Interactions, feedbacks and carbon consequences of Amazon forest edges incorporating ecosystem structure and thermal dynamics

Paulo Brando,  Yale University,  paulo.brando@yale.edu (Presenter)

Deforestation has been the main driver of forest carbon (C) losses across the Amazon
Basin. As people clear forests for the expansion of pasture, croplands, and palm oil, much
of the aboveground C is transferred to the atmosphere. In addition to directly affecting
the C cycle, deforestation indirectly contributes to the degradation of the remaining
forests, especially along forest edges neighboring agricultural fields and pasturelands.
Near these edges, hot-dry climatic conditions impose harsh environmental filters that may
cause reduced tree growth; increased mortality of large trees; increased risk of fires; and
forest grass invasion and liana overdominance, which slow forest succession and
diminish forest net C accumulation. These factors cause increased vulnerability of C
stocks in Amazon forests, with C emissions from forest degradation recently surpassing
those from deforestation. This proposal aims to quantify address the extent to which
forest edges across dry-to-wet gradients of the Amazon experience different rates of
degradation and C losses, with certain forest adjacent land uses potentially reinforcing
degradation either by increasing land surface temperatures or increasing the likelihood of
disturbances. This hypothesis stems from several studies showing drier forests are more
flammable, experience stronger soil water stress, become hotter during the dry season,
and are exposed to more severe heating stress from adjacent land uses (mostly pasture
and croplands). The overall objective of this proposal is to quantify the past, present and
future C consequences of Amazon forests' edges and to attribute the variability in forest
edge effects to different drivers.
Three comprehensive components will be used to achieve the proposal goals. First, the
project will employ over 1,000 airborne LiDAR transects and millions of Global
Ecosystem Dynamics Investigation (GEDI) LiDAR measurements, combined with
Landsat and climate data, to evaluate forest-edge-to-interior gradients in vegetation
structure along gradients of climate, land-use and time-since-edge formation. In doing so,
our proposed project will improve current and historical estimates of forest
fragmentation, edge effects, and C losses in the Amazon. Second, it will develop
empirical models to estimate the most important variables accounting for edge effects,
changes in the regional heat percolation effects from open, agricultural fields into forest
edges (from the ECOsystem Spaceborne Thermal Radiometer Experiment on Space
Station, ECOSTRESS), and forest edge disturbance history (from Landsat). These tasks
will fill the important knowledge gap of how climatic gradients within the Amazon and
different land uses influence forest edge degradation and C stocks over time. Third, the
project will use forest degradation metrics derived from LiDAR (airborne and from
GEDI) and Landsat to represent forest edge degradation into the FATES a vegetation
demographic model. Historical and future simulations based on this improved model will
be used to quantify the potential consequences of edge effects to the full C cycle and
forest flammability, leveraging a suite of scenarios of deforestation and climate change
forcings. Initializing the FATES model with forest structure across Amazonia represents
a new opportunity and major contribution to better understand and reduce uncertainties of
climate-structure-disturbance interactions and the effects of forest fragmentation on the C
cycle by employing a process-based framework utilizing multi-scale feedbacks between
vegetation structure and fire risk. Overall, this proposal will provide new estimates of the
changes in C cycling of the remaining forests of the Brazilian Amazon due to
fragmentation, forest edge responses to a changing environment

Presentation: Talk_Brando_203_35.pptx 

Presentation Type: Talk

Session: TE Plenary Morning Session 2

Talk Time: Tue (May 9) 11:30 AM