In August and October 2013, I travelled to Iceland and Finland as part of my Ph.D. dissertation research on Bayesian stock assessment methods and the integration of aboriginal and local knowledge into fisheries stock assessment and management. The trips were funded by the Canadian Fisheries Research Network (CFRN) through a Strategic Network Enhancement Initiative (SNEI) grant from the Natural Sciences and Engineering Research Council of Canada (NSERC). The objective of these international visits was to meet with some of the research groups and individuals in Iceland and Finland who are focused on stakeholder engagement and stock assessment, and particularly in the case of the Fisheries and Environmental Management (FEM) Group at the University of Helsinki, are using Bayesian methods to implement ecologically realistic and highly interdisciplinary stock assessments and fisheries management planning

ICELAND

I was fortunate that my one week visit to Iceland coincided with the summer vacation of fellow CFRN and Guelph node PhD candidate Davíð Gíslason who was in the middle of a 3 week visit to his Icelandic homeland with his son Dúi Davíðsson when I arrived at Keflavik on August 19^th. After a much needed visit to the local sundlaugar (swimming pool/spa) to get the full Icelandic geothermal hot tub and steam room experience, we consumed several pylsa (delicious Icelandic hotdogs) at Bæjarins Beztu near Reykjavík harbour. This humble hot dog shack is probably Iceland’s most famous restaurant. They take pride in a picture of U.S. President Bill Clinton scarfing down a pylsa beside the stand – mustard only – it’s called a “Clinton” if you want to order one. 

It was a great way to shed some of the jet lag before we drove 5 hours north to Skagafjörður and nearby Hólar University College where I was to meet with Dr. Skúli Skúlason former Rector of Hólar and Dr. Sigurður S. Snorrason, Head of the Faculty of Life and Environmental Sciences, University of Iceland.  

On the way to Hólar we stopped at the fishing village of Sauðárkrókur where Davíð showed me the cod head drying racks. Here, what appear to be millions of cod heads are strung, about 8-10 at a time, by short sections of polypropylene rope and hung from these racks which seem to extend for many 100s of meters along the shoreline. These open air dried cod heads are shipped to Nigeria where they are sold in local markets as an inexpensive source of protein for the making of fish head soup. I wonder if Canada ever shipped dried cod heads to Africa in the days before the moratorium.

Cod heads being dried outside in the traditional way. 

Photo Davíð Gíslason

Hólar, a remarkably beautiful, historically significant and highly collegial place, is situated in the narrow but stunning valley of the Hjaltadalsá River. Since 1106, Hólar has been the seat of Bishops (initially Catholic – the last one was beheaded, later Protestant) and administrative center in northern Iceland. The aquaculture program at Hólar University College is the Icelandic center for Arctic charr research and experimental breeding, and also produces embryos for the charr farming industry. I had a number of excellent discussions with both Skúli and Siggi about my thesis research and on-going ecological research at Hólar and University of Iceland. 

Skúli recently facilitated a stakeholder engaged process to review the management of the Icelandic commercial fishing industry and we were able to compare notes and have a good discussion about our experiences with high stakes, conflict-ridden fisheries decision problems and processes, and some implications for my research on incorporating structured decision aiding and stakeholder ecological knowledge into stock assessments. Skúli also provided an introduction and site visit to the Hólar-associated Verið Science Park and research facilities in nearby Sauðárkrókur where we met a group of international researchers who are collaborating to investigate the roles of ecology, genetics and development in generating diversity among fishes. 

Hjaltadalsá River with view of Hólar University College down the valley. Photo Davíð Gíslason

Dr. Skúli Skúlason (centre), his partner Sólrún Harðardóttir (right) and CFRN student Davíð Gíslason and son Dúi Davíðsson (left) at Hólar. Photo Kevin Reid.  

Kevin Reid (left), Dúi Davíðsson (centre) and Dr. Skúli Skúlason (right) standing in front of Nýjibær (a traditional sod farmhouse) at Hólar. Photo Davíð Gíslason

Discussing Arctic charr production with Friðrik Steinsson, Production Manager at the Hólalax fish farm near Hólar.

Photo Davíð Gíslason 

During a visit to the nearby Arctic charr farm, a couple kilometers up the valley from Hólar, I was impressed by the density of charr in the facility (80 kg per m³) and the use of geothermal water to heat the river water used in the farm to the temperature range that is optimal for charr growth. Originally part of the aquaculture teaching program at Hólar, the farm has since been privatized, but Hólar students continue to be involved in the production of charr at the farm as a component of their training. On our way up to the farm we encountered a number of Icelandic sheep, scurrying along the road in front of the car for several minutes before running off to the side to congregate around a large pipeline that provides geothermal water to Hólar.  

Icelandic sheep at the geothermal water pipeline in Hjaltadalur  

Photo Davíð Gíslason 

The black sand beach in Skagafjörður near Sauðárkrókur  

Photo Davíð Gíslason 

On the way back to Reykjavík, the black sands on the beach near Sauðárkrókur offered a striking contrast with the verdant dune grasses that have been planted to stabilize the dunes thus preventing wind-blown sand from blocking the road that runs behind the dunes.

FINLAND

In late September, I left for Finland where I was invited to discuss my thesis research and Great Lakes fisheries with the members of the Fisheries and Environmental Management (FEM) Group at the University of Helsinki. I wanted to provide my hosts with some useful and interesting insights into Great Lakes fisheries, get some feedback on what I had done so far on my thesis, and to gather some expert insight into the hypotheses, predictions and study design, for the work I have planned for chapter 3 of my dissertation. I enjoyed numerous and detailed discussions with several FEM group members over the course of my week-long visit to the very attractive Viikki campus, referred to as the “green campus” because of the high tech energy efficient design of most of its buildings. I am particularly grateful to fisheries sociologist Päivi Haapasaari, stock assessment scientists Samu Mäntyniemi and Mika Rahikainen, statistician/analyst Paul Blomstedt, ecologist Asta Audzijonyte, modeller/Bayesian network specialists Annukka Lehikoinen and FEM-associate Laura Uusitalo of the Finnish Environment Institute (SKYE).  The FEM folks were highly collegial.  I enjoyed the stimulating and useful lunch-time conversations I shared with most members of the group at the Gardenia, an on-campus greenhouse with a really nice cafeteria only a 5 minute stroll away from the FEM group’s offices.

I spent a fun evening at the home of friend of the CFRN, Laura Uusitalo, her partner Edu and their two youngsters where, during the after dinner conversation, Laura offered some very useful insight into how to deal with what I suspected to be a flaw in my value of information analysis.  

Mika Rahikainen (left)  

Mika, Päivi and I spent a really entertaining (for me at least) afternoon discussing alternative approaches to eliciting the knowledge or beliefs of stakeholders, scientists and managers, and the advantages, and particularly the limitations, of BNs for doing elicitations and incorporating the elicited knowledge into stock assessments. I came away with an improved understanding of methods used by FEM members for elicitation, knowledge incorporation via BMA, and conditioning stock assessments. I believe that we also generated some good ideas for chapter 3 of my dissertation - a simplified but robust approach to using BNs for elicitation of knowledge and the conditioning of stock assessment models on that knowledge. 

left to right Päivi Haapasaari, Samu Mäntyniemi, Kevin Reid, Riikka Venesjärvi,  Annukka Lehikoinen and Laura Uusitalo

On the morning of my last day with the FEM group I led a seminar about the CFRN, the evolution of the industry-academic partnership that lead to the establishment of the Guelph node within CFRN, and my thesis work on using state-space surplus production models to assess to the status of the Lake Nipigon whitefish fishery, communication of fishery status, uncertainty and risk via graphical methods, and value of information (VoI) analysis. I presented an earlier version of my VoI analysis, one that I had previously presented at the Canadian Conference for Fisheries Research (CCFFR) and elsewhere, and for the first time, several members of the FEM group immediately noticed the error in my earlier VoI analysis. The fact that over the course of several presentations to large academic audiences no one had previously observed the error is a pretty clear indication to me that many Canadian ecologists and fisheries scientists seem to know little about VoI analysis. I had used some of my time during the week to re-run the VoI analysis based on the mid-week insights provide by Laura, so I was able to present a revised version of the Lake Nipigon whitefish VoI analysis to the FEM group.

After lunch, Samu Mäntyniemi gently schooled me about some deficiencies he detected in the hierarchical structure of the surplus production models used in the analysis, and the need to be more clear about why I am using hierarchically structured priors in the first place, and what each of the levels of the hierarchically structured priors represents. Samu also led some useful discussion about model checking, particularly in a Bayesian context. He recommended that I rely not only on the DIC to compare the performance of the models, but to extend the process to other forms of model checks such as using Bayesian p-values. He also provided me with a copy of his Ph.D. dissertation which gave me some deeper insights into the utility of this posterior predictive approach to model checking; an approach that I now intend to incorporate into my chapter on Bayesian stock assessment methods.

At that point, several FEM members had to leave the seminar and we were joined by some others who had been unable to come in the morning. I then presented some historical and current information on the commercial fisheries of the Great Lakes which lead to some discussion about the strengths and weaknesses of a couple of the decision analysis initiatives that have been implemented in the Great Lakes basin, specifically the problems of downstream passage of American eel through hydroelectric  turbines in the Lake Ontario-upper St. Lawrence R. system and walleye management on Lake Erie.  We ended the seminar with some discussion about the similarities between the Baltic and the Great Lakes and the common environmental and fisheries-related problems confronting both ecosystems and the enormous potential for future collaborations between CFRN and the FEM Group members.

SUMMARY

Without a doubt, these working visits to Iceland and Finland were highly productive for making progress on my dissertation research, and inspirational for my development as an analyst.  I was already an enthusiastic Bayesian before these visits, but I came away even more convinced that Bayesian methods, and particularly Bayesian networks (BNs), are well suited to helping fisheries science and management become interdisciplinary. By providing the tools and techniques to integrate knowledge, and ways of knowing, across a wide range of disciplines, e.g.,  operations research, economics, ecology, ekistics, sociology, ecosociology and the other humanities including psychology, ecopsychology, ecumenics, ecosophics and other eco-studies, Bayesian methods such as BNs have potential to foster the stakeholder engagement, the emergence of human networks and interdisciplinarity; to shift fisheries science and management toward an ecogenic (sensu Regier 2013) approach that can integrate humans and human-induced uncertainty with ecosystem-based fisheries management. 

It will continue to be hard to move fisheries management and fishing industries toward an ecogenic approach or to “grow into interdisciplinarity”, as the FEM folks have put it. The FEM group have very ably shown how BNs have the potential to help in their own backyard, but more global success requires a simpler but still defensible approach, an increase in the number of skilled and experienced practitioners, and a few more worked examples, i.e., real life applications, so that decision makers, managers and stakeholders will begin to appreciate, understand and, eventually, even demand the use of BNs to incorporate interdisciplinarity into fisheries stock assessments and management. 

Thanks again to my many hosts and colleagues in Iceland and Finland. Thanks also to the CFRN Implementation Team and management, for securing and providing the SNEI funding to support these visits, and also to the Ontario Commercial Fisheries’ Association, former OCFA Executive Director Peter Meisenheimer and A/Executive Director Dennis Cartier for their generous and on-going support of my dissertation research.

References

Regier, H.A.  2013. Perspectives on an ecosystem approach to ecogenic challenges in the Great Laurentian Basin and beyond. Aquatic Ecosystem Health & Management, 16(1):6–19.