Energy expenditure in Awassi sheep grazing wheat stubble in the northern Negev Desert of Israel

The heart rate method for estimating metabolic rate: review and recommendations

Under most circumstances heart rate (f(H)) is correlated with the rate of oxygen consumption (VO(2)) and hence the rate of energy expenditure or metabolic rate (MR). For over 60 years this simple principle has underpinned the use of heart rate to estimate metabolic rate in a range of animal species and to answer questions about their physiology, behaviour and ecology. The heart rate method can be applied both quantitatively and qualitatively. The quantitative approach is a two-stage process where firstly f(H) and MR are measured simultaneously under controlled conditions and a predictive calibration relationship derived. Secondly, measurements of heart rate are made and converted to estimates of MR using the calibration relationship. The qualitative approach jumps directly to the second stage, comparing estimates of f(H) under different circumstances and drawing conclusions about MR under the assumption that a relationship exists. This review describes the range of studies which have adopted either the quantitative or qualitative approach to estimating the MR of birds, mammals and reptiles. Studies have tended to focus on species, states and questions which are hard to measure, control or define using other techniques. For example, species studied include large, wide-ranging species such as ungulates, marine predators, and domestic livestock while research questions have concerned behaviours such as flight, diving and the effects of stress. In particular, the qualitative approach has applied to circumstances and/or species where it may be hard or impossible to derive a calibration relationship for practical reasons. The calibration process itself can be complex and a number of factors such as body mass, activity state and stress levels can affect the relationship between f(H) and VO(2). I recommend that a quantitative approach be adopted wherever possible but that this may entail deriving a calibration relationship which is practical and applicable, rather than the most accurate possible. I conclude with a series of recommendations for the application and development of this method.