A MODEL OF THE GRAZING OF HILL VEGETATION BY SHEEP IN THE UK .1. THE PREDICTION OF VEGETATION BIOMASS

1. A computer model is described which predicts monthly growth, senesc ence, litterfall and standing biomass of ungrazed herbage in seven dwa rf shrub-dominated and five grass-dominated vegetation types commonly found in the hill areas of the UK. 2. Using published data, the annual and monthly dry matter (DM) production of each vegetation type is pre dicted at sea level for the temperature zone in which the relevant dat a were collected. These estimates are then adjusted to take account of the effects of altitude and temperature zone on the production of all vegetation types and the effects of fertilizer rate, likely levels of soil nitrogen, summer rainfall and available soil water capacity on r eseeded grassland. Empirically derived rules predict senescence and li tterfall and, together with predicted DM production, are used to deter mine standing biomass in each month. The model also predicts the mean undisturbed sward surface height (sward height) of the grass vegetatio n types. 3. The sensitivity of the model to variations in inputs is de scribed and the reliability of the relationships used in the model an discussed. Despite the limited availability of data suitable for model development, the model predicts DM production to lie between 1 and 11 % of the measured values for the few remaining sites with appropriate data. 4. The most significant gaps in knowledge are identified and sug gestions are made as to how the model might be further developed and t ested.