I think that any explanation should be focused on the independence and the source of variation.
A biological effect can be validated only if it arises from several different biological replicates.
If the samples are not independent, then you dont have different replicates but technical replicates.