Advice on configuring ontogenetic migration in SS3
Paola Castellanos
Dear SS3 developers,
We are developing an SS3 assessment model for Northeast Atlantic black scabbardfish (Aphanopus carbo) and would appreciate your advice on how best to represent its ontogenetic migration within the SS3 spatial framework.
We have experience developing single-area SS3 assessments, but this is our first attempt to implement a spatial model.
Current biological knowledge suggests the following life cycle: spawning occurs around Madeira, larvae and/or eggs progressively migrate northwards and recruitment occurs W of the British Isles where juveniles remain some years to grow, then move to the Portuguese mainland coast where they remain some years to grow, and finally immature adults move to Madeira to mature and spawn.
Our available data include:
- Commercial catches from three fleets (Madeira longline, Mainland Portugal longline, and French trawl).
- Length-composition data for all fleets.
- Growth and maturity information.
- No fishery age-composition data.
We are considering either a single-area model or a spatial model with multiple areas representing the main stages of the life cycle. Our main question, however, is not whether to use a spatial model, but how to configure the migration process in SS3.
Since SS3 implements movement by age, we would appreciate your advice on the following:
- What would be the most appropriate way to represent this type of ontogenetic migration in SS3?
- How should the age-specific movement chronology be configured to represent a progressive migration throughout the life cycle?
- Are there any published SS3 assessments or example models with similar ontogenetic migration patterns that could serve as useful references?
Our goal is to develop a biologically realistic model while keeping the configuration as simple as possible.
Thank you very much for your time and for the continued development of SS3.
Best regards,
Richard Methot - NOAA Federal
- What would be the most appropriate way to represent this type of ontogenetic migration in SS3?
- SS3 allows you to configure migration as the age-specific fraction of the stock in an area that moves to each of the other areas. It can be set up to allow the progressive movement that you describe. The age-specificity allows for a movement rate for young fish, a movement rate for old fish, and a linear interpolation of the rate for fish of intermediate age. This is described in section 8.4.3 of the SS3-manual.
- In general, SS3 can estimate movement rates when it has size or age composition data from the progressive areas. But, you generally need to assume asymptotic selectivity in order for the differences in length composition to inform movement rates. I attach a draft presentation from an ICES training that should be helpful.
- How should the age-specific movement chronology be configured to represent a progressive migration throughout the life cycle?
- I presume the French trawl fishery is catching fish as they move from the British Isles towards Portugal.
- The dynamic you should try to model is the degree to which the French trawl and Portuguese longline reduce the abundance of fish moving into the Madeira region. I think you could do this with just 2 areas. Area 1 being Madeira where most spawning and one fishery occurs. Area 2 being not-Madeira where recruitment occurs and two fisheries operate. In SS3, the spawning biomass is summed over all areas, so any mature fish in area 2 would be included. The recruitment process is not explicitly spatial with movement. In SS3, recruitment is simply the result of the total spawners and assigned to occur only in area 2.
- Are there any published SS3 assessments or example models with similar ontogenetic migration patterns that could serve as useful references?
- A few papers resulted from an international spatial modeling effort. SS3 was included in the model testing.
- Aaron M. Berger, Daniel R. Goethel, Simon D. Hoyle, Patrick Lynch, Caren Barceló, Alistar Dunn, Brian J. Langseth, Carolina Minte-Vera, Jemery Day, Haikun Xu, Francisco Izquierdo, Dan Fu, Nicholas D. Ducharme-Barth, Mathew Vincent, Arnaud Grüss, Jonathan J. Deroba, Giancarlo M. Correa, Jeremy McKenzie, Will Butler, Jie Cao, Craig Marsh, Teresa A'mar, Valerio Bartolino, Massimiliano Cardinale, Claudio Castillo-Jordan, Bjarki Þór Elvarsson, John Hampton, Andrea Havron, Pamela Mace, Arni Magnusson, Mark Maunder, Richard Methot, Sophie Mormede, Maria Grazia Pennino, Alfonso Perez-Rodriguez, Marta Cousido-Rocha, Thomas Teears, Agurtzane Urtizberea. ‘Building the (im)perfect beast’: Strategies for identifying appropriate spatial stock assessment model complexity from an international, blinded high-resolution simulation experiment. Fish and Fisheries. https://doi.org/10.1111/faf.70048
- Aaron M. Berger, Caren Barceló, Daniel R. Goethel, Simon D. Hoyle, Patrick D. Lynch, Jeremy McKenzie, Alistair Dunn, André E. Punt, Richard D. Methot, John Hampton, Clay E. Porch, Richard McGarvey, James T. Thorson, Z. Teresa A’mar, Jonathan J. Deroba, Bjarki Þór Elvarsson, Steven J. Holmes, Daniel Howell, Brian J. Langseth, Craig Marsh, Mark N. Maunder, Sophie Mormede, Scott Rasmussen. 2024. Synthesizing the spatial functionality of contemporary stock assessment software to identify future needs for next generation assessment platforms. Fisheries Research 275. ttps://doi.org/10.1016/j.fishres.2024.107008.
Daniel R. Goethel, Aaron M. Berger, Simon D. Hoyle, Patrick D. Lynch, Caren Barceló, Jonathan Deroba, Nicholas D. Ducharme-Barth, Alistair Dunn, Dan Fu, Francisco Izquierdo, Craig Marsh, Haikun Xu, Giancarlo M. Correa, Brian J. Langseth, Mark N. Maunder, Jeremy McKenzie, Richard D. Methot, Matthew T. Vincent, Teresa A'mar, Massimiliano Cardinale, Marta Cousido-Rocha, Nick Davies, John Hampton, Carolina Minte-Vera, Agurtzane Urtizberea. 2024. ‘Drivin' with your eyes closed’: Results from an international, blinded simulation experiment to evaluate spatial stock assessments. Fish and Fisheries. https://doi.org/10.1111/faf.12819.
Richard D. Methot Jr.
Stock Assessment Research Scientist (ST)
Northwest Fisheries Science Center
NOAA Fisheries | U.S. Department of Commerce
Office: (425) 666-9893
Mobile: (301) 830-2454
--
You received this message because you are subscribed to the Google Groups "SS3 - Forum" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ss3-forum+...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/ss3-forum/afbb11a2-2a89-44b8-a89f-604773533e52n%40googlegroups.com.
Mark Maunder
Sent: Thursday, July 2, 2026 8:54 AM
To: Paola Castellanos <paola.cas...@gmail.com>
Cc: SS3 - Forum <ss3-...@googlegroups.com>
Subject: Re: [SS3] Advice on configuring ontogenetic migration in SS3
paola.cas...@gmail.com
Dear Richard,
Thank you very much for your detailed response and for sharing the articles and training material. They are very useful. I would also like to thank Mark Maunder for the last reference he sent, which is particularly helpful to have as guidance.
Regarding the available data, we do not have age data. However, we do have CPUE information for all areas, as well as length-composition data that clearly reflect differences in size structure among areas, consistent with the expected population movement.
Your suggestion of defining two areas, with Madeira as Area 1 representing the main spawning area, seems to be a good and practical simplification. However, given that we do not have a known migration rate, I wonder whether an alternative approach could be to treat the three areas as fleets rather than explicit spatial areas, while assuming common biological parameters and spawning only in Area 1, Madeira.
Under this approach, the model would not estimate movement explicitly, but the area-specific CPUE and length compositions could still inform differences in availability and exploitation patterns.
What is your opinion of this approximation? Would it be a reasonable intermediate approach before moving towards a spatial model with explicit movement?
Many thanks again for your guidance.
Best regards,
Paola Castellanos Ossa
Mark Maunder
Sent: Tuesday, July 7, 2026 7:21 AM
To: Richard Methot - NOAA Federal <richard...@noaa.gov>
Cc: SS3 - Forum <ss3-...@googlegroups.com>; paola...@ipma.pt <paola...@ipma.pt>
Subject: Re: [SS3] Advice on configuring ontogenetic migration in SS3
|
You don't often get email from paola.cas...@gmail.com.
Learn why this is important
|
Paola Castellanos
Hi Mark,
Thank you very much for sharing this reference. It is extremely helpful, and I really appreciate you pointing me to Haikun's work.
This approach seems to fit very well with our current data situation, and I think it provides a very practical starting point. We will use the areas-as-fleets approach as the basis for configuring our first SS3 model and then evaluate whether the available data support moving towards a more explicit spatial structure in the future.
Thank you again for your guidance and for sharing these valuable references.
Best regards,
Paola
Paola Castellanos
Hi Mark,
Thank you very much for sharing this reference. It is extremely helpful, and I really appreciate you pointing me to Haikun's work.
This approach seems to fit very well with our current data situation, and I think it provides a very practical starting point. We will use the areas-as-fleets approach as the basis for configuring our first SS3 model and then evaluate whether the available data support moving towards a more explicit spatial structure in the future.
Thank you again for your guidance and for sharing these valuable references.
Best regards,
Paola