Empirical seasonal predictive models of production and cover of above ground biomass in relation to climatic parameters in a subalpine grassland of Tymfristos mountain – prefecture of Evrytania

Ecosystems’ productivity is strongly related to climatic conditions. Specifically, the productivity of Mediterranean rangeland ecosystems is depended to the degree of time- and space-irregularity of precipitations’ distribution. The use of climatic parameters in the construction of empirical mathematical models towards the prediction of cover and production of the above ground biomass is a powerful tool for management purposes. The present research deals with the exploration and construction of mathematical seasonal predictive models having above ground biomass and cover as dependent variables and climatic variables as explanatory ones. The research was conducted in 2000 in ungrazed grassland of Rovia location of Evrytania. Vegetation data were collected in a monthly basis from April to September and models were set up using climatic data obtained from the Institute of Mediterranean Forest Ecosystems & Forest Product Technology of Athens. For climatic data time depth was obtained from the start of the raining season of the previous year (September 1999) until the antecedent month of vegetation’s data collection. A total of 288 models (linear and non linear) included (and not included) a constant term in their mathematical expression were examined. It was found that only the sigmoid model with a constant term is statistically fitted to precipitation and dead biomass data. On the contrary, when the constant term is not included it was found that only 10 from 144 models were not statistically fitted to climatic and vegetation data. In the case of constant term’s exclusion the selected seasonal predictive models were the power, the “S”, and the growth ones.