On the dynamics of behavior allocation between simultaneously and successively available reinforcer sources