Measurements of transpiration isotopologues and leaf water to assess enrichment models in cotton

Abstract

SummaryThe two-pool and P eclet effect models represent two theories describing mechanistic controls underlying leaf water oxygen isotope composition at the whole-leaf level (d 18 O L ). To test these models, we used a laser spectrometer coupled to a gas-exchange cuvette to make online measurements of d and source water (f sw ) remained relatively constant with a mean AE SD of 0.11 AE 0.05 and 0.13 AE 0.05, respectively, regardless of variation in E spanning 0.8-9.1 mmol m À2 s À1 . Neither f trans nor f sw exhibited a significant difference between the SS and NSS leaves at the P < 0.05 level.Our results suggest that the simpler two-pool model is adequate for predicting cotton leaf water enrichment at the whole-leaf level. We discuss the implications of adopting a two-pool concept for isotopic applications in ecological studies.

Stable isotopes in leaf water of terrestrial plants

“…The explanation for the generally lower observed value of Δ L compared to Δ e has attracted considerable research effort, because it is important to determine which leaf water signal is most relevant to the various applications that depend upon it. Two models have been proposed to explain this pattern when steady state conditions can reasonably be expected: a two‐pool model, based on two discrete pools of water within the leaf, with one of them being unenriched xylem water (Leaney et al ; Roden & Ehleringer ; Song et al ; Yakir et al ; Yakir et al ); and an advection–diffusion, or Péclet, model (Barnes et al ; Farquhar & Gan ; Farquhar & Lloyd ).…”

“…If the two‐pool model is assumed to comprise unenriched source water and enriched evaporative site water, it can be written as (Leaney et al ; Song et al ) where φ is the proportion of leaf water that is unenriched xylem water, presumably residing mainly in the major veins and ground tissue associated with them. In this model, the overestimation of Δ L by the Craig–Gordon model is because of the contribution from the unenriched pool.…”

“…A convenient way to probe observed leaf water isotopic composition for evidence of Péclet effects is to plot the proportional difference between Δ e and Δ L as a function of transpiration rate. Such plots have yielded variable results, with some authors finding a positive relationship, as predicted by the Péclet model (Barbour et al ; Loucos et al ; Ripullone et al ), and others, either no detectable relationship or a negative relationship (Cernusak et al ; Roden & Ehleringer ; Roden et al ; Song et al ; Song et al ). For our combined dataset given in the Supporting Information and shown in Fig.…”

Online CO2 and H2O oxygen isotope fractionation allows estimation of mesophyll conductance in C4 plants, and reveals that mesophyll conductance decreases as leaves age in both C4 and C3 plants

“…This may be particularly important for C 4 species, where CA occurs throughout the cytosol. However, a recent study has suggested that a radial P eclet effect may not occur within the lamina, but rather very steep gradients in enrichment may occur over a very small spatial scale near the veins (Song et al, 2014). If this observation is shown to be general across a range of species, including C 4 species, then evaporation site enrichment calculated from transpired vapour may provide accurate values of cytosolic water.…”

Leaf vein fraction influences the Péclet effect and 18O enrichment in leaf water

“…The reporting of enrichment based on the composition of the transpired vapour may reduce bias in the estimation of evaporative site enrichment in plants not at steady state (Song et al . ); however, this assumes that the propagation of the NSS signal of evaporative sites to the bulk leaf water is instantaneous, that is, that the ratio of the bulk leaf water composition to evaporative site composition is constant as the leaf water approaches steady state (Farquhar & Cernusak ). If relatively isolated pools of water exist within the leaf leading to turnover times longer than expected for a one‐pool time constant, then asynchrony between the two isotope signals would give errors in the estimates of L .…”