Long-term transpiration patterns of detached leaves
By using detached leaves, we were able to follow the transpiration dynamics on a high-throughput basis in controlled conditions using the PHENOPSIS automaton. A major advantage of using detached leaves is that changes in (specific) transpiration rate are expected to be the direct effect of stomatal behaviour as leaves were chosen as fully expanded (no growth anymore) and leaf movements were prevented by fixing the leaf on the tube. However, detaching leaves from the plant is expected to induce stress and senescence responses and therefore most detached leaf assays take no longer than 2-3 hours (Bazzaz et al., 1974; Green et al., 1998; Pantin et al., 2013). Although, excision and senescence responses may drastically influence the transpiration dynamic on the long term, it is not known to what extent and after how much time these processes appear on detached grapevine leaves. Therefore, an experiment was carried out on 70 detached leaves from 68 different genotypes derived from a biparental cross between Syrah and Grenache.
Time after leaf cutting (h)
Figure 9: Transpiration patterns of detached grapevine leaves during 48-hours. Seventy leaves of 68 genotypes derived from a grapevine mapping population of Syrah and Grenache. The black line represents the average of all dynamics and grey period the night period.
The detached leaves weight losses were monitored during 48 hours to calculate the transpiration dynamic (Figure 9). After installation in the late morning (11 AM), transpiration increased till midday and decreased in the afternoon. Transpiration was drastically lowered at the onset of darkness, and remained almost stable during the rest of the night. After 18 hours in tubes, transpiration increased in response to light of the next day in all leaves and most leaves showed a typical diurnal transpiration pattern as in Arabidopsis thaliana: rapid increase (likely due to stomatal opening), steady increase until midday and decrease in the afternoon (Figures 7, 9). For a small number of leaves, transpiration rate increased normally but showed a steady decline afterwards suggesting that symptoms of stress appeared. These leaves exhibited mitigated responses of transpiration to darkness on the second night, suggesting a dysfunction of stomata. For most other leaves transpiration decreased as expected in responses to darkness on the second night, resulting in similar average nighttime transpiration levels as the preceding night. After 42 hours in tubes, the transpiration response to light was dampened in all leaves and did not show the similar dynamic as the day before suggesting the activation of stress and senescence processes influencing stomatal behaviour.
We conclude that detached leaves of grapevine could be used to monitor transpiration dynamics for at least 24 hours following excision before the arrival of severe stress symptoms. Using this method, the transpiration dynamic of a single leaf could be followed during day and night in controlled conditions at a high-throughput scale (70 leaves per PHENOPSIS chamber). As fully expanded grapevine leaves which had stopped growing were chosen for the analysis, differences in transpiration during the day and night corresponded almost exclusively to stomatal movements. Another advantage of the method is that other influences on transpiration, such as canopy structure or energy balance of the leaf were homogenised for all leaves by putting leaves in a similar orientation towards light. Our biological analysis could also focus on local signalling, disregarding possible influence of root signals. This allows us to analyse internal leaf processes that control stomatal behaviour.
Comparison of detached leaf and whole vine transpiration
We next examined to what extent isolation of the leaf from the rest of the plant did influence leaf patterns in transpiration compared to leaves still attached to the plant and whole-vine transpiration. We therefore compared the transpiration of detached leaves as described above with that of whole-plants and with gas-exchange measured on attached leaves using four genotypes. This experiment was repeated in two PHENOPSIS chambers: one with ambient CO2 conditions (400 ppm) and a second with elevated CO2 conditions (800 ppm). Elevated CO2 was added to investigate how detached leaf transpiration responds to different environmental conditions. Five detached leaves, two whole-plants and one attached leaf per genotype were installed at the end of the afternoon in each of the chambers. First, the average transpiration at night per genotype was calculated for every method followed by correlation analysis for comparing the methods in both conditions (Figure 10A). Detached leaf transpiration significantly correlated with gas-exchange measurements and whole-plant transpiration. This suggests that detached leaves can be considered as representative of attached leaves on whole plants for comparing mean transpiration of various genotypes and environmental conditions.
average of five detached leaves, a single leaf in a gas-exchange chamber or the average of two whole-plants for each of the two climate chambers, differing in the level of ambient CO2. Whole-plants and detached leaves were installed at the end of the afternoon. Four genotypes were used for comparison analysis of transpiration. Correlations between average transpiration values at night with statistical tests using Pearson’s correlation coefficients for both ambient and elevated CO2 concentrations (symbols indicate p < 0.05 (asterisk) and p < 0.1 (square)). B) Average transpiration dynamic of four genotypes in ambient CO2 conditions for detached leaves, whole-plants and attached leaves. Errors bars represent the standard error.
The average dynamic of transpiration rate for each method was calculated using the data of all genotypes under ambient CO2 conditions (Figure 10B). Detached leaves had a slightly lower transpiration rate at night than whole-plants but during daytime no difference was observed between both methods. Several reasons may be responsible for this higher whole-plant transpiration at night, such as the contribution of younger and older leaves having generally a higher transpiration (Hopper et al., 2014) or spatial microclimate heterogeneity within whole plant canopy (Albasha et al., 2019) which depends on the plant position regarding ventilators and humidifiers. Moreover, nighttime transpiration rate of detached leaves was more stable than for whole plants where transpiration rate continuously increased. We cannot exclude an effect of leaf expansion for this increase in whole plants since we considered the surface area as fixed in our calculation. However, the fact that gas-exchange showed a similar increase in transpiration rate suggests that this increase could be due to stomatal reopening during the night. Stomatal reopening at night has been observed in field-grown grapevines (Rogiers et al., 2009) and in Arabidopsis thaliana (Figure 7). The apparent absence of stomatal reopening in detached leaves may be caused by the stress induced by excision or the AXS fed to leaves. However, stomatal reopening at night could also be observed in a test on detached leaves of three different genotypes suggesting that the absence of stomatal reopening is not necessarily due to our detached leaf protocol (Figure 7, Supplementary Figure 4). Stomatal reopening at night may also be influenced by photosynthesis on the day prior to assay (Easlon and Richards, 2009). We cannot guarantee that detached leaves obtained the same light treatment outside as whole-plants, possibly resulting variation of subsequent transpiration dynamics. Nevertheless, only a small reopening was observed at night in grapevine compared to Arabidopsis thaliana (Figure 7C). This suggests that variation in transpiration dynamics at night has a small influence on nighttime water loss in grapevine.
Diurnal transpiration appeared as much more influenced by stomatal movements than nocturnal transpiration in grapevine (Figures 9). A progressive increase of transpiration rate until midday and gradual decrease in the afternoon has been observed in many plants in outdoors conditions closely following light intensity (Rogiers et al., 2009). This typical diurnal dynamic is also observed under steady light conditions (Lascève et al., 1997; van Wesemael et al., 2019) suggesting an internal signal that controls the dynamic of stomatal aperture. Indeed, we observed a similar diurnal dynamic of transpiration rate in whole plants of Arabidopsis thaliana (Figure 7), whole-plants of grapevine (Supplementary 5) and detached grapevine leaves (Figure 9). Several processes have been proposed for this ‘dusk’ anticipation, such as higher ABA sensitivity of the stomatal guard cells by circadian clock regulation (Pokhilko et al., 2013) or photosynthesis-derived signals from the mesophyll (Matthews et al., 2017), but the exact signal has not yet been identified. Therefore, a grapevine association panel was phenotyped for variation in diel transpiration dynamics on detached leaves to know if genetic variation existed for the level of decrease during the day.
Figure 11: Genotypic diel transpiration dynamics on detached leaves for 279 genotypes of the grapevine association panel. A) Genotypic transpiration dynamics based on the average of five repetitions. B) Genotypic transpiration dynamics were normalised for their transpiration rate at midday to observe genotypic differences in stomatal closure in the afternoon.
Most genotypes increased transpiration rate until midday and decreased slightly in the afternoon (Figure 11A). Then, the diurnal transpiration pattern was normalised for its transpiration rate at midday (representing generally the peak of diurnal transpiration rate) to analyse the genetic differences in the relative decrease of transpiration during the ‘dusk’ anticipation (Figure 11B). We observed large variation among genotypes in the level of decrease in the afternoon (Figure 11B), possibly depending on the initial transpiration rate at midday.
Genotypes differed considerably in the level of stomatal closure in the afternoon from strong to almost no closure. The percentage of stomatal closure in the afternoon had a H2 of 0.2, meaning that 20% of all phenotypic variation can be attributed to genetic variation. This confirms that our detached leaf assays can be used to detect genetic variation for dynamic transpiration traits in a grapevine association panel.
Table of contents :
A high-throughput phenotyping pipeline to detect genetic variation for diel transpiration patterns
A puzzling role for MALTOSE EXCESS 1 in stomatal dynamics
Genetic control of nighttime transpiration through MYB-specified epidermal ophistication in grapevine