'Effective growing days': A simple predictive model of the response of herbaceous plant growth in a Mediterranean ecosystem to variation in rainfall and phosphorus availability

1. Herbaceous plant production on many Mediterranean soils is severely restricted by nutrient deficiency. On such a shallow, phosphorus-deficient soil in the western Galilee of Israel with an average annual rainfall > 800 mm, a single addition of phosphorus fertilizer consistently increased herbaceous plant production in a community dominated by the dwarf shrub Sarcopoterium spinosum by a factor of 2.8-3.8 over 7 years. Regardless of phosphorus addition, interannual variation of the herbaceous vegetation remained high (C.V. = 0.33). A simple model based on climate data was developed to explain the continuing interannual variation in biomass production. 2. This interannual variation is accounted for by an 'effective growing day' index, G, based on an estimate of the number of clays throughout the growing season during which the available water in the rooting zone can balance the current potential evaporation demand. The validity of this approach was tested on data from 7 consecutive years. 3. At given phosphate (P) availability levels, the index G was found to be highly correlated with both seasonal biomass production and P uptake of the herbaceous plant, but the regression lines intersected the G axis far to the right of the origin. 4. When the adverse effect of the winter climate on growing conditions was taken into account, the regression lines passed close to the origin. Annual biomass production and P uptake could then be well described as linear functions of G with the slope related to availability of P. 5. Average intraseasonal trends gave G-values which were better predictors of both biomass production and P-uptake than annual weather (temperature or potential evaporation) fluctuations. 6. Both shoot biomass and P uptake were strongly influenced by growing conditions in autumn (beginning of growing season) and in spring (main growing season), but plant production was more sensitive than P uptake to growing conditions in autumn and vice versa in spring. 7. The close relationship between G and P uptake at different P availability levels indicates that availability of P did not diminish for at least 7 years after a single application of 4.5-9.0 g m^-2 P as superphosphate. Nutrient cycling, secondary effects of the dominant legume species, and small losses of P from the ecosystem are proposed as possible reasons for this phenomenon.