Long-term experiments in the South Island high country: an example from Mt. Possession, Canterbury

Abstract

This paper discusses the successional changes that have occurred during 13 years of a designed experiment in short tussock grassland. The 'experiment site is at Mt. Possession Station, Canterbury and was established in 1979. Nine species of legume were overdrilled and 9 phosphate rates (O-800 kg P/ha) broadcast at sowing. A vegetation survey of all plots was carried out in November 199 1 and soil sampled in the following year. Legumes were dominant during the early years of the experiment but the responses to P rates differed. Some of the differences in the present vegetation could be explained by earlier treatment effects. Exotic grasses invaded the plots which received higher levels of fertiliser with the exception of plots sown with Lupinus species where grass growth was independent of the applied P rate. The initial treatments have strongly affected present vegetation. Had an attempt been made at the beginning to predict the current composition, or even half way through the experiment, the conclusions are unlikely to have been realistic. It is suggested that the dynamics and composition of the responses are inter-related and complex and as such, they require long-term monitoring before any predictions can be made. Kewords: long-term experiment, phosphate gradient, sown legumes, species composition, tussock grasslands