Document Type

Honors Project - Open Access

Abstract

Tropical Dry Forests (TDFs) are among the world’s most threatened ecosystems. Many are biodiversity hotspots, and most are fragmented and vulnerable to deforestation, wildfires and climate change. TDFs also receive less conservation attention compared to other tropical forest types. To provide critical baseline information for their conservation, this research defines and quantifies the extent of different forest types and their respective aboveground biomass (AGB) in a fragmented TDF landscape in Manabí, Ecuador. Focusing on two forest reserves and their adjacent areas, we used tree plot inventories (N=16) and community composition analyses to define forest types based on tree community composition. We then applied PlanetScope imagery and a Random Forest classification approach to estimate the extent of forest types according to both our community composition-based and a local forest categorization. Classification results indicate that the community-based categorization of five forest classes provides the best estimate of forest type extent with 96% accuracy and a Kappa coefficient > 0.98. Using the same plot inventories, we estimated plot-level AGB using allometric equations. These data were subsequently used in a Random Forest regression with PlanetScope imagery to estimate AGB across forest types. Results indicate that the identified five forest types vary in AGB, ranging from the highest with a mean of 258.02 ± 60.05 Mg ha -1 biomass in Wet or Seasonal Evergreen Foothills Forest to 84.32 ± 23.19 Mg ha -1 in Early Successional or Degraded Season Evergreen Forest. This research highlights the importance of defining forest types to understand variation in AGB and provides critical baseline information to monitor forest change and integrate AGB in conservation initiatives.

Download

Included in

Geography Commons

Share

COinS
 
 

© Copyright is owned by author of this document