Branch MeOH and AA emission responses to experimental drought
During the drought experiment, a subset of drought (N = 6) and control (N = 6) plants were transported to the analytical laboratory in the morning and analyzed for ‘snap-shot’ branch MeOH and AA emissions for 1 hour in a constant light and temperature environment (Fig. 2a-c ). Control plants had high average rates of MeOH emissions (2.3-4.4 nmol m-2 s-1) and low, but detectable levels of AA emissions (0.1 nmol m-2s-1). In contrast, drought stressed trees showed low MeOH emissions (0.3 nmol m-2 s-1) while also showing higher average AA emissions (0.2 nmol m-2 s-1). This pattern resulted in the branch ‘snap-shot’ AA/MeOH emission ratio for the control (10 +/- 10%) being lower than the drought stressed plants (84 +/- 57%).
In contrast to the greenhouse drought experiments which showed rapid negative leaf physiological effects. A second set of drought experiments occurred in a cooler lab, where limited lighting was provided artificially via a grow-light. To confirm the pattern of decreased MeOH emissions and increased AA emissions, and consequently high AA/MeOH emission ratios during the greenhouse drought, real-time MeOH and AA emissions were characterized before, during and after the onset of drought impacts on leaf gas exchange. Five well-watered control individuals were sequentially transported to the laboratory and studied for real-time diurnal branch VOC emissions in the absence of additional soil moisture additions under a daytime (6:00-22:00) light pattern that mimicked the greenhouse lighting conditions. While variability in the timing and magnitudes of the MeOH and AA emissions was observed between the five individuals, the same general emission pattern was observed during the real-time emission studies as those from the ‘snap-shot’ studies with drought inducing a pattern of decreasing branch MeOH emissions and increasing AA emissions together with high AA/MeOH emissions ratios (Figure 2d-f, Figure 3 and supplementaryFigures S2-5 ).
When the temporal patterns of branch gas exchange during drought was analyzed in more detail, four distinct phases could be described. The first ‘growth phase’ with physiologically active foliage is characterized by high rates of transpiration, net photosynthesis, and MeOH emissions, with low AA emissions and values of AA/MeOH emission ratios < 30% (e.g. the first three days in Figure 3 ). During this ‘growth phase’, high light-dependent emissions of isoprene linked with photosynthetic CO2 assimilation occur during the day (6:00-22:00). Despite the constant daytime light environment and relatively stable laboratory temperature, a strong circadian gas exchange pattern was observed with maximum gas exchange fluxes near mid-day of transpiration (enhancing H2O concentrations in the branch chamber), net photosynthesis (drawing down the CO2 concentrations in the branch chamber), and MeOH and AA emissions. However, the high MeOH emissions relative to AA emissions from physiologically active branches in the ‘growth phase’ constrain daytime AA/MeOH emission ratios to low values, reaching maximum mid-day values of 6% (e.g. day 3 in Figure 4 ). The second phase of drought response consists of a strong suppression in MeOH emissions, apparently occurring prior to any reductions in stomatal conductance and CO2 and H2O gas exchange (e.g. day 4 in Figure 3 ). Although AA emissions remained low, similar to the well-watered active growth conditions, branch AA/MeOH emission ratios during this ‘MeOH suppression’ phase increased slightly from 18% on day 4 to 24% on day 5. The third phase of plant response to drought stress is characterized by a reduction in stomatal conductance with a suppression of transpiration and net photosynthesis rates, a continued strong suppression of MeOH emissions, together with the activation of aerobic fermentation including high branch emissions of the fermentation volatiles acetaldehyde, ethanol, acetic acid (AA), and acetone (e.g. initiated on day 5 in Figure 3 ). High rates of fermentation VOC emissions were found to be initiated both during the day and the night, depending on the individual, and so is not directly considered a light-dependent process. Emissions of acetaldehyde during this phase was far higher than those of the other fermentation products, whose emissions in general tracked the rise and fall of acetaldehyde emissions in this fermentation phase. Elevated branch fermentation VOC emissions from the individual shown in Figure 3 continued for three days, with the peak in AA/MeOH emission ratio (444%) occurring on beginning of day 6. Throughout this ‘aerobic fermentation’ phase, daytime transpiration and net photosynthesis continued to decline, likely as a consequence of decreasing stomatal conductance, with a loss of positive net carbon assimilation evident by day 7 and strongly reduced daytime isoprene emissions. Following this ‘fermentation phase’, a final ‘senescence phase’ was observed (day 7-10), likely associated with irreversible damage to cellular components including photosynthetic membranes and greatly compromised cellular structure and function. In this senescence phase which inevitably resulted in the leaves dying and falling off the tree, isoprene emissions were essentially eliminated, while MeOH emissions increased again to a high level with AA emissions also continuing at an elevated rate resulting in AA/MeOH emission ratios declining but remaining elevated reaching a low value of 50% by day 10.
To test for the potential reversibility of the branch MeOH emission suppression during drought, when another drought-stressed potted tree showed strong suppression of MeOH emissions in the laboratory, re-watering of the soil with 100 ml additions on day 4 (red arrows in supplementary Figure S6 ), resulted in a rapid (~15 min) return of high branch MeOH emissions and a dramatic reduction of the AA/MeOH emission ratios to around 1%. As the soil continued to dry through the experiment, the suppression of MeOH emissions was again rapidly relieved by a 100 ml soil moisture addition, regardless if it was added during the day or night. This watering effect of the drought stressed plant, completely altered the normal diurnal cycle in MeOH emissions which normally peak around mid-day in well-watered individuals. While the drought-stimulated aerobic fermentation emissions were observed on day 5, they were greatly reduced with maximum AA/MeOH emission ratios of 12%. This is in contrast to the five trees for which water was completely withheld (Figure 3and S2-5 ) which showed strong drought-stimulated aerobic fermentation emissions and high maximum AA/MeOH emission ratios ranging from 400-3500%.