Evaluation of two methods for minimally invasive peripheral body temperature measurement in birds

Body temperature (Tb) is a valuable parameter when assessing the physiological state of animals, but its widespread measurement is often constrained by methods that are invasive or require frequent recapture of animals. Alternatives based on automated remote sensing of peripheral Tb show promise, but little is known about their strengths and limitations. We measured peripheral Tb in great tits Parus major with subcutaneously implanted passive integrated transponders (PIT tags) and externally attached radio transmitters to determine repeatability of measurements, sensitivity of each method to variation in ambient temperature (Ta) and wind speed, the relationship between methods, and their ability to capture circadian variation in Tb. Repeatability of measurements by radio transmitters was high (> 80%) when readings were taken within 20 min, but reduced to 16% when measures were spaced 3.5 h apart. PIT tag data for the 3.5 h interval were more repeatable (33%) and less variable (cv). Data were affected by Ta with a stronger effect on the externally attached transmitters, but the influence of wind speed was small for both methods. There was a significant positive relationship between transmitter- and PIT tag temperature during both days and nights. Both methods were equally suited to detect diel changes in peripheral Tb. However, transmitters offered longer detection distance and better temporal resolution. These qualities should be considered when deciding how to collect Tb data remotely. If properly deployed, both methods allow measurement of peripheral Tb over a wide range of natural systems and conditions in small, free-ranging, birds.