Using spectral biology to explore plant diversity and function in a biodiversity hotspot

Chenge, Simcelile

Using spectral biology to explore plant diversity and function in a biodiversity hotspot

Thesis / Dissertation

2026

Publisher

University of Cape Town

Department

Department of Biological Sciences

Faculty

Faculty of Science

Abstract

Spectral biology promises to revolutionise plant ecology and ecophysiology by providing hitherto unavailable information on species' identity and distribution, and their physiological and structural properties. Such information is crucial for understanding ecosystem function, while providing tools to monitor and detect change. This thesis investigates the use of leaf scale and imaging spectroscopy to characterise plant diversity and ecophysiology in the Cape Floristic Region of South Africa, a globally recognised hotspot of plant endemism and diversity. The thesis investigates four main questions: (1) Can leaf spectra capture the physiological responses of the major plant functional types to seasonal water stress? (2) Can the entire leaf spectrum (400–2400 nm) capture seasonal differences among the major plant functional types (3) Are there consistent differences in the leaf spectral reflectance of species in the Cape Restionaceae associated with their hydrological niches? and (4) Can we map the major plant functional types with airborne imaging spectroscopy? Based on partial least squares regression models, I find that leaf spectra can be used to estimate functional responses linked to seasonal plant water stress (water potential) with high accuracy, but performance declines for unseen species. I also find that the full-range spectrum reveals differing functional responses which are more distinct during summer. Furthermore, I show that the full-range spectrum has diagnostic wavelengths that help classify Restionaceae species to their hydrological niches. Lastly, I find that unmixing imaging spectroscopy data allows for accurate mapping of plant functional types in the region. Collectively, the findings advance our understanding of spectra as an alternative or a complement to traditional measures of biodiversity or ecophysiology. In sum, this work breaks new ground in our understanding of spectral biology in a hyperdiverse region and shows the value of leaf and imaging spectroscopy for accelerating our understanding of and ability to monitor biodiversity change across scales.