Modeling sardine and anchovy low-frequency variability.

The group of small pelagics (sardines, anchovies, herrings, etc.) is the largest source of wild marine protein worldwide, providing annually between 21% and 43% of the total marine capture production during at least the last 60 y (1). Even when large-scale fisheries have been operating for decades to centuries, their assessment and management remains difficult and highly uncertain (2). Such challenges are mainly because these species do not conform to the traditional assumptions of population dynamic models, such as a steady level of unexploited stock (carrying capacity), or that the influence of climate on the population is negligible when compared to fishing mortality. In other words, it has been clear for decades that the traditional toolbox of fisheries management is clearly unsuitable to these massive resources. The paper by Lindegren et al. (3) in PNAS provides an alternative framework to study and ultimately to develop management-supporting knowledge for these types of resources. The modern era of small pelagic fishes research began exactly three decades ago, when the major sardine and anchovy fisheries catch series from remote systems in the western (Humboldt and California) and eastern (Japan) Pacific were presented together, and synchrony between them was first suggested (4). The sardine-anchovy puzzle was further complicated when alternation between sardine and anchovy stocks within each system and empirical linkages with climate were proposed, conforming to what was called The Regime Problem (5, 6). During these last 30 y, very valuable research has been conducted to document abundance fluctuations, synchrony between regions, and alternation between species observed during the 20th century, and many hypotheses have been proposed to explain the underlying mechanisms (7) or question the existence of such patterns (8). A comprehensive review of the history of scientific research on small pelagic fishes and their variability (9) crudely recognizes that the 20th century ended without a generally accepted theory to explain the nature and mechanisms governing small pelagic fishes fluctuations, but optimistically claims that the science of fisheries oceanography seems to be close to achieving the breakthrough. The paper by Lindegren et al. (3) is a solid step forward because, just as happens when new technical tools and models become available, their framework brings new opportunities to test already existing hypotheses and formulate new ones. Surprisingly, the model by Lindegren et al. (3) was able to hindcast the fluctuations of the California sardine and anchovy populations at two time scales, using independent experiments. On one hand, the model captured the major features of the sardine and anchovy population trends during the last ∼80 y, including the sardine collapse in the 1940s–1950s, its recovery during the 1980s– 1990s, and even the most recent population decrease during the last decade. On the other hand, the model properly reproduces mutidecadal fluctuations (before the industrial fishery) that have also been observed in the rates of sardine and anchovy scales sedimentation in anaerobic sediments (paleoreconstruction using abundance proxys). These robust results provide an exciting opportunity to explore the major paradigms of The Regime Problem: the large fluctuations pattern occurring at the major sardineanchovy systems, the apparent synchrony between them, and the alternation of species within each of them (Fig. 1). An effort toward exporting the model to all major systems may allow testing whether (i) the same basic climate forcing is operating in all systems [i.e., sea surface temperature (SST)], (ii) similar fluctuations occur due to other environmental forcing, such as winds and mesoscale ocean dynamics (10) or oxygen limitations (11), and (iii) they can be synchronized at the global scale. Based on the long-term variability experiments, Lindegren et al. (3) were able to detect 50–100 y dominant periodicities in the simulated fluctuations of sardines and anchovies, which grossly agrees with the observed frequencies from the scales-based paleorecords (7), as well as the dominant frequencies detected in the fisheries period (12). Interestingly, they detect differences in the dominant periods between sardines (80 y) and anchovies (60 y), which in the long term translates to nonsynchronic alternation between species (a perfect alternation would only occur under equal dominant periods). It has already been Fig. 1. Sardine and anchovy catch series from the main fishing regions of the world: California (USA and Mexico, including the Gulf of California), Humboldt (Chile and Peru), Japan (Japan, Korea, China, Russia/USSR), and Benguela (Namibia and South Africa). Data sources include the SCORWG98 (6) covering from 1920 to 1997, and FAO fisheries statistics for the 1950–2011 period (1). Dots indicate catch values reported by SCORWG98 (blue/red) and FAO (green/yellow). Curves are 11–y moving averages for sardine (blue) and anchovy (red) time series. Shaded areas represent periods where sardine (light blue) and anchovy (light red) dominated in most of the Pacific basin fisheries. The SCORWG98 California catch series do not include the Gulf of California.