Browsing by Subject "Spektralphotometrie"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Publication
    Spektralphotometrische Bestimmung des pflanzenverfügbaren Nitrats in der Bodenlösung : Entwicklung einer in-situ Messmethode zur Optimierung der Fertigation im intensiven Gemüsebau
    (2015) Mayer, Stephan; Müller, Torsten
    For many specialized cultivations, mainly in intensive horticulture, a slight N-deficiency may dramatically reduce crop quality and yield. Hence, especially in this case, fertilizer is often applied in surplus. Compared to the real N-demand of a crop, this results in a large investment in fertilizer and may also promote nitrogen loss due to leaching. At many sites, the consequences of this are nitrate contaminated ground and surface water. An in-situ method for the continuous determination of the actual and plant-available nitrate content in soil solution, which may be fundamental for adapted and culture specific fertilization, has not yet been created for practical use in horticulture. Such a method would minimize excessive nitrate loads, reduce fertilizer use as well as omit labor required for soil sampling for Nmin determination. The aim of this thesis is to develop such an in-situ method. For this purpose, nitrate was detected and quantified in soil solution by ultraviolet spectrophotometry. Based on these measurements and a comparison with the N-demand of the chosen crop, an adapted fertilization plan was established. Recovery of the necessary soil solution was carried out in the field with the aid of suction cups, which were connected to a vacuum system, which directed the solution to a measuring cell where the spectrophotometric measurement was performed. The data was collected and evaluated on an external server. After the calculation of the actual nitrate concentration, based on the spectral data, and comparing them to the N-demand, the need and amount of fertilization was determined. This process is performed automatically. Based on numerous lab experiments, one pot and two greenhouse experiments, the suitability of the nitrate-online-measurement-system (NITROM) for the determination of nitrate concentration in soil solution was tested, calibrated and validated on a total of twelve soil types, two gardening substrates and three different cultures. For the calculation of the nitrate concentration from spectral data, simple and multiple linear regressions (SLR, MLR), as well as polynomial multiple regression (PMR), were compared. Lab results of the nitrate UV absorption between 230 and 260 nm in pure nitrate standards (0 – 1000 mg NO3-- L-1) showed highly linear relationships for several wavelengths (231 und 240 nm: R2 > 0.999, p < 0.001). Low nitrate concentrations (0 – 150 mg L-1) were precisely determined between 230 and 240 nm, while high nitrate concentrations (150 – 1000 mg L-1) were determined between 240 and 250 nm. For UV measurements in soil solution with several interfering substances, no linearity was achieved (see below). The predominant interfering substances for nitrate measurements in the field are aromatic and alkene compounds in dissolved organic carbon (DOC). The complex structure of DOC may only be considered in a calibration through a multi-wavelength approach, accounting for wavelengths from the ranges of high and low nitrate concentrations as well as the reference range without nitrate absorption (250 – 260 nm). The PMR, in comparison to SLR and MLR, fit best for the estimation of nitrate concentration from spectral data. This can be seen in the field data obtained from the first greenhouse experiment (PMR: R2 = 0.963, p < 0.001; MLR: R2 = 0.948, p < 0.001; SLR 232 nm: R2 = 0.093, p = 0.047). The pot experiment with three different soil types and two gardening substrates allowed i.a. conclusions of DOC quality on different sites and confirms the need for a site-specific calibration of the measurement method. During the second greenhouse experiment, the entire NITROM technology was tested. Data obtained during half an hour intervals over a period of five weeks was evaluated online. Using a PMR calibration, a soil nitrate content curve (n = 998) was created and fertilizer loads were adapted. The fertilization events are clearly recognizable as distinct peaks in the graph. The PMR calibration (with 15 wavelength and n = 36) was highly significant (p = 0.001) and had a R2 > 0.999. The validation of the calibration reveals a relative estimation error of 6.1 %. The suitability of this method for in-situ determination of the nitrate concentration, as well as an adapted and culture specific fertilization management on the basis of measured data, can be confirmed. Further improvements to refine the measurement technology and evaluation procedure, as well as calibration of the method for DOC, are planned.