Journal of Applied Botany and Food Quality 91, 96 - 102 (2018), DOI:10.5073/JABFQ.2018.091.013 1Norwegian Institute of Bioeconomy Research, Ås, Norway 2Arctic University of Norway, UIT, Tromsø, Norway Seasonal and yearly variation of total polyphenols, total anthocyanins and ellagic acid in different clones of cloudberry (Rubus chamaemorus L.) Anne Linn Hykkerud1, Eivind Uleberg1, Espen Hansen2, Marieke Vervoort1, Jørgen Mølmann1, Inger Martinussen1 (Submitted: June 30, 2017; Accepted: March 1, 2018) * Corresponding author Summary Cloudberry (Rubus chamaemorus L.) is a wild perennial shrub growing on peatland with a circumpolar distribution. The combined berries have a high polyphenol content comprised primarily of ellagitannins. A few commercial cultivars are available, and pre- breeding trials on clonal material from different geographical origins are in progress. The objective of this study was to investigate how the content of polyphenols of four different cloudberry cultivars were affected by harvesting time and climatic variations during a 3-year- period. Plants were grown outside in plots and berries were harvested when mature. Berries were analyzed for total polyphenols and total anthocyanins by spectrophotometer. Total ellagic acid was identified and quantified using HPLC-MS after hydrolysis of the extracts. Results showed that all measured parameters; total anthocyanins, total polyphenols and ellagic acid are strongly influenced by the genetic background. Although low anthocyanin contents were present in all genotypes, they were highly affected by climatic conditions, being highest at low temperatures. However, the content of ellagic acid was less affected by environmental conditions and showed little response to changing temperatures. In conclusion, ellagitannin content was the most dominating polyphenol group observed in this study and was affected by genetics and is therefore a good breeding criterion for increased health benefit of cloudberry. Keywords: Climatic effects, Cloudberry, Ellagic acid, Gene × environment interaction, Polyphenols, Rubus chamaemorus Abbreviations EA ellagic acid, dw dryweight, GA gallic acid, GDD growing degree days, TA total anthocyanins, TP total polyphenols Introduction Cloudberry (Rubus chamaemorus L.) is a dioecious perennial rhi- zomatous plant, native to the Subarctic regions. Berries have a dis- tinct aromatic flavor and are highly valued in the Nordic countries (Thiem, 2003). In addition, the market and interest in wild berries is increasing, especially in the East Asian region (PaassilTa et al., 2009; TurTiainen and nuuTinen, 2012). Although most of the com- mercial harvesting comes from wild stands, four commercial varie- ties have been released (rapp and marTinussen, 2002). Additional- ly, there is ongoing testing of new cultivars (uleberg et al., 2011). Previous breeding strategies used the number of pistils, flowers and shoots as selection criteria (rapp, 1989). Berries are known to be rich in a variety of bioactive compounds including phytochemi- cals related to human health benefits (DuThie, 2007). Cloudberries contain a range of bioactive compounds and are especially rich in polyphenols (marTinussen et al., 2010; Jaakkola et al., 2012). The nutritional quality has become a priority for breeding and biotech- nological strategies, in order to control or to increase the content of specific health-related compounds in fruits and berries (mazzoni et al., 2015). Polyphenols have unique physical, chemical, and biological proper- ties (Manach, 2004) and polyphenol content has been shown to be highly correlated with antioxidant capacity (Dobson et al., 2012). Total polyphenols in cloudberry has been observed to be affected by genotype (uleberg et al., 2011). Ellagitannin occurrence in com- mon food is limited to a few fruit and nut species, and cloudberry is among the fruits with the highest reported levels (Koponen et al., 2007) where they are the dominating polyphenol group (Häkkinen, 1999; Koponen et al., 2007; Kylli, 2011). Ellagitannins contain ester bonds, which upon hydrolysis give ellagic acid. The gut deri- vate of ellagic acid have been linked to inhibition of prostate cancer cell growth (seeram et al., 2007) and have vascular health benefits (larrosa et al., 2010). In cloudberry, only a minor part of the total ellagic acid content is found as free ellagic acid. Most ellagitannins are bound and are released upon acid hydrolysis (Kylli, 2011). It has previously been reported that different growing locations and sea- sons affected the ellagic acid content in cloudberry (Li et al., 2016). Anthocyanins, another group of polyphenols, are sparse in cloud- berries, strongly correlated with berry color, and have higher levels at low temperatures (MarTinussen et al., 2010). Climatic conditions of high latitude areas, where cloudberries grow, differ significantly between years, as well as within season, affect- ing the phenology and quality of the berries. Light conditions vary greatly during the growing season, with midnight sun until late July and a minimum of 12 h daylight until the end of September (bliss, 1962). Effects of climatic conditions on various quality parameters, including the content of bioactive compounds, have been evaluated in previous cloudberry studies (marTinussen et al., 2010; Jaakkola et al., 2012; li et al., 2016), as well as other berries including bilberry (uleberg et al., 2012; zoraTTi et al., 2014), blueberry (connor et al., 2002) and raspberry (mazur et al., 2014). However, phyto- chemical content varies greatly among genotypes (anTonnen et al., 2005; scalzo et al., 2005; uleberg et al., 2011). By investigating the genetic and environmental effects and their in- teractions on the polyphenol levels, we will have a better insight into the regulation and variety consistency. The objective of this paper is to investigate how the content of total phenols, total anthocyanins and levels of ellagic acid of four different clones of cloudberry are affected through season and between years. Materials and methods Plant material In 2012, 2013 and 2014, cloudberries from four different clones (‘Fjordgull’, ‘Fjellgull’, ‘102’ and ‘306’) were harvested at Holt in Tromsø 69°39’N 18°57’E. The different clones were grown in two open outdoor benches in separated squares (2 × 2 m), and every clone was represented at random positions in the bench. The two replicated benches consisted of 6 plots, including the female clones ‘Fjordgull’, Seasonal and yearly variation of secondary metabolites in cloudberry 97 ‘Fjellgull’, ‘102’ and ‘306’, while the last two plots in each bench con- sisted of a mixture of three different male clones (‘Apollen’, ‘Apolto’ and ‘H510’), to ensure sufficient pollination (see Tab. 1 for the geo- graphical origin of the clones). The clones ‘102’ and ‘306’ were in- cluded in this study based on their production performances in earlier experiments (ULEBERG et al., 2011; TROST et al., 2013), in addition to the released cultivars ‘Fjellgull’ and ‘Fjordgull’. The plots with male plants were located between the plots with female plants so every plot with a female clone was next to one of the plots with male plants. Data for flowering and harvest were registered per each 2 × 2 m plot. Individual berries were harvested when ripen, weighed and frozen at storage -80 °C and freeze dried before extraction. Depending on the year, the length of the harvest season was approximately one month (Tab. 2). For each year, an early, middle and late harvest date were selected and berries from these dates were analyzed (Tab. 2). The se- lection criteria for the selected harvesting dates was the earliest and latest date where all clones had mature berries, the middle time was selected as close as possible to the mean of these two dates. Experimental conditions Weather data information was downloaded from the weather sta- tion at NIBIO Holt, Tromsø Collected data included average daily temperature as well as daily precipitation. From these data, monthly and total precipitation and monthly and seasonal mean temperatures were extracted. Daily, monthly and aggregated growing day degrees above 5 °C were calculated and used for analysis of weather impacts on cloudberry quality (Tab. 3). The berry samples consisted of the pooled berries sampled from a cultivar a given day, never less than two berries. The samples were freeze-dried for minimum 48 h. Extraction Freeze-dried powder (100 mg) from ripe harvested cloudberries were extracted according to TrošT et al. (2013) with some modifications in 5 ml 80% methanol and vortexed in 1 min and then centrifuged. The supernatant was decanted, and then added 5 ml 80% methanol and vortexed in 1 min before centrifugation. Total polyphenols Total phenolic content was determined by the Folin-Ciocalteu method (SingelTon and Rossi, 1965) with minor modifications as described by Uleberg et al. (2011), in which gallic acid was used as a calibra- tion standard in spectrophotometric measurements. Samples (20 μL) was combined with 1.58 mL of dH2O and 100 μL Folin-Ciocalteu reagent. The mixture was incubated for 5 min at room temperature before adding 300 μL Na2CO3 (saturated sodium) carbonate solution. After a 2 h incubation at room temperature, the absorbance of the mixture was measured at 765 nm in a spectrophotometer (SmartSpec Plus, Bio-Rad, Hercules, USA). A gallic acid (GA) reference absor- bance curve was subsequently used to calculate the polyphenol con- tent in mg of GA per gram dry weight (dw) sample. Hydrolysis Methanolic extracts (2 mL) were transferred to glass tubes and subject to acid hydrolysis by addition of methanol and 37% HCl to 2.5 M at 85 °C for 6 h. Hydrolyzed extracts (1 mL) were diluted with methanol (2 mL), before UHPLC-injections (5 μL). Ellagic acid Ellagic acid was identified and quantified using UHPLC-PDA-HR- MS on a Waters Acquity UPLC (Milford, MA, USA), Waters 2996 UV-detector and Waters LCT Premiere time-of-flight MS with elec- Tab. 1: Geographical origin of the different cloudberry (Rubus chamaemorus L.) clones and varieties used in the study. Clone Origin (County) Sex ‘102’ 58°30‘N, Aust-Agder Female ‘306’ 66°30‘N, Nordland Female ‘Fjordgull’ 69°06‘N, Andøya, Nordland Female ‘Fjellgull’ 70°24‘N, Ifjordfjellet, Finnmark Female ‘Apollen’ 69°06‘N, Andøya, Nordland Male ‘Apolto’ 69°06‘N, Andøya, Nordland Male ‘H510’ 69°06‘N, Andøya, Nordland Male Tab. 2: Harvesting season during 2012-2014. First flower Last flower Flower season First harvest Last harvest Harvest period Early Middle Late 2012 15.06 28.06 13 01.08 06.09 36 15.08 20.08 28.08 2013 28.05 13.06 16 08.07 16.08 39 15.07 29.07 06.08 2014 04.06 03.07 29 23.07 14.08 22 30.07 04.08 11.08 Tab. 3: Mean monthly and growing season: air temperature, precipitation and growing degree-days (GDD) in Tromsø during 2012-2014. Month Mean temperature (°C) Precipitation (mm) Growing degree days (GDD) 2012 2013 2014 2012 2013 2014 2012 2013 2014 May 4.5 8.7 5.1 107.4 12.0 42.4 30.2 127.1 41.1 June 9.1 11.7 9.1 35.6 52.1 38.4 115.2 201.4 125.0 July 10.9 11.9 15.2 90.1 130.3 30.0 184.0 214.6 314.9 August 9.9 11.9 11.8 29.1 95.0 71.0 151.7 214.5 209.5 September 7.6 10.2 7.9 83.9 40.0 97.4 84.3 157.0 96.8 Total growing 8.4 10.9 9.8 346.1 329.4 279.2 565.4 914.6 787.3 98 A.L. Hykkerud, E. Uleberg, E. Hansen, M. Vervoort, J. Mølmann, I. Martinussen trospray ionisation. MassLynx version 4.1 (Waters) was used for instrument control and data processing. The extracts were injected on a Waters Acquity ethylene bridged hybrid (BEH) C18 column (2.1 × 100 mm, 1.7 μm) and ellagic acid was separated using a gra- dient of 5-30% acetonitrile in water (both containing 0.1% formic acid) over 6 min at a flow rate of 0.5 ml/min. The column was kept at 40 °C, and 5 μL of the extracts was injected. The samples were ion- ized with negative electrospray (ESI-), and data from 150 to 1500 Da were acquired at a scan time of 0.25 s. Capillary and cone voltages were set to -2.8 kV and -50 V, respectively, whereas source and de- solvation temperatures were set to 120 °C and 350 °C, respectively. Nitrogen was used as desolvation gas at 650 L/min. The MS was tuned to a resolution of 10,000 (FWHM) and leucine-enkapheline was infused through the reference probe for internal calibration during data acquisition. For quantification of ellagic acid, UV- absorption at 258.2 nm was used. A calibration curve for ellagic acid was made using a commercial standard (Sigma-Aldrich, Darmstadt, Germany). Ellagic acid eluted at 3.55 min and was identified as a deprotonated species ([M-H]- m/z 300.9983 calculated 300.9990), a deprotonated dimer ([2M-H]- m/z 603.0041 calculated 603.0053) and a deproto- nated trimer ([3M-H]- m/z 905.0095 calculated 905.0116). The for- mation of deprotonated multimers of ellagic acid in the ion source of the MS is concentration dependent; it is therefore difficult to quantify the compounds using MS data. Ellagic acid absorbs very well in the UV-spectrum, and as no other compounds eluting close to ellagic acid (as detected with UV and ESI-), we quantified ellagic acid by integrating the peak at 258.2 nm (± 0.5 nm). Total anthocyanins Total anthocyanins were estimated by a pH differential as described by (GiusTi and WrolsTaD, 2001; Lee et al., 2005). Each sample was diluted in a 0.025 M potassium chloride buffer, pH 1.0 and a 0.4 M sodium acetate buffer, pH 4.5. Of these mixtures the λvis-max (510 nm) and 700 nm (haze) absorbance was measured. The ab- sorbance of the diluted sample then equals (A510 nm – A700 nm) pH 1.0 – (A510 nm – A700 nm) pH 4.5 and the monomeric anthocy- anin pigment in mg/L was calculated using (A × MW × DF × 1000)/ (ε × 1) with MW = 449.2 and ε = 26,900 defining the major pigment content as cyanidin-3-glucoside. The anthocyanin content for all samples was expressed in mg cyanidin-3-glucoside per gram dw. Statistical analyses Data for total polyphenols, total anthocyanins and ellagic acid were analyzed by the GLM procedure of R (r-project) in a model that included the main effects clone/genotype, harvesting time and year and their interactions. Pearson correlation analysis (Minitab 17) was used to indicate potential impact of climate factors (temperature and precipitation) on berry quality. Furthermore, Pearson correlation analysis was carried out to show relationships between the quality parameters total polyphenols, total anthocyanins and ellagic acid. Results Plant development Weather data for the growing seasons of 2012, 2013 and 2014 are summarized (Tab. 3). Heat unit accumulation evident with Growing degree-days (GDD) is presented (Fig. 1). Results illustrated that the time for early-, middle- and late-harvest varied considerably between the three years. The coldest summer, 2012, had a mean temperature of 8.4 °C and 565.4 GDD. The warmest summer was in 2013, the second year of the experiment, with a mean temperature of 10.9 °C and 914.6 GDD. In 2014 the mean temperature was 9.8 °C and ac- cumulated GDD was 787.3 (Fig. 1, Tab. 3). Mean temperature during berry ripening, June, July and August, differed from 10.0 °C in 2012 to 11.8 °C in 2013 and 12.0 °C in 2014. Thus, the highest temperature during berry ripening occurred in 2014 while 2013 was the over- all warmest summer. Time for first flowering, first and last harvest- ing and time points for early, middle and late harvest are presented (Tab. 2). Berries were harvested 2-3 times per week during the harvesting seasons. As ‘306’ and ‘Fjellgull’ are earlier than ‘Fjordgull’ and par- ticularly ‘102’, this implies that ‘306’ and ‘Fjellgull’ produced ripe berries before early harvest and ‘102’ produced ripe berries after late harvest time point. Generally, late-season harvesting involves lower light intensity, shorter day length and lower temperature; compared with early and mid-season especially in 2013 (Fig. 2, Tab. 2). Total polyphenols Total polyphenol content was significantly affected by genotype, year and harvest time (Tab. 4 and 5). In the statistical model, 17% of the variation of total polyphenols were assigned to genotype (Tab. 4). ‘Fjellgull’ (22.99 mg/GA/g dw) was the cultivar with the highest level of total polyphenols while ‘102’ (21.67 mg/GA/g dw) had the second highest level (Fig. 2, Tab. 5). Berries from cultivar ‘306’ (19.18 mg/GA/g dw) had the lowest levels of total polyphenols all years, but not statistically different from ‘Fjordgull’. Additional- ly, approximately 17% of the variation was associated with yearly variation in weather conditions (Tab. 4). Year 2012, with the lowest mean temperature, had significantly higher levels of total polyphe- nols than 2013 and 2014 (Fig. 1, Tab. 5). All cultivars showed a re- duction in levels of total phenols during the three years period of experiment, but the reduction from 2013 to 2014 was not significant (Fig. 2). The yield and mean berry sizes of the plots were higher in 2013 compared to 2012 and 2014 (data not shown). The berries that matured early had the highest levels of total polyphenols (Fig. 2, Tab. 5) and berries from early harvest had significant higher levels of polyphenols than berries harvested late in the harvesting season. Berries harvested in the mid-season were not statistically different from early and late harvested berries. There were no sig- nificant interactions between the main effects for total polyphenols. The contents of total polyphenols were negatively correlated with temperature (Tab. 6). Fig. 1: Development of Growing Degree Days (GDD) through the growing season. Time points for early harvest, middle harvest and late har- vest for the harvesting years 2012, 2013 and 2014 are shown. Seasonal and yearly variation of secondary metabolites in cloudberry 99 Ellagic acid Ellagic acid content after hydrolysis varied from 8.05 mg/g dw in ‘Fjellgull’ to 5.44 mg/g dw in ‘Fjordgull’ (Fig. 2, Tab. 4). Ellagic acid content was significantly affected by genotype (32% of the variation) (Tab. 5). ‘Fjordgull’, which contained the highest levels of total an- thocyanins, had the lowest levels of ellagic acid at all harvest points for all three years. There were also significant effects of year and har- vest time and interactions between harvesting time and year as well as between cultivar and harvesting time (Tab. 4). Ellagic acid levels were significantly higher in berries collected in early and middle sea- son compared to late season (Fig. 2, Tab. 4). The cultivars varied in response to ellagic acid content throughout the harvest season during the study, indicating that other factors than harvest time may have in- fluenced the seasonal variations. The levels were significantly higher in the coolest year, 2012, than in the warmest year, 2014, while 2013 was not statistically different from the other years (Fig. 2, Tab. 4). Still, ellagic acid content showed a small negative correlation with temperature (Tab. 5). Total anthocyanins The highest content of total anthocyanins was observed in berries harvested in 2012, the year with lowest GDD, Additionally, berries harvested early in the 2012 season had the highest level of total an- thocyanins. As much as 45% of the variation was linked to the year and harvest time (Tab. 5). The contents of anthocyanins were clearly negative correlated with temperature (Tab. 6). Levels of anthocya- nins in 2012 were 70% higher than 2013 and 63% higher than in 2014 (Fig. 2, Tab. 4). All clones had the highest levels in 2012, but the lowest levels for ‘102’ were measured in 2014 while the other clones had the lowest amounts of total anthocyanins in 2013 (Fig. 2, Tab. 4). Additionally, there was an interaction between year and harvesting time (Tab. 5). In 2013 the anthocyanin levels decreased through the harvesting season from early to middle and late. In 2013, levels were highest in the early harvest and in alignment middle and late harvest, while in 2014 the levels were high and leveled in early and middle harvest and lowest in late harvest. Thus, the middle harvest levels differed while the high contents in early season harvesting and low contents in late harvesting were evident all three years. In addition to the environmental conditions, there was also a significant effect of genotype and an interaction between genotype and year. ‘Fjordgull’ and ‘Fjellgull’ had significant higher contents than ‘102’ and ‘306’. Discussion Influence of genotype The present study confirms that genotype significantly influences the content of total polyphenols, ellagitannins and total anthocya- nins in cloudberries. This is in alignment with previous studies on Fig. 2: Yearly (from 2012-2014) and harvesting time variation in the con- tent of a) total anthocyanins (TA) mg/g dry weight (dw) sample; b) total phenols (TP) expressed as mg of GA (gallic acid) equivalents per gram dry weight (dw) of sample and c) ellagic acid (EA) mg per gram dw of sample in cloudberry (Rubus chamaemorus L.) clones ‘102’, ‘306’ and the commercial varieties ‘Fjellgull’ and ‘Fjordgull’. The growth harvesting season are divided into ‘early’, ‘middle’ and ‘late’. Tab. 4: Effects of genotype, year and harvesting time and their interactions on the quality parameters total anthocyanin (TA), total polyphenol (TP) and ellagic acid (EA) determined in cloudberry berries. The levels are expressed as mg per g dry weight (dw) of sample, total polyphenols were expressed as mg of GA (Gallic acid) equivalents per gram dw of sample. Clone TP EA TA mg/GA/g dw mg/g dw mg/g dw Genotype ‘102’ 21.67 7.28 0.14 ‘306’ 19.18 6.83 0.14 ‘Fjellgull’ 22.99 8.05 0.21 ‘Fjordgull’ 20.01 5.44 0.24 Year 2012 23.19 7.22 0.32 2013 20.51 7.11 0.10 2014 19.57 6.54 0.12 Harvest time Early 22.12 7.41 0.24 Middle 20.97 7.03 0.19 Late 20.09 6.35 0.13 100 A.L. Hykkerud, E. Uleberg, E. Hansen, M. Vervoort, J. Mølmann, I. Martinussen 20 different cloudberry clones, including the four tested here, which revealed large variation on total polyphenol levels and total antho- cyanins (Uleberg et al., 2011). In raspberry (Rubus idaeus) stu- dies, the content of total phenolics, ellagic acid, and total anthocya- nins also varied greatly between cultivars (AnTTonen et al., 2005; BobinaiTe et al., 2012; Mazur et al., 2014). Additionally, our results showed that genotype had greatest influence on total polyphenols and ellagic acid and less for anthocyanin content, which was more affected by environment. This is in line with findings in black cur- rant (Ribes nigrum) (Vagiri et al., 2013) where the large variation in total anthocyanins was mainly affected by yearly variations and less by genotype while polyphenolics to be more affected by genotype. Likewise, the polyphenol content in blackberry was highly affected by cultivar and less by yearly variations (Milošević et al., 2012). On the contrary, another study on raspberry (Mazur et al., 2014), found small yearly variations on the anthocyanin content, however, the temperature difference between the two selected years was smaller than the yearly variations in our study. Cultivar effect on anthocya- nins have also been detected in other berry species such as cranberry (Vaccinium oxycoccos) (Borowska et al., 2009) and lowbush blue- berry (Vaccinium angustifolium) (KalT et al., 1996). These results indicate that total polyphenol content is most affected by genotype while total anthocyanins are more controlled by environmental con- ditions. Ellagitannins are the major phenolic constituent in cloudberry (Kähkönen et al., 2001; MääTTä-riihinen et al., 2004; Kylli, 2011) and given this would be an interesting qualitative trait in future plant breeding activities. In this study, there were significant effects of year and harvesting time during the season, yet genotype was the factor with the greatest impact on ellagic acid content. In a Finnish study, ellagic acid content in cloudberry was found to be affected by the location, which could be due both to genetic and environmental dif- ferences (Häkkinen et al., 1999), while Li et al. (2016) did not find a location effect in four different locations in Atlantic Canada. Our results are consistent with those of AnTTonen and KarJalainen (2005) and BobinaiTe et al. (2012) that described large variations in ellagic acid content between cultivars of raspberries and ATkinson et al. (2006) that reported a strong seasonal effect on ellagic acid con- tent in strawberry. Mazur et al. (2014) reported that genotype was more important for ellagitannin content in raspberry than seasonal variation. Additionally, genotype effect was found to be smaller in red raspberry as compared to blackberry (Vrhovsek et al., 2008), indicating differences between species. The correlation between to- tal phenolics and ellagic acid suggest that selection for a high total phenolic content will be a good indirect selection criterion for el- lagic acid content. In the present study, the differences between the genotypes were less profound than findings from raspberries but still significant for all investigated compounds. Clones ‘102’ and ‘306’ were included in this study based on their performances in earlier experiments (uleberg et al., 2011; TrosT et al., 2013), in addition to the released cultivars ‘Fjellgull’ and ‘Fjordgull’. Thus, the studied clones are already selected and we expect less variation here than we would in a study with wild material. Influence of climatic conditions Total polyphenolic and ellagic acid content were significantly affec- ted by year and harvest time. Li et al. (2016) also found a yearly ef- fect on the total polyphenol content as well as the ellagic acid content in cloudberries collected at four different locations during a two-year period in Atlantic Canada. Total phenols showed a negative correla- tion with GDD and the highest polyphenol content was observed in the year with the lowest growing temperature, while the levels were reduced the warmest summers (2013 and 2014). Conversely, ellagic acid content was less correlated to temperature, but behaved similar as total phenols through the harvesting season and between years. On the contrary, Remberg et al. (2010) reported significant higher levels of the analyzed two prevalent ellagitannins, lambertianin C and sanguiin HG, in raspberries at increasing temperature. Our re- sults indicate that although ellagitannins are dominating, there are other polyphenols more affected by temperature. In cloudberries, a previous study suggests that in addition to temperature, sunlight is the main factor affecting the chemical composition of cloudberry (Jaakkola et al., 2012). Total anthocyanin levels in the four studied cloudberry clones were strongly affected by year, time of harvesting and climatic conditions as described previously by TrôsT et al. (2013), uleberg et al. (2011) and MarTinussen et al. (2010). In the present study, anthocyanin levels were negatively correlated with GDD. The anthocyanin levels were highest in 2012, the coldest summer, while the lowest contents were found in 2013, which had the warmest summer. Findings in the present study thus confirm the results described by marTinussen et al. (2010) that found enhanced levels of anthocyanins in the cultivar ‘Fjellgull’ at 9 °C or 12 °C compared to 15 °C and 18 °C Jaakkola et al. (2012) found similar results in cloudberry from dif- ferent locations over several years; in a cold and rainy summer, the Tab. 5: Percentage of total variation (Sum of squares) ascribed to clone, harvesting time, and year and their interactions on the content of total polyphenols, ellagic acid and total anthocyanins in cloudberry grown during three different years. Total polyphenol Total anthocyanin Ellagic acid % of variations p-value % of variations p-value % of variations p-value Genotype (G) 17.1 8.3 × 10-11 10.2 6.0 × 10-9 32.3 2.0 × 10-16 Year (Y) 17.2 2.0 × 10-12 34.4 2.0 × 10-16 3.4 1.1 × 10-3 Harvest time (H) 5.7 1.2 × 10-4 10.1 1.7 × 10-9 5.7 1.3 × 10-4 G × Y 1.3 0.24 3.2 3.2 × 10-3 0.12 0.94 G × H 3.6 0.07 1.6 0.31 5.2 0.01 Y × H 0.6 0.35 1.9 0.01 0.0 0.93 Tab. 6: Correlations (Pearson’s) between total anthocyanins, total phenols, ellagic acid, mean temperature and total precipitation during the rip- ening period, and berry weight. TA TP EA Temp Precipitation TP 0.548 EA 0.131 0.629 Temp -0.502 -0.434 -0.214 Precipitation -0.24 -0.003 0.161 0.062 Berry weight 0.158 0.363 0.284 -0.141 0.301 Seasonal and yearly variation of secondary metabolites in cloudberry 101 content of anthocyanins were significantly higher than in a warm and dry summer. Higher anthocyanin levels at low temperatures have been described in various species (Chalker-scoTT, 1999), and at high temperatures for others berry species (Uleberg et al., 2012; JosuTTis et al., 2012). The regulation of anthocyanins is found to be a stress response in plants. Depending on the stress conditions pro- duction of specific anthocyanins is induced (kovinich et al., 2014). In addition to the year and temperature effects, there were also in- teractions; year by genotype and year by harvesting time (Tab. 3). The sensitivity of anthocyanin accumulation at temperature changes related to different genotypes have also been found to differ signifi- cantly in flowers of Plantago lanceolata from different genotypes (STiles et al., 2007). In our study, anthocyanin levels decreased from early to late harvesting time. The accumulated GDD (Tab. 2) could explain the reduction in anthocyanin contents that was observed from early to late harvest. In a recent study BarnuuD et al. (2014), using climate-variable-based empirical models to predict quality changes of different cultivars of grapes (Vitis vinifera), found that in a warmer climate berry anthocyanins would decline. Dark inhibition of the anthocyanin biosynthesis has been described by Zhang et al. (2015) who found that light was necessary for anthocyanin produc- tion in Begonia leaves under low temperature. Effect of other environmental factors The variation between years was significant and all genotypes showed a total polyphenolic reduction during the three years period of experiment, even if the reduction from 2013 to 2014 was not sig- nificant. The explanation could be weather conditions, as discussed above or other factors as age of the plantlets, total yield or nutritional status in the plots. Cloudberry spread vegetatively so the restricted area of each plot gradually gets more and more populated, increas- ing the competition for nutrients. Nevertheless, the yields in the plots were highest in 2013 (2421 g) while similar in 2012 and 2014 (1807 g and 1648 g, respectively). Berry yields in cloudberry are highly correlated to the weather conditions during pollination, since insect pollination is necessary for fruit development in the dioe- cious cloudberry (Rapp and MarTinussen, 2002). The high mean temperature in June 2013 (11.7 °C) compared to 2012 and 2014 (9.1 °C) indicate that the conditions for pollination was better this year, which could explain the observed yield differences. The mean berry weights were similar in 2012 (1.69 g) and 2013 (1.89 g) while significantly lower in 2014 (1.18 g). Even if the berry size was re- duced, good berry yields were obtained in 2014, indicating that nu- trients were available for fruit growth and development. Conclusions To investigate the breeding potential of cloudberry clones with high content of health beneficial compounds total anthocyanins, total polyphenols and ellagic acid content were analyzed in berries har- vested from four clones of cloudberry from a three years period. The results illustrated that although climatic effects have strong impact on total anthocyanins, the levels are highly genotype dependent. Although genotype affects the polyphenolic levels in cloudberry, fi- nal content was also dependent upon environmental parameters. In addition, results indicated that the different clones respond different- ly to environmental factors. The presented study expands our know- ledge about variation in the content of polyphenols in cloudberry clones and how they are affected by environmental conditions. This knowledge may help for the selection and validation of the clones with the highest health benefit. The strong correlation between total polyphenols and ellagic acid content of cloudberry suggest that selec- tion for a high total phenolic content will be a good indirect selection criterion for ellagic acid content. Acknowledgements The work has been funded by the Czech-Norwegian Research Program CZ09, project ID 7F14122. Thanks to Kirsten Jakobsen for valuable technical assistance. References anTTonen, m.J., karJalainen, r.o., 2005: Environmental and genetic variation of phenolic compounds in red raspberry. J. Food. Compos. Anal. 18, 759-769. DOI: 10.1016/j.jfca.2004.11.003 aTkinson, c.J., DoDDs, p.a.a., ForD, y.y., le miere, J., Taylor, J.m., blake, p.s., paul, n., 2006: Effects of cultivar, fruit number and re- flected photosynthetically active radiation on Fragaria × ananassa productivity and fruit ellagic acid and ascorbic acid concentrations. Ann. Bot. 97, 429-441. DOI: 10.1093/aob/mcj046 barnuuD, n.n., zerihun, a., gibberD, m., baTes, B., 2014: Berry compo- sition and climate: responses and empirical models. I J. Biometeorol. 58, 1207-1223. DOI: 10.1007/s00484-013-0715-2 bliss, l.c., 1962: Adaptations of arctic and alpine plants to environmental conditions. Arctic 15, 117-144. BoBinaitė, R., viškelis, P., venskutonis, P.R., 2012: Variation of total phenolics, anthocyanins, ellagic acid and radical scavenging capacity in various raspberry (Rubus spp.) cultivars. Food Chem. 132, 1495-1501. DOI: 10.1016/j.foodchem.2011.11.137 chalker-scoTT, L., 1999: Environmental significance of anthocyanins in plant stress. Responses. Photochem. Photobiol. 70, 1-9. DOI: 10.1111/j.1751-1097.1999.tb01944.x connor, a.m., luby, J.J., Tong, c.b.s., Finn, c.e., hancock, J.F., 2002: Genotypic and environmental variation in antioxidant activity, total phenolic content, and anthocyanin content among blueberry cultivars. J. Amer. Soc. Hort. Sci. 127, 89-97. Dobson, p., graham, J., sTewarT, D., brennan, r., hackeTT, c.a., mcDougall, G.J., 2012: Over-seasons analysis of quantitative trait loci affecting phenolic content and antioxidant capacity in raspberry. J. Agric. Food Chem. 53, 625-634. DOI: 10.1021/jf3005178 DuThie, S.J., 2007: Berry phytochemicals, genomic stability and cancer: Evidence for chemoprotection at several stages in the carcinogenic pro- cess. Mol. Nutr. Food Res. 51, 665-674. DOI: 10.1002/mnfr.200600257 giusTi, m.m., wrolsTaD, r.e., 2001: Characterization and measurement of anthocyanins by UV-visible spectroscopy. In: Wrolstad, R.E. (ed.), Currant protocols in Food Analytical Chemistry, 1-13. John Wiley & Sons, New York. häkkinen, s., heinonen, m., kärenlampi, s., mykkänen, h., ruus- kanen, J., Törrönen, r., 1999: Screening of selected flavonoids and phenolic acids in 19 berries. Food Res. Int. 32, 345-353. DOI: 10.1016/S0963-9969(99)00095-2 Jaakkola, m., korpelainen, v., hoppulab, k., virTanena, v., 2012: Chemical composition of ripe fruits of Rubus chamaemorus L. grown in different habitats. J. Sci. Food Agric. 92, 1324-1330. DOI: 10.1002/jsfa.4705 kovinich, n., kayanJa, g., chanoca, a., rieDl, k., oTegui, m.s., groTewolD, e., 2014: Not all anthocyanins are born equal: distinct patterns induced by stress in Arabidopsis. Planta 240, 931-940. DOI: 10.1007/s00425-014-2079-1 kähkönen, m.p., hopia, a.i., heinonen, m., 2001: Berry phenolics and their antioxidant activity. J. Agric. Food Chem. 49, 4076-4082. DOI: 10.1021/jf010152t kylli, p., 2011: Berry phenolics: isolation, analysis, identification, and anti- oxidant properties. PhD thesis. University of Helsinki. EKT-series 1502. ISBN 978-952-10-7115-7 lee, J., DursT, r.w., wrolsTaD, r.e., 2005: Determination of total mo- nomeric anthocyanin pigment content of fruit juices, beverages, natural colorants, and wines by the pH differential method: collaborative study. J. AOAC Int. 88, 1269-1278. larrosa, m., garcía-conesa, m.T., espín, J.c., Tomás-barberán, F.A., 102 A.L. Hykkerud, E. Uleberg, E. Hansen, M. Vervoort, J. Mølmann, I. Martinussen 2010: Ellagitannins, ellagic acid and vascular health. Mol. Aspects Med. 31, 513-539. DOI: 10.1016/j.mam.2010.09.005 li, J., percival, D., hoyle, J., whiTe, J., heaD, k., pruski, k., 2016: Phenology and yield of native fruits cloudberry/bakeapple (Rubus cha- maemorus L.) and lingonberry/partridgeberry (Vaccinium vitis-idaea L.) grown in southern Labrador, Canada. Can. J. Plant Sci. 96, 329-338. DOI: 10.1139/cjps-2015-0131 määTTä-riihinen, k.r., kamal-elDin, a., Törrönen, r., 2004: Iden- tification and quantification of phenolic compounds in berries of Fragaria and Rubus species (Family Rosaceae). J. Agric. Food Chem. 52, 6178-6187. DOI:10.1021/jf049450 manach, c., scalberT, a., moranD, c., rémésy, c., Jimènez, l., 2004: Polyphenols: food sources and bioavailability. Am. J. Clin. Nutr. 79, 727- 747. marTinussen, i., uleberg, e., mcDougall, g.J., sTewarT, D., JunTTila, o., 2010: Development and quality of cloudberry (Rubus chamaemorus L.) as affected by female parent, male parent and temperature. J. Berry Res. 1, 91-101. DOI: 10.3233/br-2010-010 mazur, s.p., nes, a., wolD, a.b., remberg, s.F., aaby, k., 2014: Quality and chemical composition of ten red raspberry (Rubus idaeus L.) geno- types during three harvest seasons. Food Chem. 160, 233-240. DOI: 10.1016/j.foodchem.2014.02.174 mazzoni, l., perez-lopez, p., giampieri, F., alvarez-suarez, J.m., gasparrini, m., Forbes-hernanDez, T.y., Quiles, J.l., mezeTTi, b., baTTino, m., 2015: The genetic aspects of berries: from field to health. J. Sci. Food Agric. 96, 365-371. DOI: 10.1002/jsfa.7216 Milošević, t., Milošević, n., Glišić, i., Mladenović, J., 2012: Fruit qua- lity attributes of blackberry grown under limited environmental condi- tions. Plant Soil Environ. 58, 322-327. paassilTa, m., moisio, s., Jaakola, l., häggman, h., 2009: Voice of the Nordic wild berry industry – a survey among the companies. Oulu University Press, Oulu, Finland. ISBN 978-951-42-9041-1 rapp, k., 1989: Number of pistils, an alternative criterion when selecting for high productivity in Rubus. Norweg J. Agric. Sci. 3, 29-32. rapp, k., marTinussen, i., 2002: Breeding cloudberry (Rubus chamaemo- rus L.) for commercial use. Acta Hort. 585, 159-160. DOI: 10.17660/ActaHortic.2002.585.23 remberg, s.F., sønsTeby, a., aaby, k., heiDe, o.m., 2010: Influence of post flowering temperature on fruit size and chemical composition of Glen Ample raspberry (Rubus idaeus L.). J. Agric. Food Chem. 58, 9120-9128. DOI: 10.1021/jf101736q seeram, n.p., aronson, w.J., zhang, y., henning, s.m., moro, a., lee, r.-p., sarTippour, m., harris, D.m., reTTig, m., sucharD, m.a., panTuck, a.J., bellDegrun, a., heber, D., 2007: Pomegranate ella- gitannin-derived metabolites inhibit prostate cancer growth and localize to the mouse prostate gland. J. Agric. Food Chem. 55, 7732-7737. DOI: 10.1021/jf071303g scalzo, J., poliTi, a., pellegrini, n., mezzeTTi, b., baTTino, m., 2005: Plant genotype affects total antioxidant capacity and phenolic contents in fruit. Nutrition 21, 207-213. DOI: 10.1016/j.nut.2004.03.025 singelTon, v.l., rossi, J.a., 1965: Colourmetry of total phenolics with phosphomolybdic-phosphotungstic acid reagents. Am. J. Enology Viti- cult. 16, 144-158. sTiles, e.a., cech, n.b., Dee, s.m., lacey, e.p., 2007: Temperature- sensitive anthocyanin production in flowers of Plantago lanceolata. Physiol Planta. 129, 756-765. DOI: 10.1111/j.1399-3054.2007.00855.x Thiem, b., 2003: Rubus chamaemorus L. A boreal plant rich in biologically active metabolites: A review. Biol. Lett. 40, 3-13. DOI: 10.1.1.611.4317 TrošT, k., uleberg, e., marTinussen, I., 2013, Levels of individual antho- cyanins and hydroxycinnamic derivates in selected clones of cloudberry (Rubus chamaemorus) Acta Agric. Scand. B, 63, 147-152. DOI: 10.1080/09064710.2012.733726 TurTiainen, m., nuuTinen, T., 2012: Evaluation of information on wild berry and mushroom markets in European countries. Small-scale Forestry 11, 131-145. DOI: 10.1007/s11842-011-9173-z uleberg, e., rohloFF, J., Jaakola, l., TrôsT, k., häggeman, h., marTinussen, I., 2012: Effects of temperature and photoperiod on yield and chemical composition of northern and southern clones of bilberry (Vaccinium myrtillus L.). J. Agric. Food Chem. 60, 10406-10414. DOI: 10.1021/jf302924m uleberg, e., TrôsT, k., sTavang, J.a., røThe, g., marTinussen, I., 2011: Evaluation of cloudberry (Rubus chamaemorus L.) clones for selection of high-quality varieties. Acta Agr. Scand. B-S P. 61:sup1, 18-26. DOI:10.1080/09064710.2011.599330 vagiri, m., ekholm, a., oberg, e., Johansson, e., anDersson, s.c., rumpunen, K., 2013: Phenols and ascorbic acid in black currants (Ribes nigrum L.): Variation due to genotype, location, and year. J. Agric. Food Chem. 61, 9298-9306. DOI: 10.1021/jf402891s vrhovsek, u., giongo, l., maTTivi, F., viola, r., 2008: A survey of el- lagitannin content raspberry and blackberry cultivars grown in Trentino (Italy). Eur. Food Res. Technol. 226, 817-824. DOI: 10.1007/s00217-007-0601-4 zhang, k.m., guo, m.l., he, D., wu, r.h., li, y.h., 2016: The inhibition effect and excessive carbon flux resulting from blocking anthocyanin biosynthesis under darkness in Begonia semperflorens. J. Plant Growth Reg. 35, 22-30. DOI: 10.1007/s00344-015-9503-z zoraTTi, l., sarala, m., carvalho, e., karppinen, k., marTens, s., giongo, l., hägman, h., Jaakola, L., 2014: Monochromatic light in- creases anthocyanin content during fruit development in bilberry. BMC Plant Biol. 14, 377-387. DOI: 10.1186/s12870-014-0377-1 Address of the authors: Anne Linn Hykkerud, Eivind Uleberg, Marieke Vervoort, Jørgen Mølmann, Inger Martinussen, Norwegian Institute of Bioeconomy Research, P.O. Box 115, NO-1431 Ås, Norway E-mail: anne.linn.hykkerud@nibio.no Espen Hansen, Arctic University of Norway, UIT, P.O. Box 6050 Langnes, N-9037 Tromsø, Norway © The Author(s) 2018. This is an Open Access article distributed under the terms of the Creative Commons Attribution Share-Alike License (http://creative- commons.org/licenses/by-sa/4.0/).