Let's take a look at N-increase trials from a fertile loess site in Saxony over 20 years - always the same field, always the same previous crop. 

On average, 166 kg N/ha yields around 93 dt/ha. However, the range here is from 60 kg N/ha and 85 dt/ha to 240 kg N/ha and 107 dt/ha. Every year, the combination of optimal yield and corresponding N application rate is somewhere else. Nature simply does not adhere to the specifications of Brussels or Berlin. If we were to fertilize optimally in all years, this example would result in an N balance of -30 kg N/ha. On this fertile and well-mineralizing site, 50 kg N/ha straw balance fertilization would therefore be absolutely sensible, otherwise we will deplete the soil fertility.

Now I have extended the table once a little:

First, I have shown the yield increase due to mineral fertilization, i.e. optimum yield minus basic yield, which is 28 dt/ha. Secondly, I have converted the basic yield into soil nitrogen, which had to be available to produce it. This is an average of 135 kg N/ha. The subsequent supply from the soil varies between 92 and 188 kg N/ha, or about 100 kg. This means that the first 64 dt/ha yield is produced with 135 kg N/ha from the soil, the further 28 dt/ha with 166 kg N/ha. A total of 135 and 166 kg N/ha, i.e. a total of 301 kg N/ha, is therefore required for the optimum yield. If we now add the soil and mineral nitrogen together, between 232 and 347 kg N/ha must be provided optimally. Both the soil and mineral N are not known beforehand. And that is the real optimization task in crop production.

Let's return to the optimal amounts of N and the corresponding optimal yields. What happens if fertilization is aligned with an arbitrary mean value?

As an example, let's just take the mean value from this trial series: 166 kg N and 93 dt yield. If we consider a corridor of +- 20 kg N/ha as permissible, then we would have 5 years in which we fertilized on average 46 kg N/ha too much and 8 years in which we fertilized on average 42 kg N/ha too little. So in 13 years we would have missed the optimum with at least 40 kg N/ha. In this example, this means that we fertilize 75% of the areas incorrectly.

And these are only the differences between years. What we see is first ONLY the influence of the weather on the optimal fertilization and yield. The differences within a field on different subplots, i.e. due to soil differences are at least as big, rather even bigger and these heterogeneities are now added on top! So we have a total of two real confounding variables, the weather AND the soil. To be fair, we would still have to mention the different appropriation capacity of the root. So a total of three real disturbance variables.

What does this behavior lead to, i.e. to fertilize only what the legislator prescribes? 

It leads both between the years and within a year on different subplots to a constant over and under the optimum fertilizing. With a good 40 kg N/ha over or under, you have to reckon with an average yield loss of 5 dt. That is about 80 to 100 €/ha, which remain outside - as long as no storage occurs. 

Let's make an example calculation! With 500 ha grain and 75 error rate are 375 ha times in the minimum 80-100 E/ha loss. That is an annual loss of 26 to 35,900 €, which we incur every year. 

But we had this year in many places camps and that despite DVO. Enclosed are some impressions from this year:

If light to medium storage is added to this, the losses from yield, harvest and follow-up costs in the overfertilized areas increase from 80-100 €/ha*year to 150 to almost 400 €/ha. Thus we are then fast with the double to 4-fold loss amount. The gladly expressed saying, "where no eggs, there no nests" indicates actually only that some farmers are not really clear the connections.

You can only remedy the situation by critically questioning your fertilization regime. Fertilizing according to fixed guideline values, as in the DVO or according to other balance approaches or cooking recipes, always leads to the same losses. When overfertilizing and the appearance of bearings, as this year, you notice it clearly. Without bearings you do not see it visually, but nevertheless the losses are existent. 

Often you hear: "The DVO restricts us anyway, I still fertilize what I am allowed to and good with it." This statement is also wrong and ill-considered.

Before the introduction of the DVO, many or most farm managers were often behind the optimum. The result of trials with the N-sensor were N-savings between 8 and 15% and moderate yield increases of around 5 dt/ha. After the introduction, many farmers believe that the N fertilizer is no longer enough. Means, they assume that we will not reach the optimum anymore. So we are possibly before or far before the optimum. And now comes the big thinking mistake: "I won't do anything anymore. It's not worth it.". This is complete nonsense, the opposite is the case. In front of the optimum, the production functions are usually steeper than behind the optimum. Means, the yield increase with an optimization between the many production functions is far larger, than before the introduction of the DVO. We may not save any more mineral fertilizer, but we get a larger yield increase.

Especially under the conditions of the DVO you have to move to flexible fertilizer strategies with the help of the N-tester. The benefit of small-scale and annually adjusted N fertilization is GREATER after the introduction of the DVO than before. In addition, it is also an economic imperative to wisely counteract the N-related decline in yields.