I did a story in Crops magazine a couple of years ago comparing how soil mineral nitrogen results from the same sample varied between different labs - and there were certainly some interesting results!
Cores were taken (by an agronomist) at 0-30cm and 30-60cm from 2 fields - one heavy clay, another sandy loam. Samples were sent to 4 labs (NRM, Direct Labs, Lancrop Labs and Hill Court Research). When the results came back, on the heavier field (wheat after set-aside), the SMN analysis for the total of the 2 cores (0-60cm) varied from 63-132kg N/ha. For the light land, results ranged from 42-78kg N/ha. Which was closest to the truth is anyone's guess...
This opened a whole can of worms around how samples were taken in field, stored, transported, sub-sampled at the lab and analysed. You may take a bag full of soil from the field, but if the lab only selects a 40g sub-sample for analysis, there's a pretty good chance it might not be representative of the rest!
The main conclusion was to use sampling as a guide only and take unexpectedly high or low results with a very large pinch of salt!
Paul (FW arable)