Diversity dynamics across scales

• Název práce v češtině: Dynamika diverzity napříč škálami
• Název v anglickém jazyce: Diversity dynamics across scales
• Klíčová slova: phylogeny, ecology, diversification, statistics
• Klíčová slova anglicky: phylogeny, ecology, diversification, statistics
• Akademický rok vypsání: 2012/2013
• Typ práce: disertační práce
• Jazyk práce: angličtina
• Ústav: Katedra ekologie (31-162)
• Vedoucí / školitel: prof. David Storch, Ph.D.
• Řešitel: skrytý - zadáno vedoucím/školitelem
• Datum přihlášení: 20.10.2012
• Datum zadání: 20.10.2012
• Datum a čas obhajoby: 21.06.2018 15:00
• Datum odevzdání elektronické podoby: 13.03.2018
• Datum proběhlé obhajoby: 21.06.2018
• Oponenti: prof. Mgr. Vladimír Remeš, Ph.D.

Předběžná náplň práce

(1) Motivation
Integration of phylogenetic and ecological research represents one of the most promising challenges in contemporary biology. The growing amount of molecular data, publicly
accessible databases of species distributions and life histories pose novel and highly intriguing biological questions. By combining the approaches of biogeography,
phylogenetics, and statistics, we can address these novel questions and gain a more profound insight in the phenomena that have long attracted biologists’ attention (e.g. biodiversity
gradients, rates of diversification, evolution of life histories). However, bringing all this available information together often is a challenging task. For instance, biogeographic data
are generally strongly spatially autocorrelated while data on species traits are phylogenetically dependent. In these cases, conventional statistics cannot be applied, and the
correlation structure in the data must be taken into account. Reconstruction of ancestral states, analyses of evolutionary diversification, inference of speciation and extinction events,
reconstructing historical dispersals and vicariance, all these critical approaches call for specific analyses that explicitly incorporate evolutionary models and/or spatial
autocorrelation. Therefore, a number of novel methods have been designed recently in order to address these effects within likelihood and Bayesian frameworks.

(2) Objectives
The dissertation will critically summarize these progressive methods which currently emerge in the field of evolutionary ecology. The primary focus of the dissertation will be the analyses
that combine phylogenetic data and GIS: advanced evolutionary comparative analyses (PVR, GLS), analyses of diversification rates (LASER, GEIGER), historical biogeography
(LAGRANGE, DIVA), trait evolution (PIC, ACE, JUMP). Individual methods will be introduced and their advantages as well as pitfalls critically discussed. Selected methods will
be applied to specific biological problems that are in the spotlight of current ecological and evolutionary research (e.g. analysis of mechanistic determinants of biodiversity patterns,
niche evolution and niche conservatism, analysis of evolutionary diversification, reconstruction of life -histories).

(3) Methods
Modern methods of comparative evolutionary analysis, historical biogeography, trait evolution, species distribution modeling, diversification analysis will be employed. The
methods are implemented in relevant ‘R’ packages, such as ape, ouch, phytools, dismo, diversitree, etc. Mr. Machac is skilled in statistics and R programming. Computationally
exhaustive algorithms will be executed at the MetaCentrum computer cluster.

Předběžná náplň práce v anglickém jazyce

(1) Motivation
Integration of phylogenetic and ecological research represents one of the most promising challenges in contemporary biology. The growing amount of molecular data, publicly
accessible databases of species distributions and life histories pose novel and highly intriguing biological questions. By combining the approaches of biogeography,
phylogenetics, and statistics, we can address these novel questions and gain a more profound insight in the phenomena that have long attracted biologists’ attention (e.g. biodiversity
gradients, rates of diversification, evolution of life histories). However, bringing all this available information together often is a challenging task. For instance, biogeographic data
are generally strongly spatially autocorrelated while data on species traits are phylogenetically dependent. In these cases, conventional statistics cannot be applied, and the
correlation structure in the data must be taken into account. Reconstruction of ancestral states, analyses of evolutionary diversification, inference of speciation and extinction events,
reconstructing historical dispersals and vicariance, all these critical approaches call for specific analyses that explicitly incorporate evolutionary models and/or spatial
autocorrelation. Therefore, a number of novel methods have been designed recently in order to address these effects within likelihood and Bayesian frameworks.

(2) Objectives
The dissertation will critically summarize these progressive methods which currently emerge in the field of evolutionary ecology. The primary focus of the dissertation will be the analyses
that combine phylogenetic data and GIS: advanced evolutionary comparative analyses (PVR, GLS), analyses of diversification rates (LASER, GEIGER), historical biogeography
(LAGRANGE, DIVA), trait evolution (PIC, ACE, JUMP). Individual methods will be introduced and their advantages as well as pitfalls critically discussed. Selected methods will
be applied to specific biological problems that are in the spotlight of current ecological and evolutionary research (e.g. analysis of mechanistic determinants of biodiversity patterns,
niche evolution and niche conservatism, analysis of evolutionary diversification, reconstruction of life -histories).

(3) Methods
Modern methods of comparative evolutionary analysis, historical biogeography, trait evolution, species distribution modeling, diversification analysis will be employed. The
methods are implemented in relevant ‘R’ packages, such as ape, ouch, phytools, dismo, diversitree, etc. Mr. Machac is skilled in statistics and R programming. Computationally
exhaustive algorithms will be executed at the MetaCentrum computer cluster.