Elsevier

Biosystems Engineering

Volume 197, September 2020, Pages 76-90
Biosystems Engineering

Research Paper
Non-destructive and rapid measurement of sugar content in growing cane stalks for breeding programmes using visible-near infrared spectroscopy

https://doi.org/10.1016/j.biosystemseng.2020.06.012Get rights and content

Highlights

  • NIR has potential for CCS monitoring of growing cane in breeding programmes.

  • Variation in sample temperatures (25–45 °C) did not affect spectral reproducibility.

  • PLS models provided good accuracy and could be used for screening.

  • Instrument temperature, which varied in field application, caused model error.

  • Modelling should be performed based on individual measurement periods of the day.

The efficient selection of sugarcane varieties with enhanced sugar content requires simple, rapid, accurate and cost-effective assays. The objective of this research was to develop spectroscopic models for non-destructive evaluation of Commercial Cane Sugar (CCS) in growing cane stalks. A portable visible-shortwave near-infrared (Vis/SWNIR) spectrometer with a wavelength range of 570–1031 nm was applied to cane stalks grown under normal field conditions. The CCS models were developed by partial least squares (PLS) regression using spectra sets obtained at three different times (i.e. morning, afternoon and evening), based on both individual spectra and a combined set. During the in-field measurements, it was found that model performance could be affected by varying sample and instrument temperatures, especially for the combined set. The models had coefficients of determination of the prediction set (r2) of 0.76, 0.76, 0.78 and 0.69 and root mean square errors of prediction (RMSEP) of 1.01, 1.05, 0.99 and 1.17 CCS for the models constructed with the spectra sets of the morning, afternoon, evening and combined, respectively. These results indicate that the CCS models could be used for the monitoring and screening of sugarcane clone selection, except for the model obtained from the combined spectra set, which was influenced by varying instrument temperature. Thus, heat protection for the portable Vis/SWNIR instrument, allowing it to maintain a constant instrument temperature, should be considered if data collection across different periods of the day is necessary.

Introduction

Variety improvement has played an important role in sugar industries to improve sugarcane quality and especially sugar content in cane stalk. Improvement in cane quality has a high economic impact because it can increase sugar products without increasing costs of harvesting, cane transport or milling. Consequently, many countries consider the sugar content to be an important selection target in sugarcane breeding programmes. Sugar content in cane in Thailand and Australia is indicated by a measurement unit called commercial cane sugar (CCS) (Albertson & Grof, 2004). To obtain the highest profitability, sugarcane should be harvested when it reaches its highest CCS value. Thus, the knowledge of CCS accumulation in each sugarcane variety is important for both farmers and sugar mills alike.

Nowadays there are three conventional methods of CCS measurement in breeding programmes: (a) a labour-intensive extraction method for analysis of juice and residual fibre that requires skill, is time-consuming and needs some chemical substances (Albertson & Grof, 2004); (b) an extracting or shredding method for analysis of juice and residual fibre using a Fourier transform near-infrared (FT-NIR) spectrometer (Taira, Ueno, Saengprachatanarug, & Kawamitsu, 2013); (c) the genetic correlation method, which determines the correlation between CCS and physiological observations, such as cane yield (Jackson, 2005). Both methods (a) and (b) require the destruction of the cane stalks during laboratory analysis. Method (c) can be measured without destruction; however, its accuracy is yet to be quantified.

Roach (1989) has described three phases of sugarcane improvement in breeding programmes. The first phase focuses on selection and crossbreeding among clones of the main varieties. The second phase focuses on the development of interspecific hybrids, and the third phase is the exploitation of the interspecific hybrids. These three phases, when combined, commonly take more than ten years. Destructive testing cannot be used during the first and second phases because each clone has a specific characteristic representing each new variety. So, the CCS of each clone is only estimated using the relationship between CCS and physiological characteristics, which provides low accuracy. This limitation affects clone selection in both the first and second phases, which reduces the chances of obtaining a good clone. The selected clones in the second phase would be produced to confirm the exact values of CCS accumulation of each variety in the third phase. Thus, the third phase needs a large experimental area that is able to produce a sufficient number of stalks for monthly cane cutting and CCS measurements. Moreover, due to the requirement of a large experimental area, breeders can select only a few clones for studying in the third phase. All of these constraints lead to a costly and time-consuming process that leads to slow improvement in sugar content.

Besides the knowledge of CCS accumulation throughout the cane growth, breeders also need to follow up on the change in the chemical composition of an individual cane stalk when there is a change in plant growth conditions, such as climate, soil composition or rainfall. However, following up on the CCS for an individual cane stalk cannot be done using destructive methods. Accordingly, a method that can measure the CCS directly in an intact cane stalk without destruction is strongly needed for breeding programmes. A precise, non-destructive method would help breeders unlock the above limitations. Moreover, there will be a possibility to conduct experiments to improve sugar content based on a powerful control system (such as breeding programmes with a small trial size and a control system in chambers).

Presently, near-infrared (NIR) spectroscopy is widely used to determine the quality of agricultural products. It is a rapid and non-destructive method that can be applied without any sample preparation (Jamshidi, Minaei, Mohajerani, & Ghassemian, 2012; Schaare & Fraser, 2000; Sunoj, Igathinathane, & Visvanathan, 2016). NIR spectroscopy has also been used to assess the quality of fruit (Nascimento, Carvalho, Júnior, Pereira, & Teixeira, 2016; Saranwong, Sornsrivichai, & Kawano, 2004). In breeding programmes, Bailleres, Davrieux, and Ham-Pichavant (2002) and Yeh et al. (2005) used the NIR analysis as a tool for the rapid screening of some major wood characteristics in eucalyptus and pine, respectively. Their models allowed them to rapidly screen large breeding populations for a variety of phenotypic traits. Similarly, Shiroma-Kian, Tay, Manrique, Giusti, and Rodriguez-Saona (2008) and Ayvaz et al. (2016) employed Fourier transform infrared spectroscopy (FTIR) to screen potato breeding lines. In the sugar industry, however, such non-destructive methods for variety screening have not been applied. Research has been conducted on using a portable visible-shortwave near-infrared (Vis/SWNIR) instrument to predict pol and fibre content values of cane stalk with promising results (Maraphum et al., 2018; Phuphaphud, Saengprachatanarug, Posom, Maraphum, & Taira, 2019; Taira, Ueno, Saengprachatanarug, et al., 2013). Accordingly, there is a possibility to predict the CCS of growing cane stalk directly in the field, although previous studies have only been carried out in the laboratory.

However, the practicality of using NIR in the field is constrained by the influence of unknown factors on the spectra in the NIR region. According to previous studies, it is well known that NIR spectra can be very sensitive to temperature (Chapanya, Ritthiruangdej, Mueangmontri, Pattamasuwan, & Vanichsriratana, 2019; Thamasopinkul et al., 2017; Xie & Shi, 2013). Wülfert, Kok, and Smilde (1998) found that the isolated molecular features and inter- or intramolecular features, shown from vibrational spectra, were affected by sample temperature. A high sample temperature means that there is a lot of energy stored in a molecule. So, the energy absorption requirement of the molecules in the high temperature sample is less than that of the low temperature sample, despite the same structure and concentration. Many studies recommended that, when using NIR spectra, one should avoid the variation of sample temperature since the weaker hydrogen bonding of water is easily influenced by temperature, which causes changes in model accuracy (Cozzolino et al., 2007; Hageman, Westerhuis, & Smilde, 2005; Wülfert et al., 1998). Nevertheless, it is not easy to maintain samples at a constant temperature for practical use in the field.

In tropical countries, ambient temperatures can vary widely throughout the day. Spectra obtained in the morning may have different characteristics compared to spectra measured in the afternoon. Accordingly, the objective of this study was to investigate the effect of field temperature on Vis/SWNIR spectra and the predictive ability of calibration models. Commonly, field temperature can easily affect the sample temperature. Thus, the development of a calibration equation that can compensate for a variety of sample temperatures was examined in this study.

Section snippets

Sample preparation

The cane samples in this study were supplied by the Faculty of Agriculture at Khon Kaen University from the 2018–2019 harvest season. To obtain a robust model to be applied in breeding programmes, three varieties of sugarcane were included in this study. These three varieties consisted of KK3, K88-92 and UT84-12, which are the major commercial cultivars in Thailand. They are very important for breeding in Thailand because, before any new cultivars are released, their important features, such as

Overview of optical measurements of cane stalk and statistical characteristics of sugar content

The second derivative spectra of a cane sample, scanned at different sample temperatures, are presented in Fig. 2. A strong absorption band was observed at approximately 966 nm, related to the O–H stretching mode (Vanoli et al., 2012). A slight shift in peak intensity was observed, which tended to decrease with temperature. This result was similar to that obtained by Kawano, Abe, and Iwamoto (1995), who used a Vis/SWNIR spectrometer to determine Brix value in intact peaches. The effect was

Conclusions

The results of this study demonstrate that NIR spectroscopy could be used as a non-destructive method for in-field CCS prediction of growing cane stalks, which will be a helpful tool for monitoring changes in CCS values of growing cane stalks in breeding programmes. Increases in detector temperature led to shifts in spectra, which tended to increase with temperature. The results of the investigation into the effects of sample temperatures and detector temperatures on the spectral

Declaration of Competing Interest

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.

Acknowledgements

We would like to acknowledge the financial support from the Royal Golden Jubilee PhD Scholarship (PHD/0152/2561) of Thailand Research Fund (TRF), Applied Engineering for the Important Crops of the North East Research Group of Khon Kaen University, and the Northeast Thailand Cane and Sugar Research Centre, Faculty of Agriculture, Khon Kaen University for providing sugarcane samples. In addition, the authors wish to thank the Khon Kaen Field Crop Research Centre of Khon Kean, Thailand for

References (43)

  • P. Sirisomboon et al.

    Improvement of proximate data and calorific value assessment of bamboo through near infrared wood chips acquisition

    Renewable Energy

    (2020)
  • S. Sunoj et al.

    Nondestructive determination of cocoa bean quality using FT-NIR spectroscopy

    Computers and Electronics in Agriculture

    (2016)
  • J. Wang et al.

    Development of multi-cultivar models for predicting the soluble solid content and firmness of European pear (Pyrus communis L.) using portable vis–NIR spectroscopy

    Postharvest Biology and Technology Journal

    (2017)
  • Junjie Xue et al.

    Study of the influence of NIRS acquisition parameters on the spectral repeatability for on-line measurement of crop straw fuel properties

    (2014)
  • R. Zornoza et al.

    Near infrared spectroscopy for determination of various physical, chemical and biochemical properties in Mediterranean soils

    Soil Biology and Biochemistry

    (2008)
  • P.L. Albertson et al.

    The effect of hexose upon pol, brix and calculated CCS in sugarcane: A potential for negative pol bias in juice from a actively growing cane

    Journal American Society Sugar Cane Technologists

    (2004)
  • H. Bailleres et al.

    Near infrared analysis as a tool for rapid screening of some major wood characteristics in a eucalyptus breeding program

    Annals of Forest Science

    (2002)
  • H. Bornefalk et al.

    Effect of temperature variation on the energy response of a photon counting silicon CT detector

    IEEE Transactions on Nuclear Science

    (2013)
  • Bureau of sugar experiment stations

    Laboratory manual for Australian sugar mills. Australia

  • P. Chapanya et al.

    Temperature compensation on sugar content prediction of molasses by near-infrared spectroscopy (NIR)

    Sugar Technology

    (2019)
  • X. Chen et al.

    Nondestructive differentiation of panax species using visible and shortwave near-infrared spectroscopy

    Food and Bioprocess Technology

    (2011)
  • Cited by (19)

    • Real-time measuring energy characteristics of cane bagasse using NIR spectroscopy

      2022, Sensing and Bio-Sensing Research
      Citation Excerpt :

      For practical use, partial least squares regression (PLS) is a famous modeling technique. Several articles have documented its efficiency in developing rapid spectral analytical screening methods [7,14–19]. Meanwhile, the optimal spectral pre-processing depended on the characteristics of the sample and wavelength bands.

    • Fast determination of total solids content (TSC) and dry rubber content (DRC) of para rubber latex using near-infrared spectroscopy

      2022, Industrial Crops and Products
      Citation Excerpt :

      The strongest point is that it has been used to measure the sample in a non-destructive manner (Maraphum et al., 2020, 2022). Based on these advantages, it has been applied in several processes, such as assessing prices (Taira et al., 2010, 2017), breeding programmes (Maraphum et al., 2018; Phuphaphud et al., 2020), etc. The pre-processing of NIR spectral data has become an integral part of chemometrics modelling.

    • Achieving robustness across different ages and cultivars for an NIRS-PLSR model of fresh cassava root starch and dry matter content

      2022, Computers and Electronics in Agriculture
      Citation Excerpt :

      It has also been used to measure other parameters in fruits and vegetables, such as titratable acidity or pH level (Ying and Liu, 2008) and pectin constituents in Japanese ‘Housui’ pears (Sirisomboon et al., 2007). NIR spectroscopy has many advantages, and among other applications, it has been used in breeding programmes (Maraphum et al., 2018; Phuphaphud et al., 2019; Phuphaphud et al., 2020; Maraphum et al., 2020a) and for product guarantees (Kawano et al., 1992; Kawano et al., 1993; Miyamoto and Kitano, 1995). Maraphum et al. (2020b) have previously reported developing a model for directly estimating SC in cassava tubers.

    • Handheld near infrared spectrometer and machine learning methods applied to the monitoring of multiple process stages in industrial sugar production

      2022, Food Chemistry
      Citation Excerpt :

      In relation to the cane quality monitoring, classification models based on on-line NIR data from shredded cane have been developed to discriminate between typical and atypical cane samples, in order to identify samples with abnormally low quality (Sexton et al., 2020). Moreover, a number of works have investigated the use of portable instruments in the NIR and visible and shortwave near infrared (Vis/SW-NIR) regions to measure quality parameters such as fiber, brix, pol, commercial cane sugar and solid density in sugarcane stalks (Nawi et al., 2013; Phuphaphud et al., 2020; Sanseechan et al., 2018; Taira et al., 2013). These works were mainly focused on in-field applications.

    View all citing articles on Scopus
    View full text