The trial was established after winter punning 2018. On 20th of February (BBCH 00) for each species, 30 canes were selected and pruned to a length of 40-50 cm. For each cane diameter was measured 1 cm from the base of the cane and the canes were divided into 3 categories (Table 1). On 10th of May (BBCH 55) the flowers were manually thinned, leaving only the main flowers, full blooming appeared on 25th of May (BBCH 65), further on manual thinning of fruit was done on 24th of June (BBCH 74) to set the fruit load on each individual cane uniformly. The manual thinning of the fruits was carried out in such a way that each annual shoot had 1 fruit on low vigour cane, 2 fruits on medium vigour canes and 3 fruit on high vigour canes, both on A. deliciosa and A. arguta. The total number of fruits examined for their quality was 60 per species, proportionally from different type of cane vigour.
Table 1
Distribution of canes by their vigour/diameter
Species
|
cane vigour
|
cane diameter (mm)
|
|
low
|
6.00 – 7.99
|
A. deliciosa
|
medium
|
8.00 – 9.99
|
|
high
|
10.00 – 11.99
|
|
low
|
3.00 – 4.49
|
A. arguta
|
medium
|
4.50 – 5.99
|
|
high
|
6.00 – 7.49
|
2.2. Fruit size, weight and firmness measurements
The first fruit measurements (height, diameter) were taken on 24th June, immediately after thinning of the fruit, followed by measurements in one- month interval (24th June, 23rd July, 22nd August, 22nd September).The weight of the fruits was determined on the same day as harvest, i.e. for A. arguta on 5th October (133 days after full bloom (DAFB)) and on 20th October (148 DAFB) for A. deliciosa. Fruit firmness was measured at harvest with a digital hand penetrometer (T. R. Turoni, Forli, Italia) with 2.4 mm tip, as recommended by (Huajia et al., 2016). The fruit firmness for A. deliciosa was also measured after fruits were stored for 1 month at 5 °C (on 20th November).
2.3. Sample preparation
The samples were prepared according to Mikulic-Petkovsek et al. (2007). The extract for sugars and acids was prepared from 5 g of fruit pericarp without the peel of A. deliciosa and 5 g of whole fruit of A. arguta, as the fruits of A. arguta are eaten without peeling and the fruits of A. deliciosa must be peeled before consumption. Plant material was cut into small pieces and 25 ml of double distilled water was added to the test tubes of each sample. Samples were placed on a shaker for 30 minutes on room temperature. After they were put in a centrifuge (Eppendorf Centrifuge 5810 R) for 5 minutes at 5000 rpm at 4 °C. Samples were filtered through a cellulose filter Chromafil A-20/25 produced by Macherey-Nagel (Düren, Germany), transferred to a vial and stored at –20 °C until analysed by high-performance liquid chromatography (HPLC).
Extract for phenols was prepared according to Mikulic-Petkovsek et al. (2010) in test tubes from 5g of fruit pericarp without peel of A. deliciosa and 5 g of whole fruit of A. arguta, and 10 ml of methanol with 3% formic acid was added. Test tubes were places in a cooled ultrasonic bath for 60 minutes. After the sample was centrifuged for 8 minutes at 8000 rpm, supernatant was filtered through a polyamide filter Chromafil AO-20/25 produced by Macherey-Nagel (Düren, Germany), transferred to a vial and stored at – 20 °C until the start of the analysis.
2.4. Analytical methods
Primary metabolites were analysed using a HPLC system connected to a RI plus detector (Finnigan Surveyor, Thermo, San Jose, USA) for determination of individual sugars and a diode array detector (Dionex UltiMate 3000, Thermo Scientific, Waltham, USA) for analysing individual organic acids.
Analytical conditions for primary metabolites were as described by Mikulic-Petkovsek et al. (2012a). For analyses of sugars a Rezex RCM-monosaccharide Ca+ 2 % (300 mm x 7,8 mm) HPLC column produced by (Phenomenex, Califoria, USA) was used. The column was operated at 65 °C, mobile phase was double distilled water, and the flow rate was 0,6 ml/min. 20 µl of sample was injected and analysed for 30 minutes. Sugars were determined with the help of external standards for glucose, fructose and sucrose (Fluka Chemie GmBH, Buchs, Switzerland).
A HPLC column Rezex ROA-organic acid H+ 8 % (300 mm x 7,8 mm) produced by Phenomenex was used to determine the organic acid content according to Mikulic-Petkovsek et al. (2012a). The column was operated at 65 °C, mobile phase was 4 mM sulfuric acid in double distilled water, UV- detector was set to 210 nm and the flow rate was 0,6 ml/min. 20 µl of sample was injected and analysed for 30 minutes. Organic acids were determined with the help of external standards for citric, quinic and malic acid (Fluka Chemie GmBH, Buchs, Switzerland).
Quantification was based on comparison of peak areas of samples with standard solutions. Content of individual sugars and organic acids were expressed in g/100 g of fresh weight (FW) of fruits.
UV/VIS spectrometer (Lambda Bio 20, Perkin Elmer, Waltham, USA) was used to determine the total phenolic content. Total phenolic content (TPC) of extracts was assessed by the Folin–Ciocalteau phenol reagent method (Singleton et al., 1999). For TPC a 10 ml test tube was used in which 7.9 ml of water was pipetted. After 100 µl of extract and 500 µl of Folin–Ciocalteau reagent was added. The extract was in (1:2 ratio (v/v) with MeOH). The samples were placed on a room temperature for 4 minutes, after 1.5 ml of sodium carbonate (20% w/v) and the remaining water was added so we had exactly 10 ml. The tube was closed, shaked and put in the oven on 40 °C for 30 minutes. A blind sample was prepared as well, which had 100 µl of methanol instead of the extract. The absorbance was measured at 765 nm, for each sample we had 3-repetitions. Total phenolic content of kiwi fruit was expressed as gallic acid equivalents (GAE) in mg/100 g FW of fruit.
2.5. Statistical analysis
Data was arranged in Microsoft Excel (Microsoft, Redmont, USA) and statistical analysed with R commander. For the determination of statistical differences between fruit, grown on canes with different vigour, one-way variance analysis (ANOVA) was used. The statistical differences between mean at 95 % confidence level were calculated. All data in tables shows average values with standard errors and statistical differences, marked with different letters.