Scripts for creating maps and performing niche modeling in RStudio.
The general workflow is:
• mapping.R: importing Callisia graminea occurrence data into R, creating simple maps, creating customized shapefiles
• layerPrep.R: masking/clipping PRISM layers, looking for correlations between layers
• maxent.R: creating niche models using occurrence data and climate layers
• nicheOverlap.R: assessing whether niche models for different cytotypes are distinct from each other
• creatingPermutedCallisiaDatasets.R: assessing the accuracy of identification of climatic determinants of a cytotype's range
• ResponseCurves.R: creating figures that each have multiple response curve graphs representing each environmental layers included in each model constructed
• TestGainTable.R: creating table of of test gain values for environmental variables used in the model to determine which one is most important to the predictability of the model
There are a number of files and directories that are downloaded or created during this analysis. To perform all commands in layerPrep.R, you should download the PRISM 1929 layers from Recent Years and the PRISM 2011 layers from Historical Past. Store them in a convenient place (note the paths for using these layers may need to be changed). Clipped layers are included in layers for your convenience. A shapefile created in mapping.R is included in shapefiles for convenience. Models are included for both tetraploids and diploids for Callisia graminea in models for observation. Logistic Regression results are included in logistic_regression_results for observation. The ResponseCurve.R creates the publication–ready figures for each taxon in figures. Manually move the MaxentResults.csv file from a temporary files generated by Maxent to the specific model directory within models. Run TestGainTable.R to generate a test gain table that includes taxon,test AUC, and test gain values for the full model, and only for the specific environmental variables included in the model built.