Spectroscopic analysis (UV-VIS-NIR) for predictive modeling of macro and micronutrients in grapevine leaves
Assessing nutrient concentrations in grapevines is crucial not only for the overall physiology of the plant but also for the quality of the resulting wine. Accurate determinations are also relevant for enhancing nutrient use efficiency and formulating fertilizer recommendations. Hence, there is a co...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-03-01
|
| Series: | Smart Agricultural Technology |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2772375525000462 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Assessing nutrient concentrations in grapevines is crucial not only for the overall physiology of the plant but also for the quality of the resulting wine. Accurate determinations are also relevant for enhancing nutrient use efficiency and formulating fertilizer recommendations. Hence, there is a considerable demand for a swift technique to analyze vine organs. Diffuse reflectance spectroscopy coupled with chemometric methods emerges as a potent, cost-effective, and environmentally friendly analytical technique for determining nutrient concentrations in plants. The objective of this study is to ascertain the viability of wide range spectrum (190–2600 nm) spectroscopy in providing precise insights into the nutritional status of vines. Our investigation specifically targets on the determination of C, N, P, K, Ca, Mg, B, Cu, Fe, Mn, Zn, Na, and Al in vine leaves from different wine growing areas, varieties and harvest years. Partial Least Squares Regression (PLS-R) was employed to construct models for the concentrations of these nutrients based on the reflectance measurements of the leaves. The model was trained using 70 % of the samples, while the remaining 30 % constituted the independent validation. Results from the validation set indicated accurate validation for most nutrients, with determination coefficients (r2) of 0.70 for C, 0.72 for N, 0.64 for P, 0.75 for K, 0.84 for Ca, 0.48 for Mg, 0.45 for B, 0.58 for Cu, 0.26 for Fe, 0.82 for Mn, 0.50 for Zn, 0.90 for Na, and 0.69 for Al. The findings revealed that reflectances in the visible (VIS) region of the spectrum played a key role in predicting micronutrients like B, corresponding with photosynthetic pigments (chlorophylls and carotenoids). In contrast, reflectances in the near-infrared region (NIR) had a greater impact on macronutrient prediction, particularly for P and Mg, due to their stronger interaction with organic compounds. The ultraviolet (UV) range played a minor role, highlighting the predominant importance of the VIS-NIR regions in spectroscopic analyses.Finally, the results support the potential of this technique for swiftly and non-invasively predicting both macro and micronutrient levels in grapevine plants, and facilitate the fertilization planning using variety-specific reference levels, or precision viticulture adapted to site-specific demands, including spatial intra-plot variability. |
|---|---|
| ISSN: | 2772-3755 |