The paradox of defoliation: Declining tree water status with increasing soil water content

2020 
Abstract Defoliation can enhance tree water status by reducing canopy transpiration under drought. During long-lasting insect outbreaks however, this effect can be transient as reduced foliage affects not only transpiration but also the entire soil-plant-atmosphere continuum. In this study, we investigated the effects of defoliation and vapor pressure deficit (VPD) on plant and soil water status in balsam fir and black spruce defoliated by spruce budworm, Choristoneura fumiferana (Clemens). We sampled 48 fir trees and 36 spruce trees subjected to differing severities of defoliation. In May–September 2014 and 2015, we monitored the relative shoot water content (RWC) and soil volumetric water content (VWC), and midday shoot water potential (Ψmd, only in 2015). We applied linear mixed models (LMMs) to assess changes in RWC, Ψmd, and VWC to defoliation and VPD and we ran structural equation models (SEM) to determine the causal relationships between the measured variables in relation to defoliation and VPD. In LMMs models, defoliation and VPD, as individual factors, reduced Ψmd in both balsam fir and pooled species models but did not affect RWC. Defoliation alone increased VWC in balsam fir and in pooled models. We observed no interaction between VPD and defoliation on tree water status, but significant effect on VWC (in balsam fir and pooled models), indicating that both factors had independent and additive effects on plants but not on soil. However, in SEM models, RWC was negatively correlated to defoliation, suggesting a hydraulic safety margin. Under conditions of multiple-years of natural defoliation during a spruce budworm outbreak, the decrease in Ψmd reflects the amount of internal water capacitance that could be caused by both a lower Ψmd due to larval feeding and a negative feedback between defoliation and xylem vulnerability.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    90
    References
    4
    Citations
    NaN
    KQI
    []