PFRP Trophodynamics Projects

Integrating Conventional and Electronic Tagging Data into the Spatial Ecosystem and Population Model SEAPODYM

Progress Reports: FY 2012, FY 2011

Project Summary

The Eulerian model SEAPODYM was developed for describing the spatio-temporal dynamics of tuna populations under the influence of environment and fishing pressure (Lehodey et al., 2008). Using a rigorous numerical methodology incorporating catch and length frequencies data within the model allowed significant improvement of the agreement between observations and model predictions (Senina et al 2008). However, in a number of cases fishing data do not allow estimating all model parameters. Among reasons are weak or absent signal, seasonality of fishing effort and seasonal changes of fishing practices within a fishery. This proposal seeks to integrate conventional and electronic tagging data explicitly linking observed movement to the model in order to improve an estimation of movement and habitat parameters. We have already tested two different approaches to incorporate tagging data into a simplified one-cohort version of SEAPODYM: 1) using conventional tag release-recapture dataset in a Eulerian framework; 2) employing SEAPODYM’s movement rates by a Lagrangian model and estimating the best track via unscented Kalman filter. A third approach is envisaged. It consists in estimating from the tagging datasets a space/time vector field of movement (Preisler 2004, Brillinger 2007) and incorporating these intermediary results into SEAPODYM’s likelihood formulation. Thus the objectives of this project are to validate and compare the performance of these approaches integrated into the full population dynamics model; to develop the fully operational computer model with parameter estimation from fishing and tagging data; and to conduct two case studies, with a focus on Pacific skipjack and yellowfin tuna populations.

Objectives

This proposal aims on integrating conventional and electronic tagging data with SEAPODYM habitat-based population dynamics model. We strongly believe that merging the tagging data and fishing data in the data assimilation framework will facilitate the parameter estimation, improve the model predictions and enable us to move toward higher spatial and temporal resolutions. Thus, we propose to develop a fully-operational model allowing assimilation of fishing and tagging data, both conventional and electronic, and to validate it performing two case studies, with a focus on Pacific skipjack and yellowfin tuna populations.

Electronic tagging data provide detailed information on the movement of individual fish that are likely influenced by small to mesoscale oceanographic features. Therefore to reduce as far as possible the mismatch between observation and prediction, it becomes essential to increase the realism of physical and biogeochemical forcing fields used to drive the model SEAPODYM, and in particular the movement of fish. We propose to use the most realistic available oceanic environment for driving SEAPODYM simulations at resolutions varying from 1/4° x 6 days to 1° x month.

Since the 1980s, the Oceanic Fisheries Programme of the Secretariat of the Pacific Communities (SPC) has conducted three large scale conventional tagging experiments, with the largest number of tags released on skipjack and yellowfin. During the last decade these tagging experiments included also electronic archival and satellite pop-up tags. Similarly the Japanese Fisheries Research Agency (FRA) has released conventional tags on skipjack to investigate the seasonal migration of this species along the coast off Japan and in the Kuroshio extension. In addition the domestic Japanese pole-and-line fishery provides high resolution catch and effort and size frequency data that can be included in the model optimization and evaluation process. We propose to apply the new version of SEAPODYM to Pacific skipjack and yellowfin using tagging and fishing data and the new enhanced environmental configuration to provide more realistic population assessment at high resolution.

Expected Outcomes

Expected outcomes for Year 1 include:

SEAPODYM version 3. The final product of this study will be a new operational upgraded version of SEAPODYM with fishing and tagging data assimilation. The development phase will take place during the first year of the project. The computer application should be able to use different types of data with a flexible construction of the total likelihood. The application will be tested on smaller spatio-temporal domain and artificial dataset, the strengths and weaknesses of each proposed approach will be explored.

Tagging System Simulation Experiment (TSSE). An interesting by-product of the test phase associated to the development of the new SEAPODYM version should be a Tagging System Simulation Experiment (TSSE) that would be the equivalent of an Observing System Simulation Experiment (OSSE) applied to tagging data. OSSE experiments are routine studies in physical oceanography to examine the impact of simulated data from a given observational platform(s), and thus to decide sampling strategies, i.e., where to release these platforms to retrieve the best information to assimilate in an ocean circulation model. The advantages include easy control of the experiments, precise knowledge of the data properties and errors, and knowledge of the truth. Given the high cost of both conventional and archival tagging experiments, such tool would be certainly useful to test various scenarios of tagging at the scale of the ocean.

Model Configurations. The physical (temperature and currents), biochemical (primary production, euphotic depth and dissolved oxygen) and biological (Mid-Trophic Level) environmental variables required to run SEAPODYM tuna applications will be prepared for the Pacific ocean domain at original resolution of ¼°x 6 day and degraded resolution of 1°x month for the period 1998-2010. Physical variables and primary production will be used to predict production and biomass of MTL functional groups.

During year 2 of the project, researchers will conduct methods validation and comparison and prepare tagging data set (SPC-OFP) for skipjack and yellowfin tunas. High resolution simulations will also be run to get detailed descriptions of spatio-temporal dynamics of skipjack and yellowfin on a basin scale.

Funding for this project to be available late 2010.

[ Top of the page ]

References

Brillinger, D.R. (2007) Learning a potential function from a trajectory, Signal Processing Letters 14, 867-870.

Lehodey P., Senina I., Murtugudde R. (2008). A Spatial Ecosystem And Populations Dynamics Model (SEAPODYM) - Modelling of tuna and tuna-like populations. Progress in Oceanography, 78: 304-318.

Preisler, H. Ager, A., Johnson B. and J. G. Kie (2004) Modelling animal movements using stochastic differential equations, EnvironMetrics, 15:7 643-657.

Senina I., Sibert J., Lehodey P. (2008). Parameter estimation for basin-scale ecosystem-linked population models of large pelagic predators: application to skipjack tuna. Progress in Oceanography, 78: 319-335.

Principal Investigators
Dr. Inna Senina Marine Ecosytems Modeling and Monitoring by Satellites (MEMMS) Direction Oceanographie Spatiale CLS 8-10 rue Hermes 31526 Ramonville St Agne France email: isenina@cls.fr Co-PIs: Dr. Patrick Lehodey, CLS email: plehodey@cls.fr Dr.Francois Royer email: froyer@cls.fr	Collaborators: Dr. John Hampton, SPC email: johnh@spc.int Dr. Simon Nicol, SPC email: simonn@spc.int Dr. Miki Ogura, NRIFSF, Japan email: ogura@affrc.go.jp Dr. Hidetada Kiyofuji, NRIFSF, Japan email: hkiyofuj@affrc.go.jp

Download free Adobe Acrobat Reader to view PDF files.