the lower elevation. When species were examined individually, several increased flower production with warming at both elevations. Warming did not change the timing of flowering for any species, contrary to other published results. Warming treatments consistently decreased soil moisture at both elevations. In comparison to non-vegetated areas, the presence of vegetation further decreased soil moisture during rain-free periods at the lower elevation but not at the higher elevation. Most other physical and biological variables show large differences between the two elevations but little or no response yet to experimental warming or to the different chamber designs. Because different results were obtained at the two elevations, we conclude that the use of small-scale warming experiments to make predictions about ecosystem responses to climate change need to take landscape level variation into account.