Transport of water to leaves implies whole-plant coordination of hydraulic and photosynthetic traits

Abstract

In simple terms, photosynthesis can be viewed as having a water cost linked to the trade-off of opening stomata for carbon dioxide uptake and the associated loss of water through transpiration. This can be measured at the leaf scale as instantaneous water use efficiency (WUE), whereby photosynthetic rate is divided by stomatal conductance (Lawson & Vialet-Chabrand, 2019). This measure of WUE is highly dynamic according to time of day, soil moisture and other environmental conditions and plant adaptations. WUE can also be measured at scales of the whole plant and ecosystems, with plants and ecosystems with higher WUE generally maintaining productivity during dry periods (Medrano et al., 2015). While WUE is strongly influenced by dynamics of stomata controlling transpiration and CO₂ uptake, less is known about the role of the rate of water delivery to leaves (measured as hydraulic architecture of roots, stems and branches) in WUE variation at the whole-plant level. Furthermore, nitrogen use efficiency (NUE) also interacts with WUE since higher concentrations of foliar N infer higher investment in Rubisco carboxylation capacity which may enhance WUE.

As an alternative to the concept of WUE, the interplay of water cost, foliar N and the resulting photosynthetic rate can be explained with the ‘least cost’ theory where the optimal combination of water and nitrogen produce the photosynthetic rate at least total cost according to environmental conditions (Wright et al., 2003). However, until now, least cost theory has not been applied to plant investment in hydraulic architecture, or the movement of water through a plant. By combining branch hydraulic traits with measures of photosynthetic performance, Chhajed et al. have extended least cost theory to explore how hydraulic traits of branches influence the cost of gaining and using water in photosynthesis and how these traits might influence photosynthetic rates. Specifically, significance of and reasons for variation in the ratio of internal and ambient CO₂ concentration (c_i : c_a, an indicator of relative carbon supply and demand during photosynthesis) amongst co-occurring species was explored. Species with higher leaf area relative to sapwood area had higher c_i : c_a values, while branch water storage interacted with the daily range of water potentials to influence c_i : c_a.

Questions remain about the role of environmental conditions and biological factors in the patterns found. Measurements in this study were taken under inferred well-watered conditions, most likely after spring-time leaf flush. How soil moisture and other seasonal factors might influence the interplay between hydraulic and photosynthetic traits remains unclear. Macinnis-Ng et al. (2004) found similar Australian species showed seasonal variation in Huber value (inverse of leaf area per unit sapwood area), hydraulic conductivity and photosynthesis. Huber values were higher in winter while hydraulic conductivity per sapwood area and photosynthetic rates were both higher in summer. Seasonal variation was attributed to increased solar radiation and evaporative demand in summer in addition to changes in microclimate across ecosystem types (McClenahan et al., 2004). Whether seasonal changes in hydraulic and photosynthetic traits would maintain relative ratios consistent with least cost theory needs further exploration. Similarly, plant leaf water potentials and stomatal conductance rates were highly responsive to soil moisture availability, so comparing results reported by Chhajed et al. with measurements from plants in drier soils would further improve understanding of the constraints of the patterns attributed to least cost theory. Other environmental drivers relevant to a changing climate include increasing temperatures (likely causing increased photosynthesis until water stress occurs) and increasing vapour pressure deficit (likely decreasing stomatal conductance, photosynthesis and water potentials). Finally, investigating combinations of different environmental drivers is also essential.

Hydraulic traits are significant from an evolutionary perspective and are very informative in describing global patterns in drought-induced tree mortality (Anderegg et al., 2016). Despite the explanatory value of these traits, they are highly technical, time-consuming and rarely measured in conjunction with leaf gas exchange, so the dataset presented by Chhajed et al. is unique. While most studies of plant hydraulic traits focus on branch measurements, there are limitations to this approach. Specifically, measurements of branch hydraulics may not be representative of the entire plant hydraulic network, particularly in large trees (McCulloh et al., 2019). Bench-top water release curves also have limitations because they may not accurately represent stem water storage used during the day and in drier periods. That is, just because a plant has a certain amount of stored water, it does not mean that water is used each day in transpiration. For instance, during drought, large trees may use more stem capacitance which can eventually result in the development of tree water deficits. Furthermore, stem capacitance includes elastic water storage in living tissues (Zweifel & Häsler, 2001) and inelastic storage of water replaced by air due to cavitation (Steppe, 2018). Branch water release curves can differentiate the three phases of water inelastic release. At high xylem water potentials during Phase I, water comes from capillary release from xylem vessels and intercellular spaces. As water potential declines during Phase II, capillary release continues and elastic storage from living tissues increases. Finally, water is released from xylem vessels during cavitation in Phase III (Skelton, 2019). Therefore, analysis of branch capacitance might not be indicative of whole-plant capacitance because some storage compartments may not be used in intact plants (Steppe, 2018).

McCulloh et al. (2019) highlight the importance of root and soil properties and hydraulic coordination between plant organs as less-frequently measured hydraulic traits that strongly influence plant water uptake, storage and flow through the plant. Using point dendrometers on living trees allows the calculation of tree water deficits to capture a clearer picture of stem water stores (Zweifel & Häsler, 2001). Another relevant approach is the use of process-based models to simulate plant water pools and points of resistance to water flow in the soil–plant–atmosphere continuum (e.g. Williams et al., 1996). Finally, measures of tissue relative water content provide an indication of the relative water status of plants (Sapes & Sala, 2021).

While Chhajed et al. found hydraulic traits explained variation in photosynthetic traits at a single site, understanding whether these patterns hold across sites with different environmental conditions (especially under a range of soil moistures) and ecosystems with different suites of species needs further exploration. Understanding the potential for extending the least cost hypothesis across plant functional types and biomes is essential if these insights are going to be integrated into global vegetation models. Traits-based approaches provide insights into plant water and carbon budgets. However, integrating trait measurements with more dynamic field-based measures such as tree water deficit and relative water content and including key environmental measurements (including soil moisture content and vapour pressure deficit) will improve our understanding of whole-plant processes. Overall, this will improve modelling, enhancing the prediction of carbon uptake rates in a changing climate and understanding vulnerability to extreme events including droughts and heatwaves. Extension of datasets such as AusTraits (Falster et al., 2021) and SAPFLUXNET (Poyatos et al., 2020) by including dynamic and traits measurements, respectively, is an ideal opportunity for extending this analysis across ecosystems.