July 2, 2013 — Why is the 90% figure still being used? Is it because it is a convenient number for fund-raising campaigns? A convenient truth? I think so. Groups that hang on to simplistic (and erroneous) numbers are doing a disservice. Management needs to be science-based, not faith-based.
Last week, I had the privilege of presiding over the defense of the Ph.D. dissertation of Maria José Juan-Jordá in La Coruña, Spain. Maria José is a bright young researcher and has already published several chapters of her dissertation in the peer-reviewed literature.
The first chapter in her dissertation is the last in a series of peer-reviewed scientific papers that demonstrate that the combined biomass of large oceanic predators (mostly tunas) has not declined by 90% as stated in another scientific paper ten years ago.
The 90% decline figure came from an analysis published in 2003 which concluded that “large predatory fish biomass today is only about 10% of pre-industrial levels” [Myers and Worm, 2003]. Those analyses relied heavily upon catch rates from a single fishing gear type (longline) and aggregated catch across species to estimate trends in “community biomass”. The paper quickly became high-profile in the tuna world. Many environmental groups embraced it as proof that all tunas, not just the bluefins, were in serious trouble. On the other hand, tuna scientists who were actually conducting stock assessments, especially for tropical tunas knew immediately that the 90% number was totally wrong.
Over the next few years, a number of peer reviewed publications showed that the conclusions in Myers and Worm (2003) were fundamentally flawed. Two of the most important reasons for this are: The aggregation of data, and the use of data from a single fishing method. The end result is that the 90% decline is an overestimate. This process of rebuttal is a natural part of the way science develops. Sometimes scientists reach conclusions that are wrong, for whatever reason, and other scientists discover flaws and point to them. A paper, once published, is not necessarily immortal.