Banca de DEFESA: TAYNÁ BARBOZA FERRARI

Uma banca de DEFESA de MESTRADO foi cadastrada pelo programa.
DISCENTE : TAYNÁ BARBOZA FERRARI
DATA : 22/05/2020
HORA: 14:00
LOCAL: PPG Ecologia e Conservação - videoconferência
TÍTULO:

Determinants of the genetic diversity of lizards in the Brazilian Cerrado

PALAVRAS-CHAVES:

Biogeography, hierarchical Bayesian analysis, INLA, machine-learning, Squamata.

PÁGINAS: 40
GRANDE ÁREA: Ciências Biológicas
ÁREA: Ecologia
SUBÁREA: Ecologia Teórica
RESUMO:

Geographic patterns of genetic diversity have the potential to provide a better understanding of natural populations demographic trajectory and the evolutionary processes that shaped biodiversity in complex biogeographic systems. Given the complex recent geologic and climatic history of the Neotropical region, elucidating these mechanisms is essential to provide a more complete evolutionary picture of its biodiversity. Here, we tested the influence of historical, ecological and environmental predictors in determining the genetic diversity of Brazilian Cerrado lizards. To this end, we compared populations that differed in altitude, distance from the center and periphery of the biome, and in areas with distinct stability levels of vegetation distribution over the time. We also used body size as a surrogate for dispersal potential of the different lizard species. We hypothesized that populations at higher altitudes, close to the center of the Cerrado, with larger body sizes and in areas of greater vegetation stability will depict higher levels of genetic diversity. We compiled mitochondrial DNA sequences from previously published works on 34 lizard species collected within the limits of the Brazilian Cerrado. We defined sampling units using the geographic location of the species and a Bayesian implementation of the general mixed Yule-coalescent model (bGMYC), a method that investigates species limits and estimates groups that are potentially evolving separately. Subsequently, we calculated nucleotide diversity, a metric used to measure the genetic diversity of each sample unit, which we termed “populations”. We used the variables altitude, distance to the center of the Cerrado, distance to the margins of the Cerrado, lizard snout-vent length, and the number of times that climatic changes altered the Cerrado landscape in vegetation terms as predictors of genetic diversity in a machine learning multivariate analysis using Random Forests. We also employed two powerful modelling tools (SPDE – Stochastic Partial Differential Equations, and INLA – Integrated Nested Laplace Approximation) to create hierarchical spatial models that allowed the estimation of genetic diversity for non-sampled regions. Our results indicate that populations at higher and lower altitudes have higher genetic diversity compared to populations at intermediate altitudes. In addition, populations further from the Cerrado’s periphery and with smaller body sizes exhibited higher genetic diversity. Finally, genetic diversity values were lesser explained by vegetation stability than any other variable. Hence, the environmental determinants and ecological characteristics of the species better explained the genetic diversity of lizards in the Brazilian Cerrado than the historical stability.

MEMBROS DA BANCA:
Presidente - 005.506.021-86 - FABRÍCIUS MAIA CHAVES BICALHO DOMINGOS - NENHUMA
Interno - 132029001 - EDDIE LENZA DE OLIVEIRA
Externo à Instituição - ROSANE GARCIA COLLEVATTI - UFG