Seleção de progênies e estimativas de parâmetros genéticos em Ageratum conyzoides L. (ASTERACEAE) para caracteres morfo-agronômicos

Abstract

Mentrasto, an Asteraceae originated from South and Central America (The Caribbean), holds analgesic and anti-inflammatory properties. The presence of active principles also allows different uses in popular medicine such as: anti-rheumatic, anti-diarrheal, febrifuge, carminative, hemostatic and healing. This work had the objective to start a genetic improvement program aiming at increasing the amount of essential oil and other characters of agronomic interest, by means of agronomic characterization of Ageratum conyzoides L accessions and progenies selection, by means of massal selection with progenies test. The experiment was carried out at the Experimental Unit of Horto Florestal (Universidade Estadual de Feira de Santana - UEFS). The accessions were collected from distinct rural areas, summing up an original population of 700 A. conyzoides L plants. It was applied the agronomic characterization of the accessions, following the Completely Random Design (CRD), with two distinct tests. The following evaluation was done: the plant height (cm), the stalk diameters (øC) in (mm) and the plant diameter (øP) in (cm), number of branches, fresh and dry matter (g) (FM) and (DM), leaf characterization of the accessions (leaf and petiole lengths (cm), leaf width (cm), fresh dry and dry masses (g)), phenology, format of the top, growing habit. The data obtained were submitted to variance analysis and the averages were compared with Tukey test (p≤0,01 e 0,05), along with the genetic divergence study, adopting multivariate methods with the GENES program, besides the pre selection of 72 progenies. In the second stage of collection, the progenies were evaluated in experimental design in square lattice 9 X 8, with three repetitions and 9 plants/useful per portion representing each progenies, aiming at selecting 10% of the best progenies based on the characteristics previously cited. The variance analysis plan followed the randomized block design, the statistical analysis were achieved with statistical program Genes and a variance analysis as well as F test, the average and experimental variance coefficient were performed. The following genetic parameters were estimated: phenotypic (σ2f), genotypic (σ2g), and environmental (σ2e) variances, the broad heritability h2, genetic variance coefficient (CVg%), genetic variance coefficient /experimental variance coefficient ratio (CVg/CV), Intra Class correlation, phenotypic 13 (rf), environmental (ra), and genetic (rg) correlations. Concerning the agronomic characterization, in general, the variance coefficients for some characteristics were low, indicating a good experimental accuracy in both tests. On the contrary, others had 13 medium and high variance coefficients. There is a big phenotypic variability among the accessions for all the characteristics evaluated, what is very promising for a pre-selection work. There was formation of different genetic groups in the multivariate analysis. In the field test, four and three groups were formed in polyethylene sacks. The specimen presented a different agronomic behavior in both growing environments for the experimental conditions. After this study, it was possible to pre-select progenies of A. conyzoides L for the next stage. In the last stage, it is verified the significance among the accessions, with results showing plant genetic variability for some characteristics evaluated. The progenies present heritability between low and medium with a great influence of the environment upon the characters studied. The genotypic correlations M.F. and N.R. (0,99); M.S. and N.R (0,94); ØP and A.P (0,64) were very strong, indicating that the selection of a characteristic influences positively the other characteristics. There is a big variability among the progenies for the agronomic characteristics evaluated, with a great selections potential of a group of plants with good harvest characteristics, so as to improve the specie vegetal productivity. According to the heritability values considered between medium and high, along with the ratio genetic variance coefficient and the genetic variance coefficient/environmental variance coefficient medium and low, the selection of the genetic material must be judicious due to the huge environmental influence. After these results, it was possible to select progenies of Ageratum conyzoides L with interesting agronomic characteristics for future selection of superior genetic material.