Sustainability studies are increasingly needed to determine management
systems that protect the environment and maintain production potentia
ls. Whether these analyses are performed by field experiments or by co
mputer modelling, defining the appropriate spatial and temporal scales
is essential. For three management types within one soil series in Th
e Netherlands, the land quality indicator ''transpiration ratio'', E-r
atio (ratio between actual and potential transpiration), was determine
d using a simulation model of water flow in the unsaturated zone. Soil
s were characterised by eight profiles per management type in which me
asurements of the retention and hydraulic conductivity characteristics
were made. Variation in weather data was described by 30 years of his
torical data. Soil hydraulic characteristics were used as individual d
ata and as averaged data. Differences in E-ratio obtained by using ind
ividual soil data rather than averaged data were not significant, but
E-ratio values for the three management types were significantly diffe
rent (0.79, 0.81 and 0.83). The results of the 30 years were used to a
nalyse the effect of the time over which an experiment is evaluated, o
n the E-ratio. The differential of variance indicated that the length
of an experiment should be at least seven years in order to reduce the
effect of the variable weather conditions. (C) 1997 Elsevier Science
B.V.