dc.contributorJoao Renato Stehmann
dc.contributorAdriano Pereira Paglia
dc.contributorAlexandre Salino
dc.contributorAry Teixeira de Oliveira Filho
dc.contributorMarinez Ferreira de Siqueira
dc.creatorLuciana Hiromi Yoshino Kamino
dc.date.accessioned2019-08-13T02:16:25Z
dc.date.accessioned2022-10-04T00:53:06Z
dc.date.available2019-08-13T02:16:25Z
dc.date.available2022-10-04T00:53:06Z
dc.date.created2019-08-13T02:16:25Z
dc.date.issued2009-11-13
dc.identifierhttp://hdl.handle.net/1843/TJAS-89QQCQ
dc.identifier.urihttp://repositorioslatinoamericanos.uchile.cl/handle/2250/3837251
dc.description.abstractABSTRACTKAMINO, L.H.Y. (2009). Potential Models of Species Geographical Distribution: anessay with threatened species from Atlantic Forest. Ph.D. Thesis Instituto CiênciasBiológicas, Universidade Federal de Minas Gerais, Belo Horizonte. Advisor: JoãoRenato Stehmann.Potential Models of Species Distribution have been used as important tools inconservation biology, producing maps of expected distribution of rare, endemic andthreatened species. One of the most used algorithms is Maxent, that can predictdistribution with a few occurrence data. The main goal of this work is to evaluate thepredictive models generated by Maxent, as a tool to help finding new populations ofplants threatened of extinction in the Atlantic Forest. In the first chapter two tree specieswith large extent of occurrence are evaluated (Melanoxylon brauna Schott e Dalbergianigra (Vell.) Allemao ex Benth., Fabaceae), and the data of occurrence acquired fromherbarium sheets were of low geographic precision. To generate the models were usedthe Maxent LPT and T10 methods. The resulting maps exhibit na acceptable approachof the species potential distribution, with less then 25 ocurrence data. The set rate for M.brauna was of 63% (LPT) and 83% (T10), and for D. nigra of 40% (LPT) and 75%(T10), respectively. According to the results of the field surveys, the maps generated bythe T10 algorithm had a better accuracy than the LPTs. Considering the level offragmentation and the history of over exploitation of the occurrence area of the species,they are both recommended to be maintained as Vunerable, according to IUCN criteriafor threatened species. The second chapter evaluate the predictive model for one herbspecie of restricted distribution (Petunia mantiqueirensis T. Ando & Hashim.,xixSolanaceae), with only eight known registries, but with more precise references than thetrees of the first chapter. The models showed a set rate of 87%, what are consideredstatistically significant. Nine new populations of P. mantiqueirensis were found inpredicted areas of the model. The Potential Models of Geographical Distributionassociated with field trips were useful to enlarge the knowledge about the occupancyarea of threatened plant species.KEYWORDS: Potential Models of Species Geographical Distribution, Maxent,Threatened Species, Atlantic Forest, Melanoxylon brauna, Dalbergia nigra, Petuniamantiqueirensis.
dc.publisherUniversidade Federal de Minas Gerais
dc.publisherUFMG
dc.rightsAcesso Aberto
dc.subjectModelos de distribuição geográfica potencial
dc.subjectFloresta Atlântica
dc.subjectMAXENT
dc.subjectespécies ameaçadas de extinção
dc.subjectMelanoxylon brauna
dc.subjectPetunia mantiqueirensis
dc.subjectDalbergia nigra
dc.titleMODELOS DE DISTRIBUIÇÃO GEOGRÁFICA POTENCIAL:APLICAÇÃO COM PLANTAS AMEAÇADAS DE EXTINÇÃO DAFLORESTA ATLÂNTICA
dc.typeTese de Doutorado


Este ítem pertenece a la siguiente institución