Yield Components as Indirect Selection Criteria for Late‐Planted Soybean Cultivars

Cultivar selection for late‐planted soybean [ Glycine max (L.) Merr.] in the wheat–soybean doublecropping system is an important production problem. Top‐yielding cultivars for late plantings could be selected more efficiently by identification of yield components that indicate top yield, compared with the traditional combine‐harvested plot yield method. Our objective was to identify yield components that could be used as indirect selection criteria to identify top‐yielding cultivars for late planting dates. A 2‐yr study (1998 and 1999) with 26 cultivars was conducted at a mid‐July planting at Baton Rouge, LA (30° N, 90° W). A similar study involving 27 different cultivars was conducted for 1 yr at Los Oliveros, Santa Fe Province, Argentina (32°48′ S, 62° W), planted in early January 1999. Experimental designs were randomized complete blocks with four replications and one factor (cultivar). Data were obtained on combine‐harvested plot yield, seed m −2 , seed size, seed per pod, pods m −2 , pods per reproductive node, and reproductive node m −2 . Across years, yields at Baton Rouge ranged from 1183 to 2992 kg ha −1 , while yields in Argentina ranged from 1688 to 2809 kg ha −1 . Yield at Baton Rouge increased with maturity group, whereas in Argentina there was no relationship between yield and maturity group. For both phenotypic and genotypic levels, selection for either seed m −2 or pods m −2 identified top‐yielding cultivars, although seed m −2 was more accurate.