Construction of a classification tree for the study of the variability present in a working collection of Pisum sativum L.?

ESPÓSITO MA ; DUARTE C; COINTRY E,; GATTI I

BIOCELL

INST HISTOL EMBRIOL-CONICET

Año: 2020 vol. 44 p. 12 - 12

0327-9545

In breeding programs is necessary to have characterized and evaluated the working collections of germplasm to allow choosing parents with desirable characteristics to form, by crosses, segregating generations. Classification trees present a non-parametric statistic tool with high robustness in the presence of outliers and easy interpretation in the analysis of the variability present in these collections, allowing assigning accessions to homogeneous groups with defined characteristics, and are useful for the initial exploration of a large number of data. In order to systematize the available variability, ninety accessions of the working collection of peas (Pisum sativum L.), were evaluated during 2017 and 2018 in the Campo Experimental Villarino, FCA-UNR in Zavalla. A randomized complete block design was used with two repetitions, with experimental plots of 20 plants each. The analyzed characters were: days up to 50% of flowering plants (DF), first knot with pod (PNV), days to harvest (DC), plant height (Alt), pod length (LV), pod width (AV), grain caliber (C), number of pods (NV) and grains (NG) per plot and plot yield (RE). Highly significant differences (P