I have a long data series of species abundance at 3 different sites. I think I can not model species distribution because of the spatial scale of the samples (discrete points, not a grid), but because I have robust environmental data, I'm looking for a model that can explain the temporal variability considering the nature of my data.