Estimating migration patterns of fish from otolith chemical composition time series

Understanding migration patterns and habitat use is of great importance for management and conservation of marine living resources. The chemical composition of otoliths is influenced by the surrounding environment; therefore, they are indispensable data archives. To extract migration patterns and historical habitat use of individual fish, we analyse otolith chemical compositions obtained by laser ablation inductively coupled plasma mass spectrometry by a regime-switching state-space model. The state-space model filters the measurement noise from the environmental signal. In turn, the filtered signal is converted to geographical positions through a calibration of strontium to salinity. The method is validated by a simulation study and applied to 404 Atlantic cod (Gadus morhua) otoliths.