Repeatability of mammalian physiology: Evaporative water loss and oxygen consumption of Dipodomys merriami

Phenotypic measurements (e.g., body mass, metabolic rate) made at one point in time are often assumed to represent measurements made at other times. However, if traits are not repeatable over time, inferences about ecological, environmental, or other correlates of the trait could be misleading. We found that evaporative water loss, oxygen consumption, resting metabolic rate, and their residuals from body mass were significantly repeatable (intraclass correlations generally >0.6) in Merriam's kangaroo rats (Dipodomys merriami). Nevertheless, our data, plus a review of the scant data on repeatability of mammalian physiological traits, show that repeatability typically is less than perfect (i.e., less than one). Consequently, investigators studying physiological performance of mammals should not assume that the traits they study are good indicators of performance at other points in time.