Full text: Download
© 2016 Rogers et al.Biological markers (biomarkers) are invaluable and widely adopted in ecology, archaeology, and anthropology. Serially sampling biomarkers along continuously growing inert tissue, such as vibrissae, hair, nail, horn, or baleen, is an ideal method by which to capture the changes in an individual's diet, environment, climate, health, and stress levels. However, there are complications in calibrating the time periods that samples represent. Here, we assess how the choice of model and the input data used affect the prediction of time span. We use the Antarctic leopard seal, Hydrurga leptonyx, as our model, with stable isotope data from the vibrissae of captive and wild animals. We show that for tissues with nonlinear growth, and where fine-scale data on tissue growth are unavailable, modeling over the life span of the tissue is a simple and easily adopted approach. In this method, growth parameters are derived from surveys of the study population and the shed.replacement cycle to calibrate the time span of the biomarker data. This model performed better than linear and nonlinear models, which use parameters that have been derived from growth measurements documented over short periods relative to the life span of the tissue. These latter approaches performed in a similar fashion to each other as they tended to overestimate the life span of the tissue. Whether growth is linear or nonlinear, not accounting for position-specific differences in growth (rates or coefficients) and asymptotic length (L∞) resulted in different interpretations of biomarker data across the samples collected from the same individual. In species where metabolically inert tissue grows in a progressive fashion, the ability to account for time-specific information refines our ability to interpret the biomarker data. We recommend that this approach be adapted for tissue, such as the vibrissae and hair of mammals and the hair of humans, which exhibit the predictable growth.