Effects of Pinus taeda leaf anatomy on vascular and extravascular leaf hydraulic conductance as influenced by N-fertilization and elevated CO2

International audienceSilvicultural practices (e.g., nitrogen addition through fertilization) and environmental changes (e.g., elevated [CO 2 ]) may alter needle structure, impacting mass and energy exchange between the biosphere and atmosphere through alteration of stomatal function. Hydraulic resistances in leaves, controlling the mass and energy exchanges, occur both in the xylem and in the flow paths across the mesophyll to evaporation sites, and therefore largely depends on the structure of the leaf. We used the Free-Air Carbon dioxide Enrichment (FACE) experiment, providing a unique setting for assessing the interaction effects of [CO2] and nitrogen (N) supply to examine how leaf morphological and anatomical characteristics control leaf hydraulic conductance (K leaf) of loblolly pine (Pinus taeda L.) trees subjected to ambient or elevated (+200 ppmv) CO 2 concentrations (CO 2 a and CO 2 e , respectively) and to soil nitrogen amendment (N). Our study revealed that CO 2 e decreased the number of tracheids per needle, and increased the distance from the xylem vascular bundle to the stomata cavities, perturbing the leaf hydraulic system. Both treatments induced a decrease in K leaf , and CO 2 e also decreased leaf extravascular conductance (K extravascular), the conductance to water flow from the xylem to the leaf-internal air space. Decline in K leaf under CO 2 e was driven by the decline in K extravascular , potentially due to longer path for water movement through the mesophyll, explaining the decline in stomatal conductance (g s) observed under CO 2 e. This suggests that the distance from vascular conduits to stomata sub-cavity was a major constraint of leaf water transport. Across treatments our results showed that needle vein conductivity was slightly more limited by the lumen than by the bordered-pits, the latter accounting for 30-45% of vein resistance. CO 2 e-induced reduction in K leaf was also consistent with an increased resistance to xylem collapse due to thicker cell wall. In addition, stomatal closure corresponded to the water potential inducing a reduction in 50% of leaf vascular conductance (K vascular) via xylem wall rupture. The water potential that was estimated to induce complete xylem wall collapse was related to the water potential at turgor loss. Our study provided a framework for understanding the interaction between CO 2 e and N availability in affecting leaf anatomy, and the mechanisms for the response of K leaf to the treatments. These mechanisms can be incorporated into predictive models of g s , critical for estimating forest productivity in water limited environments in current and future climates and a landscape composed of sites of a range in soil N fertility