A simulation model for the prediction of the productivity of Lolium perenne
L. grasslands is described and validated. Simulated key processes are ligh
t utilization, leaf formation, leaf elongation, tillering, and carbon parti
tioning (storage, shoot, root). Source- and sink-limited growth are simulat
ed independently. Sink-limited growth is characterized by temperature-depen
dent leaf expansion and tiller development, whereas source-limited growth i
s determined by photosynthetic light-use-efficiency of the canopy and the r
emobilization of stored carbohydrates in the stubble. At each integration s
tep, commonly 1 day, the available amount of carbon from the source is comp
ared with the carbon required by the sink. The actual growth is determined
by the minimum value of either the sink or the source. If the source is in
excess of the sink, the surplus is allocated to storage carbohydrates in th
e stubble. This storage carbon is available for remobilization at times tha
t the sink requires more carbohydrates than are available from photosynthes
is. In contrast to previous grassland models, LINGRA describes regrowth aft
er defoliation in a mechanistic way, balanced by temperature-driven remobil
ization of stored carbohydrates. In order to validate LINGRA, an extensive
set of experimental data was used, derived from measurements at 35 sites in
Europe. The average error between the observed and predicted yields was 14
% at the level of irrigated, and 19% at the level of non-irrigated, treatme
nts for the whole of Europe. (C) 1998 Elsevier Science B.V. All rights rese
rved.