Hello,
I´d like to know, how can I interpret the results of variation partitioning when I run it separately for two different datasets. For example, I have bacterial and fungal communities and I test the effect of variables X, Y and Z on their composition. In the case of bacteria, I get following proportions: X = 20%, Y = 5%, Z = 0.05%, in the case of fungi: X = 10%, Y = 5%, Z = 13%. Can I say that variable X has more influence on bacteria than on fungi or it´s not possible to compare two independent variation partiioning analyses?
Thanks a lot!
I believe so, that is the purpose of this method of analysis. Note that the difference between 10% and 20% may be within some range of error from your samples or within the analysis. Additionally, a 10% difference is also relatively small. I think the Z value might be more telling because it seems to be 260x more correlated with fungi than with bacteria.
It might be worthwhile to try Principal components analysis (PCA) as well since that is easier to interpret and is more robust.
- Badges
- Science topic
- Similar topics
- Biological Science
- Kinesiology
- Running