EDIT
"The whole picture of fisheries that we have is basically wrong," says lead author Daniel Pauly, of the University of British Columbia in Vancouver. "The underreporting is of such magnitude that it boggles the mind." The results, say Pauly and his colleagues, have major implications not only for how fisheries are managed in these areas, but also for how long local subsistence fishers can continue to support their families.
Fishy statistics
The United Nations' Food and Agriculture Organization (FAO) maintains the only global database of fisheries statistics, which spans the period from 1950 to 2004. Numbers used are voluntarily reported by individual countries and are generally gleaned from fish sales, rather than scientific surveys. In island nations especially, the system overlooks fish caught and consumed by those who catch them because this leaves no economic trail.
To explore the problem, Pauly, along with coauthor Dirk Zeller and others from the University of British Columbia, began by studying the basis for existing FAO statistics to identify gaps. Then, working with local collaborators whenever possible, the team searched for any fishing data they could unearth for a given location from sources not included in conventional scientific literature, such as local surveys of residents, that might fill those gaps. "The more you dig, the more you discover,” says Zeller. Using the historical information they found, making various assumptions and interpolations, and combining this with the FAO statistics, the researchers estimated true catches.
The worst problem was found in American Samoa, where reported catches for the FAO period were 1,525 metric tons, but the authors' reconstruction put at 25,380 tons. Three- and fourfold underestimates were common. The authors believe further analysis will in many cases lead to even higher results. In a few cases catches were overreported, but this was due to countries such as Vanuatu allowing foreign vessels to register there, leading to foreign catches being counted in a country's official tally. The authors recognize that their results include large uncertainties, but say that data reviews like this, although often met with initial scepticism by local managers, have ultimately been considered plausible.
EDIT
http://www.nature.com/news/2008/080709/full/news.2008.942.html