Exploring spatial transferability of expert-derived river ecosystem indicators and driver-response relationships in southern Africa

Abstract

Interactive, ecosystem-based environmental flow models provide stakeholders with a wide range of useful information, such as scenario analysis of trade-offs between social and ecological impacts and river infrastructure expansion for various development pathways. However, their adoption remains restricted in data-limited environments partly because of the difficulty in developing the driver-response relationships that form the heart of ecosystem-based models. Such relationships describe ecological responses to environmental drivers and are perceived to have limited transferability from previously studied river basins. To test this perception, this study extracted and synthesized expert-derived ecosystem indicators and driver-response relationships developed for 63 sites across 20 rivers in southern Africa and evaluated the factors that determined transferability of indicators and relationships between river sites. The assessment revealed that, in general, ecosystem indicators and responses were not dictated by river type (in terms of longitudinal zone, broad habitat type, and valley slope) as calculated by continental scale datasets available for southern Africa. Instead meso-habitats played a key role in determining the ecosystem indicators and links between them. Riparian vegetation and macroinvertebrate indicator guilds had functionally similar links even as the species and underlying river type varied between sites. Expert-derived driver-response relationships were found to be convergent across a range of specialist teams and project time allocations. These findings support the spatial transferability of driver-response relationships for scenario analysis across southern Africa. Further, they provide the foundation for the development of generic ecosystem indicators and driver-response relationships for geomorphology, riparian vegetation, macroinvertebrate and fish indicators that can be used for rapid environmental flow assessments of rivers to support informed decision making related to river infrastructure development.