Abstract/Summary

Ephemeral river systems refer to rivers that only visibly contain water for a few days or weeks in a year, at most. This is usually because they are situated in hot, dry climates with very little rainfall. The groundwater in these systems may maintain a ‘linear oases’ of riparian vegetation. As such they can be invaluable to communities of humans and wildlife alike because of the relative scarcity of water available to them from other sources. Namibia has 12 large ephemeral rivers running westward through the country which support rich and unique vegetation, provide agricultural services, and act as a linear oasis for migrating and endemic megafauna (terrestrial vertebrates with a notable trophic/engineer role). Unfortunately, megafauna, which likely play an important functional role in these ecosystems, are now an uncommon sight in most of the Namibian ephemeral rivers, largely due to hunting and other human-wildlife conflicts. The absence of megafauna in some of these 12 rivers but not others, has provided the basis of a unique large-scale ‘natural’ experiment, which this PhD exploits through a remote sensing approach. Across these river systems I have been able compare the disparity in riparian habitat heterogeneity (diversity in proximate habitat types, that may result from multiple functions provided by megafauna herbivory), and subsequently infer some of the possible consequences of local extirpation on ephemeral river ecosystem functioning. Initially, I adopt a remote sensing approach (using high resolution Normalised Difference Vegetation Index data) to examine determinants of riparian habitat heterogeneity. The results suggest that the presence of African elephant (Loxodonta africana) may increase ephemeral river vegetation spatial variability. Next, I expand my approach to explore annual and seasonal variation in these megafaunal effects by comparing patterns of habitat heterogeneity over time. These results indicate that megafauna presence may increase the seasonality of vegetation heterogeneity. I then aim to explore how megafauna presence/absence might interact with episodic river flow (flash floods) to generate the observed patterns of habitat heterogeneity along ephemeral rivers, but limitations in the available data meant that the results of this analysis were inconclusive. Finally, I investigate the likely future scenarios of increasing temperatures and decreasing precipitation in Namibia under different emission scenarios and climate model projections using a ‘climate narrative’ approach, and discuss the future consequences of megafauna loss for ephemeral rivers in these scenarios. This project highlights the potential ©University of Reading 2023 Page 4 benefits but also challenges of reintegrating lost megafauna ecological processes into ephemeral river ecosystems.