I want to see if the environmental parameters as explanatory variables and species density as response variable can be modelled into generalised linear mixed model. But the env parameters are not fixed treatments and they are different everytime i measure in the quadrates. Is using just the bioenv procedure is enough?