Banca de DEFESA: LEONARDO CAMPOS VELOSO

Uma banca de DEFESA de MESTRADO foi cadastrada pelo programa.
STUDENT : LEONARDO CAMPOS VELOSO
DATE: 13/10/2020
TIME: 08:30
LOCAL: Sala virtual no google meet
TITLE:

MODELING OF MORTALITY IN A FOREST MANAGED IN THE CITY OF PARAGOMINAS, PARÁ

KEY WORDS:

sustainability; timber harvesting; damaged trees; mortality modelling

PAGES: 38
BIG AREA: Ciências Agrárias
AREA: Recursos Florestais e Engenharia Florestal
SUMMARY:

Timber harvesting, although well planned, causes damage to the remaining trees, which can lead to the mortality of some individuals. Understanding the circumstances that influence the mortality of these trees can help in planning logging activities, in order to reduce this impact and contribute to the sustainability of forest management. In this context, this work aimed to study the mortality of trees after logging, through the characterization of the mortality dynamics, estimation of individual mortality of the remaining trees and projection of the average accumulated tree mortality by damage category, for a 35-year cutting cycle. The study was carried out in an experimental area located in a forest management unit at Fazenda Rio Capim, municipality of Paragominas, state of Pará. 18 permanent plots of 1 ha each were established and a census of trees with a diameter at breast height (DBH) ≥ 20 cm was performed. Trees with 10 cm ≤ DBH <20 cm were measured in two contiguous 25m x 25m sub-plots chosen at random within each plot. Seven measurements were made: in 2004 (before logging), 2005 (after logging) and the other every two years until 2014. The mortality model was developed by "logit" according to individual and population variables of the managed forest. To fit this model, thirteen plots with DBH trees ≥ 20 cm were randomly chosen, the remaining five were used for validation by graphical comparison and Student t-means comparison tests. The projected cumulative mortality for each damage category was calculated based on the average values found in the individual mortality model. It was found that the annual mortality rate of trees was maximum in the year following logging, with values of 16.8% for trees with 10≤DBH<20cm and 9.9% for those with DBH ≥ 20cm, compared to the data of the pre-logging inventory. The mortality rate decreased over the years after the logging, reaching a value of 3% in the tenth year for these two sizes of trees.  It was observed by means of a comparison test of means at 95% confidence level that the mortality of damaged trees is more expressive than that of undamaged trees and that from the fourth year after the logging the average rate between them is equal. The individual tree mortality model indicated three explanatory variables: category of damage (DC), density of trees per plot (N) and time elapsed after logging (TEAL). Among the damage categories with the highest probability of mortality in the first year after logging, trees with broken trunk at height <3m (64%), uprooted (52%), and lean trees with inclination angle of 20 °-45 ° (30%).

BANKING MEMBERS:
Interno - 745.945.679-91 - ADEMIR ROBERTO RUSCHEL - EMBRAPA
Externo à Instituição - EDSON MARCOS LEAL SOARES RAMOS - UFPA
Interno - 2315025 - HASSAN CAMIL DAVID
Presidente - 142.060.489-91 - JOSE NATALINO MACEDO SILVA - Oxford