Research ArticleOCEANOGRAPHY

Climate-driven aerobic habitat loss in the California Current System

See allHide authors and affiliations

Science Advances  15 May 2020:
Vol. 6, no. 20, eaay3188
DOI: 10.1126/sciadv.aay3188
  • Fig. 1 Metabolic Index (Φ) distribution and historical variability in the CCS.

    Φ is computed using physiological parameters (hypoxia tolerance Ao = 25 atm−1, temperature sensitivity Eo = 0.4 eV) that are typical of marine species, including those found in the CCS, and for environmental conditions (temperature and O2) from ROMS model simulations. (A) Mean Φ at 200 m depth, and (B) alongshore depth section (200 km from shore; 12 km resolution hindcast). (C) Offshore depth section (from 4 km resolution ROMS hindcast). Distributions of Φ remain qualitatively similar for species with alternate physiological parameters: Values of Φ scale uniformly higher for species with a higher hypoxia tolerance (higher Ao), whereas for species whose hypoxia tolerance is more temperature sensitive (i.e., higher Eo), vertical Φ gradients are weaker and latitudinal Φ gradients are stronger. In the Southern California Bight, CalCOFI monthly Φ anomalies (for Ao = 30 atm−1, Eo = 0.6 eV) are calculated from historical hydrographic observations at (D) 0 to 200 m depth and (E) 200 to 400 m depth. Data are overlaid by 3-year moving averages of Φ (lines) where both temperature and O2 vary as observed, as well as Φ expected from changes in temperature or O2 independently.

  • Fig. 2 Aerobic habitat in the CCS for varied marine ecophysiotypes.

    (A) Time-mean aerobic habitat volume and (B) variability in time of aerobic habitat volume (standard deviation), spatially averaged as a fraction of the total volume over the coastal, epipelagic CCS (27°N to 51°N, 0 to 200 km offshore, 0 to 200 m depth; 4 km resolution ROMS hindcast). The range of hypoxia tolerance, Ao, and temperature sensitivity, Eo, evaluated are based on a global species trait compilation; known species traits are marked with black symbols for specific (sub)phyla (legend). This calculation assumes a ratio of active to resting metabolic rates Φcrit = 3.5, near the inter-specific mean. Native CCS species are marked with open symbols, or, if the species-specific Φcrit is known, with a larger marker and fill color corresponding to the actual aerobic habitat volume in the specified depth range (Dp, Desmophyllum pertusum; Pp, Pandalus platyceros; Sl, Stenobrachius leucopsarus; Tc, Tarletonbeania crenularis). Lower Φcrit results in higher aerobic habitat volume for the same physiological traits.

  • Fig. 3 Anchovy distributions and abundance predicted by the Metabolic Index (Φ).

    (A) Anchovy observations (circles, n = 1917 from 1888 to 2015, 1° bins), including observations not used in the distribution analysis (all south of 32°N). Also, May aerobic habitat thickness (shading), where Φ ≥ Φcrit, for the distribution-derived ecophysiotype (from ROMS 12 km resolution climatology, black line contours 10 m thickness). (B) Frequency of anchovy observations associated with climatological O2 and temperature conditions, for samples paired with observed and modeled hydrographic data (n = 1075, north of 32.63°N only). The resulting temperature-dependent hypoxia threshold for the best-fit ecophysiotype (ecological hypoxia tolerance Aocrit = 5.4 atm−1 or 0.053 kPa−1, temperature sensitivity Eo = 0.4 eV or 6 × 10−20 J; black line) demarcates the space below which anchovy can sustain resting but not active energetic demands. Cross-shore distributions of anchovy observations, aerobic habitat thickness, and net primary productivity (ROMS 12 km output) in the SCB, in (C) spring, peak upwelling season, and (D) fall. Nearshore aerobic habitat thickness is roughly half as thick, and the mean location of observations (dashed black lines) shifts offshore during upwelling.

  • Fig. 4 Anchovy response to historical aerobic habitat variability.

    (A) For the SCB, standardized anomalies (Δ) in time of spawning season anchovy larval abundance are correlated with annual Φ and aerobic habitat volume predicted by the distribution-inferred ecophysiotype (lines between consecutive annual data). (B) Relationship between larval abundance and aerobically suitable habitat volume (fraction of upper 100 m), from the weighted mean response across ecophysiotypes (both normalized to CalCOFI 1956–2011 mean; dashed line is 95% confidence interval for the slope, dotted line is 95% prediction interval for annual values). Median spawning season abundance was 379.1 larvae dm−2 (1956–2011), and dropped to historically low levels (1.9 larvae dm−2; red line) with a 15% decrease in aerobic habitat volume. Decadal means (filled circles) align closely with the slope derived from the interannual data (bars show the interannual range for each decade).

  • Fig. 5 Projected anchovy range contraction due to aerobic habitat loss.

    Aerobic habitat volume is computed at each latitude under historical climate (1995–2010; black line) and future climate (2100; dashed lines), and normalized by the mean value in the SCB. Thresholds for potential extirpation are defined using an ecological criterion (historically low larval abundance in the SCB; flat red line) and a more stringent physiological criterion (zero aerobic habitat in the SCB; flat black line). Latitudinal range limits are denoted by circles at the intersection of these extirpation thresholds with the computed habitat volume in each period. The 95% confidence intervals based on Fig. 4B are marked with gray bars.

  • Fig. 6 Projected aerobic habitat loss by 2100 for varied marine ecophysiotypes.

    Relative habitat volume loss (2100–2000) across physiological traits (for assumed ratio of active to resting metabolic rates Φcrit = 3.5) in the coastal CCS (27°N to 51°N, 0 to 200 m offshore; 12 km resolution ROMS with Representative Concentration Pathway 8.5 climate forcings). Percent habitat loss from (A) 0 to 200 m depth and (B) 200 to 400 m depth. Gray indicates no current aerobic habitat in the depth range. Known species traits are marked with black symbols for specific (sub)phyla (legend). Native CCS species are marked with open symbols, or, if the species-specific Φcrit is known, with a larger marker and fill color corresponding to the species-specific aerobic habitat volume loss in the specified depth range. Lower Φcrit results in higher aerobic habitat volume for the same physiological traits. The inferred anchovy ecophysiotype (assuming Φcrit = 3.5 to translate Aocrit onto plot) is marked with a large filled marker with white outline (Em, Engraulis mordax).

Supplementary Materials

  • Supplementary Materials

    Climate-driven aerobic habitat loss in the California Current System

    Evan M. Howard, Justin L. Penn, Hartmut Frenzel, Brad A. Seibel, Daniele Bianchi, Lionel Renault, Fayçal Kessouri, Martha A. Sutula, James C. McWilliams, Curtis Deutsch

    Download Supplement

    This PDF file includes:

    • Supplementary Text
    • Figs. S1 to S9
    • Table S1
    • References

    Files in this Data Supplement:

Stay Connected to Science Advances

Navigate This Article