The Interpretation and Misinterpretation of Mortality Rate Measures

Ecologists frequently measure and compare mortality rates and other count-dependent rates of change. The simplest measures employ mortality counts for predetermined populations over a defined census interval (e.g. Harper 1977; Putz & Milton 1983; Connell et al. 1984; Hubbell & Foster 1990; Turner 1990; Osunkoya et al. 1992). More complex formulations are required to allow comparison over varying time periods, because these measures require a knowledge or assumption of how probabilities of death change over time. In many ecological applications this probability is taken to be constant and can therefore be used to define a rate. In this paper we examine some potentially misleading discrepancies that occur in the recent ecological literature. In its simplest form a constant mortality is modelled by exponential population decline: