Looking at three subsets of the dataset: Mediterranean and Tropical and Subtropical Dry Broadleaf Forest and Temperate Grassland biomes. The overall objectve is to assess whether total population (persons per km^2) and land cover type (crops, pasture, and urban) affect native species (N).
Assess whether data between each land cover type and biome are autocorrelated using variograms and calculating Moran’s I.
Compare the ordinary least squares (OLS) and generalalized least squares (GLM) regression models for spatially autocorrelated data. I’m assuming this combination of variables will be autocorrelated based on an analysis of the Tundra subset.
Evaluate the (spatial autoregressive) SAR model for each biome and cover