Resilience of acid subalpine grassland to short-term liming and fertilisation

A fertilisation experiment was started in the French Alps on an acid grassland at 2000 m in 1989 where lime as calcium carbonate (“liming”) and Thomas Slag enriched by potassium chloride (“fertilisation”) were added in a random block design until 1992. Since then, no further amendments were applied.Fifteen years after the last application, we revisited the experiment and observed that soil pH was still significantly higher on limed plots, while nitrogen (N) concentrations were lower. On fertilised plots, soil carbon (C) and N concentrations were lower compared to unfertilised plots. However, litter quality (C and N concentrations, near infrared spectroscopy [NIRS] data) was similar for both treatments. Vegetation composition, but not species richness, nor Shannon–Wiener or evenness differed between limed and unlimed plots, and fertilised and unfertilised plots. Liming explained about 18% and fertilisation about 6% of the variability of the vegetation composition. These changes in the vegetation composition are probably due to lower abundances of former dominant grass species and to an increase in generalist grasses. However, these changes did not influence the total above-ground productivity, which was similar for all treatments. Tissue N and C concentrations and NIRS data indicated a changed chemical composition of the biomass which persisted during time.We conclude that the 3 years of fertilisation and liming did substantially influence the vegetation composition at our site and lead to an increase in the agricultural value of the grassland. These changes are long-lasting as they changed key features of the functioning in the soil of grasslands ecosystems. From an ecological point of view, specialised vegetation was replaced by generalist species leading to a trivialisation of the vegetation.