Niche Modeling of the economical important Mahanarva species in South and Central America (HEMIPTERA, CERCOPIDAE)

• Main Author: Sch?bel, Christian
• Other Authors: Carvalho, Gerv?sio Silva
• Format: Others
• Language: English
• Published: Pontif?cia Universidade Cat?lica do Rio Grande do Sul 2018
• Subjects: Spittlebug; Sugarcane; Pasture; Agricultural Damage; Cana de A??car; Pastagem; Dano Agr?cola; CIENCIAS BIOLOGICAS::ZOOLOGIA
• Online Access: http://tede2.pucrs.br/tede2/handle/tede/8004

Submitted by PPG Zoologia ([e-mail]) on 2018-04-20T12:08:51Z No. of bitstreams: 1 Disserta??o corre??o final - Christian.pdf: 2735367 bytes, checksum: 73fdc67c9898d886fdbd0e0cce1dfad3 (MD5) === Approved for entry into archive by Caroline Xavier ([e-mail]) on 2018-05-08T16:47:52Z (GMT) No. of bitstreams: 1 Disserta??o corre??o final - Christian.pdf: 2735367 bytes, checksum: 73fdc67c9898d886fdbd0e0cce1dfad3 (MD5) === Made available in DSpace on 2018-05-08T16:53:43Z (GMT). No. of bitstreams: 1 Disserta??o corre??o final - Christian.pdf: 2735367 bytes, checksum: 73fdc67c9898d886fdbd0e0cce1dfad3 (MD5) Previous issue date: 2018-02-22 === Conselho Nacional de Pesquisa e Desenvolvimento Cient?fico e Tecnol?gico - CNPq === Mahanarva fimbriolata, M. spectabilis, M. liturata and M. posticata (Hemiptera: Cercopidae) s?o conhecidas como pragas de planta??es de cana-de-a??car e pastagem em todo Brasil. Por alimentarem-se diretamente da seiva das plantas, esses cercop?deos causam fitotoxicidade e devido a isso diminuem a produ??o. A modelagem da distribui??o de esp?cies permite analisar a poss?vel occurencia das quatro esp?cies na Am?rica do Sul e Central. Para criar modelos de distribui??o de esp?cies foram utilizados em R, os algoritmos Bioclim, Domain, diferentes modelos lineares generalizados e Maxent. Nesses modelos foram utilizadas vari?veis bioclim?ticas atuais e futuras, al?m da eleva??o e outras vari?veis agr?colas. As vari?veis clim?ticas futuras s?o para os anos 2050 e 2070 com diferentes repentant concentration pathways. As esp?cies apresentam habitats adequados em diferentes pa?ses da Am?rica do Sul e Central, onde as planta??es de cana-de-a??car s?o abundantes. Os resultados das an?lises clim?ticas futuras n?o apresentaram diferen?as em rela??o ?s an?lises clim?ticas atuais. No geral, o algoritmo Maxent mostrou os maiores valores de AUC e o Bioclim os menores. As vari?veis que mais contribu?ram para os modelos s?o: eleva??o, isothermality e diferentes vari?veis de precipita??o. As mudan?as clim?ticas e ciclos de vida de insetos adicionais n?o t?m impacto em habitats adequados dos insetos. Em geral, o Maxent ? o melhor algoritmo para realizar modelos de distribui??o de esp?cies com um n?mero baixo de pontos de ocorr?ncia e an?lises de mudan?as clim?ticas. === Mahanarva fimbriolata, M. spectabilis, M. liturata and M. posticata (Hemiptera: Cercopidae) are known pests for sugarcane and pasture plantations throughout Brazil. By direct sap feeding on the plants they cause phytotoxicity and due to this they decrease the production of plantations. With species distribution modeling it is possible to analyze the possible occurence of the four species in South and Central America. To create species distribution models the algorithms Bioclim, Domain, different generalized linear models and Maxent were used in R. For those models current and future bioclimatic variables as well as elevation and other agricultural variables were used. The future climatic variables are for the years 2050 and 2070 with different repentant concentration pathways. The species show suitable habitats in different countries in South and Central America where sugarcane plantations are abundant. The results of the future climate analyzes do not show differences compared to the current climate analyzes. Overall the Maxent algorithm showed the highest AUC scores and Bioclim the lowest. The variables which contributed the most to the models are elevation, isothermality and different precipitation variables. Climate change and therefore additional insect lifecycles do not have an impact on the insect?s suitable habitats. Overall Maxent is the best algorithm to perform species distribution models with a low number of occurrence points and for climate change analyzes.