Model estimates (±SE) of the probability of a particular bug being attacked first, in Experiment C where all spiders had prior exposure to milkweed bugs.

Values are relative to the intercept which represents brown milkweed bugs. Z and P values are indicated for each fixed effect. The lack of a significant effect of palatability treatment suggests that spiders were not biasing their attacks away from milkweed bugs, as would be expected if the presence of unpalatable milkweed toxins had an aversive odor that deterred spider attack. Note that there is a non-significant trend in the opposite direction, where spiders are more, rather than less, likely to attack milkweed-fed bugs.