TY - BOOK

T1 - Using big data to understand carbon recovery and habitat availability in the Amazon biome

AU - Smith, Charlotte Caroline

PY - 2022

Y1 - 2022

N2 - The scale of information required to inform global climate change and biodiversity initiatives goes beyond traditional environmental monitoring and into the realms of big data. Halting deforestation and restoring the world’s forests is key to the success of such initiatives and there is growing recognition of the potential of large-scale restoration in the Amazon as a “nature-based solution” for both climate change and biodiversity loss. But our understanding of forest loss and recovery in the Amazon is incomplete. In this thesis I use MapBiomas, a 30-m resolution annual timeseries of Amazonian landcover from 1985 to 2020, to address knowledge gaps surrounding secondary forests and their role in carbon accumulation and habitat provisioning. Chapter 2 maps the extent, age, and carbon stocks of secondary forest in the Brazilian Amazon and explores their distribution relative to key variables known to influence secondary forest carbon accumulation. The findings show that, in 2017, despite occupying 20% of deforested land, secondary forests had offset less than 10% of deforestation emissions. Furthermore, they were typically situated in contextsthat are less favourable for carbon accumulation. These results demonstrate thatold-growth forest loss remains the most important factor determining the carbon balance of the Brazilian Amazon. Chapter 3 evaluates spatial and temporal trends in forest loss and recovery across all nine Amazonian countries. The findings reveal a strong, negative spatial relationship between old-growth forest loss and recovery by secondary forests, showing that regions with the greatest area available for large-scale restoration are also those that currently have the lowest recovery. This chapter also highlights the variation between countries; Brazil has both the highest percentage of deforestation and the lowest percentage of secondary forest recovery. Chapter 4 explores the co-location of old-growth and secondary forests. It finds that while 41% and 94% of secondary forests are adjacent or connected to old-growth forests, these values decline to 20% and 57% when considering adjacency and connectivity with structurally intact and extensive old-growth forest. It also reveals that secondary forests buffer over 40% of old-growth forest edges and reduce the number of isolated old-growth fragments by ~2 million. Chapter 5 explores the impactof deforestation, disturbance, and regeneration on habitat availability for species with different tolerances for disturbance. It reveals that, although old-growth forest cover has only reduced by 8.6%, there has been a 40% decline in biome-wide habitat for disturbance-sensitive species since 1985, with 79% of the loss due to changes in forest condition rather than extent. Overall, this thesis provides new insights into changes in forest cover and condition in the Amazon biome and demonstrates the power of big data for answering environmental questions at large spatial scales.

AB - The scale of information required to inform global climate change and biodiversity initiatives goes beyond traditional environmental monitoring and into the realms of big data. Halting deforestation and restoring the world’s forests is key to the success of such initiatives and there is growing recognition of the potential of large-scale restoration in the Amazon as a “nature-based solution” for both climate change and biodiversity loss. But our understanding of forest loss and recovery in the Amazon is incomplete. In this thesis I use MapBiomas, a 30-m resolution annual timeseries of Amazonian landcover from 1985 to 2020, to address knowledge gaps surrounding secondary forests and their role in carbon accumulation and habitat provisioning. Chapter 2 maps the extent, age, and carbon stocks of secondary forest in the Brazilian Amazon and explores their distribution relative to key variables known to influence secondary forest carbon accumulation. The findings show that, in 2017, despite occupying 20% of deforested land, secondary forests had offset less than 10% of deforestation emissions. Furthermore, they were typically situated in contextsthat are less favourable for carbon accumulation. These results demonstrate thatold-growth forest loss remains the most important factor determining the carbon balance of the Brazilian Amazon. Chapter 3 evaluates spatial and temporal trends in forest loss and recovery across all nine Amazonian countries. The findings reveal a strong, negative spatial relationship between old-growth forest loss and recovery by secondary forests, showing that regions with the greatest area available for large-scale restoration are also those that currently have the lowest recovery. This chapter also highlights the variation between countries; Brazil has both the highest percentage of deforestation and the lowest percentage of secondary forest recovery. Chapter 4 explores the co-location of old-growth and secondary forests. It finds that while 41% and 94% of secondary forests are adjacent or connected to old-growth forests, these values decline to 20% and 57% when considering adjacency and connectivity with structurally intact and extensive old-growth forest. It also reveals that secondary forests buffer over 40% of old-growth forest edges and reduce the number of isolated old-growth fragments by ~2 million. Chapter 5 explores the impactof deforestation, disturbance, and regeneration on habitat availability for species with different tolerances for disturbance. It reveals that, although old-growth forest cover has only reduced by 8.6%, there has been a 40% decline in biome-wide habitat for disturbance-sensitive species since 1985, with 79% of the loss due to changes in forest condition rather than extent. Overall, this thesis provides new insights into changes in forest cover and condition in the Amazon biome and demonstrates the power of big data for answering environmental questions at large spatial scales.

KW - forest

KW - carbon

KW - Big Data

KW - habitat

KW - amazon

U2 - 10.17635/lancaster/thesis/1678

DO - 10.17635/lancaster/thesis/1678

M3 - Doctoral Thesis

PB - Lancaster University

ER -