Detecting cycles and delayed density dependence: a comment on Hunter & Price (1998)