We tested the effect on diversity estimates of two alternative sampling conventions.
If plot size is <1 m2, the two systems yield significantly different values.
The amount of the difference varies with grassland type and physiognomy.
Methodological consistency is essential even for simple indicators.