Seemingly unpredictable weather conditions each spring inevitably bring up questions on placement of fertilizer with the seed.

Starter fertilizer has played an important role in nutrient management in corn in Minnesota. However, tools for deciding on how much that can safely be applied have not been widely available. While these tools can be used common sense is still needed in making a decision on what should be done.

Dr. Ron Gelderman at South Dakota State University developed a decision guide for multiple fertilizer sources and crops that can be helpful in deciding on rates to apply.  The guide is based on a Microsoft Excel spreadsheet that allows the used to vary soil moisture conditions, soil texture, and accepted level of risk for stand loss.  All three factors are important in making a decision on rate.

The guide also includes a variety of crops.  Popup fertilizer use on soybeans is always a question on some of the soils in western MN that have a higher potential of tie-up of phosphorus.  However, the use of pop-up fertilizer on soybeans IS NOT recommended due to the greater risk for seedling damage on soybean relative to other crops.  Knowing the tolerance of a specific crop to seed placed fertilizer is important due to the varying potentials for stand damage.

FERTILIZER SEED DECISION AID (link can be found in this page)

We recently finished up a greenhouse study using several liquid and dry fertilizer sources for seed placement on corn. Through funding provided by the MN Agricultural Fertilizer Research and Education Council we were not only able to look at the effects on emergence, but also the effects on nutrient uptake to obtain a better picture of optimum rates for uptake and maximum rates that could be applied before stand damage occurred. 

Three soils were used in this study, a Le Sueur clay loam, Zimmerman fine sand, and a Floyd loam.  It is important to note that soil moisture was kept at 80% of field capacity to keep soil moisture as NOT LIMITING. Therefore, any data presented would be of a situation of the best case scenario and would not be representative of a dry spring.

Our attempt was to develop a single model that could be used over a number of fertilizer sources to predict rates. We used two approaches: 1) based on fertilizer rate; and 2) based on the amount of nutrients applied. 

Our approach was to use and index value based on plants emerged after 14 days and total above ground growth based on the total number of seeds planted. Both factors together should give a better picture of what was happening in the soil. One thing we did notice on some fertilizer materials that even if stand was not reduced, plant growth may be reduced. 

This was apparent with low rates of ammonium thiosulfate (ATS). The effect of ATS could be readily seen in the plant mass but not in the emergence. Both measurements were used to develop indexes that were based on the control where no fertilizer was applied (set at 100% relative value).

With approach 1, the fertilizer rate alone was not at all predictive of damage potential due to the relative injury potential of the various sources. The fertilizer salt index was developed as a way to gauge the effects of fertilizer against each other but not predict the overall rate. In order to weight the sources the salt index of the fertilizer source was multiplied times the rate of fertilizer applied. 

There still was some variability in the data, but the overall relationship between the salt index weighted rates appeared to work well across fertilizer sources. However, I do have some concerns which will be discussed when I present the data on predicted rates.

One factor that we wanted to study was the old rule of thumb of no more than 10 lbs N+K2O to be applied with the seed in medium and fine textured soils. When we factored in the amount of N+K2O in a prediction model the rate that returned 100% that of the control was right on 10 lbs N+K2O which validates the old rule of thumb. The value for the sandy soil was 4-5 lbs N+K2O, roughly half of the other two soils. 

Two exceptions were found to this rule. First, ATS rates predicted by this model are extremely high and SHOULD NOT be used. When the total amount of N and S were factored for ATS, both models returned similar rates but the rate model would be better since it is slightly more conservative in it's prediction. The thiosulfate ion in ATS does pose serious risk for damage, but when potassium thiosulfate (KTS) was tested it did not appear to have the same effect as ATS. 

Therefore, ATS should be considered differently than other sources and approached with extreme caution. I would always prefer to see ATS applied in a band on the soil surface to the side of the row to lessen the risk for stand loss. 

The other issue occurred when predicting the rate of a low salt source, 9-18-9. For 9-18-9, the K in the source does not appear to have the same effect and the N+K2O rule appears to be highly conservative when predicting rate. 

The two models predicted similar rates of 9-18-9 when only the N portion was considered without the K2O.  I still would err with some caution on some of the NPK mixes especially if urea is added to increase the nitrogen concentration. Urea can be extremely damaging especially in situations where high amount of ammonia (NH3) are liberated in the soil.

The table below shows the calculated rates for several sources used in the study based on 30" row spacing. 

As was mentioned before, some caution should be used when viewing this data since it was developed under highly controlled circumstances.  Since the model used was developed across all sources it predicts a non-zero application rate for all sources.  For ATS and urea, when models were generated for each product individually, no rate could be safely applied. 

However, the rates predicted are relatively low and could be used if some risk is acceptable.  The SDSU fertilizer decision guide is useful in visualizing the risk associated by modifying the tolerated stand loss in the worksheet.  As with any management decision, sound judgment should always be used. 

The Low and High rates in the table were developed based around the confidence interval of the mean and are meant represent the relative risk for damage to the seed (low end of the range, risk is lower but not zero; high end, risk is higher).  It does not mean that damage will occur, but damage will be more likely when rates are applied above the high end of the range in the table.  When using models such as these for products like ATS, AMS, urea, and KTS, values near the low end of the range would be a better choice since the potential for damage is high. 

The table below is meant as a general guideline for the damage potential for a fertilizer source and not present a direct recommendation on how much to apply since the exact rate that can be safely applied depends on many conditions at planting.

Table Abbreviations (AMS, ammonium sulfate; ATS, ammonium thiosulfate; APP, 10-34-0; DAP, 18-46-0; KCl, potash (0-0-60); KTS, potassium thiosulfate; MAP, 11-52-0).

Always err with some caution when applying fertilizer with the seed.  There is no 100% safe fertilizer source for seed placement. 

While some may believe a product is 100% safe, statements on safety are always predicated with the rate applied. 

At low rates, some fertilizer sources may pose a lower risk for stand damage but there always is some risk associated with this practice depending on soil moisture conditions and soil texture.  In the end, too much of a good thing can be very bad. 

However, if some caution is maintained the use of starter fertilizer can be a beneficial practice for corn.