Artifacts arising from sampling interval in dive depth studies of marine endotherms

Abstract

Most depth recorders used to study the diving behaviour of polar marine endotherms record depth data at specific time intervals. The length of recording interval can have potentially profound implications for the interpretation of the data. We used data acquired on the diving behaviour of king penguins, Aptenodytes patagonicus, to examine the validity of various analyses routinely conducted on depth data. In our experiments, increasing the sampling interval led to an underestimation of the number of dives performed, an overestimation in mean dive duration and substantial changes in the form of the dive profile. Our analysis indicates that depth data should be recorded at a minimum rate corresponding to 10% of the total dive duration and that conventional dive profile categorization may be inappropriate. Alternatives that are less subjective, and based on curve fits of dive depth versus time, are proposed.