The worldsheet fermions have to do with internal degrees of freedom, namely the spin -- therefore better name for the superstring is the more old-fashioned "spinning string" (since worldsheet SUSY should not be confused with spacetime SUSY). The worldsheet fermions generate multiplets of some internal symmetry group. If you want those internal degrees of freedom generated by WS fermions to transform under spacetime Lorentz Transformations, rather than an independent internal symmetry, you need to correlate the Lorentz transformations of the worldsheet bosons and fermions. This is what worldsheet SUSY does for you.
All of this is not specific to string theory. If you want to first-quantize a field theory, a "bosonic" worldline theory will give you a (free) scalar field theory. Adding fermions and the corresponding worldline supersymmetries will generate (free) higher spin fields. It is probably a useful exercise to get e.g. classical (free) Maxwell field from a (N=2 SUSY) worldline theory in order to appreciate precisely what the worldsheet structures mean precisely. Wish I had a good reference, but maybe someone can help me out.
This post has been migrated from (A51.SE)