Main authors: R.K. Laursen, F. Bondgaard, P. Schipper, K. Verloop, L. Tendler, R. Cassidy, L. Farrow, D. Doody, F. A. Nicholson, J. R. Williams, I. Wright, J. Rowbottom, I. A. Leitão, A. Ferreira, B. Hasler, M. Glavan, A. Jamsek, N. Surdyk, J. van Vliet, P. Leendertse, M. Hoogendoorn and L. Jackson-Blake.
Editor: Jane Brandt
Source document: R.K. Laursen et al. (2019) Evaluation of Decision Supports Tools. FAIRWAY Project Deliverable 5.2 216 pp

 

A comprehensive review and survey of Decision Support Tools (DSTs) currently in use in the FAIRWAY case studies is described in »Survey and review of existing decision support tools. Of the 36 DSTs  identified as most relevant, 12 were selected for further investigation to see if a tool developed in a particular national context could be used or provide inspiration elsewhere (»Evaluation of decision support tools). Here we describe the tools evaluated for potential use in the Lower Saxony case study.


Contents table
1. Selection of DSTs to evaluate in Lower Saxony case study
2. Mark Online  
3. NDICEA 
[Note: Because of the resolution of the images, it is difficult to see the detail in some of the figures and tables. See the »full report for more legible originals.]

1. Selection of DSTs to evaluate in Lower Saxony case study

The German case study site is located in southwest Lower Saxony (»Lower Saxony, DE case study). The production on the farms is exclusively arable with a focus on wheat, canola and sugar beet. In particular, bread (milling) wheat production comprises 45-60 % of cropping on arable land. Compared to the average farm size in the region, the test farms cultivate a land area which is above average (150 – 350 ha) with favourable soils. Fertilizer practice is based on mineral fertilizers. Application of manure is mainly restricted to some biogas residues and (in some cases) organic manure (pig slurry, poultry solid manure) imported from the western region. However, the results here represent arable farms only and do not consider famers in the western part of Lower Saxony where a lot of animal production (especially pig and poultry farming) takes place. 

In Germany, the amendment of the fertilizer legislation (DüV 2017) requires the documentation of crop fertilizer needs at the field level. Increases or decreases to this plant specific nitrogen (N) need are based on soil nutrient contents (mineral N (Nmin) in spring, and available phosphorus (P)), precrop and catch crops, fertilizer history and yield level. In our case study and in many parts of Lower Saxony we currently work with the software Düngeplanung. This software goes beyond the legal requirements by creating individual fertilizer plans (incl. the specific fertilizers used) before fertilizer application takes place. Usually farm advisors and famers work together to fine-tune the individual fertilizer strategies. FAIRWAY researchers were interested to see how fertilizer planning in other countries works and how the DSTs are designed.

When selecting DSTs to be tested, it was a priority to select tools, which could be potentially integrated into our advisory work (so they should be of interest to the farmers). Hence, we applied the following criteria:

  • It should be a DST dealing with fertilizer management, which indirectly reflects on nitrate leaching. The Lower Saxony case study provides a feasible dataset and the test results could be compared with the results of Düngeplanung directly.
  • The DST should be applicable at farm-level and consequently directly illustrate and/or influence farmer’s management practices.
  • The DST should originally come from a case study site with comparable climate, soils and agricultural structure.
  • The DST should have the possibility to integrate and assess both obligatory and voluntary measures/environmental restrictions.

We selected the Danish tool Mark Online and the Irish Teagasc NMP Online which are used for fertilizer planning. Unfortunately, the Irish software developers did not provide access for Teagasc NMP Online despite several requests. In addition, we selected the Dutch DST NDICEA, which is an advisory tool to, among others, estimate N-mineralization in the soil during the growing period. Thus, it can be used to additionally adjust fertilizer plans. In Germany, similar software called ISIP is available. A second Dutch program called ANCA could not be tested since its use is restricted to dairy farms only which do not exist within our case study area (for details of these last two see »Decision support tool short list).

2. Mark Online

2.1 Assessment

Mark Online is the most widely used DST for fertilizer planning, optimization and documentation in Danish crop production. It covers all aspects of crop management including soil tillage and crop protection. It is a modularly built and web-based software and is maintained by SEGES, the most important test and research organization in Denmark.

The Danish agricultural system is generally known to be quite restrictive with respect to fertilizer practices but has been proven to show very positive environmental effects. The effects of such strict limitations on N fertilization on farming (like decreasing protein contents in winter wheat) have been discussed on a quite emotional basis within Germany. However, details about the legislation and its implementation are not widely known among German farmers. Hence, it was very interesting to assess the on farm-level implications in the German case study area if farmers had to follow Danish law. We were especially eager to know how implementation of Danish law with the help of Mark Online looks in practice. The SEGES staff were very helpful providing free software access in the context of this project. Danish farmers pay a yearly fee amounting to about 200 EUR/year.

Beforehand and while testing the software, some challenges had to be tackled:

  • Since Mark Online always reflects prevailing law, we also had to pay attention to the specific legal frame conditions in Denmark in order to be able to interpret the results in a sensible way.
  • Denmark designates some sensitive areas with additional restrictions concerning animal rearing, catch crop establishment and/or phosphorus application. For our German test farms, we assumed that they are not located in any of these specific areas.
  • Climate and soil conditions in Denmark and parts of Lower Saxony are only comparable to a limited extent. Especially in the farms located in the very south of Lower Saxony, agri-environmental conditions can deviate (continental instead of maritime climate).
  • The software is only available in the Danish language; hence exploring some of the software details was very time-consuming.
  • On one of the test farms, no stable web access was available.

2.2 Testing

In total, we mapped four farms in our case study in southwest Lower Saxony with Mark Online. All the farmers were very interested to get to know the tool and were supportive with the provision of data and additional information. Within individual farm visits, we discussed details of crop production and compared recent cropping and fertilizer practices to legal requirements in Denmark and Lower Saxony.

Qualitative findings mainly concerning the manageability or Mark Online are described in the following. Generally, our test famers appreciate the well-structured way Mark Online is built. The modular design in particular helps to stepwise tailor farm management to the complex rules. In addition, it is web-based and continuously updated, hence when using Mark Online, farmers can be sure, they are working with the most recent software covering most recent legal restrictions. Furthermore, information has to be only entered once (and not repeatedly into a number of different applications). Therefore it is possible to easily produce long-term analysis of data (e.g. concerning yields, fertilizer used, expenses, etc.). However, our test farmers criticised the fact that they had to expend much effort to first enter the relevant data and further maintain the documentation since many bureaucratic requirements already exist. In addition, worries concerning data security were expressed in this context. Due to the software’s complexity, farmers confirmed, they would need the help of an advisor. Especially the integration their own measurements (which can sometimes differ a lot from general numbers) was found to be difficult to handle.

Table 22. Advantages and disadvantages of the application of Mark Online on German test farms.

Advantages Disadvantages
User-friendly design, clear structure Very complex, help of advisor needed
Supplementary information provided (manual Vejledning om gødsknings og harmoniregler) to answer the most frequent FAQ Entering and maintaining of data is time-consuming
Centralized and holistic approach, data has to be only entered once. Concerns about data privacy
Software is always up-to-date (Farmers can rely on information provided within the software) Infrastructure (Stable and fast web-access has to be available)
Multiannual analysis of data easily possible Relies very much on general numbers
Graphical illustration of some elements provided (green check marks, management of manure within a tank, etc.) Software has to be tailored to conditions in Lower Saxony (climate, legal system, language, etc.)
Various ways for data output (Excel-sheet, pdf, etc.) No freeware

In order to produce and interpret some of the quantitative test results, the different legislation and its execution in both Denmark and Lower Saxony (Germany) had to be considered. Since the Danish system is very comprehensive, we had to limit our focus to some selected legal requirements (farm level) that could be compared between the two countries and which are of greatest interest to the farmers. This is most feasible for the:

  • N-quota according to the Danish system,
  • limits for P-fertilizer use and
  • additional environmental requirements.

The results from our four test farms are summarized below. However, we must stress that these test results should not be extrapolated to farms beyond the case study area. For this, a more comprehensive analysis is needed, e.g. including farms with different site conditions (e.g. poor soil quality) and different focus of agricultural production (dairy, pig farming, etc.).

Calculation of farm-specific N-quota

Both Mark Online and Düngeplanung attribute specific N-needs as nutritional demand for different crops (Grundnorm in Danish, N-Bedarfswert in German) which are legally binding. Both systems allow an increase of this value, if a higher than average yield level can be proven for a specific crop (5-year’s average in Denmark, 3-year’s average in Germany). In the German system, obligatory reductions have to be made if the average yield level falls below the standard value. In the Danish system, the expected yield level is closely linked to soil texture, and thus to soil quality, which is not the case in Germany. Therefore, soil analysis data available for the farms was used to classify the soils accordingly (Figure 12).

D5.2 fig12
Figure 12
D5.2 fig13
Figure 13

Depending on the preceding precipitation and crop growth during winter, the amount of mineral nitrogen in the soil (Nmin) at the start of vegetative growth can vary substantially. In the Danish system, this effect is accounted for by tailoring the crop N-need to the soil type. In addition, in early spring the so-called N-prognose for the whole of Denmark’s territory is published in order to correct values on a regional level. The N-prognose is based on both climatic data and several field trials. In contrast, Lower Saxony requires that soil Nmin is directly subtracted from the calculated plant N-need. To come up with appropriate values, so-called soil climate zones (Bodenklimaräume) are defined. Within each zone, reference soil samples for different crops are taken and analysed for Nmin (0-90 cm depth). The individual farmer can take either their own samples or use the published Nmin-values (Figure 13).

Both, the Danish and the German system account for the effect of the precrop including catch crops; however, values differ. Accounting for fertilizer history (e.g. application of manure for many years) differs between the countries.

Moreover, in the Danish system the N-quota can be further reduced, if specific environmental regulations (like a defined share of catch crops) are not met. Based on this information, Mark Online calculates a farm-specific N-quota, i.e. the amount N a farmer is legally allowed to purchase. In contrast to the German system, retailors of mineral fertilizers have to report sales of fertilizers directly to the authorities.

Below, the average result of the four farms tested is graphed (Figure 14).

D5.2 fig14
Figure 14

  • Column 1 represents the N-quota calculated according to the more restrictive Danish agricultural legislation for the year 2015. The average amount of ca. 200 kg/ha N is further reduced by the average N-prognose (orange) and by some further reduction attributed to N-mineralization of previous catch crops.
  • Column 2 represents the change in the legislation (with higher N-Grundnorms) since 2016. Consequently the average N-quota standardized by hectare increased by about 20 kg/ha.
  • Column 3 represents the average maximum amounts of N in the four test farms that are allowed to apply according to current German law. The maximum amounts of N to be applied in autumn (crosshatched light yellow) is added to the maximum fertilizer applied in spring and summer (light yellow). The average Nmin-contents in early spring are subtracted (orange).
  • Column 4 represents the average amount of fertilizer the four test farms purchased during winter 2017/18 and spring/summer 2018. Data originates from sales accounts and individual farm records.

The results show that the four test farms were able to comply with the Danish N-quota, both for the “old” and recent regulation. Compared with the recent Danish legislation, the four farms purchased only 78 % of the N they would have been allowed to. This corresponds with 87 % for the “old” Danish legislation.

Farm-specific limits for P-fertilization

Both the Danish and the German systems aim at establishing a balance between P-fertilizer inputs and P export from the field. The Danish system defines a limit of 30 kg/ha P-fertilizer on-farm average (corresponds to ca. 68 kg/ha P2O5) for arable farms. If manure is used, this threshold is slightly increased, depending on the type of livestock farming. Furthermore, some P-sensitive areas are mapped, and in these P-fertilizer is even more restrictive. In contrast, in Germany, soil analysis is considered and fields with high P-contents will receive only the amount of P which will be exported by harvest.

The four test farms complied with both systems. On farm-average P-application was 22-30 kg/ha P. Since they primarily apply expensive mineral P-fertilizers such as triple superphosphate or di-ammonium-phosphate, the total amount of P applied is limited and does not exceed plant P-uptake. In regions with intensive livestock production and high amounts of P in organic fertilizers, results could deviate.

Environmental demands

Both countries, Denmark and Germany, define some additional environmental targets in order to fulfil good agricultural practice. This includes the diversification of cropping systems, i.e. number of crops produced on the farm should be at least three, and the share of the two most important is restricted (Krav om flere afgrøder in Danish). The crop rotation of our test farms involved 4-6 different crops.

Furthermore, both systems force farmers to manage some area in an especially environmental-friendly way (Miljøfokusområder in Danish, Ökologische Vorrangfläche in German). For both countries this areas should be about 5 % of the farmed area (only arable in Germany, both arable and grassland in Denmark). Measures to comply with this request include e.g. provision of fallow areas or buffer strips or establishment of catch crops. Depending on the individual farm structure, in both countries farms can choose which fits best for them.

In Denmark, however, depending on the number of animals kept and on the location of the farm, some additional area has to be attributed for the establishment of catch crops, fallow land, multiannual crops for energy production, etc. The test farms, which neither keep animals nor are located in environmentally vulnerable areas (designated by the Danish system) hence have the obligation to establish catch crops on at least 10 % of their summer-harvested crop area (Pligtige efterafgrøder). Depending on the number of animals kept and on the location of the farm, the share of obligatory and voluntary catch crops can also be substantially higher. Again, the share of catch crops can be replaced by alternative measures. Three out of four of our test farms complied accordingly. However, some of the measures the farmers applied (e.g. buffer strips with flowering plants) are financially reimbursed in Lower Saxony (payments for agri-environmental measures). Without these compensation payments, the respective areas would be probably substantially smaller. Since none of our test farms keeps animals, regulations concerning the storage of manure (Lagerreglen) are not relevant in the frame of the testing.

D5.2 fig15
Figure 15

2.3 Implementation

Generally, testing was very successful, although some difficulties occurred. Some elements of Mark Online could be integrated easily into the German system, although the legal framework conditions certainly have to be respected. Mark Online creates clear overviews of the four test farms which helps considerably in advisory work. However, this advisory session would have to be an individual session in order to cope with the complexity of the software.

The tested farms of the case study almost completely meet the requirements concerning fertilizer use and environmental-friendly crop rotations. Admittedly, these are comparatively large arable farms with favourable soil conditions. Hence, ideas on how to implement Mark Online are deduced from the results of these farms only. For livestock farms and/or farms with poor soils additional aspects could be relevant.

In the Danish system a lot of information is linked (fertilizers sales accounts, transport of manure, number of animals, etc.). In addition, it is based on numbers, which can be crosschecked (amounts of fertilizers, numbers of animals, etc.). For that reason, it is much easier for the authorities to control if a farmer meets the requirements. Most information of that kind is also available for Lower Saxony but not linked in the same way. The establishment of the Danish system in Lower Saxony may risk violating data privacy rights; however, it would make fertilizer regulation much more transparent. A farm-specific N-quota can limit the total amount of fertilizers to be applied (and thus better control the risk of reactive N being emitted to the environment) but at the same time rely on farmer’s expertise when it comes to allocation of nutrients in an agronomically sensible way. The way this is realized within Mark Online looks very promising. By linking the yield level to the soil type (and thus to the assigned N-demand), N-quotas are closely linked to the individual nutrient need of farms. In contrast, in the German system the influence of different mineral soil types on field-level is largely disregarded.

Some technical issues hamper the one-to-one implementation of Mark Online. The language of the software is Danish, and many terms are abbreviated which complicates the situation further, especially for details such as the classification of a certain crop. Furthermore, the software is calibrated on the basis of Danish site conditions. Numbers are derived with the help of climatic measurements in Denmark and field trials on Danish research sites (e.g. N-prognose, yield levels etc.). Before implementing, it would have to be harmonized with the German site conditions and numbers. Moreover, the software would need to be continuously updated and maintained.

3. NDICEA

3.1 Assessment

The software NDICEA stikstofplanner provides an integrated assessment of the nitrogen availability in the soil. It goes beyond simple nitrogen budgeting for each crop since it accounts for the complex interaction of the soil-crop-management system. By integrating live weather data, the most variable influence factor for crop development is also factored in.

In Germany, a comparable system is used, which is part of the web-based platform Integrated Plant Production System (ISIP) (see »Decision support tool short list). It estimates nitrogen availability to crops in order to optimize N-fertilizer use and hence improve N-efficiency. Although the German model produces clear graphical representations of N-availability to crops, measured values and field observation sometimes substantially differ from model predictions. Certainly, high spatial variability of precipitation events are a big challenge since they have a large influence on model performance. In addition, the data input into the ISIP system may not be sufficiently comprehensive, e.g. the effects of soil tillage are not accounted for. Since NDICEA also considers a lot of additional information on soil properties and soil tillage, it might be more precise in assessing N-dynamics in the soil.

During first assessment, some basic challenges arose:

  • Integration of live weather data is only possible for the Netherlands, Flanders, England, Denmark and Spain. Even if the software developer added some German weather stations subsequently, the problem of the high spatial variability of precipitation events would remain.
  • Output crucially depends on the quality of the input data. For our test farms site-calibrated descriptions on soil structure in top- and subsoil was unfortunately not available.

3.2 Testing

During the testing phase, we first compiled some data for our test farms. Since NDICEA needs information for at least five years of cropping in order to provide output, this a comparatively comprehensive task. Management of the software was quite convenient (see screenshot in Figure 16) and in most cases no paper guidance was needed to use NDICEA. However, not all types of mineral fertilizers applied by our test farms were listed.

D5.2 fig16
Figure 16
D5.2 fig17
Figure 17

NDICEA calculates plant available N and crop N-uptake for five years of simulation (Figure 17). From the drop-down menu of NDICEA, we chose the Dutch weather station Gelderland-Oost, since it is the one which is closest to our test field (ca. 300 km distance). Although climate conditions here on average resemble those in the south eastern part of Lower Saxony, they are too inaccurate to provide reliable results.

The calculated plant N-uptake by NDICEA of 210 kg/ha N in the year 2018 was somewhat below the results estimated by ISIP (240 kg/ha N-uptake; Figure 17).

Table 23. Advantages and disadvantages of the use of NDICEA to estimate nitrogen mineralization in the soil.

Advantages Disadvantages
User-friendly design, self-explanatory application, additional dialogue boxes provided No additional information on assumptions made
Many influencing factors are considered Comprehensive data input needed (at least five years of cropping)
Own data (soil analysis, crop quality, etc.) can be optionally be integrated (if not, default values are used) Calibration needed
Clear graphical representation provided No availability of local weather data
Freeware  

3.3 Implementation

NDICEA is an advisory tool to provide additional information on N-availability to crops. The level of adoption by farmers depends on the quality of the calculated result. Unfortunately, in the latest version input data (especially live weather data) is much too inaccurate to use for fertilizer recommendations.

Even if the software could be locally calibrated for our test farms, the model results should be continuously verified by measurements in the field (optical and colorimetric measurements). For that reason there is no potential for the implementation of NDICEA on the test farms at present.

 


Notes:

For full references to papers quoted in this article see » References

Download the full report for the Annex and original figures and tables

Go To Top