Bayesian
Hierarchical Models to Study a Spatial Distribution about Broca Infestation of
Local Coffee Plantations
Ramiro Ruíz, Clarice
Demetrio, Renato Assuncao & Roseli Leandro
Abstract
Studying the spatial
distribution of agricultural pests can provide important information about the
species dispersion mechanisms and its interaction with environmental factors.
It also helps the development of sampling plans, the integrated pest management
and planning of experiments. This work compared several models for studying the
spatial variation of the coffee berry borer infestation in order to produce
risk maps and identify areas of low/high levels of infestation. Firstly spatial
analysis was carried out using different combinations of random effects
representing spatially structured and unstructured variability. Also different
neighborhood schemes were used to represent the spatial correlation of the
data. Mixture models allowing for the excess of zeros in the first months were
also considered. The model fitting was done using MCMC methods. The results are
presented as a sequence of risk maps.
Key words: Markov chain
PDF (Portuguese)