• Refine Query
  • Source
  • Publication year
  • to
  • Language
  • 1
  • Tagged with
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • About
  • The Global ETD Search service is a free service for researchers to find electronic theses and dissertations. This service is provided by the Networked Digital Library of Theses and Dissertations.
    Our metadata is collected from universities around the world. If you manage a university/consortium/country archive and want to be added, details can be found on the NDLTD website.
1

Padr?es espaciais e temporais da fenologia de tr?s esp?cies arb?reas na regi?o do Alto Jequitinhonha

Magalh?es, Mariana Rodrigues 26 July 2013 (has links)
Submitted by Rodrigo Martins Cruz (rodrigo.cruz@ufvjm.edu.br) on 2015-01-07T11:11:35Z No. of bitstreams: 2 mariana_carolina_magalhaes.pdf: 2267644 bytes, checksum: d8a5e80ce31e3ecca7f32cb94d6d755b (MD5) license_rdf: 23898 bytes, checksum: e363e809996cf46ada20da1accfcd9c7 (MD5) / Approved for entry into archive by Rodrigo Martins Cruz (rodrigo.cruz@ufvjm.edu.br) on 2015-01-07T11:12:01Z (GMT) No. of bitstreams: 2 mariana_carolina_magalhaes.pdf: 2267644 bytes, checksum: d8a5e80ce31e3ecca7f32cb94d6d755b (MD5) license_rdf: 23898 bytes, checksum: e363e809996cf46ada20da1accfcd9c7 (MD5) / Approved for entry into archive by Rodrigo Martins Cruz (rodrigo.cruz@ufvjm.edu.br) on 2015-01-07T11:12:17Z (GMT) No. of bitstreams: 2 mariana_carolina_magalhaes.pdf: 2267644 bytes, checksum: d8a5e80ce31e3ecca7f32cb94d6d755b (MD5) license_rdf: 23898 bytes, checksum: e363e809996cf46ada20da1accfcd9c7 (MD5) / Made available in DSpace on 2015-01-07T11:12:17Z (GMT). No. of bitstreams: 2 mariana_carolina_magalhaes.pdf: 2267644 bytes, checksum: d8a5e80ce31e3ecca7f32cb94d6d755b (MD5) license_rdf: 23898 bytes, checksum: e363e809996cf46ada20da1accfcd9c7 (MD5) Previous issue date: 2013 / Este trabalho teve como objetivo estudar o comportamento de tr?s esp?cies do Cerrado em tr?s ?reas distintas com base na rela??o com as condi??es meteorol?gicas e utilizar Redes Neurais Artificiais (RNAs) para prever a ocorr?ncia de eventos fenol?gicos. As esp?cies Caryocar brasiliense, Bowdichia virgilioides e Plathymenia reticulata foram selecionadas devido ao seu potencial em recuperar ?reas degradadas e multiprodutos. Os locais de estudo para a sazonalidade foram os munic?pios de Diamantina (DIA), S?o Gon?alo do Rio Preto (SGRP) e Itamarandiba (ITA) e para as RNAs foram DIA e ITA. Foram selecionados 12 indiv?duos de cada esp?cies em cada local. As visitas fenol?gicas foram realizadas quinzenalmente entre 2008 e 2011, sendo registradas as fenofases seguindo a escala de Fournier. Os dados meteorol?gicos foram fornecidos pelas Esta??es Climatol?gicas Principais de Diamantina e Itamarandiba. Para a rela??o entre as vari?veis meteorol?gicas e eventos fenol?gicos utilizou-se correla??es de Spearman (rs) e estat?stica circular para testar a ocorr?ncia de sazonalidade. As fenofases referentes ? folhagem e ?s fases reprodutivas foram quantificadas pelo do ?ndice de atividade. Nas tr?s ?reas os eventos relacionados ? folhagem e flora??o das tr?s esp?cies apresentaram sincronia e as fenofases referentes ? frutifica??o apresentaram sincronia para a Bowdichia virgilioides e Plathymenia reticulata, j? o Caryocar brasiliense apresentou sincronia entre DIA e SGRP. A maioria dos eventos apresentaram sazonalidade pelo menos para um dos per?odos estudados. As esp?cies apresentaram rela??o com as vari?veis meteorol?gicas, sendo que os intervalos parciais apresentaram tend?ncia diferente dos per?odos totais demostrando ? necessidade de uma s?rie hist?rica mais longa. Para a an?lise das RNAs, os dados referentes ?s fases reprodutivas foram processados para obten??o da estimativa atual e futura dos valores referentes ? porcentagem dos eventos fenol?gicos reprodutivos. As redes apresentaram-se como um grande apoio no estudo da fenologia, uma vez que com um banco de dados menor que tr?s anos os valores estimados foram muito pr?ximos dos observados principalmente quando utilizada uma fenofase por rede. A estimativa futura indicou a possibilidade de previs?o de eventos fenol?gicos possuindo apenas dados meteorol?gicos. / Disserta??o (Mestrado) ? Programa de P?s-Gradua??o em Ci?ncia Florestal, Universidade Federal dos Vales do Jequitinhonha e Mucuri, 2013. / ABSTRACT This study aimed to understand the behavior of three species of the Cerrado in three distinct areas based on the relationship between meteorological conditions and use Artificial Neural Networks (ANN) to predict the occurrence of phenological events. The species Caryocar brasiliense, Bowdichia virgilioides and Plathymenia reticulata were selected because of their potential to recover degraded areas and multiproducts. The study locations for seasonality were the cities of Diamantina (DIA), S?o Gon?alo do Rio Preto (SGRP) and Itamarandiba (ITA) and for ANN were DIA and ITA. We selected 12 individuals of each species at each site. The phenological visits were every two weeks between 2008 and 2011 and the were registered phenophases following scale Fournier. Meteorological data were provided by Climatological Stations Main Diamantina and Itamarandiba. For the relationship between the meteorological variables and phenological events used Spearman correlations (rs) and circular statistics to test the occurrence of seasonality. The stages refer to the foliage and the reproductive phases were quantified using the activity index. In the three areas the events related foliage and flowering of three species presented synchrony and phenophases related to fruiting presented synchrony for Bowdichia virgilioides and Plathymenia reticulata, already the Caryocar brasiliense presented synchrony between the DIA and SGRP. Most events presented seasonality in the at least one of the study periods. The species presented relationship with meteorological variables studied, and the partial ranges presented different tendency of total periods demonstrating the need for a longer time series. For the analysis of the ANN, the data relating to reproductive phases of the studied species were processed to obtain estimates of current and future values ??for the percentage of reproductive phenological events. The networks presented as a lot of support in the study of phenology, because with a database of less than three years, the estimated values ??were very close to those observed especially when used one phenophase by network. The estimated future indicated the possibility of predicting phenological events having only meteorological data.

Page generated in 0.1006 seconds