Estimation of Muscle Mass by Ultrasonography Differs between Observers and Life States of Models in Small Birds

Ultrasonography has proven to be a valuable noninvasive method of measure of muscle size in birds, but validation of its use in birds as small as black-capped chickadees (Poecile atricapillus; 11 g) is scarce. The effect of observers and life state (dead or alive) of models used for calibration on measurement quality is also poorly documented. Using 31 dead and 22 live chickadees, linear regressions between ultrasound and dissection measurements of pectoral and thigh muscles were fitted and compared between five different observers. R-2 values varied greatly between observers and were generally weaker in live birds, ranging between 0.02 and 0.59, despite high repeatability of measurement. Using equations calculated from dead birds to estimate muscle mass of live birds yielded much higher measurement errors (9%-18%) than when using equations calculated from live birds (5%-8%). Our results suggest that with careful training and using only calibration from live birds, ultrasonography can be a useful but limited tool to estimate muscle size of birds as small as the black-capped chickadee.