Onsite nutritional diagnosis of tea plants using micro near-infrared spectrometer coupled with chemometrics

https://doi.org/10.1016/j.compag.2020.105538Get rights and content

Highlights

  • Smartphone-based NIRS technique was used for onsite nutritional diagnosis.

  • Two tea plant varieties widely cultivated in China were analyzed and studied.

  • Two novel algorithms, VCPA and VCPA-GA, were introduced to simplify the models.

  • The simplified models for three photosynthetic pigments yielded good performance.

Abstract

A rapid and accurate diagnosis of nutritional status in field crops is crucial for site-specific fertilizer management. The micro near-infrared spectrometer (Micro-NIRS) is an extremely portable optical device that can be connected to a smartphone through a Bluetooth connection. In this study, a Micro-NIRS was used to evaluate pigment contents, namely chlorophyll a (Chl-a), chlorophyll b (Chl-b), and carotenoid (Car) in two varieties of field tea plants. A variable combination population analysis (VCPA), genetic algorithm (GA), and VCPA-GA hybrid strategy were used to select characteristic wavelengths; a partial least squares regression (PLSR) algorithm was employed for modeling. Results indicated that the simplified VCPA-GA-PLSR models provided the most favorable performance among all models for Chl-a, Chl-b, and Car content prediction; the correlation coefficients in prediction (Rps) were 0.9226, 0.9006, and 0.8313, respectively; the root mean square errors in prediction (RMSEPs) were 0.0952, 0.0771, and 0.0373 mg/g, respectively; the relative prediction deviations (RPDs) were 2.55, 1.92, and 1.79, respectively. Extracted characteristic variables occupied <13.63% of full spectra. The current work provided a useful example for implementing a smartphone-based Micro-NIRS system that can diagnose plant nutrition rapidly, nondestructively, and at low cost.

Introduction

Tea (Camellia sinensis L.) is a commercial crop, which is widely cultivated in China with a cultivation area of 2,930,500 ha in 2018. Fertilizer application, especially nitrogen application, considerably increases the yield and improves the quality of tea plant leaves. However, excessive or ineffective nitrogen application not only increases agricultural costs but also causes pollution of the environment and underground water (Saraswathy et al., 2007). The physiological indexes of each tea plant are closely related to the growth status and yield of tea leaves from that plant. Photosynthetic pigments, including chlorophyll and carotenoids, play essential roles in photosynthesis, and can be used as crucial indicators of plant nutrition status (Nishio, 2000). Studies have indicated that foliar chlorophyll concentrations are positively correlated with nitrogen content, which can help farmers to assess the nitrogen status of a crop and to optimize fertilization (Wang et al., 2019). Carotenoids provide much complementary information on vegetation physiological status and protect chlorophyll molecules from photo-oxidation under excessive light (Ge et al., 2011). The accurate diagnosis of photosynthetic pigment content in tea leaves is conducive for understanding the nutritional status of tea plants and providing rational fertilization.

Currently, many methods are used for quantitatively determining the contents of photosynthetic pigments in plants, such as spectrophotometry and high-performance liquid chromatography (HPLC) (Solymosi et al., 2012). However, these methods are laboratory-based, destructive to samples, time-consuming, and demand technical skills. Moreover, they require complex sample preparation and cannot provide determination. Therefore, accurate, rapid, and nondestructive determination methods must be developed. Near-infrared (NIR) spectroscopy is a mature detection technology with the advantages of rapidity, nondestructiveness, accuracy, and reliability. NIR spectroscopy can be used to describe the overtone and combination bands of C–H, O–H, and N–H groups in the spectral wavelength range of 780–2500 nm for analyzing most chemical compounds (Guo et al., 2016, Cui et al., 2019). Because photosynthetic pigments such as chlorophyll and carotenoids have hydrogen-containing groups, NIR can capture the relevant information of these pigments. Based on these principles, NIR was used by researchers to assess the photosynthetic pigment content of various plants (Zhang et al., 2016). He et al. reviewed the progress of spectral techniques applied to crops for the diagnosis of nutrient status. They concluded that NIR combined with chemometrics can be used to accurately predict the pigment content in various crops (He et al., 2015). However, most of the studies employed laboratory-based benchtop NIR spectrometers, which are costly and immobile. A small number of studies have used portable NIR spectrometers to diagnose pigments, such as ASD field spectrometers (Liu et al., 2019). However, the high cost of these instruments and the need of operation skills make it difficult for them to become routine testing tools for farmers. Therefore, it is necessary and advantageous for farmers to search for low-cost and extremely portable diagnostic tools for plant pigments for scientific crop management.

The micro near-infrared spectrometer (Micro-NIRS) is an extremely portable optical device that can be connected to a smartphone through a Bluetooth connection. It can acquire, record, and store spectral data, then upload data to cloud servers. Due to advantages over lab-based desktop NIRS in portability, price, and environmental requirements, Micro-NIRS systems have received considerable attention from researchers. Malegori et al. assessed the feasibility of one of the smallest NIR spectrometers on the market (MicroNIR 1700) for the evaluation of acerola fruit quality (titratable acidity and ascorbic acid content) during ripening. The results indicated that the predictive ability of the MicroNIR 1700 was comparable with that of a desktop FT-NIR spectrometer (Malegori et al., 2017). Similarly, Sun et al. used a MicroNIR 1700 for determining the glucosamine content in fermentation. The results of Passing–Bablok regression and paired t testing indicated no significant differences between the MicroNIR 1700 and FT-NIR spectrometers (Sun et al., 2018). Coronel-Reyes et al. employed a low-cost Micro-NIRS connected to a smartphone to determine egg storage time. Their success demonstrated that such devices can quantify an egg's freshness to an industrial standard of precision (Coronel-Reyes et al., 2018). These results indicated that the performance of the Micro-NIRS was comparable to that of any desktop NIRS in the qualitative and quantitative analysis of food quality. However, few studies have been conducted on evaluations for the stress diagnosis and nutritional diagnosis of plants.

Thus, the main goals of this study were to investigate the analytical performance of a smartphone-based Micro-NIRS and to evaluate the prediction accuracy in terms of direct applicability in the field. This study explored the effectiveness of a smartphone-based Micro-NIRS for estimation of tea plant photosynthetic pigment content under field conditions. The results of this novel study is expected to help in the onsite, rapid, and low-cost evaluation for the photosynthetic pigment content in tea plants, which will quickly diagnose and evaluate the nutritional status of tea plants. These promising results can provide support for the development of smartphone applications in the near future.

Section snippets

Sample preparation

Two varieties of tea plant, namely Nongkangzao (NKZ) and Longjing 43 (LJ 43), were selected as research materials. Both varieties have been identified and registered at the national level and have large areas of cultivation in China, where they are often used for green tea. We conducted our trials at High-tech Agricultural Garden in Anhui Agricultural University, Hefei city, Anhui province (31.90° N, 117.23° E).

Spectral data acquisition

In this study, we used a micro-NIRS (NIR-S-R2; InnoSpectra Corporation, Taiwan,

Pigment content distribution in the canopy leaves of the two tea plant varieties

Photosynthetic pigments, including chlorophyll and carotenoids, are crucial nutrient components of tea plants and can indicate their growth status. Moreover, chlorophyll is an essential green-color related component of green tea, which is positively correlated with the quality of green tea. We conducted statistical analysis of photosynthetic pigment content in the canopy leaves of two commercially noteworthy varieties of green tea, namely NKZ and LJ 43. As listed in Table 1, the Chl-a values in

Conclusion

A plant’s level of photosynthetic pigment can reflect the physiological status of that plant, and can provide information to guide rational fertilization. In this study, a smartphone-based Micro-NIRS was used for the on-site evaluation of photosynthetic pigment content in tea plants. Full-PLSR models were established for Chl-a, Chl-b, and Car. Moreover, VCPA, GA, and VCPA-GA algorithms were employed to select the characteristic wavelengths. Results showed that the simplified VCPA-GA-PLSR models

Declaration of interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

CRediT authorship contribution statement

Yu-Jie Wang: Conceptualization, Data curation, Formal analysis, Methodology. Shan-Shan Jin: Formal analysis, Methodology, Software. Meng-Hui Li: Data curation, Formal analysis, Methodology. Ying Liu: Data curation, Formal analysis. Lu-Qing Li: Visualization, Investigation, Supervision. Jing-Ming Ning: Writing - review & editing, Validation, Supervision. Zheng-Zhu Zhang: Funding acquisition, Project administration, Resources.

Acknowledgements

This work has been financially supported by the National Key Research and Development Program of China (2016YFD0200900, 2017YFD0400800), and Major scientific and technological projects of Anhui Province (18030701149, 18030701153).

References (24)

Cited by (0)

View full text