干旱区生态环境知识资源中心

Impacts of climate change were simulated over five summer seasons in a high arctic polar semi-desert at Ny Alesund, Svalbard, by using polythene tents to increase temperature, and by increasing precipitation and soil nutrient (NPK) availability. The effects of these treatments on vegetation cover were assessed at the start of the 1991, 1993, and 1995 field seasons, and at peak biomass in the same years. Over the first season of the experiment (1991), changes in percentage total living vegetation cover were significantly greater, and changes in dead vegetation cover significantly lower, in the tented treatments. In subsequent seasons, changes in total living cover were also greater under treatments simulating climate change, although the significant factors and interactions were year-specific. Between years, at both the early and mid-season sampling periods, the fertilizer application had the strongest effect on changes in plant cover, significantly decreasing cover of living Dryas octopetala, Saxifraga oppositifolia, and bare ground between 1991 and 1995, while increasing cover of bryophytes, Salix polaris, Polygonum viviparum, and total dead vegetation. Although cover of D. octopetala was greater during the first three years of fertilizer addition, marked winter injury occurred in this species on fertilized plots during winter 1993-1994. This resulted in reductions in total live cover and D. octopetala cover and an increase in total dead cover (by up to 22%) in watered and fertilized plots between 1991 and 1995. Seedlings of nitrophilous "immigrant" species were established naturally on bare ground in fertilized plots in the third year of the study and subsequently increased in number, so that after five seasons the community tended more toward bird-cliff vegetation rather than polar semi-desert vegetation. The tent treatment and the simulated increase in summer precipitation had little effect between seasons on the plant community, in comparison with the fertilizer treatment.