A Bayesian nested patch occupancy model to estimate steelhead movement and abundance

Area of Science:

• Ecology
• Conservation Biology
• Fisheries Science

Background:

• Anthropogenic impacts on rivers raise concerns about aquatic species' population status.
• Passive integrated transponder (PIT) tag studies with in-stream detectors are used to monitor species movements in river networks.
• Quantifying animal movements using imperfect detection data in river networks is challenging.

Purpose of the Study:

• To develop a flexible Bayesian analytic framework for modeling imperfectly detected movements of tagged individuals in river networks.
• To provide probabilistic estimates of upstream migration routes and convert them into abundance estimates.
• To evaluate model performance and inform future monitoring and management decisions.

Main Methods:

• A flexible Bayesian framework was developed to model imperfectly detected movements in a nested PIT tag array river network.
• The model estimates probabilistic upstream migration routes based on nested state variables.
• A simulation framework was used to evaluate model performance based on tagging rates and array detection probabilities.

Main Results:

• The framework provides probabilistic estimates of individual movement routes within river networks.
• Movement estimates can be converted into abundance estimates when local abundance is known.
• Model performance was evaluated using steelhead (Oncorhynchus mykiss) data from the Upper Columbia River basin.

Conclusions:

• The proposed Bayesian framework offers a robust method for analyzing PIT tag data in river networks.
• The approach can improve the accuracy of abundance estimates for species of concern.
• Simulation results inform decisions on optimal PIT tag array configurations and tagging strategies for effective riverine species monitoring and management.