PREDICTING SOYBEAN SEED-GERMINATION DURING WAREHOUSE STORAGE

Germination of soybean (Glycine max (L.) Merrill) seeds declines more rapidly during storage than other grain crops. This investigation was conducted to determine if the Ellis and Roberts (1980a) seed deteriora tion model could be used to predict changes in soybean seed germinatio n in the variable temperature and moisture conditions encountered duri ng warehouse storage. Seventeen seed lots that were produced in 1987 w ere stored for 22 months in multiwall paper bags in seed warehouses at four locations in Kentucky and Indiana, U.S.A. Warehouse temperature was monitored and seed moisture and germination was determined at thre e month intervals. A controlled seed deterioration test was conducted at 40-degrees-C and 15% seed moisture (fresh weight basis) to establis h.an initial seed viability constant (Ki) for all seed lots prior to s torage. This constant was used with average storage temperature and se ed moisture in the Ellis-Roberts seed deterioration model to predict s eed germination at monthly intervals. The model accurately predicted g ermination (+/- 10 percentage points) after 16 months of storage for 1 6 of the 17 seed lots compared across the four storage locations. The model also accurately predicted the germination of 15 of 17 seed lots after 4 months storage. Two seed lots (Beck 333 and Lawrence) which we re not accurately predicted had high levels of mechanical seed-injury which were not accounted for by the model. The model predicted the ger mination (+/- 15 percentage points) of the lowest quality seed lot, Ri pley, which declined from an average of 93 (Dec. 1987) to 49% (Oct. 19 89). The model offers potential for predicting the germination of seed s in normal warehouse storage, if warehouse temperature and seed moist ure can be determined.