SylvCiT - An AI-based support to urban forest resilience

Area of Science:

• Urban Ecology
• Computational Ecology
• Forestry

Background:

• Urban trees provide essential ecosystem services, including carbon sequestration and heat island mitigation.
• A lack of tree diversity in urban environments compromises their resilience and ability to deliver services.
• Current urban forestry practices often overlook species functional traits and spatial relationships.

Purpose of the Study:

• To introduce SylvCiT, a novel system for recommending diverse and suitable urban tree species.
• To maximize functional diversity in urban forests at various spatial scales.
• To enhance the user experience and transparency of tree recommendation systems.

Main Methods:

• Development of a machine learning and optimization-based system (SylvCiT) integrating functional traits, planting location, and neighboring tree data.
• Analysis of urban forest structure, diversity, and carbon storage in a Montreal neighborhood.
• Assessment of species and functional group diversity in Montreal parks, simulating the impact of recommended species.

Main Results:

• SylvCiT effectively recommends tree species to increase functional diversity.
• Analysis of Montreal's urban forest revealed specific diversity and structure characteristics.
• Simulations demonstrated that planting recommended species significantly enhanced species and functional group richness in parks.

Conclusions:

• SylvCiT offers a powerful tool for improving urban forest diversity and resilience.
• Integrating functional traits and spatial context is crucial for effective urban tree selection.
• The system's focus on human-machine interfaces promotes user acceptance and transparency in urban forestry decisions.