Are whole canopies optimized for carbon gain? How wasteful water use in shade leaves of Eucalyptus trees constrain theoretical relationships of photosynthesis and resource distribution

Background/Question/Methods

It is commonly assumed that stomatal conductance (g_s) is optimised to maximise photosynthesis (A) while minimizing water loss, which has been confirmed for individual leaves under various manipulations of environmental drivers. When optimized, g_s should be proportional to A along canopy light gradients, and thus the intercellular CO₂ concentration (C_i) should not differ between sun and shade leaves (provided vapour pressure deficit is constant). Evidence from leaf C isotope discrimination (δ¹³C), however, generally indicates a lower C_i in sun leaves.  Few datasets have compared C_i estimated from leaf gas exchange between sun and shade leaves, which is a more direct measure of stomatal behaviour. Although δ¹³C is used to infer C_i and water use efficiency, it actually represents the CO₂ concentration inside the chloroplast (C_c), and the draw-down between C_i and C_c depends on mesophyll conductance (g_m), and A. If g_m is unaccounted for then relationships of stomatal control of A derived from δ¹³C along canopy gradients may be flawed. Here we investigate differences in stomatal behaviour between sun and shade leaves in Eucalyptus tereticornis trees grown in climate-controlled whole-tree chambers. We tested the optimization theory hypothesis that C_i is not different between sun and shade leaves when accounting for potential differences in g_m. We studied the relationship between C_i and leaf nitrogen canopy distribution – a proxy for photosynthetic capacity, and hydraulic conductance (K) – a measure of water supply capacity.  Additionally, g_m was measured with online δ¹³Cmeasurements, using a tunable diode laser.

Results/Conclusions

We found that A was reduced by ca. 40% in shade leaves associated with a 75% reduction in photosynthetically active radiation.  Increases in A were proportional to g_s in sun leaves, but this relationship was uncoupled in shade leaves. Photosynthetic capacity was lower in shade leaves (ca. 20% lower Vc_max and J_max), but K did not differ. Leaf C_i, estimated from both δ¹³C and gas exchange, was lower in sun leaves. We reject the hypothesis that C_i is optimized throughout the canopy because shade leaves used more water with lower rates of A, as compared to sun leaves. This study provides evidence that theoretical relationships between A and optimal resource distribution may not hold true for all leaves within a canopy, and has consequences for models that scale A to the canopy.