Modelling marine populations from physics to evolution - An Advanced Course from Nordic Marine Academy Organizers: Christian Jørgensen and Øyvind Fiksen, BIO, University of Bergen

Participants and venue The course was attended by 35 students and 11 lecturers. The students were distributed among 10 countries: Spain (1), Poland (1), Switzerland (1), Canada (1), Austria (1), Finland (3), Iceland (2), Denmark (3), Sweden (6), Norway (16). Most of the students were PhD students, but some were Master students, post-docs and young researchers. Lecturers and students were accommodated at Marine Biological Station at Espegrend (http://www.ifm.uib.no/lsf/inst2.html) about 20 km outside Bergen. As the picture shows, the weather was sunny and warm (as always during October in Bergen). Some were even swimming in the sea, repeatedly!

Content and lecturing The overarching theme of the course was modelling practices of marine systems, both physical and biological. These traditions interact at various scales and through many pathways and this course therefore spans across physics, ecology and evolutionary theory. We touched upon 1) large scale circulation models and biogeochemical cycles; 2) small-scale physics and predator-prey interactions; 3) evolutionary modelling, including behavioural models, life history theory, and fisheries-induced evolution. Practical training of students was limited to 2) and 3), and to methods where formal courses are not available for students in any Nordic country. Large-scale circulation models and biogeochemical models: Our understanding of the effects of global change on marine systems and organisms rely to a large extent on global and regional circulation models and climatology. These models are improving rapidly and large efforts are placed in their development and applications. They are invaluable for inferences about climate change. For fisheries scientists, biological oceanographers and marine biologists they pose a range of novel challenges: How will, for instance, productivity, marine communities and fish stocks react to the predicted reduction in ice-cover in the Arctic? How will primary production and cycling of carbon be distorted as wind, currents, and vertical stability of the water column changes? At the same time they open new eras of opportunities. Studies of recruitment variability of fish can now be studied in three-dimensional ocean models with currents and environmental forcing, and provide trajectories of particles, their origin and spatial drift patterns. We are in the early phase of the scientific unravelling of the fine-tuned adaptations of plankton and fish to large scale circulation patterns. Working on such models requires specialised team-work, but interpretation, utilisation of outputs and offline data are becoming more common, and available also to biologists. Prof. Helge Drange (Nansen Environmental and Remote Sensing Center, Bergen, Norway)gave an introduction to global circulation models, and Prof. Christoph Heinze (Bjerknes Centre for Climate Research, Bergen, Norway) introduced us to marine biogeochemical modelling. These two lectures reviewed the state-of-the-art within the large-scale circulation of oceans and matter. Small-scale physics and predator-prey interactions: Our ability to observe and model processes at the scale of individual zooplankton and larval fish has improved immensely over the last decade. With these observations came also a realisation of how important physical properties of the water-column are in predator-prey interactions, and an appreciation of the role of behaviour on individual modification of growth and predation risks. Today, detailed mechanistic physical-biological models of predator-prey interactions are emerging from bacteria to fish and jellyfish. With global circulation models becoming more precise and reliable, the next move forward will be to integrate large-scale circulation models with the biological scale of the water-column and predator-prey processes. As lecturers in these topics we had Senior scientist Andy Visser (Danish Institute for Fisheries Research, Copenhagen, Denmark) and Prof. Per Jonsson (Tjärnö Marine Biological Laboratory, Göteborg University, Sweden). Visser and Jonsson are both among the leading scientists within the subject of how physical properties of water affect planktonic organisms, and are excellent lecturers as well. Visser provided 4x45 minutes of lectures, and alsoled a practical course in Matlab on how models are

formulated and explored with this software. Jonsson gave give 4x45 minutes of lectures. Evolutionary algorithms: The abiotic and biotic environments in which marine organisms live change at unprecedented rates. Anthropogenic climate change affects temperature, circulation, turbulence and turbidity, to name a few important physical characteristics that are important for marine life. At the same time, steadily increasing exploitation of marine living resources has for most harvested species increased mortality dramatically, which has repercussions on how a species should time growth, maturation and reproduction. The basic problem when dealing with such changes is that biology changes incessantly because of evolution, while we know the biology of a species only as it is today or has been in the recent past. We might expect nonlinearities and dramatic shifts between several stable ecosystem configurations, but when, why, or how to forecast them? To be able to predict evolutionary change, there is no shortcut but to understand the mechanisms that shape the adaptive landscape of individuals, and at all levels integrate a full understanding of the evolutionary process. The introduction to this section covered life history theory, the theoretical basis allowing integration of the physical and biological environment with individual strategies with respect to both behaviour as well as growth and maturation strategies. Thereafter, this section focussed on methods for evolutionary modelling. Evolutionary individual-based genetic-algorithm models and adaptive dynamics are still awaiting the publication of comprehensive introductions, and while introductions to dynamic programming exist in textbooks, the most recent contributions are published scattered as part of journal articles. Consequently, students in evolutionary modelling learned the techniques favoureded by their respective research groups, while acquisition of alternative methodologies is often underrepresented. We included broad introductions, coupled with practical workshops, to the three most widespread evolutionary algorithms by experts working with development of the respective methodologies. The techniques were: 1) Adaptive dynamics, 2) Individual-based genetic algorithm models, and 3) Dynamic programming. In addition, we covered one statistical method for assessing evolutionary change, namely the probabilistic reaction norm for maturation. The reaction norm method bridges nicely with modelling, and has proved important in assessing the degree of contemporary human-induced evolution in fish. Eventually, we hope the evolutionary perspective can foster a fuller integration of physical processes with individual strategies, which in turn propagate via populations and population interactions to ecosystem characteristics. The following lecturers contributed to this section: - Dr. Ulf Dieckmann (International Institute for Applied Systems Analysis, Laxenburg, Austria), who repeatedly has contributed to the foundation of adaptive dynamics theory and among the principal spokesmen for fisheries-induced evolution. He is currently a central actor to many fields of evolutionary ecology and theoretical biology. - Dr. Bruno Ernande (Laboratoire Ressources Halieutiques, IFREMER, Port-enbessin, France), who is experienced with theoretical and applied adaptive dynamics, fisheries-induced evolution, as well as evolutionary biology in general. - Dr. Mikko Heino (Institute of Marine Research, Bergen, Norway), who was central in devising the probabilistic maturation reaction norm as a means to separate evolutionary change form phenotypic plasticity, and has applied this method to many and diverse fish stocks. He has also contributed repeatedly to many fields in evolutionary ecology and theoretical biology.

- Dr. Erin Dunlop (International Institute for Applied Systems Analysis, Laxenburg, Austria), who has been investigating fisheries induced evolution in individual based eco-genetic models (she was a student at the course as well). - Dr. Geir Huse (Institute of Marine Research, Bergen, Norway), who developed individual-based neural-network genetic algorithm (ING) models, and is currently working also with other simulation-based evolutionary methods. - Dr. Espen Strand (University of Bergen, Norway), who has combined behavioural and life-history evolution in ING models. - Dr. Øyvind Fiksen (University of Bergen, Norway), who has long experience with complex dynamic programming and has used this technique to build an evolutionary understanding of zooplankton behaviour and life history strategies. - Mr Christian Jørgensen (University of Bergen, Norway), who is working with life history evolution and fisheries-induced evolution in cod with several of the above. He is using dynamic programming, a technique he also has extended to become evolutionary consistent. Timetable:

Food:

The food was traditional Norwegian – and students enjoyed it!

Course material Lectures, exercises, reading list, contact information and pictures can be downloaded from: http://www.ifm.uib.no/Download/Fiksen/. List of participants Name Erin Dunlop Paul Venturelli Gert Mehl Virenfeldt Martin Pedersen Kristine Skovgaard Madsen Jan Hesuchele Inga Hense Päivi Laine Jónas Páll Jónasson Gudni Magnús Eiríkson Ivo Orellana Raul Primicerio Paolo Simonelli Øystein Varpe Shahaama Sattar Kjersti Eline Larsen Ingrid H. Ellingsen Anne Maria Eikeset Cecilie Hansen Tian Tian Dorothy Jane Dankel Sigrunn Eliassen Trond Kristiansen Frode Vikebø Thomas Torgersen Jostein Starrfeldt Kryszof Switek Agurtzane Urtizberea Sarah Robinson Wolrath Elin Lindehoff Wanderson F de Carvalho Geir Halnes Karin Nilsson Davnah Urbach Olivia Langahamer

Institution IIASA U Toronto DIFRES DTU/DIFRES U Copenhagen U Helsinki Finnish Environ Inst U Turku U Iceland U Iceland IMR U Tromsø U Bergen U Bergen U Bergen U Tromsø SINTEF U Oslo Nansen-senteret U Bergen U Bergen U Bergen U Bergen U Bergen U Bergen U Oslo MIR ATZI U Uppsala U Kalmar U Kalmar U Uppsala U Umeå U Lausanne U Uppsala

Country Austria Canada Denmark Denmark Denmark Finland Finland Finland Iceland Iceland Norway Norway Norway Norway Norway Norway Norway Norway Norway Norway Norway Norway Norway Norway Norway Norway Poland Spain Sweden Sweden Sweden Sweden Sweden Switzerland Sweden

Nansen-senteret U Bergen U Bergen U Bergen DFU IMR Bergen U Bergen U Gøteborg IIASA IFREMER IMR Bergen

Norway Norway Norway Norway Denmark Norway Norway Sweden Austria France Norway

Lecturers: Helge Drange Christoph Heinze Øyvind Fiksen Christian Jørgensen Andy Visser Geir Huse Espen Strand Per Jonsson Ulf Dieckmann Bruno Ernande Mikko Heino

Course evaluation

NMA COURSE EVALUATION Name of course: Modelling marine populations from physics to evolution - autumn 2005 A total of 23 course evaluations were delivered.

Summary of evaluation: 1.

Correspondence between the objective and content of the course has been:

Very good 14

2.

2

0

Very bad 0

How was the work load compared to the number of credits (studiepoeng)?

Much too high 0

3.

7

4

16

3

Much to low 0

Where there areas where you didn’t have necessary previous knowledge?

No

Yes 4

18

If yes: in: - ING- hard to grasp - Physical aspects such as climate change, climate modelling - Mathematics - Many new concepts, a prioritized readinglist would have been nice - Mathematics and programming languages - Theoretical ecology - Most of them - Oceanography - Mathematical models - Marine physics, climate models - Oceanography, modelling workshops - Biology - Software, genetics, evolution - Physics, mathematics - Evolutionary ADT

4.

How will you assess the lectures?

Very useful 10

9

4

0

Not useful at all 0

Comments: - Overall very good, some accents where a little hard to understand - Although some had quite high standards, for people who did not have prior backgrounds. Lectures were good. - The a…. was not always very good……apparently very high. In some lectures more general aspects would have been better. - Generally very high standard, but some lectures didn’t relate so much to my field

-

-

Sometimes I experienced the program as to tight. The tight schedule didn’t encourage real discussions Varying degree of level. To little focus on the model building procedure, to much focus on results Good lectures - especially Ulf Dieckmann who presents very clearly Generally high standard Not top graded since not all lectures focused on methods Some of them not directly related to my work, but nevertheless useful to give new ideas Some of the lectures it was better to see some results of modelling instead what be good to model e.g. Per Jonsson lectures – it was more biology lecture than model lecture Generally high pedagogic level, with few exceptions mainly due to time constraints Some lecturers missed the pint. Tech the tool and its application

5.

How will you assess the field course/practicals?

-

Very useful 5

9

4

3

Not useful at all 0

Comments: - Good to do not just listen and practice what you have just learned - Quite good, although a very brief introduction - Some times hard to understand the instructions preliminary for …may have been useful, e.g. in the form of abstracts. - Ok, but a bit to little time to go through the real essentials. More lab would be ok - Not focused enough on the methods just on results of some models - The time was to short to appreciate them. By the way the help was good and it allowed to follow the practical also to the persons who were not skilled in models - Good workshops , coding and modelling - It might have been an ides to spend more time in one language and one topic? - Given the constraint, very well thought practicals giving a sense of the heuristics for the different modelling approaches - The more we got to play with the models on our own the better get a feel for what tools is capable of/useful for. Present participants with problems to solve to give some structure 6.

How will you assess the seminars/students presentation?

Very useful 9

10

4

0

Not useful at all 0

Comments: (on the pedagogics, the standard, did you get the necessary help) - Good opened for discussions - Was a good opportunity to know those who are doing the same area of work for future collaboration - To many presentations took a large part of the course time. Presentations however very interesting - Some were nice and interesting for me - It was very good to be able to know what the others students were working on in their different topics - Very good to hear about colleagues work - Good feedback after the presentation - It is great to share some experience with others in the similar field and also to learn some new - Nice to see what others are doing. Generated a lot of discussion

7. How will you assess the syllabus (extent and content)? Too big 1 11 9

0

Too small 0

Very good 4

0

Very bad 0

15

2

Comments: - It would have been nice if the order of some lectures are changed (Saturday before Friday), but of course it is sometimes not possible due to organization - For a week length course, it was a lot of material to cover, but the content was very good - It takes a lot to put in 4 ECTS in one week of lectures, by the end of the week it has been difficult to keep up concentration. A day off during the week would be help. Or maybe a field trip - Ok - Have a smaller list of required reading and a large list of continued studies in each area

8.

How will you assess the organizing and practical arrangement of the course?

Very good 19

4

0

0

Very bad 0

Comments: - A job well done! Food, accommodation, reception - perfect. It has been a fun week - Everything very fluent, but sometimes to strict. Some kind of get to gether program would make the start easier - Great organization, great food - Congratulations! - Perfect. Good food, nice place (although it was a bit hard to get anywhere) - Everything very well organized. The food was great; the general social atmosphere was great! - Fun to be at Espegrend - Excellent organization - Very good food and good arrangement - Excellent; logistics, time schedule, organization of topics - I am very impressed

9.

Your general impression of the course:

Very good 13

8

0

0

Very bad 0

10. Did this course readily combine with other courses that you took this term? Yes, very much so 5

5

3

Comments: - The demands were higher here, but also the outcome - No other courses taken - Well, it is always hard to get a week off, but once you are there.. - This is my first course this term - Did not take courses so far

0

No, not at all 0

11. General comments / suggestion for improvement / criticism: -

-

-

-

-

Instead of 4 small workshops. 1 extended workshop with clear task definition I have never been to Norway or in a marine environment like this one. A little boattrip would have been nice Very useful course, but a lot of information in a short time. Might be better if it was spread out over a longer period than a week with shorter days Thank you! I really liked the course. I just would like to have a bit more examples about my field. Maybe for the next you could evaluate the proposal earlier and try to bring somebody that can talk more about all subjects Maybe a less tight schedule. They days were kind of exhausting, and it was hard to concentrate through all lectures. I know it is hard to make it happen, but more discussions in the classrooms would be a fresh addition. Thanks for a good week. More methods. Practicals were not always that good. Good job ☺ The food was great! Great relaxed atmosphere. Very good to network with many “famous” researchers. Organized boat trip would have been nice. It was really great! Focus more on methods instead pf results. This applies for a few of the lectures. Great organization, motivation supply for the next year Very good back ground of the course. Some lectures were very good, but some other topic were a bit high for my background Main comments are on the time available for the practicals. The many students presentations took away pc-lab time. In the future the only solution will probably be to limit the number of attendants The course structure is very good. I feel satisfied with the lecture and practical in physical processes, marine dynamics modelling and IBM. Evolution theory is necessary to my knowledge when I do some study in population dynamic. And I like the atmosphere for young people to communicate each other. Last but important the food let me know more about Norway. Extremely valuable. Excellent opportunity to bring young minds together to learn, being exposed to different ideas. Good atmosphere and food. Reduce length of day to prevent burnout. More sessions to after dinner? Or have small group brainstorming sessions in evening. Could also be these to discuss a topic/tools so that we are sure to really get it. Thank you!