Our research campaing involved sampling of 4 rivers (at three different altitude stations each river). At every station we collected three different samples in a longitudinal 100 m-transect of the river taking special care to sample the full heterogeneity of substrata, and analyzed for benthic macroinvertebrates.
At the same time we evaluated numerous catchment variables in order to test the relevance of the land use and catchment properties on the macroinvertebrate community.
Therefore, we ended up with 4 rivers * 3 stations * 3 samples per station = 36 samples.
My question is Whether all samples could be wisely included as cases in a Random Forest Model (n=36)?…, or should I instead average macroinvertebrate samples per station to avoid pseudoreplication (n=12)?
I would greatly appreciate any help and advice on this issue.
Salud, y gracias
Manuel
Your actual replicates are 'river' i.e. they are your 'sites' or 'areas'. But they must be significantly spatially separated to be true replicates. Then 'altitude' is a 'subsample' of the replicate. So when you construct your ANOVA table, you will have the factors 'river' (fixed), 'altitude' (fixed, hopefully), and then your catchment variables. Your ANOVA table will automatically work out degrees of freedom (i.e. n–1) so don't worry about calculating n in advance. Do NOT average your macro-invertebrate samples – put ALL the data into one big ANOVA/MANOVA/GLM table (each sample should correspond to a station in your table). If you average your data before entering it in your table, the software has nothing to analyse lol. Pseudo-replication is only a problem if you sampled one river, but you have 4 of them lol :)
- Badges
- Science method
- Similar topics
- Mathematical Sciences
- Statistics
- Sampling Design