150 https://doi.org/10.37580/JSR.2019.2.9.

97-118 Journal of Sugarcane Research

https://doi.org/10.37580/JSR.2019.2.9.150-163
RESEARCH ARTICLE
PREDICTION MODELS FOR NON-DESTRUCTIVE ESTIMATION OF TOTAL
CHLOROPHYLL CONTENT IN SUGARCANE

Krishnapriya Vengavasi*, R. Arunkumar, R. Gomathi and S. Vasantha

Abstract
Total chlorophyll content of sugarcane is an important indicator of plant health, directly correlated to the
photosynthetic potential of the crop. With recent technological advancements, portable chlorophyll meters
have largely replaced biochemical chlorophyll estimation, requiring laborious extraction procedure with
solvents like acetone and dimethyl sulphoxide. Chlorophyll meters determine only ‘greenness’ index, which
has to be converted into scientifically standard units in order to make the data comprehensive. Prediction
models for inter-conversion of chlorophyll units are available for crops like rice, wheat, sorghum, barley,
maize, etc., but not for sugarcane till date. In the present study, total chlorophyll content was recorded in
diverse sugarcane germplasm and commercial hybrids using both non-destructive and destructive sampling
methods. A strong positive correlation was observed between meter readings (SPAD and CCI) with total
chlorophyll content estimated using 80% acetone (r = 0.800 and 0.793) and dimethyl sulphoxide (r = 0.915
and 0.868). Regression models for the best fit curve between meter reading and extracted chlorophyll values
of the tested sugarcane germplasm and hybrids were non-linear, polynomial equations of the second order.
The model developed was validated in an independent experiment wherein sugarcane variety Co 86032 was
subjected to increasing nitrogen levels. Highly significant linear regression was found between observed
and predicted values of all estimates of total chlorophyll content with almost negligible prediction error.
Thus, the model calibrated and validated for sugarcane germplasm and commercial hybrids would be a
small yet significant step towards aiding high-throughput phenotyping in sugarcane thereby accelerating
crop improvement programmes.

Keywords: Acetone; CCI; Chlorophyll; DMSO; SPAD; Sugarcane

Introduction and hence was widely accepted by researchers as

Chlorophyll content in leaves is one of the it reduced the burden of manual labour for tissue
most essential physiological traits which gives homogenisation (Holm-Hansen 1978; Barnes et
basic information about the crop’s potential to al. 1992; Porra et al. 1989; Porra 2002; Ritchie
photosynthesise and thereby indirectly correlated 2006). In the recent times, with technological
to overall growth and development. Classical advancements leading to high-throughput
experiments emphasised the importance of phenotyping, portable chlorophyll meters have
chlorophyll content estimation by destructive largely replaced the biochemical assays. One
sampling which required rigorous maceration of of the earliest of such hand-held meters was
the leaf tissue with solvents like acetone, methanol designed by the Soil-Plant Analysis Development
and ethanol to extract the pigments. Extraction unit of Minolta Camera Co. Japan (Rodriguez and
techniques involving better solvents like dimethyl Miller 2000), popularly called as the SPAD meter.
sulphoxide and dimethyl formamide completely The principle of chlorophyll estimation in the
bleached the chlorophyll pigments from the leaves portable meter is measurement of the ratio of light

Krishnapriya Vengavasi, R. Arunkumar, R. Gomathi and S. Vasantha

Plant Physiology Section, Division of Crop Production, ICAR-Sugarcane Breeding Institute, Coimbatore – 641 007,
Tamil Nadu, India
*Corresponding author: [email protected]
Submitted: 2 November 2019; Accepted: 2 March 2020
Journal of Sugarcane Research https://doi.org/10.37580/JSR.2019.2.9.97-118 151
https://doi.org/10.37580/JSR.2019.2.9.150-163

intensities transmitted from the leaf in response (Pisum sativum), lettuce (Lactuca sativa) and
to sequential illumination at two wavelengths i.e. pepper (Capsicum sp.).
red (650 nm) and infrared (940 nm). Many such
Sugarcane (Saccharum sp.) is a cash crop
meters are commercially available today with
cultivated mainly for its commercial value in
slight variation in wavelength of emitted light
the form of sugar and bio-fuel, predominantly
and the unit in which the measured chlorophyll
grown in Brazil, India and China. As in other
is expressed. For example, certain meters give
plant species, chlorophyll content in sugarcane
the ratio of transmittance at 940 nm to 660 nm,
influences photosynthesis and thereby biomass
called the Chlorophyll Content Index (CCI)
accumulation, partitioning and cane productivity.
(Richardson et al. 2002). In essence, the relative
Limited reports are available as on date about
chlorophyll content or the ‘greenness’ index in
the relationship between SPAD and extracted
the leaf is only being captured by non-destructive
chlorophyll in sugarcane under natural field
sampling. Converting these ratios or indices into
condition (Radhamani and Kannan 2016) or when
scientifically standard units of chlorophyll content
subjected to water stress (Jangpromma et al. 2010;
would require comparison of methodologies
Silva et al. 2013). Though these papers report
across crop species, varieties and a wide range of
positive correlation between meter recorded
leaf greenness per se. Development and validation
indices and extracted chlorophyll values, a
of such models/equations have been attempted in
reliable model and/or equation to convert one unit
several agricultural crops including rice (Oryza
to the other is yet to be established in sugarcane.
sativa) (Monje and Bugbee 1992; Yuan et al.
Development and validation of such a statistical
2016), wheat (Triticum aestivum) (Monje and
model to estimate the leaf chlorophyll content
Bugbee 1992; Castelli et al. 1996; Samsone et
in sugarcane by rapid non-destructive sampling
al. 2007; Uddling et al. 2007; Zhu et al. 2012),
methods is essential so as to reduce the resources
sorghum (Sorghum bicolor) (Yamamoto et al.
and time involved in extraction procedures. The
2002), barley (Hordeum vulgare) and maize (Zea
objective of this investigation was to develop such
mays) (Castelli et al. 1996; Zhu et al. 2012),
a model, with regression equations which would
soybean (Glycine max) (Monje and Bugbee 1992;
be valid across the sugarcane germplasm and
Castelli et al. 1996), pigeon pea (Cajanus cajan)
commercial hybrids. In the era of high-throughput
(Yamamoto et al. 2002), field bean (Phaseolus
phenotyping or ‘phenomics’, such rapid non-
vulgaris) (Samsoneet al. 2007), potato (Solanum
destructive techniques which could act as proxies
tuberosum) (Uddlinget al. 2007; Zhu et al. 2012),
to time- and resource-consuming biochemical
muskmelon (Cucumis melo) (Azia and Stewart
assays, are most essential to accelerate varietal
2001) and Amaranthus vlitus (Kapotis et al. 2003).
selection in breeding programmes.
This relationship has also been worked out in
some fruit trees (Schaper and Chacko 1991) and Materials and Methods
endangered plant species (Hawkins et al. 2009). Experimental details
Parry et al. (2014) have given an exhaustive
review and analysis of the relationship between Data for deriving the relationship between
destructive and non-destructive chlorophyll methods of chlorophyll estimation in sugarcane
sampling in 22 plant species including agricultural was recorded in five independent experiments at
crops viz. rice, wheat, maize, sorghum, barley, ICAR-Sugarcane Breeding Institute, Coimbatore
soybean, tomato (Solanum lycopersicum), pea during the cropping season 2018-19. Experiment
152 https://doi.org/10.37580/JSR.2019.2.9.97-118 Journal of Sugarcane Research
https://doi.org/10.37580/JSR.2019.2.9.150-163

I comprised 14 sugarcane germplasm clones V, sugarcane variety Co 86032 was subjected to

[Saccharum officinarum: Sarawak Unknown, increasing levels of nitrogen in semi-hydroponic
Cebu Light Purple, Bontha, Hawaii Original 24; culture under glasshouse condition with controlled
S. spontaneum: US 59-1-1; S. barberi: Katha environment. Single budded setts of Co 86032
Coimbatore, White Pindaria, S. sinense: Reha; S. raised in polybags were transferred to porcelain
robustum: NG 77-237; S. robustum inter-specific pots containing 4 kg gravel chips and nutrient
hybrid: GUK 14-7; Pennisetum sp.: IS 76-166; solution (1mM CaCl2, 4 mM MgSO4, 2 mM
Erianthus arundinaceus: IJ 76-503; E. bengalensis: H3PO4, 2 mM KCl, 100 µM EDTA-FeNa, 10 µM
IND 82-281; Saccharum×Bamboo inter-generic H3BO3, 2 µM MnCl2, 5 µM ZnSO4, 2 µM CuSO4
hybrid: GUK 14-530] raised in polybags of 3 feet and 0.2 µM MoO4). The sole source of nitrogen
height and 2 feet diameter containing 5 kg of potting Ca(NO3)2 was added to the solution to create 12
mixture (red soil:farm yard manure:sand::2:2:1). varying levels of nitrogen (0, 5, 10, 25, 50, 75,
Chlorophyll estimation was carried out in the 100, 250, 500, 750, 1000, 2000 µM). The solution
4-month old crop. Experiment II was screening for pH was adjusted between 6.8 to 7.0, and added
water use efficiency with 18 hybrids (Co 10026, to the pots every alternate day. The chlorophyll
Co 15018, Co 16018, Co 85019, Co 15015, Co measurement was recorded in 4-month old crop
12009, Co 95020, Co 13014, CoM 0265, Co when the deficiency symptoms were visible.
86032, Co 09004, Co 0212, Co 15021, Co 15007,
Total chlorophyll estimation
Co 8021, Co 11015, Co 14025, Co 14002) grown
in field condition with all recommended package Total chlorophyll content was recorded in
of practices, except irrigation which was limited sugarcane crop in the physiologically active
by 50% of the volume required based on pan leaf (+3) i.e., third fully opened leaf with visible
evaporation rates. Chlorophyll in the leaves was dewlap. An area of approximately 60 cm2
recorded in 8-month old crop. Experiment III was marked in the mid portion of the leaf for
included 12 commercial hybrids suitable for measurement of chlorophyll by non-destructive
cultivation in the tropical and sub-tropical zones of as well as destructive sampling. Non-destructive
India viz., tropical: Co 86032, Co 0212, Co 14012, measurement was recorded from 0900 to 1200
Co 06022, Co 11015, Co 13006 and sub-tropical: h using portable hand-held chlorophyll meter
Co 0238, Co 15027, Co 15023, Co 98014, BO 91, (MC-100, Apogee Instruments, Utah, USA)
CoLk 8102. Experiment IV included 10 sugarcane independently for the two units (i) SPAD and
hybrids in the drought screening programme of (ii) CCI. Average of five readings recorded in the
advanced varietal trial (AVT: Co 13002, Co 13003, 60 cm2 leaf area excluding midrib was taken for
Co 13004, Co 13006, Co 13008, Co 13009, Co individual leaf sample. For destructive sampling,
13013, Co 13014, Co 13018, Co 13020) and two the entire 60 cm2 leaf area excluding midrib was cut
standard checks Co 86032 and Co 99004. Drought into uniform squares (approximately 0.25 mm2),
was imposed from 60 to 150 days after planting from which 100 mg representative tissue was used
by withholding the irrigation, to select clones to determine (iii) total chlorophyll extracted with
that could sustain growth and yield under severe 80% acetone according to Arnon (1949) (C_ACE)
drought. Plants in experiments III and IV were and (iv) total chlorophyll extracted with dimethyl
grown in the field following the recommended sulphoxide by the method given by Hiscox and
package of practices and chlorophyll estimation Israelstam (1979) (C_DMSO). Three technical
was done in the 6-month old crop. In experiment replicates were taken for the destructive method
Journal of Sugarcane Research https://doi.org/10.37580/JSR.2019.2.9.97-118 153
https://doi.org/10.37580/JSR.2019.2.9.150-163

Table 1. Chlorophyll content in sugarcane germplasm

Non-destructive measurement Destructive measurement

Chlorophyll content Chlorophyll content
(mg g-1 FW) in acetone (mg g-1 FW) in DMSO
Genotype SPAD CCI (C_ACE) (C_DMSO)
Sarawak Unknown 38.43 19.12 2.288 2.032
Cebu Light Purple 40.70 33.08 2.234 1.978
Bontha 36.90 18.52 1.906 1.893
Hawaii Original 24 36.67 19.24 1.955 1.880
US 59-1-1 32.37 14.34 1.705 1.314
Katha Coimbatore 26.67 9.64 1.170 1.164
White Pindaria 41.50 29.50 2.271 2.089
Reha 43.17 26.47 2.156 2.084
NG 77-237 31.30 16.50 1.964 1.421
GUK 14-7 25.93 11.07 1.227 1.087
IS 76-166 34.10 18.30 1.811 1.722
IJ 76-503 28.77 12.60 1.635 1.399
IND 82-281 37.90 20.60 1.962 1.551
GUK 14-530 46.43 30.65 2.388 2.382
Mean 35.77 19.97 1.905 1.714

CV (%) 9.13 9.55 9.16 9.22

MSE 10.657 3.642 0.030 0.025
LSD 5.460 3.192 0.292 0.264
F value 10.877 44.958 13.759 18.686
p value 9.8E-08*** 3.0E-15*** 7.1E-09*** 2.0E-10***

Data for non-destructive measurement is the average of five meter readings, and three technical replicates for
destructive measurement. CV, MSE and LSD denote coefficient of variation, mean square error and least significant
difference, respectively. F and p values from analysis of variance indicate significant (α = 0.001) difference among
genotypes.

of chlorophyll estimation, which was expressed as by recording the same four traits in experiment
mg g-1 fresh weight (FW). V. The total chlorophyll values predicted from
Statistical data analysis SPAD and CCI using the calibration equation was
compared against the observed values to validate
Initial equations for calibration model was the developed model. Prediction or forecast error
developed by recording the above four traits viz., was calculated according to Kumar and Bhar
SPAD, CCI, C_ACE and C_DMSO in sugarcane (2005) with slight modification as below.
hybrids in experiments I to IV. The statistical models
developed from these experiments were validated
154 https://doi.org/10.37580/JSR.2019.2.9.97-118 Journal of Sugarcane Research
https://doi.org/10.37580/JSR.2019.2.9.150-163

Data analysis for variance, correlation, regression, installed with ‘agricolae’ package. Graphs were
model development and validation was carried plotted in Microsoft Excel 2016.
out using the statistical software R (version 3.6.0)

Table 2. Chlorophyll content in commercial sugarcane hybrids grown with irrigation limited by
50% volume based on pan evaporation rates

Non-destructive
Destructive measurement
measurement
Chlorophyll content (mg g-1 Chlorophyll content (mg g-1
Clone SPAD CCI
FW) in acetone (C_ACE) FW) in DMSO (C_DMSO)
Co 10026 25.56 10.02 1.098 1.299
Co 15018 26.76 10.36 1.424 1.054
Co 16018 26.08 9.24 1.112 1.101
Co 85019 31.54 17.22 1.652 1.342
Co 15015 34.86 18.58 1.460 1.484
Co 12009 33.74 15.50 1.688 1.846
Co 95020 32.24 14.28 1.638 1.417
Co 13014 35.68 14.86 1.522 1.784
CoM 0265 25.86 10.10 1.420 0.893
Co 86032 31.30 16.04 2.112 1.858
Co 09004 39.22 22.32 2.688 2.227
Co 0212 28.10 12.06 1.424 1.400
Co 15021 33.38 15.82 1.888 1.796
Co 15007 41.18 27.38 2.178 1.898
Co 8021 40.06 27.22 2.612 2.438
Co 11015 43.02 27.64 2.123 2.085
Co 14025 28.08 11.50 1.489 1.307
Co 14002 43.44 27.88 2.094 2.289
Mean 33.34 17.11 1.757 1.640

CV (%) 9.14 9.62 9.29 9.31

MSE 9.284 2.708 0.027 0.023
LSD 5.045 2.725 0.270 0.253
F value 11.807 48.634 23.850 26.026
p value 5.1E-10*** 2.2E-16*** 1.2E-14*** 3.0E-15***
Data for non-destructive measurement is the average of five meter readings, and three technical replicates
for destructive measurement. CV, MSE and LSD denote coefficient of variation, mean square error and
least significant difference, respectively. F and p values from analysis of variance indicate significant (α =
0.001) difference among clones.
Journal of Sugarcane Research https://doi.org/10.37580/JSR.2019.2.9.97-118 155
https://doi.org/10.37580/JSR.2019.2.9.150-163

Table 3. Chlorophyll content in commercial sugarcane hybrids adapted to tropical and

sub-tropical zones of India

Non-destructive
Destructive measurement
measurement
Chlorophyll content Chlorophyll content
Clone SPAD CCI (mg g-1 FW) in acetone (mg g-1 FW) in DMSO
(C_ACE) (C_DMSO)
Tropical clones
Co 86032 29.24 13.00 1.527 1.412
Co 0212 26.07 10.45 1.388 1.237
Co 14012 28.73 12.72 1.613 1.337
Co 06022 31.26 16.64 1.971 1.569
Co 11015 24.95 9.32 1.400 1.108
Co 13006 32.06 15.39 1.823 1.694
Sub-tropical clones
Co 0238 28.48 11.14 1.594 1.406
Co 15027 29.81 13.39 1.895 1.586
Co 15023 31.25 14.83 2.241 1.671
Co 98014 28.25 11.61 1.970 1.391
BO 91 25.50 9.91 1.647 1.082
CoLk 8102 31.72 14.37 1.720 1.658
Mean 28.94 12.73 1.732 1.429

CV (%) 9.03 9.13 9.09 9.09

MSE 6.829 1.352 0.025 0.017
LSD 4.404 1.960 0.265 0.219
F value 2.625 11.814 7.908 8.026
p value 2.3E-02*** 3.5E-07*** 1.3E-05*** 1.2E-05***
Data for non-destructive measurement is the average of five meter readings, and three technical replicates
for destructive measurement. CV, MSE and LSD denote coefficient of variation, mean square error and
least significant difference, respectively. F and p values from analysis of variance indicate significant (α =
0.001) difference among clones.

Results and Discussion 530 which is a hybrid between Saccharum and

Variation in total chlorophyll content recorded in bamboo recorded the highest values for SPAD
sugarcane germplasm (46.43), CCI (30.65), C_ACE (2.388 mg g-1 FW)
and C_DMSO (2.382 mg g-1 FW). S. robustum
Significant variation was observed in the total inter-specific hybrid GUK 14-7 recorded the least
chlorophyll content of sugarcane germplasm SPAD value (25.93) and C_DMSO (1.087 mg g-1
(Table 1). Among the genotypes tested, GUK 14- FW), while lowest CCI (9.64) and C_ACE (1.170
156 https://doi.org/10.37580/JSR.2019.2.9.97-118 Journal of Sugarcane Research
https://doi.org/10.37580/JSR.2019.2.9.150-163

mg g-1 FW) was recorded in S. barberi clone Katha actual chlorophyll content in sugarcane leaf, and
Coimbatore. thereby assessing the overall plant health under
various biotic and abiotic stress conditions has
Variation in total chlorophyll content of
been commonly practiced in the recent years
commercial sugarcane hybrids
(Silva et al. 2007; Radhamani and Kannan 2013;
Among the sugarcane hybrids grown with Vasantha et al. 2017; Kohila and Gomathi 2018).
irrigation limited by 50% volume based on pan SPAD was used to assess leaf senescence in
evaporation rates, highest indices of SPAD (43.44) sugarcane cultivar RB867515, which had potential
and CCI (27.88) was recorded in Co 14002 (Table applications for biofuel production (Martins
2). C_ACE was highest in Co 09004 (2.688 mg g-1 et al. 2016). Correlation between the different
FW), while C_DMSO was maximum (2.438 mg estimates of total chlorophyll content in sugarcane
g-1 FW) in Co 8021. Co 10026 showed least SPAD viz., SPAD, CCI, C_ACE and C_DMSO was
(25.56) and C_ACE (1.098 mg g-1 FW), while positive and highly significant, ranging from
CCI (9.24) and C_DMSO (1.101 mg g-1 FW) was 0.793 to 0.953 (Table 5). Jangpromma et al.
lowest in Co 16018. Results from experiment III (2010) reported a significant correlation (r =
revealed significant variation in total chlorophyll 0.78) between SPAD index and chlorophyll
content among hybrids adapted to tropical and sub- content extracted with dimethyl formamide,
tropical zones (Table 3). Highest SPAD (32.06) although the number of sugarcane hybrids were
and C_DMSO (1.694 mg g-1 FW) was observed limited to 10. The correlation coefficient seemed
in Co 13006, while highest CCI and C_ACE to increase during drought imposition (r = 0.90)
was recorded in Co 06022 (16.64) and Co 15023 and recovery (r = 0.98), although number of data
(2.241 mg g-1 FW), respectively. Lowest values for points was low to establish a strong relationship.
SPAD (25.50), CCI (9.91) and C_DMSO (1.082 Significant correlation (r = 0.90) between SPAD
mg g-1 FW) was observed in BO 91, while Co meter reading and total chlorophyll content
0212 recorded least C_ACE (1.388 mg g-1 FW). was observed among 24 commercial sugarcane
With regard to sugarcane hybrids in the advanced hybrids (Radhamani and Kannan 2013). Silva
varietal trial for drought screening, SPAD and CCI et al. (2013) reported significant correlation (r
values varied from 24.18 to 30.24 and 10.31 to = 0.68) between SPAD and total chlorophyll
15.01, respectively (Table 4). Similarly, C_ACE content in drought susceptible sugarcane cultivar,
ranged between 1.306 mg g-1 FW (Co 13006) to but this relationship did not hold true for tolerant
2.053 mg g-1 FW (Co 13002), while C_DMSO cultivar. Iron being an important constituent of the
varied from 0.939 mg g-1 FW in Co 13013 to chlorophyll molecule, metabolically active iron
1.671 mg g-1 FW in Co 13020. Wide variability content in sugarcane leaves was also positively
has been reported in total chlorophyll content of correlated to SPAD meter reading as well as to
commercial sugarcane hybrids measured in terms total chlorophyll content (Radhamani and Kannan
of both SPAD and extracted chlorophyll values 2013), stressing upon the use of non-destructive
(Vasantha and Rajalakshmi 2009; Radhamani and methods for chlorophyll estimation as a time and
Kannan 2013; Vasantha et al. 2017; Kohila and resource saving exercise. In the present study, the
Gomathi 2018). experiments were chosen carefully so as to include
Model calibration a wide variability of sugarcane germplasm as well
as commercial hybrids, grown with recommended
Use of chlorophyll meter index as a proxy for
package of practices (Exp I and III) subjected
Journal of Sugarcane Research https://doi.org/10.37580/JSR.2019.2.9.97-118 157
https://doi.org/10.37580/JSR.2019.2.9.150-163

Table 4. Chlorophyll content in commercial sugarcane hybrids in the advanced varietal trial for
drought screening
Non-destructive measurement Destructive measurement
Chlorophyll content Chlorophyll content
Clone SPAD CCI (mg g FW) in acetone (mg g-1 FW) in DMSO
-1

(C_ACE) (C_DMSO)
Co 13002 30.08 13.73 2.053 1.636
Co 13003 29.12 12.53 1.432 1.257
Co 13004 26.11 10.86 1.351 1.171
Co 13006 28.47 10.87 1.306 1.238
Co 13008 29.30 13.59 1.715 1.532
Co 13009 26.39 10.35 1.401 1.113
Co 13013 24.81 10.36 1.139 0.939
Co 13014 26.86 10.87 1.335 1.136
Co 13018 30.24 15.01 1.918 1.524
Co 13020 33.14 15.47 1.929 1.671
Co 86032 24.18 10.31 1.371 1.219
Co 99004 30.18 12.68 1.710 1.333
Mean 28.24 12.22 1.555 1.314
CV (%) 9.04 9.10 9.15 9.12
MSE 6.511 1.236 0.020 0.014
LSD 4.300 1.874 0.240 0.202
F value 3.151 8.598 13.034 10.895
p value 9.1E-03*** 6.4E-06*** 1.4E-07*** 7.5E-07***
Data for non-destructive measurement is the average of five meter readings, and three technical replicates
for destructive measurement. CV, MSE and LSD denote coefficient of variation, mean square error and
least significant difference, respectively. F and p values from analysis of variance indicate significant
(α = 0.001) difference among clones.

to limited (Exp II) and severe water stress (Exp of determination (R2) among the traits observed
IV). Regression models for the best fit curve was statistically significant as presented in Figures
between meter reading and extracted chlorophyll 1 and 2. Regression between the two indices SPAD
values of the tested sugarcane clones were non- and CCI (Fig. 1A and 2A) also fitted a non-linear,
linear, polynomial equations of the second order, polynomial curve in agreement to Richardson
in concurrence with reports in wheat, rice and et al. (2002). Few reports indicate a linear
vsoybean (Monje and Bugbee 1992), muskmelon relationship between SPAD and total chlorophyll
(Azia and Stewart 2001), maize (Gitelson and content in Augustine grass (Rodriguez and Miller
Merzlyak 2004), wheat and potato (Uddling et al. 2000), sorghum and pigeon pea (Yamamoto et
2007). The data from experiments I to IV were al. 2002), but in the present investigation non-
pooled to derive the regression models. Coefficient linear polynomial equations emerged as the best
158 https://doi.org/10.37580/JSR.2019.2.9.97-118 Journal of Sugarcane Research
https://doi.org/10.37580/JSR.2019.2.9.150-163

Fig. 1. Polynomial regression models (n = 56) of(A) SPAD against CCI, (B) SPAD against chlorophyll content
in acetone and (C) SPAD against chlorophyll content in DMSO observed in germplasm and commercial hybrids
of sugarcane, (D) equations to predict CCI, C_ACE and C_DMSO from SPAD values. Calculated and adjusted
coefficient of determination (R2) and p values indicate statistical significance (α = 0.001) of the predicted models

fit. Similarly, Castelli et al. (1996) demonstrated fitting methods, polynomial exponential function
that among linear, quadratic and exponential curve best described the relationship between SPAD
and total chlorophyll content recorded in wheat,
Table 5. Correlation between estimates of total
maize, soybean and tobacco (Nicotiana tabacum).
chlorophyll content in sugarcane by non-
Hence, the model calibrated in sugarcane is in line
destructive and destructive sampling
with other major crops.
C_
SPAD CCI C_ACE Model validation
DMSO
SPAD 0.953 0.800 0.915 To validate the accuracy of model prediction,
CCI 1.04E- 0.793 0.868
chlorophyll measurement by both destructive and
29***
C_ACE 1.36E- 3.20E- 0.869 non-destructive sampling was done in sugarcane
13*** 13*** variety Co 86032 raised in semi-hydroponic
C_ 5.60E- 3.91E- 3.85E- culture conditions subjected to increasing nitrogen
DMSO 23*** 18*** 18*** levels (Table 6). This experiment was intended as
Data in upper panel denotes Pearson’s correlation to obtain wide range of total chlorophyll values
coefficient (r) between the traits, lower panel as the leaves showed slight to severe yellowing
denotes p value indicating statistical significance symptoms due to varying nitrogen levels. Nitrogen
(α = 0.001) of r (n=56). fertilisation to sugarcane had a greater effect to
Journal of Sugarcane Research https://doi.org/10.37580/JSR.2019.2.9.97-118 159
https://doi.org/10.37580/JSR.2019.2.9.150-163

Table 6. Chlorophyll content in sugarcane variety Co 86032

subjected to increasing nitrogen levels

Non-destructive measurement Destructive measurement

Nitrogen Chlorophyll content (mg g-1 Chlorophyll content (mg g-1

SPAD CCI
level FW) in acetone (C_ACE) FW) in DMSO (C_DMSO)

0 µM 14.23 1.58 0.439 0.172

5 µM 19.11 4.37 0.796 0.403
10 µM 21.29 6.13 0.973 0.637
25 µM 21.96 6.61 1.016 0.706
50 µM 22.23 6.81 1.034 0.733
75 µM 29.05 12.25 1.457 1.353
100 µM 32.00 14.65 1.618 1.578
250 µM 33.30 15.89 1.697 1.668
500 µM 37.29 20.67 1.978 1.916
750 µM 37.88 21.40 2.019 1.948
1000 µM 38.59 22.79 2.094 1.986
2000 µM 41.99 27.63 2.344 2.145
Mean 29.08 13.40 1.455 1.271
CV (%) 9.39 10.50 9.67 10.17
MSE 7.460 1.981 0.020 0.017
LSD 4.603 2.372 0.237 0.218
F value 33.146 107.410 54.262 87.823
p value 8.1E-12*** 2.2E-16*** 3.2E-14*** 2.2E-16***

Data for non-destructive measurement is the average of five meter readings, and three technical replicates
for destructive measurement. CV, MSE and LSD denote coefficient of variation, mean square error and
least significant difference, respectively. F and p values from analysis of variance indicate significant (α =
0.001) difference among nitrogen levels.
SPAD index as well as total chlorophyll content µM nitrogen showed highest values of C_ACE
(Silva et al. 2017), hence gradually decreasing (2.344 0.439 mg g-1 FW) and C_DMSO (2.145
the dose of nitrogen application would yield 0.439 mg g-1 FW). Highly significant linear
in visible chlorosis symptoms. In the present regression was found between observed and
study, with increasing nitrogen level SPAD and predicted values of all estimates of total chlorophyll
CCI values showed an increasing trend from content with almost negligible prediction error
14.23 to 41.99 and 1.58 to 27.63, respectively. (Fig. 3). Thus, the equations developed could be
Similarly, lowest C_ACE (0.439 mg g-1 FW) employed for rapid estimation of total chlorophyll
and C_DMSO (0.172 0.439 mg g-1 FW) was content in sugarcane germplasm and commercial
recorded at 0 µM nitrogen level, whereas 2000 hybrids.
160 https://doi.org/10.37580/JSR.2019.2.9.97-118 Journal of Sugarcane Research
https://doi.org/10.37580/JSR.2019.2.9.150-163

Fig. 2. Polynomial regression models (n = 56) of (A) CCI against SPAD, (B) CCI against chlorophyll content in
acetone and (C) CCI against chlorophyll content in DMSO observed in germplasm and commercial hybrids of
sugarcane, (D) equations to predict SPAD, C_ACE and C_DMSO from CCI values.Calculated and adjusted coefficient
of determination (R2) and p values indicate statistical significance (α = 0.001) of the predicted models

Conclusion eqs. IV to VI for CCI recorded using portable

meters) have been developed and validated across
Non-destructive methods to assess plant health
diverse germplasm, commercial hybrids, crop age
is in vogue in order to save resources, time and
and/or stress condition.
labour. This includes developing alternative, rapid
techniques to substitute for chlorophyll extraction Alongside these equations, important precautions
with solvents which require more time and effort. to follow during non-destructive sampling is also
With this intention, keeping in view the need to emphasised i.e., optimum time of day for sampling
(0900 to 1200 h), physiologically active leaf (+3
i.e., third fully opened leaf with visible dewlap),
optimum leaf area (60 cm2 laminar area in the
middle portion of leaf, excluding mid-rib) and
technical replicates (minimum five). Though total
leaf chlorophyll content is only one of the several
physiological traits important for crop selection,
interpret total chlorophyll in sugarcane not as models developed herein would definitely aid
relative indices, but as the actual content in the high-throughput phenotyping in sugarcane for
leaf, following equations (Eqs. i to iii for SPAD, accelerating crop improvement programmes.
Journal of Sugarcane Research https://doi.org/10.37580/JSR.2019.2.9.97-118 161
https://doi.org/10.37580/JSR.2019.2.9.150-163

Fig. 3. Validation ofobserved against predicted values (n = 12) of chlorophyll content in sugarcane; (A) Observed
CCI vs CCI predicted from SPAD, (B) observed C_ACE vs C_ACE predicted from SPAD, (C) observed C_DMSO vs
C_DMSO predicted from SPAD, (D) observed SPAD vs SPAD predicted from CCI, (E) observed C_ACE vs C_ACE
predicted from CCI and (F) observed C_DMSO vs C_DMSO predicted from CCI. Purple plus symbol indicates
prediction error plotted in secondary axis

Acknowledgments Azia F, Stewart KA (2001) Relationships between

extractable chlorophyll and SPAD values
We acknowledge the financial support, field and
in muskmelon leaves. Journal of Plant
laboratory facilities provided by Director, ICAR- Nutrition 24(6):961-966.
Sugarcane Breeding Institute for carrying out this
Barnes JD, Balaguer L, Manrique E, Elvira
research work. Authors also acknowledge the
S, Davison AW (1992) A reappraisal of
Early Career Research Award [ECR/2017/000738]
the use of DMSO for the extraction and
of Science and Engineering Research Board, determination of chlorophylls a and b in
Department of Science and Technology, New lichens and higher plants. Environmental
Delhi for funding. We thank our technician Ms. and Experimental Botany 32(2):85-100.
Aswathy and students Ms. Karpagam, Ms. Swathy, Castelli F, Contillo R, Miceli F (1996) Non-
Ms. Keerthana, Mr. Karthik and Ms. Subashree destructive determination of leaf
for chlorophyll extraction. chlorophyll content in four crop species.
References Journal of Agronomy and Crop Science
177(4):275-283.
Arnon DI (1949) Copper enzymes in isolated
Gitelson AA, Merzlyak MN (2004) Non-
chloroplasts: Polyphenol oxidase in Beta
destructive assessment of chlorophyll,
vulgaris. Plant Physiology 24(1):1-15.
162 https://doi.org/10.37580/JSR.2019.2.9.97-118 Journal of Sugarcane Research
https://doi.org/10.37580/JSR.2019.2.9.150-163

carotenoid and anthocyanin content Buckeridge MS, Kobayashi AK, Quirino

in higher plant leaves: principles and BF, Molinari HBC (2016) Characterization
algorithms. Papers in Natural Resources of sugarcane (Saccharum spp.) leaf
263. senescence: implications for biofuel
production. Biotechnology for Biofuels
Hawkins S, Gardiner ES, Comer GS (2009)
9:153.
Modelling the relationship between
extractable chlorophyll and SPAD-502 Monje OA, Bugbee B (1992) Inherent limitations
readings for endangered plant species of non-destructive chlorophyll meters:
research. Journal for Nature Conservation a comparison of two types of meters.
17:123-127. Hortscience 27(1):69-71.

Hiscox JD, Israelstam GF (1979) A method for the Parry C, Blonquist JM, Bugbee B (2014) In
extraction of chlorophyll from leaf tissue situ measurement of leaf chlorophyll
without maceration. Canadian Journal of concentration: analysis of the optical/
Botany 57:1332-1334. absolute relationship. Plant, Cell and
Environment 37:2508-2520.
Holm-Hansen O, Riemann B (1978) Chlorophyll
a determination: improvements in Porra RJ, Thompson WA, Kriedemann PE (1989)
methodology. Oikos 30:438-447. Determination of accurate extinction
Jangpromma N, Songsri P, Thammasirirak coefficients and simultaneous equations
S, Jaisil P (2010) Rapid assessment of for assaying chlorophylls a and b extracted
chlorophyll content in sugarcane using a with four different solvents: verification of
SPAD chlorophyll meter across different the concentration of chlorophyll standards
water stress condition. Asian Journal of by atomic absorption spectroscopy.
Plant Science 9(6):368-374. Biochimica Biophysica Acta 975:384-394.
Kapotis G, Zervoudakis G, Veltsistas T, Salahas Porra RJ (2002) The chequered history of the
G (2003) Comparison of chlorophyll meter development and use of simultaneous
readings with leaf chlorophyll concentration equations for the accurate determination
in Amaranthus vlitus. Russian Journal of of chlorophylls a and b. Photosynthesis
Plant Physiology 50(3):395-397. Research 73:149-156.
Kohila S, Gomathi R (2018) Adaptive Radhamani R, Kannan R (2016) Nondestructive
physiological and biochemical response of and rapid estimation of leaf chlorophyll
sugarcane genotypes to high-temperature content of sugarcane using a SPAD meter.
stress. Indian Journal of Plant Physiology International Journal of Science and
23(3):245-260. Research 5(4):2392-2397.
Kumar A, Bhar L (2005) Forecasting model for
Richardson AD, Duigan SP, Berlyn GP (2002)
yield of Indian mustard (Brassica juncea)
An evaluation of noninvasive methods to
using weather parameter. Indian Journal of
estimate foliar chlorophyll content. New
Agricultural Sciences 75(10):688-690.
Phytologist 153:185-194.
Martins MTB, Souza WR, Cunha BADB, Basso
Ritchie RJ (2006) Consistent sets of
MF, Oliveira NG, Vinecky F, Martins PK,
Oliveira PA, Arenque-Musa BC, Souza AP, spectrophotometric chlorophyll equations
Journal of Sugarcane Research https://doi.org/10.37580/JSR.2019.2.9.97-118 163
https://doi.org/10.37580/JSR.2019.2.9.150-163

for acetone, methanol and ethanol solvents. affected by nitrogen applied via subsurface
Photosynthesis Research 89:27-41. drip irrigation. Journal of Environment and
Agricultural Sciences 11:15-28.
Rodriguez IR, Miller GL (2000) Using a
chlorophyll meter to determine the Uddling J, Gelang-Alfredsson J, Piikki K, Pleijel
chlorophyll concentration, nitrogen H (2007) Evaluating the relationship
concentration and visual quality of St. between leaf chlorophyll concentration
Augustinegrass. Hortscience 35(4) : 751- and SPAD-502 chlorophyll meter readings.
754. Photosynthesis Research 91(1):37-46.
Samsone I, Andersone U, Vikmane M, Levina Vasantha S, Rajalakshmi R (2009) Progressive
B, Pakarna G, Levinsh G (2007) changes in biochemical characters of
Nondestructive methods in plant biology: sugarcane genotypes under salinity
an accurate measurement of chlorophyll stress. Indian Journal of Plant Physiology
content by a chlorophyll meter. Acta 14(1):34-38.
Universitatis Latviensis 723:145-154.
Vasantha S, Gomathi R, Brindha R (2017) Growth
Schaper H, Chacko EK (1991) Relations between and nutrient composition of sugarcane
extractable chlorophyll and portable genotypes subjected to salinity and drought
chlorophyll meter readings in leaves of eight stresses. Communications in Soil Science
tropical and subtropical fruit-tree species. and Plant Analysis 48(9):989-998.
Journal of Plant Physiology 138(6):674-
677. Yamamoto A, Nakamura T, Adu-Gyamfi JJ, Saigusa
M (2002) Relationship between chlorophyll
Silva MA, Jifon JL, Da Silva JAG, Sharma V
content in leaves of sorghum and pigeonpea
(2007) Use of physiological parameters as
determined by extraction method and by
fast tools to screen for drought tolerance
chlorophyll meter (SPAD-502). Journal of
in sugarcane. Brazilian Journal of Plant
Plant Nutrition 25(10):2295-2301.
Physiology 19(3):193-201.
Yuan Z, Cao Q, Zhang K, Ata-Ul-Karim ST, Tian
Silva MA, Jifon JL, Santos CM, Jadoski CJ,
Y, Zhu Y, Cao W, Liu X (2016) Optimal leaf
Silva JAG (2013) Photosynthetic capacity
positions for SPAD meter measurement in
and water use efficiency in sugarcane
genotypes subjected to water deficit during rice. Frontiers in Plant Science 7:719.
early growth phase. Brazilian Archives of Zhu J, Tremblay N, Liang Y (2012) Comparing
Biology and Technology 56(5):735-748. SPAD and atLEAF values for chlorophyll
Silva MA, Rhein AFL, Barbosa AM (2017) assessment in crop species. Canadian
Physiology and productivity of sugarcane as Journal of Soil Science 92:645-648.