Sources of error in advisory soil tests

Abstract
The different sources of variation in soil tests estimated in the previous paper in this series are discussed, and methods of increasing precision considered. It is shown that simply increasing the number of cores taken effects only a slight improvement in precision.