High xylem redundance in the branches buffers the water household of trees against changes in the Huber Value

A common effect of drought is the reduction of the hydraulic capacity as xylem is embolised. Furthermore, trees can prematurely shed leaves during more severe droughts, presumably to protect xylem from excessive embolisms. Both of these effectively change the ratio of water supply and demand (i.e. the ratio of sapwood to leaf area; Huber value). While changes in this ratio during drought should therefore affect a trees’ water household, this has never been experimentally tested at the branch level. 

We asked: How does experimentally changing the branch Huber value affect the branch water household in deciduous and evergreen trees? To address this question, we experimentally changed the Huber value of trembling aspen (Populus tremuloides) and subalpine fir (Abies lasiocarpa) branches and measured regular stomatal conductance and water potentials. We did this with a fully factorial experiment either removing half of the leaf area, cutting through half of the xylem area, or both on branches in situ.  At the end of the experiment, we conducted native and max hydraulic conductivity and dye perfusion measurements. We hypothesized that reducing leaf area leads to increased area-specific stomatal conductance, resulting in constant whole-branch transpiration. We also hypothesized that reducing sapwood area leads to a decrease in water potentials, stomatal conductance and hydraulic conductivity. 

We found that after leaf removal there was a small increase of stomatal conductance in aspen but not enough to keep whole-branch transpiration constant, otherwise we did not see any effects. Surprisingly, after cutting through the xylem area there was no difference in any measured traits. This implies that removing leaf area, at least in aspen, has a greater effect on the water household than removing xylem area. We found that in aspen, the xylem was transporting much less water than its potential, implying high xylem redundance. This pattern was not as strong in subalpine fir, as they were operating closer to their potential. These different responses between the species may be explained by their different anatomical types, as fir xylem is more resistant and less conductive than aspen wood. The high degree of branch xylem redundancy found here shows that the water household of trees can be buffered against substantial changes in the Huber value, indicating that drought-related seasonal changes in xylem or leaf area may not affect water relations as much as hitherto assumed.