Browsing by Subject "Nahinfrarotspektroskopie"

Now showing 1 - 2 of 2
  • Thumbnail Image
    Publication
    The potential of near-infrared spectroscopy (NIRS) as a rapid method to evaluate the aroma profiles of winemaking grapes
    (2023) Gehlken, Jana; Zörb, Christian
    The production of high-quality wines requires the use of high-quality grapes. For wines produced from certain grape varieties or grapes grown in specific cultivation regions, certain aroma impressions are expected, which often result from compounds already present in the grapes. Other compounds may negatively influence odour or flavour of the wine in their original form or as precursors for off-odours or off-flavours. Therefore, a rapid evaluation of grape aroma quality directly upon receival at the winery is advantageous. Grape aroma is usually evaluated by tasting, which leads to subjective results. Common instrumental analysis for aroma evaluation is time- and cost-intensive, which makes it inappropriate for rapid quality evaluation. The use of near-infrared (NIR) spectroscopy allows the simultaneous determination of various compounds without previous sample preparation. Moreover, the measurement is non-destructive, which allows an easy integration into existing processes and real-time analysis without additional sampling. Preliminary studies on the use of NIR spectroscopy for the evaluation of fruits and fruit products, with special regards to grapes, wine and wine aroma, were reviewed. In fruits and fruit products, a large number of different compounds was examined for maturity and quality evaluation. Only in few studies, sensory attributes were examined for quality evaluation and no published reports were available about the determination of specific fruit aroma compounds. In grapes and wine, NIR spectroscopy was mainly used for maturity determination and control of the alcoholic fermentation. For the evaluation of wine quality, numerous different compounds as well as the differentiation of grape varieties and cultivation regions were examined. For aroma evaluation, both specific aroma compounds and sensory attributes were examined in finished wines. Grape aroma was only evaluated in very few studies and in all of them glycosylated compounds were examined. Although these compounds play a key role in wine aroma, grapes contain numerous other compounds from various chemical classes, which may also considerably influence the aroma of the wine. Therefore, it is beneficial to acquire precise knowledge about the total aroma profile of winemaking grapes. Moreover, the majority of the preliminary studies was only performed under laboratory conditions. Based on these findings, the aroma profiles of grapes upon receival at wineries were examined. Grape mash samples from different wineries were examined for aroma compounds by gas chromatography-mass spectrometry (GC-MS). Instead of previous extraction of the aroma compounds, the conditions for measurement were chosen as similar as possible to those in the mouth (e.g. temperature, pH value, enzymes), to detect only compounds, which are released during tasting. The determined concentrations of the aroma compounds were related to the corresponding NIR spectra to calculate calibration models. Additionally, selected aroma compounds were grouped according to their flavour descriptions and calibration models for the determination of sensory attributes were developed. The aroma profiles of several grape mash samples showed clear differences. In some grape varieties, high concentration of four different terpenoids were measured, while other varieties showed considerably lower concentrations or no detection was possible. The concentrations of these four compounds allowed a rough classification of the samples into different aroma types. The phytosanitary status of the grapes also influenced the aroma profiles. Samples, which had been classified as “rotten” in previous examinations, showed significantly higher concentrations of various aroma compounds than samples from “sound” grapes. The addition of artificial saliva solution before analysis had no significant effect on the concentrations of the aroma compounds. The calibration models for the single aroma compounds showed varying results. Separate calibration models for cultivation region and/or grape colour lead to increased prediction accuracies compared to global calibration models. For the determination of sensory attributes only global calibration models could be developed due to the small number of samples. All models already showed very high prediction accuracies. The calibration models for the sensory attributes were compared to the result from a sensory evaluation of grape model solutions, which contained the same aroma compounds used for the calculation of the reference values for the development of the calibration models. The results from the sensory evaluation showed a large variation of the single values, however, the average values for the single odour and flavour impressions partly showed clear differences. In most cases, this could be explained by high concentrations of certain aroma compounds, which were related to the same sensory impressions for calibration model development (e.g. linalool for “floral”). However, grouping of the results according to the different model solutions was not possible. The calibration models for the single aroma compounds allow a semi-quantitative determination of numerous compounds by NIR spectroscopy, partly even a quantitative determination is possible. The intensity of various sensory attributes can also be determined. However, to ensure sufficient prediction accuracies, further samples need to be evaluated to extend the calibration models. The use of NIR spectroscopy upon receival at the winery enables an immediate evaluation of grape aroma quality, which can be used as an additional quality parameter. Furthermore, decisions can be made about the treatment of the grapes based on the results to choose the optimum conditions for the production of high-quality wines.
  • Thumbnail Image
    Publication
    Ruminal degradation characteristics of barley, rye, and triticale grains assayed in situ and in vitro, and by near-infrared spectroscopy
    (2017) Krieg, Jochen; Rodehutscord, Markus
    The milk yield of dairy cows and related energy and protein requirements have steadily increased in the last few decades. Since feed intake has not increased to the same extent as nutritional requirements, the concentration of nutrients in mixed rations had to be increased. An increase in energy concentration is often achieved by the inclusion of high levels of cereal grains. In the EU—apart from wheat—barley, rye, and triticale are widely cultivated cereal grains. Starch (ST), followed by crude protein (CP), is the main constituent of cereal grains. The rate and extent of ruminal CP and ST degradation can influence the performance and health of dairy cows, but data that can enable the comparison of ruminal degradation within and between barley, rye, and triticale grains are scarce. Commonly used techniques to explore ruminal degradation of feed are in situ and in vitro incubations. Both techniques require ruminal-fistulated animals, but alternative methods are being demanded by the community, in order to reduce the number of animal trials. An approach with the potential to estimate the nutritional value of various feeds is near-infrared spectroscopy (NIRS). The present thesis has two major parts. In the first part, ruminal degradation parameters and the effective degradability (ED) of DM, CP, and ST from barley, rye, and triticale grains are investigated using standardised in situ and in vitro incubation techniques. A total of 20 genotypes per grain species were used. In the second part, NIRS calibrations were developed with the aim of estimating the CP and ST concentrations of cereal grains and their incubation residues. Subsequently, data from in situ experiments were used to establish the calibrations for estimating the ruminal in situ degradation of cereal grains from their spectral data. In situ degradation studies have been conducted by ruminal incubation, utilising three lactating cows. Ruminal degradation parameters and ED (ruminal passage rate = 8%//h) were calculated. For in vitro incubations, the samples were incubated in a rumen fluid-buffer mixture (‘Hohenheim Gas Test’). The gas production was recorded for estimating gas production kinetics. In vitro gas production—in combination with crude nutrient concentrations—was used to estimate the metabolisable energy concentration (ME) and digestibility of organic matter (dOM). The degradation rates differed between and within the grain species for DM, CP, and ST. The variation within grain species was not reflected in the ED of CP and ST, due to the relatively fast and almost complete degradation of the grains. The ED of CP was 77% (69–80%) for barley, 85% (83–86%) for rye, and 82% (79–84%) for triticale. The corresponding ED of ST was 86% (82–88%), 95% (92–96%), and 94% (90–95%). Accordingly, the estimated ME (barley: 13.5 MJ/kg DM, rye: 13.9 MJ/kg DM, triticale: 13.5 MJ/ kg DM) showed only relatively minor variation within one grain species. The dOM was overall at a high level (barley: 91.3%, rye: 95.3%, triticale: 95.8%). The relatively small variation within one grain species could not be explained by the chemical and physical characteristics of the samples. Hence, it was concluded that it is feasible to use mean values for every species in feed formulation and ration planning. In the second part of this thesis, it was shown that it is possible to replace chemical CP and ST analyses of samples from in situ studies by NIRS without affecting the calculated ruminal degradation characteristics. NIRS could be used to estimate the ED of CP and ST from cereal grains. The sample set to establish the calibrations included barley, durum, maize, rye, triticale, and wheat grains. Calibrations for the CP and ST concentration were extended to pea samples. The calibrations with the best validation performance for CP and ST concentration were obtained by using the wavelength segment of 1250 to 2450 nm and the first derivative of the spectra (CP: R2 = 0.99; SEP = 0.46% DM. ST: R2 = 0.99; SEP = 2.10% DM). The results of in situ studies did not differ, irrespective of whether chemical or NIRS analysis was used. Like the CP and ST concentration, the ED was estimated with a high accuracy (ED8 CP: R2 = 0.95; SEP = 2.43%. ED8 ST: R2 = 0.97; SEP = 2.45%). However, calibrations need to be extended before they can be recommended for routine use. The present thesis demonstrates that the ED of CP and ST of barley, rye, and triticale grains differ between the species, but variation within one grain species is relatively small and not related to the chemical and physical characteristics of the grain. Hence, under the prevailing cultivation conditions, the mean values for each grain species in feed evaluation are deemed adequate. It was demonstrated that NIRS has the potential to facilitate the evaluation of the nutritive value of cereal grains for ruminants.