Should You Variable Rate Nitrogen Spatially by Yield in Corn?

I spoke with an agronomist about a grower who started variable rating nitrogen (N) in corn. Logically, given that high yielding corn takes up more N than low yielding corn, the grower was going to variable rate N by yield across the field - more in high yielding areas, less in low yielding areas.

The agronomist was cautious on this strategy however... while they agreed high yielding areas take up more N than low yielding areas, their concern was the strategy ignores variability in N supply across the field. For instance, some areas may yield high because of deeper topsoil and more organic matter, conditions which likely already mineralize more N. And some areas may yield low because of little topsoil or organic matter, conditions which likely already mineralize less N.

Observing N Response Down the Length of a Field

I looked at past trials pairing zero-N and N-rich strips (often called "delta yield" strips) down the length of a field. Strips are harvested with yield monitors so we can graph yields down the length of the field. Pairing these strips tells a nitrogen story, at least for that part of the field for that growing season:

  • Zero-N strips indicate yields where N is only provided from soil mineralized N, and should partly reflect differences in the ability of soil to provide N
  • The N-rich strip (fertilizer N applied at rate not expected to limit yields) indicates yield potential if N was not limiting
  • The difference ("delta yield") between the two indicates yield that can be captured with N fertilizer... large differences indicate higher responses/needs for fertilizer N while small differences indicate lower responses/needs for fertilizer N (Figure 1)

Figure 1. Conceptual yield responses (delta yields) to nitrogen.

Figure 1. Conceptual yield responses (delta yields) to nitrogen.

An Example Field

Figure 2 is one example from a past trial. Zero-N yield, N-rich yield and elevation are graphed down the length of the field.

Figure 2. Yields of zero-N and N-rich strips and elevation down the length of a field.

Figure 2. Yields of zero-N and N-rich strips and elevation down the length of a field.

A couple things stand out:

  • Not only is there significant variability in N-rich yields (we know this already when we harvest across any field), there is also significant variability in zero-N yields... and the two don't necessarily follow the same pattern
  • There is significant variability in N response (difference between N-rich and zero-N) across the field... in this example response to fertilizer N ranges from 0 to 50 bu/ac
  • This field has significant topography which appears to be a strong driver of N response patterns, particularly for zero-N yields:
    • depressions often result in local yield increases for zero-N ("1" in Figure 2)
    • knolls often result in local yield decreases for zero-N ("2" in Figure 2

Presumably, soil formation and erosion (tillage, water) result in more topsoil, organic matter and N mineralization in depressional areas and less on knolls. Soil cores can reveal the secrets below ground. Pulling 4' cores in another field with significant topography, we could find knolls with only a couple inches of topsoil (and too many stones to insert a full 4' core) and depressions with at least 4' of topsoil (Figure 3) (Field Crop News, 2020).

Figure 3. Soil profile of knoll (upper slope) and depression (lower slope) in field with significant topography near Rockwood, ON.

Figure 3. Soil profile of knoll (upper slope) and depression (lower slope) in field with significant topography near Rockwood, ON.

Is Relative Yield a Good Predictor of Fertilizer N Response Across a Field?

If relative yield is a good predictor of N response across a field, the yield response between N-rich and zero-N strips should be correlated to N-rich yields... big yields should have big responses, small yields should have small responses.

Looking at high yielding areas in Figure 2, some occurred in depressions where there was little response to N - corn yielding 175 bu/ac and strips with zero-N yielded similar to N-rich ("3" in Figure 2). However, other high yielding areas had some of the largest N responses in the field, often elevated areas or knolls producing high yields but very low zero-N yields ("4" in Figure 2).

Similarly, there were low yielding areas with little N response where zero-N yields were similar to N-rich ("5" in Figure 2). N-rich yields and potential N response were perhaps limited by other issues (drought stress, drainage, pH etc.). There were also low yielding areas with very large N responses, particularly knolls and side slopes where N-rich yields were low, but zero-N yields were exceptionally low ("6" in Figure 2).

Clearly in this example, the agronomist's caution was warranted - while corn yield demand plays an important part of N response, it is only one part of the equation. Yield response to N across the field does not appear to be well correlated to yield alone. While only one example was shown, these patterns are common in other fields too. Figure 4 shows similar N responses for a field with less topography from a different county and year.

Figure 4. Yields of zero-N and N-rich strips and elevation down the length of a different field, different year.

Figure 4. Yields of zero-N and N-rich strips and elevation down the length of a different field, different year.

Notes on Delta Yield Strips

Delta yield strips provide great insight, there are also some things to know:

  • one delta yield strip in one year only gives one snapshot in time, and patterns should be taken in context with conditions that year
    • within even a small area, optimum N rate varies year to year, driven by changes in corn yield potential (Deen et al, 2015) and/or N-loss
    • high or low N response areas within fields can change across years (Mamo et al, 2003)... yield potential, N loss and N-response of knolls, soils with poor water holding capacity or drainage or depressions may vary depending on precipitation
  • yield loss from field length zero-N strips can be expensive, especially with high corn prices... using low N rates instead of zero, or targeting areas of interest (e.g. topography, soil types) with mini-strips may indicate general N response with less yield loss

References

Deen, B., K. Janovicek, J. Lauzon and T. Bruulsema. 2015. Optimal rates for corn nitrogen depend more on weather than price. Better Crops. 99:16-18.

Field Crop News. 2020. Understanding and managing field variability in soybean production. Youtube. (accessed 11 Aug. 2021).

Mamo, M., G.L. Malzer, D.J. Mulla, D.R. Huggins and J. Strock. 2003. Spatial and temporal variation in economically optimum nitrogen rate for corn. Agron. J. 95:958-964.


For more information:
Toll Free: 1-877-424-1300
E-mail: ag.info.omafra@ontario.ca