Alpine tundra ecosystems such as those which are dominated by ericaceous dwarf shrubs in northern Scandinavia are characterised by low productivity, and this is due in part to low availability of nutrients and retardation of those microbial processes required for nutrient cycling. We conducted an eight-year field experiment in an alpine tundra in northern Sweden, in which eleven treatments aimed at alleviating possible stresses were applied to field plots; these included addition of various forms of nitrogen and other nutrients, addition of lime, addition of available carbon, and reduction of possible adverse effects of secondary metabolites produced by the dwarf shrub Empetrum hermaphroditum. Nearly all of the treatments had statistically significant effects on at least some of the plant species present in the experiment. Addition of nitrogen and liming both had important effects in reducing E. hermaphroditum cover and in these treatments Deschampsia flexuosa rapidly became dominant. Manipulations that reduced E. hermaphroditum or its effects frequently also stimulated Vaccinium myrtillus and V. vitis-idaea. Fertilisation and liming treatments also often caused decreases in the mosses Dicranum sp. and Pleurozium schreberi, and the lichen Cladina spp. Ordination analysis revealed that vascular plant community structure was most significantly altered by treatments involving mineral nitrogen addition and liming, moss community structure by treatments involving addition of available carbon and lichen community structure by treatments involving addition of some forms of nitrogen, lime and reduction of effects of E. hermaphroditum. Nearly all treatments significantly reduced total plant diversity (species richness) and several treatments reduced diversity of each of the vascular plant, lichen and moss groups. This reduction in diversity was frequently associated with monopolisation of plots by D. flexuosa. Decomposition rates of litter added to the plots were generally only weakly influenced by treatments, but several treatments (most notably those involving nitrogen addition) induced litter nitrogen immobilisation and increased litter microbial biomass levels. Humus nitrogen and microbial levels were also sometimes enhanced by the nitrogen addition treatments. This suggests that those plots dominated by D. flexuosa were likely to result in greater conservation of nitrogen in the litter relative to those dominated by E. hermaphroditum. An important exception to this pattern was the effects of the liming treatment which, although responsible for an increase in D. flexuosa, caused net losses of nitrogen and reduced microbial biomass both in the litter and the humus. Our data therefore suggest that treatments which have vastly differing consequences for humus properties and the decomposer subsystem (and ultimately conservation of nitrogen in the soil), but which alleviate an inherent stress, have similar consequences above-ground, e.g. E. hermaphroditum decline, D. flexuosa enhancement, and associated reductions of plant diversity.