Written by Nereus fellow, Fernando Gonzalez Taboada
And They Still Say Fish are Expensive! (¡Y aún dicen que el pescado es caro!) was painted by Joaquín Sorolla in 1894. This vivid painting represents a scene where two elderly fishermen relieve their younger partner, who lies fatally injured in the hold of a rudimentary fishing vessel. The portrait is considered one of the best representatives of the social realism movement in Spain. Its title hides a reference to a dialogue in a novel describing daily life in a small fishing village in the late nineteenth century. Contemporary to the emergence of labor movements, the scene calls attention to the precarious conditions and low revenue prevailing in some jobs during the industrial revolution. However, the most striking aspect of this painting is how its message remains valid today.
A close look at the portrait would hopefully inspire curious reflections in anyone. In my case, it brings to mind memories of a fairly limited seagoing experience, accumulated mostly in large oceanographic research vessels in open waters and many short sampling trips near the coast. Even if my time at sea has little resemblance with pre-industrial fishing, a couple of days at sea are more than enough to realize that a vessel is not the safest working environment. It immediately becomes evident why fishermen and other people working at sea still consider themselves to be at the mercy of the elements.
Nowadays, fishing remains one of the professional sectors with greatest morbidity and occupational injury levels, even in developed countries. As crude as it is, these risks generally do not correspond in terms of an elevated income or social protection. The overall situation contrasts with a rising demand in marine fish products that will continue to grow in the foreseeable future, with human numbers continuing to increase and many communities relying on wild caught fish as their main source of protein.
As one devotes more time to observe the portrait, longer excursions lead the mind towards digression and pondering deeper questions. What leads fishermen to expose themselves to the risks of fishing? How often do they operate under unfavorable weather conditions? Why do they need to travel long distances and spend many days at sea to find their catch instead of staying safe closer to shore? There are no simple answers to these questions, and when one digs into the literature there are a plethora of potential explanations that are rarely accompanied by practical solutions.
With a background in population biology and environmental sciences, I seek answers to the questions above by examining potential relationships between changes in ocean conditions and the dynamics of fish populations. If we are able to advance changes in fish abundance, then it should be easier to regulate fisheries and ensure a sustained income for fishermen – fewer risks and better lives.
Most often the approach relies on the analysis of long series of observations of changes in fish abundance and environmental conditions. In some ways, it can be regarded as a kind of forensic exam; a careful assessment of available evidence reveals the causes of a fisheries collapse, or what happened before a species became so abundant to support the establishment of a new fishery – explaining past changes as a way to predict the future. It seems appropriate, but unfortunately this type of approach is often disregarded in practical management settings.
Two main arguments are invoked to explain why dismissing environmental information and detailed mechanisms do not impair fisheries management. First, it is widely accepted that the effects of potential environmental stressors on fish can be ephemeral or intermittent; they do not last for a long time and thus they are often difficult to characterize and are unreliable. However, it is also true that careful analysis and high quality data usually reward robust relationships with predictive power, especially if there is a mechanistic basis to support the proposed relationships.
The second family of arguments is more pragmatic, and in my opinion gets to the essence of the problem. Since numerous processes affect the amount of fish in the sea, from changes in ocean currents to fluctuations in fishing effort, it might be almost impossible to identify the mechanisms behind changes in fish abundance. Furthermore, the effect of different processes does not just add up, but instead they interact in involved and tricky ways that result in unexpectedly large responses to small changes in the prevailing conditions (i.e. the total is more than the sum of the parts).
Under these circumstances, it can be argued that empirical approaches that blindly attempt to relate past fluctuations to future abundance become more appropriate . Indeed, it has been proposed that they can outperform traditional approaches that incorporate biological constraints and mechanisms. A more radical interpretation asserts that the dynamics of single species are just unpredictable, thus fisheries managers should forget about single species and start thinking in terms of entire ecosystem properties like the total fish biomass. This way of thinking is behind approaches like balanced harvesting and no-discard fishing (be happy with what you find, and do not alter the structure of the ecosystem to keep it working).
In parallel to these discussions, there is a movement in fisheries science bringing in a new way of thinking. The approach takes advantage of the development of forecasts of ocean conditions based on models similar to those employed to anticipate the impact of climate change, but instead focusing on management-relevant scales (weeks to months, or even years in advance). Environmental forecasts are used to assist managers with improved predictions of fish abundance and distribution. These forecasts can then be used to optimize fishing activities, closing the circle back to our beautiful portrait.
It is a small revolution, but evidence favoring the use of environmental forecasts is steadily accumulating. Positive testimonies can help to improve the management and safeguarding of marine living resources. Just like the brightness of Sorolla paintings lets us think beyond the canvas, predictions based on mechanistic models leading to improved management might be pointing toward a new way for fisheries modeling.
Fernando González Taboada is a fellow of the Nereus Program at the Atmospheric and Ocean Sciences Program, Princeton University
 The original painting by Sorolla is part of the permanent collection of The Prado Museum in Madrid. There is also a wonderful website with tons of information about Joaquín Sorolla and its paintings (there are more about fish).
 The novel «Flor de mayo» by Blasco Ibáñz is also available in English as «The Mayflower».
 FAO cites 24000 causalities per year in their 2016 report about The State of World Fisheries and Aquaculture (SOFIA). In a country like USA, fatal occupational accidents are roughly 30 times more likely in the fishing sector (see CDCP 2010).
 For instance, FAO and the International Labor Organization also warn about the comparatively large incidence of child labor in fisheries (see FAO & ILO 2013).
 See the SOFIA reports cited above.
 See for example Taboada and Anadón 2015.
 Skern-Mauritzen et al 2015 estimated that only 2% of the fish stocks managed worldwide incorporate ecosystem drivers in their assessments of fisheries status and short term production.
 This is at least my lecture of the widely cited review by Myers 1998, which is commonly cited to support the opposite view.
 See for instance the commentary by DeAngelis and Yurek 2015.
 García et al 2012
 The review paper by Tommasi et al 2017 is a very good start for a recent overview of the potential application of seasonal to decadal forecasts in fisheries.
 Cvitanovic and Hobday 2018 stress the importance of «bright spots» to identify the factors behind the success of alternative environmental management practices.