Effects of shoot age on biological and chemical properties of red currant (Ribes rubrum L.) cultivars

In this study, 11 red currant cultivars, namely ‘Jonkheer van Tets’ (Jvt), ‘Junifer’ (Jun), ‘London Market’ (Lok), ‘Makosta’ (Mak), ‘Mirana’ (Mir), ‘Redpoll’ (Red), ‘Rolan’ (Rol), ‘Rondom’ (Ron), ‘Rovada’ (Rov), ‘Slovakia’ (Slo) and ‘Stanza’ (Sta), were evaluated over a 6-year period (2013–2018). The experimental fields were located in Mislođin village near Belgrade, Serbia, between 44°30′ and 44°45′ north latitudes. The altitude was between 80 and 90 m. Mislođin is located almost in the centre of the northern warm temperate belt, with a climate milder and more continental than that of typical Pannonia. The average annual temperature in this area is about 11°C, and the amount of annual rainfall is about 640 L · m⁻². Planting was done in spring 2007 with 1-year-old bushes, with distances of 1.8 m between rows and 0.8 m within the row, on sandy loam soil type with average pH 6.1. The orchards were managed according to the integral cultivation system. During the drought season, drip irrigation was used. During the vegetative phase, contact herbicides were used for weed control. At the beginning of the growing seasons, 200 kg · ha⁻¹ mineral fertiliser (N-P-K, 14-7-17) was applied. Cattle manure was applied every third year (10.0 kg · m⁻²). The experiment was organised randomly with five repetitions (four bushes per repetition). On each bush, three 2-year-old and three 3-year-old shoots were left after winter pruning. Berries were hand-harvested in the months of June–July, depending on the commercial ripening time for each cultivar (90% fruit surface colour).

According to the UPOV (The International Union for the Protection of New Varieties of Plants) Code for Ribes_Rub (2012), the following phenological properties were studied: bud burst (when 10% of total buds showed burst), beginning of flowering (10% open flowers) and harvesting time (when the fruit starts to be easily removable from the plant). For all cultivars, the beginning of bud burst, the beginning of flowering and the harvesting time were recorded on 3-year-old shoots and were compared with the properties of 2-year-old shoots.

The following generative traits were evaluated: cluster weight, cluster length, number of flowers per cluster, number of berries per cluster, berry weight and fruit set (number of fruits per 100 flowers), number of clusters per shoot and bush, and yield per shoot and bush. The total weights of the clusters and the berries were measured and expressed in grams, whereas cluster length was measured with a ruler and expressed in centimetres. Measurements included both types of shoots.

A new generative parameter for red currant, the ‘Yield efficiency’ (YE) of the cultivar, is defined. This parameter shows the number of clusters required for having a 1-kg yield per bush. The equation for calculating the YE is given below: YE=1000 g/(BW×NFC×FS){\rm{YE}} = 1000\,{\rm{g}}/({\rm{BW}} \times {\rm{NFC}} \times {\rm{FS}}) where BW is the berry weight, NFC is the number of flowers per cluster and FS is the fruit set.

After harvesting, the berries were stored at 3°C and analysed within 24 h. In total, 500 g of undamaged berries were selected and manually crushed to obtain juices. For analyses, the juices were centrifuged at 14,000 × g for 20 min. The total soluble solid (TSS) content of the berries was determined by using a refractometer (pocket PAL-1; Atago, Kyoto, Japan). Titratable acidity (TA) was determined by titrating the berry juice with 0.1 N NaOH up to pH 7.0. Acidity was expressed as the percentage of malic acid. Total sugars (TS), invert sugar (IS) and sucrose (SUC) levels were determined by the Luff-Schoorl method in percentage. Vitamin C was quantified using the reflectometer set of Merck Co. (Merck RQflex, Darmstadt, Germany). Fruit sample (5 g) and 20 mL oxalic acid (1%) were mixed, homogenised for 1 min and filtered. Polyvinylpolypyrrolidone (PVPP) (500 g) was added to 10 mL of the filtered sample to remove phenols, and 6–7 drops of H₂SO₄ (25%) were added, to reduce the pH level to <1.14. Results were expressed as milligrams of ascorbic acid per 100 g fresh weight (FW). The primary metabolites were analysed in the whole fruit without seeds. For each cultivar, three repetitions per sampling date were carried out.

The total content of phenols was estimated by the Folin–Ciocalteu method with slight modifications (Waterman and Mole, 1994). Berries (10 g) were subjected to extraction in an ultrasonic bath with MeOH for 30 min and then filtered. The extracts (200 μL) were then added to 1 mL of 1:10 diluted Folin–Ciocalteu reagent. After 4 min, 800 ml of sodium carbonate (75 g · L⁻¹) was added. After 2 h of incubation at room temperature, the absorbance at 765 nm was measured. Gallic acid (0–100 mg · L⁻¹) was used for calibration of a standard curve. The results were expressed as milligrams of gallic acid equivalents (GAE) per 100 g FW (mg GAE · 100 g⁻¹ FW).

Total anthocyanin content was investigated according to the procedure described in European Pharmacopoeia 6.0 (2008) with slight modifications. Fresh berries (50 g) were crushed extemporaneously. To about 10 g of crushed, accurately weighed berries or juice (10 g equivalent of fresh berry), methanol (95 mL) was added and mechanically stirred for 30 min, then filtered into a 100 mL volumetric flask. The filtrate was rinsed and diluted to 100 mL with methanol. A 50-fold dilution of this solution in a 0.1% v/v solution of hydrochloric acid in methanol was prepared. The absorbance of the solution was measured at 528 nm, using a 0.1% v/v solution of hydrochloric acid in methanol as the compensation liquid. The percentage content of anthocyanins, expressed as cyanidin-3-glucoside chloride, was calculated using the following expression: Anthocyanin (%)=(A×5000)/(718×m){\rm{Anthocyanin}}\,(\% ) = (A \times 5000)/(718 \times m) where A = absorbance at 528 nm; 718 = specific absorbance of cyanidin-3-glucoside chloride at 528 nm; and m = mass of the substance to be examined [in grams]).

The analysis was performed according to the procedure described by Nyman and Kumpulainen (2001), with slight modifications. About 5 g of crushed and accurately weighed berries was extracted with 40 mL of a mixture of 50 mL MeOH, 33 mL H₂O and 17 mL 37% HCl using an ultrasonic bath for 20 min. The extract was filtered through 0.45 mm filters into polypropylene vials and hydrolysed for 60 min in a water bath at 100°C. After quick cooling, the samples were injected into a high-performance liquid chromatography (HPLC) column. Analyses were carried out on Agilent series 1200 RR system equipped with a diode array detector, on a reversed-phase Lichrospher RP-18 analytical column (250 × 4 mm inner diameter; particle size 5 mm; Agilent). The mobile phases were A (H₂O containing 10% HCOOH) and B (MeCN). The extracts were separated by gradient elution according to the following scheme: 1% B 0–0.5 min; 1–7% B 0.5–1 min; 7% B 1–4 min; 7–10% B 4–7.5 min; 10–14% B 7.5–11.5 min; 14% B 11.5–15.5 min; 14–18% B 15.5–18.5 min; 18% B 18.5–22 min. The flow was adjusted to 1 mL × min⁻¹ and detection was carried out at 290, 350 and 520 nm. Quantification was done using the calibration curves of authentic standards (Fluka, Germany).

All data on the phenological properties of the red currant cultivars are presented as mean values ± standard deviations. The physical attributes of the cluster are presented as the means of five repetitions (20 clusters per repetition) and, for chemical traits, as the means of triplicate determination. The generative properties of the cultivars are presented as the mean of five repetitions (four bushes per repetition). Analysis using a one-way analysis of variance (ANOVA) was done with STATISTICA 9.0 software package. The significant differences between means, determined at p < 0.05, were measured with LSD (the least significant difference) tests.

Principal component analysis (PCA) was applied to visualise the structure of the data and to identify important variables. PCA was carried out using a demo version of the PLS Toolbox software package for MatLab 7.8.0 (Eigenvector Research, Inc., Wenatchee, WA, USA). All data were group-scaled, i.e. groups of variables were scaled to grand standard deviation, before PCA.

Junifer’ had the earliest beginning of bud (average date) burst, while ‘Rovada’ had the latest (Figure 1A and 1B). In all cultivars, earlier beginning of bud burst was recorded in 3-year-old shoots compared with 2-year-old shoots. This was probably due to the fact that old shoots possess a higher level of nitrogen and carbohydrate storage, which caused earlier start of vegetation (Palacio et al., 2018). The highest time difference in the beginning of bud burst among 3- and 2-year-old shoots was for Slovakia (6 days) and the lowest was for ‘Mirana’ (2 days). ‘Junifer’ and ‘Stanza’ had the earliest time of flowering for 3-year-old shoots and the latest was observed for ‘Rolan’ (Figure 2A and 2B). All cultivars showed delayed flowering time on 2-year-old shoots compared to 3-year-old shoots. ‘Rolan’, ‘Slovakia’ and ‘Jonkheer van Tets’ did not have flower buds on 2-year-old shoots at all. At harvesting, the cultivars were divided into three groups: early (‘Junifer’, ‘Jonkheer van Tets’ and ‘Stanza’), which ripened from June 9 to 20; medium (‘Rondom’, ‘London Market’, ‘Rolan’ and ‘Makosta’), which ripened in the period June 21–28; and late (‘Mirana’, ‘Slovakia’, ‘Rovada’ and ‘Redpoll’), which ripened from June 29 up to July 6. Similar results were found by Georgiev et al. (2008), who described similar ripening times for the same cultivars. All cultivars manifested later ripening of berries on 2-year-old shoots (Figure 3A and 3B). The highest difference in the beginning of harvest between 3- and 2-year-old shoots was observed for ‘Stanza’ (6 days) and the lowest was for ‘Mirana’ and ‘Redpoll’ (2 days). Furthermore, cultivars with the highest number of clusters per plant had the largest difference in harvesting times when comparing shoots with different ages. This could be either genotype dependent or due to the vigour and/or the number of clusters per bush.

Figure 1

Figure 2

Physical attributes of clusters and berries were significantly dependent on shoots’ age and cultivars (Table 1). On 2-year-old shoots, the cluster weight ranged between 5.4 g (‘Makosta’) and 13.8 g (‘Redpoll’). Cluster weight ranged between 3.6 g (‘London Market’) and 13.5 g (‘Slovakia’) on 3-year-old shoots. However, all cultivars showed significantly higher results in terms of cluster length, the number of flowers and berries per cluster and the berry weight on 2-year-old shoots. The highest number of berries per cluster was found in ‘Redpoll’ and ‘Slovakia’ (24.6 and 21.2, respectively), while ‘Jonkheer van Tets’ and ‘Junifer’ (9.1 and 8.7, respectively) had the lowest. This finding confirmed previous reports by Giongo et al. (2008), who had evaluated several red currant cultivars and showed that ‘Redpoll’ had the highest number of berries compared to other cultivars. The number of berries per cluster was slightly lower than that published by Milivojević et al. (2012) for the same cultivars. The differences between the results could be attributed to the significantly longer life span of the plants studied herein and the different environmental conditions.

The cultivar Rondom (1.02 g) had the largest berry weight, while ‘London Market’ (0.32 g) had the lowest. All cultivars, except ‘Stanza’, had larger berries on 2-year-old shoots, compared to 3-year-old shoots. Marketable berries (diameter >10 mm) were found in ‘Junifer’, ‘Rondom’ and ‘Mirana’ on both types of shoots, in ‘Redpoll’ on 2-year-old shoots and in ‘Slovakia’, ‘Rolan’ and ‘Jonkheer van Tets’ on 3-year-old shoots. Berries from both 2- and 3-year-old shoots did not yield any marketable fruits in ‘Rovada’, ‘London market’, ‘Stanza’ and ‘Makosta’. The fruit set ranged from 66% to 91.1% on 2-year-old shoots and from 70.9% to 91.8% on 3-year-old shoots. The highest percentage of fruit set on 2-year-old shoots was in ‘Redpoll’ (91.1%), while ‘Slovakia’ (91.8%) had the highest fruit set on 3-year-old shoots. Only ‘Stanza’ and ‘Makosta’ had higher fruit sets on older shoots. Fruit set in this experiment was at the same level as reported by Giongo et al. (2008).

Figure 3

All studied cultivars had a higher number of clusters on 3-year-old shoots compared to 2-year-old shoots (Table 2). The highest number of clusters was found in ‘Junifer’ (142.3) and the lowest number of clusters was in ‘Rovada’ (53.6). The cultivar ‘Junifer’ had the highest number of clusters on 2-year-old shoots (13). The highest-yielding cultivar was ‘Slovakia’ (3.9 kg per bush) despite cluster development only on 2-year-old shoots, while the lowest yield per bush was in ‘Makosta’ (0.7 kg per bush). The highest yield on 3-year-old shoots was found in ‘Slovakia’ (1.29 kg per bush) and ‘Mirana’ (0.97 kg per bush) and the lowest was in ‘Rovada’ (0.41 kg per bush) and ‘Makosta’ (0.23 kg per bush). Cultivars London Market and Junifer were the only two cultivars with >10% of the total yield on 2-year-old shoots. Similar results were obtained by Milošević and Milošević (2018), who studied numerous red currant cultivars and showed that ‘Slovakia’ had the highest yield per bush. But generally, it is known that the yield potential can be modified by unfavourable conditions (mostly temperatures during flowering), some biotic stresses or the nutrition and carbohydrate status of the plant (Krüger et al., 2011).

The highest generative potential, expressed as the total number of clusters per bush, was observed for ‘Junifer’ (465.89), while the lowest was recorded in ‘Rovada’ (164.4). However, ‘Slovakia’ had the highest YE and ‘Makosta’ had the lowest. To achieve a yield of 1 kg berries per bush, ‘Slovakia’ needed 73.9 clusters, while ‘Makosta’ needed 267.4 clusters.

The TSS content varied among cultivars and with shoot age (Table 3). The highest average values of TSS during the period under study were for ‘Mirana’ (12.6%) and the lowest for ‘Junifer’ (9.5%) on 2- and 3-year-old shoots. The highest TSS was on older shoots in ‘Redpoll’ (12.7%) and on younger shoots in ‘Mirana’ (12.1%). Late-ripening cultivars had higher contents of TSS in berries compared to early-ripening cultivars in both types of shoots. Moreover, early-ripening cultivars had higher values of TSS in 2-year-old shoots, while all late-ripening cultivars had higher TSS in 3-year-old shoots. The highest difference in TSS was found between the shoots in cultivars Makosta (0.80%) and Redpoll (0.8%). Kaldmäa et al. (2013) noticed that both genotypes and microclimate conditions significantly influenced the TSS values and that late-ripening cultivars had higher TSS.

‘Redpoll’ (1.6%) had the highest level of total acids (TA) and ‘Rolan’ had the lowest (0.5%) level of TA. All cultivars had a higher level of TA in the berries on older shoots. A higher yield has a negative influence on bioactive compounds (Krüger et al., 2011; Woznicki et al., 2016). The average values for TA in all red currant cultivars studied herein were 1.5–2.8 folds lower compared to values reported by other researchers (Milošević and Milošević, 2018). The differences between the results could be due to the influence of environmental conditions and agricultural practices. Rubinskiene et al. (2006) and Zheng et al. (2009) proved that environmental conditions, especially temperature, humidity, number of cloudy days and precipitation, have a strong influence on the amount of TA in berries.

The cultivar Mirana had the highest content of TS (10.3%), while ‘Junifer’ had the lowest (6.4%). Several cultivars had a higher content of TS on young shoots, which have lower crop load. Milivojević et al. (2012) recorded a lower level of TS when bushes developed more clusters and berries. The highest values of sugar/acid ratio were found in ‘Stanza’ (13.2) and the lowest in ‘Redpoll’ (4.8), both found in berries on 2-year-old shoots. Sugar/acid ratio was 3–5.2 folds higher compared to those obtained by Zorenc et al. (2017), who studied the contents of primary metabolites in three currant cultivars, primarily due to the high level of TS.

A similar pattern was recorded for vitamin C values. All cultivars had a higher level of vitamin C in berries on 2-year-old shoots. Early-ripening cultivars had higher values of vitamin C compared to medium-and late-ripening cultivars. Berk et al. (2020) recorded that the content of vitamin C varied depending on the ripening stages. Redalen (1993) found that the ascorbic acid content of black currants is reduced by elevated temperature. The highest content of this vitamin was found in ‘Stanza’ (66.9 mg · 100 g⁻¹ on 2-year-old shoots) and the lowest in ‘London Market’ (24.6 mg · 100 g⁻¹ on 3-year-old shoots). A similar low value of vitamin C was recorded by Pantelidis et al. (2007) in ‘London Market’. According to Kevers et al. (2014), environmental conditions, growing systems and genotypes have strong influence on the vitamin C content in berries. Similarly, Vagiri et al. (2013) noticed that the genotype affected the total level of vitamin C the most (50%), followed by locality (17.1%) and growth conditions (9.3%). Crus-Rus et al. (2010) established that the synthesis of vitamin C depended on age, type of plant tissue and cell compartment.

Content of total phenols ranged from 101.2 mg GAE · 100 g⁻¹ FW (‘Makosta’) to 325.2 mg GAE · 100 g⁻¹ FW (‘Redpoll’) (Table 4) and was much higher compared to values noted in some studies (Gavrilova et al. 2011; Cristina et al. 2013). The highest values were found in ‘Redpoll’ (325.2 mg GAE · 100 g⁻¹) in the berries on older shoots, while ‘Makosta’ had the lowest (101.2 mg GAE · 100 g⁻¹ FW) on the same shoots. Moreover, Giongo et al. (2008) recorded significantly higher values of total phenols in berries of ‘Redpoll’ compared to other red currant cultivars. A significant variation in the total phenol values among black currant cultivars was also described by Krüger et al. (2011). Milivojević et al. (2012) recorded lower contents of total phenols in the same red currant cultivars and under the same environmental conditions. This discrepancy might be due to the different years of study and the fact that bushes were in the period of full production. In the current study, all cultivars had a higher level of total phenols in the berries on older shoots. The older part of bushes stores higher contents of carbohydrates, which influence the synthesis of phenolic compounds (Machida et al. 2019). Late-ripening cultivars had higher contents of total phenols. Earlier harvest dates of cultivars resulted in lower contents of phenolics due to a shorter maturity period (Sahamishirazi et al. 2017). Zorenc et al. (2017) found lower contents of total phenols with late harvest dates. According to Vagiri et al. (2015), phenolic content varied between harvest dates and with the position of the leaf on the shoot.

Total anthocyanin contents ranged between 12.9 and 33.2 mg · 100 g⁻¹ FW. Cultivar Redpoll (33.2 mg · 100 g⁻¹ on 3-year-old wood) had the highest value, while ‘Makosta’ had the lowest (12.9 mg · 100 g⁻¹ on 2-year-old wood). Redpoll had the highest level of anthocyanins with the smallest berry weight. Similar observation was reported in another study on black currant and blueberry cultivars by Krüger et al. (2011), who postulated that anthocyanins are mainly localised in the berry skin, and with an increase in fruit size, the surface area/volume ratio decreases together with the concentrations of anthocyanins.

The total anthocyanin contents of the berry fruits were higher compared to the results reported by other authors analysing the same cultivars (Pantelidis et al., 2007). However, Mitić et al. (2011) and Donno et al. (2013) recorded two- to three-fold-higher content of total anthocyanins in ‘Rondom’, ‘Rovada’, ‘Jonkheer van Tets’, Junifer’ and ‘Redpoll’ cultivars. These differences could be genotype dependant, due to agronomic practices or due to the use of a specific solvent for extraction (Djordjević et al., 2010; Nour et al., 2013). According to Wang et al. (2003), environmental pre-harvest temperature (especially warm night/day temperatures) influences the accumulation of anthocyanins in strawberries. In this experiment, four out of the eight studied cultivars had a higher content of total anthocyanins in the berries on 2-year-old shoots. Likewise, it was observed that early-ripening cultivars had a higher content of total anthocyanins in the berries on younger shoots compared to the content in late-ripening cultivars. The difference in the content of total anthocyanins in berries between 2- and 3-year-old shoots was lower in early-ripening cultivars compared to that in the late-ripening cultivars. Mattila et al. (2016) demonstrated a significant negative correlation between berry size and anthocyanin content of red currant cultivars.

Cyanidin and delphinidin were the most ubiquitous anthocyanin aglycones, and their contents varied among cultivars and with the ages of the shoots. All the studied cultivars (except ‘Rolan’) had higher cyanidin levels compared to delphinidin levels. Cyanidin 3-xylosylrutinoside is the most common anthocyanin aglycone in red currant berries (Stój et al., 2006; Gavrilova et al., 2011; Milojevic et al., 2012). In acidic conditions, cyanidin and delphinidin are the more stable glycosides of anthocyanidin. The red currant cultivars had a higher level of cyanidin glycosides probably due to the acidic nature of the berries (Khoo et al., 2017). However, many researchers did not find delphinidin in red currant in their studies (Milivojevic et al., 2012; Veberic et al., 2015; Jara-Palacios et al., 2019). Contents of cyanidin glycosides ranged between 4.2 (‘Rolan’) and 20.1 (‘Redpoll’) mg · 100 g⁻¹ FW, while values of delphinidin ranged between 3.2 (‘Makosta’) and 14.2 (‘Rolan’) mg · 100 g⁻¹ FW. Late-ripening cultivars had higher values of cyanidin in berries on 3-year-old shoots, while early-ripening cultivars had higher contents in berries on 2-year-old shoots. According to Turturică et al. (2015), the stability of the anthocyanin aglycones depends on the type of anthocyanin pigment, the co-pigments, weather conditions, enzymes and oxygen.

PCA was used to establish similarity/dissimilarity among the 2- and 3-year-old shoots from different red currant cultivars. PCA was applied on the data matrix of 19 samples (eight currant cultivars that bear berries on both 2- and 3-year-old wood, and three cultivars that have berries on just 3-year-old wood) × 24 traits (phenology, morphology, pomology and chemistry), while using the covariance matrix with auto-scaling. All studied traits were included except for the total number of clusters per bush, total yield per bush, and YE due to the inability to separate the results between 2- and 3-year-old wood specimens.

The obtained six-component model explained 86.89% of all the data variations. The first principal component (PC1) represents the greatest degree of variance (36.04%), while the second PC accounted for 15.13%, the third for 13.81%, the fourth for 10.75%, the fifth for 5.88%, and the sixth for 4.28% of the total variation in the data.

Figure 4A shows the projection of the investigated 2- and 3-year-old woods in the current cultivars on the PC1/PC2 plane. The existence of two groups along the PC2 axis is noticeable, with 2-year-old wood is separated from the 3-year-old wood. From the PCA loading plot, it is possible to identify the most influential variables responsible for the clustering (Figure 4B). Bud burst (BB), blooming time (BT), harvesting time (HT), cluster length (CLE), number of flowers per cluster (NFC), and number of berries per cluster (NBC) were the most important factors for the discrimination of 2-year-old wood from 3-year-old wood samples. On the other hand, samples from 3-year-old wood were grouped based on their number of clusters per shoot (NCS), TA, yield per shoot (Y) and percentage of shoot yield in total yield (SYTY).

Figure 4

PC1 was mainly attributed to blooming time (0.82), harvesting time (0.84), cluster weight (0.83), cluster length (0.85), number of flowers per cluster (0.88), number of berries per cluster (0.94), fruit set (0.89) and total anthocyanins (0.71). The summarisation of these traits in one component indicated a strong correlation among these traits reciprocally. Similar results for different black, red and white currant cultivars were obtained by Djordjevic et al. (2014), where bud burst, cluster weight, numbers of flowers and berries per cluster and the total anthocyanin content correlated among each other on PC1. PC2 is dominated by the number of clusters per shoot, yield per shoot and percentage of shoot yield in total yield, with strong factor loadings of 0.76, 0.91 and 0.82, respectively. Rakonjac et al. (2015) also found a significant correlation between yield and cluster number per shoot. The largest score of PC3 was due to TS (0.72). PC4, PC5 and PC6 did not have any strong factor loadings.