Using Leaf Spectra to Elucidate the Taxonomy and Ecology of Living and Dried Plants

<p>Identifying plants and understanding of changes in plant communities are crucial to the conservation and management of nature. The shortwave spectral reflectance of leaves is a promising tool for rapidly identifying species at different taxonomic ranks and predicting important plant functional traits. However, the spectral reflectance of leaves changes in response to biotic and abiotic conditions.</p> <p>The aim of this dissertation is to investigate the potential of shortwave spectral reflectance to investigate (Chapter 1) how its variation affects the accuracy of methods used to predict plant taxonomies and what environmental factors most influence biophysically predicted traits, (Chapter 2) what are the changes within the spectral signature of leaves after controlled leaf desiccation, and what are the consequences for the prediction of leaf traits through radiative transfer models and inference of leaf-soil relationships, and finally (Chapter 3) how I can predict the past water content using only desiccated leaves, and what can I do to improve the prediction.</p> <p>In Chapter One, I found that the leaf reflectance classification accuracy improved when I used its natural variance in the classification model. I also found that species' relatedness does not influence biophysically predicted traits, and environmental factors affect the biophysically predicted leaf traits in both magnitude and direction. In Chapter Two, I found that the reflectance of leaves changes the most in the NIR wavelengths after leaf desiccation but not that much in known absorption features. I also found that after leaf desiccation, the uncertainties around the prediction of PROSPECT leaf traits such as photosynthetic pigments increased while in phenolics and structural leaf traits decreased, and that PROSPECT traits predicted from dried leaves are better at inferring relationships with the nutritional composition of the rhizosphere. Finally, in Chapter Three, I predicted past water content using the empirically derived traits leaf dry mass per area, nitrogen, and phenolic content, from the spectral reflectance of dry leaves. I observed that some samples had a lower prediction error than others, and I attribute these outcomes to the tightly shared trait space and environmental spaces of these samples compared to the sparsity of the others.</p> <p>In conclusion, I have shown the potential of using the spectral reflectance of leaves in plant taxonomy, ecology, and trait reconstruction by using the spectral reflectance of leaves in fresh and dry states. I believe that further investment in the field of spectral biology will be of great use to aid the quick identification of plants, generate a deeper understanding of their ecologies, and increase the application of herbarium samples in the study of plant life.</p>

Name thesis:degree_name

Doctor of Philosophy

Level thesis:degree_level

Doctoral

Discipline thesis:degree_discipline

Biology

Grantor

The Graduate School and University Center of The City University of New York

Year dc:date.available

2024

Author dc:creator

• Quinteros Casaverde, Natalia L

Advisors dc:contributor.advisor

• Douglas C. Daly
• Shawn P. Serbin

Committee members dc:contributor.committeemember

• Amy Berkov
• Ana C. Carnaval
• Fabián Michelangeli
• Kyle Dexter

• Biodiversity
• Botany
• Other Ecology and Evolutionary Biology
• Leaf spectroscopy
• hyperspectral reflectance
• taxonomy
• functional traits
• plant-soil relationships

Repository record dc:identifier

https://academicworks.cuny.edu/gc_etds/5616

OAI identifier oai:identifier

oai:academicworks.cuny.edu:gc_etds-6720