Estimate of genetic diversity in germplasm of elephant grass accessions in Brazil using the Gower algorithm
Management of variability in germplasm banks is essential for genetic improvement, so that the breeder can estimate the genetic similarity between cultivars, as well as maintain genetic diversity in breeding programs. Elephant grass is a forage crop plant native to Africa of great socio-economic and environmental importance; it can be used for animal feed and for bioenergy production. Understanding the genetic variability of elephant grass is essential for breeding programs. In this context, we examined the genetic divergence of elephant grass accessions using the Gower algorithm. The experiment was conducted under field conditions in the municipality of Campos dos Goytacazes-RJ, Brazil. All 85 elephant grass accessions belonging to the Active Elephant Grass Germplasm Bank of Embrapa Gado de Leite were included. These genotypes are commercial varieties from various countries. They were evaluated for morphoagronomic, morphological and phenological characteristics. The experimental design was randomized blocks with two repetitions. The plots were composed of 5.5-meter rows, with 2.0-meter spacing between the planting rows, totaling 11.0 m2. The useful area was a sample in the center of the plot. We generated an illustrative dendrogram, obtained by the UPGMA method and the Tocher clustering, based on the Gower algorithm. Data were examined by means of the GENES statistical programs and the R program. According to the dissimilarity matrices based on the Gower algorithm, the genetic distances varied between 0.08 and 0.56, and the mean distance of the 85 evaluated accessions was 0.25, suggesting, consequently, that there is wide genetic variability between the accessions. Of the 85 genotypes, seven presented genetic distances smaller than 0.1, being indicative of duplicates in the germplasm bank, which could be eliminated without risk of loss of genetic variability.