Influence of harvesting factors on sensory attributes and phenolic and aroma compounds composition of Cymbopogon citratus leaves infusions

    Influence of harvesting factors on sensory attributes and phenolic and aroma compounds composition of Cymbopogon citratus leaves infusions M. Coelho, C. Rocha, L.M. Cunha, L. Cardoso, L. Alves, R.C. Lima, M.J. Pereira, F.M. Campos, M. Pintado PII: DOI: Reference:

S0963-9969(16)30284-8 doi: 10.1016/j.foodres.2016.07.008 FRIN 6335

To appear in:

Food Research International

Received date: Revised date: Accepted date:

24 February 2016 13 July 2016 15 July 2016

Please cite this article as: Coelho, M., Rocha, C., Cunha, L.M., Cardoso, L., Alves, L., Lima, R.C., Pereira, M.J., Campos, F.M. & Pintado, M., Influence of harvesting factors on sensory attributes and phenolic and aroma compounds composition of Cymbopogon citratus leaves infusions, Food Research International (2016), doi: 10.1016/j.foodres.2016.07.008

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Influence of harvesting factors on sensory attributes and phenolic and aroma compounds composition of Cymbopogon citratus leaves infusions

T

Coelho, M.1*, Rocha, C.2,3*, Cunha, L.M.2,4, Cardoso, L.5, Alves, L.5, Lima, R.C.3, Pereira, M. J.1*, Campos, F.M.1, Pintado, M.1**

IP

1

NU

SC R

CBQF/Escola Superior de Biotecnologia da Universidade Católica Portuguesa, Rua Arq. Lobão Vital, 4200-072 Porto, Portugal 2 LAQV-REQUIMTE/DGAOT, Faculty of Sciences, University of Porto, Campus Agrário de Vairão, Vila do Conde, Portugal 3 SenseTest, Lda., Rua Zeferino Costa, 341, 4400 – 345 Vila Nova de Gaia, Portugal 4 GreenUP/CITAB-UP, Campus Agrário de Vairão, Vila do Conde, Portugal 5 Cantinho das Aromáticas Lda., Rua do Meiral, 508, 4400-501 Canidelo – Vila Nova de Gaia, Portugal

MA

*equal contribution **Corresponding author: [email protected]

Abstract

AC

CE P

TE

D

Herbal infusions are receiving increasing attention for the number of physiological benefits that can bring to human health. Cymbopogon citratus is one the most used plants in traditional medicine besides its characteristics and unique aroma appreciated by the consumers; however, little is known about the effects of harvesting on functional and sensory properties of this plant. The present work aimed to assess the lemongrass infusions, which were harvested following a factorial plan according to the type of cut (manual and mechanical) and part of the plant (tips and 2nd half leaves). Total phenolic content, antioxidant capacity, aroma compounds composition (terpenoids) and sensory profiles and consumer perception of different samples were assessed. The Quantitative Descriptive Analysis (QDA) and Check-All-That-Apply (CATA) methods were used to describe the lemongrass infusion, complemented with the overall liking evaluation, using a 9-point scale. Results showed that there were significant differences between type of cut and part of the plant concerning phenolic compounds: samples obtained by mechanical cut showed in general higher content of individual phenolic compounds (in particular for chlorogenic acid) and tips showed in general the highest contents for both types of cut. Consumers’ panel didn’t find significant differences between samples. Generally, consumers refer a bitter taste in all infusions when the content of the phenolic compounds was higher, in particular for p-coumaric acid. Concerning the aroma compounds no significant differences were observed between type of cut and part of the plant, and citral was the terpenoid present in higher quantity. In sensory profile methods, it was found that QDA and CATA were both good methods to describe this infusion. Considering consumers evaluation of product chain, the infusion prepared with plant tips of lemongrass was selected as the premium herbal tea.

1. Introduction Nowadays, there is an increased interest by policy makers, researchers, food industries and consumers on foods and food ingredients that provide benefits beyond their traditional nutritional value (Hoekstra et al., 2008). In this regard, the use of

ACCEPTED MANUSCRIPT

AC

CE P

TE

D

MA

NU

SC R

IP

T

medicinal plants is increasing, since they are considered “natural” and therefore commonly perceived as “safe” (Goey, Mooiman, Beijnen, Schellens, & Meijerman, 2013; Tascilar, de Jong, Verweij, & Mathijssen, 2006). Furthermore, many beneficial properties of plants have been associated to the presence of bioactive compounds, like vitamins, minerals, polyphenols and plant sterols (Aruoma, 2003; Hertog, Hollman, & Putte, 1993; Hollman & Katan, 1999). Those components are important for plant development and play an important role in the defence mechanisms. The regular consumption of plant infusions may be valuable to human due to confirmed healthrelated benefits namely, antioxidant and anti-inflammatory capacity, cholesterol regulation, improvement of vascular disorders and anti-mutagenic properties (Duarte Silva et al., 2000; Gião et al., 2007; Pennington, 2002). For these reasons, it is crucial to characterize bioactive compounds present in the plant infusions to better understand their health benefits. Despite their widespread use, the aroma compounds composition of most plant infusions is not well known, since few scientific studies have been published on the subject. However, the connoisseur consumer of herbal infusions expects to find their characteristic and unique aroma that stimulates various sensations leading to pleasure and thus promoting its consumption. Cymbopogon citratus, popularly known as citronella grass or lemongrass, belong to the family of Poaceae and is probably originally from India or Sri Lanka, being cultivated in regions of the world with tropical and subtropical humid environments. It has been used by humans for thousands of years due to its characteristic and valuable flavour and for the treatment or prevention of health disorders and diseases (Barbosa et al., 2008; Carteri Coradi, de Castro Melo, & Pereira da Rocha, 2014). Its essential oil is characterized by 65-85% citral (constituted by the isomers of neral and geranial), βmyrcene, linalool, geraniol, 6-methylhept-5-en-2-one, citronellol and isovaleric acid esters (Cunha, Ribeiro, & Roque, 2007; Nur Ain, Zaibunnisa, Halimahton Zahrah, & Norashikin, 2013). A recent review work (Lasekan & Lasekan, 2012) cites only two studies on the characterization of aroma lemongrass infusions, although there are several publications on the volatile composition of its essential oil (Kasali, Oyedeji, & Ashilokun, 2001; Saleem, Afza, Anwar, Hai, & Ali, 2003; Schaneberg & Khan, 2002). One study has reported the effect of the cutting height (Ganjewala & Luthra, 2010) and another the effect of water stress (Singh‐Sangwan, Abad Farooqi, & Singh Sangwan, 1994) on the essential oil composition of this plant, but no study has been addressed on the impact of production and processing factors in the sensory profile and aroma composition of herbal infusions prepared from this plant. So, the present work was aimed at evaluating the impact of the type of cut (manual or mechanical) and the plant part (tips and 2nd half leaves) on the phenolics, aroma compounds composition (terpenoids) and sensory profiles, and consumer acceptance of lemongrass infusions. With this evaluation was intended to obtain an herbal infusion which under the selected conditions studied may be classified as premium infusion.

2. Materials and Methods 2.1 Samples Leaves of lemongrass (Cymbopogon citratus) were collected at Cantinho das Aromáticas (Vila Nova de Gaia, Portugal) following a 2k factorial plan according to the type of cut (manual and mechanical) and part of the plant (tips and 2nd half leaves, referred as 2nd half). All plants came from the same plantation lot and were harvested in June and July 2014, totalizing four different batches performed in triplicate. The company's production method is based on organic farming, which is certified by the organization Ecocert Portugal, since 2005. The samples were dried at a temperature of

ACCEPTED MANUSCRIPT 37 °C for 3 days on a drying chamber (AromaDry AR-4000®, Portugal) and stored in hermetic containers before being used for infusion preparation.

2.2 Herbal infusions

SC R

IP

T

Each batch of Cymbopogon citratus leaves was obtained according to the experimental design and used to prepare the infusions. Each infusion was prepared using 4.5 g of whole dried part of plant (tips and 2nd half) infused in 1.5 L of hot natural mineral water. Water was heated up to 99 ° C, very close to a full boil, and immediately poured over the leaves. The infusion was let to simmer for 8.5 min, and leaves removed. (Rocha et al., 2015) For sensory evaluation, after simmering, each infusion was cooled to 65 °C and then placed on thermally-insulated flasks until the time of serving. For all other analysis the samples were cooled to room temperature and stored at – 80 °C until further analysis.

NU

2.3 Antioxidant Capacity and Total Phenolic Content

CE P

TE

D

MA

Total antioxidant capacity was determined using the ABTS method as described by Elsner et al. (2013). The sample, diluted when needed, was added to a coloured solution of 2,2′-azinobis-(3-ethylbenzothiazoline-6-sulfonic acid radical cation) (ABTS•+). The initial optical density (OD) of the ABTS•+ solution, measured at 734 nm using a UVmini 1240 UV-Vis spectrophotometer (Shimadzu, Japan), was adjusted to an absorbance of 700±0.020. After 6 min, the final absorbance was measured and the results were calculated as ascorbic acid equivalents. In order to evaluate the phenolic component of the extracts the Folin – Ciocalteu method was adapted based on the method described by (Singleton & Rossi, 1965)). To 50 µL of sample, diluted when needed, 50 µL of Folin – Ciocalteu reagent, 1 mL of sodium carbonate at 75 g/L (Sigma) and 1.4 mL of distilled water were added. After 1 h the absorbance was read at 750 nm using an UVmini 1240 UV-Vis spectrophotometer (Shimadzu, Japan) and the phenolic content was calculated as gallic acid equivalents.

2.4 HPLC analysis

AC

Analysis was carried out according to the method proposed by (Cabrera, Giménez, & López, 2003) with slight modifications. Qualitative and quantitative profiles of polyphenols were conducted on a Waters Liquid Chromatograph (Waters Series 600. Mildford MA. USA). A C18 guard column (Symmetry® C18) and an Alltech Adsorbosil C18 reversed-phase packing column (250 x 4.6 mm i.d. 5 μm particle size and 125 Å pore size) were used for separation throughout this study. The PDA acquisition wavelength was set in the range of 200–400 nm, analog output channel A at wavelength 320 nm and analog output channel B at 350 nm both with 2 nm bandwidth. The solvent gradient varied from mixture solvent A (water–acetic acid, 97:3 v/v) to B (methanol), with a flow rate of 1 mL min−1. The mobile phase composition started at 100% solvent A for 1 min, followed by a linear increase of solvent B to 63% in 38 min, and then back to the initial conditions in 2 min for the next run. All the prepared solutions were filtered through 0.45 μm membranes (Fisher Scientific) and the mobile phase was degassed before injection into the HPLC. Calibration curves were obtained at detection wavelengths of 320 nm and 350 nm. Standards solutions over the concentration range from 0.10 to 100.00 mg/L were prepared for the identification and quantification of the following compounds: gallic, chlorogenic, caffeic, ferulic, p-coumaric, and ellagic acids and epicatechin (Sigma-Aldrich, Steinheim, Germany) expressed as micrograms per mL of herbal tea. All calibration curves were linear over the concentration ranges tested, with correlations coefficients higher than 0.999.

2.5 Extraction and GC analysis of volatile (terpenoids) components

ACCEPTED MANUSCRIPT

SC R

IP

T

An internal standard solution containing 45.0 µg.L-1 of 3-octanol in 20% (m/v) ethanol was added (50 µL) to the samples (50.0 mL). The mixture was consecutively extracted with 4.0, 2.0 and 2.0 mL of a diethyl ether/hexane mixture (1:1 v/v) and stirred for 5 min. The organic phases were collected, combined and concentrated under a stream of N2 to approx. one-third of the original volume. Volatile compounds were analysed by GC-FID according to the method described by Bertrand (1981). 1 µL of extract was injected in split/splitless mode (0.5 min), at 200 °C in a FFAP capillary column (50 m × 0.22 mm i.d.; 0.2 µm film thickness) using hydrogen as carrier gas. The oven temperature programme used was: 60 °C (5 min) – 3ºC/min – 200ºC (30 min). The FID temperature was set to 250 °C.

2.6. Sensory Analysis 2.6.1. Sensory Panels

CE P

2.6.2. Overall liking

TE

D

MA

NU

An untrained panel of 75 naïve panellists were recruited from Sense Test’s consumer database. The company ensure the protection and confidentiality of data through the authorization 2063/2009 of the National Data Protection Commission and an accomplished internal code of conduct. Sensory evaluation was carried out in a special room equipped in accordance with ISO 8589:2007 Sensory analysis—General guidance for the design of test rooms. Samples at 65 °C were presented to each taster on white porcelain cups identified by a three-digit random number, in individual booths, under normal white lighting. Panellists were provided with a porcelain spittoon, a glass of bottled natural water and unsalted crackers. They were also allowed to add castor sugar cubes (2.5 g), according to their regular usage, and instructed to use the same amount for all samples under evaluation. All panellists were asked to chew a piece of cracker and to rinse the mouth with water before testing each sample.

AC

For each infusion, overall liking was evaluated using a 9-point scale, going from 1“dislike extremely” to 9- “like extremely” (Peryam & Pilgrim, 1957). To compensate for eventual carry-over effects, each panellist received the set of four samples following a monadic and sequential order of presentation, with a balances presentation order (Macfie, Bratchell, Greenhoff, & Vallis, 1989). For each sample, overall liking was immediately followed by the evaluation of the sensory profile, through Check-All-ThatApply (CATA) methodology, as recently suggested (Ares & Jaeger, 2013; King, Meiselman, & Carr, 2013) to reduce bias.

2.6.3. CATA (Check-All-That-Apply) The untrained panellists were asked to perform a CATA question with a list of 78 sensory attributes divided into four dimensions: appearance (11), odour (37), texture (5) and taste (25) (Table 1). This list of terms was created combining previous research for herbal teas in general (Cardoso, 2013) with the descriptors generated by the trained panel (Table 2). This list was presented in two different orders to the consumers following a direct and an inverse alphabetic order within each dimension (King et al., 2013). The CATA question “Please select out of the following list of terms those that characterized the tasted sample” was answered in a “yes/no” response scale, indicating if they recognize the presence or absence of such attributes. This option for the response scale increases the focus of respondents on each attribute (Jaeger et al., 2014).

ACCEPTED MANUSCRIPT 2.6.4 Quantitative Descriptive Analysis (QDA)

MA

NU

SC R

IP

T

A panel of eight tasters was selected and trained to evaluate the infusions, according to ISO 8586:2012 Sensory analysis -- General guidelines for the selection, training and monitoring of selected assessors and expert sensory assessors, and Meilgaard et al. 2010 (Meilgaard, Carr, & Civille, 2010). The initial training, which started with 50 tasters, consisted of a set of recruitment tests, such as pairing, triangle, ordering and description tests, the Ishihara’s test for colour blindness and evaluation of personality traits, as described by Rocha and colleagues (2013), allowing the selection of the best eight assessors for evaluation of infusions. This initial training lasted up to four months. The eight selected tasters were gathered to discuss the list of attributes that best describe lemongrass infusions. After three discussion sessions, the panel members reached a consensus about the attributes list to describe the Cymbopogon citratus herbal infusion. The final list of attributes (Table 2) had 11 attributes distributed by the four sensory categories: appearance (1), odour (3), texture (1) and taste (6). After deciding the final list, all panel members trained the use of the scale for each of the attributes, defining their references for the anchors. At the final stage of the process, the samples were evaluated, in triplicate, on 10 cm unstructured scales anchored at the extremes.

2.7. Statistical analysis

AC

CE P

TE

D

Data analyses were performed using XL-STAT 2016® system software (Addinsoft, USA). All chemical data (antioxidants, phenolic and aroma compounds) were evaluated applying a two-way ANOVA with interaction (type of cut and part of plant). Additionally, for the set of phenolic compounds, when significant effect were identified, aggregated data was used to perform a PCA, identifying major relationships between phenolic compounds and samples. To evaluate the results of the overall liking tests, descriptive statistics (mean and standard deviation) were used, and a two-way (type of cut and part of plant) ANOVA with blocks (panellists) was applied at a 95 % confidence level. For CATA evaluations, the Cochran test was applied to identify which attributes were discriminating among samples (Parente, Manzoni, & Ares, 2011). Then, the frequency of use of each attribute was determined, calculating the number of panellists who have used each attribute to describe the samples. Over this frequency matrix, a correspondence analysis (CA) was applied. Such analysis provides a sensory map of the samples, allowing the perception of the similarities and differences between samples and their sensory characteristics (Ares, Barreiro, Deliza, Giménez, & Gàmbaro, 2010; Ares, Deliza, Barreiro, Giménez, & Gámbaro, 2010; Ares, Varela, Rado, & Giménez, 2011). QDA data were analysed using a three-way ANOVA with interaction (panellists, samples and sessions) in order to determine which attributes were discriminating between products (Albert, Varela, Salvador, Hough, & Fiszman, 2011; Varela & Ares, 2012). The Multiple Factor Analysis (MFA) was applied to assess the consensus among the different variables analysed: antioxidants, phenolic and aroma compounds, QDA and CATA. This statistical methodology allows the analysis of the consensus between different variables that differ in number and data nature and is an improved PCA, which analyses the individual differences instead of average differences (Louw et al., 2013; Nestrud & Lawless, 2008).

3. Results and discussion

ACCEPTED MANUSCRIPT 3.1. Antioxidant Capacity and Total Phenolic Content

AC

CE P

TE

D

MA

NU

SC R

IP

T

The total phenolic content and antioxidant capacity of different lemongrass infusions are summarized in Table 3. In terms of antioxidant activity, the ascorbic acid equivalent concentration of different lemongrass infusions ranged from 0.45±0.01 mg AAE g-1 dry leaf for the tip leaves by manual cut to 0.55±0.01 mg AAE g-1 dry leaf for 2nd half leaves by manual cut. The antioxidant activity for different lemongrass infusions was in the following order: 2nd half leaves by manual cut > 2nd half leaves by mechanical cut > tips leaves by mechanical cut > tips leaves by manual cut. Statistical analysis of these results demonstrated that only the plant area had significant effects on the antioxidant activity of lemongrass infusions. Concerning the phenolic content assessment, the gallic acid equivalent (GAE) concentration ranged from 23.07 ± 0.028 mg GAE g -1 dry leaf for the tips of leaves by manual cut to 16.09 ± 1.62 mg GAE g-1 dry leaf for 2nd half leaves by manual cut. These results are in accordance with those of Gião et al. (2007), which found a total phenolic content of 66 mg GAE L-1 and, which reported a total phenolic content of 15.3 mg GAE g-1 of lemongrass infusion. The amount of phenolic compounds found in the different lemongrass infusions was in the following order: plant tips with manual cut > plant tips with mechanical cut > 2nd half leaves with mechanical cut > 2nd half leaves with manual cut. The statistical analysis of the results demonstrated that both factors type of cut and plant part – have significant effects on the total phenolic content of lemongrass infusions. So, it’s possible to conclude that regarding the total phenolic content, the tips of leaves by manual cut yielded infusions richer in these compounds. These results suggest that the antioxidant properties of some polyphenols are not so notable, having distinct activities. In addition, it should be taken into consideration that there may be antagonistic or synergistic interactions between phenolic and other compounds, including carbohydrates, which represents a limitation of Folin-Ciocalteau method (Odabasoglu et al., 2005; Ranković, 2015). The obtained results regarding the phenolic content of lemongrass infusions was expected, since this type of secondary metabolites are involved in plant defense mechanisms and mainly the cut process induce stress to the plant. The mechanical cut was more aggressive for the plant causing additional stress and in response to the injury caused the plant produces a higher amount of phenolic compounds. These compounds are synthesized by plants in response to physical injury, infection or when stressed by suitable elicitors such as heavy metal-salts, temperature (Lattanzio, Lattanzio, Cardinali, & Amendola, 2006). Plants in the field are exposed to ambient solar UV-B radiation (280–320 nm) that is an environmental challenge negatively affecting DNA, proteins and membranes, thus leading to altered metabolism through the generation of reactive oxygen species (ROS). Plants protect themselves from this harmful radiation by synthesizing phenolic compounds, which act as a screen inside the epidermal cell layer, and by adjusting the antioxidant systems at both cell and whole organism level (Lattanzio et al., 2006). Therefore, being more exposed to sunlight, it is natural that plant tips had a higher phenolic compounds than other parts of the plant. Additionally, phenolic compounds have a allelopatic function: from leaves, roots, and decaying litter, plants release a variety of primary and secondary metabolites into the environment, reducing the growth of nearby plants to increase its access to light, water and nutrients (Li, Wang, Ruan, Pan, & Jiang, 2010).

3.2 HPLC analysis A representative HPLC chromatogram of phenolic compounds present in lemongrass infusion (2nd half leaves obtained by mechanical cut) is depicted in Fig. 1. Analysis of

ACCEPTED MANUSCRIPT

NU

SC R

IP

T

peaks (retention time and λmax) allowed to identify five hydroxycinnamic acids (caffeic, ρ-coumaric, ferulic, chlorogenic and rosmarinic) and a flavonoid (quercitrin). Quantification of each phenolic compound identified in all lemongrass infusions is shown in Table 4. All the compounds were previously described in lemongrass by Roriz, Barros, Carvalho, Santos-Buelga, & Ferreira (2014) (Carvalho Rodrigues, Silva, Santos, Zielinski, & Haminiuk, 2015; Nambiar & Matela, 2012; Sharoba, El Mansy, El Tanahy, El Waseif, & Ibrahim, 2015). In general, the type of cut and plant influenced significantly (p<0.05) the profile of phenolic compounds identified in the infusions. Applying this analysis the total content of these compounds was in the following order: 2nd half leaves by mechanical cut> leaves tips by mechanical cut> leaves tips by manual cut> 2nd half leaves by manual cut. Among the different compounds identified, the caffeic acid reached the highest concentration, and 2nd half of manual cut sample was significantly lower. Concerning the chlorogenic acid content significantly differences between the samples were found, being the content tips-mechanical cut sample significantly higher.

TE

D

MA

The PCA plot of aggregated data (content of individual phenolic compounds and total phenolic compounds), shown in Fig. 2, provides an overview of the similarities and differences between the different lemongrass infusions. The sum of principal components 82.5% of variation among all the variables tested and two main principal components (PC1 and PC2) accounted for 33.9 and 48.7% of the variance, respectively. These results showed that plant part has the major effect on the total and individual phenolic compounds, with better results on the plant tips. However, concerning the individual compounds, the type of cut showed higher effect, in which the mechanical cut resulted in a higher effect on the phenolic content.

AC

CE P

As relate previously, individual phenolic compounds may have different antioxidant properties (Odabasoglu et al., 2005; Ranković, 2015). Furthermore, plants respond to environmental stress through distinct morphological, biochemical, and molecular mechanisms to offset and keeping the balance. The defensive compounds are either produced constitutively or in response to plant damage(War et al., 2012). An example is the secretion of secondary metabolites, like flavonoids, terpenoids, and alkaloids by glandular trichomes of plants that allows a structural and chemical defence. Thus phenols, including chlorogenic acid, act as a defensive mechanism and their qualitative and quantitative alterations, as well as elevation in activities of oxidative enzyme in response to type of cut is a general phenomenon (Morant et al., 2008; Nuessly, Hentz, Beiriger, Snook, Widstrom, 2007; War et al., 2012)

3.3. Extraction and GC analysis of volatile (terpenoids) components Lemongrass leaves are rich in volatile compounds. The major aroma compounds identified in the studied infusions are citral (geranial and neral), 6-methyl-6 heptone-2one, linalool, geraniol, nerol and eucalyptol. As shown in Table 5, the concentration of terpenoids compounds varied with type of cut (manual and mechanical) and the harvested part of plant. Our results are consistent with earlier reports concerning the presence of the two geometric isomers of citral, geranial (44.5%) and neral (31.5%); as well as, geraniol and linalool as major components of this plant (Andrade, Zoghbi, & Lima, 2009; Sessou, , Farougou, & Kaneho, 2012). Since the major volatile compound of lemongrass infusions was citral, it was used this compound, through its isomers geranial and neral, as a markers to analyse the harvesting effects. The results suggest that concerning the geranial and neral content there are no significant differences between type of cut and leaf area (p>0.05). Biosynthesis of monoterpenes in plants is mainly due to the presence of specialized secretory structures (e.g. glandular trichomes, oil and resin ducts and glandular

ACCEPTED MANUSCRIPT epidermis) thus different parts of the plant may have different aroma compounds composition (Lewinsohn et al., 1998).

3.4. Sensory Analysis

T

3.4.1 Overall Liking

SC R

IP

Results from the two-way ANOVA show that there is no significant effect for type of cut (F(1,298) = 0.166, p = 0.684), part of the plant (F(1,298) = 0.225, p=0.636) and interaction type of cut * part of the plant (F(1,298) = 0.618, p = 0.432). These results can be explained by the high quality of the original plant and its production mode, with infusions yielding, on average high values for overall liking: tips-manual cut = 7.7(±0.9), tips – mechanical cut = 7.5(±0.8), 2nd half – manual cut = 7.5(±0.9), and 2nd half – mechanical cut = 7.6(±1.0).

NU

3.4.2 Quantitative Descriptive Analysis (QDA)

CE P

TE

D

MA

Fig. 3 shows the mean values of QDA evaluation of lemongrass infusions, according to the three-way ANOVA model (assessors, panellists and sessions). There are significant differences between samples in attributes: colour intense (F(3,6,2)=32.26, p<0.001), citric odour (F(3,6,2)=3.06, p=0.036), bitter taste (F(3,6,2)=2.45, p=0.073), after-taste (F(3,6,2)= 3.48, p=0.022) and odour intensity (F(3,6,2)=3.84, p=0.014) with a higher mean value for tips samples. Significant differences between samples in citric taste (F(3,6,2)= 4.24, p=0.009) and taste intensity (F(3,6,2)= 4.56, p=0.006) were found, with a higher mean value for ‘mechanical-tips’ samples. Samples harvested with manual cut had a higher mean value for dry grass odour attribute, with significant differences (F(3,6,2)=3.36, p=0.025). For herbal odour (F(3,6,2)=1.05, p=0.377), astringency (F(3,6,2)=0.66, p=0.580), acid taste (F(3,6,2)= 0.86, p=0.466), and dry grass taste (F(3,6,2)= 1.08, p=0.366) attributes, there were no significant differences between samples. These results show that the judges perceived the samples of the plant tips as the samples with more citric odour and citric taste and with a higher perception of intensity of colour, odour and after-taste.

AC

3.4.3. Check-All-That-Apply (CATA) Table 1 shows the significant values of CATA attributes on the sample differentiation. From Cochran’s test results a total of thirteen attributes with a significant value less than 0.10: intense (0.021), mild (<0.001), vivid colour (0.003) and transparent (0.072), from appearance attributes; lavender (0.026), eucalyptus (0.042), fresh (0.071), intense (0.070), lime (0.034) and mild (0.009), from odour attributes; light texture (0.077); and intense (0.015) and menthol (0.002) taste. Fig. 4 shows the configuration of the samples in the first and second dimension of correspondence analysis, which explained 93.7% of the data variation. The representation shows that consumers of lemongrass infusions clustered the samples in three different groups. The group of mechanical cut samples, is characterized by attributes such as vivid colour, lime and intense odour, and menthol taste. The other group, which integrate the sample ‘manual-2nd half’, is characterized by attributes like transparency, mild colour and odour. Finally, ‘tips-manual cut’ sample is characterized by intense colour, aspic and eucalyptus odour, and taste intensity. These results show that a higher percentage of the data are explained by the CATA method, in which samples of mechanical cut are related with odour attributes (intensity and citric) and fresh taste (menthol). ‘Manual-2nd half samples are mostly characterized with appearance attributes (transparency and colour) and ‘tips-manual cut’ are characterized by floral odours (aspic and eucalyptus) and taste intensity.

ACCEPTED MANUSCRIPT

3.5. Multiple Factor Analysis (MFA)

AC

CE P

TE

D

MA

NU

SC R

IP

T

Multiple factor analysis was performed to compare results from total antioxidant capacity, phenolic and aroma compounds and QDA and CATA evaluations. Table 6 shows the Rv coefficients resulting from MFA. For antioxidant activity, phenolic content, phenolic compounds, QDA and CATA the Rv value in MFA configuration were very high and closer to the unit, 0.880, 0.740, 0.910, 0.826, 0.923, 0.920, respectively. These results show that the patterns of the samples in all parameters were similar. Total phenolic content presented a lower value of Rv coefficient (0.740) in MFA analysis, nevertheless also representing a good agreement, according to Tang and Heymann (2002) who reported 0.68 as the lower limit of an acceptable Rv coefficient value. When comparing, individually, this configuration with the others the values of Rv are very low for phenolic compounds (0.426), aroma compounds (0.634), CATA (0.476) QDA (0.513) showing that the association between the total phenolic content and the other variables is weak, except for comparison with total antioxidant activity in which Rv value is higher (0.848), as expected. This lower correlation between total phenolic content and others variables mean that the samples configuration was different. Probably as the colour yield in Folin-Ciocalteu method depends on the redox potential of the reference standard used as well as the phenolic compounds and other interfering compounds (sugars, aromatic amines) present in the plant, increasing the data deviation (Singleton & Rossi, 1965). Consequently, the results of total phenolic compounds have to be interpreted carefully and should be seen as total polyphenol estimate rather than total polyphenol content. Aroma compounds had also a weak association to total antioxidant activity (0.573), CATA and QDA (0.672 and 0.590, respectively). The results showed a strong (0.981) correlation between CATA and QDA meaning that there was a high similarity in the CATA and QDA configurations. In the contribution for MFA configuration the Rv is higher too (Albert et al., 2011). These results suggest that the accuracy and reproducibility of sensory profile obtained by consumers with CATA and QDA is similar, supporting that both approaches were valid methodologies for lemongrass samples evaluation. Worch and colleagues (2010) in their research compared the configurations of QDA and CATA, and obtained a Rv value similar but lower (0.870) than the Rv of the present research, supporting that CATA and QDA were reliable methodologies for sensory product description. Thus, CATA, QDA and aroma compounds had a strong association with phenolic compounds. Total phenolic content and antioxidant activity were higher related. Fig. 5 shows the two first dimensions of the MFA consensus, which explains 86.5% of the variability of the experimental data. A closer look into Fig. 5 indicates that there was a positive correlation between geranial and neral, which together represented the citral, reinforcing the quality of the configuration. In opposition, bitter taste and taste intensity (QDA attributes) are associated with p-coumaric acid and total phenolic content. These phenolic compounds are described in the literature as compounds which provide bitter and astringent tastes to food products (Cheynier, 2005; El Gharras, 2009; Scharbert, Holzmann, & Hofmann, 2004). Moreover, lavender and eucalyptus odour (CATA) had a positive association with linalool. This aroma compound is described in the literature as a sweet, floral, herbaceous and lavender aroma (Aprotosoaie, Hăncianu, Costache, & Miron, 2014). Nevertheless, ferulic acid, quercitrin, caffeic acid, rosmarinic acid and chlorogenic acid (phenolic compounds) are associated to intensity of taste (CATA) and citric odour and taste, and aftertaste (QDA). This intensity and aftertaste are possibly caused by the presence of these compounds responsible for a bitter taste (Umar Lule & Xia, 2005).

ACCEPTED MANUSCRIPT 4. Conclusions

NU

SC R

IP

T

The premium concept of lemongrass infusions was validated by consumers with the high extraction of compounds associated to a higher overall liking expressed. From the results of all methodologies applied to evaluate the effect of type of cut and plant area, is possible to conclude that the sample with maximum content obtained in analytical methods did not totally coincide with the sample with a higher overall liking value. However, being the consumer the final judge of the product decision chain, it has been considered as premium lemongrass tea the sample obtained from plant tips with manual cut. Moreover, in consumers and judges assessment phenolic compounds provided to samples a bitter taste, a characteristic that generally was rejected. This research also shows that CATA applied to herbal infusions yields an adequate sensory description of samples, revealing similar sample configurations to the reference QDA results. Moreover, the sensory analysis is in agreement with phenolic content extraction. Acknowledgments

CE P

TE

D

MA

The authors are grateful to the ANI institution, who supported the project Infusão Premium – Criação de lotes de plantas aromáticas premium com propriedades sensoriais e funcionais validadas com elevada aceitação pelo consumidor”, through Sistemas de Incentivos à Investigação e Desenvolvimento Tecnológico do QREN. Author C. Rocha acknowledges Industry Doctoral grant No. SFRH/BDE/100483/2014, funded by FCT, Portuguese Foundation for Science and Technology (FCT). This work also received financial support from the national funds by FCT, under projects PEstOE/EQB/LA0016/2013, UID/AGR/04033/2013 and Pest-C/EQB/LA0006/2013.

4. References

AC

Albert, A., Varela, P., Salvador, A., Hough, G., & Fiszman, S. (2011). Overcoming the issues in the sensory description of hot served food with a complex texture. Application of QDA, flash profiling and projective mapping using panels with different degrees of training. Food Quality and Preference, 22(5), 463-473. Andrade, E. H. a., Zoghbi, M. D. G. B., & Lima, M. D. P. (2009). Chemical Composition of the Essential Oils of Cymbopogon citratus (DC.) Stapf Cultivated in North of Brazil. Journal of Essential Oil Bearing Plants, 12(1), 41–45. Aprotosoaie, A. C., Hăncianu, M., Costache, I.-I., & Miron, A. (2014). Linalool: a review on a key odorant molecule with valuable biological properties. Flavour and Fragrance Journal, 29(4), 193-219. Ares, G., & Jaeger, S. R. (2013). Check-all-that-apply questions: Influence of attribute order on sensory product characterization. Food Quality and Preference, 28(1), 141-153. Ares, G., Barreiro, C., Deliza, R., Giménez, A. N. A., & Gàmbaro, A. (2010). Application of a check-all-that-apply question to the development of chocolate milk desserts. Journal of Sensory Studies, 25, 67-86. Ares, G., Deliza, R., Barreiro, C., Giménez, A., & Gámbaro, A. (2010). Comparison of two sensory profiling techniques based on consumer perception. Food Quality and Preference, 21(4), 417-426. Ares, G., Varela, P., Rado, G., & Giménez, A. (2011). Are consumer profiling techniques equivalent for some product categories? The case of orange flavoured powdered drinks. International Journal of Food Science and Technology, 46(8), 1600-1608.

ACCEPTED MANUSCRIPT

AC

CE P

TE

D

MA

NU

SC R

IP

T

Aruoma, O. I. (2003). Methodological considerations for characterizing potential antioxidant actions of bioactive components in plant foods. Mutation Research/Fundamental and Molecular Mechanisms of Mutagenesis, 523524(0), 9-20. Barbosa, L. C. a, Pereira, U. A., Martinazzo, A. P., Maltha, C. R. Á., Teixeira, R. R., & Melo, E. D. C. (2008). Evaluation of the chemical composition of Brazilian commercial Cymbopogon citratus (D.C.) stapf samples. Molecules, 13(8), 1864–1874. Bertrand A. (1981). Formation de substances volatiles au cours de la fermentation alcoolique. Incidence sur la qualité des vins. In Colloque Societé Française de Microbiologie (pp. 251–267). France, Reims. Cabrera, C., Gimenez, R., & Lopez, M. C. (2003). Determination of tea components with antioxidant activity. Journal of Agricultural and Food Chemistry, 51(15), 4427-4435. Cardoso, L. (2013). Sensory evaluation of infusions of aromatic and medicinal plants: influence of post-harvest factors and preparation process [In Portuguese]. Agricultural Engineering MSc final report, Faculty of Sciences, University of Porto, Portugal. Carteri Coradi, P., de Castro Melo, E., & Pereira da Rocha, R. (2014). Mathematical modeling of the drying kinetics of the leaves of lemon grass (Cymbopogon citratus Stapf) and its effects on quality. Idesia (Arica), 32(4), 43–56. Carvalho Rodrigues, V., Silva, M. V., Santos, A. R., Zielinski, A. A., & Haminiuk, C. W. (2015). Evaluation of hot and cold extraction of bioactive compounds in teas. International. International Journal of Food Science & Technology, 50(9), 2038– 2045. Cheynier, V. (2005). Polyphenols in foods are more complex than often thought. The American Journal of Clinical Nutrition, 81(1), 223S-229S. Cunha, A. P. d., Ribeiro, J. A., & Roque, O. R. (2007). Aromatic plants in Portugal. Characterization and practices [In Portuguese]. Lisboa: Fundação Calouste Gulbenkian. Duarte Silva, I., Gaspar, J., Gomes da Costa, G., Rodrigues, A. S., Laires, A., & Rueff, J. (2000). Chemical features of flavonols affecting their genotoxicity. Potential implications in their use as therapeutical agents. Chemico-Biological Interactions, 124(1), 29-51. El Gharras, H. (2009). Polyphenols: food sources, properties and applications – a review. International Journal of Food Science & Technology, 44(12), 2512-2518. Ganjewala, D., & Luthra, R. (2010). Essential oil biosynthesis and regulation in the genus Cymbopogon. Natural Product Communications, 5(1), 163–172. Gião, M. S., González-Sanjosé, M. L., Rivero-Pérez, M. D., Pereira, C. I., Pintado, M. E., & Malcata, F. X. (2007). Infusions of Portuguese medicinal plants: Dependence of final antioxidant capacity and phenol content on extraction features. Journal of the Science of Food and Agriculture, 87(14), 2638-2647. Goey, A. K. L., Mooiman, K. D., Beijnen, J. H., Schellens, J. H. M., & Meijerman, I. (2013). Relevance of in vitro and clinical data for predicting CYP3A4-mediated herb-drug interactions in cancer patients. Cancer Treatment Reviews, 39(7), 773– 783. Hertog, M., Hollman, P., & Putte, B. v. d. (1993). Content of potentially anticarcinogenic flavonoids of tea infusions, wines and fruit juices. Journal of Agricultural and Food Chemistry, 41, 1242-1246. Hoekstra, J., Verkaik-Kloosterman, J., Rompelberg, C., van Kranen, H., Zeilmaker, M., Verhagen, H., & de Jong, N. (2008). Integrated risk-benefit analyses: Method development with folic acid as example. Food and Chemical Toxicology, 46(3), 893–909. Hollman, P. C. H., & Katan, M. B. (1999). Dietary Flavonoids: Intake, Health Effects and Bioavailability. Food and Chemical Toxicology, 37(9-10), 937-942.

ACCEPTED MANUSCRIPT

AC

CE P

TE

D

MA

NU

SC R

IP

T

ISO (2006). Sensory analysis: General guidance for the staff of a sensory evaluation laboratory - Part 1: Staff responsibilities, ISO Standard 13300-1:2006. Geneva, Switzerland: International Organization for Standardization. ISO (2006). Sensory analysis - General guidance for the staff of a sensory evaluation laboratory - Part 2: Recruitment and training of panel leaders, ISO Standard 13300–2:2006. Geneva, Switzerland: International Organization for Standardization. ISO (2007). Sensory Analysis: General guidance for the design of test rooms, ISO Standard 8589. Geneva, Switzerland: International Organization for Standardization. ISO (2012). Sensory analysis. General guidelines for the selection, training and monitoring of selected assessors and expert sensory assessors, ISO Standard 8586:2012. Geneva, Switzerland: Internantional Organiztion for Standardization. Jaeger, S. R., Cadena, R. S., Torres-Moreno, M., Antunez, L., Vidal, L., Gimenez, A., Ares, G. (2014). Comparison of check-all-that-apply and forced-choice Yes/No question formats for sensory characterisation. Food Quality and Preference, 35, 32-40. doi: 10.1016/j.foodqual.2014.02.004 Kasali, A. A., Oyedeji, A. O., & Ashilokun, A. O. (2001). Volatile leaf oil constituents of Cymbopogon citratus (DC) Stapf. Flavour and Fragrance Journal, 16(5), 377378. King, S. C., Meiselman, H. L., & Carr, B. T. (2013). Measuring emotions associated with foods: Important elements of questionnaire and test design. Food Quality and Preference, 28(1), 8-16. Lasekan, O., & Lasekan, A. (2012). Flavour chemistry of mate and some common herbal teas. Trends in Food Science & Technology, 27(1), 37-46. Lattanzio, V., Lattanzio, V. M. T., Cardinali, A., & Amendola, V. (2006). Role of phenolics in the resistance mechanisms of plants against fungal pathogens and insects. Phytochemistry (Vol. 661). Lewinsohn, E., Dudai, N., Tadmor, Y., Katzir, I., Ravid, U., Putievsky, E., & Joel, D. M. (1998). Histochemical Localization of Citral Accumulation in Lemongrass Leaves (Cymbopogon citratus (DC.) Stapf., Poaceae). Annals of Botany, 81, 35–39.Li, Z.-H., Wang, Q., Ruan, X., Pan, C.-D., & Jiang, D.-A. (2010). Phenolics and plant allelopathy. Molecules, 15(12), 8933-8952. Louw, L., Malherbe, S., Naes, T., Lambrechts, M., van Rensburg, P., & Nieuwoudt, H. (2013). Validation of two Napping® techniques as rapid sensory screening tools for high alcohol products. Food Quality and Preference, 30(2), 192-201. Macfie, H. J., Bratchell, N., Greenhoff, K., & Vallis, L. V. (1989). Designs to balance the effect of order of presentation and first-order carry-over effects in hall tests. Journal of Sensory Studies, 4(2), 129-148. Meilgaard, M. C., Carr, B. T., & Civille, G. V. (2010). Sensory evaluation techniques (3th ed.). New York: CRC Press. Morant, A. V., Jørgensen, K., Jørgensen, C., Paquette, S. M., Sánchez-Pérez, R., Møller, B. L., & Bak, S. (2008). β-Glucosidases as detonators of plant chemical defense. Phytochemistry. Nambiar, V. S., & Matela, H. (2012). Potential Functions of Lemon Grass (Cymbopogon citratus) in Health and Disease. International Journal of Pharmaceutical & Biological, 3(5), 1035–1043. Nestrud, M. A., & Lawless, H. T. (2008). Perceptual mapping of citrus juices using projective mapping and profiling data from culinary professionals and consumers. Food Quality and Preference, 19(4), 431-438. doi: http://dx.doi.org/10.1016/j.foodqual.2008.01.001 Nuessly, G.S., Hentz, M.G., Beiriger, B., Snook, M.E., & Widstrom, N.W. (2007). Resistance to spodoptera frugiperda (Lepidoptera: Noctuidae) and euxesta

ACCEPTED MANUSCRIPT

AC

CE P

TE

D

MA

NU

SC R

IP

T

stigmatias (Diptera: Ulidiidae) in sweet corn derived from exogenous and endogenous genetic systems. Journal of Economic Entomology, 100(6), 9. Nur Ain, A. H., Zaibunnisa, A. H., Halimahton Zahrah, M. S., & Norashikin, S. (2013). An experimental design approach for the extraction of lemongrass (Cymbopogon citratus) oleoresin using pressurised liquid extraction (PLE). International Food Research Journal, 20(1), 451-455. Odabasoglu, F., Aslan, A., Cakir, A., Suleyman, H., Karagoz, Y., Bayir, Y., & Halici, M. (2005). Antioxidant activity, reducing power and total phenolic content of some lichen species. Fitoterapia, 76(2), 216–219. Parente, M. E., Manzoni, A. V., & Ares, G. (2011). External preference mapping of commercial antiaging creams based on consumers' responses to a check-allthat-apply question. Journal of Sensory Studies, 26(2), 158-166. Pennington, J. A. T. (2002). Food Composition Databases for Bioactive Food Components. Journal of Food Composition and Analysis, 15(4), 419-434. Peryam, D. R., & Pilgrim, F. J. (1957). Hedonic scale method of measuring food preferences. Food Technology, 11 (Suppl. 9), 9–14. Ranković, B. (2015). Lichen secondary metabolites: Bioactive properties and pharmaceutical potential. Lichen Secondary Metabolites: Bioactive Properties and Pharmaceutical Potential. Rocha, C., Laranjeira, A., Moura, A. P., Lima, R. C., & Cunha, L. M. (2013). On the selection of consumers for a descriptive panel group to evaluate new products from the application of innovative food processing technologies. [Poster Presentation] 10th Pangborn Sensory Science Symposium, 11- 15 August 2013, Rio de Janeiro, Brazil. Rocha, C., Madeira, A., Lima, R. C., Cardoso, L., Alves, L., & Cunha, L. M. (2015). Optimization of processing conditions for improved sensory acceptance of herbal teas: effect of time and temperature. [Oral Communication] 9th International Conference on Culinary Arts and Sciences 13-16 October, New Jersey. Roriz, C. L., Barros, L., Carvalho, A. M., Santos-Buelga, C., & Ferreira, I. C. F. R. (2014). Pterospartum tridentatum, Gomphrena globosa and Cymbopogon citratus: A phytochemical study focused on antioxidant compounds. Food Research International, 62, 684–693. Saleem, M., Afza, N., Anwar, M. A., Hai, S. M. A., & Ali, M. S. (2003). A comparative study of essential oils of Cymbopogon citratus and some members of the genus Citrus. Natural product research, 17(5), 369-373. Schaneberg, B. T., & Khan, I. A. (2002). Comparison of extraction methods for marker compounds in the essential oil of lemon grass by GC. Journal of Agricultural and Food Chemistry, 50(6), 1345-1349. Scharbert, S., Holzmann, N., & Hofmann, T. (2004). Identification of the astringent taste compounds in black tea infusions by combining instrumental analysis and human bioresponse. Journal of Agricultural and Food Chemistry, 52(11), 34983508. Sessou, P., S. Farougou, S. Kaneho, S. D. and G. A. A. et al. (2012). Bioefficacy of Cymbopogon citratus essential oil against foodborne pathogens in culture medium and in traditional cheese wagashi produced in Benin. Internantional Research Journal Microbiology, 3, 406–415. Sharoba, A. M., El Mansy, H. A., El Tanahy, H. H., El Waseif, K. H., & Ibrahim, M. A. (2015). Chemical Composition, Antioxidant and Antimicrobial Properties of the Essential Oils and Extracts of Some Aromatic Plants. Middle East Journal of Applied Sciences, 5(2), 344–352. Singh‐Sangwan, N., Abad Farooqi, A., & Singh Sangwan, R. (1994). Effect of drought stress on growth and essential oil metabolism in lemongrasses*. New phytologist, 128(1), 173-179.

ACCEPTED MANUSCRIPT

AC

CE P

TE

D

MA

NU

SC R

IP

T

Singleton, V. L., & Rossi J A Jr. (1965). Colorometry of total phenolics with phosphomolybdic-phosphotungstic acid reagents. American Journal of Enology and Viticulture, 16(3), 144–158. Tang, C., & Heymann, H. (2002). Multidimensional sorting, similarity scaling and freechoice profiling of grape jellies. Journal of Sensory Studies, 17(6), 493509.Tascilar, M., de Jong, F. A., Verweij, J., & Mathijssen, R. H. J. (2006). Complementary and Alternative Medicine During Cancer Treatment: Beyond Innocence. The Oncologist, 11(7), 732-741. Umar Lule, S., & Xia, W. (2005). Food Phenolics, Pros and Cons: A Review. Food Reviews International, 21(4), 367-388. Varela, P., & Ares, G. (2012). Sensory profiling, the blurred line between sensory and consumer science. A review of novel methods for product characterization. Food Research International, 48(2), 893-908. War, A. R., Paulraj, M. G., Ahmad, T., Buhroo, A. A., Hussain, B., Ignacimuthu, S., & Sharma, H. C. (2012). Mechanisms of plant defense against insect herbivores. Plant Signaling & Behavior, 7(10), 1306–1320. Worch, T., Lê, S., & Punter, P. (2010). How reliable are the consumers? Comparison of sensory profiles from consumers and experts. Food Quality and Preference, 21(3), 309-318.

ACCEPTED MANUSCRIPT Legends for Figures

T

Figure 1 - HPLC–DAD chromatogram representing the phenolic compounds profile of lemongrass infusion of half-parts of leaves obtained by mechanical cut. Detection at 320 nm. Peaks: (1) chlorogenic acid; (2) caffeic acid; (3) p-coumaric; (4) ferulic acid; (5) quercitrin; (6) rosmarinic acid.

SC R

IP

Figure 2 - Principal component analysis of aggregated data from different lemongrass infusions. Infusions of 2nd half leaves with mechanical cut, 2nd half leaves with manual cut, tips with mechanical cut, and tips with manual cut are represented as Mec_half. Man_half. Mec_tip and Man_tip, respectively. Total phenolic compounds are represented as TPC.

NU

Figure 3 - Mean values of Quantitative Descriptive Analysis (n = 8) performed to lemongrass infusions according to the ANOVA model of three factors (samples. tasters and sessions). (***significant at p<0.001; ** p<0.05. *p<0.10). Figure 4 - Dimensions 1 and 2 of Correspondence Analysis of CATA (Check-All-ThatApply) questionnaire applied to lemongrass samples.

AC

CE P

TE

D

MA

Figure 5 - Dimension 1 and 2 of Multiple Factor Analysis applied to total antioxidant (red lines), total phenolic compounds (blue lines), individual phenolic compounds (green lines) aromatic compounds (purple lines), QDA (orange lines) and CATA (grey lines) data.

ACCEPTED MANUSCRIPT

AC

CE P

TE

D

MA

NU

SC R

IP

T

Figure 1

ACCEPTED MANUSCRIPT Figure 2

T

Biplot (axes D1 and D2: 82.54 %) after Varimax rotation

IP

1.2

Man_tips TPC

SC R

0.8

p-Coumaric acid

Mec_tips

Ferulic acid

0.4

NU

-0.4

Man_2ndhalf

Caffeic Acid Rosmarinic Acid

-1.2 -1.2

-0.8

CE P

-1.6

TE

D

-0.8

Quercitrin

MA

0

AC

D2 (33.86 %)

Chlorogenic Acid

-0.4

Mec_2ndhalf

0

D1 (48.68 %)

0.4

0.8

1.2

1.6

ACCEPTED MANUSCRIPT Figure 3 Colour***

T_intensity***

NU

Aftertaste**

O_citric**

T

T_acid

O_dry grass**

IP

T_dry grass

SC R

9.00 8.00 7.00 6.00 5.00 4.00 3.00 2.00 1.00 0.00

T_citric***

O_herbal

O_intensity**

Astringency

MA

T_bitter*

Mechanical - 2nd half

AC

CE P

TE

D

Manual - 2nd Half

Manual - Tips

Mechanical - Tips

ACCEPTED MANUSCRIPT Figure 4 Symmetric plot (axes F1 and F2: 93,69 %)

T_Menthol O_Lime

Transparent

Intense Colour

O_Eucalyptus T_Intense O_Aspic

-0.05

Mild Colour

Manual - 2nd Half

O_Mild

MA

Manual - Tips

O_Fresh Light Texture

NU

0

O_Intense

Mechanical - Tips Vivid Colour

SC R

Mechanical - 2nd Half

-0.05

CE P

TE

-0.1

D

-0.1

AC

F2 (14,67 %)

0.05

IP

T

0.1

0

0.05

F1 (79,01 %)

0.1

0.15

ACCEPTED MANUSCRIPT Figure 5 Variables (axes F1 and F2: 86,48 %) 6-methyl-6 heptone-2one

1

Geranial Neral

0.75

T

O_Intense O_Lime Quercitrin

IP

T_Menthol

Chlorogenic Acid T_citric Rosmarinic Acid

Transparent

0

Mild Colour -0.25

O_Fresh

-0.5

-1

-0.5

CE P

-0.75

TE

-0.75

AC

Total antioxidant activity

T_intensity p-Coumaric acid O_Lavender

TPC

D

O_dry grass

-1

O_intensity T_bitter Colour Linalool O_Eucalyptus

Gallic acid

Geraniol O_Mild

Caffeic Acid O_citric Vivid Colour Aftertaste Ferulic acid T_Intense Intense Colour

NU

TAC

MA

F2 (25,86 %)

0.25

SC R

0.5

-0.25

Light Texture

0

0.25

0.5

0.75

1

F1 (60,62 %)

Total phenolic content

Phenolic compounds

Aromatic compounds

CATA

QDA

ACCEPTED MANUSCRIPT Legends for Tables Table 1 – CATA attributes and p-value of Cochran’s Test.

SC R

IP

T

Table 2 - Attributes used in quantitative descriptive analysis. Table 3 - Total antioxidant activity and total phenolic content (average ± standard deviation) of different lemongrass infusions. Table 4 – Content of phenolic compounds identified in lemongrass infusions (µg/kg dry leaf)1 obtained by HPLC. Table 5 – Chemical composition of volatile compounds identified in lemongrass infusions.

AC

CE P

TE

D

MA

NU

Table 6 - Rv coefficients resulting from Multiple Factor Analysis applied to antioxidant, phenolic and aromatic content, QDA and CATA data (p -value is shown in brackets).

ACCEPTED MANUSCRIPT Table 1 Appearance (11)

p-value

Odour (cont.)

p-value

Taste (25)

p-value

n. a.

Dry straw

0.580

Acid

0.753

Golden colour

n. a.

Herbal

0.675

Aftertaste

0.634

Grass-yellow colour

0.948

Honey

n. a.

Bitter

0.392

Green colour

0.373

Intense

0.070

Citric

0.958

Intense colour

0.021

Lemon

Lemon-yellow colour

0.697

Lime

Mild colour

<0.001

Medicinal

No particles

0.172

Menthol

Transparent

0.072

Metallic

Vivid colour

0.003

Yellow colour

0.937

IP

SC R

0.506

0.034

Fresh

0.857

0.245

Fresh cut grass

0.414

0.801

Herbal

0.325

n. a.

Intense

0.015

Mild

0.009

Soil

n. a.

Mint

0.365

Lemon

0.691

Mould

n. a.

Medicinal

0.873

Naphthalene

n. a.

Menthol

0.002

MA

NU

Floral

TE

p-value

0.110

D

Odour (37)

T

Dark colour

n. a.

Orange flower

0.224

Metallic

n. a.

Anise

n. a.

Peach

n. a.

Mild

0.115

0.026

Pineapple

n. a.

Mint

0.576

n. a.

Spice

0.748

No aftertaste

n. a.

n. a.

Sweet

0.182

No taste

n. a.

n. a.

Tangerine

n. a.

Peppery

0.607

0.518

Thyme

0.801

Spice

n. a.

Fresh cut grass

0.881

Typical

0.837

Spicy

n. a.

Dry

0.543

Vegetable

0.349

Sweet

0.753

Eucalyptus

0.042

Wood

n. a.

Thyme

n. a.

Floral

0.724

Typical

0.121

Fresh

0.071

Texture (5)

Wood

n. a.

Fruity

0.365

Astringent

0.801

Grapefruit

0.629

Fresh

0.930

Light

0.077

CE P

Algae

Lavender

Camphor Cedar Citric

AC

Black pepper

p-value

ACCEPTED MANUSCRIPT Smooth

0.649

Soft

0.290

AC

CE P

TE

D

MA

NU

SC R

IP

T

n. a. – not applicable, maximum count less than five.

ACCEPTED MANUSCRIPT Table 2 - Attributes used in quantitative descriptive analysis. Description

Colour intensity

Magnitude of typical colour of lemongrass infusions

Herbal odour

Magnitude of herbal odour of plants

Dry straw odour

Strength of dry grass odour

Global odour

Strength of global odour of lemongrass infusions

Astringency

Feeling of roughness and dryness of the tongue associated with the presence of tannins

Bitterness

Magnitude of basic taste of bitterness as found in caffeine solution

Sourness

Magnitude of basic taste of acidity as found in tartaric acid solution

Dry straw taste

Magnitude of dry grass taste

Citric taste

Magnitude of citric taste

Global taste

Magnitude of global taste

Aftertaste

Intensity of residual taste

AC

CE P

TE

D

MA

NU

SC R

IP

T

Attributes

ACCEPTED MANUSCRIPT Table 3 -

Manual cut

Mechanical cut-

2nd half

Tips

2nd half

0.45±0.01a

0.55±0.01c

0.47±0.02a,b

0.51±0.01b,c

23.07±0.28a

16.09±1.62c

19.91±0.65b

ABTS●+

Folin-Ciocalteu

SC R

(mg AAE/g dry leaf)

CE P

TE

D

MA

NU

(mg GAE/g dry leaf)

AC

IP

T

Tips

16.49±0.17c

p-value (from ANOVA) Type of cut

Part of Interplant action

0.705

0.000

0.004

0.027

0.000

0.008

ACCEPTED MANUSCRIPT Table 4 – Compounds

Manual cut

Mechanical cut

(µg/kg dry leaf)

(µg/kg dry leaf)

p-value (from ANOVA)

2

RT --

2nd half

Type of cut

Part of plant

Interaction

8.54±0.70b

0.000

0.000

0.000

Tips

2nd half

13.38±0.36a

T

Tips 20

4.29±0.02c

Caffeic acid

23

20.76±2.80a 11.85±2.28b

19.96±1.45a

22.80±0.40a

0.002

0.027

0.001

p-Coumaric acid

30

3.01±0.31a

2.28±0.04c

2.93±0.02a,b

2.54±0.05b

0.375

0.000

0.108

Ferulic acid

33

0.66±0.03b

0.47±0.06c

0.84±0.06a

0.66±0.03b

0.000

0.000

0.816

Quercitrin

40

1.21±0.08a

0.90±0.16

1.37±0.10a

1.41±0.06 a

0.001

0.055

0.022

Rosmarinic acid

45

3.82±0.34a

2.34±0.47b

2.76±0.10b

4.39±0.07 a

0.021

0.674

0.000

SC R

NU

MA

∑

2.72±0.36d

33.75

20.56

IP

Chlorogenic acid

40.34

41.25

Σ, sum of individual phenolic compounds identified; Values are expressed as mean±standard deviation of three assays.

2

RT, retention time (min).

AC

CE P

TE

D

1

ACCEPTED MANUSCRIPT Table 5 Manual cut (%) RT1 2nd half

Tips

Type of cut

Part of plant

Interaction

1.26±0.06a

0.720

0.990

1.20±0.01a

0.731

0.007

0.013

2nd half

T

Tips

p-value (from ANOVA)

6-methyl-6 heptone-2-one

16

1.20±0.03 a

1.22±0.13 a

0.565

Linalool

24

1.27±0.02 a

1.11±0.04a

1.21±0.02a

IP

Compounds

Mechanical cut (%)

Neral

30

35.73±0.59a

35.54±0.50a

35.61±0.65a

37.02±0.63a

0.181

0.225

0.124

Geranial

33

48.40±0.86a

48.56±1.37a

48.50±1.00a

50.41±0.95a

0.305

0.278

0.352

Nerol

37

0.20±0.01 a

0.22±0.01a

0.20±0.01 a

0.00±0.00a

1.000

0.066

n.a.

Geraniol

39

6.77±0.24b

7.56±0.30a

6.63±0.12 b

6.90±0.08 b

0.061

0.021

0.191

Eucalyptol

57

0.03±0.01 b

0.01±0.00b,c

0.00±0.00c

0.014

0.122

n.a.

SC R

NU

MA

1

1.24±0.02a

RT, retention time (min).

0.08±0.00a

AC

CE P

TE

D

n.a. – not applicable, due to lack of replicates.

ACCEPTED MANUSCRIPT Table 6 Total Total antioxidant phenolic Phenolic Aromatic activity content compounds compounds

0,848

1,000

Total phenolic content

SC R

(0,083)* 0,667

0,426

1,000

(0,458)

(0,833)

0,573

0,634

(0,708)

(0,583)

0,735

0,476

(0,208)

(0,417)

(0,000)***

(0,625)

0,812

0,513

0,904

0,590

0,981

(0,042)**

(0,208)

(0,083)*

(0,750)

(0,042)**

0,740

0,910

0,826

0,923

Phenolic compounds

***

Indicates that the RV coefficient is significant at p < 0.05

CE P

*

Indicates that the RV coefficient is significant at p < 0.01

Indicates that the RV coefficient is significant at p < 0.10

AC

**

0,880

TE

MFA

D

QDA

1,000

(0,500) 0,932

MA

CATA

0,764

NU

Aromatic compounds

MFA

T

1,000

QDA

IP

Total antioxidant activity

CATA

0,672

1,000

1,000

0,920

1,000

ACCEPTED MANUSCRIPT

AC

CE P

TE

D

MA

NU

SC R

IP

T

Graphical Abstract

ACCEPTED MANUSCRIPT Highlights

CE P

TE

D

MA

NU

SC R

IP

T

Influence of harvesting conditions in sensory, phenolic and aroma properties; Association of phenolic compounds with bitterness in sensory perception; Cymbopogon citratus harvested from tips of the plant yielded premium infusions.

AC

  