Principle Response Curves with R
In my first post I want to reproduce the results of:
Van den Brink, P. J. & Braak, C. J. F. . Principal response curves: Analysis of time-dependent multivariate responses of biological community to stress. Environmental Toxicology and Chemistry 18, 138–148 (1999).
Principle Response Curves (PRC) are commonly used for analyzing ecotoxicological mesocosm experiments. Here I will show how to run such an analysis with R. I won’t dig in the mathematics behind it, so this is up to the reader.
The data comes with the vegan package which we will also use for the PRC:
So rows are samplings and columns are the species (with abbreviated names), a normal species x sites matrix. The colnames code treatment and time, but we must create these two factors as well as a factor coding the ditch:
With this at hand we can calculate and plot the PRC using the prc function:
This comes quite near to Figure 3 in the paper, except some scaling on the y-Axis and all the species scores are displayed.
Man, that was complicated… In Figure 3 they plotted only the species with scores greater than 0.5 or less than -0.5. We can access the species scores with the summary function:
And then select only those species with greater or less scores:
OK, that’s it for now. I think in following posts I will reproduce their tables and also show alternatives to the Principle Response Curves.