A Critical Examination of Winkler's Method for Determining Dissolved Oxygen in Respiration Studies with Aquatic Animals