Using Metabolic Energy Density Metrics to Understand Differences in Ecosystem Function During Drought

Abstract

Terrestrial ecosystems obtain energy in the form of carbon‐containing molecules. Quantifying energy acquisition and dissipation throughout an ecosystem may be useful for describing their resistance and resilience to disturbances. Three longleaf pine savannas with different vegetation composition—a result of variation in soil moisture and land use legacy—were used as a case study to test energy‐based metrics of ecosystem metabolic function. Available energy from gross ecosystem exchange of CO2 and its dissipation into metabolic energy density (EM) and energy storage were used to identify differences in drought recovery over an 8‐year period. Sites with higher plant functional diversity in the understory stored more energy and lowered their EM by ~20% when adapting to drought. In contrast, the site with greater abundance of woody understory and overstory species relied on stored energy twice as often as the more diverse sites. The absence of native understory species, due to anthropogenic legacy, prolonged ecosystem‐scale drought recovery by 1 year. This study provides the tools to understand differences in site metabolic energy dynamics and has the potential to identify site characteristics that indicate greater vulnerability to disturbances. Metabolic energy density can be applied to any global ecosystem and provides a first step to describe coupled carbon and energy allocation in ecosystems, which may be used to further refine ecological theory and its management implications.

Uncertainty in parameterizing a flux‐based model of vegetation carbon phenology using ecosystem respiration

“…While the xeric site also experienced abnormal winter warming, the phenological model of Re was still able to yield an estimate for EOR. This difference in phenological response could be a consequence of the site's forest structure and species composition (Wiesner et al, 2020(Wiesner et al, , 2021. As climate change has given rise to more pronounced winter warming at mid-latitudes, and winter warming has been shown to impact Re more than summer warming (Kreyling et al, 2019), this result points to the need for models and methods that can account for shifts in weather extremes.…”

Characterizing Growing Season Length of Subtropical Coniferous Forests with a Phenological Model

“…The study was conducted at the Jones Center at Ichauway (JCI) in southwestern Georgia, USA, using data from three sites collected during 2009-2017. The 11,000-hectare JCI property (31.22 • N, 84.47 • W; Figure 1) has a relatively flat terrain with long-term average annual precipitation of 1310 mm; air temperature ranges from −3 • C to 23 • C in winter, and 16 • C to 31 • C in summer [37]. The regional climate is classified as subtropical.…”

“…Where possible, the first correction method was used; however, when the phenology model yielded unrealistic predictions for the EOS correction (e.g., estimated EOS exceeded the established time scale), the second method was used. Since low-intensity prescribed fires occurred during the dormant season between winter and early spring (January-March) [37,45], the phenology model should capture the potential impacts of fire on SOS.…”

“…Where possible, the first correction method was used; however, when the phenology model yielded unrealistic predictions for the EOS correction (e.g., estimated EOS exceeded the established time scale), the second method was used. Since low-intensity prescribed fires occurred during the dormant season between winter and early spring (January-March) [37,45], the phenology model should capture the potential impacts of fire on SOS. is shown as the fitted red curve, the black dotted line is the default end of growing season (EOS) output of the phenology model using the first maximum GPP decline rate caused by short-term summer drought as the slope of ASL(t) to predict EOS, the solid blue line on the right is the corrected EOS using the second maximum GPP decline rate as the slope of ASL(t) to predict EOS; (b) k(t) is shown as the fitted red curve, the black arrow points to the first maximum GPP decline rate caused by weather disturbance, the blue arrow points to the second maximum GPP decline rate after weather disturbance in autumn.…”

“…There was no summer short-term drought or extreme air temperature events in the mesic site in 2009 [37] and thus this site-year was representative of phenological characteristics associated with normal weather conditions (Appendix B, Figure A2). The spring recovery line (RL, A RL (t)) had a slope of 0.08 g C m NC: no recovery, GPP: Gross Primary Production at the day (g C m −2 d −1 ).…”