People's influence on forest structure and composition in Sri Lankan rainforest fragments

Date of Award

Spring 1-1-2025

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Forestry and Environmental Studies

First Advisor

Ashton, Mark

Abstract

Tropical rainforests, some of the most biodiverse and carbon-rich ecosystems on earth, are threatened by land conversion for agriculture and development. Despite global efforts to slow deforestation, forest loss and land conversion rates remain high. A ripple effect of land conversion is forest fragmentation or the breakup of contiguous forests into smaller and smaller isolated patches. Forest fragmentation leads to further carbon and biodiversity losses inside the remaining forests in fragmented landscapes. Environmental changes, such as higher temperatures and wind speeds, stress and kill trees, while increased access to resources and reduced perceived value of degraded fragments drive further human disturbance. As a result, forest successional pathways are complex in fragmented forests because these chronic disturbances alter growth and development heterogeneously across the landscape. Forest development pathways are particularly uncertain in fragmented smallholder landscapes with relatively large rural populations reliant on forest resources because human values change rapidly, and few published descriptions of these forests exist. In this dissertation, I add to the growing body of work on fragmentation in smallholder landscapes by describing 18 small forest fragments (< 50 ha) within a mosaic agricultural landscape in the rainforest region of Sri Lanka.In the first chapter, I examine forest structure and aboveground biomass (AGB) of overstory trees (≥ 10 cm DBH) and compare them to the primary forest. Fragments retain about 60% of the AGB stored in the primary forest, but I suggest they are large carbon reservoirs compared to other tiles in the landscape mosaic. I also find that standard height-diameter models tend to overestimate tree heights in the fragments, which results in biased AGB estimates without field-measured tree heights. Lastly, land tenure and human disturbance are important predictors of AGB concentrations within the landscape, with surprisingly more AGB stored in privately owned fragments than government-owned and protected fragments. The second chapter explores tree species composition in fragments compared to the primary forest. Fragments exhibit higher species richness and more variability at the landscape scale. Most of this variability is related to the relatively large distance between fragment plots, but fragmentation and land tenure contribute substantially. These patterns are related to the presence of early successional species and utilitarian species in fragments, which were proportionally more abundant in private fragments than in government-owned forests and the primary forest. In the third chapter, I investigate sapling composition to evaluate whether fragments might recover primary forest structure on their current successional trajectory. Fragments had dramatically lower sapling establishment and a lower proportion of canopy trees in this stratum than in the primary forest. These results suggest that fragments may stabilize as novel ecosystems with an upper stratum comprised of subcanopy trees and early successional species, and active management may be required to restore primary forest structure and composition. My dissertation describes a smallholder landscape shaped by fragmentation and past disturbance. Differences in AGB and species composition between land tenure types, i.e., privately owned and government-owned fragments, indicate that land tenure and subsequent human management influence forest structure and composition. Patterns in sapling establishment indicate that past canopy disturbance strongly influences tree regeneration. Ultimately, these results suggest that people living in the fragmented landscape significantly influence forest fragment successional pathways. I suggest that people and their values are more frequently included in future succession models in fragmented smallholder landscapes.

This document is currently not available here.

Share

COinS