An experiment using five early cycle, eight medium cycle, and five semilate and late cycle soybean cultivars was conducted in Cruz Alta, in the growing seasons of 1994/95 up to 1997/98, and Passo Fundo, in 1994/95 and 1997/98, Planalto Médio region of the State of Rio Grande do Sul, Brazil, to evaluate the behavior in seedings made between 10 and 30 December. An experimental design of complete randomized blocks with three replications was used. To avoid interference by plants of different cultivars, neighboring plots were kept 1.0m apart and also one row in each side of the plots was eliminated prior to harvest. Joint variation analysis for yield was made by splitting the cultivars in early cycle, medium cycle, semilate and late cycle cultivars, and between cycles. The same procedure was followed for the cultivar x environments interaction. Highly significant differences were obtained between environments and among cultivars. Nevertheless, no significant differences were found between cultivar cycles nor between cultivars x environments interaction. Likewise, no limitations related to plant architecture were observed. The results show that the soybean cultivars available currently have similar yield potential when sowed from December 10 to December 30, regardless of their cycle.
Glycine max; seeding date; cultivars
