Is precision agriculture the future or an expensive aberration?
On the occasion of the 26th Brandenburg Fertilizer Day, Peer Leithold explains the basics of precision farming and takes a stand on the question of whether precision farming is the future or just an expensive aberration.
Earnings development in Germany
Chart 1 shows the development of winter wheat yields in Germany since 1950. Below the blue trend line, the average yield increase or decrease in the decade is shown in dt/ha*year. We see a long phase of increase starting around 1950. Around 2000, stagnation can be seen and since 2010, we have been losing an average of 0.7 d/ha*year. It should be noted that the graph looks at all-German yield trends. The figures may vary from farm to farm. If the importance of wheat is taken into account, this has major economic consequences. This is because the costs incurred, such as wages, rents, machinery or depreciation, increase at the same time.
The graph shows the development of winter oilseed rape yields in Germany. Here, too, stagnation can be seen around 2000/10. After that, yields decline.
The last crop we consider is grain corn. Graph 3 shows a long increase over the 50 years. Towards 2010, yields are also falling in this crop.
The same can be observed in other countries. In summary, this means on the one hand that 20 years ago the trend began that the positive yield development is weakening. On the other hand, a negative earnings trend has been evident for 10 years. Logically, this is followed by the question of the explanatory pattern. As possible causes are mentioned:
- Climate change
- Breeding
- structural and utilization changes
- Soil cultivation
- Plant nutrition
Another explanatory approach names the reasons:
- Conversion of society to climate-neutral
- Green Deal: Reduction of fertilization and crop protection
Where are we in precision farming?
In digital crop production, we have three major segments.
1. machines and equipment are increasingly being digitized. The goal is to automate field work. There is a long trend over many years that human labor is increasingly decreasing. At the same time, processes are being automated and handled by machines and technology.
2. digital tools are increasingly being used in management, administration and documentation.
3. the third segment comprises field cultivation. Here, suitable algorithms are used to support plant growth. This segment is "precision farming."
The three segments exchange data with each other.
Background & Motivation
What motivates farm managers to deal with the topic of precision farming? In recent years, three major points have crystallized:
- Economic interest (increasing efficiency throughout the farm, i.e. "field and office").
- Positive influence on the environment (keeping the environmental impact of land management low)
- Gain of free time for the farm manager (streamline management processes and reduce time spent)
Distribution
Peer Leithold defines customers as users who use at least one precision farming method and do so on the majority of the areas of at least one crop. In addition, he must use the method every year and can do so without outside help. It is estimated that about 10,000 farms, each with more than 200 hectares, have the potential to use precision farming.
Figure 4 shows the estimated breakdown of precision farming applications. The largest part with about 800 users is in N-fertilization, 300 in basic fertilization and 200 in liming. Then the number of users decreases rapidly.
In summary, it can be said that we are still in an early phase in precision farming.
Introduction Precision Farming
Until now, fields have been managed uniformly. We have the same costs everywhere, but the yields fluctuate. And that's where precision farming comes in. Precision Farming is the solution to an agronomic and crop production problem. To do this, it is necessary to work according to agronomic rules of integrated crop production, because this provides the tools to solve this agronomic question. The agronomic question is solved based on information from the soil and the plant. This solution is then applied to the subplots.
Precision farming using the example of optimal liming
At the beginning, the question is: "What is the optimal amount of lime on each subplot?". This leads to the second question, "Is there an agronomic rule for this, a knowledge base to answer this agronomic question correctly?" Correct means that there are experiments that determine the optimum. In liming, there is that in the form of the fertilizer rule. In liming, we need two pieces of information to answer the agronomic question. We need to know the pH of the soils and the soil group. These two pieces of information need to be generated.
For all other practices used in crop production today, such as N fertilization, growth regulators, fungicides, base dressing, organic fertilization, and seeding, there are functional relationships today between the information we need and an algorithm tested in field trials that will get me to the answer to my agronomic question. To answer the agronomic question, the information must satisfy eight factors:
- Correctness (causality and high correlation).
- Absolute ("more or less of something" is not enough)
- Objective and reproducible
- Sufficient in the sense of the decision
- Easy to use
- High resolution
- Digital
- Inexpensive
This mix of the eight quality requirements then brings us precisely to a selection of tools to carry out the procedures of crop production. These are:
- N uptake (measurable by N sensor)
- Soil groups and nutrients (soil samples and laboratory analysis)
- Soil differences (EM-38)
- Relief (GPS)
Requirements depending on the fruit and the course of the year
For Peer Leithold, one of the most fundamental undesirable developments in precision farming in recent years has been the management zones. Originally, yield management or management zones came from the USA. However, these zones are not stable, because the individual sub-zones are year-specific, crop-dependent and change within a season.
Process approach
If one wants to manage the subarea, several 10,000 agronomic decisions have to be made constantly. As a consequence, one must be prepared to partially automate processes. If you're not ready for that, you can't do precision farming. And we can only automate processes that have already been standardized. In the end, crop production must be automated, standardized and rule-based.
This in turn brings us to the process approach. As in all automation processes, it also runs in agriculture according to the so-called EVA principle (input - processing - output). This means that first information must be generated, then it must be packed into a rule, i.e. an algorithm, and then it is applied to the machine. Figure 5 shows the process flow.
Introduction Precision Farming by Agricon
When a farm has decided to use a precision farming technique, we start with basic training in the first year. Topics include understanding crop production, operating the technology and data management. After that, it's on to on-site consultations. In this way, 25 years of knowledge about digital crop production is transferred, questions are answered and advice is given. After one year, the user masters the system. The goal is not to build lasting consulting relationships. After the first year, Agricon offers several services with licenses, ongoing knowledge transfer, service and regional technicians, and a hotline.
Figure 6 shows each of Agricon's precision farming practices. The customer decides which procedure to start with. They are all compatible with each other, because all procedures work with the same data management - the agriPORT.
Outlook Precision Farming in 5 years
The processes are well developed, yet we are only in the process of transforming them into farms. The training of farm managers will be crucial. On the process side, there will be no revolution, much more an evolution. There will be detail improvements in individual procedures. The emphasis will be on automation of data processing. In addition, bogus or immature processes will be weeded out.
