Abstract: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.

Measuring carbon in forests: current status and future challenges

Evidence for a recent increase in forest growth

Allometric equations for integrating remote sensing imagery into forest monitoring programmes

Mapping carbon accumulation potential from global natural forest regrowth

An empirical, integrated forest biomass monitoring system

Review of Remote Sensing-Based Methods for Forest Aboveground Biomass Estimation: Progress, Challenges, and Prospects

Past and prospective carbon storage in United States forests

Retrieving forest biomass through integration of CASI and LiDAR data

Shifting Forests and Carbon: Linking Community Composition and Aboveground Carbon Attributes

Remote sensing in forestry: current challenges, considerations and directions

Pulse-labelling trees to study carbon allocation dynamics: a review of methods, current knowledge and future prospects.

Vicarious calibration of GEDI biomass with landsat age data for understanding secondary forest carbon dynamics

A systematic review of remote sensing and machine learning approaches for accurate carbon storage estimation in natural forests

Beyond biomass to carbon fluxes: application and evaluation of a comprehensive forest carbon monitoring system

Airborne Lidar Estimation of Aboveground Forest Biomass in the Absence of Field Inventory

Forest structure and vegetation dynamics as a driver of global carbon uptake

Controls of Initial Wood Decomposition on and in Forest Soils Using Standard Material

A comparison between data requirements and availability for calibrating predictive ecological models for lowland UK woodlands: learning new tricks from old trees

Carbon flux from decomposing wood and its dependency on temperature, wood N2 fixation rate, moisture and fungal composition in a Norway spruce forest

A solution for reducing high bias in estimates of stored carbon in tropical forests (aboveground biomass)

Benchmarking simulations of forest regrowth across Europe