Influence of Rheum taxa and harvesting date on the content of L-ascorbic acid and oxalic acid in the climatic conditions of South Moravia (Czech Republic)

The influence of Rheum taxa and harvesting time on the L-ascorbic and oxalic acid content was evaluated in 16 accessions of the Rheum genetic resources, containing 13 accessions of R. rhabarbarum, representing the most cultivated species of culinary rhubarb in the Czech Republic, two accessions of R. rhaponticum and one hybrid of R. palmatum × wittrockii (Table 1). The input taxonomic classification of the studied accessions was stated according to the plant providers. The rhubarb plants were cultivated on experimental fields at Mendel University in Brno, the Faculty of Horticulture in Lednice, Czech Republic (48°47′36″ N, 16°47′48″ E). Plants have been grown on this plot since 2015 and the 42H7500047 accession since 2018.

Meteorological data were recorded with automatic sensors located near the experimental field. Recorded data comprise the amount of precipitation (mm), air temperature (°C) and sunshine duration, defined as the number of sunny hours. The long-term average values (2015–2021) of selected climate characteristics and the values of the studied period (2022) are shown in Table 2 (CHMI, 2022).

According to the agrochemical analyses of the soil from the experimental field in 2022, the nitrogen content was 0.146%, the content of phosphorus was 37.794 mg · kg^-1 and the potassium content was 198 mg · kg^-1. The calcium and magnesium contents were 4961 mg · kg^-1 and 198 mg · kg^-1, respectively. The pH (H₂O) of the soil at a depth of 0.3 m was 7.71. The area of cultivation of each accession was 2.9 m². Plants were grown without additional irrigation.

Sixteen accessions of Rheum genetic resources were evaluated during the years 2015, 2019, 2020, 2021 and 2022 using the morphological characteristics of petioles and the content of L-ascorbic acid, oxalic acid and soluble solids. All accessions were evaluated in full maturity (May till July) and marketable quality. Ten petioles from several plants of the same accessions were hand-harvested by breaking out. Petioles with removed leaf blades were washed and evaluated in the laboratory. For the evaluation, the descriptor list of Turečková et al. (2001) was used. The classification consists of six morphological parameters (petiole skin colour above the base, skin colour at the base, shape of the petiole base in cross-section, surface of the petiole abaxial side, flesh colour at the petiole base and petiole thickness diameter) and three plant content analyses (L-ascorbic acid, oxalic acid and total soluble solids) (Table 3).

The concentration of L-ascorbic acid was determined by the modified high-performance liquid chromatography (HPLC) method according to Arya et al. (2000). Three parts were taken from each petiole - the basal, central and upper parts - and pooled and then homogenized. For analyses, 10 g of the mixture was mixed in the blender with 20 mL of 0.1 M oxalic acid. The homogenate was filtered and adjusted with 0.1 M oxalic acid to a volume of 100 mL. Then 20 mL of the mixture solution was centrifuged at 3500 rpm for 10 min. The supernatant was filtered through a microfilter (PVDF (polyvinylidene difluoride membrane) 0.45 μm) and used for the determination (ECOM, 1999). The analyses were performed by RP-HPLC (reversed-phase high-performance liquid chromatography) (ECOM, Czech Republic) at 254 nm using a UV-VIS (ultraviolet visible) detector on column (C18). All samples were evaluated in three technical replicates. The amount of L-ascorbic acid was expressed as mg · 100 g^-1 FW.

In a 4-year evaluation, the standard titration with the potassium permanganate method as described by Karamad et al. (2019) was used. In 2022, a more precise method to quantify oxalic acid content was applied. The content was determined by a modified HPLC method according to Rahman et al. (2007). The petioles were cut, dried at 70°C in a forced-air dryer for 48 hr and ground into a powder. The sample (0.5 g) was extracted by 15 mL of 1 M hydrochloric acid solution. The sample suspension was heated in a bath of boiling water for 18 min. After cooling, the mixture was filtered, rinsed with distilled water and adjusted to 50 mL. The filtrate of the acid extract was adjusted to pH 3.0 using a 5 M sodium hydroxide solution. The filtrates were further filtered through a syringe filter with a 0.45 μm hydrophilic membrane before analysis by RP-HPLC (ECOM, Czech Republic). A 10 μL sample was injected at a column (C18) using a mobile phase of 15 mM NaH₂PO₄ (pH 2.7), a flow rate of 0.5 mL · min^-1 and a wavelength of 210 nm. Measurement of each accession was performed in three technical replicates. The amount of oxalic acid was expressed as mg · 100 g^-1 FW.

The content of total soluble solids was evaluated according to Zbíral (2005). The samples were sliced and pressed, and the extracted juice was dropped on the measuring glass plate of the refractometer. Measurements were then conducted on the digital optical instrument HI 96801 (Hanna Instruments Ltd., USA) measuring refractive index, which was converted to the sucrose sugar concentration (%) in water suspensions. For each sample, three technical replicates (measurements) were performed.

In 2022, a detailed observation of the development of L-ascorbic and oxalic acid content and soluble solids content in rhubarb petioles was done. The evaluation was performed at six-time points: May 3^rd, May 17^th, May 31^st, June 14^th, June 28^th and July 12^th, 2022.

Based on the morphological diversity of the studied R. rhabarbarum accessions, a genetic analysis based on ISSR (inter simple sequence repeats) reaction was performed. Total genomic DNA was extracted from young frozen leaves (0.1 g) using the DNeasy Plant Mini Kit (Qiagen, Germany). The DNA quality was estimated using electrophoresis on a 0.8% agarose gel compared with lambda DNA standards containing 200 ng · μL^-1 and 400 ng · μL^-1 DNA. The concentration was measured by the Modulus Single Tube Multimode Reader (Turner BioSystems, USA) and adjusted to the final concentration of 12.5 ng · μL^-1.

Initially, 17 ISSR primers (Tabin et al., 2016) were screened for polymorphism and finally 15 polymorphic primers (Table 4) were selected for generating ISSR profiles of the Rheum samples. The reaction mixture consisted of 1.5 μL DNA, 4 μL 5 × GoTaq Flexi Buffer, 0.5 μL dNTPs (10 mM), 1.2 μL MgCl₂ (25 mM), 1.5 μL primer (10 μM), 0.7 μL 5 × Green GoTaq Flexi Buffer, 0.2 μL GoTaq G2 Flexi DNA polymerase (5 U · μL^-1, Promega, USA) and nuclease-free water to make a final volume of 25 μL. The PCR (polymerase chain reaction) cycles were: one activation cycle at 94°C for 5 min, followed by 45 cycles at 94°C for 1 min, annealing temperature for each primer given in Table 4 for 1 min and extension at 72°C for 2 min. The final extension cycle was carried out at 72°C for 7 min. PCR products were separated on a 1.5% agarose gel stained with Midori green Advance dye (Nippon Genetics, Japan) and the size of each fragment was estimated by comparison with a 100 bp DNA ladder (New England BioLabs, United Kingdom) and a TrackIt 1 Kb Plus DNA Ladder (Invitrogen, USA).

Morphological data analysis was performed based on the six leaf-petiole quality characters obtained for the same input material as for the L-ascorbic and oxalic acid content. The representative values of individual descriptors for each accession were calculated using the 4-year dataset. These data were subjected to distance-based clustering analysis using the unweighted pair-group method with arithmetic averages (UPGMA) and Jaccard’s similarity coefficient by the FreeTree v.0.9.1.50 software (Hampl et al., 2001). The phylogenetic tree was visualized by FigTree v1.4.4 (Rambaut, 2010) software.

Data of L-ascorbic and oxalic acid content were transformed by the decimal logarithm and analysed by repeated measures analysis of variance (ANOVA) and Tukey’s multiple comparison tests with an α level of 0.05 by the Statistica v.12.0.0 software (StatSoft, USA). The obtained results were examined by residual analysis using the same software. The putative role of L-ascorbic acid as the oxalate precursor in R. rhabarbarum accessions was examined by the correlation of L-ascorbic and oxalic acid measured in individual sampling terms using the regression analysis of Statistica v.12.0.0 software with an α level of 0.05. For ordination analysis to evaluate the relationship between accessions and sampling terms and the correlation of L-ascorbic acid and oxalic acid content, the program Canoco 5 (Biometris, Wageningen University and Research Centre, Wageningen, The Netherlands; University of South Bohemia in České Budějovice, České Budějovice, Czech Republic) was used.

For genetic analysis, each amplified fragment of the ISSR reaction was scored manually and converted into a binary matrix based on the presence (1) and absence (0) of the respective band. Subsequently, the data were used to construct a difference/similarity matrix based on UPGMA and Jaccard’s similarity coefficient and to construct a phylogenetic tree using FreeTree v. 0.9.1.50 (Hampl et al., 2001) and FigTree v. 1.4.4 (Rambaut, 2010) software. Based on the taxonomical relationships, the accession 42H7500012 (R. palmatum × wittrockii) was used as the outgroup.

The gradient for the response data in the ordination analysis was 0.6 units long, and redundancy analysis (RDA) was chosen as a statistical method. The statistical significance of the results was calculated with the Monte-Carlo permutation test (499 permutations). Results are significant at a significance level of p = 0.002.

Four-year observation of six morphological characters of petioles and the content of L-ascorbic acid, oxalic acid and soluble solids in 16 accessions of Rheum genetic resources brought the following results. The R. palmatum × wittrockii petioles (42H7500012) were characterized by red skin on the base to green skin above the base, a semi-circular shape in the cross-section, a ribbed abaxial surface, green flesh and a petiole diameter up to 20 mm (Table 5). R. rhaponticum petioles (42H7500014) are distinguished from the Rheum hybrid accession by the green to rose flesh and green colour of the skin on both evaluated parts (at and above the base). The rest of the petiole traits were identical. For accession 42H7500020 (R. rhaponticum), the affiliation with the R. rhabarbarum species was presumed based on the morphological analysis.

The majority of the evaluated accessions of R. rhabarbarum showed red or dark red petiole colour at the base and from green (42H7500011, 42H7500016, 42H7500025, 42H7500044) to rose (42H7500001, 42H7500003, 42H7500005, 42H7500006, 42H7500008) and red stripped (42H7500004, 42H7500023, 42H7500030, 42H7500047) colouring above the base. The cross-sections were mostly reinformed, with both smooth and ribbed shapes. The predominant colour of the flesh was green. Except for accessions 42H750001, 42H750004, 42H750006 and 42H750044, whose petioles showed a diameter from 20 mm to 25 mm, the R. rhabarbarum plants had values of petiole diameter similar to those of R. rhaponticum and R. palmatum × wittrockii species (<20 mm).

According to the average content of L-ascorbic acid, oxalic acid and soluble solids, evaluated accessions created groups independently of the groups based on taxonomy or morphological characteristics. In the case of L-ascorbic acid, eight accessions (42H7500001, 42H7500003, 42H7500004, 42H7500006, 42H7500012, 42H7500014, 42H7500016 and 42H7500020) showed values of 5–15 mg · 100 g^-1 FW, representing a low content of L-ascorbic acid. The value of L-ascorbic acid content reported for R. rhabarbarum cultivated in Latvia was 15.62–68.80 mg · 100 g^-1 FW, representing rhubarb growing areas further north of the Czech Republic (Duma et al., 2016). The rest of the accessions showed intermediate values (15–25 mg · 100 g^-1). Compared to Romania, placed further south, where L-ascorbic acid gained a content of 339–438 mg · 100 g^-1 FW (Stoleru et al., 2019); all the evaluated accessions had lower L-ascorbic acid content. In Table 5, the descriptor value representing the given content range of L-ascorbic acid presents a 4-year average (2015, 2019, 2020 and 2021).

Similarly, in Table 5, the 4-year average of oxalic acid content measured in the harvest season is expressed by a descriptor value according to Table 3. For all accessions, a very high content of oxalic acid (>450 mg · 100 g^-1 FW) was found. Significantly lower content was presented by Stoleru et al. (2019) ranging from 256 mg · 100 g^-1 to 377 mg · 100 g^-1 FW in rhubarb grown in Romania. The oxalic acid content in our study was comparable to experiments from Latvia (Duma et al., 2016) or New Zealand (Nguyen and Savage, 2020). The content of oxalic acid is strongly variable, probably depending on local climate and soil conditions, and taxonomy. Moreover, the variability in content may be affected by the selected analysis type (titration, HPLC, catalytic kinetic spectrophotometry).

For the content of soluble solids, the intermediate content (3.6%–5%) was the most prevalent. The results found in Slovakia (3.54–4.37°Bx) (Mezeyová et al., 2021) were close to our values.

The UPGMA analysis based on morphological data created a phylogenetic tree with four main clades, gathering 12 of 13 R. rhabarbarum accessions into Group I, II and IV (Figure 1). Interestingly, the species R. palmatum × wittrockii (42H7500012) and R. rhaponticum (42H7500014) created Group III together with R. rhabarbarum ‘Victoria’ (42H7500003) and R. rhabarbarum (42H7500008). However, the cultivar ‘Victoria’ is referred to as the R. rhaponticum species by several authors (Walkey and Matthews, 1979; Noonan and Savage, 1999; Zhao et al., 2009) and its categorization into Group III based on visual similarities was possible to expect.

Figure 1.

The first clade (Group I) linked R. rhabarbarum accessions 42H7500004 (‘Dawes Chalenge’), 42H7500005 (‘The Sutton’) and 42H7500023 to R. rhaponticum 42H7500020 and R. rhabarbarum 42H7500006 (‘Holsteiner Blut’) as the least similar ones. Accessions with the closest relationship shared the same values for four of the six evaluated traits. The second and fourth clade (Group II, Group IV) consisted of the other seven R. rhabarbarum accessions, 42H7500001 (‘Jara’) and 42H7500044 (both Group II) and 42H7500011, 42H7500016 (‘Timperley Early’), 42H7500025 (‘Glaskin Perpetual’), 42H7500030 (‘Krupnochereshkovyj’) and 42H7500047 (Group IV). Group IV included the most homogenous subclades, where five of the six characteristics were identical. Nevertheless, the common characteristics of the whole Group IV were only the shape of the petiole cross-section and the petiole diameter.

To evaluate the development of L-ascorbic and oxalic acid, the content of those acids was evaluated in six sampling terms during the year 2022: May 3^rd, May 17^th, May 31^st, June 14^th, June 28^th and July 12^th, covering the common rhubarb harvesting period in South Moravia (Czech Republic). Regarding the climate conditions, values of the average daily temperature and the sum of sunshine did not differ from average values recorded in the last 6 years (2015–2021, Table 2). A different value was determined for the precipitation, where the long-term precipitation in May (56.3 mm) differed from 2022 by almost 30 mm (26.4 mm in 2022). Contrary to the June long-term value of 46.8 mm, the precipitation in 2022 was markedly higher (76.9 mm) (CHMI, 2022). The lowest sum of precipitation was measured in the 2 weeks before the third sampling term (9.5 mm) and the highest before the fourth term (49.7 mm). Some accessions showed a significant increase in L-ascorbic acid content between those two terms. The content of L-ascorbic acid increased by around 50% or more in R. rhabarbarum accessions 42H7500003, 42H7500005, 42H7500006 and 42H7500030. The highest daily average temperature and maximum daily temperature were recorded in the 2 weeks before the fourth sampling term (21.16°C and 28.24°C, respectively) (Supplementary Table 1), which could contribute to higher L-ascorbic acid content. However, most of the accessions showed only a slight increase and, in some cases, a significant decrease was recorded (42H7500020). The results of Duma et al. (2016) also suggest a low influence of the sudden precipitation, as the presented value of L-ascorbic acid obtained for ‘Victoria’ in May was 15.62 mg · 100 g^-1 FW. The same cultivar in our study (42H7500003) showed the value of 11.98 mg · 100 g^-1 FW. Generally, studies assessing temperature effects on L-ascorbic acid content in other agricultural crops (e.g. potatoes or tomatoes) suggest that higher temperature growth conditions might increase the L-ascorbic content in plants (Hamouz et al., 2018; Hernández et al., 2018). Nevertheless, not many studies investigating the effects of climate conditions on rhubarb have been conducted yet.

The content of L-ascorbic acid developed as follows: in most (10 of 16 accessions), the increase of L-ascorbic acid between the first and second sampling terms was recorded, with the most significant increase in the R. rhabarbarum accessions 42H7500004, 42H7500006 and 42H7500023. Next, the decreasing tendency of L-ascorbic acid in the third term, followed by an increase in the fourth term, was detected. Similarly, the content of L-ascorbic acid decreased in the fifth sampling term in most accessions. R. rhabarbarum accessions 42H7500016, 42H7500020 and 42H7500047 and R. rhaponticum 42H7500014 did not show such a decrease in the fifth sampling term. During the sixth sampling term, the content of L-ascorbic acid increased in all accessions except R. rhabarbarum accession 42H7500047 (Table 6 and Supplementary Graph 1). The development of L-ascorbic acid in Rheum species was monitored only by Duma et al. (2016) where the increase of L-ascorbic acid from May to June, followed by a little decrease in July, was reported.

The statistical analysis of the influence of taxonomy and harvesting time on the content of L-ascorbic acid showed significant differences for both factors. In the Rheum taxonomy, R. palmatum × wittrockii and R. rhaponticum significantly differ from the tested R. rhabarbarum accessions, except for accessions 42H7500004 and 42H7500044 in the case of R. palmatum × wittrockii and accessions 42H7500016 and 42H7500023 for R. rhaponticum. On the other hand, four accessions of R. rhabarbarum (42H7500005, 42H7500008, 42H7500020 and 42H7500047) significantly differed from all other accessions. The other R. rhabarbarum accessions were similar to the maximum of two other accessions, indicating individual differences within the R. rhabarbarum species in the case of L-ascorbic acid content (Supplementary Table 2). The sampling term was a significant factor in most of the evaluated accessions. The highest values of L-ascorbic acid were obtained for R. rhabarbarum 42H7500008, where the content of L-ascorbic acid was >20 mg · 100 g^-1 FW during the whole evaluated period. These results suggest that the content of L-ascorbic acid in Rheum plants depends on both taxonomy and harvesting date. Nevertheless, the development of L-ascorbic acid during the evaluated period showed similar trends.

The content of oxalic acid in the Rheum accessions was significantly influenced by the taxonomy and harvesting time. The accessions of R. rhaponticum (42H7500014) and three R. rhabarbarum (42H7500003, 42H7500016 and 42H7500044) significantly differed from all of the other accessions. Accessions 42H7500014 and 42H7500044 represented Rheum with a high content of oxalic acid during the evaluated period, whereas accessions 42H7500003 (R. rhabarbarum ‘Victoria’) and 42H7500016 (R. rhabarbarum ‘Timperley Early’) showed low values. The content of oxalic acid in the first sampling ranged from 296.99 mg · 100 g^-1 ± 0.008 mg · 100 g^-1 FW (R. rhabarbarum accession 42H7500016) to 942.77 mg · 100 g^-1 ± 0.031 mg · 100 g^-1 FW (R. rhabarbarum accession 42H7500044). For most of the accessions, the content of oxalic acid was slightly increasing in time, with the highest values on June 14th (the fourth sampling, Supplementary Graph 2). However, for 12 out of 16 (total) accessions, the decrease during the fifth sampling term (June 28th) was recorded. In four accessions; R. rhabarbarum 42H7500003, 42H7500004, R. rhaponticum 42H7500014 and R. palmatum × wittrockii 42H7500012; the content of oxalic acid increased in the fifth sampling term. The first three listed herein above belong to the same group (Group III) according to molecular data from 15 ISSR markers, suggesting possible similarities based on genetics. The content of oxalic acid in R. rhabarbarum from organic production was observed by Stoleru et al. (2019). The presented values in this study ranged from 256 mg · 100 g^-1 to 377 mg · 100 g^-1 FW, approximately two to three times lower than in our study. The possible differences may be caused by different methodological approaches when catalytic kinetic spectrophotometry was used by Stoleru et al. (2019), while titration and HPLC (high-performance liquid chromatography) were in our study. Another factor causing this difference may be climatic and weather conditions affecting the content of oxalic acid. Several studies reported a higher content of oxalic acid as crystals of calcium oxalate (druse crystals) in plant tissues (e.g. taro, sweet potatoes) under stress conditions such as drought (Woolfe, 1992; Oscarsson and Savage, 2007; Gouveia et al., 2018).

The ordination diagram (Figure 2) shows relationships between accessions, sampling terms and the content of oxalic acid and L-ascorbic acid. The content of L-ascorbic acid was the highest at the last, sixth, sampling term. The diagram also indicates an obvious negative correlation between oxalic acid content and the first sampling term, representing the lowest content of oxalic acid at the beginning of the harvesting season. Accessions R. rhabarbarum 42H7500008 and 42H7500005 were among those with the highest content of L-ascorbic acid, and R. palmatum × wittrockii 42H7500012, R. rhabarbarum 42H7500020, 42H7500023 had the lowest ones. Regarding oxalic acid, a notable negative correlation between accession R. rhabarbarum 42H7500047 and oxalic acid content was found. In the fourth sampling term, both oxalic acid and L-ascorbic acid content increased. No correlation between oxalic acid and L-ascorbic acid was found in this analysis.

Figure 2.

The measured values of both acids obtained from May to July were evaluated from the perspective of a possible role of L-ascorbic acid as a precursor of oxalate formation in R. rhabarbarum accessions. The regression analysis performed for individual sampling terms suggested the presence of a negative correlation for accession 42H7500001 on 3^rd May and a positive correlation for accessions 42H7500006, 42H7500025 and 42H7500044 on 17^th May and for 42H7500016 on 28^th June (Supplementary Table 3). These results do not support the conversion of L-ascorbic acid to the oxalate form as was described for other oxalate-accumulating plants such as spinach, wood sorrel shamrock and begonia by Yang and Loewus (1975) or in yucca Horner et al. (2000). On the other hand, the increase of L-ascorbic acid with unchanged content of oxalic acid in low-temperature stressed spinach plants was also published (Proietti et al., 2009). However, a more comprehensive dataset would be required for the confirmation of these findings in the case of Rheum species.

The analysis based on the molecular ISSR markers clustered 15 Rheum accessions into 3 main clusters (Figure 3), where the first cluster consisted of 3 closely related groups (Groups Ia, Ib and Ic). The only groups that remained from the morphological analysis were R. rhaponticum (42H7500014) and R. rhabarbarum ‘Victoria’ (42H7500003). These accessions were clustered into Group III of the molecular-based tree with a small genetic distance from R. rhabarbarum 42H750004 (‘Dawes Chalange’) and 42H7500023. Nevertheless, the visually similar R. rhabarbarum 42H7500008 occurred in close Group II, and the hybrid species R. palmatum × wittrockii was used as the outgroup of ISSR data analysis. The rest of the studied accessions did not show similarities between morphological and genetic results. The differences in phylogenetic results based on morphology and molecular characteristics are generally presented, as well as the close similarities (Persson et al., 2000). ISSR primers from our study were used for the diversity assessment in the cases of R. emodi, R. spiciforme and R. webbianum by Tabin et al. (2016). The analysis not only clustered individual species into groups under their geographical location but also indicated differences in the genetic background of the included populations.

Figure 3.

Results of RDA ordination analysis evaluating relationships between the content of oxalic acid and L-ascorbic acid and groups based on molecular data (Figure 3) and sampling terms indicate several correlations (Figure 4). Accessions from Group II had a higher content of L-ascorbic acid at the sixth sampling term. A negative correlation between oxalic acid content and Group Ia suggests a lower content of oxalic acid in this group. Conversely, some positive correlation indicating a higher content of oxalic acid was observed in Group Ib and Group III. The results of this analysis can explain the possible effect of taxonomy in a few groups (Group 2, Group 1a) on L-ascorbic and oxalic acid content development during the harvesting season. In other groups, the correlation was weaker. The content of L-ascorbic and oxalic acid in accessions from Group 1c was probably affected by other factors not included in this analysis.

Figure 4.