INTRODUCTION TO HIERARCHiCAL modeling

(Austral) Fall, 25-student, 6-day intensive course, ECP00142, 3 Cr.

This intensive course is the empirical complement to the semester-long theoretical Population Ecology class on the right. The course provides a sampling-design and statistical modeling basis for obtaining estimates of population parameters that are essential for testing population dynamics theory. Students will study the foundations of hierarchical modeling and its applications to site-occupancy, capture-recapture and count data analysis. Course contents include an overview of the treatment of uncertainty in ecology, generalized linear modeling, and basic principles of maximum-likelihood and Bayesian data analysis. Students will learn to program simple R functions to analyze simulated data. 

This class is not a pre-requisite but it is recommended to UFRGS students who will take 'Population Ecology'. Both classes require R programing but this one contains more programing basics—if you want to take both, start with Hierarchical Modeling. If you are at UFRGS, it is also a very good idea—though not at all a requirement— to take Prof. Pablo Inchausti's Métodos Estatísticos Avançados em Ecologia e Evolução (ECP00140)—before taking this class. You can perfectly well take this class on its own too, and we welcome students from other schools in Brazil our abroad.

Note: Currently the formal Portuguese name of this class at UFRGS is Introdução à Modelagem Hierárquica

Population Ecology

25-30 students, (Austral) Fall semester, ECP0055, 2 Cr.

This class offers an overview of the theoretical foundations of population dynamics. The course opens with a presentation of exponential growth and its simplifying assumptions as a null model of ecological population dynamics.  Subsequently, we withdraw the assumptions one by one to study stochasticity, density-dependence, interspecific interactions, age-structure and spatial structure as they affect the demography of wild animal and plant populations (syllabus). Every topic will include readings of classic and contemporary papers, as well as an R practicum where students will implement numerical simulations of the population models. 

This class has been offered 20 times to more than 500 students at INPA, USP, UFRJ, UFMT, and UFRGS. The contents have changed since the first edition, at INPA, in 2004, but there's been a consistent emphasis on the formulation of theoretical ideas that can be tested under the empirical approach taught in the companion 'Introduction to Hierarchical Modeling'.  


Graduate Student mentorship

Graduate training at the Ferraz lab emphasizes quantitative reasoning, critical thinking, and breadth of understanding. Students are challenged to present their ideas in a mathematical framework, to question every scientific statement and to approach their research problems from a variety of perspectives. Ongoing lab projects center on the analysis of animal population responses to landscape change in the study area of the Biological Dynamics of Forest Fragments Project, near Manaus, in the Brazilian Amazon. Students can develop their projects around ongoing field work and/or analysis of existing Amazon data; students with research interests outside the scope of ongoing work are encouraged to develop original projects, conditional on funding availability. Prospective students are welcome to visit the lab in Porto Alegre, or send email with questions. If you know what you want from graduate school and think that this lab is the right place to get it, try composing a personal statement and send it to Gonçalo's Email.  

Lab team at the 2014 retreat, with collaborators, graduate, and undergraduate students. From left to right: (standing) Gilberto J. Fernandez, Renata Ilha, Pâmela Friedemann, Pedro Martins, Alejandra Pizarro, Carolina Schuch; (seating) Bento C. Gonçalves, Murilo Guimarães, Gonçalo Ferraz, and Bianca Darski.