Wisconsin Study Evaluates Starter Fertilizer Effects
Greg Roth and Doug Beegle
Department of Crop and Soil Sciences
Penn State
February 1999
At the recent American Society of Agronomy meetings in Baltimore, University of Wisconsin researchers Larry Bundy and Todd Andraski reported on an extensive study of starter fertilizer on corn. The researchers compared no starter to a 10-20-20 application of N+P2O5+K2O/acre for the effects on corn growth and yield. They conducted 100 replicated field trials in 30 counties and on 64 farms during a three year period. The soils were predominately in the excessively high soil test category for P and K.
Averaged over all sites, they found a 4 bushel/acre advantage to the starter fertilizer. Yield responses varied from -10 bushels/acre to +42 bushels/acre. Yield increases were statistically significant at 10% of the sites at the 90% level of probability, 20% of the sites at the 80% level of probability, and 30% of the sites at the 70% level of probability. At the 90% level of probability, we would expect that the yield differences observed at a site would be due to the fertilizer 9 times out of 10. As a result, depending on what you considered a significant effect, the starter fertilizer increased yield in 10 to 30% of the cases on these high testing soils.
The researchers also calculated the number of times there was an economic response to starter. They assumed you needed a yield increase of 4.5 bushels/acre to pay for the starter. They found that the starter was economical on 40% of these sites. Averaged across all of the sites, the use of starters would not have been economical since the average yield increase of 4 bushels/acre was less than the 4.5 bushels per acre need to pay for the cost.
The researchers found that grain moisture averaged 24.7% without starter and 24.6% with starter. At sites where an economic response to starter occurred, the difference in grain moisture at harvest was greater: 25.2% without starter compared to 24.7% with starter. This suggests that one benefit of the starter was to accelerate plant growth and hasten plant maturity.
Bundy and Andraski then correlated various site characteristics to the economic responses to starter. They found that soil test K and hybrid maturity were two factors that contributed to the response the most. Subsequent analyses showed that when soil K levels were less than 140 PPM, the beginning of their excessive range for K, response was more likely. They also found that later hybrids responded more (53% of the time) than early (less than 100 day RM ) hybrids (33% of the time), especially when planted later in May.
These results suggest that starter fertilizer responses may be greatest for long season hybrids planted later in the season because of the potential effect of the starter to hasten maturity of the crop before frost and help it achieve near maximum yields. This is in contrast to conventional wisdom regarding starters that suggests the fields that are the best candidates for response would be those planted early.
Most of the results of this study are consistent with our observations in Pennsylvania except for the increased response on later planted corn. This might be more a pronounced effect in Wisconsin due to the shorter growing season there than it would be in many of the longer season areas of Pennsylvania. Nevertheless, this is an area where some research is needed. This study confirmed some of our observations on these high testing P and K soils: the yield response varies considerably from site to site, the grain moisture is sometimes reduced and there is often a plant growth advantage to using a starter. The study also points out a key consideration in using starters: identifying situations that are most responsive is a key to using starters economically on these high testing soils.