The Spruce and Peatland Responses under Climatic and Environmental Change (SPRUCE) experiment (http://mnspruce.ornl.gov/) was initiated in 2015 to assess implications of climate change on a boreal forest ecosystem at the southern edge of its range and considered vulnerable to increases in temperature. The project exposes 12 m diameter plots of a black spruce bog ecosystem to long-term temperature (T) (0 to +9 °C) and elevated CO2treatments (ambient or + 500 ppm) using a forced air system. The T treatments have led to progressive increases in atmospheric vapor pressure deficit (VPD) as expected, but nighttime re-radiation from the chamber walls has eliminated normal nighttime dew formation in the control plots.
We examined impacts of these treatments on plant water relations of Picea mariana (black spruce), Larix laricina (tamarack), Rhododendron groenlandicum (Labrador tea) and Chamaedaphne calyculata(leatherleaf), including seasonal patterns of water potential (ψ), stomatal conductance, and in addition for the trees, seasonal and diurnal patterns of sapflow and stem diameter. Sapflow was assessed using Granier-style thermal dissipation sensors that had been previously calibrated in situ (outside plots) by cutting instrumented trees and measuring their actual water uptake.
Pre-treatment measurements of spruce specific leaf conductivity declined as seasonal water stress increased beyond -1.2 MPa, with the turgor loss point reached by -2.2 MPa. Post-treatment late summer midday water potentials approached this turgor loss point for all treatments, yet stomata closure allowed full hydraulic recovery in predawn water potential. In contrast, larch foliage had difficulty rehydrating overnight with predawn water potentials ranging from -1.5 to -2.5 MPa in the warmest plots. Pre-treatment sap flow typically began by late May and was continued into fall until and ceasing as temperatures dropped below zero. Once the T treatments began, sap flow began earlier in the spring and continued later in the fall indicating an expanded physiological season. Annual stem growth increment was initiated earlier in the spring and completed earlier for warmer plots, linked to accelerated springtime water use and onset of physiological drought. There was substantial visual evidence of increasing foliar desiccation with warmer plots in late summer, with large heterogeneity, yet the most vigorous foliage in all plots had similar predawn and midday water potentials. Initial results indicate the potential for species-specific shifts in seasonal water stress dynamics, with the possibility for progressive changes in canopy structure and community composition.