The issues are similar, whatever problem is being investigated. Biological systems are variable in space and time at many scales, due to many interacting
processes. It is the responsibility of biologists and ecologists to understand the consequences of such variation and the ensuing interactions and non-independent
patterns so that sampling can be planned to take all of the issues into account.
Where ‘ideal’ sampling is not possible, what is uncertain in the data can and must
be considered before any conclusions are reached. This will ensure that unplanned
difficulties or accidental loss of samples do not prevent interpretations being of
value. If major problems are inevitable with some design, alter the design or investigate more preliminary hypotheses in order to unravel the problems.