No abstract

No abstract

No abstract

Regimes of high abundance of sardine (Sardinops sagax and Sardina pilchardus) have alternated with regimes of high abundance of anchovy (Engraulis spp.) in each of the five regions of the world where these taxa co‐occur and have been extensively fished. When one taxon has been plentiful, the other has usually been at a reduced level of abundance, and vice versa. Changes in the four heavily fished regions that support S. sagax–the Japanese, Californian, Humboldt, and Benguela systems–from a regime dominated by one taxon to a high level of abundance of the other have occurred more or less simultaneously. In the Pacific Ocean, sardines have tended to increase during periods of increasing global air and sea temperatures and anchovies to decrease. The Japanese system is dominated by sardines to a greater extent than the other systems, and sardines off Japan appear to increase as the Kuroshio Current cools. At the eastern edge of the Pacific Ocean, sardines colonize cooler areas during periods of warming. The Benguela system is out of phase with the three Pacific systems. The four systems all appeared to be in a state of flux in the 1980s. Increased abundance of the subdominant taxon is often one of the first signs of change. Sardines are relatively sedentary in refuge areas when scarce but change behavior to become highly migratory and colonize cooler areas when abundant. Anchovies, by contrast, expand around a fixed geographic center.

Climate change and overfishing are increasingly causing unanticipated changes in marine ecosystems (e.g. shifts in species dominance). In order to understand and anticipate these changes, there is a crucial need for indicators that summarise large quantities of information into a few relevant and accessible signals. Seabirds have been suggested as good candidates for ecological indicators of the marine environment; however, few studies have critically evaluated their value as such. We review the role of seabirds as ecological indicators, and discuss their limitations and drawbacks, as compared to other types of indicators. In addition, we highlight the statistical consequences of inverse inference when using seabird data as indicators. We discuss the use of integrated indices and the use of seabirds as autonomous samplers of the marine environment. Finally, we highlight the necessary steps preceding the use of seabirds as indicators. We conclude that, in order to use seabird time series properly, the use of recent advances both in statistics and in remote sensing is a way to move forward. This, along with the assessment of their usefulness, should enable us to use seabird indicators appropriately for managing urgent conservation problems.