COS 178-6 - How do degradation-driven changes in forest structure in the Amazon impact the carbon and water cycles: an integrated remote-sensing and model approach

Wednesday, August 9, 2023

11:15 AM – 11:30 AM PDT

Location: B117

Presenting Author(s)

ML

Marcos Longo

Research Scientist
Lawrence Berkeley National Laboratory, United States

Coauthor(s)

MK

Michael Keller

International Institute of Tropical Forestry, USDA Forest Service, Puerto Rico
SS

Sassan Saatchi

Senior Research Scientist
Jet Propulsion Laboratory, United States
KB

Kevin Bowman

Jet Propulsion Laboratory, California Institute of Technology, United States
PM

Paul Moorcroft

Harvard University, United States
XX

Xiangtao Xu

Cornell University, United States
AK

Alexandra Konings

Stanford University, United States
AF

António Ferraz

Research Scientist
Jet Propulsion Laboratory, United States
OC

Ovidiu Csillik

Post-doctoral Researcher
Jet Propulsion Laboratory
Pasadena, California, United States
EP

Ekena Pinagé

Post-doctoral Researcher
Oregon State University, United States
EO

Elsa Ordway

Assistant Professor
University of California, Los Angeles, United States
EL

Erik Larson

Harvard University, United States
JO

Jean Ometto

Instituto Nacional de Pesquisas Espaciais, Brazil
LX

Liang Xu

Jet Propulsion Laboratory, California Institute of Technology, United States
LK

Lara M. Kueppers

University of California, Berkeley; Lawrence Berkeley National Lab, United States

Abstract: Nearly 20% of the Brazilian Amazon has been cleared, and a similar area has been degraded through selective logging, fires, and fragmentation. Forest degradation changes forest structure and composition, yet how forest structure changes may further impact ecosystem function and resilience to hot droughts remains uncertain. To study the effects of forest structure on its sensitivity to climate variability, we combined field plots, airborne lidar, and a vegetation demography model to study how forest structure across the Brazilian Amazon mediates the carbon and water cycles. We pooled data from 541 airborne lidar transects (375 ha each), which were acquired at random locations across the Brazilian Amazon in 2016, to generate distributions of forest structure at 1° resolution. We used these distributions to initialize the Ecosystem Demography Model (ED2), which represents the dynamics of vertically and horizontally heterogeneous forest canopies. ED2 has been extensively benchmarked in the Amazon. We drove ED2 with the WFDE5 reanalysis under three forest structure scenarios: (1) retained initial structure, without further changes; (2) forest recovery by excluding all future deforestation and degradation; (3) expansion of selective logging and deforestation using a published land use scenario (2010-2039) that accounts for the expansion of the road network and logging profitability. To quantify the forest sensitivity to environmental drivers, we computed marginal changes of monthly gross primary productivity (GPP) and evapotranspiration (ET) as functions of monthly averages of vapor pressure deficit (VPD) and incoming shortwave radiation (DSWR). We found that at the most degraded forests along the Southern and Eastern Brazilian Amazon, both ∆GPP/∆VPD and ∆GPP/∆DSWR are negative when VPD and DSWR are intermediate or high, indicating that forests are frequently drought-stressed. In contrast, ΔGPP/ΔVPD and ΔGPP/ΔDSWR are always positive in the minimally disturbed forests in the Northwest, suggesting that these forests remain light-limited even under the driest conditions. Across the Amazon, high-biomass forest patches showed little deleterious effect of high VPD on GPP and ET, whereas ET at low-biomass forests peaked at intermediate levels of VPD (17–22 hPa). Likewise, forest degradation towards the central Amazon increased the areas where GPP and ET respond negatively to high values of VPD and DSWR. Together, these results suggest that the local forest structure of forests is critical for understanding the ecosystem response to climate variability, and that the loss of canopy trees in the Amazon through forest degradation could increase and expand forest vulnerability to droughts.