Resilience Indicators for Tropical Rainforests in a Dynamic Vegetation Model

Area of Science:

• Earth System Science
• Ecology
• Climate Science

Background:

• Tropical forests, especially the Amazon, are potential Earth system tipping elements.
• Declining forest resilience may precede biome shifts, indicated by increasing biomass autocorrelation.
• Satellite data show this trend, with higher autocorrelation in drier regions, but underlying processes are unclear.

Purpose of the Study:

• Investigate the processes determining tropical forest resilience using the LPJmL dynamic global vegetation model.
• Understand the scales at which these processes operate and explain observed trends in resilience indicators.

Main Methods:

• Utilized the state-of-the-art dynamic global vegetation model LPJmL.
• Constructed a reduced version of LPJmL to isolate and test specific ecological processes.
• Disabled and enabled various model processes to assess their impact on forest resilience indicators.

Main Results:

• Autocorrelation of biomass time series is higher in dry climates than wet climates, aligning with observational data.
• This pattern is linked to climate-dependent population dynamics operating on different timescales.
• Forest resilience indicators are sensitive to carbon allocation strategies, particularly during stress events.

Conclusions:

• Observed spatial variations and trends in tropical forest resilience can be explained by local physiological and ecological mechanisms.
• These findings suggest tropical forests may respond to climate change through local adaptations rather than necessarily approaching a single, large-scale tipping point.
• The identified processes (population dynamics, carbon allocation) are often simplified or missing in current Earth system models, highlighting areas for improvement.