Identification of volatile compounds in salep (Serapias vomeracea) tubers and effects of harvest time and drying method on composition variation

ABSTRACT Serapias vomeracea is an economically valuable plant belonging to the Orchidaceae family; particularly, its tubers are consumed as hot drinks. Past studies on S. vomeracea have focused on volatile components only in the aerial parts of this orchid. This study is the first to investigate the volatile compounds present in S. vomeracea tubers, which have a high commercial value. Additionally, we determined the effect of harvest time and drying method on the volatile compound profile. The tubers were harvested on two different days (May 17, 2022, and June 6, 2022) and subjected to four different drying conditions (shade drying, oven drying, shade drying after boiling, and lyophilizer drying). To determine the volatile compound profile, sample extracts were prepared by solid-phase microextraction (SPME) and analyzed by gas chromatography-mass spectrometry. Overall, 22 volatile compounds were identified: 3 esters, 3 aldehydes, 3 alcohols, 3 terpenes, 3 alkanes, 3 carboxylic acids, 1 phenol, 1 ketone, 1 lactone, and 1 furan. The results revealed that p-cresol was the main compound. During harvesting and subsequent drying, some compounds were lost, and some new compounds were formed. Moreover, harvest time had quite limited effects on volatile compounds in the lyophilizer method. Cluster analysis revealed that the combination of harvest time and drying was effective in the distribution of volatile compounds in salep powder.


INTRODUCTION
The salep plant belongs to the Orchidaceae family (Christenhusz;Byng, 2016).This family includes plants with tuberous and tuberless roots (Clifford;Lavarack, 1974).Serapias is a member of the genus comprising plants with tuberous roots (Caliskan;Kurt;Odabas, 2020).Serapias vomeracea (Burm.f.) Briq. is an important species and an emerging crop that is increasingly being used in Turkey.Harvested tubers are dried and ground into a fine powder before use (Turgay;Çınar, 2017).Salep is an important raw material in the pharmaceutical and food industries owing to the presence of glucomannan.In Turkey, "salep" refers to the traditional beverage prepared from dried and fine-ground tubers (Tamer;Omeroglu;Copur, 2019).Salep powder is used as a raw material to prepare Maraş ice cream (Sezik, 2002).Hossain (2011) reported that orchids are used as medicines to cure many diseases, such as paralysis, stomach ailments, asthma, inflammation, tuberculosis, rheumatism, and chest pain since prehistoric times.
Understanding the diversity and proportional values of fragrance compounds is essential in revealing the differences between products and applications.Therefore, solid-phase microextraction (SPME) is used for this purpose.SPME (Xu et al., 2016) is proven as an effective tool available within a short time; it is also an inexpensive and solvent-free extraction technique and can be used for flavor analysis in foods (Souza-Silva;Gionfriddo;Pawliszy, 2015).This method has been widely used in many studies.Several studies have focused on the detection of volatile compounds in the aerial parts (parts present above the ground) of various orchid species.Erik et al. (2020) investigated the profile of volatile compounds present in solvent extracts of Serapias orientalis subsp.orientalis plants.The major compound identified in that study was limonene (76.5%), followed by heptanal (10.4%), pelargonaldehyde (7%), and capronaldehyde (2.1%).Moreover, components whose levels proportionally increased and decreased were determined by solvent extraction.The authors reported that new compounds were formed, such as α-methoxy-p-cresol (52.9%) in water extract, decanal (7.5%), and 2(E)-heptenal (2.1%) in methanolic extract, and octanal (3.6%) in hexane extract.Mecca et al. (2022) conducted a study on the volatile compounds present in the flowers of Dactylorhiza species belonging to the Orchidaceae family, and the highest area values were attributed to verbenone (28.86%) in Dactylorhiza viridis, β-ocimene (18.69%) in Dactylorhiza romana, pentadecane (28.40%) in Dactylorhiza incarnata, caryophyllene (17.38%) in Dactylorhiza saccifera, and β-sesquiphellandrene (32.16%) in Dactylorhiza sambucina.
Past studies have paid attention to the volatile components present in the aerial parts of orchids obtained from natural flora.Although the tuber is the main part of the salep plant that is consumed and used, only a few studies have determined the volatile compounds present in salep tubers by SPME analysis.Salep plant harvesting in Turkey is usually carried out during the full flowering period of the plants or later for various reasons.Harvested tubers are traditionally cleaned, boiled, and then dried in the shade.
Different drying methods can be applied to tubers.To the best of our knowledge, the present study is the first to determine the volatile compounds present in S. vomeracea tubers.Moreover, given that the influence of late harvest on the volatile compound profile remains unknown, this study determined the effects of harvest time and drying method on the composition variation of volatile compounds in salep tubers by SPME and gas chromatography-mass spectrometry (GC-MS) analysis.

Material preparation for analysis
This study was conducted in the Pazar District of Rize Province.S. vomeracea tubers were used in this study.The seedlings that emerged from tubers were transplanted on December 2, 2021, in plots filled with media comprising field soil, stream sand, and tea fiber (in a 2:1:1 ratio, respectively).Before transplanting the seedlings, a basal dose of 50-kg NPK per hectare (15-15-15) was applied to the plots.
Tubers from half of the grown plants were harvested on May 17, 2022 (first harvest time: H1), during the full flowering period.The remaining plants were harvested on June 6, 2022, which was 20 days after the first harvest (H2).The harvested tubers were washed to remove dust, after which they were dried.Ten tubers were dried.Drying was carried out in four different methods: shade drying, oven drying, shade drying after boiling, and lyophilizer drying.
(1) Shade drying: The samples were dried under room conditions and in the shade at 25 ±1 °C for 20 days.
(2) Oven drying: The samples were dried in an oven at 40 °C for 10 days.
(3) Shade drying after boiling: This method is traditionally used in Turkey.First, the samples were boiled for 7 min, after which they were filtered and dried in the shade at 25 ±1 °C under room conditions for 10 days.(4) Lyophilizer drying: Labconco Lyophilizer (0.12 mbar vacuum and adjusted to −84 °C) was used for the drying process.In this method, the drying process took 42 h.The dried tubers were ground into a fine powder and preserved in a refrigerator at 4 °C until further analysis.
The extract of volatile compounds was prepared by SPME using fiber (50/30-µm film thickness, 2-cm length, divinylbenzene/carboxen/polydimethylsiloxane).After samples were weighed to 0.1 g, they were placed in vials and placed on a hotplate at 65 °C for 5 min.Before the analysis, fiber was desorbed into the GC-MS injector at 65 °C for 5 min.SPME analyses were carried out in two replications.
The volatile compounds present in the samples were identified using a GC-MS system (PQ2010 Ultra; Shimadzu, Japan).Volatile compounds were separated using a fused-silica capillary column Rxi-5Sil MS (30.0 m, 0.25 mm ID, 0.25-µm thickness).The injection was carried out in split mode (1:25), and the column temperature was 250 °C.The GC-MS interface and ionization source temperatures were 210 °C and 250 °C, respectively.The column oven temperature was set at 40 °C.The carrier gas used was high-purity helium.Its flow rate was 1.00 mL min −1 .Detection was realized in electronic ionization in the pulse mode, and the ionization voltage was 70 eV.All compounds were identified by the NIST 1992 and Wiley mass spectral libraries.The relative intensity of each compound was calculated as the ratio between the area of the specific molecule and the sum of the areas of all identified peaks (% peak area normalization method) in the chromatogram (Kafkas et al., 2006).

Statistical analysis
The data of volatile compounds were obtained using SPME/GC-MS from the library and confirmed against references.Graphs were constructed using the obtained data with Microsoft Excel 2016.Hierarchical cluster analysis was performed using SPSS 20.0 software.Additionally, before PC analysis, the relative area value of each compound was recalculated using Log+1 transformation using the SPSS program.Cluster analysis (Ward method with squared Euclidean distances) was used for classifying a dataset into homogeneous groups.The results of processed data are displayed as a diagram, called a dendrogram, which belongs to hierarchical clustering.

RESULTS AND DISCUSSION
Twenty-two volatile compounds were detected in eight samples and identified by SPME/GC-MS analysis (Table 1).The compounds and their groups are listed in Table 1.Only three (phenol, aldehyde, and alcohol) of the ten compound groups have an area ratio of more than 5% (Figure 1).The areal distribution of volatile compounds (above 3%) according to harvest time and drying is shown in Figure 2. The data of the variation in the volatile compounds of the salep tuber are evaluated below.

Harvest time
Volatile compounds showed a wide variation in number and percentage area depending on the harvest time.The oven drying method, followed by shade drying after boiling, shade drying, and lyophilizer drying, yielded the highest number of compounds (18 compounds) in tubers at the first harvest time (H1).In all drying conditions, the main compound obtained was p-cresol (minimum: 64.96%, maximum: 87.32%; Table 1).In the second harvest time (H2), shade drying, followed by oven drying, shade drying after boiling, and lyophilizer drying, yielded the highest number of compounds.Moreover, p-cresol values (minimum: 70.71%, maximum: 90.21%) showed increases in all the methods.Therefore, p-cresol was the primarily volatile compound in S. vomeracea tubers.
In H1, the average percentage area of the phenolic group was the highest at 74.53%, followed by the average percentage area of the aldehyde and alcohol groups at 8.15% and 5.90%, respectively.In H2, the ranking was similar, and the corresponding average values of these groups were 83.17%, 5.73%, and 5.39% (Figure 1).
Depending on the interaction of the drying × harvest time, the number of volatile compounds with an area ratio of above 3% was 10 (Table 1).Except for 2-ethyl hexanol, nine compounds (acetoin, ethyl lactate, capronaldehyde, p-methyl anisole, phenylmethanol, p-cresol, pelargonaldehyde, capraldehyde, and ethylene brassylate) were effective in H1.In H2, the number of volatile compounds with an area ratio of above 3% was only 5 in at least one sample.These compounds were capronaldehyde, 2-ethyl hexanol, phenylmethanol, p-cresol, and pelargonaldehyde.
In general, late harvest by 20 days caused an increase in the ratio of p-cresol, which had an animal, ink, or feceslike odor (Beauchamp;Zardin, 2017) and was the only phenolic volatile compound obtained in the study.Among the drying conditions, shade drying after boiling yielded the highest increase in this compound.p-Cresol exists at a high concentration mostly in animal products such as milk (Faulkner et al., 2017), cheese (Kilcawley et al., 2018), meat (Begić;Forto;Krvavica, 2022), and yogurt (Cheng et al., 2022).Interestingly, this compound also exists in a significant amount in coffee, depending on the roasting process (Ayseli et al., 2021).p-cresol concentrations in essential oils from flowers of different salep species were 38.10% in Anacamptis morio, 15.28% in Himantoglossum robertianum, 12.75% in Ophrys sphegodes, and 12.99% in Orchis purpurea (Robustelli della Cuna et al., 2022).Dalar et al. (2015) reported that the p-cresol ratio in the extracts of Dactylorhiza chuhensis flowers was 4.4%, which was much lower than that obtained in this study.This is because both the investigated species and the analyzed plant part are different from each other.

Shade drying after boiling
The p-cresol ratio was 64.96% in tubers that were harvested during full flowering and then shade-dried after boiling, followed by capronaldehyde (6.19%), pelargonaldehyde (4.27%), capraldehyde (4.02%), ethyl lactate (3.80%), and others.As for the percentage area, 5 of 16 compounds existed at a concentration below 1%.In tubers belonging to the second harvest period, the total number of compounds detected decreased to 13.In this method, while the p-cresol ratio increased to 88.30%, the ratios of other compounds decreased.As for the ranking of compounds, p-cresol was the highest, followed by 2-ethyl hexanol (1.91%), pelargonaldehyde (1.56%), hexadecane (1.47%), and others (Table 1).on the harvest time and drying method.This trend of volatile compounds revealed in the present study is consistent with the findings of Ye et al. (2019) for Anoectochilus roxburghii from the Orchidaceae family, those of Díaz-Maroto, Pérez-Coello, and Cabezudo (2022) for Laurus nobilis (Mr. Leaf) from the Lauraceae family, those of Zhang et al. (2016) for Capsella bursapastoris from the Brassicaceae family, and those of Polat et al. (2022) for Daucus carota (black carrot) from the Apiaceae family, indicating that volatile compounds differed depending on the drying conditions.Some volatile compounds in salep tubers were specifically seen only in the harvest time-drying method combination.3-Methylvaleric acid, ethyl hexanoate, and dodecane are examples of such compounds (Table 1).3-Methylvaleric acid was obtained only by boiling and shade-drying of the tubers harvested at full bloom time.Ethyl hexanoate compound was detected only in tubers harvested at H2 and shade-dried.Dodecane was detected only in tubers harvested at H2 and dried in the lyophilizer.
In the shade drying method, five new volatile compounds (isovaleric acid, ethyl hexanoate, p-methyl anisole, limonene, and ethylene brassylate) were identified in the late harvest (Table 1).In the shade drying method after boiling, three compounds, namely, ethyl lactate, 3-methylvaleric acid, and ethylene brassylate, were seen in tubers from the first harvest time but disappeared in tubers from the second harvest time.This effect of harvest time may have resulted particularly from ontogenetic variability.Similarly, many researchers, such as Stefanakis et al. (2022), Li andZidorn (2022), andHaghighi et al. (2019), have pointed out the importance of seasonal variation regarding harvest time.
The drying method in an oven at 40 °C caused the loss of four compounds (ethyl lactate, 2-pentylfuran, p-methyl anisole, and ethylene brassylate) in tubers from the second harvest time.By contrast, in the lyophilizer drying method, two compounds (diethyl phthalate and ethylene brassylate) were lost in tubers of the second harvest time, but one new compound (dodecane) appeared (Table 1).Díaz-Maroto, Pérez-Coello and Cabezudo (2002) reported the least change in the volatile composition of Petroselinum crispum from the Apiaceae family when the plants were dried at an ambient temperature; additionally, they reported that oven drying (at 45 °C) and freeze drying led to a decrease in most of the volatile compounds.In a similar approach, according to Ding et al. (2012), drying at higher temperatures may result in the decomposition of components, whereas drying at lower temperatures may inhibit the appearance of new compounds.The appearance of new compounds after drying and the realization of losses in some compounds may have been caused by glycoside hydrolysis and esterification (Bhatt et al., 2018).Hierarchical clustering of volatile compounds considering harvest time and drying conditions.The clustering of volatile compounds based on the drying conditions applied to the tubers harvested at H1 (first harvest time) and H2 (second harvest time) was performed using the hierarchical clustering Ward method.
The results of the dendrogram obtained with the examined traits showed the formation of two separate clusters (Figure 3).One of these clusters was separated into two sub-clusters, wherein one of the subgroups consisted of H1-Lyophilizer and H2-Lyophilizer.The other cluster formed two sub-clusters within itself and a total of three different subgroups.The first subgroup consisted of H1-Shade and H2-Etuv, while the second subgroup consisted of H1-Etuv, H2-Shade, and H1-ShadeB.In the third subgroup, H2-ShadeB was found alone.H1-Shade and H2-Etuv had highly similar properties in the cluster.Finally, a total of 4 clusters were formed, namely, A, B, C, and D.
The volatile compounds and area values obtained from samples dried in the lyophilizer were not affected much by the harvest time, unlike the situation in other drying conditions.The results of the SPSS cluster analysis indicated that, generally, the differences among the drying conditions in terms of the volatile compounds obtained varied depending on the harvest time.
Hierarchical clustering analysis of compound groups depending on the harvest time and drying conditions.In the clustering analysis considering the compound groups, initially, two clusters had formed (Figure 4).Phenol group alone formed a cluster, whereas other groups were divided into two subgroups.Aldehyde and alcohol groups formed one of these clusters.The other cluster was divided into two more subgroups.Finally, a total of 4 clusters, namely, A, B, C, and D, had formed.Cluster analysis was performed to reveal the relationships between volatile component groups present in the tubers of S. vomeracea.The dendrogram of the hierarchical clustering analysis showing the component groups acting together is illustrated in Figure 4.
The first of the four clusters, i.e., Cluster A, consisted of the aldehyde and alcohol groups.The second of the four clusters (Cluster B) consisted of ester, lactone, carboxylic acid, furan, terpene, and alkane groups.The third one (Cluster C) consisted of only the ketone group, and the last cluster (Cluster D) consisted of the phenol group alone (Figure 4).

CONCLUSIONS
This study demonstrated that the type of volatile compounds present in the tubers of S. vomeracea grown under field conditions changed depending on the harvest time and drying conditions.Some compounds were lost or new compounds appeared depending on the application.This study presents original information about the formation and diversity of volatile compounds under different drying conditions.The findings of this study can serve as an important resource for researchers in obtaining more specific salep powder.

Figure 1 :
Figure 1: Areal distribution of compound groups depending on drying conditions in the first harvest time (A) and second harvest time (B).

Figure 3 :
Figure 3: Hierarchical clustering analysis of volatile compounds considering harvest time × drying method.Clusters A, B, C, and D were identified at cutting point 10.Eight combinations were obtained: H1 Shade = shade drying in the first harvest time, H2Etuv = oven drying in the second harvest time, H1Etuv = oven drying in the first harvest time, H2Shade = shade drying in the second harvest time, H1ShadeB = shade drying in the first harvest time, H2ShadeB = shade drying after boiling in the second harvest time, H1Lyoph = lyophilizer drying in the first harvest time and H2Lyoph = Lyophilizer drying in the second harvest time.

Figure 4 :
Figure 4: Hierarchical clustering of compound groups.Clusters A, B, C, and D were identified at cutting point 3.

Table 1 :
Results of identification harvest time and drying conditions, retention index, and a chemical class of the volatile compounds.