/AnMtgsAbsts2009.55210 Is There A Greater Risk of Low Soybean Yield From Low Seeding Rates? Analyses From Illinois Seeding Rate Studies.

Wednesday, November 4, 2009: 2:00 PM
Convention Center, Room 333, Third Floor

Vince Davis, Emerson Nafziger and Ralph Esgar, Crop Sciences, Univ. of Illinois, Urbana, IL
Abstract:
The seeding rate needed to maximize crop yields and economic return is an important farming decision.  Because established plant populations differ from seeding rate, farmers increase seeding rate to account for unfavorable environmental conditions.  Historically, the cost of soybean seed was minimal due to the use of public cultivars and the saving of seed by farmers themselves.  Therefore, many farmers planted up to twice as many seeds as might be required to maximize yield under favorable conditions.  Seed costs have rapidly increased due to privately-developed cultivars and the patent protection of genetically-modified traits.  This has renewed interest in lowering seeding rates to maximize economic return to seed.  The objective of this study was to evaluate soybean yields from seeding rates and calculate economically optimal rates.  Rates of 124,000, 247,000, 371,000, and 494,000 seeds ha-1 were used in replicated trials planted in 76-cm rows at 33 diverse locations over four years in Illinois.  Soybean cultivars were adapted to maturity zone of particular sites.  Site and cultivar showed no interaction with seeding rate at P = 0.1.  Over all sites, yields at 124,000 seeds ha-1 averaged 3,760 kg ha-1, and successive increments of 124,000 seeds increased yield by 350, 80, and 20 kg ha-1, thus showing a rapidly diminishing return to added seed.  At current prices of soybean seed and soybean grain, economically optimal seeding rates based on these averages are about 280,000 to 320,000 seeds ha-1.  These results provide support for modest lowering of seeding rate by many producers.  One question, though, is under ‘normal’ conditions, is there a greater risk of low soybean yields form low seeding rates?  The coefficient of variation was uniform for all four rates, so the question is, how many ways should agronomists analyze data to estimate the risks’ associated from data like these?