Infestation and hydraulic consequences of induced carbon starvation

Abstract

Drought impacts on forests, including widespread die-off, are likely to increase with future climate change, although the physiological responses of trees to lethal drought are poorly understood. In particular, in situ examinations of carbon starvation and its interactions with and effects on infestation and hydraulic vulnerability are largely lacking. In this study, we conducted a controlled, in situ, repeated defoliation experiment to induce carbon stress in isolated trembling aspen (Populus tremuloides) ramets. We monitored leaf morphology, leaves per branch, and multitissue carbohydrate concentrations during canopy defoliation. We examined the subsequent effects of defoliation and defoliation-induced carbon stress on vulnerability to insect/fungus infestation and hydraulic vulnerability the following year. Defoliated ramets flushed multiple canopies, which coincided with moderate drawdown of nonstructural carbohydrate reserves. Infestation frequency greatly increased and hydraulic conductivity decreased 1 year after defoliation. Despite incomplete carbohydrate drawdown from defoliation and relatively rapid carbohydrate recovery, suggesting considerable carbohydrate reserves in aspen, defoliation-induced carbon stress held significant consequences for vulnerability to mortality agents and hydraulic performance. Our results indicate that multiyear consequences of drought via feedbacks are likely important for understanding forests' responses to drought and climate change over the coming decades.

Figure 1.

Canopy characteristics (mean ± se) of defoliated ramets after first canopy flush (C1), second canopy flush (C2), and third canopy flush (C3). A, Average area per leaf (cm²). B, Average number of leaves per branch. C, Leaf area index (LAI; m² m⁻²).

Figure 2.

Distribution of average canopy mortality of SAD-affected ramets. Average height of mortality within the canopy (left) and direction of mortality (right) are shown.

Figure 3.

Starch levels (mean ± se) of branch, xylem, bark, and root tissues in control ramets (white bars) and defoliated ramets (gray bars) over the course of the experiment. Sampling events were preleaf flush (P-L), first canopy flush (C1), second canopy flush of defoliated ramets (C2), third canopy flush of defoliated ramets (C3), and the next year (N-Y) after defoliation. Note that next-year samples were not taken from xylem/bark tissues.

Figure 4.

Suc levels (mean ± se) of branch, xylem, bark, and root tissues in control ramets (white bars) and defoliated ramets (gray bars) over the course of the experiment. Sampling events were preleaf flush (P-L), first canopy flush (C1), second canopy flush of defoliated ramets (C2), third canopy flush of defoliated ramets (C3), and the next year (N-Y) after defoliation. Note that next-year samples were not taken from xylem/bark tissues.

Figure 5.

Frequency (mean ± se) of fungus or insect attacks in control (white bars) and defoliated (gray bars) ramets 1 year after defoliation for Cytospera canker, black canker, poplar borer, and aspen bark beetle.

Figure 6.

A, Refilled basal area-specific hydraulic conductivity (mean ± se; g mm⁻¹ kPa⁻¹ s⁻¹) in control (white bars) and defoliated (gray bars) ramets in 2010 prior to defoliation and in 2011. B, Percent loss of conductivity of control (white circles) and defoliated (black circles) ramets as a function of branch water potential.