Harvard Forest >

Harvard Forest Symposium Abstract 2018

  • Title: Long-Term Soil Warming Reduces Water Holding Capacity and Thermal Buffering Capacity, Leading to Warmer and Drier Soils even Without the Warming Manipulation
  • Primary Author: William Werner (Marine Biological Laboratory)
  • Additional Authors: Michael Bernard (Marine Biological Laboratory); Francis Bowles (Marine Biological Laboratory); Jerry Melillo (Marine Biological Laboratory)
  • Abstract:

    Long-term soil warming at Harvard Forest has led to reduced soil organic matter (SOM). Soil warming can directly affect water balance by increasing evaporation from surface soils (Figure 1a), and possibly by increasing transpiration due to increased N availability enhancing tree productivity. Because soil water holding capacity (WHC) is related to SOM in SOM-rich soils, a reduction in SOM can lead to increased runoff and/or deep percolation, indirectly lowering soil moisture (Figure 1b). Since water has a much higher thermal capacity than dry soil, lower soil moisture reduces the soil’s thermal buffering capacity, leading to warmer summer soil temperatures.

    We investigated these feedbacks at the Barre Woods Soil Warming Experiment during the summer of 2017, when we temporarily suspended the warming treatment for 104 days. Neither soil temperature nor soil moisture fully converged between the control and formerly-warmed plots, despite 2017 being a relatively cool and wet summer at Harvard Forest. During the end of the off period, we measured a mean reduction in soil moisture of -0.024 cm3 H2O cm-3 soil, compared to a reduction of -0.048 cm3 H2O cm-3 soil prior to suspending the warming treatment. The fact that we observed partial but not full convergence in soil moisture provides evidence for both the direct (increased evapotranspiration) and indirect (reduced hydrological buffering) feedbacks. We also measured a mean temperature increase of +0.39°C in the formerly-warmed plot relative to the control plot, as well as a +0.28°C increase in the amplitude of the diurnal soil temperature cycle.

    We measured reduced SOM and WHC in the warmed plot relative to the control plot for both organic and mineral soil. Surprisingly, the measured SOM reductions could only statistically account for part of the measured WHC reductions. We hypothesize that soil warming may have affected SOM quality as well as SOM quantity, with more hydrophilic compounds being preferentially lost, rendering bulk SOM more hydrophobic.

    We see three major consequences of these results. First, due to Harvard Forest’s climate and soil characteristics, soil moisture is not typically a primary control on ecosystem processes. However, experiments have demonstrated that there are times and/or threshold points when it becomes more important. A generally drier soil regime would expand these time periods and more frequently exceed these thresholds. Second, we predict that on net these feedbacks will increase runoff. Projected changes in precipitation for the northeastern US are likely to amplify this effect, as precipitation is projected to come in fewer but larger events, and to become more temporally dislocated from potential evapotranspiration; both phenomena should make WHC more important to water balance at the same time that the feedbacks outlined here are reducing it. Third, these hydrologic feedbacks that further raise soil temperature could amplify the previously recognized self-reinforcing feedback between soil warming and SOM loss via the global carbon cycle, so long as the marginal effect of warmer soil temperatures on SOM decomposition remains stronger than the marginal effect of lower soil moisture.

  • Research Category: Large Experiments and Permanent Plot Studies; Soil Carbon and Nitrogen Dynamics

  • Figures:
  • LTER Abstract 2018 Figure.pdf