Investigating the Effect of Multitemporal Land Use Changes to the Middle Magdalena River Valley on the Abundance and Distribution of Important Plant Species

Abstract

This thesis investigates changes to the lowland rainforest of the Middle Magdalena River Valley, Colombia, a region high in species diversity that is under severe anthropogenic pressures. Geographic Information Systems are utilised to conduct forest change analysis and Human Footprint Index (HFI) calculations between 2000 and 2018. Seven plant species are selected for reasons of ecological and economic importance, and their abundances and distributions are calculated over the same time scale. A Generalised Linear Model is created to test for correlation between each species’ abundance and distribution, and the measures of forest change and HFI, to identify which land use changes are having a statistically significant impact upon their abundance and distribution. The HFI of the study site consistently increased from 2000-2018, and the area of forest consistently decreased, indicating that the study site is being increasingly degraded by anthropogenic activities, particularly after the recent end of the Colombian armed conflict. The Serranía de Las Quinchas was found to be an important relic within the study site that has been better preserved, potentially owing to its status as a Regional Natural Park. Slope and Normalised Difference Vegetation Index (NDVI) are the two most correlated explanatory variables for species abundance and distribution, but the presence of clouds makes conclusions based on NDVI uncertain. Lack of accessibility is found to be a significant barrier to deforestation, with forests on steeper slopes being better preserved. The abundance and distribution of some species reduced, suggesting little resistance to landscape changes. Other species are more robust to changes, but this may be due to planting by roadside or in agroforestry systems. Regional conservation classifications for the species vary from their global IUCN classifications, highlighting the importance of regional assessments.