Quantification of the genetic divergence among guava accessions using Ward-MLM strategy/Quantificacao da divergencia genetica entre acessos de goibeira por meio da estrategia Ward-MLM

The present study aimed at evaluating the genetic divergence among 138 guava accessions coming from the Germplasm Collection of the Universidade Estadual do Norte Fluminense Darcy Ribeiro (UENF) based on morphological, agronomic and physic-chemical traits by Ward - Modified Location Model (MLM). Thirteen descriptors were evaluated, five qualitative (flesh color, fruit surface, shape at stalk end, neck width and uniformity of the flesh color) and eight quantitative (fruit weight, fruit diameter, fruit length, pulp revenue, soluble solids contents, total titratable acidity, total soluble solids/total titratable acidity ratio and ascorbic acid). High genetic variability was detected considering morphological, agronomic and physicchemical traits in the 138 guava accessions studied. The procedure of the likelihood function showed that the ideal numbers of groups was eight, with an increment value of 67.51. Group III was considered the most distant, while groups I, II, IV, V and VI the closest. The Ward- MLM procedure is a useful tool for detecting genetic divergence and cluster accessions using both qualitative and quantitative variables.