Recent statistical advances and applications of species distribution modeling
- Autores: Joaquín Martínez Minaya
- Directores de la Tesis: David Valentín Conesa Guillén (dir. tes.), Antonio Vicent Civera (codir. tes.)
- Lectura: En la Universitat de València ( España ) en 2019
- Idioma: inglés
- Tribunal Calificador de la Tesis: Havard Rue (presid.), Anabel Forte Deltell (secret.), Raquel Menezes (voc.)
- Programa de doctorado: Programa de Doctorado en Estadística y Optimización por la Universitat de València (Estudi General) y la Universitat Politècnica de València
- Materias:
- Texto completo no disponible (Saber más ...)
- Resumen
In the world that we live, we produce approximately 2.5 quintillion bytes of data per day. This huge amount of data comes from social media, internet, satellites, etc. All these data, which can be recorded in time or in space, are information that can help us to understand the spread of a disease, the movement of species or the climate change.
The use of complex statistical models has recently increased in the context of species distribution behavior. This complexity has made the inferential and predictive processes challenging to perform. The Bayesian approach has become a good option to deal with these models due to the ease with which prior information can be incorporated along with the fact that it provides a more realistic and accurate estimation of uncertainty.
This Thesis is devoted to provide an updated vision of the use of the latest statistical tools that have been emerging in the application of species distribution models (SDMs) in real contexts from a Bayesian perspective, and to develop new methodological tools to solve some statistical problems appeared in that process.
With regard to the application of the latest statistical tools in the context of SDMs, the particular objectives have been model the production of Plurivorosphaerella nawae ascospores in persimmon leaf litter; to study the spatial and climatic factors associated with the distribution of the citrus black spot disease; to analyze the effects of geographic genetic structure and spatial autocorrelation on species distribution range shifts; and to study the bottlenose dolphin (Tursiops truncatus) distribution.
Two goals have guided the most methodological part of the thesis: a review with the focus in the statistical issues in Species Distribution modeling, and the implementation of Bayesian Dirichlet regression in the context of the integrated nested Laplace approximation (INLA).
These two main objectives provide the following structure to the Thesis, which is a compendium of eight papers. The first four chapters are devoted to present a general introduction including a description of the objectives (Chapter 1), the basis of the methodology employed (Chapters 2 and 3) and a description of the results obtained (Chapter 4). The next eight chapters are dedicated to display all the papers which compose this compendium.
The final part of the Thesis includes Chapter 13, where some conclusions and lines of future research are presented, and a generic bibliography corresponding to the introductory chapters.
