Quantifying losses in productivity by the rust in eucalypt plantations in Brazil

The eucalyptus plantations in Brazil have been great importance in economy with 1.3% of the Gross Domestic Product, so their productivity has recently been extensively researched by the forestry. This productivity is affected by abiotic and biotic factors. Among the biotic factors we can highlight the rust, which attacks and generates large long-term losses in large scale. The aim of this paper is to measure the effect of the rust disease in eucalyptus stand by the statistical-models approach. Firstly, a new index named Rust Severity Index - (RSI) has been proposed. The index is a weighted average of the severity scale by the inverse of age in days. The generalized linear model has been used to model the effect of region/farm and severity of disease in maximum sample volume (m(3)) of the plantations. We also fit a generalized linear regression model to maximum sample volume (m(3)) of the plantations by the rust severity index. Finally, Gompertz nonlinear regressions models were fitted to the volume growth according to region/farm and rust severity. Then, it has been possible to measure the losses in volume (m(3)) due the rust severity by the percentage difference between the rust severity categories at 51st and 72nd months.