Fig. 3. NSC does not increase in J. monosperma and P. edulis during drought.

As ${\psi }_{pd}$ becomes more negative, both species reallocate carbohydrates from the canopy to the roots (C). In J. monosperma, this is likely to increase water supply to the roots and allow stomata to remain open. Additionally, J. monosperma increased the sugar concentrations of apical tissues, consistent with osmotic adjustment under drought. In P. edulis, osmotic adjustment of apical tissues did not occur as stomata closed relatively early. Instead, increased sugars in the bole and roots likely supported metabolism during extended periods of stomatal closure. These tissue-specific patterns are reflected by the unweighted averages shown in (A, B). Points in (A, B) are observations of canopy sugar (blue points and line) and starch (yellow points and line). Lines in (A, B) represent point estimates of regression coefficients for the line of best fit, determined using the Akaike Information Criterion (Supplementary Table 11). C Bars represent point estimates of the regression coefficient for each NSC component and tissue against predawn water potential. Gray areas around lines in (A, B) and error bars in (C) indicate 95% confidence intervals. A separate analysis of needle, twig, bole, and root NSC for June only can be viewed in Supplementary Fig. 3. Tissue-specific regressions can be found in Supplementary Figs. 8–10.