Estimation of daily energy expenditure from heart rate and doubly labeled water in exercising geese

We investigated whether daily O2, consumption (Vo2) could be predicted from heart rate (f(H)) in five exercising barnacle geese (Branta leucopsis) and compared the accuracy of this method with that of the doubly labeled water (DLW) method. The regressions of Vo2 on f(H), based on incremental speed tests, differed among individual birds. The O2 pulse (i. e., Vo2/f(H)) progressively increased with exercise level from 0.22 mL O2 heartbeat-1 during resting to an estimated 0.47 mL O2 heartbeat-1 during flight. Daily Vo2 was generally underestimated (-3.9%) by (individual) resting O2 pulses but overestimated (+8.4%) by linear regressions of Vo2 on f(H). However, it was well predicted (+0.8%) by the O2 pulses appropriate for each exercise level. When using relationships derived from the group of birds, the estimations were generally improved (-3.3% for resting O2 pulse, -0.03% for appropriate O2 pulse) but poorer (+13.6%) for the group linear regression. Some of these predictions were better than the estimation of daily CO2 production (Vco2) by the two-compartment model of the DLW method (average algebraic error of +0.9%). We conclude that f(H) can be used to estimate daily energy expenditure in birds accurately provided that (1) its application is limited to the range of exercise levels in which f(H) has been calibrated against Vo2 and (2a) Vo2-f(H) relationships are determined for each individual bird or (2b) the f(H) measurements of several free-ranging birds are averaged. Heart rate can also be used to indicate within-day variation in energy expenditure.