Supplementary Materials for - science.sciencemag.org · Materials and Methods Data collection Many reviews and analyses have examined the temperature tolerance of marine and freshwater

science.sciencemag.org/content/369/6499/65/suppl/DC1

Supplementary Materials for

Thermal bottlenecks in the life cycle define climate vulnerability of fish

Flemming T. Dahlke*, Sylke Wohlrab, Martin Butzin, Hans-Otto Pörtner* *Corresponding author. Email: [email protected] (F.T.D.); [email protected] (H.-O.P.)

Published 3 July 2020, Science 369, 65 (2020) DOI: 10.1126/science.aaz3658

This PDF file includes: Materials and Methods Figs. S1 to S10 Tables S1 to S6 References

1

Materials and Methods Data collection

Many reviews and analyses have examined the temperature tolerance of marine and freshwater fish species (20, 31, 38, 50-55), as well as the temperature dependence (thermal responsiveness) of development rates (31, 51, 56) and oxygen consumption rates (8, 57-59), mainly by focusing on a particular life stage. Here we merged and updated these existing data compilations with peer reviewed results (published until March 2020) to explore the degree of ontogenetic variation in thermal tolerance and thermal responsiveness.

Thermal tolerance metrics

Thermal tolerance data of spawners, embryos, larvae and adults were organized into four temperature threshold metrics: Upper temperature limits (Tmax), lower temperature limits (Tmin), tolerance ranges (Trange = Tmax – Tmin), and range midpoints (Tmid = (Tmax + Tmin) /2). Tolerance metrics of spawners reflect the temperatures at which ripe males and females (advanced stage of gonad maturation) have been observed in the field. In some cases (n = 91), only a preferred spawning temperature was reported, which was regarded as Tmid. To assess the extent to which in situ (behavioral) tolerance data differ from temperatures that exert negative effects on gonad maturation and/or subsequent offspring survival (i.e., physiological limits), we compared in situ and experimental observations of spawner’s Tmax, including 15 species from sub-arctic, temperate and tropical regions (fig. S1). This dataset indicates that in situ behavioral estimates of spawner’s Tmax are slightly lower (Δ0.35°C) than experimental estimates, although the difference was statistically not significant (Two-sided paired t-test, P = 0.189, n = 15, fig S1, table S1). Tolerance metrics of embryos are based on laboratory incubation experiments with at least stable temperature treatments. High or low temperatures reported to cause a statistically significant change in survival relative to the optimum value were defined as Tmax or Tmin. If no significant change was observed over the applied temperature range, only the range midpoint (Tmid) was noted (n = 116). For larvae and adults, Tmax and Tmin are based on studies using established lethality tests (60), where organisms are exposed to either a dynamic change in temperature (constant rate) until loss of equilibrium (CTmin/max), or a static holding temperature treatment until mortality occurs (e.g. 50% or 100% mortality, LT50/100). The influence of acclimation temperature on larval and adult thermal tolerance was accounted for by considering the value determined under the highest/longest acclimation treatment (60). We excluded experimental studies using farmed fish or insufficient replication (n < 3 groups for each temperature treatment). For each observation, the sampling site (latitude/longitude) was noted. If no geo-information was reported, the longitudinal/latitudinal midpoint and the habitat range were extracted from FishBase (www.fishbase.org) or Aquamaps (www.aquamaps.org).

Phylogenetic data imputation

Phylogenetic data imputation is based on ancestral state reconstruction and missing data estimations by a likelihood-based phylogenetic imputation approach (61, 62). This approach showed high reliability for trait data estimation (63) and has been previously used to e.g. estimate climatic vulnerability of freshwater and marine fishes (20) or to assess climate-change impacts on interacting species (64). Although this approach also allows the estimation of traits for species for which no other information than their phylogenetic placement is known, we have restricted our imputation approach to species for which at least one trait parameter has been

2

experimentally validated. In addition, prior to the imputation procedure, we tested for the presence of a significant phylogenetic signal for each parameter in our experimental trait data set (table S2). Data imputation was performed for Tmax and Tmin values for each life stage (spawners, embryos, larvae and adults). Our imputation input matrix contained a total of 694 species, for which an average of 2.5 trait values per species was derived experimentally. We determined the best-fitting likelihood model of evolution based on the Akaike and Bayesian Information Criterion (AIC and BIC, respectively) of a Brownian motion, Pagel’s, Ornstein-Uhlenbeck, Early-Burst and a star phylogeny model (table S6). Models were based on the most comprehensive time-calibrated phylogeny available for fishes (24), retrieved with the aid of the R package ‘fishtree’ (65). Based on the AIC and BIC, we used the Pagel’s λ model to impute missing data and ancestral states in our trait matrix. Tests for phylogenetic signals and data imputation were performed with the R packages ‘phylosignal’, ‘phylobase’ and ‘Rphylopars’ (66) in R version (version 3.6.2, https://www.R-project.org/). We used the R packages ‘ggtree’, ‘ggplot2’ and ‘viridis’ (67) for graphical representation of the imputed and experimental trait data.

We compiled a final dataset of thermal tolerance measures for Tmin and Tmax from which we also derived Trange (Tmax − Tmin). The final dataset contained imputed and experimental data for four life stages for 694 fish species, and additional experimental data for each life stage for species not represented on the time resolved phylogeny (table S2). Differences in thermal tolerance metrics between life stages and associated significance levels were estimated with Generalized Additive Models (GAMs) with the R package ‘mgcv’ (68). Due to the strong influence of geographical location on thermal tolerance metrics, means of absolute latitude values per species were used as smooth term in the GAM models. Means of absolute latitude values per species were derived from sampling locations or latitude midpoints from original publications if present, otherwise from FishBase (www.fishbase.de) or Aquamaps (www.aquamaps.org). We accounted for uncertainty due to the imputation procedure in our dataset by including weights in our GAMs for each thermal tolerance value as inverse of its variance estimated during the imputation procedure. Weights for Trange were calculated from the sum of variances from Tmin and Tmax. For experimental data, weights were set equal to the highest weights obtained from the imputed data. In addition, we tested if experimental data alone would have led to contrasting results (i.e. no significant effect for life stage specific differences in thermal tolerance metrics), but this was never the case. Our final GAMs allowed for life stage specific smooth terms across absolute latitudes for each thermal tolerance measure, and accounted for fixed differences among life stages (spawner, embryo, larvae, adult) and realm (marine, freshwater). Means, standard errors and confidence intervals for each combination of life stage and habitat, as well as post-hoc comparisons for life stages were extracted from model outputs (table S3) with the aid of pre-defined contrast matrices and the R package ‘multcomp’ (69). Differences between life stages were called significant if adjusted P-values from pairwise comparisons were P < 0.05 (see table S4 for model outputs and summaries). For graphical purposes only, GAMs without fixed effects for habitat were used to predict mean values across life stage and latitude, including confidence bands. Predicted values are plotted in fig. S3.

Thermal responsiveness

In some cases, the authors provided only the regression equation of the development or oxygen consumption rate-temperature relationships, which was then used to calculate the specific rates for each temperature treatment. Embryo development days (D) from fertilization to

3

hatching were transformed into developmental rate (1/D). Measurements of MO2 using anesthetics were excluded. Activation energies (Ea) of monotonic, exponential responses were estimated based on the Arrhenius-Boltzmann model (1) by fitting a linear ordinary least square (OLS) regression of the log-transformed trait values on the reciprocal absolute temperature (28). Some MO2 data sets (<10%) followed a unimodal response shape due to falling respiration rates at upper terminal temperatures, indicating thermal constraints on aerobic metabolism. In these cases, extreme temperature treatments were removed until monotonicity was observed. All developmental responses of embryos to temperature were monotonic. In cases where the authors reported increased mortality at lower or upper extremes (despite monotonicity), these temperature treatments were excluded. Responses with r2> 0.5 and P < 0.05 were considered as sufficiently well fitted (15). All measurements of embryonic MO2 (and development) are based on acclimated animals (incubation temperature = assay temperature), while assessments of larval and adult MO2 employed variable acclimation procedures. However, the effect of acclimation time (more versus less than 5 days) on thermal responsiveness was not significant (fig. S10). The hypothesis of no systematic difference between Ea of development rate and MO2 was confirmed by paired observations in six experiments with five species (Two-sided paired t-test: P = 0.187, fig. S4). Statistical analyses of Ea data are based on species-specific means if multiple observations were available. Stage-specific differences in Ea were assessed using a linear mixed-effect models [LMMs ‘lme4’ (69)]. In this analysis, the "life stage" was treated as a fixed factor, and to account for phylogenetic non-independence, "order", "family" and "genus" were included as nested random effects. The hypothesis that stenothermal organisms with narrow temperature ranges are thermally more responsive than eurythermal ones was explored based on the correlation between Ea and Trange. In cases (n = 39 out 151) where no empirical or imputed Trange data was not available, habitat temperature ranges from original publications or from FishBase (www.fishbase.org) and Aquamaps (www.aquamaps.org) were used instead. The correlation between Ea and Trange was assessed using a Generalized Additive Mixed-effect Model [GAMM in ‘mgcv’ (68)], with "order", "family" and "genus" included as nested random effects.

Thermal safety margins

Thermal safety margins (TSMs) were calculated for spawners, embryos and adults as the difference between species- and life stage-specific upper temperature limits (empirical and imputed Tmax) and seasonal habitat temperatures (Thab) within the native distributional range: (1) TSM = Tmax – Thab Due to the low density of empirical tolerance data for larval stages and therefore relatively high uncertainty around imputed tolerance data (table S3), larvae were excluded from the TSM analysis.

Spatial habitat maps were collected for a total of 630 species for which we had either experimental or imputed stage-specific Tmax data. The spatial habitat data are based on information provided by Aquamaps (www.aquamaps.org) and the International Union for Conservation of Nature’s Red List of Threatened Species IUCN (www.iucnredlist.org). We retrieved spatial data for 287 marine fish species in NetCDF format from Aquamaps published on THREDDS (72) (downloaded on 24.02.2020) with the R package ‘thredds’ (Bocinsky 2020, https://rdrr.io/github/bocinsky/thredds/f/README.md) and another 51 NetCDF files manually from the Aquamaps (downloaded 24.02.2020). All Aquamaps NetCDF files included layer information about the occurrence probability (OP) for each fish species and we always considered the area with OP ≥ 80%. Aquamaps was used as primary source for spatial habitat

4

information. For species not represented or accessible at Aquamaps, spatial habitat data were retrieved from IUCN (downloaded on 24.02.2020) as ESRI shape files. With this approach, we gathered additional spatial habitat data for 272 freshwater fish species, and another 20 marine fish species. We used the r package ‘letsR’ (72) to subset all IUCN ESRI shape files so that they contain only spatial information for native habitats and extant species occurrences. Afterwards, shape files were gridded on a mesh with 1°×1° resolution and converted into NetCDF files with the R packages ‘sp’ (73), ‘rgdal’ (Bivand et al. 2019, r-project. org/web/packages/rgdal/index. Html), ‘raster’ (Hijmans 2020, https://CRAN.R-project.org/package=raster) and ‘ncdf4’ (Pierce 2019, https://CRAN.R-project.org/package=ncdf4). Since the IUCN shape files, in contrast to the spatial habitat data from Aquamaps, do not contain specific occurrence probabilities, probabilities were set equal to 100% for the subsequent analyses.

Habitat temperatures were derived from climatological temperatures averaged over the habitat area relevant for the respective life stage. While adult TSMs consider the entire distributional area of a species (estimated as described above), TSMs of spawners and embryos are based on a subset of that area, i.e. potential spawning habitat area. Potential spawning habitats were constrained based on the species-specific thermal tolerance range (Trange of spawners and embryos), thereby excluding areas where reproduction is currently impossible. We also considered information on spawning seasons as well as species- and stage-specific depth preferences provide by FishBase (www.fishbase.org) and the literature (26). Deep sea species with preferred water depths greater than 500 m were excluded. For all TSMs, we considered 20-year averages of monthly air and seawater temperatures for the periods 1981–2000, 2021–2040, 2041–2060, and 2081–2100. Recent (1981–2000) values originate from atmospheric and ocean (74, 75) reanalyzes. Future temperatures were derived from climate simulations produced during Phase 6 of the World Climate Research Programme Coupled Model Intercomparison Project (CMIP6) (76, 77). For our analysis we selected all models which were forced with greenhouse gas concentrations according to Shared Socioeconomic Pathway (SSP) scenarios SSP 1-1.9, SSP 4-6.0, and SSP 5-8.5 (23). The models are CanESM5 (ensemble member r1p1f1) (78-81), CNRM-ESM2-1 (ensemble member r1p1f2) (82-85), IPSL-CM6A-LR (ensemble member r1p1f1) (86-89), MIROC6 (ensemble member r1p1f1) (90-94), and MRI-ESM2 (ensemble member r1p1f1) (95-101). To account for potential model biases, we diagnosed for each model the differences between simulated and observed temperatures for 1981–2000 and subtracted these anomalies from the results of the SSP simulations. We then considered the ensemble median as well as upper and lower bounds of the simulation results. All involved datasets (i.e. species distribution maps, reanalyzed and simulated temperatures) where remapped to the same horizontal resolution of 1°×1°. Marine temperatures were averaged over the depth ranges 0–50 m, 50–200 m and 200–500 m. Temperatures of freshwater habitats were estimated from surface air temperatures (Tair) applying the global regression model by Punzet et al. (102): (2) Thab = C0 / [1 + exp(C1 Tair + C2)], where C0 = 32.0 °C, C1 = -0.13 / °C, and C2 = 1.94 °C. We neglected regional variations of the coefficients as the effect is small in the typical temperature range of most climate zones. Equation (2) is probably inaccurate at specific sites, but we were not able to quantify its uncertainties. Statistical comparison of stage-specific TMSs followed the same procedure (generalized additive models) as for thermal tolerance metrics described above. Test results are summarized in table S5.

5

fig. S1. Upper temperature thresholds estimates (Tmax) of spawners (n = 15 species, see table S1) based on field observations (y-axes, behavioral threshold) and rearing experiments (x-axes). Field-based Tmax values are the maximum temperatures at which spawning and/or the presence of males/females at the final gonad maturation stage has been observed (during spawning or spawning migration). Experimental estimates are sublethal temperatures at which negative effects on gonad development and/or subsequent offspring viability have been detected. The standardized major axes regression (red line) for the correlation between behavioral and experimental Tmax values indicates that the slope is not different to 1 (P = 0.785) and the intercept is not different to 0 (P = 0.776), which means that the behavioral Tmax of spawners is not significantly lower (Δ0.35°C) than the experimental Tmax (Two-sided paired t-test: t = 1.382, P = 0.189, n = 15).

6

fig. S2. Pairwise correlations of thermal tolerance metrics indicate consistency between experimental and imputed data. Blue dots represent experimental temperatures, green dots represent imputed temperatures. Correlation coefficients are marked with stars according to their significance level, i.e. *** P < 0.001. Pairwise correlations, linear regression fits and significance levels were computed and plotted with the R package ‘psych’ (Revelle 2019, https://CRAN.R-project.org/package=psych).

7

fig. S3. (A−C) Colored symbols are means (±95% confidence intervals, 95% CIs) of (A) upper tolerance limits (Tmax), (B) lower tolerance limits (Tmin) and (C) tolerance ranges (Trange) of spawners (black) embryos (blue), larvae (orange) and adults (red) of marine and freshwater species (circles, left and squares, right, respectively). Differences between marine and freshwater species are significant for each tolerance metric and life stage (Generalized Additive Model, GAM: P < 0.0001, n = 698 − 735 for each life stage).

8

fig. S4. Boxplots (boxes: 25th to 75th percentiles, lines: median, whiskers: 10th and 90th percentiles) show activation energies (Ea) of embryonic oxygen consumption rates (MO2) and embryonic development rates estimated in the same species (circles linked by lines; (1) Boreogadus saida, (2) Gadus morhua, (3) Coregonus clupeaformis, (4) Gadus morhua, (5) Oncorhynchus mykiss, (6) Clupea harengus) and the same experiment (29, 103-106), and all embryonic Ea data (n = 83) presented in the main text. The difference between the mean activation energy of MO2 and development is not significant (two-sided paired t-test: t = 1.526, P = 0.187, n = 6).

9

fig. S5. (A) The correlation between thermal responsiveness (activation energy, Ea in eV, data from Fig. 3C of the main text) and thermal tolerance range (Trange) is significant across embryos and adults (GAMM: P < 0.001), but not across larvae (P = 0.712). (B) Arrhenius plot of oxygen consumption rates (MO2) indicating stage-specific responsiveness (slope = Ea, color code as in A) in Atlantic cod (Gadus morhua). MO2 data (colored circles) are based on experiments with different populations (embryos: n = 2, larvae: n = 4, adults: n = 2) (29, 58, 59, 103). Stippled color-coded lines denote the corresponding stage-specific Trange, demonstrating an inverse relationship between Ea and Trange at the species-level.

10

fig. S6. Potential ecological consequences associated with different levels of embryonic thermal responsiveness in Atlantic cod. Solid lines indicate the change in development time (fertilization to hatching) as a function of temperature considering a theoretical Ea of 0.65 eV (14) (green line) and the empirical Ea of 0.92 eV (blue line) for Atlantic cod embryos in the Barents Sea (29). The intercept of the theoretical response was adjusted so that both curves have the same intercept (development time, 21.7 days) at the mean spawning temperature (4°C) of this population. The inlet shows the cumulative embryo survival over time according to an observed mortality rate (Z) of 0.17 per day (107). Considering the observed Ea (0.92 eV, blue), a tolerable temperature increase of 3°C (4 to 7°C) causes a decrease in development time of 7.6 days compared to 4.3 days with an Ea of 0.65 eV (green), as indicated by stippled lines. Accordingly, all other things being equal, metabolic theory (Ea of 0.65 eV) would underestimate the reduction in cod embryo development times due to a temperature increase of 3°C by the associated cumulative survival until hatch by ~80%.

11

fig. S7. Projected changes in thermal safety margins (TSMs) of spawners, embryos and adults (colored symbols) under the Shared Socioeconomic Pathway 1-1.9 for the years 2040, 2060, 2080 and 2100. TSMs of the upper, middle and lower panel plots consider the upper (90%), median and lower (10%) percentile bounds of temperatures generated by the climate model ensemble (5 models), respectively. Yellow dashed lines mark the 0.0°C safety margin cut-off used for risk assessments. Smooth lines are based on generalized additive model fits for each life stage and scenario.

12

fig. S8. Projected changes in thermal safety margins (TSMs) of spawners, embryos and adults (colored symbols) under the Shared Socioeconomic Pathway 4-6.0 for the years 2040, 2060, 2080 and 2100. TSMs of the upper, middle and lower panel plots consider the upper (90%), median and lower (10%) percentile bounds of temperatures generated by the climate model ensemble (5 models), respectively. Yellow dashed lines mark the 0.0°C safety margin cut-off used for risk assessments. Smooth lines are based on generalized additive model fits for each life stage and scenario.

13

fig. S9. Projected changes in thermal safety margins (TSMs) of spawners, embryos and adults (colored symbols) under the Shared Socioeconomic Pathway 5-8.5 for the years 2040, 2060, 2080 and 2100. TSMs of the upper, middle and lower panel plots consider the upper (90%), median and lower (10%) percentile bounds of temperatures generated by the climate model ensemble (5 models), respectively. Yellow dashed lines mark the 0.0 °C safety margin cut-off used for risk assessments. Smooth lines are based on generalized additive model fits for each life stage and scenario.

14

fig. S10. Boxplots (boxes: 25th to 75th percentiles, white lines: median, whiskers: 10th and 90th percentiles) show activation energies of individual measurements of larval and adult MO2 using different temperature acclimation procedures. Experiments with less than 5 days of acclimation were considered to represent “acute” responses following Lefevre et al. (58). The difference between acute and acclimated responses (> 5 days of acclimation, light grey) is not significant (linear mixed-effect model: P = 0.610, n = 79).

15

table S1. Experimental and behavioral Tmax values for 15 marine and freshwater species presented in fig. S1.

Species Realm Experimental Tmax (°C) Reference Behavioral

Tmax (°C) Reference

Thymallus thymallus Freshwater 8 (108) 7.2 (108) Hippoglossus hippoglossus Marine 8 (109) 7 (109) Salvelinus alpinus Freshwater 8 (110) 8.6 (111) Gadus morhua Marine 9 (112) 7 (113) Anarhichas lupus Marine 12 (114) 9 (114) Cyclopterus lumpus Marine 14 (115) 14 (115) Salvelinus fontinalis Freshwater 16 (116) 19 (117) Scophthalmus maximus Marine 17 (118) 17.2 (119) Salmo salar Freshwater 18 (120) 15.5 (121) Oncorhynchus mykiss Freshwater 18 (122) 18 (122) Pagrus major Marine 20 (123) 22 (124) Solea senegalensis Marine 22 (125) 23 (125) Odontesthes bonariensis Freshwater 24 (126) 21 (126) Acanthochromis polyacanthus Marine 30 (127) 28.5 (127) Dascyllus aruanus Marine 33 (38) 30 (38)

16

table S2. Phylogenetic signal indices and significance per thermal tolerance metrics and life stage n Albouheif’s

Cmean Blomberg’s

K Pagel’s λ Albouheif’s Cmean p-value

Blomberg’s K p-value

Pagel’s λ p-value

Tmax Spawner 239 0.6305 0.3495 0.9751 0.0002 0.0002 0.0002 Embryo 215 0.5760 0.2552 0.9380 0.0002 0.0002 0.0002 Larvae 65 0.6177 0.5509 1.0041 0.0002 0.0002 0.0002 Adult 460 0.6833 0.3123 0.9415 0.0002 0.0002 0.0002

Tmin Spawner 241 0.5795 0.3836 0.9718 0.0002 0.0002 0.0002 Embryo 208 0.5758 0.3354 0.9330 0.0002 0.0002 0.0002 Larvae 35 0.2470 0.7492 0.9980 0.0154 0.0002 0.0002 Adult 84 0.5275 0.4051 0.8559 0.0002 0.0002 0.0002

Trange Spawner 236 0.1564 0.1060 0.4400 0.0008 0.0330 0.0004 Embryo 204 0.1402 0.1185 0.3723 0.0026 0.0086 0.0192 Larvae 30 -0.1841 0.4592 0.0001 0.9048 0.0382 1.0000 Adult 81 0.5085 0.2840 0.6581 0.0002 0.0156 0.0002

Tmid Spawner 88 0.6161 0.5664 0.9752 0.0002 0.0002 0.0002 Embryo 103 0.4648 0.6022 1.0008 0.0002 0.0006 0.0002

17

table S3. Composition of the dataset used to estimate life stage differences in thermal tolerance metrics with mean standard deviations of phylogenetic imputation results. Species totals for Trange values are equal to species totals for Tmin. Tolerance measure / life stage

Species totals

Imputed /experimental (%)

Freshwater / marine (%)

Experimental (missing in

phylogeny) (%)

Standard deviation

imputation (°C) Tmax

Spawner 712 63.9 / 36.1 49.6 / 50.4 2.5 2.5 Embryo 714 67.1/ 32.9 48.6 / 51.4 2.8 2.8 Larvae 706 89.1 / 10.9 49.7 / 50.3 1.7 4.0 Adult 735 31.9 / 68.1 50.5 / 49.5 5.6 1.1

Tmin Spawner 712 63.6 / 36.4 49.6 / 50.4 2.5 2.2 Embryo 714 68.1 / 31.9 48.6 / 51.4 2.8 2.6 Larvae 698 94.4 / 5.6 49.6 / 50.4 0.6 4.9 Adult 700 87.1 / 12.9 49.4 / 50.6 0.9 3.3

18

table S4. Summaries of GAM analysis, estimated means and post-hoc comparisons of thermal tolerance metrics. Tmin GAM Thermal tolerance metrics Tmin Formula: T_min ~ 0 + life_stage + realm + s(Lat, by = life_stage, k = 5) Parametric coefficients: Estimate Std. Error t value Pr(>|t|) life_stage_Spawner 15.3086 0.1607 95.28 <2e-16 life_stage_Embryo 15.7983 0.1688 93.62 <2e-16 life_stage_Larvae 10.6322 0.2769 38.39 <2e-16 life_stage_Adult 7.4656 0.2025 36.86 <2e-16 Realm: Marine -1.7353 0.1735 -10 <2e-16 Significance of smooth terms: edf Ref.df F p-value s(Lat):life_stage_Spawner 3.779 3.972 466.9 <2e-16 s(Lat):life_stage_Embryo 3.82 3.981 386.8 <2e-16 s(Lat):life_stage_Larvae 1.157 1.296 267.7 <2e-16 s(Lat):life_stage_Adult 2.783 3.308 169.9 <2e-16 R-sq.(adj) 0.668 Deviance explained 91.10% GCV 3.1339 n 2824 Values for each Variable Estimate Std. Error Spawner-marine 13.5733 0.1545 Spawner-freshwater 15.3086 0.1607 Embryo-marine 14.063 0.1546 Embryo-freshwater 15.7983 0.1688 Larvae-marine 8.8969 0.2777 Larvae-freshwater 10.6322 0.2769 Adult-marine 5.7302 0.2024 Adult-freshwater 7.4656 0.2025 Posthoc comparison (Adjusted p values reported -- single-step method) Linear Hypotheses: Estimate Std. Error z value Pr(>|z|) Spawner-Embryo 0.4896 0.1893 2.587 0.0455 Spawner-Larvae -4.6764 0.2945 -15.88 <0.001 Spawner-Adult -7.8431 0.2254 -34.799 <0.001 Embryo-Larvae -5.166 0.2968 -17.406 <0.001 Embryo-Adult -8.3327 0.2283 -36.492 <0.001 Larvae-Adult -3.1667 0.3207 -9.873 <0.001

19

Tmax GAM Thermal tolerance metrics Tmax Formula: T_max ~ 0 + life_stage + realm + s(Lat, by = life_stage, k = 5) Parametric coefficients: Estimate Std. Error t value Pr(>|t|) life_stage_Spawner 22.8342 0.1954 116.88 <2e-16 life_stage_Embryo 24.1342 0.2093 115.31 <2e-16 life_stage_Larvae 33.8176 0.2856 118.41 <2e-16 life_stage_Adult 35.9991 0.1508 238.69 <2e-16 Realm: Marine -2.5895 0.1849 -14.01 <2e-16 Significance of smooth terms: edf Ref.df F p-value s(Lat):life_stage_Spawner 3.739 3.961 344.9 <2e-16 s(Lat):life_stage_Embryo 3.382 3.797 218.1 <2e-16 s(Lat):life_stage_Larvae 2.985 3.508 113.7 <2e-16 s(Lat):life_stage_Adult 3.938 3.998 311.1 <2e-16 R-sq.(adj) 0.782 Deviance explained 97.80% GCV 5.7583 n 2867 Values for each Variable Estimate Std. Error Spawner-marine 20.2447 0.1888 Spawner-freshwater 22.8342 0.1954 Embryo-marine 21.5446 0.1936 Embryo-freshwater 24.1342 0.2093 Larvae-marine 31.2281 0.2885 Larvae-freshwater 33.8176 0.2856 Adult-marine 33.4095 0.1658 Adult-freshwater 35.9991 0.1508 Posthoc comparison (Adjusted p values reported -- single-step method) Linear Hypotheses: Estimate Std. Error z value Pr(>|z|) Spawner-Embryo 1.2999 0.2454 5.298 <1e-06 Spawner-Larvae 10.9834 0.3198 34.344 <1e-06 Spawner-Adult 13.1649 0.2124 61.996 <1e-06 Embryo-Larvae 9.6835 0.3257 29.728 <1e-06 Embryo-Adult 11.8649 0.2216 53.551 <1e-06 Larvae-Adult 2.1814 0.3005 7.259 <1e-06

20

Trange GAM Thermal tolerance metrics Trange Formula: TRange ~ 0 + life_stage + realm + s(Lat, by = life_stage, k = 5) Parametric coefficients: Estimate Std. Error t value Pr(>|t|) life_stage_Spawner 7.8623 0.1653 47.575 <2e-16 life_stage_Embryo 8.8843 0.1758 50.534 <2e-16 life_stage_Larvae 23.3485 0.2658 87.856 <2e-16 life_stage_Adult 28.4434 0.1534 185.401 <2e-16 Realm: Marine -1.4246 0.1681 -8.474 <2e-16 Significance of smooth terms: edf Ref.df F p-value s(Lat):life_stage_Spawner 2.654 3.186 7.510 4.06E-05 s(Lat):life_stage_Embryo 2.605 3.131 2.081 0.11 s(Lat):life_stage_Larvae 3.201 3.676 14.202 1.98E-10 s(Lat):life_stage_Adult 3.925 3.997 55.210 < 2e-16 R-sq.(adj) 0.852 Deviance explained 95.40% GCV 7.6166 n 2824 Values for each Variable Estimate Std. Error Spawner-marine 6.4376 0.1605 Spawner-freshwater 7.8623 0.1653 Embryo-marine 7.4597 0.1627 Embryo-freshwater 8.8843 0.1758 Larvae-marine 21.9239 0.268 Larvae-freshwater 23.3485 0.2658 Adult-marine 27.0187 0.1621 Adult-freshwater 28.4434 0.1534 Posthoc comparison (Adjusted p values reported -- single-step method) Linear Hypotheses: Estimate Std. Error z value Pr(>|z|) Spawner-Embryo 1.022 0.2024 5.049 <1e-05 Spawner-Larvae 15.4863 0.2893 53.538 <1e-05 Spawner-Adult 20.5811 0.1934 106.443 <1e-05 Embryo-Larvae 14.4642 0.2931 49.357 <1e-05 Embryo-Adult 19.5591 0.1992 98.198 <1e-05 Larvae-Adult 5.0948 0.2862 17.799 <1e-05

21

table S5. Summary of GAM analysis of thermal safety margin (TSM) data. GAM

Thermal safety margins

Formula: Ttol ~ life_stage + Realm + s(MeanLat, by = life_stage, k = 5) Parametric Terms:

df F p-value

life_stage 2 1014.847 <2e-16

Realm 1 3.377 6.63E-02

Formula: Ttol ~ 0+life_stage + Realm + s(MeanLat, by = life_stage, k = 5) Parametric coefficients: Estimate Std. Error t value Pr(>|t|) life_stageTtol_spawn 3.9327 0.1889 20.82 <2e-16 life_stageTtol_egg 4.367 0.2026 21.558 <2e-16 life_stageTtol_adult 11.4421 0.1538 74.377 <2e-16 RealmMarine 0.3305 0.1798 1.838 0.0663

Significance of smooth terms: edf Ref.df F p-value s(MeanLat):life_stageTtol_spawn 2.559 3.082 5.652 0.000659 s(MeanLat):life_stageTtol_egg 1.534 1.876 13.341 2.71E-06 s(MeanLat):life_stageTtol_adult 3.228 3.701 11.285 4.71E-08

R-sq.(adj) 0-.565

Deviance explained 87.40%

GCV 3.7148

n 1678

Values for each Variable

Estimate Std. Error

a_Spawner-marine 4.2632 0.1653

a_Spawner-fresh 3.9327 0.1889

b_Embryo-marine 4.6975 0.1699

b_Embryo-fresh 4.367 0.2026

d_Adult-marine 11.7726 0.1429

d_Adult-fresh 11.4421 0.1538

Posthoc comparison

(Adjusted p values reported -- single-step method)

Linear Hypotheses: Estimate Std. Error z value Pr(>|z|) a_Spawner-b_Embryo 0.4343 0.2207 1.968 0.119 a_Spawner-d_Adult 7.5094 0.1922 39.064 <1e-04 b_Embryo-d_Adult 7.0751 0.2005 35.281 <1e-04

22

table S6. Akaike and Bayesian Information criterion for Brownian motion, Pagel’s, Ornstein-Uhlenbeck, Early-Burst and a star phylogeny models. Phylogenetic model AIC BIC Brownian motion 10035 10335 Ornstein-Uhlenbeck 9660 9966 Early-Burst 10046 10352 Pagel’s 9451 9756 Star phylogeny 10139 10439

References

1. A. Clarke, Principles of Thermal Ecology: Temperature, Energy and Life (Oxford Univ. Press, 2017).

2. H.-O. Pörtner, A. P. Farrell, Physiology and climate change. Science 322, 690–692 (2008). doi:10.1126/science.1163156 Medline

3. J. M. Sunday, A. E. Bates, N. K. Dulvy, Thermal tolerance and the global redistribution of animals. Nat. Clim. Change 2, 686–690 (2012). doi:10.1038/nclimate1539

4. H.-O. Pörtner, Climate variations and the physiological basis of temperature dependent biogeography: Systemic to molecular hierarchy of thermal tolerance in animals. Comp. Biochem. Physiol. A 132, 739–761 (2002). doi:10.1016/S1095-6433(02)00045-4 Medline

5. H.-O. Pörtner, R. Knust, Climate change affects marine fishes through the oxygen limitation of thermal tolerance. Science 315, 95–97 (2007). doi:10.1126/science.1135471 Medline

6. H.-O. Pörtner, C. Bock, F. C. Mark, Oxygen- and capacity-limited thermal tolerance: Bridging ecology and physiology. J. Exp. Biol. 220, 2685–2696 (2017). doi:10.1242/jeb.134585 Medline

7. H.-O. Pörtner, M. A. Peck, Climate change effects on fishes and fisheries: Towards a cause-and-effect understanding. J. Fish Biol. 77, 1745–1779 (2010). doi:10.1111/j.1095-8649.2010.02783.x Medline

8. P. J. Rombough, in Fish Physiology, vol. 11 (Elsevier, 1988), pp. 59–161.

9. A. Clarke, Costs and consequences of evolutionary temperature adaptation. Trends Ecol. Evol. 18, 573–581 (2003). doi:10.1016/j.tree.2003.08.007

10. H.-O. Pörtner, A. F. Bennett, F. Bozinovic, A. Clarke, M. A. Lardies, M. Lucassen, B. Pelster, F. Schiemer, J. H. Stillman, Trade-offs in thermal adaptation: The need for a molecular to ecological integration. Physiol. Biochem. Zool. 79, 295–313 (2006). doi:10.1086/499986 Medline

11. D. H. Janzen, Why mountain passes are higher in the tropics. Am. Nat. 101, 233–249 (1967). doi:10.1086/282487

12. J. Sunday, J. M. Bennett, P. Calosi, S. Clusella-Trullas, S. Gravel, A. L. Hargreaves, F. P. Leiva, W. C. E. P. Verberk, M. Á. Olalla-Tárraga, I. Morales-Castilla, Thermal tolerance patterns across latitude and elevation. Philos. Trans. R. Soc. London Ser. B 374, 20190036 (2019). doi:10.1098/rstb.2019.0036 Medline

13. H.-O. Pörtner, M. Lucassen, D. Storch, Metabolic biochemistry: Its role in thermal tolerance and in the capacities of physiological and ecological function. Fish Physiol. 22, 79–154 (2005). doi:10.1016/S1546-5098(04)22003-9

14. J. H. Brown, J. F. Gillooly, A. P. Allen, V. M. Savage, G. B. West, Toward a metabolic theory of ecology. Ecology 85, 1771–1789 (2004). doi:10.1890/03-9000

15. A. I. Dell, S. Pawar, V. M. Savage, Systematic variation in the temperature dependence of physiological and ecological traits. Proc. Natl. Acad. Sci. U.S.A. 108, 10591–10596 (2011). doi:10.1073/pnas.1015178108 Medline

16. T. P. Smith, T. J. H. Thomas, B. García-Carreras, S. Sal, G. Yvon-Durocher, T. Bell, S. Pawar, Community-level respiration of prokaryotic microbes may rise with global warming. Nat. Commun. 10, 5124 (2019). doi:10.1038/s41467-019-13109-1 Medline

17. J. Brett, Some principles in the thermal requirements of fishes. Q. Rev. Biol. 31, 75–87 (1956). doi:10.1086/401257

18. L. Ciannelli, K. Bailey, E. M. Olsen, Evolutionary and ecological constraints of fish spawning habitats. ICES J. Mar. Sci. 72, 285–296 (2015). doi:10.1093/icesjms/fsu145

19. M. L. Pinsky, A. M. Eikeset, D. J. McCauley, J. L. Payne, J. M. Sunday, Greater vulnerability to warming of marine versus terrestrial ectotherms. Nature 569, 108–111 (2019). doi:10.1038/s41586-019-1132-4 Medline

20. L. Comte, J. D. Olden, Climatic vulnerability of the world’s freshwater and marine fishes. Nat. Clim. Change 7, 718–722 (2017). doi:10.1038/nclimate3382

21. C. H. Trisos, C. Merow, A. L. Pigot, The projected timing of abrupt ecological disruption from climate change. Nature 580, 496–501 (2020). Medline

22. See supplementary materials.

23. M. Gidden et al., Global emissions pathways under different socioeconomic scenarios for use in CMIP6: A dataset of harmonized emissions trajectories through the end of the century. Geosci. Model Dev. 12, 1443–1475 (2019). doi:10.5194/gmd-12-1443-2019

24. D. L. Rabosky, J. Chang, P. O. Title, P. F. Cowman, L. Sallan, M. Friedman, K. Kaschner, C. Garilao, T. J. Near, M. Coll, M. E. Alfaro, An inverse latitudinal gradient in speciation rate for marine fishes. Nature 559, 392–395 (2018). doi:10.1038/s41586-018-0273-1 Medline

25. E. W. Goolsby, J. Bruggeman, C. Ané, Rphylopars: Fast multivariate phylogenetic comparative methods for missing data and within‐species variation. Methods Ecol. Evol. 8, 22–27 (2017). doi:10.1111/2041-210X.12612

26. See the acknowledgments (data and materials availability).

27. R. M. Sibly, J. H. Brown, A. Kodric-Brown, Metabolic Ecology: A Scaling Approach (Wiley, 2012).

28. M. E. Dillon, G. Wang, R. B. Huey, Global metabolic impacts of recent climate warming. Nature 467, 704–706 (2010). doi:10.1038/nature09407 Medline

29. F. T. Dahlke, M. Butzin, J. Nahrgang, V. Puvanendran, A. Mortensen, H.-O. Pörtner, D. Storch, Northern cod species face spawning habitat losses if global warming exceeds 1.5°C. Sci. Adv. 4, eaas8821 (2018). doi:10.1126/sciadv.aas8821 Medline

30. O. Hoegh-Guldberg, D. Jacob, M. Taylor, T. Guillén Bolaños, M. Bindi, S. Brown, I. A. Camilloni, A. Diedhiou, R. Djalante, K. Ebi, F. Engelbrecht, J. Guiot, Y. Hijioka, S. Mehrotra, C. W. Hope, A. J. Payne, H.-O. Pörtner, S. I. Seneviratne, A. Thomas, R. Warren, G. Zhou, The human imperative of stabilizing global climate change at 1.5°C. Science 365, eaaw6974 (2019). doi:10.1126/science.aaw6974 Medline

31. P. J. Rombough, in Seminar Series—Society for Experimental Biology, vol. 61 (Cambridge Univ. Press, 1997), pp. 177–224.

32. A. Hamdoun, D. Epel, Embryo stability and vulnerability in an always changing world. Proc. Natl. Acad. Sci. U.S.A. 104, 1745–1750 (2007). doi:10.1073/pnas.0610108104 Medline

33. F. Dahlke, M. Lucassen, U. Bickmeyer, S. Wohlrab, V. Puvanendran, A. Mortensen, M. Chierici, H.-O. Pörtner, D. Storch, Fish embryo vulnerability to combined acidification and warming coincides with a low capacity for homeostatic regulation. J. Exp. Biol. 223, jeb212589 (2020). doi:10.1242/jeb.212589

34. R. Przeslawski, M. Byrne, C. Mellin, A review and meta-analysis of the effects of multiple abiotic stressors on marine embryos and larvae. Glob. Change Biol. 21, 2122–2140 (2015). doi:10.1111/gcb.12833 Medline

35. D. R. Barneche, D. R. Robertson, C. R. White, D. J. Marshall, Fish reproductive-energy output increases disproportionately with body size. Science 360, 642–645 (2018). doi:10.1126/science.aao6868 Medline

36. B. Björnsson, A. Steinarsson, The food-unlimited growth rate of Atlantic cod (Gadus morhua). Can. J. Fish. Aquat. Sci. 59, 494–502 (2002). doi:10.1139/f02-028

37. R. S. Wu, in Fish Physiology, vol. 27 (Elsevier, 2009), pp. 79–141.

38. N. W. Pankhurst, P. L. Munday, Effects of climate change on fish reproduction and early life history stages. Mar. Freshw. Res. 62, 1015–1026 (2011). doi:10.1071/MF10269

39. B. S. Walsh, S. R. Parratt, A. A. Hoffmann, D. Atkinson, R. R. Snook, A. Bretman, T. A. R. Price, The impact of climate change on fertility. Trends Ecol. Evol. 34, 249–259 (2019). doi:10.1016/j.tree.2018.12.002 Medline

40. B. A. Seibel, J. C. Drazen, The rate of metabolism in marine animals: Environmental constraints, ecological demands and energetic opportunities. Philos. Trans. R. Soc. London Ser. B 362, 2061–2078 (2007). doi:10.1098/rstb.2007.2101 Medline

41. M. J. J. Angilletta, Thermal Adaptation: A Theoretical and Empirical Synthesis (Oxford Univ. Press, 2009).

42. M. L. Pinsky, B. Worm, M. J. Fogarty, J. L. Sarmiento, S. A. Levin, Marine taxa track local climate velocities. Science 341, 1239–1242 (2013). doi:10.1126/science.1239352 Medline

43. A. Bruge, P. Alvarez, A. Fontán, U. Cotano, G. Chust, Thermal niche tracking and future distribution of Atlantic mackerel spawning in response to ocean warming. Front. Mar. Sci. 3, 86 (2016). doi:10.3389/fmars.2016.00086

44. M. Byrne, S. A. Foo, P. M. Ross, H. M. Putnam, Limitations of cross- and multigenerational plasticity for marine invertebrates faced with global climate change. Glob. Change Biol. 26, 80–102 (2020). doi:10.1111/gcb.14882 Medline

45. H.-O. Pörtner et al., Ocean systems. In Climate Change 2014: Impacts, Adaptation, and Vulnerability. Part A: Global and Sectoral Aspects. Contribution of Working Group II to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change (Cambridge Univ. Press, 2014).

46. A. B. Neuheimer, B. R. MacKenzie, M. R. Payne, Temperature-dependent adaptation allows fish to meet their food across their species’ range. Sci. Adv. 4, eaar4349 (2018). doi:10.1126/sciadv.aar4349 Medline

47. R. J. Wootton, C. Smith, Reproductive Biology of Teleost Fishes (Wiley, 2014).

48. C. M. Duarte, S. Agusti, E. Barbier, G. L. Britten, J. C. Castilla, J.-P. Gattuso, R. W. Fulweiler, T. P. Hughes, N. Knowlton, C. E. Lovelock, H. K. Lotze, M. Predragovic, E. Poloczanska, C. Roberts, B. Worm, Rebuilding marine life. Nature 580, 39–51 (2020). doi:10.1038/s41586-020-2146-7 Medline

49. H.-O. Pörtner, C. Bock, R. Knust, G. Lannig, M. Lucassen, F. C. Mark, F. J. Sartoris, Cod and climate in a latitudinal cline: Physiological analyses of climate effects in marine fishes. Clim. Res. 37, 253–270 (2008). doi:10.3354/cr00766

50. J. M. Bennett, P. Calosi, S. Clusella-Trullas, B. Martínez, J. Sunday, A. C. Algar, M. B. Araújo, B. A. Hawkins, S. Keith, I. Kühn, C. Rahbek, L. Rodríguez, A. Singer, F. Villalobos, M. Ángel Olalla-Tárraga, I. Morales-Castilla, GlobTherm, a global database on thermal tolerances for aquatic and terrestrial organisms. Sci. Data 5, 180022 (2018). doi:10.1038/sdata.2018.22 Medline

51. S. Tsoukali, A. W. Visser, B. R. MacKenzie, Functional responses of North Atlantic fish eggs to increasing temperature. Mar. Ecol. Prog. Ser. 555, 151–165 (2016). doi:10.3354/meps11758

52. R. Motani, P. C. Wainwright, How warm is too warm for the life cycle of actinopterygian fishes? Sci. Rep. 5, 11597 (2015). doi:10.1038/srep11597 Medline

53. M. Moyano, C. Candebat, Y. Ruhbaum, S. Álvarez-Fernández, G. Claireaux, J.-L. Zambonino-Infante, M. A. Peck, Effects of warming rate, acclimation temperature and

ontogeny on the critical thermal maximum of temperate marine fish larvae. PLOS ONE 12, e0179928 (2017). doi:10.1371/journal.pone.0179928 Medline

54. K. E. F. Hokanson, A National Compendium of Freshwater Fish and Water Temperature Data. Volume II: Temperature Requirement Data for Thirty Fishes (U.S. Environmental Protection Agency Environmental Research Laboratory, Duluth, MN, 1990).

55. S. S. Hasnain, B. J. Shuter, C. K. Minns, Phylogeny influences the relationships linking key ecological thermal metrics for North American freshwater fish species. Can. J. Fish. Aquat. Sci. 70, 964–972 (2013). doi:10.1139/cjfas-2012-0217

56. M. A. Peck, K. B. Huebert, J. K. Llopiz, Intrinsic and extrinsic factors driving match-mismatch dynamics during the early life history of marine fishes. Adv. Ecol. Res. 47, 177–302 (2012). doi:10.1016/B978-0-12-398315-2.00003-X

57. A. Clarke, N. M. Johnston, Scaling of metabolic rate with body mass and temperature in teleost fish. J. Anim. Ecol. 68, 893–905 (1999). doi:10.1046/j.1365-2656.1999.00337.x

58. S. Lefevre, Are global warming and ocean acidification conspiring against marine ectotherms? A meta-analysis of the respiratory effects of elevated temperature, high CO2 and their interaction. Conserv. Physiol. 4, cow009 (2016). doi:10.1093/conphys/cow009 Medline

59. M. A. Peck, M. Moyano, Measuring respiration rates in marine fish larvae: Challenges and advances. J. Fish Biol. 88, 173–205 (2016). doi:10.1111/jfb.12810 Medline

60. W. I. Lutterschmidt, V. H. Hutchison, The critical thermal maximum: Data to support the onset of spasms as the definitive end point. Can. J. Zool. 75, 1553–1560 (1997). doi:10.1139/z97-782

61. E. W. Goolsby, Rapid maximum likelihood ancestral state reconstruction of continuous characters: A rerooting-free algorithm. Ecol. Evol. 7, 2791–2797 (2017). doi:10.1002/ece3.2837 Medline

62. J. Bruggeman, J. Heringa, B. W. Brandt, PhyloPars: Estimation of missing parameter values using phylogeny. Nucleic Acids Res. 37 (suppl. 2), W179–W184 (2009). doi:10.1093/nar/gkp370 Medline

63. C. Penone, A. D. Davidson, K. T. Shoemaker, M. Di Marco, C. Rondinini, T. M. Brooks, B. E. Young, C. H. Graham, G. C. Costa, Imputation of missing data in life‐history trait datasets: Which approach performs the best? Methods Ecol. Evol. 5, 961–970 (2014). doi:10.1111/2041-210X.12232

64. M. Schleuning, E. L. Neuschulz, J. Albrecht, I. M. A. Bender, D. E. Bowler, D. M. Dehling, S. A. Fritz, C. Hof, T. Mueller, L. Nowak, M. C. Sorensen, K. Böhning-Gaese, W. D. Kissling, Trait-based assessments of climate-change impacts on interacting species. Trends Ecol. Evol. 35, 319–328 (2020). doi:10.1016/j.tree.2019.12.010 Medline

65. J. Chang, D. L. Rabosky, S. A. Smith, M. E. Alfaro, An R package and online resource for macroevolutionary studies using the ray‐finned fish tree of life. Methods Ecol. Evol. 10, 1118–1124 (2019). doi:10.1111/2041-210X.13182

66. F. Keck, F. Rimet, A. Bouchez, A. Franc, phylosignal: An R package to measure, test, and explore the phylogenetic signal. Ecol. Evol. 6, 2774–2780 (2016). doi:10.1002/ece3.2051 Medline

67. G. Yu, D. K. Smith, H. Zhu, Y. Guan, T. T. Y. Lam, ggtree: An R package for visualization and annotation of phylogenetic trees with their covariates and other associated data. Methods Ecol. Evol. 8, 28–36 (2017). doi:10.1111/2041-210X.12628

68. S. N. Wood, Fast stable restricted maximum likelihood and marginal likelihood estimation of semiparametric generalized linear models. J. R. Stat. Soc. B 73, 3–36 (2011). doi:10.1111/j.1467-9868.2010.00749.x

69. T. Hothorn, F. Bretz, P. Westfall, Simultaneous inference in general parametric models. Biometr. J. 50, 346–363 (2008). doi:10.1002/bimj.200810425 Medline

70. D. Bates, M. Mächler, B. Bolker, S. Walker, Fitting linear mixed-effects models using lme4. arXiv 1406.5823 [stat.CO] (23 June 2014).

71. P. Scarponi, G. Coro, P. Pagano, A collection of Aquamaps native layers in NetCDF format. Data Brief 17, 292–296 (2018). doi:10.1016/j.dib.2018.01.026 Medline

72. B. Vilela, F. J. M. E. Villalobos, letsR: A new R package for data handling and analysis in macroecology. Methods Ecol. Evol. 6, 1229–1234 (2015). doi:10.1111/2041-210X.12401

73. R. S. Bivand, E. J. Pebesma, V. Gomez-Rubio, E. J. Pebesma, Applied Spatial Data Analysis with R (Springer, 2008).

74. Copernicus Climate Change Service, ERA5: Fifth generation of ECMWF atmospheric reanalyses of the global climate. Copernicus Climate Change Service Climate Data Store (CDS), date of access: 2020-03-16. https://cds.climate.copernicus.eu/cdsapp#!/dataset/ecv-for-climate-change?tab=form

75. H. Zuo, M. Alonso Balmaseda, S. Tietsche, K. Mogensen, M. Mayer, The ECMWF operational ensemble reanalysis–analysis system for ocean and sea ice: A description of the system and assessment. Ocean Sci. 15, 779–808 (2019). doi:10.5194/os-15-779-2019

76. V. Eyring, S. Bony, G. A. Meehl, C. A. Senior, B. Stevens, R. J. Stouffer, K. E. Taylor, Overview of the Coupled Model Intercomparison Project Phase 6 (CMIP6) experimental design and organization. Geosci. Model Dev. 9, 1937–1958 (2016). doi:10.5194/gmd-9-1937-2016

77. B. C. O’Neill, C. Tebaldi, D. P. van Vuuren, V. Eyring, P. Friedlingstein, G. Hurtt, R. Knutti, E. Kriegler, J.-F. Lamarque, J. Lowe, G. A. Meehl, R. Moss, K. Riahi, B. M. Sanderson,

Scenario Model Intercomparison Project, ScenarioMIP for CMIP6. Geosci. Model Dev. 9, 3461–3482 (2016).

78. N. C. Swart et al., CCCma CanESM5 model output prepared for CMIP6 CMIP historical, Version 20190429, Earth System Grid Federation, doi:10.22033/ESGF/CMIP6.3610 (2019).

79. N. C. Swart et al., CCCma CanESM5 model output prepared for CMIP6 ScenarioMIP ssp119, Version 20190429, Earth System Grid Federation, doi:10.22033/ESGF/CMIP6.3682 (2019).

80. N. C. Swart et al., CCCma CanESM5 model output prepared for CMIP6 ScenarioMIP ssp460, Version 20190429, Earth System Grid Federation, doi:10.22033/ESGF/CMIP6.3693 (2019).

81. N. C. Swart et al., CCCma CanESM5 model output prepared for CMIP6 ScenarioMIP ssp585, Version 20190429, Earth System Grid Federation, doi:10.22033/ESGF/CMIP6.3696 (2019).

82. R. Seferian, CNRM-CERFACS CNRM-ESM2-1 model output prepared for CMIP6 CMIP historical, Version 20181206, Earth System Grid Federation, doi:10.22033/ESGF/CMIP6.4068 (2018).

83. A. Voldoire, CNRM-CERFACS CNRM-ESM2-1 model output prepared for CMIP6 ScenarioMIP ssp119, Version 20190328, Earth System Grid Federation, doi:10.22033/ESGF/CMIP6.4182 (2019).

84. A. Voldoire, CNRM-CERFACS CNRM-ESM2-1 model output prepared for CMIP6 ScenarioMIP ssp460, Version 20190328, Earth System Grid Federation, doi:10.22033/ESGF/CMIP6.4217 (2019).

85. A. Voldoire, CNRM-CERFACS CNRM-ESM2-1 model output prepared for CMIP6 ScenarioMIP ssp585, Version 20191021, Earth System Grid Federation, doi:10.22033/ESGF/CMIP6.4226 (2019).

86. O. Boucher, S. Denvil, A. Caubel, M. A. Foujols, IPSL IPSL-CM6A-LR model output prepared for CMIP6 CMIP historical, Version 20180803, Earth System Grid Federation, doi:10.22033/ESGF/CMIP6.5195 (2018).

87. O. Boucher, S. Denvil, A. Caubel, M. A. Foujols, IPSL IPSL-CM6A-LR model output prepared for CMIP6 ScenarioMIP ssp119, Version 20190410, Earth System Grid Federation, doi:10.22033/ESGF/CMIP6.5261 (2019).

88. O. Boucher, S. Denvil, A. Caubel, M. A. Foujols, IPSL IPSL-CM6A-LR model output prepared for CMIP6 ScenarioMIP ssp460, Version 20190506, Earth System Grid Federation, doi:10.22033/ESGF/CMIP6.5268 (2019).

89. O. Boucher, S. Denvil, A. Caubel, M. A. Foujols, IPSL IPSL-CM6A-LR model output prepared for CMIP6 ScenarioMIP ssp585, Version 20190903, Earth System Grid Federation, doi:10.22033/ESGF/CMIP6.5271 (2019).

90. H. Tatebe, M. Watanabe, MIROC MIROC6 model output prepared for CMIP6 CMIP historical, Version 20181212, Earth System Grid Federation, doi:10.22033/ESGF/CMIP6.5603 (2018). (air temperature)

91. H. Tatebe, M. Watanabe, MIROC MIROC6 model output prepared for CMIP6 CMIP historical, Version 20190311, Earth System Grid Federation, doi:10.22033/ESGF/CMIP6.5603 (2018). (water temperature)

92. H. Shiogama, M. Abe, H. Tatebe, MIROC MIROC6 model output prepared for CMIP6 ScenarioMIP ssp119, Version 20190807, Earth System Grid Federation, doi:10.22033/ESGF/CMIP6.5741 (2019).

93. H. Shiogama, M. Abe, H. Tatebe, MIROC MIROC6 model output prepared for CMIP6 ScenarioMIP ssp460, Version 20190807, Earth System Grid Federation, doi:10.22033/ESGF/CMIP6.5766 (2019).

94. H. Shiogama, M. Abe, H. Tatebe, MIROC MIROC6 model output prepared for CMIP6 ScenarioMIP ssp585, Version 20190627, Earth System Grid Federation, doi:10.22033/ESGF/CMIP6.5771 (2019).

95. S. Yukimoto et al., MRI MRI-ESM2.0 model output prepared for CMIP6 CMIP historical, Version 20190222, Earth System Grid Federation, doi:10.22033/ESGF/CMIP6.6842 (2019). (air temperature)

96. S. Yukimoto et al., MRI MRI-ESM2.0 model output prepared for CMIP6 CMIP historical, Version 20191205, Earth System Grid Federation, doi:10.22033/ESGF/CMIP6.6842 (2019). (water temperature)

97. S. Yukimoto et al., MRI MRI-ESM2.0 model output prepared for CMIP6 ScenarioMIP ssp119, Version 20190222, Earth System Grid Federation, doi:10.22033/ESGF/CMIP6.6908 (2019). (air temperature)

98. S. Yukimoto et al., MRI MRI-ESM2.0 model output prepared for CMIP6 ScenarioMIP ssp119, Version 20200120, Earth System Grid Federation, doi:10.22033/ESGF/CMIP6.6908 (2019). (water temperature)

99. S. Yukimoto et al., MRI MRI-ESM2.0 model output prepared for CMIP6 ScenarioMIP ssp460, Version 20190222, Earth System Grid Federation, doi:10.22033/ESGF/CMIP6.6926 (2019). (only air temperature)

100. S. Yukimoto et al., MRI MRI-ESM2.0 model output prepared for CMIP6 ScenarioMIP ssp460, Version 20200120, Earth System Grid Federation, doi:10.22033/ESGF/CMIP6.6926 (2019). (water temperature)

101. S. Yukimoto et al., MRI MRI-ESM2.0 model output prepared for CMIP6 ScenarioMIP ssp585, Version 20191108, Earth System Grid Federation, doi:10.22033/ESGF/CMIP6.6929 (2019).

102. M. Punzet, F. Voß, A. Voß, E. Kynast, I. Bärlund, A Global Approach to Assess the Potential Impact of Climate Change on Stream Water Temperatures and Related In-Stream First-Order Decay Rates. J. Hydrometeorol. 13, 1052–1065 (2012). doi:10.1175/JHM-D-11-0138.1

103. F. T. Dahlke, E. Leo, F. C. Mark, H.-O. Pörtner, U. Bickmeyer, S. Frickenhaus, D. Storch, Effects of ocean acidification increase embryonic sensitivity to thermal extremes in Atlantic cod, Gadus morhua. Glob. Change Biol. 23, 1499–1510 (2017). doi:10.1111/gcb.13527 Medline

104. E. Leo, F. T. Dahlke, D. Storch, H.-O. Pörtner, F. C. Mark, Impact of Ocean Acidification and Warming on the bioenergetics of developing eggs of Atlantic herring Clupea harengus. Conserv. Physiol. 6, coy050 (2018). doi:10.1093/conphys/coy050 Medline

105. J. Eme, C. A. Mueller, R. G. Manzon, C. M. Somers, D. R. Boreham, J. Y. Wilson, Critical windows in embryonic development: Shifting incubation temperatures alter heart rate and oxygen consumption of Lake Whitefish (Coregonus clupeaformis) embryos and hatchlings. Comp. Biochem. Physiol. A 179, 71–80 (2015). doi:10.1016/j.cbpa.2014.09.005 Medline

106. C. Melendez, thesis, California State University, San Marcos (2019).

107. Ø. Langangen, L. C. Stige, N. A. Yaragina, F. B. Vikebø, B. Bogstad, Y. Gusdal, Egg mortality of northeast Arctic cod (Gadus morhua) and haddock (Melanogrammus aeglefinus). ICES J. Mar. Sci. 71, 1129–1136 (2013). doi:10.1093/icesjms/fst007

108. F. Lahnsteiner, M. Kletzl, The effect of water temperature on gamete maturation and gamete quality in the European grayling (Thymalus thymallus) based on experimental data and on data from wild populations. Fish Physiol. Biochem. 38, 455–467 (2012). doi:10.1007/s10695-011-9526-8 Medline

109. N. Brown, R. Shields, N. Bromage, The influence of water temperature on spawning patterns and egg quality in the Atlantic halibut (Hippoglossus hippoglossus L.). Aquaculture 261, 993–1002 (2006). doi:10.1016/j.aquaculture.2006.08.025

110. C. Gillet, Egg production in an Arctic charr (Salvelinus alpinus L.) brood stock: Effects of temperature on the timing of spawning and the quality of eggs. Aquat. Living Resour. 4, 109–116 (1991). doi:10.1051/alr:1991010

111. E. Baroudy, J. Elliott, Racial differences in eggs and juveniles of Windermere charr, Salvelinus alpinus. J. Fish Biol. 45, 407–415 (1994). doi:10.1111/j.1095-8649.1994.tb01323.x

112. H. Tveiten, Temperature influence on reproductive development and gamete quality in Atlantic cod (Gadus morhua). Cybium 32, 195 (2008).

113. D. A. Righton, K. H. Andersen, F. Neat, V. Thorsteinsson, P. Steingrund, H. Svedäng, K. Michalsen, H. H. Hinrichsen, V. Bendall, S. Neuenfeldt, P. Wright, P. Jonsson, G. Huse, J. van der Kooij, H. Mosegaard, K. Hüssy, J. Metcalfe, Thermal niche of Atlantic cod Gadus morhua: Limits, tolerance and optima. Mar. Ecol. Prog. Ser. 420, 1–13 (2010). doi:10.3354/meps08889

114. H. Tveiten, S. Solevåg, H. Johnsen, Holding temperature during the breeding season influences final maturation and egg quality in common wolffish. J. Fish Biol. 58, 374–385 (2001). doi:10.1111/j.1095-8649.2001.tb02259.x

115. S. M. Pountney, I. Lein, H. Migaud, A. Davie, High temperature is detrimental to captive lumpfish (Cyclopterus lumpus, L) reproductive performance. Aquaculture 522, 735121 (2020). doi:10.1016/j.aquaculture.2020.735121

116. K. E. Hokanson, J. H. McCormick, B. R. Jones, J. H. Tucker, Thermal requirements for maturation, spawning, and embryo survival of the brook trout, Salvelinus fontinalis. J. Fish. Board Canada 30, 975–984 (1973). doi:10.1139/f73-158

117. M. Baril, P. Magnan, in Ecology, Behaviour and Conservation of the Charrs, Genus Salvelinus (Springer, 2002), pp. 175–181.

118. N. Devauchelle, J. Alexandre, N. Le Corre, Y. Letty, Spawning of turbot (Scophthalmus maximus) in captivity. Aquaculture 69, 159–184 (1988). doi:10.1016/0044-8486(88)90194-9

119. L. Barbut, C. Groot Crego, S. Delerue-Ricard, S. Vandamme, F. A. M. Volckaert, G. Lacroix, How larval traits of six flatfish species impact connectivity. Limnol. Oceanogr. 64, 1150–1171 (2019). doi:10.1002/lno.11104

120. N. W. Pankhurst, H. R. King, Temperature and salmonid reproduction: Implications for aquaculture. J. Fish Biol. 76, 69–85 (2010). doi:10.1111/j.1095-8649.2009.02484.x Medline

121. A. Baisez, J. M. Bach, C. Leon, T. Parouty, R. Terrade, M. Hoffmann, P. Laffaille, Migration delays and mortality of adult Atlantic salmon Salmo salar en route to spawning grounds on the River Allier, France. Endanger. Species Res. 15, 265–270 (2011). doi:10.3354/esr00384

122. N. Pankhurst, G. Purser, G. Van Der Kraak, P. Thomas, G. Forteath, Effect of holding temperature on ovulation, egg fertility, plasma levels of reproductive hormones and in vitro ovarian steroidogenesis in the rainbow trout Oncorhynchus mykiss. Aquaculture 146, 277–290 (1996). doi:10.1016/S0044-8486(96)01374-9

123. B.-S. Lim, H. Kagawa, K. Gen, K. J. F. P. Okuzawa, Biochemistry, Effects of water temperature on the gonadal development and expression of steroidogenic enzymes in the gonad of juvenile red seabream, Pagrus major. Fish Physiol. Biochem. 28, 161–162 (2003). doi:10.1023/B:FISH.0000030511.73742.35

124. S. Battaglene, R. B. Talbot, Induced spawning and larval rearing of snapper, Pagrus auratus (Pisces: Sparidae), from Australian waters. N. Z. J. Mar. Freshw. Res. 26, 179–185 (1992). doi:10.1080/00288330.1992.9516513

125. Á. García-López, E. Pascual, C. Sarasquete, G. Martínez-Rodríguez, Disruption of gonadal maturation in cultured Senegalese sole Solea senegalensis Kaup by continuous light and/or constant temperature regimes. Aquaculture 261, 789–798 (2006). doi:10.1016/j.aquaculture.2006.09.005

126. M. Elisio, T. Chalde, L. A. Miranda, Effects of short periods of warm water fluctuations on reproductive endocrine axis of the pejerrey (Odontesthes bonariensis) spawning. Comp. Biochem. Physiol. A 163, 47–55 (2012). doi:10.1016/j.cbpa.2012.05.178 Medline

127. J. Donelson, P. Munday, M. McCormick, N. Pankhurst, P. Pankhurst, Effects of elevated water temperature and food availability on the reproductive performance of a coral reef fish. Mar. Ecol. Prog. Ser. 401, 233–243 (2010). doi:10.3354/meps08366