ethanwhite
commented
6 years ago One possibility is to use something like a Mantel test to determine if the environment is generally getting more different from the last year of data.
We should just stick to weighting all environmental variables equally, normalize them all, and use a Euclidean Distance (or some other distance if there's a clearly superior one) to quantify the change.
If the environment is relatively stable we might not expect richness to change much, but if it's shifting a lot then we might expect meaningful shifts. This is a related problem to #253, but is a multivariate change instead of a univariate one. Since we are interested in the environment as it drives richness we should also consider whether we want to weight all environmental factors equally or give more weight to those shown to correlate with richness.