Modeling life history parameters in marine mammals: Bayesian hierarchical analysis of ontogenetic dietary shifts in Indo-Pacific humpback dolphins
Preprint: 
Mean δ15N isotope values (‰) across different ages (represented by growth layer groups) in 38 Indo-Pacific humpback dolphin carcasses stranded ashore in the Pearl River Delta region between 2007 and 2018. Vertical bars represent standard deviation (± ‰). All analyzed individuals had at least five growth layer groups identified in their teeth, of which 14 were males (M), 18 were females (F), and the sex of 6 individuals was undetermined (Un).
Weaning represents a pivotal ontogenetic process for mammals, marking the transition from parental provisioning to independent foraging. In monophyodont species, distinct growth layer groups (GLGs) that are deposited in their teeth as the animals age represent a permanent chronological archive of physiological events across the animals’ lifetime. Thus, biochemical analysis of annual dentine increments provides a means to explore animal physiological history. We examined the age-specific pattern, individual-level variations, and sex-related differences in dentine nitrogen isotopic values in the Indo-Pacific humpback dolphin (Sousa chinensis) using tooth samples collected from 38 carcasses that stranded ashore in the Pearl River Delta region, southeast China, between 2007 and 2018. The longitudinal isotopic records archived in dolphin teeth offered insight into their foraging ecology and individual ontogenetic dietary life history. The overall pattern of δ15N isotopic values in the incremental layers, analyzed under a hierarchical Bayesian framework, indicates that humpback dolphins typically undergo an ontogenetic dietary shift (i.e., wean) before reaching the age of three years (mean 2.394 ± 0.143 years), albeit there is a considerable individual heterogeneity (ranging from 1.548 to 4.180 years), with males weaning consistently ~3.5 months earlier than females. Our study underscores the importance of quantifying ontogenetic parameters at the individual level as overlooking individual variations in life history events (such as the age of weaning) may introduce biases in the broader population-level life history metrics. The application of hierarchical Bayesian modeling proved effective in quantifying individual heterogeneity and factoring it into the estimates of ontogenetic dietary shifts — an important component in analyses of broader population-level processes.
All source code used to generate the results and figures in this paper is available in EEP-sousa-weaning.ipynb.
The data supporting the findings of this study are currently retained for ongoing analysis in preparation for future publications. We plan to make the data publicly accessible upon completion of the study. For interim data requests or metadata inquiries, please contact Yuen-Wa Ho.
Yuen-Wa Ho, Leszek Karczmarski*, Wenzhi Lin, Mandy C.Y. Lo, Yuping Wu, David M. Baker* (In Press as of December 2024). Modeling life history parameters in marine mammals: Bayesian hierarchical analysis of ontogenetic dietary shifts in Indo-Pacific humpback dolphins. Ecological and Evolutionary Physiology