Temporal profiles of pod production and pod set in soybean

The distribution of flower and pod production during flowering may be an important determinant of pod and seed number in grain crops. We characterized the dynamics of small pod production and survival to maturity on indeterminate and determinate soybean [Glycine max (L.) Merrill)] cultivars growing in the field or greenhouse. Two soybean cultivars (maturity group IV, indeterminate and determinate) were grown in the field near Lexington. KY (38 degrees N latitude) in 2001 and 2002 in 0.76 cm rows using late May and late June (2002 only) planting dates, and normal (24 plants m(-2)) and low (9 plants m(-2), 2002 only) plant populations. Cultivar Elgin 87 (indeterminate, maturity group II) was grown in a greenhouse in 3.0 L pots with one plant per pot. All unmarked pods that were >= 10 mm long were marked with acrylic paint at the base of the pod at 3-day intervals. Paint color was changed at each marking to provide a temporal profile of pod production and pod survival. The pod production (marked pods) period was longer in the indeterminate cultivar (nearly 50 days after R1) than the determinate cultivar (<= 40 days after RI). Delayed planting shortened the pod-production period, but a two- to three-fold difference in pods per plant, created by changing plant population, did not affect it. The temporal distribution of small pods that survived to maturity (full sized pods with at least one normal seed) closely followed the distribution of pod production in all experiments. Some surviving pods initiated growth after the beginning of seed filling (i.e., between growth stage R5 and R6), but most of the pods were initiated in a much shorter interval (up to 84% were initiated in <40% of the period) before R5. Abortion of pods >10 mm long was relatively low (20-30%), so production of a pod >= 10 mm long seems to be a key event in the pod set process. The average length of the pod set period at individual nodes on the main stem was larger for the determinate cultivar (14 days) than for the indeterminate (9 days), so the longer total period in the indeterminate cultivar resulted from the delay in initating pod production at the upper nodes on the main stem. Temporal profiles of pod production and pod set seem to be more sensitive to changes in flower and main stem node production than to changes in photosynthesis per plant (created by varying plant population). These results provide some of the information needed to integrate time into models predicting pod and seed number. (c) 2005 Elsevier B.V. All rights reserved.