The validity of capture-recapture estimates, like any other statistical procedures, critically depends upon the model assumptions. If the assumptions were not followed, the estimates are questionable.