On the stability of a delay differential population model