CLIMATIC VARIABILITY AND GRASSLAND COMMUNITY COMPOSITION OVER 10 YEARS - SEPARATING EFFECTS ON MODULE BIOMASS AND NUMBER OF MODULES

1. Data from permanent grassland plots, collected yearly over a 10 year period were analysed for correlation between weather variables and plant performance. Number of modules and mean biomass of the module were recorded separately. Because the two variables are inter-related, both with each other and with the values in preceding years, regression residuals were used to quantify the response in the particular year. The performance variables were correlated with the monthly averages of temperature and bimonthly averages of precipitation. 2. There was a large year-to-year component of variation in species performance. For biomass, the year-to-year variation is largely explained by correlations with weather; in number of modules, some species show variation which is not explained by the weather variables studied. Weather variables of both current and preceding years showed significant correlations with species performance. 3. Species responded characteristically to weather variables and in any one species, the number of modules and the biomass do not necessarily respond in the same way. This may be an important mechanism creating time lags in the community response to weather variables and, more generally, may be responsible for non-linearity of the community response to the weather.