Hi @kims -
Actually, none of the results from the nutrient model are related to retention.
x_load_tot (in the watershed shapefile) is the amount of nutrient produced from each pixel, summed across the whole watershed. it does not take into account the rest of the landscape.
x_exp_tot (in the watershed shapefile) is the amount of x_load that actually reaches a stream, summed across the whole watershed. It does take into account the ability of the vegetation downslope to retain some of the nutrient.
x_export.tif is the amount of nutrient produced on a particular pixel (the nutrient load from that pixel) that actually reaches a stream. This result is used to produce x_exp_tot in the watershed shapefile, by summing the pixel values for export.