I have a large dataset with coordinates of survey locations, and results for presence or absence of a species. These surveys are spatially clustered, with four in a line separated by 1km each, and the lines separated by about 20km. I made a simple dataset in R to illustrate my problem (see code and plot 1).
I want to know if there is spatial auto-correlation in presence-absence between survey sites; specifically within each line (the sites separated by 1km), and at which point spatial auto-correlation along the line is no longer an issue.
I have used gstat's variogram() to test autocorrelation of residuals of a model, but so far geoR's variog() is the only one I found which will work to test to spatial-autocorrelation in the raw data (no model).
Here is the code I have tried in R
X
I want to know if there is spatial auto-correlation in presence-absence between survey sites; specifically within each line (the sites separated by 1km), and at which point spatial auto-correlation along the line is no longer an issue.
I have used gstat's variogram() to test autocorrelation of residuals of a model, but so far geoR's variog() is the only one I found which will work to test to spatial-autocorrelation in the raw data (no model).
Here is the code I have tried in R
X