Artificial neural networks: Modeling tree survival and mortality in the Atlantic Forest biome in Brazil.

Models to predict tree survival and mortality can help to understand vegetation dynamics and to predict effects of climate change on native forests. The objective of the present study was to use Artificial Neural Networks, based on the competition index and climatic and categorical variables, to predict tree survival and mortality in Semideciduous Seasonal Forests in the Atlantic Forest biome. Numerical and categorical trees variables, in permanent plots, were used. The Agricultural Reference Index for Drought (ARID) and the distance-dependent competition index were the variables used. The overall efficiency of classification by ANNs was higher than 92% and 93% in the training and test, respectively. The accuracy for classification and number of surviving trees was above 99% in the test and in training for all ANNs. The classification accuracy of the number of dead trees was low. The mortality accuracy rate (10.96% for training and 13.76% for the test) was higher with the ANN 4, which considers the climatic variable and the competition index. The individual tree-level model integrates dendrometric and meteorological variables, representing a new step for modeling tree survival in the Atlantic Forest biome.

Fuzzy logic applied to prospecting for areas for installation of wood panel industries.

Prospecting for suitable areas for forestry operations, where the objective is a reduction in production and transportation costs, as well as the maximization of profits and available resources, constitutes an optimization problem. However, fuzzy logic is an alternative method for solving this problem. In the context of prospecting for suitable areas for the installation of wood panel industries, we propose applying fuzzy logic analysis for simulating the planting of different species and eucalyptus hybrids in Espírito Santo State, Brazil. The necessary methodological steps for this study are as follows: a) agriclimatological zoning of different species and eucalyptus hybrids; b) the selection of the vector variables; c) the application of the Euclidean distance to the vector variables; d) the application of fuzzy logic to matrix variables of the Euclidean distance; and e) the application of overlap fuzzy logic to locate areas for installation of wood panel industries. Among all the species and hybrids, Corymbia citriodora showed the highest percentage values for the combined very good and good classes, with 8.60%, followed by Eucalyptus grandis with 8.52%, Eucalyptus urophylla with 8.35% and Urograndis with 8.34%. The fuzzy logic analysis afforded flexibility in prospecting for suitable areas for the installation of wood panel industries in the Espírito Santo State can bring great economic and social benefits to the local population with the generation of jobs, income, tax revenues and GDP increase for the State and municipalities involved. The proposed methodology can be adapted to other areas and agricultural crops.

Distribuição espacial dos danos de Heilipodus naevulus em plantio de clones de eucalipto / Spatial distribution of damage caused by Heilipodus naevulus in a clonal eucalyptus plantation

As espécies Heilipodus naevulus (Mannerheim, 1836) foi detectado, em 2008, no município de Barão de Cocais, MG, em plantios comerciais de clones de eucalipto, danificando o ponteiro principal e o lateral. Devido ao seu hábito noturno, o seu tamanho reduzido, coloração acinzentada, e , por se abrigar na área de serapilheira durante o dia, a sua detecção é dificultada, sendo a sua presença apenas evidenciada através dos ponteiros tombados. Este trabalho teve como objetivo conhecer a distribuição desta espécie durante o primeiro ano de crescimento clones eucalipto, a fim de fornecer subsídios para a elaboração de um plano de amostragem. O trabalho foi realizado no município de Barão de Cocais, MG, a partir de outubro/2008 a março/2009. Nos dois talhões de avaliação foram estratificados em baixada, encosta e topo. Foram distribuídas 18 porções, de 10 x 10 árvores, sendo nove por talhão, e, três por estrato. Foi adotado como indicador da presença de H.naevulus na planta a presença ponteiro recém danificado. Para o cálculo da distribuição espacial foram empregados a Razão Variância/Média, o Índice de Morisita, o Parâmetro k da distribuição binomial negativa, o Coeficiente de Green, a Distribuição de Poisson e a Lei de potência de Taylor. Ao longo do período de avaliação da presença de H. naevulus, os índices calculados apresentaram um padrão agregado de ocorrência dos danos, indicando que a espécie apresenta um padrão agregado de distribuição, mesmo em baixa incidência.(AU)

The species Heilipodus naevulus (Mannerheim, 1836) was detected, in 2008, in the municipal district Barão de Cocais, MG, in commercial plantings of eucalypt clones, damaging the main and the lateral pointer. Due to its night habit, its reduced size, grayish coloration, and, for sheltering in the litter area during the day, its detection is hindered, being its presence just evidenced through tumbled pointers. This work aimed to know the distribution of this species during the first year of eucalypt clones growth in order to supply subsidies for elaborating of a sampling plan. The work was carried out in the municipal district from Barão de Cocais, MG, from October/2008 to March/2009. The two plots of evaluation were stratified into lowland, slope and top. Were distributed 18 portions, of 10 x 10 trees, being nine per plot, and three for stratum. It was adopted as indicator of H. naevulus presence in the plant the pointer presence newly damaged. For calculating of the space distribution were used the Variance/Mean Ratio, Morisita Index, Parameter k of the negative binomial distribution, the Green's Coefficient, Poisson's Distribution and the Law of Taylor's potency. Along the evaluation period of the H.naevulus presence, the calculated indexes introduced an aggregated pattern of damage occurrence. That Indicates that the species presents a aggregated pattern distribution, even in low incidence.(AU)