Shifting Forests and Carbon: Linking Community Composition and Aboveground Carbon Attributes
- Download PDF (869.0 KB)
- This publication is available only online.
Forest communities—assemblages of tree species at stand- to landscape-scales—vary in their capacity to capture and store carbon, but previous attempts using discrete forest types to quantify this relationship often fail to consider gradual changes to the seedling, live tree, and standing dead tree demographic components of the community. We analyzed two decades of plot remeasurements (20,680 remeasurements of 5594 unique plots) from the United States Department of Agriculture Forest Service Forest Inventory and Analysis (FIA) Program across the Great Lakes region of the eastern U.S. to assess shifts in forest communities using Latent Dirichlet Allocation, a continuous model of forest community composition. We tested for links between aboveground live and standing dead tree carbon dynamics and shifts in community composition (the relative proportion of different communities within a stand) of three demographic components (seedlings < 2.54 cm, live trees ≥ 2.54 cm, and standing dead trees ≥ 2.54 cm). Live tree carbon varied across six unique communities ranging from 5.2 Mg C ha-1 (SE = 0.77 Mg C ha- 1) in the spruce-tamarack community to 88.4 Mg C ha-1 (SE = 0.65 Mg C ha-1) in the sugar maple-basswood community. On average, forests contained less standing dead tree carbon (2.2 Mg C ha-1) than live tree carbon (40.4 Mg C ha-1), because of lower stem density and fewer species of standing dead trees. Both live and standing dead tree carbon increased over time across the study area (on average 0.274 and 0.045 Mg C ha-1 year-1, respectively). Remeasurements of FIA plots revealed that shifts in community composition of the live and standing dead demographic components were associated with decreases in live tree carbon. Conversely, larger degree of divergence between the community composition of the live and standing dead demographic components and shifts in stand structure between remeasurements (mean, variance, and skewness of the live tree diameter distribution) were associated with increases in live tree carbon. Shifts in stand structure were significant yet weaker predictors of standing dead tree carbon dynamics. Instead, shifts in community composition of the standing dead component over time and degree of divergence between the community composition of the live tree and standing dead tree components were more closely linked to changes in standing dead tree carbon stocks. The effects of shifting communities on carbon dynamics are important for understanding carbon cycling in forests and management under continued climate change.
Keywordscarbon; forest type; forest community; Forest Inventory and Analysis (FIA); Latent Dirichlet Allocation; standing dead trees; live trees; seedlings; structure
Knott, Jonathan; Domke, Grant; Woodall, Christopher; Walters, Brian; Jenkins, Michael; Fei, Songlin. 2022. Shifting Forests and Carbon: Linking Community Composition and Aboveground Carbon Attributes. Ecosystems. 16 p. https://doi.org/10.1007/s10021-022-00765-6.