|Edition description:||2nd ed.|
|Product dimensions:||8.50(w) x 10.90(h) x 0.90(d)|
About the Author
Read an Excerpt
Global Atlas of Marine Fisheries
A Critical Appraisal of Catches and Ecosystem Impacts
By Daniel Pauly, Dirk Zeller
ISLAND PRESSCopyright © 2016 Daniel Pauly and Dirk Zeller
All rights reserved.
ON THE IMPORTANCE OF FISHERIES CATCHES, WITH A RATIONALE FOR THEIR RECONSTRUCTION
Sea Around Us, Fisheries Centre, University of British Columbia, Vancouver, BC, Canada
Fishing must generate a catch, whether it is done by West African artisanal fishers supplying a teeming rural market, by the huge trawler fleets in Alaska that supply international seafood markets, by women gleaning on a reef flat in the Philippines to feed their families, or by an Australian angler bragging about it in a bar. Indeed, a fishery is defined by the amount and kind of fish caught and by their monetary value. This is how we judge a fishery's importance, compared with other fisheries and other sectors of the economy. It seems clear that the health of a fishery should by measured by changes in the magnitude and species composition of catches, along with other information, such as the growth and mortality of the fish that are exploited. Yet a debate has been recently raging about whether to use catch data to infer the status of fisheries, causing great confusion among fisheries scientists and managers. If the muddle continues, it could undermine the credibility of fisheries science.
The key role of catch data is the reason why the Food and Agriculture Organization (FAO) first began compiling fishery statistics soon after the agency was founded in 1945. A part of the United Nations' attempt to "quantify the world" (Ward 2004), these compendia turned, in 1950, into the much appreciated FAO Yearbook of Fisheries Catch and Landings. The findings are based on annual data submissions by FAO member countries, vetted and harmonized by its staff. In contrast with the many international databases used to track major food crops (e.g., rice, wheat, maize), the Yearbook has been, to this day, the only global database of wild-caught fish and other marine species. As such, the Yearbook is widely cited as the major source for inferences on the status of fisheries in the world (Garibaldi 2012).
However, in many countries, particularly in the developing world, the government's role in monitoring their fisheries seems to end with the annual ritual of filling in catch report forms and sending them to FAO, as parodied in Marriott (1984). For others, mainly developed countries, collecting catch data from fishing ports and markets is only a start, and the bulk of their fishery-related research is in the form of "stock assessments." This term refers to a series of analytic procedures using a variety of data, often time series of commercial catch (figure 1.1), complemented by information on the age, size, or structure of the fish in that catch, tag and recapture data, stock abundances deduced through mathematical models or by fishery research vessels, and so on. The purpose is to infer the biomass of the populations or stocks that are being exploited and to propose levels of total allowable catch (TAC, or quota).
However, traditional stock assessments are extremely expensive, ranging from roughly US$50,000 per stock (assuming 6 months for experts to analyze existing data) to millions of dollars when fisheries-independent data are required (Pauly 2013). Along with a worldwide scarcity of expertise, this is why 20% at most of the more than 200 current maritime countries and associated island territories perform regular stock assessments. Moreover, these assessments deal only with the most abundant or most valuable species exploited. For some countries or territories, this may be one species, a dozen, or about two hundred, as in the United States. In all cases, this is only a small fraction of the number of species that are exploited, if only as unintended by catch, which is often discarded.
Therefore, FAO always encouraged the development of methods that would allow scientists to infer the state of fisheries without stock assessments, or with limited ones (see Gulland 1969, 1971, 1983). This practice was driven by FAO's mission to inform policy makers about the state of fisheries in all countries of the world, including those without access to stock assessment expertise and the costly research vessels needed to collect fisheries-independent data.
To this end, FAO developed what are now called stock-status plots (SSPs; figure 1.2), which showed the status of the various fish stocks over time (Grainger and Garcia 1996). The status of each fishery was inferred from the shape of its catch time series. Essentially, increasing or stable catches meant fisheries were okay, and declining catches meant fisheries were in trouble. These SSPs or equivalent graphs were interpreted vertically, by comparing the percentage of stocks in a given state (e.g., "developing," "developed," "fully exploited," "overfished") in different years. The information was reported in press releases and in issues of the State of the World Fisheries and Aquaculture (SOFIA), a biannual narrative interpretation of the FAO fishery statistics. In successive SOFIAs (including the last available; see figure 13 in FAO 2014, p. 37), FAO notes that these percentages tend to get worse but does not analyze the SSPs further. However, much of what people throughout the world think they know about global fisheries originates from SSPs and similar approaches.
SSPs were also adopted and modified by researchers outside FAO, first by Rainer Froese of the GEOMAR Institute in Kiel, Germany (Froese and Kesner-Reyes 2002) and later by the group of which I am principal investigator, the Sea Around Us, whose work is featured in this volume. Jointly, we demonstrated that an increased number of the stocks had "collapsed," meaning that catches were less than 10% of their historic maximum. Moreover, the transition from one state (e.g., "fully exploited") to another (e.g., "overexploited") was occurring at a faster rate than previously thought. These were dramatic findings, yet they generated little press and even less action.
The world finally started paying attention to the SSP findings in 2006, when Boris Worm and his colleagues published "Impacts of Biodiversity Loss on Ocean Ecosystem Services" in Science magazine (Worm et al. 2006). For the first time, the authors used these trends to project a date by which all stocks would "collapse": 2048. The expert press release that accompanied the article (see Baron 2010) focused on this newsy aspect of what was a broad study, triggering an enormous amount of press coverage on all continents. The headlines were uniformly alarmist: "Fisheries collapse by 2048" (The Economist), "Seafood may be gone by 2048" (National Geographic), and "The end of fish, in one chart" (The Washington Post), among many more.
A strong pushback emerged, including wide and understandable criticism of the precise date, 2048, which mingled Mayan (2012) and Orwellian (1984) undertones. Stock assessment experts mocked the projection, which was mistaken for a prediction, with many arguing that scientists shouldn't extrapolate beyond the data. Yet good science always implies some inference beyond one's data; otherwise, it would consist only of descriptions. Moreover, most critiques overlooked the fact that "collapsed" stocks can continue to be fished. Indeed, this is what already occurs in vast areas of the ocean. Two notorious examples are the Swedish west coast, where a long-collapsed Atlantic cod stock continues to be exploited (Sterner and Svedäng 2005), and the Gulf of Thailand, where the demersal fish biomass was reduced in the 1990s to less than 10% of its value in the early 1960s, when trawling began, yet also continues to be exploited (Pauly and Chuenpagdee 2003). This is 2048 — now.
Still, the criticism was so strong that several co-authors of the study opted not to defend it publicly. Consequently, fisheries scientists such as me, who are concerned with the state of global fisheries, had to either duck or defend the spirit of the 2048 projection, even if we did not agree with all its particularities.
To its credit, the projection was based on catch time series from virtually the entire world. The overwhelming majority showed that peak catches occurred several decades ago, with current catches increasingly derived from "overexploited" and "collapsed" stocks (figures 1.2 and 1.3). Although there is no way to predict where anything will be in 2048 or even 10 years from now, it would certainly be better if we could reverse current trends. So far we have not done so, even though some stocks are rebuilding (figures 1.3 and 1.4).
Before this defense could be mounted, the detractors began focusing on another criticism of the 2048 projection, claiming that catch data do not contain any information about stock status. In interviews, keynote lectures, and other outlets, they argued that full-fledged stock assessments are essential to understanding fisheries; without them, we are essentially left in the dark.
This is a case of allowing the perfect to become the enemy of the good. Even without perfect data, we can infer when fisheries are in serious trouble and make efforts to conserve them. After all, maintaining catches is the raison d'être of fisheries science. One can and should infer, at least tentatively, the status of fisheries from the catch data — if this is all we have (see figure 1.1; Froese et al. 2012, 2013; Kleisner et al. 2013). It is a mistake to assume that we must remain in Muggle-like ignorance unless we have access to the magic of stock assessments.
Accepting this doctrine would put us at the mercy of stock assessment models that can be fatally flawed. For example, the models used to study the Canadian northern cod fishery in the 1990s (Walters and Maguire 1996) were considered the best in the world. In fact, experts thought the models were so good that it was not necessary to consider the catch data from the coastal trap fisheries, which could not, like the trawlers, follow the cod to where they retreated as their numbers declined. Thus, the stock assessment experts were as surprised as the general public when the fishery had to be closed. The trawlers had decimated the stock under their noses, which they could have seen if they had analyzed the coastal trap data. Note that it is not even faulty stock assessments that are at issue here; it is the notion that one type of approach is so good that it makes all other approaches superfluous.
More importantly, this doctrine would discourage efforts to improve the quality of fisheries statistics worldwide, which is bemoaned by FAO in successive issues of SOFIA. It would also thwart attempts to manage, to the extent possible, the fisheries of developing countries. If leading fisheries scientists claim that catch data are useless, why would resource-starved governments invest in reforming and improving their statistical systems?
This flawed thinking would affect not only developing countries but also the community of stock assessment experts themselves. Without the collection of catch data, experts could end up either with beautiful stock assessment models applied to lousy data, as in the northern cod example above, or needing more of the costly fishery-independent data that can be used to correct for misreported commercial catch data (Beare et al. 2005).
We gain nothing from the notion that only a select group has the key to understanding fisheries, especially if that key cannot open any doors outside a small number of developed countries. Such claims undermine the credibility of the many fisheries scientists throughout the world who attempt to extract actionable insights from sparse data and to advise their governments on how to manage their fisheries even if they cannot afford formal stock assessments.
Fortunately, there is a solution: We all agree that many stocks need to be rebuilt and that doing so would lead to sustainable increases of catches and economic benefits (Sumaila et al. 2012). In fact, the more depleted the stocks currently are, the more is to be gained by rebuilding them.
Moreover, our systematic reevaluation of the FAO statistics suggests that developed countries tend to underreport their catches by about 30%–50% (Zeller et al. 2011), and many developing countries underreport by 100%–500% (Cisneros-Montemayor et al. 2013; Pauly and Zeller 2014; Zeller et al. 2007, 2015). (One notable exception is China, which overreports its catches because officials are rewarded for high yields.) This new perspective suggests that fisheries play a far more important role in the rural economy of developing countries than previously assumed and that rebuilding depleted fish populations on a grand scale would have greater benefits than so far imagined (other implications are presented in chapter 14).
Consequently, more attention should be given to the reliable collection of catch data throughout the world. In particular, we need to devise cost-effective systems to acquire accurate fisheries catch data, along with ancillary data on fishing effort, and its economic equivalent, catch value and fishing cost.
These ideas have been apparent to me since my first field experience in Ghana in 1971 and in Indonesia in 1974 and 1976. They were reinforced in 1979 when J. A. Gulland, a world-renowned scientist and senior staff member at FAO, commented that fisheries experts should emphasize three things: "the catch, the catch and the catch." Yet often catch data seem to be entirely missing from certain areas of countries or territories, particularly for informal, small-scale fisheries. In such cases, catch statistics can be reconstructed from other data.
The text below, slightly modified from an article I wrote in 1998, provides the rationale for such reconstructions. It was inspired by discussions that took place at a conference held by the Fish Base Project in Trinidad in May and June 1998.
THE CATCH IN USING CATCH STATISTICS
It is widely recognized that catch statistics are crucial to fisheries management. However, the catch statistics routinely collected and published in most countries are deficient in numerous ways. This is particularly true of the national data summary sent annually by the statistical offices of various Caribbean and Pacific countries to the FAO for inclusion in their global statistics database (see Marriott 1984).
A common response to this situation has been to set up intensive but short-term projects devoted to improving national data reporting systems. Their key products are detailed statistics covering the (few) years of the project. However, without statistics from previous periods, these data are hard to interpret. This is a major drawback, because it is the changes in a dataset that demonstrate important trends.
Therefore, reconstructing past catches and catch compositions is a fundamental task for fisheries scientists and officers. In fact, it is necessary to fully interpret the data collected from current projects. For example, suppose that the fisheries department of Country A establishes, after a large and costly sampling project, that its reef fishery generated catches of 5 and 4 t/km/year for the years 1995 and 1996, respectively. The question now is, are these catch figures low values relative to the potential of the resource, thus allowing an intensification of the fishery, or high unsustainable values, indicative of an excessive level of effort?
Clearly, one approach would be to compare these figures with those of adjacent Countries B and C. However, these countries may lack precise statistics or have fisheries that use different gears. Furthermore, Country A's minister in charge of fisheries may be hesitant to accept conclusions based on comparative studies and may require local evidence before making important decisions affecting her country's fisheries. One approach to deal with this very legitimate requirement is to reconstruct and analyze time series, covering the years preceding the recent period for which detailed data are available and going as far back in time as possible (e.g., to the year 1950, when the aforementioned annual FAO statistics begin). Such data make it possible to quickly evaluate the status of fisheries and their supporting resources and to evaluate whether further increases in effort will be counterproductive (box 1.1).
Excerpted from Global Atlas of Marine Fisheries by Daniel Pauly, Dirk Zeller. Copyright © 2016 Daniel Pauly and Dirk Zeller. Excerpted by permission of ISLAND PRESS.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.