Thursday, October 22, 2009

Case #4 South Georgia Patagonian toothfish longline

aka Chilean sea bass
Certified as sustainable in March 2004.
Recertified on the 17th September 2009.

Thanks to Claire (Comment, Oct 19, 2009) who suggested we look at this fishery.

The Patagonian toothfish lives in deep waters (from 300 to 3,500 metres) on seamounts and continental shelves and around most sub-Antarctic islands. It is a slow-growing, late maturing species and therefore not reslient to fishing pressure. The Patagonian toothfish fisheries off Argentina and around Prince Edward and Marion Islands (under South African jurisdiction) were fished to commercial extinction over short period of time and illegal fishing around Crozet Islands has also substantially reduced that population.

Obviously there would be concerns about the South Georgia fishery as well. A paper in Russian by Shust and Kozlov in 2006 (Вопросы ихтиологии 46(6):791-298) found a decrease in average size and a high percentage catch of immature fish compensating for a reduction in the fraction of large fish over time in the commercial fishery. These could be interpreted as signs for concern.

However, the stock assessments produced by CCAMLR (an RFMO) and used by the consultants for the MSC certification suggest a stock that is being fished down from high unexploited levels towards Bmsy. While not yet at Bmsy, it is projected to reach that level in about 2040 at the current harvest of about 4kt. The last assessment was in 2007 and there will be a new one this year (2009).

After 1985 the longline fishery moved from summer day-setting to winter night-setting to mitigate seabird bycatch. Navy patrols deter IUU and there is strict catch documentation with 100% observer coverage.

The CCAMLR objective for the stock is to maintain spawning biomass above 50% of unfished levels with probability 0.5 and to ensure probability of spawning biomass being below 20% of unfished levels is never greater than 0.1. These conditions appear to be currently met and should continue to be met if the catch continues at the current level of about 4kt.

So, on the face of it, the MSC certification consistent with a sustainble fishery.
Please comment on errors or ommissions – Fishyfellow is not very familiar with this fishery.

Latest assessment from ppt by R. Hillary, C. Edwards, Division of Biology, Imperial College London

Projection from CAMLR assessment in 2007 (broken lines indicate 80% credible limits)


  1. My problem is that they never even discussed the Shust and Kozlov paper, which I know they must have come across in their assessment. To me, that makes them look a bit shady and selective in their analysis - if that paper's not enough evidence of overfishing, fine, but at least they should explain why they think the evidence supports their conclusion and not that of Shust and Kozlov.

    I do wish I knew more about fisheries science and this whole 50% of the unfished biomass thing - I wonder if for a vulnerable, late-maturing fish if it would be more appropriate to aim for a higher percentage of estimated unfished biomass.

  2. The basis for this idea comes from Benny Schaefer, scientist at the Inter-American Tropical Tuna Commission in La Jolla in the 1960s. He developed the Schaefer model for fish populations, which in really is just another expression of logistic population growth, an idea that has been around for a while. It is a modification of the Malthusian exponential population growth model developed by Verhulst in the 1800s for cases where resources (e.g. space or food) become limiting as populations increase in number. This is called density dependence, and typically influences things like birth rate or survival rate in a population.

    If the change in the rate is linearly related to density (number of animals within a finite space) then the logistic curve is symmetrical and the Schaefer production model applies. The idea is that an unfished population is at carrying capacity, the top of the logistic population growth curve, where growth has ceased because births match deaths. At this point there is no surplus yield. However, if you reduce the population size, births will exceed deaths as the population compensates by either increasing its birth rate or decreasing its death rate or both (because we normally work in biomass, change in body growth rate is also a factor that has to be taken into account). If you keep the population at this level by harvesting this excess then you are taking the surplus yield. Under the Schaefer model this surplus yield is maximized at 50% of the original unexploited population size, the so called Maximum Sustainable Yield. At this population size the yield, a function of the numbers of animals in the population multiplied by their per capita production rate (births - deaths) is at a maximum.

    Detractors of this approach will argue that it is a single species view and there is no such thing as a "surplus" in an ecosystem that has come into some kind of stability over millions of years (no free lunch). There are theoretical studies that show that you can't take MSY from a bunch of interacting species (predator-prey and competition interactions). Others will also argue that production depends as much or more on the physical environment as it does on the population. With regard to 50% of the unexploited level being safe, it is will known that you can't actually take MSY when there are random environmental effects - the population will collapse, eventually. Despite these shortcomings, a population that is fluctuating at the biomass level that gives MSY (50% of the unexploited biomass for the Schaefer model), or above, is generally considered to be sustainable. A population that has fallen to less than 10% of its unexploited state is generally considered to have collapsed. For those between 10% and 50% of the unexploited biomass, there can be some debate. Some have suggested that 80% of the biomass at MSY(or 40% of the unexploited biomass ) is the minimum requirement for a sustainable fishery. To complicate things further, for some fish populations maximum yield occurs at less than 50% of the unexploited biomass (i.e. these don’t follow the Schaefer model), allowing an argument to be made for lowering the definition of a sustainable fishery in these cases. These are typically cases where recruitment stays high until a very low biomass is achieved (i.e. populations that are resilient to recruitment overfishing)

    Whether 50% of the unfished level is really a safe level for a fish with the life history of the Patagonian toothfish is an interesting question. Would one want a higher biomass to ensure the viability of this population?