‘Grab and stab’ methods have become standard in the measurement of insect body temperatures. The gradient of the best-fit regression of body temperature on ambient temperature is often used as a measure of the thermoregulatory ability of a species. The temperatures recorded are commonly accepted as slight underestimates of actual values prior to capture due to passive cooling between capture and insertion of the thermocouple. Here we present laboratory experiments involving tethered flight which show that bees often warm up on cessation of flight, and that errors due to warm-up over the time interval typically associated with ‘grab and stab’ sampling may be significant. More importantly, the errors due to warm-up in two species are shown to change with ambient temperature, thus affecting the form of the relationship between ambient and body temperatures. We compare laboratory and field data to illustrate the way in which warm-up errors may exaggerate apparent thermoregulatory ability, and we urge greater caution in the interpretation of ‘grab and stab’ data.

This content is only available via PDF.