Térmitas de duas áreas de floresta atlântica brasileira: uma análise do desempenho de estimadores não paramétricos

• Ano de defesa: 2013
• Autor(a) principal: Ernesto, Matilde Vasconcelos
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal da Paraíba; Brasil; Engenharia e Meio Ambiente; Programa de Pós-Graduação em Ecologia; UFPB
• Palavras-chave em Português: Isoptera; Protocolo de Transecto; Enviesamento; Precisão; Acurácia; Transect Protocol; Bias; Precision; CIENCIAS BIOLOGICAS::ECOLOGIA
• Link de acesso: https://repositorio.ufpb.br/jspui/handle/tede/7617

Estimators of species richness are alternatives to characterize the local biodiversity, facing the limitations of complete fauna inventories. Due to continuous urbanization and deforestation of the Atlantic Forest, studies on local biodiversity are important as records of fauna and as a way to subsidize conservation measures. It was evaluated the effectiveness of five nonparametric estimators of species richness, from the calculation bias, precision and accuracy of termites of Floresta Nacional da Restinga de Cabedelo (FNRC), and the Área de Preservação Permanente Mata do Buraquinho (APPMB). It was utilized a standardized sampling protocol of termites, which consists of five 5 x 2m plots distributed along six transects of 65 x 2m. Five protocols were applied in FNRC and 12 in APPMB totaling a sampling effort of 5100m² and x 510h per person, and with uniques and duplicates absent in the data. Forty-five termite morphospecies were found, belonging to at least 26 genera and three families, exceeding any records ever published throughout the biome of Atlantic Forest. It was found that the Jackknife1 and Jackknife2 should only be used for databases with a number of uniques over 25% of the richness, and it was observed that the use of the Bootstrap estimates and Chao2 are indicated in the remaining cases, when the number of rare species is lower than this. It is suggested the need of further evaluation of ICE performance before indicating it to studies on the diversity of termites under any conditions. It was sampled around 65-70% of the real richness of the areas with the application of a protocol and there was a robust representation of the taxonomic composition and trophic proportions, stressing the efficiency of this method to determine the termites diversity in two areas of Atlantic Forest investigated.