Sardine populations, crucial for both commercial fisheries and marine ecosystems, exhibit dramatic boom-and-bust cycles at multidecadal scales. While these fluctuations correlate with basin-scale climate indices like the Pacific Decadal Oscillation (PDO), the underlying mechanisms remain unclear. A striking difference exists in how sardine populations in western and eastern boundary current systems respond to climate change; western North Pacific sardines (JP sardines) thrive during cooler periods, while eastern North Pacific sardines (CA sardines) flourish during warmer periods. This study aims to investigate the life-history traits of these populations and their responses to temperature variability to understand these contrasting responses. Understanding these mechanisms is crucial for predicting population responses to future climate change and for sustainable fisheries management. Each female sardine produces hundreds of thousands of eggs annually, but the vast majority die during larval and juvenile stages. Early life growth rate is strongly correlated with recruitment, particularly during low-biomass periods. Faster growth is hypothesized to increase survival through larger size, shorter vulnerable life stages, and resilience to predation. The relationship between growth, energy acquisition and expenditure, and temperature is key to understanding this, particularly within the context of the varying metabolic responses of marine ectotherms to temperature. The contrasting nursery environments of western (warmer) and eastern (cooler) boundary currents significantly impact fish physiology and food availability, potentially leading to divergent growth responses. However, a lack of techniques to track the early life environments and physiological conditions has previously hindered research in this area.

Numerous studies have identified links between habitat temperature and sardine biomass, recruitment, or early life survival. However, the contrasting responses of JP and CA sardines to temperature have been a long-standing enigma. While some studies have found correlations between temperature and sardine abundance, the mechanisms underlying these correlations remain poorly understood. Previous research has highlighted the importance of early life growth rate in determining year-class strength, suggesting that environmental variability affecting early growth is critical. Existing literature also emphasizes the role of aerobic scope, the capacity for increased metabolic rate above maintenance levels, in influencing growth in marine ectotherms. Aerobic scope is maximized at a specific optimal temperature and declines above or below this point; however, this pattern varies across life stages. The impact of temperature can be both direct, through metabolism, and indirect, via food availability. A previous study on South African sardine populations, spanning both western and eastern boundary systems, showed faster growth in warmer waters on the southeast coast compared to the cooler west coast, suggesting that this pattern might be widespread in subtropical boundary current systems. However, exploring this concept has been hampered by a lack of techniques to track the environment experienced by fish during their early life stages and their associated physiological conditions.

This study employed novel otolith analyses to overcome previous limitations. Sardine otoliths form daily growth increments during larval and juvenile stages, allowing inference of somatic growth trajectories from increment widths. Further, stable isotope ratios (δ¹⁸O and δ¹³C) in otoliths provide estimates of ambient water temperature and field metabolic rate (energy expenditure). High-resolution isotope analyses allow for examination of variability at intervals as small as 10 days. Otolith samples were collected from 420 age-0 JP sardines (Oyashio region, 2006–2010, 2014, 2015) and 136 age-1 CA sardines (Southern California Bight, 1987, 1991–1998, 2005–2007). Daily increment widths, δ¹³C, and δ¹⁸O were measured from hatching to 120–150 days post-hatch (dph), encompassing the larval and most of the juvenile stages. δ¹³C measurements, along with data on δ¹³C of sardine prey and dissolved inorganic carbon (DIC), allowed estimation of M_oto, a proxy for field metabolic rate. δ¹⁸O measurements, seawater δ¹⁸O data, and Argo float data were used to reconstruct individual habitat temperature histories. These data were compared with those of South African sardine subpopulations to infer general features of western and eastern boundary current populations. Interannual variability was assessed by comparing year-class mean experienced water temperatures to the Pacific Decadal Oscillation (PDO) and the Coastal Upwelling Transport Index (CUTI). Statistical analyses included linear mixed-effects models, generalized linear models, and Pearson's correlation tests to assess differences in growth, metabolic rates, optimal temperatures, and relationships between temperature and growth across populations and life stages. Back-calculated length at different life stages was used along with otolith growth as growth proxies. Finally, the relationship between sea surface temperature (SST) and early life survival was assessed using log recruitment residuals from Ricker models.

JP sardines exhibited higher mean otolith radii and M_oto (metabolic proxy) throughout early life stages compared to CA sardines. Growth histories were similar between JP sardines and South African sardines from the southeast coast (western boundary current) and between CA sardines and South African sardines from the west coast (eastern boundary current). JP sardines experienced a gradually declining temperature profile (19–16 °C in 120 days), reflecting northward movement, while CA sardines experienced relatively constant temperatures (15–16 °C). Year-class mean experienced water temperature in JP sardines was negatively correlated with the PDO, and in CA sardines with the CUTI. M_oto generally increased with temperature, but JP sardines consistently showed greater energy expenditure than CA sardines. Optimal temperatures for JP sardines declined with age (>20 °C in larvae to ~15 °C in late juveniles), while those for CA sardines remained at ~14–15 °C. Both populations showed positive correlations between larval growth and temperature; however, in JP sardines, the relationship became dome-shaped in early juveniles and negative in late juveniles, contrasting with the consistently positive relationship observed in CA sardines. Analysis of sea surface temperature (SST) and survival indices revealed generally negative correlations in the western North Pacific and positive correlations in the eastern North Pacific.

The findings demonstrate that the contrasting responses of JP and CA sardines to climate variability are rooted in fundamental differences in their life-history strategies. JP sardines, adapted to a warmer, western boundary current system, exhibit higher metabolic rates and faster growth at cooler temperatures due to wider aerobic scope and expanded access to prey-rich waters, while the CA sardines, in the cooler upwelling system of the east, are adapted to warmer temperatures. The results highlight the importance of considering interactions between optimal temperature and trophodynamic factors in understanding sardine population fluctuations. The study also demonstrates the power of using otolith analysis to understand the early life history of fish, providing invaluable insights into population dynamics and responses to environmental changes.

This study elucidates the contrasting responses of western and eastern North Pacific sardine populations to climate variability. Differences in habitat temperature, growth rates, metabolic rates, and the relationship between temperature and growth, are key factors driving these distinct patterns. These findings underscore the importance of considering species-specific life-history traits and metabolic responses when projecting the impact of climate change on marine populations. Future research could focus on exploring the genetic basis of these differences and examining the broader implications for species interactions and ecosystem dynamics.

The study primarily focuses on two sardine populations, limiting generalizability. The sample sizes for some analyses were relatively small, especially when considering interannual variability. The use of literature values for estimating M_oto could introduce uncertainty, although sensitivity analysis suggests that the results remain robust. Furthermore, there are assumptions made when back calculating length, although analysis with otolith radius was comparable.