Generic names and mislabeling conceal high species diversity in global fisheries markets

Overexploitation is one of the principal threats to coral reef diversity, structure, function, and resilience [1, 2]. Although it is generally held that coral reef fisheries are unsustainable [3-5], little is known of the overall scale of exploitation or which reefs are overfished [6]. Here, on the basis of ecological footprints and a review of exploitation status [7, 8], we report widespread unsustainability of island coral reef fisheries. Over half (55%) of the 49 island countries considered are exploiting their coral reef fisheries in an unsustainable way. We estimate that total landings of coral reef fisheries are currently 64% higher than can be sustained. Consequently, the area of coral reef appropriated by fisheries exceeds the available effective area by approximately 75,000 km(2), or 3.7 times the area of Australia's Great Barrier Reef, and an extra 196,000 km(2) of coral reef may be required by 2050 to support the anticipated growth in human populations. The large overall imbalance between current and sustainable catches implies that management methods to reduce social and economic dependence on reef fisheries are essential to prevent the collapse of coral reef ecosystems while sustaining the well-being of burgeoning coastal populations.

COI DNA barcoding is increasingly recognized as a significant new tool for the recognition and identification of animal species. Here, publicly available barcode data are compiled and analysed for birds (657 species) and fishes (1088 species). The proportion of species that cannot be barcode-distinguished by this marker is approximately 6.4% for birds and 2.1-2.5% for fishes. At all hierarchical taxonomic levels (species, genera, family, order, class), fish show greater mean COI divergence than birds. If two samples are barcode-identical, then for both birds and fishes, the probability that they are from the same species is 98-99%. The probability of conspecificity rapidly drops as divergence increases. At 2% COI divergence, this probability approximates to 1% for birds and 3% for fishes. The apparent difference between birds and fishes might partially reflect currently unrecognized cryptic species complexes in the latter. These probability estimates derive from pooled samples of birds and pooled samples of fishes, and will not apply in all situations. Recently evolved species complexes will have higher proportions of species that are barcode-identical. As barcode data accumulate, more refined statistical analyses will become possible. © 2009 Blackwell Publishing Ltd.