Fig 1.
The biokinetic spectrum for temperature.
The observed rate of growth for all 1627 strains versus temperature consisting of 10956 data points. We highlight as a visual indication the distribution of the data using dashed lines labeled ascending curve and descending curve. We indicate the location of the Mesophile-Thermophile Gap (MTG) described in the text and of a possible secondary peak. We also show an examples of growth curves for three strains (dashed red), and the curve described by Eppley [8] (solid green) and over the same temperature range he used. The inset shows a histogram of Topt of the strains.
Fig 2.
The observed rate of growth for the 10956 data points strains versus temperature and posterior mean fits for the quantiles. The percentile values per curve are shown on the left hand side. We show the fitted quantile curves as black lines and the data as points coloured according to the quantile in which they appeared. The inset shows the observed rate of growth for those strains with Topt ≤ 50°C (left) Topt > 50°C (right) with fitted quantile curves for each case.
Table 1.
Quantile curve parameters for the whole data set.
Table 2.
Quantile curve parameters for strains with Topt ≤ 50 or Topt > 50.
Table 3.
CIs for differences of b and d for Topt ≤ 50 and Topt > 50.
Table of 99% credible intervals (CIs) for differences of b and d between strains with Topt ≤ 50 and Topt > 50. A 99% CI that includes zero indicates that there is a 99% probability that the two parameters do not differ.
Fig 3.
Quantile curves for temperature bins ≤30, …, 19–49°.
Shown are the observed rates for all strains plotted as separate overlapping bins based on the observed strain Topt. The figure also shows the fitted quantile curves for temperature bins ≤30, …, 19–49°.
Fig 4.
Quantile curves for temperature bins 20–50, …, 39–69°.
Shown are the observed rates for all strains plotted as separate overlapping bins based on the observed strain Topt. The figure also shows the fitted quantile curves for temperature bins 20–50, …, 39–69°.
Fig 5.
Quantile curves for temperature bins 40–70, …, 59–89°.
Shown are the observed rates for all strains plotted as separate overlapping bins based on the observed strain Topt. The figure also shows the fitted quantile curves for temperature bins 40–70, …, 59–89°.
Fig 6.
Quantile curves for temperature bins 60–90, …, >75°.
Shown are the observed rates for all strains plotted as separate overlapping bins based on the observed strain Topt. The figure also shows the fitted quantile curves for temperature bins 60–90, …, >75°.
Fig 7.
Thermodynamic parameter trends.
The parameter values for quantile curves versus the midpoint of the temperature bin. The posterior mean values are shown according to their respective quantiles. The numbers at the top and bottom of each plot are the temperature bins for the plots shown in Figs 3–6. For simplicity we concentrate on the 97.5% quantile. The 97.5% quantile for the a parameter remains steady until the upper bound reaches 60°C above which it declines until the lower bound reached 60°C; b remains steady until the upper bound reaches 75°C (lower bound 45°C) above which it declines until the lower bound reaches 60°C; c declines gradually until the upper bound reaches 75°C (lower bound 45°C) above which it rose, then dips between 60°C and above; while d rises to a peak at 33°C (lower bound 7.5°C) then dips until 42.5°C (lower bound 12.5°C) above which it rises, then remains steady after 60°C (lower bound 30°C).
Fig 8.
Fitted quantile curves for strains by respiration status.
Fitted quantile curves for strains by respiration status. Shown are the observed data and 95% quantile curves for each group.
Table 4.
Probabilities that b or d are equal by respiration status.
Shown is a symmetric matrix of the probabilities that b or d are equal for respiration statuses for the 95% quantile. Those probabilities less than 0.01 are bolded.
Fig 9.
Fitted quantile curves for strains by trophic status.
Shown are the observed data and the 95% quantile curves.
Table 5.
Probabilities that b or d are equal by trophic status.
Probabilities that b or d are equal by trophic status for the 95% quantile. Those probabilities less than 0.01 are bolded.
Fig 10.
Shown are the occurrences of microbial phyla by Topt.
The horizontal lines indicate the temperature span of each phylum. The circles indicate the occurrence of strains at each temperature and their diameters are proportional to the square root of their prevalence. The vertical block indicates the approximate position of the MTG.
Fig 11.
Over-plotted fitted growth curves.
Over-plotted fitted growth curves from the thermodynamic model for the 694 strains that had at least 5 data points and well defined peaks and with quantile curves for all the 10956 data points. The quantiles are the same as in Fig 2 and are calculated from the observed data. The inset shows the same fitted thermodynamic curves coloured differently for those strains with Topt ≤ 50 or Topt > 50 along with the corresponding quantile curves for each group.
Fig 12.
Thermodynamic parameters trends.
Trends in the mean thermodynamic parameters (, ΔCP and n) for strains above (solid lines) and below (dashed lines) quantile curves versus temperature. Curves are cubic-spline smoothed with df = 5 on the top row and df = 10 on the bottom row.
Fig 13.
Predicted spectra and observed data.
Shaded areas are the predicted limits for the quantiles 50, 60, 70, 80, 90, 92.5, 95, 97.5 generated by the thermodynamic model and assuming smooth trends in the thermodynamic parameters. The observed data are indicated by black dots. The upper plot (a) is for smoothed parameters with df = 5 and the lower plot (b) is for df = 10.
Table 6.
Priors for thermodynamic model parameters.