Model estimates (±SE) of the probability of a particular termite being attacked first, in Experiment B with both naïve spiders and those previously exposed to DB.

Values are relative to the intercept which represents naïve spiders that were offered brown control termites. Z and P values are indicated for each fixed effect. The lack of any significant interaction between palatability treatment and prior DB exposure suggest that both naïve and exposed spiders responded similarly to the DB treatment. The lack of significance of the main effect of palatability treatment suggests that there are no aversive odors associated with our DB treatment that deter attacks from spiders.