Optimizing the use of morpho-agronomic descriptors of cassava in multivariate analysis

Abstract

This paper aims to evaluate and characterize the cassava accessions that are being introduced into the germplasm collection of UNEMAT present in the experimental field EMPAER at the city of Cáceres, Mato Grosso, Brazil. It was also analyzed the use of morpho-agronomic descriptors as its efficiency with the use of multicategory variables through multivariate analysis. In each accession were analyzed and assessed 40 morphological and agronomic traits, which are considered essential to the tests for distinctness, uniformity and stability of Manihot accessions ssp. The study showed that excluding the group of least descriptors obtained results from about 92% similarity compared with the use of all words. Results obtained from 38 accessions of cassava germplasm bank of UNEMAT Caceres campus, show genetic variability.