I've run a standard RDA analysis for my data. I am struggling to find a formal analysis for some important insights (i.e. I need a way to report these statements in my paper)
Consider the red x lying in between Shrub and WaterCont at ~(-0.5,-0.5). I want to estimate if Shru or WaterCont explains more of the variation. How can I calculate:
The length of each loading vector (WaterCont and Shrub)
The distance or red, x from each loading to find out which contributed to more variation?
# Load mite species abundance data data("mite") # Load environmental data data("mite.env") # Hellinger transform the community data mite.spe.hel <- decostand(mite, method = "hellinger") # Standardize quantitative environmental data mite.env$SubsDens <- decostand(mite.env$SubsDens, method = "standardize") mite.env$WatrCont <- decostand(mite.env$WatrCont, method = "standardize") mite.spe.rda.signif <- rda(mite.spe.hel ~ WatrCont + Shrub + Substrate + Topo + SubsDens, data = mite.env) # Find the adjusted R2 of the model with the retained env # variables RsquareAdj(mite.spe.rda.signif)$adj.r.squared anova.cca(mite.spe.rda.signif, step = 1000, by = "term") # Scaling 2 ordiplot(mite.spe.rda.signif, scaling = 2, main = "Mite RDA - Scaling 2") # Load mite species abundance data data("mite") # Load environmental data data("mite.env") # Hellinger transform the community data mite.spe.hel <- decostand(mite, method = "hellinger") # Standardize quantitative environmental data mite.env$SubsDens <- decostand(mite.env$SubsDens, method = "standardize") mite.env$WatrCont <- decostand(mite.env$WatrCont, method = "standardize") mite.spe.rda.signif <- rda(mite.spe.hel ~ WatrCont + Shrub + Substrate + Topo + SubsDens, data = mite.env) # Find the adjusted R2 of the model with the retained env # variables RsquareAdj(mite.spe.rda.signif)$adj.r.squared anova.cca(mite.spe.rda.signif, step = 1000, by = "term") # Scaling 2 ordiplot(mite.spe.rda.signif, scaling = 2, main = "Mite RDA - Scaling 2")