The WWF says seafood species critical to human food security will soon be lost unless we halt overfishing and other anthropogenic threats to marine life. The Scombridae family is particularly highlighted as falling by 74% between 1970 and 2012. Overall 1,234 ocean species were assessed in this WWF study and found to have cumulatively declined 49% in that same 42 year period.
The author of this article, Fiona Harvey, touches on a myriad of declining marine species and acknowledges that pollution, plastic detritus, climate change factors and the loss of marine habitat are also effecting these fish populations, not solely overfishing.
But overfishing is the focus of this short article. Louise Heaps, chief advisor on marine policy at WWF UK is frequently quoted and remarks that, “It’s not all doom-and gloom. There are choices we can make. But it is urgent.” She calls for more partnerships between private fishing fleets and governments and advises consumers to, “only eat fish certified as sustainable by the Marine Stewardship Council.”
Comment by Victor Restrepo, International Seafood Sustainability Foundation
A new publication, the Living Blue Planet Report, states that the populations of tunas and their mackerel and bonito relatives (i.e., the family Scombridae) declined 74% between 1970 and 2010. This certainly caught my attention because I am familiar with the stock assessments of tunas conducted by the scientific bodies of the Tuna Regional Fisheries Management Organizations (TRFMOs) and their combined status does not seem to support that magnitude of decline. In a meta-analysis of global population trajectories of tunas and their relatives, Juan-Jordá et al. used published stock assessment results (from the TRFMOs and other agencies) and found that the adult biomass of 26 populations of tunas, mackerels and Spanish mackerels declined by 52% between 1954 and 2006.
The Living Blue Planet Report makes use of the marine Living Planet Index (LPI), which is being used to measure trends in thousands of vertebrate marine, freshwater and terrestrial species. The LPI has a data portal that allows anyone to view or contribute to the time series of data used.
A visit to the LPI data portal reveals that not a single one of the TRFMO estimates of population trends for tunas has been used to construct the LPI. Of the 51 species of scombrids, I could only find stock assessment population estimates for Atlantic mackerel.
The majority of the time series used in the LPI are of two types: (1) The stock size trends in the stock-recruitment database that Ram Myers began to compile in the early 2000s (which is based on stock assessment results), and (2) time series of catch-per-unit-effort (CPUE). Unfortunately, there are problems with the way these two are being used:
- The Myers database time series used in the LPI are not up to date. The last year of data in the series used in the LPI ranges from the late 1980s to the mid-1990s. Most of these series can still be found in the original Ransom Myers’ Stock-Recruitment database website which was last updated in August, 2004. That database is no longer being updated and has been superseded by the RAM Legacy Stock Assessment Database, a compilation of stock assessment results for commercially exploited marine populations from around the world.
- The CPUE time series used in the LPI are, in most cases, “nominal” data. That is, catch divided by some nominal measure of fishing effort, without taking into account any factors that can bias CPUE as an index of relative population size (e.g., vessel effects, spatial effects, etc.). Without some form of appropriate standardization to take such factors into account, raw CPUE cannot be expected to measure abundance. But, even after standardization, using CPUE in isolation to infer relative abundance can be problematic (see for example this paper by Maunder et al.). Furthermore, there does not appear to be a systematic way in which the CPUE series in the LPI were selected, and I do not know if there is a reason for why the (hundreds? thousands?) other CPUE series available in the gray and peer-reviewed literature were not also used.
In short, the majority of the data used as the basis for the Living Blue Planet Report, as far as scombrids are concerned, are either outdated or potentially biased. The LPI would very much benefit from exercising sound quality control so that more up-to-date and relevant data are used. In my view, for scombrids, it should use the abundance estimates from the TRFMOs and the relevant national fisheries management agencies that assess these stocks. Such assessments are generally subject to intense scrutiny and utilize multiple sources of data including biological information, which LPI does not.
Having said that, I don’t know if a 74% decline in all of these scombrid populations between 1970 and 2010 is necessarily bad. That assertion depends on so many things … Their population dynamics are so varied: Some are long-lived, like the bluefin tunas and some short-lived like skipjack and many mackerels; some grow faster, some more slowly; some inhabit neritic ecosystems and some live in oceanic ones; some species predate on others … I personally prefer to look at the status of one stock at a time, and to look at it in relation to some measure of its productivity. For example, this report on the status of the 23 stocks of major commercial tunas presents estimates of current abundance relative to the level that would support Maximum Sustainable Yield (MSY), a common practice in the TRFMOs. And, if I wanted to see their status as a group, I would prefer to see the spread of the results for all populations, instead of a single number.
Victor Restrepo is Chair of the Scientific Advisory Committee of the International Seafood Sustainability Foundation. Contact him here: firstname.lastname@example.org
Comment by Ray Hilborn, University of Washington (@hilbornr)
The Living Blue Planet Report and associated press coverage describing trends in fish stock abundance do a discredit to ZSL and WWF by using data that are biased and inaccurate, and not using anything approximating best scientific information. The results were clearly not reviewed by anyone who knows the specific fisheries well.
Key issues I criticize are
- The trends for the most sensational claim of collapse of scombrids are wrong.
- Many of the declines that have been observed are part of a normal pattern of fishery development that occurred in the 1970s and 1980s where stocks were intentionally fished down to the level that produces long term maximum sustainable yields, and those declines are certainly not a problem. These declines do not constitute a “threat” to the populations.
A comparison of trends:
Using the RAM Legacy data base (www.ramlegacy.org) which contains most scientific assessments of trends in fish abundance, I have calculated the trends in abundance for all stocks where we have data, and one for the tunas and other scombrids. We have 503 exploited fish and invertebrate stocks (almost all of them fish) in our data base, whereas Living Planet Index (LPI) compiled by ZSL and WWF reports 1,463 stocks. We have 58 tunas and scombrids in our data and LPI also reports 58 although there is little overlap in the two data sets.
The biggest difference is clearly in the trends in tuna and scombrids. LPI reports a 70% decline, RAM stocks show a 38% decline. This is a very significant difference and is most likely due to differences in the data used in the analysis.
Looking through the data sets in the LPI it is clear that no care or knowledge was used in assembling the data. In the assessments contained in the Ram Legacy Database extensive effort is made to evaluate the trends in abundance by scientific teams. In contrast, the LPI data sets appear to use any source that appears to provide an index of abundance, with many different indices often for the same stock, and using the totally discredited use of longline CPUE (see Juan-Jordá 2011 and below) appearing frequently. We are concerned that the failure to use the most up to date stock assessments and the lack of scientific peer review of the results has lead WWF and ZSL to publish incorrect results on the decline of scombrids. Given their large platforms and influence within the conservation community, we feel there should have been greater effort to make sure the results were scientifically accurate.
As well as looking at tuna and mackerel, we also used the RAM legacy database to examine decline in all global fish stock. For all fish stocks we estimate a 42% decline between 1970 and 2010, compared to 50% in the LPI.
The data base on actual fish stock abundance in the RAM Legacy Data Base suggest the declines are largely over and stability is the norm. This stability is also found in the LPI index for marine populations as a whole, but their indices for scombrids and all fishes show continued declines.
In general this is good news. The declines between 1970 and 1990 are associated with increases in widescale industrial fishing. For the last 25 years both the RAM database and the LPI show stability in global fish stocks as effective fisheries management came into existence. Due to management policies, declines in major assessed fish stocks have largely halted. Individual stocks such as some scombrid species are still declining and in order to effectively manage them we need the best available science. Misleading results like those published in the Living Blue Planet report are of minimal use if we are to base management and conservation policies on rigorous science.
Is fully exploiting a stock a problem?
The report states ”Today, the world’s fish stocks are under considerable pressure, with 29 per cent classified as overfished and a further 61 per cent as fully exploited, with no ability to produce greater harvests (FAO 2014b). This is a big problem for future global food security.”
Ray Hilborn is a Professor in the School of Aquatic and Fishery Sciences at the University of Washington. Find him on twitter here: @hilbornr
It very common to see overexploited and fully exploited stocks clumped together to show that the majority of fish are in peril Fully exploited means that the stock abundance is in the range where long term sustainable yield is thought to occur. The clear of the text in the Living Blue Planet report is the implication that being “fully exploited” is a problem on par with being overexploited. However being fully exploited is very different than being over exploited. In fact, the objective of most fisheries management agencies is to fully exploit a stock. In other words, to extract as much food from the stock as is sustainable in the long term. In order to feed a growing global population we need to find ways to sustainably harvest fisheries as much as possible, not under harvest them. “No ability to produce greater harvests’ does not necessarily mean that the stock is about to collapse. It means that we are currently extracting as much food from that stock as is sustainable, which is a important step towards global food security.
Comment by Maria José Juan-Jordá, Simon Fraser University, Canada, and Iago Mosqueira, European Commission Joint Research Center, Italy
Taking a better look at the status of tuna stocks
What is the global status of tunas and their relatives the mackerels? How do scientists determine the status of these species? You might be surprised to know that we have a pretty good idea of the global status of tunas, since all major tuna stocks are routinely evaluated with sophisticated methods by their respective Regional Fisheries Management Organizations (RFMOs). These routine evaluations give us an idea of the current exploitation status of stocks and are crucial to set management advice to ensure stocks are maintained at healthy levels of exploitation. In a nutshell, these sophisticated methods, commonly known as fisheries stock assessments, use a great variety of information available over a number of years (commonly since the beginning of industrial fisheries) to reconstruct the trajectory of a population (e.g. how the size, in numbers or biomass, changes over time) and trajectory of fisheries (e.g. how fishing effort and overall fishing mortality changes over time). These assessments also are also key to determine the current exploitation status of a population, whether it is currently being overfished and whether overfishing is occurring. However, the reconstruction of trajectories and status is not an exact science as there is much uncertainty around the estimates, in part due to incomplete and biased information about the population and their fisheries. Typically, the primary data needed to complete an assessment and reconstruct their trajectories involves catches, catch rates or Catch-per-Unit-Effort (CPUE), and fishing effort from all the fleets and nations targeting it, together with basic life history information of the population. Thus, the quality and completeness of information that goes into a fisheries stock assessment is crucial to understand and determine with precision the status of any tuna stock.
So, what is our current knowledge on the global status of tuna stocks? According to the last report of the International Seafood Sustainability Foundation (ISSF), 86% of the world tuna catches comes from healthy tuna stocks. In terms of abundance levels of each individual stock, 52% of the tuna stocks are at healthy level of abundance above optimum levels, 35% are overfished with levels of abundance below optimum levels, and 13% are at an intermediate level. In terms of exploitation, 52% of the stocks are experiencing a low fishing mortality rate, 9% are experiences overfishing with levels of fishing mortality above optimum levels, and 39% have a high fishing mortality that is being managed adequately. This ISSF report summarizes the exploitation status of tunas derived from the most recent RFMO stock assessments by comparing their current levels of abundance and fishing mortality rates to two commonly used biological reference points or benchmarks, BMSY, the biomass that would provide the maximum sustainable yield (MSY), and FMSY, the fishing mortality to obtain MSY. Overall, the current evidence suggest that the majority of tuna stocks are either above MSY levels or have been fished to around MSY levels and therefore are fully exploited globally.
Another way to look at status is by comparing the current abundance levels of tunas relative to a period in the past, for example, when industrial fisheries started in the 50s or 60s. In 2011 a meta-analysis of biomass trajectories derived from fisheries stock assessments estimated tuna stocks have declined in biomass by 49% from 1954 to 2006 globally. When tunas, and their relatives the mackerels, were included in the meta-analysis, total biomass declines were estimated to be around 52% from 1954 to 2006 globally. This meta-analysis also concluded that the majority of tunas and their relatives had been fished down to MSY levels within the last 50 years and therefore in 2011 the majority of stocks were fully exploited globally.
Composite indices such as the Living Planet Index (LPI) can be also very valuable to track changes in the size of a group of populations and understand what is the status of species around the globe. Ideally, the LPI combines time series of population size, density or abundance in order to track changes in the size of a group of species, and when these are not available, it uses the available proxies of abundance. Recently, the LPI was estimated for the tunas and their relatives the mackerels, revealing a global decline of 74% between 1970 and 2010. This is being taken in the social media as evidence of global depletion of tunas and their relatives. As also pointed out by Victor Restrepo above, this assertion depends on so many things such as the population dynamics of the species, their biology and behavior, among many other factors, and not just a bold number.
Now, you might be wondering, why the LPI portrays such large global declines not captured in the other two global studies? A closer examination of the LPI for tunas and their relatives reveals this index is mostly based on proxies of abundance, such as time series of CPUE, and hence, not based on the most recent abundance estimates derived from stock assessments conducted by tuna RFMOs. Below we illustrate two examples of two populations, the Indian Ocean yellowfin and bigeye tunas, and what data was used to estimate the LPI. These examples show how CPUEs for these two populations have been selected unmethodically from potentially 10s or 100s of existing CPUEs available in the literature, instead of using abundance estimates derived from RFMO assessments.
Not only were CPUE selected unmethodically, it is also well known that CPUEs are notoriously problematic when used as a proxy of abundance since they do not reflect changes in abundance. Therefore, their use as the main source of information to calculate the LPI may explain such big declines compared with the most current evidence of the global status of tuna stocks. While we do not argue that composite indices such as the LPI are valuable to track changes in the status for groups of species globally, we would argue that it is extremely important to estimate such indices using the best data available at hand to provide the most accurate picture of global status.
Ideally, what type of information should be used to track changes on the global status of fish populations? In the case of tunas and their relatives, which are a highly commercial group of species and one of the few taxonomic groups in the world that are assessed routinely with sophisticated methods, biomass estimates derived from the RFMOs fisheries stock assessments should be the preferred data used to characterize their status and track changes in population size over time. Only when biomass estimates are not available, proxies of abundance such as CPUEs or catches should be carefully used to infer status.
To conclude, global attempts to track the impacts of fishing on marine species and ecosystems with composite indices such as the LPI need to be careful about selecting the information available to infer status, since their credibility might be at stake. We believe the LPI would benefit from the peer-review process by experts working with the species being analyzed. Furthermore, we also fervently emphasize the need for more cross pollination among fields, increase accessibility and sharing of datasets critical to understand status of species, and increase the visibility of the work done by RFMOs in charge of managing tunas, in support of global efforts to track the global health of our species and oceans and conserve them for future generation to come.