Amazonian forest-savanna bistability and human impact

Area of Science:

• Ecology
• Remote Sensing
• Mathematical Modeling

Background:

• A bimodal distribution of tropical tree cover has been attributed to fire-induced bistability.
• This study investigates the role of human influence versus natural factors in tropical vegetation patterns.

Purpose of the Study:

• To differentiate between human-induced and natural drivers of tropical tree cover bimodality.
• To re-evaluate the evidence for fire-induced bistability in tropical ecosystems.

Main Methods:

• Subdividing satellite vegetation data into human-unaffected and human-influenced (near cultivation) regions.
• Developing a spatiotemporal mathematical model incorporating logging rates and fire diffusion.
• Analyzing vegetation data for bimodal distribution patterns.

Main Results:

• Bimodality was largely absent in human-unaffected areas but significantly enhanced near cultivated zones.
• The mathematical model successfully reproduced observed patterns, suggesting anthropogenic edge effects and fire diffusion.
• A predictable spatial boundary (Maxwell point) was identified, separating naturally selected forest and savanna states.

Conclusions:

• The observed bimodality is primarily explained by anthropogenic edge effects and spatial heterogeneity, not solely fire-induced bistability.
• While human influence is a major driver, a small degree of bistability persists in human-unaffected areas, albeit at smaller scales.
• The findings refine our understanding of tropical ecosystem dynamics and the impact of human activities on vegetation patterns.