INVESTIGADORES
PEICHOTO Maria elisa
congresos y reuniones científicas
Título:
PREDICT AREAS OF OCCURRENCE OF Lonomia obliqua WALKER 1855 (SATURNIIDAE, HEMILEUCINAE) IN THE PROVINCE OF MISIONES, ARGENTINA
Autor/es:
MARÍLIA MELO FAVALESSO; MILENA GISELA CASAFÚS; MARÍA ELISA PEICHOTO; ANA TEREZA BITTENCOURT GUIMARÃES
Lugar:
Buenos Aires
Reunión:
Congreso; Reunión Conjunta de Sociedades de Biociencias; 2017
Institución organizadora:
Sociedades de Biociencias
Resumen:
The Lonomia obliqua moth is a species of broad medical interest because its larvae are etiological agents of lonomism, a form of erucism which can lead the patient to death. Accidents have been common in Brazil since the end of the 80s, appearing in the province of Misiones (Argentina) by the beginning of this millennium. In order to find areas of high suitability for L. obliqua in the province of Misiones, Species Distribution Modeling algorithms were used: Bioclim, Gower, Mahalanobis, Maxent and SVM. For the model, 19 biogeoclimatic variables, the mean annual solar radiation and 9 soil variables were used. In order to avoid multicollinearity of the variables, a Principal Component Analysis was performed to create new orthogonal variables, resulting in seven new variables (cumulative variance> 90%). For the extraction of the niche information, 38 points of occurrence from Misiones and Brazil were selected and partitioned by the jackknife n-1 technique. The background contemplated the geographic extension that refers to the biomes where the presences of L. obliqua were obtained: Pampa, Atlantic forest, Caatinga and Cerrado. From it, 10% of the information was sampled randomly to be considered as pseudo-absences in the model. The Ensemble technique was also applied from the frequency for the construction of a map that shows the appropriate areas in concordance among the models. The Ensemble took place from the threshold for the minimal presence of the species. The model was adequate (sensitivity=99%) and better than a null model (p