Application of the British national vegetation classification to the communities of the park grass experiment through time

The Park Grass Experiment at Rothamsted consists of a series of fertilizer treatments which have been applied to a species-rich hay meadow annually since 1856. Detailed botanical analyses of the plots between 1862 and 1992 were classified to one of the National Vegetation Classification (NVC) communities using the similarity measured by Czekanowski coefficient using computer program MATCH. Results indicated that “control” unfertilized plots were a relatively stableCynosurus cristatus-Centaurea nigra dicotyledon-rich grassland. However mineral fertilized plots moved towards a species-richArrhenatherum elatius grassland MG1e after 50–80 years and remained there. Plots receiving nitrogen fertilizer moved first to MG1e then on to aLolium perenne-Alopecurus pratensis grassland MG7D. Perhaps the most interesting plot was one which received a low rate of the soil acidifying ammonium sulphate fertilizer; the species composition changed towards a MG7D (Lolium perenne-Alopecurus pratensis) community between 1856 and 1903 but since then a liming treatment has been applied to part of the plot and the species composition on that part has recovered to a MG5 community resembling the control plots, taking 70–90 years to do so.