Ecosystem heterogeneity and diversity mitigate Amazon forest resilience to frequent extreme droughts

Marcos Longo (1, 2) , Ryan G. Knox (3) , Naomi M. Levine (4) , Luciana F. Alves (5) , Damien Bonal (6) , Plinio B. Camargo (7) , David R. Fitzjarrald (8) , Matthew N. Hayek (1) , Natalia Restrepo-Coupe (9, 10) , Scott R. Saleska (10) , Rodrigo da Silva (11) , Scott C. Stark (12) , Raphael P. Tapajós (11) , Kenia T. Wiedemann (1) , Ke Zhang (13) , Steven C. Wofsy (1) , Paul R. Moorcroft (1)

1 Harvard University
2 JPL - Jet Propulsion Laboratory
3 LBNL - Lawrence Berkeley National Laboratory [Berkeley]
4 USC - University of Southern California
5 UCLA - University of California [Los Angeles]
6 SILVA - SILVA
7 CENA - Centro de Energia Nuclear na Agricultura
8 University at Albany [SUNY]
9 C3 - Plant Functional Biology and Climate Change Cluster
10 University of Arizona
11 UFOPA - Universidade Federal do Oeste do Pará [Santarém]
12 Michigan State University [East Lansing]
13 Hohai University

The impact of increases in drought frequency on the Amazon forest's composition, structure and functioning remain uncertain. We used a process- and individual-based ecosystem model (ED2) to quantify the forest's vulnerability to increased drought recurrence. We generated meteorologically realistic, drier-than-observed rainfall scenarios for two Amazon forest sites, Paracou (wetter) and Tapajos (drier), to evaluate the impacts of more frequent droughts on forest biomass, structure and composition. The wet site was insensitive to the tested scenarios, whereas at the dry site biomass declined when average rainfall reduction exceeded 15%, due to high mortality of large-sized evergreen trees. Biomass losses persisted when year-long drought recurrence was shorter than 2-7yr, depending upon soil texture and leaf phenology. From the site-level scenario results, we developed regionally applicable metrics to quantify the Amazon forest's climatological proximity to rainfall regimes likely to cause biomass loss >20% in 50yr according to ED2 predictions. Nearly 25% (1.8 million km(2)) of the Amazon forests could experience frequent droughts and biomass loss if mean annual rainfall or interannual variability changed by 2 sigma. At least 10% of the high-emission climate projections (CMIP5/RCP8.5 models) predict critically dry regimes over 25% of the Amazon forest area by 2100.

Mots clés: GLOBAL VEGETATION MODELS; ABOVEGROUND LIVE BIOMASS; TROPICAL RAIN-FORESTS; EARTH SYSTEM MODELS; SOUTH-AMERICA; global change; CLIMATE-CHANGE; DRY-SEASON; PHOTOSYNTHETIC SEASONALITY; CARBON FLUXES; FRENCH-GUIANA

Mots clés: forêt amazonienne; résistance à la sécheresse; biomasse; changement climatique

Domaines: Sciences du Vivant [q-bio]

Marcos Longo, Ryan G. Knox, Naomi M. Levine, Luciana F. Alves, Damien Bonal, et al.. Ecosystem heterogeneity and diversity mitigate Amazon forest resilience to frequent extreme droughts. New Phytologist, 2018, 219 (3), pp.914-931. ⟨10.1111/nph.15185⟩. ⟨hal-01896160⟩