Dear Sofia,
No, you do not need all the columns listed in the bookdown.
You can directly create yourself a dataset with the same structure as the aggregated dataset, thereby bypassing the need to use the function `prepsources()`.
So one line per location and source_ID (a name for the location, can also be the lat and long pasted together or anything else), mean_source_value (the mean dC13 at the location), var_source_value (the variance across dC13 measurements at the location), n_source_value (the sample size at the location), lat (the latitude), long (the longitude).
That assumes that you have some replicates per sampling location.
If that is not the case, please let me know.
Best,
Alex