1. Introduction
Efficient use of organic fertiliser is an indispensable component of global crop production. Organic fertilisers are on-farm available resources that contain nutrients that are important for the plant [1]. They include the macronutrients nitrogen (N), phosphorus (P), potassium (K), magnesium (Mg), calcium (Ca), and sulphur (S), as well as micronutrients. The efficient use of these organic manures and residues is an important strategic component to reduce the need for mineral fertilisers. In this way, finite resources such as phosphorus and potassium can be conserved and the consumption of fossil fuels (especially natural gas) for the Haber–Bosch process for the synthesis of ammonia can be significantly reduced. In contrast to mineral fertilisers, major parts of the nutrients in organic fertilisers are gradually released into the soil solution thereby ensuring a balanced nutrient supply in the soil for the growth of a healthy crop [2].
In addition to plant nutrients, organic fertilisers such as compost, manure, slurry and biogas residues contain organic matter, which is the basis for maintaining the soil organic matter content and humus formation. They are also of particular importance for the biological activity in the soil [3]. As such, organic fertilisers serve as a source of energy for soil microbes which, among other things, improve the soil structure and thus indirectly also plant growth. A higher soil organic matter content also leads to better water infiltration and water storage capacity and improved overall soil fertility [3].
It is also relevant that these organic fertilisers are mostly free of charge for the farmer or, if they are bought in, that they are significantly cheaper than conventional mineral fertilisers since they are a by-product of animal husbandry or biogas plants.
However, improper use of organic fertilisers leads to either over-fertilisation or nutrient deficiencies in the soil. Excess nutrients, especially nitrate and phosphorus, can lead to nutrient pollution of ground or surface waters, which can result in environmental problems such as eutrophication of non-agricultural ecosystems or human health problems due to excessive nitrate concentrations in drinking water [4]. Responsible use of organic fertilisers is therefore important to avoid these effects and achieve sustainable yields [2].
2. Composition of Liquid Organic Manures
Liquid organic manures are produced in particular during intensive livestock farming and fermentation in biogas plants on farms [5]. For example, almost 190 million m3 of LOM were applied to arable land in Germany in 2020. The three most important LOM types were cattle slurry with 95 million m3, biogas digestate with 63 million m3, and pig slurry with 28 million m3 [6].
The composition of LOMs varies greatly. This is because many factors such as the type of animal, the husbandry system as well as the type of feedstuff influence the nutrient composition [7]. Depending on the digestibility of a feed ration, often well over half of the nutrients are excreted again (approx. 60–80% for N, 50–85% for P, and 85–95% for K). The exact amount of nutrients excreted depends, among other things, on the choice of feed and feed additives, the livestock species, age, and performance [7]. The quantity and selection of bedding material and the water input as a dilution factor [8] additionally affects the nutrient content and flow properties.
According to the official advisory service in Lower Saxony [9] cattle slurry is in the range of 4–10% dry matter (DM), 3.7–4.8 kg m−3 total nitrogen (TN), 1.7–2.6 kg m−3 ammonium nitrogen (AN), 0.7–1 kg m−3 P, and 3.7–4.8 K, while pig slurry typically contains 4–7% DM, 3.7–7.2 kg m−3 TN, 2.3–4.3 kg m−3 AN, 0.8–1.6 kg m−3 P and 2–3.7 kg m−3 K. Digestates consist of various substrates that have been fermented in a biogas plant [10]. The nutrient composition of such digestates can vary greatly depending on the quantity and type of the substrate used, biogas technology, temperature, and duration of the fermentation process [11]. Typical liquid fermentation residues from renewable raw materials (e.g., maize silage) contain 4.6–10.1% DM, 3.3–7.1 kg m−3 TN, 1.0–4.5 kg m−3 AN, 0.4–1.6 kg m−3 P, and 3–5.5 kg m−3 K [12].
Compared to mineral fertilisers, there is a significantly higher risk of ammonia emissions and nitrate displacement when using liquid farm fertilisers. Especially the time and quantity of N release are much more difficult to estimate due to the heterogeneity of these organic fertilisers [13]. In terms of the total agricultural ammonia emissions, 50% originate from manure spreading, 36% from stables and manure storage facilities, and 14% from grazing [14].
The nutrient content of farm-based fertilisers should be determined shortly before field application and as accurately as possible in order to determine the amount of LOM to cover the specific nutrient requirements of the crop [15]. The aim is to apply fertiliser in line with requirements, for example, to prevent excessive ammonia emission and to reduce nitrate leaching into the groundwater. Targeted application is also important as the use of manures can lead to a strong, temporary increase in available phosphorus in the top soil layer [16].
Furthermore, N losses in the form of ammonia in the stable and during storage can change the ammonium concentration of LOMs. Possible physical methods for reducing ammonia losses during storage and after the application of liquid organic fertilisers include covering during storage or, separation and dilution of LOM prior to application and incorporation into the soil during application.
3. Methods for Determining Nutrient Concentrations in Liquid Organic Manures
There are various ways of determining the nutrient concentrations of LOMs. According to estimates by various laboratories that regularly test LOMs, as well as agricultural advisors, many farmers in Germany still use tabular reference values published by official advisory bodies for different LOM types in their region to determine the LOM application rate for a particular field or crop. This is likely to be the case for many farmers in other countries as well. The advantage of this option is that no costs are incurred, but it has to be kept in mind that these table values can deviate greatly from the actual element contents of the LOM due to the specific production situation [17].
Alternatively, LOM samples can be sent off for laboratory analysis resulting in very precise data for the nutrient composition. However, for the farmer, representative sampling is not easy to do and it takes some time and effort (e.g., diesel or electricity for mixing) to homogenise the LOM storage. Improper sampling can lead to considerable deviations from the actual DM or nutrient concentrations, which would limit the optimal nutrient supply for the crop and might cause environmental problems. Further disadvantages of this approach are higher costs and the delay until the farmer receives the analysis results [18].
The third option is to determine the nutrient concentrations on site [17,19]. Such on-site nutrient analysis can be done using a large range of different methods. To measure the AN content, which is the most important constituent in LOMs as this N form is directly plant available, two devices are on the market with different designs, but the same measuring principle. In the Quantofix-N volumeter (Terraflor GmbH, Iserlohn, Germany), ammonium is converted into N2 by adding a Na hypochlorite solution into the LOM. In the device, the produced N2 gas displaces a certain amount of water which is used to quantify the gas production [20]. The ammonium nitrogen content in the sample can then be read directly from the calibrated scale of the measuring cylinder in kg m−3 based on the water level [17]. The Agros Nova device (Agros, Lidköping, Sweden) is based on the same principle and it consists of a PVC tube with a reaction chamber and a manometer. Here, the ammonium-N content can be read directly from the manometer [21].
Electric conductivity (EC) is an indirect measure of the electron flow due to the concentration of cations and anions in a solution. This measurement technique is used to determine ammonium and potassium concentrations, which are usually the dominant cations in LOMs [17]. EC probes or so-called ‘pens’ are employed for such conductivity measurements. The values of the EC meter and the ‘pens’ show good agreement with the laboratory values for NH4+ [19]. However, both measurement techniques tend to overestimate NH4+ at high concentrations and underestimate it at low concentrations [19]. According to Piepel et al. [17] the accuracy of ammonium measurement using electrical conductivity is lower than that of the Quantofix-N volumeter.
Total nitrogen (TN) quantification can be based on the measured values of the Quantofix-N volumeter in combination with the gravimetrically determined dry matter concentration [17]. In an evaluation of 391 pig slurries from practical farms in north-western Germany, the combination of these two measurement techniques for TN determination showed a very high correlation (R2) of 0.95 with the laboratory values [17]. Also, for the determination of TN or P concentrations in slurries, hydrometers are being used. In this approach, the TN or P concentration is based on the dry matter content, but the comparability of this method with laboratory values is rather poor [17,19].
The use of these rapid tests in practice at the farm level is often too imprecise and usually only a single parameter can be determined [20]. For this reason, near-infrared spectroscopy (NIRS) is seen as promising, and it is already widely used for agricultural testing in many areas [22]. NIRS is well-established, for example, for the determination of feedstuff constituents such as crude protein [23], crude fibre [24], and crude fat, for the oil content in rapeseed grains [24] or the protein content in legumes [25]. Even the infestation with pathogens in cereals can be determined [26]. It is therefore very meaningful to evaluate the performance of NIRS for the determination of nutrient contents in LOMs.
4. Near-Infrared Spectroscopy—Measuring Principle
In general, spectroscopic techniques are used to obtain chemical, physical, qualitative, and quantitative information on a given sample through the interaction of electromagnetic radiation with atoms and molecules. This information is located in a specific wavelength or frequency of the energy spectrum [27]. In near-infrared spectroscopy, the sample is irradiated with light in the wavelength range of 800–2500 nm. The combination and harmonic bands of the OH, NH, and CH bonds in the samples are set in motion by this irradiation. The interaction of the radiation with the components of the sample leads to different proportions of light reflection, transmission, and absorption, but the absorption part can only be determined indirectly by measuring the reflection and/or transmission [28].
In the reflection measurement mode, part of the incident light is scattered back from the sample surface and passed on to the NIR detector. When measuring transflection, the signal received by the NIR detector consists of two parts. One part of the incident light can be transmitted through the sample and back to the NIR detector, while the second part of the signal comes from the light that is reflected back from the sample surface. For liquid samples with few particles, the transmission of light would dominate and the use of the transflectance method may be appropriate. However, for liquid samples with a high concentration of particles such as LOMs, transflectance measurement is difficult because the suspended particles would drastically reduce the transmission of light through the sample [29].
After recording the spectra during an NIRS measurement of a sample, a mathematical relationship between the spectra and the corresponding reference values is determined using a regression model in order to be able to predict the constituents on the basis of the spectra. Complex chemometrical methods are available for the interpretation of the spectra. Often several preprocessing methods are used on the whole spectra or selected wavelengths of the spectra to correct signals for noise, remove baseline shifts, or reduce data size before conducting the final statistical analysis [30]. After the preprocessing step, a regression model must be established. This involves finding a reliable mathematical relationship between the spectra and the corresponding values for the reference samples being analysed.
For this purpose, the method of partial least squares regression (PLSR) was developed [31]. In PLSR, the prediction variables (i.e., the wavelengths of the NIR spectra) are reduced to a few uncorrelated components. The least squares errors are then determined in order to calculate the best regression equation. PLSR is often chosen when there are many strongly collinear predictor variables associated with a small number of samples [32]. Based on ‘artificial intelligence’ and ‘neural networks’, further regression techniques suitable for processing spectral data (e.g., LASSO regression, ride regression, least-angle regression, random forest) have been developed in recent years [30].
The main advantage of NIRS is the fast and contactless measurement process. Furthermore, no chemicals or high-risk ranges of radiation are required for the measurement, meaning that the analysis does not pose risks to humans or the environment during the analysis [33]. Further advantages of the NIRS technique are that the nutrients are determined continuously, with comparatively little technical effort, and the information on the constituents is immediately available. The costly development and maintenance of the NIRS calibrations and the indispensable and labour-intensive quality assurance are seen as certain disadvantages [18].
As near-infrared spectroscopy is an indirect measurement method, alternative analysis techniques—typically wet chemical methods—are required as reference for creating the calibration [22]. Calibrations must be developed which show a close correlation between the results of the NIRS measurements and the results of the reference methods. It should also be noted that in near-infrared spectroscopy, the accuracy is strongly dependent on the physical state of the sample, i.e., on the particle size, volume density and temperature, among other things [34].
5. Use of NIRS for the Characterisation of Liquid Organic Manures in the Laboratory
All components in LOMs have their individual properties (e.g., due to hydrogen bonds) and therefore a unique NIR spectrum. These special characteristics are the spectral basis for predicting the nutrient composition of LOMs using NIRS [18].
In their review, Horf et al. [30] described that under controlled conditions in the laboratory, various types of equipment are used to analyse the properties of LOMs. For example, there are spectrometers with a dispersive grating and scanning monochromator such as the NIR system 6500 from FOSS (Hilleroed, Denmark). This type of spectrometer has a high resolution, but a slow scanning speed. Furthermore, it has moving parts that are sensitive to vibrations, which clearly limits the on-farm use. Dispersive diode array NIR spectrometers (e.g., Corona®45, Carl Zeiss Spectroscopy GmbH, Jena, Germany) have—in contrast to the aforementioned spectrometer type—a high scanning speed. These robust instruments are suitable as benchtop devices for laboratory applications, but also as a portable version for on-farm use. Fourier transform spectrometers have the advantage over dispersive spectrometers in that they are faster and have a higher resolution. Although they were originally intended for use in the mid-infrared range only, devices are now also available in the near-infrared range (e.g., NIRFlex N-500, BÜCHI Labortechnik AG, Flawil, Switzerland). Furthermore, so-called bandpass spectrometers with special filters for selected wavelengths are used for laboratory measurements (e.g., MM55 device, NDC Infrared Engineering Ltd., Maldon, England). These are comparatively, low-cost NIRS devices, but they can only carry out measurements in very limited wavelength ranges.
There is much evidence in the literature that near-infrared spectroscopy works technically well for measurements of LOMs under laboratory conditions [35,36,37,38,39]. The accuracy of NIRS measurements is usually evaluated using correlation coefficients (R2) which indicate the relationship between values predicted by NIRS measurements and those measured using the reference method [40]. According to Williams [41], a value for R2 between 0.66 and 0.81 means that an ‘approximate quantitative’ prediction is possible, while a value for R2 between 0.82 and 0.90 means a ‘good’ prediction. Calibration models with a value for R2 above 0.91 are considered ‘excellent’. The root mean squared error of prediction (RMSEP) is an additional statistical measure used to assess the quality of a model’s prediction. A low RMSEP value indicates a good prediction model [41].
Over the last two and a half decades, several studies have been conducted that focus on the performance of NIRS to measure LOM characteristics relevant to farmer decisions on application rates. Most studies deal with the analysis of cattle or pig LOMs. In addition, two studies with biogas residues are listed. The number of samples ranges from 62 to 584 samples (Table 1).
Overall, these studies show a very good prediction accuracy for DM (R2 values above 0.9). Also, the contents of TN and AN in LOMs can generally be estimated well using NIRS technology. The correlation coefficients for these two constituents were just above 0.6 only in the study by Millmier et al. [24], which could be due to the choice of three quite different types of organic manures, with two of them being in liquid form and one in solid form.
Based on the evaluated studies, the accuracy of P and K determination by NIRS is less reliable. In the studies conducted by Cabassi et al. [36], Dagnew et al. [37], De Ferrari et al. [38], Horf et al. [30], Saeys et al. [35], and Zimmermann and Hartung [44], P was determined with acceptable correlation coefficients (R2 > 0.7). Millmier et al. [24] and Reeves and Van Kessel [20], on the other hand, showed no acceptable correlations between the reference data and the NIRS data. Determination of K failed completely in two of eleven studies [37,40] and was only determined very well in the studies by Zimmermann and Hartung [44], Horf et al. [30], Mouazen et al. [42], and Williams et al. [43].
One possible reason for the different prediction accuracies of the NIRS sensors in the individual studies may also be the type of sample preparation (i.e., liquid or dried) which is then measured using NIRS. In the laboratory, LOM samples are usually preprocessed before being measured with NIRS. Horf et al. [45] found that neither filtering, centrifuging, nor drying the liquid manure samples improved the accuracy of prediction of the nutrient and dry matter content compared to the raw LOM samples. The results were also independent of whether the concentrations were based on the fresh or dry matter basis.
It should also be noted that the results can vary depending on the type of evaluation, for example which algorithm was used for the data calculation. There are various programmes that mathematically evaluate the relationship between the spectra and the measured values. Various multivariate analysis methods are used to analyse NIR spectra, the entirety of which is referred to as chemometrics. Not just one data point, but certain sections or the entire spectrum can be used for calibrations. Due to the amount of structural information used to create the method, even the smallest differences in the spectra must be recorded and analysed. In the future, AI-based algorithms might also lead to significant improvements in spectra evaluations.
Overall, it can be said that NIRS is suitable for the determination of DM and important nutrients in LOMs. If this method could also be used online and on-farm, a much more targeted use of LOMs would be possible.
6. Use of NIRS for the Characterisation of LOMs On-Farm
Although NIRS is obviously suitable for the characterisation of LOMs, it must be taken into account that to date, most of the studies have dealt with experiments in the laboratory, i.e., under controlled conditions. Only very few studies on the reliability of data collected with NIRS sensors have actually been carried out on farms under real life conditions.
6.1. On-Farm Testing of NIRS Sensors for Nutrient Measurement in LOMs
In the experiments conducted by Zimmermann et al. [46], a NIRS system was mounted on an 18 m3 slurry tanker for online monitoring of the nutrient concentration during the application of LOMs. The tests took place in Schleswig-Holstein, Germany, and different slurry types of cattle, pig, and biogas digestates were used. The optimal installation position of the NIRS measuring head was determined in preliminary tests. A Polytec PSS 1720 NIR spectrometer (Polytec GmbH, Waldbronn, Germany) was used, which covers a wavelength range from 850 to 1650 nm. A special housing was designed to protect the sensor during the in-field application tests. Reference samples were taken via a sampling valve and then analysed in an accredited laboratory. The NIRS system showed reproducible results and correlation coefficients between the laboratory and NIRS for DM, TN, AN, P, and K were calculated. High correlations were found between the NIRS values and the reference values that were analysed in the laboratory (R2 = 0.95–0.99).
In 2009, Zimmermann and Hartung [44] published a comprehensive report. The aim of these tests was to evaluate the feasibility of calibration for a NIRS sensor for mobile nutrient detection on-farm. The reference samples were taken during the ongoing NIR measurements. The NIRS calibrations were created using a Partial Least Squares Algorithm and validated using the so-called ‘full cross’ validation method based on a total of 206 slurry samples. The highest correlations for DM (R2 = 0.95) were achieved with cattle slurry (n = 61). For the parameters TN, AN, P, and K, 49 pig slurries were analysed. The following correlation coefficients were obtained: TN = 0.96, AN = 0.94, P = 0.86. The best result for mixed slurry (n = 96) was shown for K with an R2 of 0.86.
In the study by Stenberg and Gustafsson [47], 73 dairy, pig, and biogas slurries were analysed in a test rig simulating a real slurry spreader. In addition, 13 biogas slurries were analysed during field application. An NIR sensor from Tec5 (Oberursel, Germany) with a PGS-NIR1.7 (Carl Zeiss AG, Oberkochen, Germany) detector was used for these experiments. On the test rig the NIRS measuring head was attached to a 100 mm tube and connected to a container containing 100 L of the test slurry. The slurry was pumped through the pipe system of the rig to simulate the slurry flow in a real spreader. Each slurry was analysed at three different pumping rates to increase the robustness of the calibrations. For the reference analysis, 1 L samples were taken during the circulation of the slurry and analysed in the laboratory. During the field application with the biogas slurry, 38 spectra were recorded from each load. The calibrations were created using PLSR and worked best for DM. Overall, nutrient concentrations in cattle slurry showed better predictions than in pig slurries or biogas digestates. The following correlation coefficients for DM were obtained for the individual slurries: Cattle slurry 0.90 (n = 22), pig slurry 0.74 (n = 27), and biogas digestate 0.67 (n = 24). For TN and AN, results were only given for cattle slurry (R2 = 0.81 and 0.44) and pig slurry (R2 = 0.46 and 0.24). For phosphorus, the correlation coefficients were: Cattle slurry 0.61 (n = 22), pig slurry 0.57 (n = 27), and biogas digestate 0.39 (n = 24).
Thiessen et al. [48] built a test rig on which they installed five different NIRS sensors. A slurry tanker was used to homogenise the slurries by circulating them in the test rig. This system was tested on 13 farms in northern Germany. A total of 51 different slurry types were sampled, including pig, cattle, biogas, and mixed slurries. For the measurement, roughly 6 m3 of slurry was pumped through the test rig until the slurry was well homogenised. The NIRS sensors achieved average deviations of 22–23% from the reference result for DM. For TN, four out of five sensors showed deviations of 24–38%. For the parameters AN, P, and K, the sensors showed similarly high deviations of 38–54%, 36–69% and 25–72% respectively. According to the authors, the good results for DM were due to the relationship between water and DM, as water leads to strong signals in the NIR spectrum. The authors assume that the results for the other nutrients were poor compared to other studies because the sensors measured unknown samples under on-farm conditions and the measured values did not fit well into the existing calibration models.
6.2. Test Procedures for Commercially Available NIRS Sensors in Germany
For more than 15 years, several companies have offered NIRS sensors in Germany for on-farm use and mounting on stationary pumps for filling transport vehicles for liquid manure directly at the storage facility or on application vehicles to measure the LOMs passing by. The three relevant suppliers include John Deere GmbH & Co., KG (Walldorf, Germany), m-u-t GmbH (Wedel, Germany) in cooperation with Zunhammer GmbH (Traunreut, Germany) and Dinamica Generale S.p.A. (Poggio Rusco, Italy). The slurry pumps or tankers with the corresponding sensor technology are either operated by contractors or by the farmers themselves. One motivation for farmers to buy such NIRS sensors is to be able to use the measured NIRS values for the documentation of the LOM constituents in accordance with the provisions of fertiliser legislation when placing LOMs on the market and immediately before LOM application in the field. It is therefore important for the farmers using NIRS sensors that the quality of the devices is certified, as only scientifically recognised techniques are permitted for official documentation in Germany.
The German Agricultural Society (DLG) conducts tests under conditions close to on-farm situations. They test mobile NIRS sensors for the determination of constituents in LOMs (i.e., liquid farm manure from animal husbandry (cattle slurry, pig slurry, and mixed slurry from cattle and pig slurry) and biogas residues from the fermentation of cattle and pig slurries with the addition of renewable raw materials) in order to determine equivalence with laboratory analysis using officially recognised methods [49]. This DLG test covers the parameters: DM, TN, AN, P, and K and for these tests, five typical LOMs from different farms are selected based on the dry matter content (cattle slurry with 4–9% DM, pig slurry with 2–7% DM, and digestate with 5–8% DM). For each test, a quantity of 3 to 5 m3 of LOM is pumped into a container equipped with a pump and a standard tubing system. The sensors to be tested and a bypass for taking the laboratory reference sample are then connected to the tubing system. In the first step, the LOM is thoroughly homogenised by continuous pumping. The NIRS sensors are then switched on. As soon as the displayed measured values are stable, the NIRS data are documented and samples are taken via the bypass for reference analysis in the laboratory. The LOM samples are deep-frozen at −18 °C immediately after collection and sent to five different accredited laboratories for analysis. The relative deviation of the sensor value for each parameter compared to the reference value (calculated as the arithmetic mean of the five laboratory results) is then used to validate the measurement accuracy of the NIRS sensors. Individual parameters of a LOM type are certified according to the criteria catalogue developed by the DLG test centre (Table 2). As a minimum requirement, the DLG has specified that the TN content must be reliably determined with the NIRS sensor, i.e., the test must therefore be passed for this constituent.
Table 3 summarises DLG test reports for NIRS sensors from three manufacturers that offer corresponding devices for on-farm use in Germany, categorised by the respective constituents for the LOM types cattle slurry, pig slurry, and biogas digestate.
Among the three commercially available NIRS sensors for on-farm use in Germany, DM determination in cattle slurry is classified as ‘good’ for the sensors from m-u-t GmbH and John Deere according to the DLG classification. In contrast, the sensor from Dinamica Generale S.p.A. only passed the test for DM in cattle slurry. For the analysis of TN and AN in cattle slurry, only the Harvestlab 3000 from John Deere received a ‘very good’ rating. The sensor from m-u-t GmbH achieved ‘passed’ ratings in the test for TN and AN. According to the DLG test, the EVO NIR 4.0 from Dinamica Generale S.p.A. was ‘good’ for TN determination, while no data are available for AN. Regarding K in cattle slurry, two out of the three sensors received a ‘passed’ rating, while data are unavailable for the sensor from Dinamica Generale S.p.A. For P, only the sensor from m-u-t GmbH was rated ‘very good’, whereas the other two sensors received only a ‘passed’ rating for this parameter.
In the case of pig slurry, DM can be determined with ‘good’ or ‘very good’ rating by two sensors, yet no data are available for the sensor from Dinamica Generale S.p.A. for this parameter. All sensors successfully passed the test for TN in pig slurry. For AN, the sensor from m-u-t GmbH received a ‘passed’ rating, whereas no data are available for the other two sensors. For K in pig slurry, two sensors completed the test with a ‘passed’ rating, while data are unavailable for the sensor from John Deere. According to the DLG assessment, only the sensor from John Deere reliably measured P in pig slurry. The other two sensors either did not pass the test for phosphorus (m-u-t GmbH) or did not provide data (Dinamica Generale S.p.A.).
For biogas residues, the DM measurement test received a ‘passed’ rating for two sensors, while no data are available for the sensor from John Deere. TN in biogas digestate was determined as ‘good’ by the sensors from m-u-t GmbH and John Deere, whereas the sensor from Dinamica Generale S.p.A. only completed the test with a ‘passed’ rating. For AN, the ratings varied among the three sensors: the m-u-t GmbH sensor was rated ‘good’, the John Deere sensor was rated ‘passed’, and no data are available for the Dinamica Generale S.p.A. sensor. According to the DLG assessment, P and K in biogas residues were rated ‘very good’ only for the Dinamica Generale S.p.A. sensor. The other two sensors received a ‘passed’ for K. Regarding P, the NIR speedspy onboard manure from m-u-t GmbH was rated ‘failed’, and no data are available for the Harvestlab 3000 from John Deere.
6.3. Factors Affecting the Reliability of On-Farm NIRS Measurements
When using NIRS sensors for practical applications on farms or directly for application in the field, there are many factors that can influence the measurement and must therefore be taken into account. In contrast to NIRS measurement in the laboratory with controlled conditions, the ambient temperature, movements and vibrations of the machines, the temperature of the slurry being measured, and the limited skills of not sufficiently trained users, for example, can play an important role under practical farming conditions and influence the measurement with the sensors and therefore the reliability of the measured values.
The homogenisation of the LOM prior to measurement can also influence the NIRS measurement and the reliability of the values determined. Liquid organic manures separate quickly, as heavier components sink to the bottom and lighter components form a floating layer at the top. The solid phase mainly contains organically bound elements in higher concentrations such as P, Mg, Ca, and N, while AN and K are dissolved in the liquid phase [14]. Good homogenisation is, therefore, important because the more evenly the components are distributed in the LOM, the better the NIRS sensors can measure the constituents, as the small-scale heterogeneity in the LOM is lower, no very short-term fluctuations in concentration occur.
In the laboratory operation of NIRS sensors, measurement as well as documentation and interpretation of the measured values are carried out by trained personnel in order to minimise measurement errors. Furthermore, a continuous quality surveillance procedure is mandatory in accredited laboratories. On-farm use of NIRS sensors involves many different people, such as farm managers and employees of agricultural contractors, who need to learn how to use the sensors to ensure reliable measurements. This requires the operators’ openness and willingness to familiarise themselves with the new technology. This is particularly crucial because the main focus of contractors, for example, is usually not on taking high-quality measurements, but on applying a certain amount of nutrients with the help of a NIRS sensor. Regular training and further education in this respect can also help to ensure that the sensors are operated correctly.
As LOMs vary greatly in their composition, the user should know which LOM type is to be measured with the sensor, i.e., in which range DM and nutrient concentrations are likely to be expected. It should also be noted that NIRS sensors are only suitable to a very limited extent for measuring special LOMs, such as LOM with a very low or very high DM content, biogas residues from biogas plants with a high intake of dry chicken manure, production waste from the food industry or similar substances. Physically or chemically acting additives to LOMs such as clay minerals, acids, iron-III-chloride or nitrification inhibitors as well as physically acting measures such as separation and massive addition of water (dilution) change the optical spectrum of the LOMs. LOMs treated in this way can therefore not be determined using NIRS.
6.4. Pros and Cons of NIRS Measurement Technology for LOM Testing
The continuous recording via NIRS measurements during the application of LOM in the field enables the targeted application of a predetermined quantity of nutrients in kg per hectare (and not just a certain LOM quantity per hectare) to an area. The sensor allows the application rate to be adjusted on the go according to the measured nutrient concentrations in the LOM and the speed of the vehicle, and the amount of nutrients applied is also automatically documented. Based on previously created application maps, variable nutrient application within a field is also possible [57]. In that way, for example, differences in the supply of certain nutrients to the soil or the crop stand can be compensated for. Most probably, the application would then be focussed on the plant-available nitrogen (i.e., the AN content) in the LOM to be applied, as N is of paramount importance for crop growth in the current vegetation period and finally for crop yield and quality. The need for other nutrients could then be met using precision application of mineral fertilisers.
The disadvantages of using NIRS sensors on-farm are that the measurement only covers certain parameters, and the accuracy, as well as the reproducibility of the NIRS values, depend heavily on the quality of the calibration (i.e., the strength of the relation-ship between the laboratory and NIRS values and the representativeness of the LOM samples used for this calibration). The costs of using a NIRS system must also be taken into account. These are incurred directly by the farmer when purchasing NIRS systems or are passed on indirectly to the farmer by the contractor via the price for the application. As mentioned before, the range of parameters that can be recorded when measuring LOMs using NIRS technology is limited. In addition, it has to be kept in mind that different animal species, feedstuffs, or slurry additives influence the measurement, so comprehensive calibration models must be available for each type of slurry [58]. For the calibration, a regression of spectral data with the lab data for the respective constituents of the corresponding reference samples is calculated [59]. The more accurate the reference method, the greater the accuracy of the NIRS measurement [59].
7. Conclusions
Near-infrared spectroscopy is well established in laboratories to analyse many different agricultural products including LOMs. The advantages of NIRS are that it is fast, contactless and that it is capable of continuously generating a large number of measured values.
In order to increase accuracy, the calibration of the sensors must be further developed. This is the most important way to provide the farmer with robust data. Moreover, the reliability of NIRS technology in practical use must be further improved. For example, the effect of different installation positions of the NIRS sensor (especially with regard to the so-called pressure side compared to the suction side) must be evaluated. In addition, extreme deviations from the usual standard values for a particular LOM type should be indicated to the user. Furthermore, representative LOM samples should also be taken several times during a LOM application period when using NIRS technology. These samples are then analysed in the laboratory in order to compare the nutrient content with the values measured by the NIRS sensor and, if necessary, to readjust the NIRS values using a correction factor.
It should also be noted that currently to the best of our knowledge, there is no standardised approval procedure for NIRS devices in Germany, Europe, or other regions of the world. For example, there are different regulations in the various federal states in Germany regarding the extent to which NIRS is approved for recording nutrient content in LOMs or whether the values determined using NIRS may be used for documentation purposes within the framework of the German Fertiliser Ordinance. Currently, some official advisory bodies (e.g., Bavaria, North Rhine-Westphalia) allow the use of NIRS data for documentation purposes, while this is not permitted in other regions (e.g., Lower Saxony).
In order to make optimum use of the potential of the NIRS measurement technology in the future, calibrations need to be expanded and improved so that the sensors can deliver accurate and reproducible results for the various LOM types. Regular maintenance and inspection of the technical properties of NIRS devices by an independent organisation should also be implemented. To achieve a high level of acceptance among farmers, an official quality assurance for NIRS sensors should therefore be introduced in practice (comparable to the mandatory technical inspection of crop protection sprayers in Germany). Based on the review presented here, we see that this is the only way to ensure that reliable results (i.e., nutrient content values that correspond well with the laboratory values) are achieved so that nutrient analyses using NIRS can finally also be used for improved application accuracy and for documentation purposes.
Conceptualisation, C.H., K.D. and H.-W.O.; formal analysis, C.H.; investigation, C.H.; data curation, C.H.; writing—original draft preparation, C.H.; writing—review and editing, K.D. and H.-W.O.; visualisation, C.H. All authors have read and agreed to the published version of the manuscript.
No new data were created or analysed in this study.
The authors declare no conflicts of interest.
Footnotes
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.
Studies using NIRS for the characterisation of liquid organic manures (LOM)—representation of R2 and RMSEP.
| R2 (RMSEP) | |||||||
|---|---|---|---|---|---|---|---|
| Reference | LOM | Number of Samples (n) | DM | TN | AN | P | K |
| Cabassi et al., 2015 [ | cattle | 99 | 0.92 (1.01) | 0.86 (0.47) | 0.78 (0.27) | 0.74 (0.15) | - |
| Dagnew et al., 2004 [ | pig | 88 | 0.92 (na) | 0.89 (na) | - | 0.79 (na) | 0.68 (na) |
| De Ferrari et al., 2005 [ | cattle | 101 | 0.95 (0.95) | 0.90 (0.32) | 0.83 (0.18) | 0.78 (0.13) | 0.19 (0.67) |
| Horf et al., 2022 [ | cattle, pig, digestate | 62 | 0.92 (0.86) | 0.91 (0.06) | 0.94 (0.03) | 0.91 (0.04) | 0.91 (0.05) |
| Millmier et al., 2000 [ | pig, lagoon effluent | 100 | 0.90 (na) | 0.69 (na) | 0.62 (na) | 0.61 (na) | 0.71 (na) |
| Mouazen et al., 2005 [ | pig | 195 | 0.91 (1.24) | 0.89 (1.22) | 0.78 (1.21) | 0.85 (0.90) | 0.84 (0.98) |
| Reeves & Van Kessel 2000 [ | cattle | 107 | 0.93 (1.45) | 0.92 (0.05) | 0.86 (0.03) | 0.48 (0.08) | 0.70 (0.07) |
| Saeys et al., | pig | 584 | 0.91 (1.20) | 0.86 (1.23) | 0.76 (1.13) | 0.75 (1.12) | 0.69 (1.15) |
| Sørensen et al., 2005 [ | cattle, pig | 130 | 0.97 (na) | 0.95 (na) | 0.92 (na) | - | - |
| Williams et al., 2020 [ | pig | 304 | 0.96 (0.24) | 0.95 (0.09) | - | 0.90 (0.03) | 0.94 (0.03) |
| Zimmermann & Hartung 2009 [ | cattle, pig, digestate | 267 | 0.95 (0.54) | 0.96 (0.29) | 0.94 (0.22) | 0.86 (0.32) | 0.86 (0.21) |
DM = Dry matter, TN = Total nitrogen, AN = Ammonium nitrogen, P = Phosphorus, K = Potassium. na = not available.
DLG assessment scheme for a single parameter for each of the three pre-defined liquid organic manure (LOM) types (cattle slurry, pig slurry, and biogas digestate).
| Symbol | Rating | Explanation |
|---|---|---|
| ++ | ‘passed, very good’ | Four out of five paired values within a LOM type show relative deviation of ≤10% and none > 20% |
| + | ‘passed, good’ | Four out of five paired values within a LOM type show a relative deviation of ≤15% and none > 25% |
| o | ‘passed’ | Three out of five paired values within a LOM type show a relative deviation of ≤25% and none > 35% |
| − | ‘not passed’ | the test was not passed |
Compilation of selected DLG test reports for NIRS sensors (for explanations of the symbols see
| Manufacturer | m-u-t GmbH | John Deere | Dinamica Generale S.p.A. |
|---|---|---|---|
| Sensor | NIR Speedspy Onboard Manure | Harvestlab 3000 | EVO NIR 4.0 |
| Software | V14.3.1 | various calibration models | various calibration models |
| Test report no. | 7122 [ | 6886 [ | 7139 [ |
| Cattle slurry | |||
| DM | + | + | o |
| TN | o | ++ | + |
| AN | o | ++ | n.a. * |
| K | o | o | n.a. |
| P | ++ | o | o |
| Pig slurry | |||
| DM | ++ | + | n.a. |
| TN | o | o | o |
| AN | o | n.a. | n.a. |
| K | o | n.a. | o |
| P | − | o | n.a. |
| Biogas digestate | |||
| DM | o | n.a. | o |
| TN | + | + | o |
| AN | + | o | n.a. |
| K | o | o | ++ |
| P | − | n.a. | ++ |
DM = dry matter, TN = total nitrogen, AN = ammonium nitrogen, P = phosphorus, K = potassium. * n.a. = not available.
References
1. Herrera, A.; D’Imporzano, G.; Zilio, M.; Pigoli, A.; Rizzi, B.; Meers, E.; Schouman, O.; Schepis, M.; Barone, F.; Giordano, A. et al. Environmental performance in the production and use of recovered fertilizers from organic wastes treated by anaerobic digestion vs synthetic mineral fertilizers. ACS Sustain. Chem. Eng. J.; 2022; 10, pp. 986-997. [DOI: https://dx.doi.org/10.1021/acssuschemeng.1c07028] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/35087697]
2. Shaji, H.; Chandran, V.; Mathew, L. Organic fertilizers as a route to controlled release of nutrients. Control. Release Fertil. Sustain. Agric.; 2021; pp. 231-245.
3. Lin, Y.; Ye, G.; Kuzyakov, Y.; Liu, D.; Fan, J.; Ding, W. Long-term manure application increases soil organic matter and aggregation, and alters microbial community structure and keystone taxa. Soil Biol. Biochem.; 2019; 134, pp. 187-196. [DOI: https://dx.doi.org/10.1016/j.soilbio.2019.03.030]
4. Withers, P.; Haygarth, P.M. Agriculture, phosphorus and eutrophication. Soil Use Manag.; 2007; 23, pp. 1-4. [DOI: https://dx.doi.org/10.1111/j.1475-2743.2007.00116.x]
5. Saeys, W.; Watté, R.; Postelmans, A. The Role of NIR Spectroscopy in Precision Fertilisation with Manure; IFS Proceedings 840, 1-22 International Fertiliser Society: York, UK, 2019.
6. DESTATIS. Wirtschaftsdünger tierischer Herkunft in landwirtschaftlichen Betrieben/Landwirtschaftszählung—Fachserie 3 Reihe 2.2.2—2020. Available online: https://www.destatis.de/DE/Themen/Branchen-Unternehmen/Landwirtschaft-Forstwirtschaft-Fischerei/Produktionsmethoden/Publikationen/Downloads-Produktionsmethoden/wirtschaftsduenger-2030222209004.pdf?_blob=publicationFile (accessed on 19 August 2024).
7. Chen, L.; Xing, L.; Han, L.; Yang, Z. Evaluation of physicochemical models for rapidly estimating pig manure nutrient content. Biosyst. Eng.; 2009; 103, pp. 313-320. [DOI: https://dx.doi.org/10.1016/j.biosystemseng.2009.04.007]
8. Martínez-Suller, L.; Azzellino, A.; Provolo, G. Analysis of livestock slurries from farms across Northern Italy: Relationship between indicators and nutrient content. Biosyst. Eng.; 2008; 99, pp. 540-552. [DOI: https://dx.doi.org/10.1016/j.biosystemseng.2007.12.002]
9. Landwirtschaftskammer Niedersachsen. Nährstoffgehalte in Organischen Wirtschaftsdüngern und Mineralischen Düngemitteln. Available online: https://www.duengebehoerde-niedersachsen.de/duengebehoerde/news/32460_Naehrstoffgehalte_in_organischen_Wirtschaftsduengern_und_mineralischen_Duengemitteln (accessed on 1 October 2024).
10. Boxberger, J. Gülleaufbereitung. Praxishandbuch Oganische Düngung; Boxberger, J.; Mayer, J.; Möller, K.; Pöllinger, A. Erling Verlag: Clenze, Germany, 2020; pp. 502-520.
11. Alburquerque, J.A.; de La Fuente, C.; Bernal, M.P. Chemical properties of anaerobic digestates affecting C and N dynamics in amended soils. Agric. Ecosyst. Environ.; 2012; 160, pp. 15-22. [DOI: https://dx.doi.org/10.1016/j.agee.2011.03.007]
12. DLG-Merkblatt 397: Gärreste im Ackerbau effizient nutzen. Available online: https://www.dlg.org/mediacenter/dlg-merkblaetter/dlg-merkblatt-397-gaerreste-im-ackerbau-effizient-nutzen (accessed on 1 October 2024).
13. Mahmud, K.; Panday, D.; Mergoum, A.; Missaoui, A. Nitrogen losses and potential mitigation strategies for a sustainable agroecosystem. Sustainability; 2021; 13, 2400. [DOI: https://dx.doi.org/10.3390/su13042400]
14. Saeys, W. Technical Tools for the Optimal Use of Animal Manure as a Fertiliser: On-line Composition Measurement and Manure Injection Control. Ph.D. Dissertation; Katholieke Universiteit Leuven: Leuven, Belgium, 2006.
15. Piepel, M.-F.; Olfs, H.-W. Development of a physicochemical test kit for on-farm measurement of nutrients in liquid organic manures. Agriculture; 2023; 13, 477. [DOI: https://dx.doi.org/10.3390/agriculture13020477]
16. Lugo-Ospina, A.; Dao, T.H.; Van Kessel, J.A.; Reeves, J.B. Evaluation of quick tests for phosphorus determination in dairy manures. Environ. Pollut.; 2005; 135, pp. 155-162. [DOI: https://dx.doi.org/10.1016/j.envpol.2004.09.007]
17. Piepel, M.-F.; Dittert, K.; Olfs, H.-W. Evaluation of physicochemical on-farm quick tests for estimating nutrient concentrations in pig slurry and development of an application for mobile devices. Agronomy; 2022; 12, 2809. [DOI: https://dx.doi.org/10.3390/agronomy12112809]
18. Chen, L.; Xing, L.; Han, L. Review of the application of near-infrared spectroscopy technology to determine the chemical composition of animal manure. J. Environ. Qual.; 2013; 42, pp. 1015-1028. [DOI: https://dx.doi.org/10.2134/jeq2013.01.0014]
19. Van Kessel, J.S.; Thompson, R.B.; Reeves, J.B. Rapid on-farm analysis of manure nutrients using quick tests. J. Prod. Agric.; 1999; 12, pp. 215-224. [DOI: https://dx.doi.org/10.2134/jpa1999.0215]
20. Van Kessel, J.S.; Reeves, J.B. On-farm quick tests for estimating nitrogen in dairy manure. J. Dairy Sci.; 2000; 83, pp. 1837-1844. [DOI: https://dx.doi.org/10.3168/jds.S0022-0302(00)75054-5]
21. Singh, A.; Bicudo, J.R. Dairy manure nutrient analysis using quick tests. Environ. Technol.; 2005; 26, pp. 471-478. [DOI: https://dx.doi.org/10.1080/09593332608618541] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/15974265]
22. Chakravarthy, K.; Nekwaya, L.; Nabeneza, S.; Tillard, E.; Marblé, Y.; Salgado, P.; Lecomte, P. Literature review on the applicability of near infrared reflectance spectroscopy (NIRS) to predict the chemical composition of forages, soil and animal organic matter samples. Proceedings of the International Conference ‘INDIGENOUS’; Hyderabad, India, 28–31 January 2016.
23. Noack, P.O. Einsatz von Multi- und Hyperspektralsensoren in der Landwirtschaft. Available online: https://www.dgpf.de/src/tagung/jt2018/proceedings/proceedings/papers/16_PFGK18_Noack.pdf (accessed on 7 October 2024).
24. Millmier, A.; Lorimor, J.; Hurburgh, C.J.R.; Fulhage, C.; Hattey, J.; Zhang, H. Near-infrared sensing of manure nutrients. Trans. ASAE; 2000; 43, pp. 903-908. [DOI: https://dx.doi.org/10.13031/2013.2986]
25. Quiñones, M.d.C.S.; Martinez, L.A.O.; Herrera, S.M.G.; Quiñones, O.M.R.; Laredo, R.F.G. Near-infrared spectroscopy (NIRS) applied to legume analysis: A review. Int. J. Eng. Innov. Technol.; 2018; 8, pp. 29-39.
26. Cozzolino, D.; Fassio, A.; Fernández, E.; Restaino, E.; La Manna, A. Measurement of chemical composition in wet whole maize silage by visible and near-infrared reflectance spectroscopy. Anim. Feed Sci. Technol.; 2006; 129, pp. 329-336. [DOI: https://dx.doi.org/10.1016/j.anifeedsci.2006.01.025]
27. Scotter, C.N.G. Non-destructive spectroscopic techniques for the measurement of food quality. Trends Food Sci. Technol.; 1997; 9, pp. 285-292. [DOI: https://dx.doi.org/10.1016/S0924-2244(97)01053-4]
28. Schmidt, W. Optical Spectroscopy in Chemistry and Life Sciences; Wiley-VCH Verlag: Weinheim, Germany, 2005.
29. Tsenkova, R.; Atanassova, S.; Kawano, S.; Toyoda, K. Somatic cell count determination in cow’s milk by near-infrared spectroscopy: A new diagnostic tool. J. Anim. Sci.; 2001; 79, pp. 2550-2557. [DOI: https://dx.doi.org/10.2527/2001.79102550x]
30. Horf, M.; Vogel, S.; Drücker, H.; Gebbers, R.; Olfs, H.-W. Optical spectrometry to determine nutrient concentrations and other physicochemical parameters in liquid organic manures: A review. Agronomy; 2022; 12, 514. [DOI: https://dx.doi.org/10.3390/agronomy12020514]
31. Wold, H. Soft modelling by latent variables: The non-linear iterative partial least squares (NIPALS) approach. J. Appl. Probab.; 1995; 12, pp. 117-142. [DOI: https://dx.doi.org/10.1017/S0021900200047604]
32. Mevik, B.-H.; Wehrens, R. The pls package: Principal component and partial least squares regression in R. J. Stat. Softw.; 2007; 18, pp. 1-23. [DOI: https://dx.doi.org/10.18637/jss.v018.i02]
33. Burns, D.A.; Ciurczak, E.W. Handbook of Near-Infrared Analysis; CRC Press: Boca Raton, FL, USA, Taylor & Francis Group: New York, NY, USA, 2008; Volume 35.
34. Restaino, E.A.; Fernández, E.G.; La Manna, A.; Cozzolino, D. Prediction of the nutritive value of pasture silage by near infrared spectroscopy (NIRS). Chil. J. Agric. Res.; 2008; 69, pp. 560-566. [DOI: https://dx.doi.org/10.4067/S0718-58392009000400011]
35. Saeys, W.; Mouazen, A.M.; Ramon, H. Potential for onsite and online analysis of pig manure using visible and near infrared reflectance spectroscopy. Biosyst. Eng.; 2005; 91, pp. 393-402. [DOI: https://dx.doi.org/10.1016/j.biosystemseng.2005.05.001]
36. Cabassi, G.; Cavalli, D.; Fuccella, R.; Marino Gallina, P. Evaluation of four NIR spectrometers in the analysis of cattle slurry. Biosyst. Eng.; 2015; 133, pp. 1-13. [DOI: https://dx.doi.org/10.1016/j.biosystemseng.2015.02.011]
37. Dagnew, M.; Crowe, T.; Schoenau, J. Measurement of nutrients in Saskatchewan hog manures using near-infrared spectroscopy. Can. Biosyst. Eng.; 2004; 46, pp. 33-37.
38. De Ferrari, G.; Gallina, P.M.; Cabassi, G.; Bechini, L.; Maggiore, T. Near infrared spectral analysis of cattle slurries from Lombardy (Northern Italy) breeding farms. NZNIRSS—The New Zealand Near Infrared Spectroscopy Society (Eds.) NIR 2005-NIR in Action. Making a Difference; NZNIRSS: Auckland, New Zealand, 2005; pp. 376-383.
39. Sørensen, P.; Birkmose, T.S.; Hansen, M.; Sørensen, L.K. Rapid analysis of animal slurry by NIR spectroscopy. Proceedings from the Seminar Manure—An Agronomic and Environmental Challenge; Stenberg, M.; Nilsson., H.; Brynjolfsson, R.; Kapuinen, P.; Morken, J.; Søndergaard, B. Nordiska Jordbruksforskares Förening (NJF): Uppsala, Sweden, 2005; pp. 94-95.
40. Evangelista, C.; Basiricò, L.; Bernabucci, U. An overview on the use of near infrared spectroscopy (NIRS) on farms for the management of dairy cows. Agriculture; 2021; 11, 296. [DOI: https://dx.doi.org/10.3390/agriculture11040296]
41. Williams, P. Near-Infrared Technology—Getting the Best Out of Light; PDK Grain: Nanaimo, BC, Canada, 2003.
42. Mouazen, A.M.; Saeys, W.; Xing, J.; de Baerdemaeker, J.; Ramon, H. Near infrared spectroscopy for agriculture materials: An instrument comparison. J. Near Infrared Spectrosc.; 2005; 13, pp. 87-97. [DOI: https://dx.doi.org/10.1255/jnirs.461]
43. Williams, P.; Eising, E.; Malley, D.F. Industrial-scale continuous on-line analysis of liquid hog manure by NIRS. NIR News; 2020; 31, pp. 25-29. [DOI: https://dx.doi.org/10.1177/0960336020978714]
44. Zimmermann, A.; Hartung, E. Nährstoffgesteuerte Gülleausbringung; Christian-Albrechts-Universität, Institut für Landwirtschaftliche Verfahrenstechnik: Kiel, Germany, 2009; pp. 1-130.
45. Horf, M.; Gebbers, R.; Olfs, H.-W.; Vogel, S. Effects of sample pre-treatments on the analysis of liquid organic manures by visible and near-infrared spectrometry. Heliyon; 2024; 10, e27136. [DOI: https://dx.doi.org/10.1016/j.heliyon.2024.e27136] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/38463763]
46. Zimmermann, A.; Thiessen, E.; Hartung, E. Near-Infrared Spectroscopy (NIRS) for a Nutrient Based Slurry Application. Available online: https://www.cabidigitallibrary.org/doi/pdf/10.5555/20103074871 (accessed on 27 August 2024).
47. Stenberg, B.; Gustafsson, K. On-line measurement of animal and bio slurry quality variations with near infrared spectroscopy. Precis. Agric.; 2013; 13, pp. 337-342.
48. Thiessen, E.; Eslamifar, M.; Kock, R.; Hartung, E. On farm validation of different NIR sensors for manure sensing. VDI-Berichte; 2022; 2406, pp. 9-16.
49. DLG. Qualität auf den Punkt gebracht. Available online: https://www.dlg.org/tests/landtechnik-betriebsmittel/dlg-auszeichnungen-im-ueberblick#c30255 (accessed on 7 October 2024).
50. DLG. Test Report 7122 (m-u-t GmbH). Available online: https://pruefberichte.dlg.org/filestorage/7122.pdf (accessed on 21 August 2024).
51. DLG. Test Report 6886 (John Deere). Available online: https://pruefberichte.dlg.org/filestorage/6886.pdf (accessed on 21 August 2024).
52. DLG. Test Report 6887 (John Deere). Available online: https://pruefberichte.dlg.org/filestorage/6887.pdf (accessed on 21 August 2024).
53. DLG. Test Report 7281 (John Deere). Available online: https://pruefberichte.dlg.org/filestorage/7281.pdf (accessed on 21 August 2024).
54. DLG. Test Report 7139 (Dinamica Generale S.p.A.). Available online: https://pruefberichte.dlg.org/filestorage/7139.pdf (accessed on 21 August 2024).
55. DLG. Test Report 7140 (Dinamica Generale S.p.A.). Available online: https://pruefberichte.dlg.org/filestorage/7140.pdf (accessed on 21 August 2024).
56. DLG. Test Report 7235 (Dinamica Generale S.p.A.). Available online: https://pruefberichte.dlg.org/filestorage/7235.pdf (accessed on 21 August 2024).
57. Schlagge, B. Was NIR-Sensoren können. Überzeugt Nahinfrarotspektroskopie in der Düngepraxis?. Bauernblatt; 2022; 22, pp. 8-10.
58. Hoffmann, A.; Lichti, F.; Olfs, H.-W.; Tillmann, P.; Severin, K.; Rest, T. Die Nahinfrarotspektroskopie (NIRS) zur Untersuchung von Güllen und Gärresten. Standpunkt des Verband Deutscher Landwirtschaftlicher Untersuchungs- und Forschungsanstalten e.V. (VDLUFA). Available online: https://www.vdlufa.de/wp-content/uploads/2022/06/2019-Standpunkt-NIRS-zur-Untersuchung-von-Guellen-und-Gaerresten-1.pdf (accessed on 7 October 2024).
59. Rubenschuh, U.; Volz, F. Nährstoffgehalte in Gülle online mit Sensoren bestimmen. Available online: https://www.dlg.org/fileadmin/downloads/Kompakt/DLGKompakt_08_19.pdf (accessed on 4 October 2024).
You have requested "on-the-fly" machine translation of selected content from our databases. This functionality is provided solely for your convenience and is in no way intended to replace human translation. Show full disclaimer
Neither ProQuest nor its licensors make any representations or warranties with respect to the translations. The translations are automatically generated "AS IS" and "AS AVAILABLE" and are not retained in our systems. PROQUEST AND ITS LICENSORS SPECIFICALLY DISCLAIM ANY AND ALL EXPRESS OR IMPLIED WARRANTIES, INCLUDING WITHOUT LIMITATION, ANY WARRANTIES FOR AVAILABILITY, ACCURACY, TIMELINESS, COMPLETENESS, NON-INFRINGMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Your use of the translations is subject to all use restrictions contained in your Electronic Products License Agreement and by using the translation functionality you agree to forgo any and all claims against ProQuest or its licensors for your use of the translation functionality and any output derived there from. Hide full disclaimer
© 2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.
Abstract
Nutrient levels in liquid organic manures (LOM) vary greatly, so it is important to determine the concentrations before field application in order to ensure that fertilisation is tailored to the crop requirements. Precise knowledge of the nutrient content in LOMs is a basic prerequisite for the optimum supply of these nutrients to crops and for avoiding environmental problems caused by over-fertilisation. The constituents of LOMs can be determined on site using various methods. One possibility is near infrared spectroscopy (NIRS). This method is already a common procedure for use in the laboratory. This review deals with the suitability of the use of NIRS for the characterisation of LOMs on farm. For on-farm applications, there are many factors such as the ambient temperature or movements and vibrations of the machines which can influence the measurement with the sensors and thus also the measured values. The influencing factors should therefore be taken into account. The reliability of NIRS systems for the on-farm analysis of liquid manure is verified by the German Agricultural Society. For the tests, various LOMs from different farms are measured with NIRS sensors and the quality of the agreement of the NIRS data with laboratory tests is certified for the respective ingredients for each LOM type. In order to exploit the full potential of the NIRS technology in the future, the indispensable calibrations need to be expanded and improved so that the sensors deliver precise and reproducible results for the different LOM types in practical applications.
You have requested "on-the-fly" machine translation of selected content from our databases. This functionality is provided solely for your convenience and is in no way intended to replace human translation. Show full disclaimer
Neither ProQuest nor its licensors make any representations or warranties with respect to the translations. The translations are automatically generated "AS IS" and "AS AVAILABLE" and are not retained in our systems. PROQUEST AND ITS LICENSORS SPECIFICALLY DISCLAIM ANY AND ALL EXPRESS OR IMPLIED WARRANTIES, INCLUDING WITHOUT LIMITATION, ANY WARRANTIES FOR AVAILABILITY, ACCURACY, TIMELINESS, COMPLETENESS, NON-INFRINGMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Your use of the translations is subject to all use restrictions contained in your Electronic Products License Agreement and by using the translation functionality you agree to forgo any and all claims against ProQuest or its licensors for your use of the translation functionality and any output derived there from. Hide full disclaimer
Details
; Hans-Werner Olfs 3
1 Section of Plant Nutrition and Crop Physiology, Department of Crop Sciences, Georg-August-Universität Göttingen, Carl-Sprengel-Weg 1, 37075 Göttingen, Germany;
2 Section of Plant Nutrition and Crop Physiology, Department of Crop Sciences, Georg-August-Universität Göttingen, Carl-Sprengel-Weg 1, 37075 Göttingen, Germany;
3 Department Plant Nutrition and Crop Production, Faculty of Agricultural Science and Landscape Architecture, University of Applied Sciences, Am Krümpel 31, 49090 Osnabrück, Germany;