Statistical test and correlation analysis.

This table summarizes the results from the statistical analysis conducted on our results. Tests between two parameters (under x- and y-column) are considered to be significant at a significance level (alpha = 5%) or p-value lower than 0.05. ANOVA and Kruskal-Wallis test amongst the four groups reveal significant differences in all parameters (except for growth rate). Hence, there is no significant difference in growth rate among groups 1, 2, 3 and 4. The groups were defined in the main text as well as in figure 5. Pairwise t-tests performed separately on all combinations of groups, on all variables (except growth rate) reveal significant differences between all group combinations (except between groups 2–3, and groups 3–4). Once more, t-test failed to detect any significant differences in growth rates between groups 1, 2, 3 and 4, and the results indicate that group 3 can be seen as an intermediate between group 2 and 4 (see also figure 5). A highly significant correlation between glucose consumption rate and biomass yield is seen on both parametric and non-parametric tests. Thus, our data support a linear model that explains 49% of the variation (See also table S3 for comparison within groups). Furthermore a significant correlation between glucose consumption rate and growth rate, and no significant correlation between growth rate and ethanol yield can be seen with both parametric and non-parametric tests.