Context dependency and generality of fever in insects.

Fever can reduce mortality in infected animals. Yet, despite its fitness-enhancing qualities, fever often varies among animals. We used several approaches to examine this variation in insects. Texas field crickets (Gryllus texensis) exhibited a modest fever (1 °C increase in preferred body temperature, T pref) after injection of prostaglandin, which putatively mediates fever in both vertebrates and invertebrates, but they did not exhibit fever during chronic exposure to heat-killed bacteria. Further, chronic food limitation and mating status did not affect T pref or the expression of behavioural fever, suggesting limited context dependency of fever in G. texensis. Our meta-analysis of behavioural fever studies indicated that behavioural fever occurs in many insects, but it is not ubiquitous. Thus, both empirical and meta-analytical results suggest that the fever response in insects 'is widespread, although certainly not inevitable' (Moore 2002). We highlight the need for future work focusing on standardizing an experimental protocol to measure behavioural fever, understanding the specific mechanism(s) underlying fever in insects, and examining whether ecological or physiological costs often outweigh the benefits of fever and can explain the sporadic nature of fever in insects.