- Joined
- Oct 15, 2019
- Messages
- 385
- Reaction score
- 1,612
Sampling is important to think about. The idea of replicate samples from the same tank is not unreasonable at all. For some kinds of measurements, it's critical. Of course it increases costs, so the question becomes, does it provide enough benefit for the cost, in this test?My gut feeling is that (again unless he has data to the contrary) - and data as to how reproducible the tests are depending on the factors I already mentioned - I don't know what to make of a report based on a single test from a single water area.
The importance of replication depends on the variation between replicates.
In this post, I showed how similar two samples from the same tank are (A1 and A2). Really, really similar. And this makes sense - the water is well mixed, and we pull a large volume from it. (It would have been easier and cheaper to sample 5 ml instead of 60 ml..) So we expect that two replicate samples will be similar, and observe this. For the same reasons, a single sample is typically used for ICP analysis, right?
Don't get me wrong. Replication is always valuable. Given unlimited resources, measure everything half a dozen times. But in a world of limited resources we focus replication on the levels that will vary most.
Since multiple samples drawn at the same time are very similar, I think it makes the most sense to sample repeatedly over time or before and after a change, rather than spending these resources on getting a really really precise estimate of the community at a single point in time.