EXPERIMENTAL DESIGN IN STUDIES OF SNAKE BEHAVIOR

Snakes have unique qualifications as experimental subjects, and the use of these animals in behavioral research has gradually increased. They have been models for comparative studies of innate prey preferences and for studies on sensory signals involved in reproductive behavior. Examples from these studies indicate the importance of keeping in mind both statistical methods and the natural behavior of the particular species when designing experiments on snakes. Randomized designs are the most powerful but require larger sample sizes, and so repeated or blocked designs have more often been used in snake studies. How replication affects the inferences drawn is rarely discussed. Nor have most authors evaluated the role of factors that may increase variation in the behavior of snakes, i.e., temperature, stress, feeding condition, shedding cycle, and general health and temperament of the individual subject. Future research on snakes will supply increased understanding of the behavioral ecology of this interesting group, but more care in the design of experiments will be important.