I'm having a difficult time interpreting PCA loadings. I am using the vegan package in R to create a PCA of environmental variables.

env.transformed <- decostand(env_clean, "standardize")
pcas <- rda(env.transformed)
loadings <- vegan::scores (pcas, display = 'species', scaling=0,choices = 1:3)

I am primarily confused about the scaling option in the last line. I understand that scaling = 0 gives raw scores, but are these considered the "loadings" that I would report to explain which variables are contributing most to each PCA axis?

When I set scaling = 2, some of the loadings become >1. Is this normal, or has something gone wrong? Which is more appropriate to report?

Thank you!