Effects of genotype and harvest year on phytochemical and fruit quality properties of Turkish fig genotypes

This study was conducted over three harvest years to determine effects of Turkish fig genotypes grown in the east ern Mediterranean region of Turkey on phytochemical and fruit quality characters. Fruit quality characters such as fruit weight, fruit width, fruit neck length, total soluble solids (TSS), pH, acidity, and TSS/acidity were examined. Total phenolics (TP), total anthocyanins (TA), antioxidant capacity (TAC), fructose (FRUC), glucose (GLUC), sucrose (SUC), and variables describing fruit skin and flesh colors (L*, a*, C , and hº ) were also determined. Analysis of the data obtained from 12 fig genotypes, and three harvest years demonstrated a highly significant influence of genotype on phytochemical and fruit quality characters. ‘Bursa Siyahı’, which has dark black fruit skin, had the highest levels of TA (113.6 μg cy-3-rutinoside g –1 fw), TP (105.2 mg GAE/100 g fw), and TAC (10.9 mmol Fe 2+ kg –1 fw). Three yearly averages of fruit weight ranged from 22.8 g (‘Sarı İncir’) to 57.5 g (‘Bursa Siyahı’), and ostiole width ranged from 0.9 mm (‘Bursa Siyahı’) to 3.7 mm (31-IN-16). These results indicate that TP ( r = 0.77) contents correlated moderately to TAC, more so than with TA ( r = 0.56). Both FRUC and SUC were negatively correlated with TA ( r = –0.34 and r = –0.42, respectively). These data demonstrate that genotype is the main influence on the phytochemical and fruit quality characters of figs.

Palabras clave adicionales: año de la cosecha; capacidad antioxidante; caracteres frutales; color del fruto; Ficus carica; genotipo. Influence of genotype and harvest year on phytochemical and fruit quality of figs (2011) reported that considerable variation exists in certain phytochemical and antioxidant properties of fig accessions grouped according to fruit skin color groups. Some purple-and black-skinned fig fruits contained 2-fold higher TAC, 15-fold higher TA and 2.5-fold higher TP than the green-and yellow-skinned accessions. However, none of these earlier reports examined the effect of harvest year on these characters.
Turkey is the world's largest producer of dried as well as fresh figs about 21% of the world's annual production. In recent years, the production of fresh figs for export has rapidly increased, mainly in the Mediterranean region of Turkey. However, phytochemical and fruit quality characteristics of important Turkish fig genotypes for fresh consumption have not been investigated in detail. In this work, the influences of year and climatic variables on phytochemical and fruit quality characters were investigated by comparing these characters for each genotype across three years. This information could be useful for the future improvement of phytochemical and fruit quality in fig genotypes.

Plant material
This study was carried out by the Agriculture Faculty at the Mustafa Kemal University, Dörtyol Experimental Farm (36°54'N, 36°13'E, 198 m elevation) in Hatay, Turkey, during 2008. This study was designed using eight cultivars ('Bursa Siyahı', 'Göklop', 'Morgüz', 'Sarılop', 'Sarı Zeybek', 'Sarı İncir', 'Yediveren' and 'Yeşilgüz'), and four genotypes , which are promising genotypes originated from the eastern Mediterranean Region of Turkey (Ozkaya, 1997). The cultivars 'Bursa Siyahı' and 'Yediveren' are very popular as fresh figs in Turkey. The remaining fig cultivars, except for 'Sarılop', 'Sarı Zeybek' and the unimproved genotypes, are also used for fresh market consumption. The Mediterranean region where this study was conducted research is not suitable for production of dried figs due to weather conditions such as high relative humidity. Therefore, fig genotypes were

Introduction
The fig, Ficus carica L., has been of great importance as a source of human food ever since its earliest cultivation as a fruit tree. Seventy percent of the world's figs are produced in the Mediterranean countries, where figs are an important part of the Mediterranean diet, that is thought to be related to health and longevity (Trichopoulou et al., 2006).
In the last decade, there has been refreshed interest in studying and quantifying the phenolic metabolites of fruits and vegetables due to their health-promoting properties. Antioxidant compounds, such as phenolics, organic acids, vitamin E, and carotenoids scavenge free radicals, thus inhibiting the oxidative mechanisms that may lead to degenerative illnesses (Silva et al., 2004). Anthocyanins and other phenolic compounds are responsible for many health benefits (Duthie et al., 2000). Anthocyanins have been identified as key contributors to antioxidant activity in vitro and in vivo (Wang et al., 1997). Phenolic compounds are common plant secondary metabolites with physiological functions in plants and favorable effects on human health, due to their antioxidant activities. Phenolics may serve this purpose by reducing or donating hydrogen to other compounds, scavenging free radicals, and quenching singlet oxygen (Costa et al., 2009).
A great benefit of the Mediterranean diet is its high level of natural antioxidants derived from fruits, including figs, and vegetables, which contribute antioxidant vitamins (Solomon et al., 2006). The green, yellow, brown, purple and the black colors in figs originate from carotenoid and anthocyanin pigments produced in the fruits during maturation. There have now been more than 50 metabolites identified in fig fruit. The consumption of these health-promoting compounds in figs may provide protection against several human diseases (Oliveira et al., 2010). Figs also are an excellent source of fiber, minerals, and polyphenols. They are low in sodium and have no fat or cholesterol (Vinson, 1999). Previous studies published on fig have examined phytochemical components such as phenolics (Solomon et al., 2006;Oliveira et al., 2009), anthocyanins (Solomon et al., 2006;Dueñas et al., 2008), sugars (Melgarejo et al., 2003;Çalişkan & Polat, 2011), and antioxidant capacity (Solomon et al., 2006;Çalişkan & Polat, 2011). Solomon et al. (2006) showed that the higher the polyphenol content, particularly anthocyanins, in fig fruit, the higher their antioxidant activity. Çalişkan & Polat According to climate data, the minimum temperatures were below 0°C in January during the experimental years in Dörtyol, Hatay (Fig. 1). The mean temperatures and sunshine duration in harvesting periods, August to September, were higher in 2010 than in 2008 and 2009. Also, no precipitation was recorded in August, 2010.

Fruit extraction
Fig fruits from each variety were harvested at their fully mature stage in triplicate. Then, 500 g of fruit from each genotype within each replicate were homogenized in a blender at room temperature. The homogenate was centrifuged at 10,000 rpm for 10 min at 22°C. The supernatant was removed and stored frozen at -22°C until analysis for phytochemical properties. The phytochemical characters measured included total anthocyanins, total phenolics, and antioxidant capacity following a single extraction procedure as described by Beccaro et al. (2006).

Determination of total phenolics
Total phenolics were determined according to the Folin-Ciocalteu reagent method (Slinkard & Singleton, 1977). In this method, each extract was mixed with Folin-Ciocalteu phenol reagent and water 1:12 (v/v) and incubated for 8 min at room temperature, followed by the addition of 10 mL of 15% (w/v) sodium carbonate, and was allowed to stand for 2 h at room temperature. The absorbance of each sample was measured at 750 nm in a spectrophotometer (Shimadzu UV-1208, Japan). Gallic acid was used as a standard. Results are expressed as mg of gallic acid equivalents (GAE/100 g fresh fruit weight (fw)).

Determination of total anthocyanins
Total anthocyanins were calculated according to the pH differential method (Cheng & Breen, 1991). Absorbance was measured at 520 and 700 nm in buffers at pH 1.0 and pH 4.5 where Results were expressed as cyanidin-3-rutinoside, molar extinction coefficient of 28.8 and molecular weight of 595.2, µg equivalents g -1 fresh weight of fruit.

Sugar analyses
Analyses of sugars were performed according to the method described by Camara et al. (1996). Fig fruit homogenates (10 g) were diluted with distilled water (40 mL) to prepare solution for detection of sugar composition, and were centrifuged at 10,000 rpm for 10 min and the supernatants were then removed and filtered through Whatman No. 42 filter paper and membrane filters (0.45 µm, Millipore, USA) before subjecting samples to HPLC. All samples and standards injections were repeated three times for HPLC on an EC 250/4 Nucleosil C18 carbohydrate column (250 mm-4.0 mm i.d.) was used (Macherey-Nagel, USA). The elution solvent contained 75% acetonitrile and 25% deionized water. The column was operated at 30 °C with a flow rate of 1.8 mL min -1 .

Fruit quality analysis
Fifty fruits were randomly sampled from each genotype at their fully mature stage. There were five replicates each consisting of ten fruits. The pomological methods used for fruit weight, fruit width, neck length, ostiole width, total soluble solids (TSS), pH, acidity, and TSS/acidity determination have been described by . Maturity index (MI) also was determined and calculated as TSS/acidity, and varieties were categorized as least sweet, MI < 50; less sweet, MI 51-100; sweet, MI 101-150; or sweetest genotypes, MI > 150.

Fruit skin and flesh color measurements
Fruit skin and flesh colors were measured using a colorimeter (Chroma Meter CR-300, Minolta Co., Osaka, Japan). Color characters were expressed as L*, a*, C, and h° (reference). Fruit skin color was measured at three random positions per fruit, and the flesh color was measured at two random positions per fruit.

Statistical analysis
Data were analyzed using SAS software and procedures (SAS, 2005). Analysis of variance (ANOVA) tables were constructed with Tukey's Honestly Significant Difference (HSD) method at p < 0.05. Restricted maximum likelihood (REML) was used to estimate variance components. Pearson correlation coefficients and their levels of significance were calculated on a genotype mean basis using PROC CORR.

Phytochemical properties
Phytochemical characters of Turkish fig genotypes are presented in Table 1. These characters were affected to different degrees by genotype, harvest year, and genotype-by-year interaction, but the effect of the genotype on TA, TP and TAC (50% to 96%) was greater than for harvest year (accounting for 1% of total variation compared to 9% of total variation), and genotype-by-year interaction (accounting for 2% of total variation compared to 16% of total variation).
The fig genotypes in this study exhibited great diversity in values for TA (ranging from 2.7-113.6 μg cy-3-rutinoside g -1 fw), TP (ranging from 57.9 to 105.2 mg GAE/100 g fw), and TAC (ranging from 5.2 to 10.7 mmol Fe 2+ kg -1 ). 'Bursa Siyahı', which has dark black fruit skin, had the highest levels of TA (113.6 μg cy-3-rutinoside g -1 fw), TP (105.2 mg GAE/100 g fw), and TAC (10.9 mmol Fe 2+ kg -1 fw). The lowest TA and TAC values were found in the genotype 'Göklop', which has green fruit skin color (2.7 μg cy-3-rutinoside g -1 fw and 5.2 mmol Fe 2+ kg -1 fw, respectively). Similarly, the other fig genotypes with darker fruit skin have the higher contents of TA, TP, and TAC, as previously reported by other authors (Solomon et al., 2006;Çalişkan & Polat, 2011). In those studies, total anthocyanins were not detected in some fig genotypes with light fruit skin colors. However, TA was found in low concentrations in genotypes with light fruit skin colors in this study. This could be related to the fruit flesh a* values greater than 10 in our data. Also, the TA values of samples in the present study were lower than those in other studies on commercial fig cultivars (Dueñas et al., 2008) and fig accessions (Çalişkan & Polat, 2011), but higher than those reported by Solomon et al. (2006). The levels of TP compared to those found in this study were in agreement with those of Piga et al. (2008), but lower than found by Çalişkan & Polat (2011). The TAC levels in the present study were much higher than those reported by Solomon et al. (2006). The effects of the harvest year and genotype-by-year interaction on TP and TAC were also statistically significant (p < 0.05). The TA and TAC values were higher in 2010, whereas TP was higher in 2009. This result could be due to higher mean temperature (>29 °C) and an absence of rainfall at maturity in August, 2010. Sunshine duration also was higher in July and August 2010 (322.9 and 331.4 h per month, respectively) than in other years. Moretti et al. (2010) re-  (Guerrero et al., 2010;Hegedús et al., 2010).
In previous studies of antioxidant activity and total phenolics contents of fruit species, the effect of harvest year was significant (Connor et al., 2005). Therefore, these researchers suggested that antioxidant activity and total phenolics contents should be investigated over more than one year to obtain reliable results. The present study also confirmed that there is variation in these metabolites due to year. Analyses of variance for FRUC, GLUC, and SUC demonstrated significant (p < 0.01) variation depending on genotype, year, or genotype-by-year interaction ( Table 1). Analysis of the components of variance indicated that genotype (29% to 67%) contributed more to overall variation in FRUC, GLUC, and SUC than did year (1% to 11%) or genotype-by-year interaction (17% to 49%). The genotypes 'Sarılop', 'Göklop', and 'Yediveren' had the highest FRUC (10.7, 10.6, and 10.6 g/100 g fw, respectively) and GLUC (7.8, 9.0, and 8.4 g/100 g fw, respectively) values. SUC values were very low in figs, but FRUC contents were higher than the values reported by Melgarejo et al. (2003). Our results were similar to those of Çalişkan & Polat (2011).
Fruit quality properties of fresh figs include fruit size, skin and flesh color, flavor, sugar content, and acidity (Flaishman et al., 2008). The sugar composition of figs, especially fructose, can influence perceived fruit sweetness (Setser, 1993;Çalişkan & Polat, 2011). The genotype 'Sarılop' had the highest fructose and maturity index values. 'Sarılop' had a particularly sweet taste compared to the other genotypes in this study.
The FRUC, GLUC, and SUC contents were statistically significant (p < 0.01) for three harvest years ( Table 1). The highest levels of FRUC and GLUC were found in 2010, whereas level of SUC was the lowest in 2010, suggesting that sugar contents are influenced by environmental variables. The year-to-year variation in sugar composition may be explained by differences in climate and crop loads during the harvest year (Brooks et al., 1993). Also, harvest time could introduce variability among years and consequently influence the antioxidant capacity and sugar compositions of fruits among genotypes (Abidi et al., 2011).

Fruit quality characters
Results of analyses of variance showed that genotype was the main factor (p < 0.01) affecting fruit quality characters (Table 2). Genotypic effects, which accounted for the largest proportion of variation in all fruit quality traits, varied between 33.1% and 69.  (Aksoy et al., 2003), and were higher than those described by . Fruit neck length was the longest for the genotypes 'Bursa Siyahı' (5.6 mm) and 'Sarılop' had great variability in fruit neck length, pH, and acidity, but these characters were not affected by harvest year. The pH of the fruit juice was the highest for 31-IN-01 (5.2) and 31-IM-05 (5.1), while it was the lowest for 'Sarılop' (4.6) and 'Sarı İncir' (4.6). 'Sarı İncir' (0.24%), 'Morgüz' (0.23%), and 'Yediveren' (0.22%) juices had the highest acidity, while juice acidity was lowest for 'Sarı Zeybek' (0.14%).The maturity index (TSS/acidity) can be a good indicator of good fruit taste; furthermore, this ratio can be a valuable descriptive character in selecting cultivars for specific uses . The effect of harvest year on fruit weight, fruit width, ostiole width and TSS content, and genotype-by-year interaction on fruit width and neck length was significant (Table 2). However, no significant effect of harvest year on neck length, pH, acidity, and TSS/acidity was observed. Mean values for fruit weight (44.8 g), fruit width (44.6 mm), ostiole width (2.5 mm), and TSS (22.4%) were highest in 2010, whereas TSS/acidity (118.3) was highest in 2008. This can be explained by changing environmental conditions and fruit-bearing loads from year to year (Botti et al., 2003;Polat & Caliskan, 2009).  (Table 3). Genotype had the main effect (p < 0.01) on fruit skin and flesh color properties, and its effects were ranged from between 16% (hº) and 73% (a*) of the variation for fruit skin color and 11% (C) and 43% (a*) of the variation for fruit flesh color.

Fruit skin and flesh color measurements
Fruit skin and flesh color of fresh figs are very important for consumer preferences. Fresh figs with pink and red flesh color are preferred by consumers. In addition, fruit skin and flesh color are used to assess ripening in figs (Tsantili, 1990)  Different letters within columns indicate significant differences by Tukey's test at p < 0.05. **: Significant at p < 0.01; *: Significant at p < 0.05; ns: not significant. TSS: total soluble solids. Influence of genotype and harvest year on phytochemical and fruit quality of figs red color, was positive for 'Bursa Siyahı' (13.6) and 31-IM-05 (11.0), whereas it was negative (green color) for 'Sarı Zeybek' (-21.6). The darkest fruit skin color, C, was found for 'Bursa Siyahi' (20.2) and 31-IM-05 (23.7). Fruits with lower values for hº (60.8 for 'Bursa Siyahı' and 47.5 for 31-IM-05) are redder, and some researches have indicated that hº values would be suitable for estimating carotenoid levels of fruit species (Ruiz et al., 2005). 'Göklop' and 'Sarı Zeybek', the genotypes with the highest hº values can be regarded as having the lowest carotenoid levels in this study, whereas 31-IM-05 and 'Bursa Siyahı' with the lowest hº values are likely to be the richest in carotenoids.
The lightest fruit flesh L* value found in this study was 48.3 for 'Sarı Zeybek', while the a* value, which indicates the redness of fruit flesh, was highest for 'Yeşilgüz' (21.8). The genotypes 'Bursa Siyahı ' (27.5) and 'Sarı İncir' (27.9) had the darkest flesh colors with the lowest hº values, followed by the genotypes ' Yediveren' and 'Sarılop' (30.5 and 30.6,respectively). These genotypes with reddish flesh color tend to be preferred by consumers for fresh figs.
The effect of year on fruit skin and flesh color was highly significant (p < 0.01), except for flesh L* value, but the contribution of this effect to total variation was minor (Table 3). The harvest year had the lowest effect on both fruit skin and flesh color a* values (1% and 2%, respectively). The lowest fruit skin C values (lower values are darker in color) of these fig genotypes were found in 2010. This color density may be related to temperatures that reached 29.7 °C, and an absence of precipitation at maturity in August, 2010. Fruit color can vary from year to year depending on light and temperature. Temperature in particular has a strong effect on anthocyanin synthesis that provides coloration in fruit species (Wang et al., 2011).

Correlation analyses
We found significant positive correlations (p < 0.01) between either TA or TP contents and TAC (r = 0.56 and r = 0.77, respectively), are important components of the antioxidant capacity of figs, as found by Solomon et al. (2006) and Çalişkan & Polat (2011). Fruit skin L*, a*, and C values were found to be moderately correlated with TA, TAC, and TP levels (Table 4). Lightness (L*) and C values were negatively related to TA, TAC, and TP contents in these fig genotypes, whereas a* values were positively correlated with TA, TAC, and TP levels. The phenolic compounds, especially anthocyanins (cy-3-rutinoside), cinnamic acid, and flavonoids, are prominent in fig fruit skins (Del ; therefore, a correlation between fruit skin colors and TP or TAC can be expected. Moreover, FRUC and SUC values were showed negative significant correlations with TA (r = -0.34 and r = -0.42, at p < 0.01). In contrast to our results, Abidi et al. (2011) reported a positive correlation in grapes between sugar content and total anthocyanins. This is likely because fig genotypes with green or brown fruit skin color have higher FRUC, GLUC, and SUC than do genotypes with black fruit skin color (Çalişkan & Polat, 2011). In agreement with the above results, fruit skin a* values were negatively correlated with FRUC, GLUC, and SUC contents, whereas fruit skin C values were positively related to FRUC and SUC. These findings support to the relationship between either FRUC or SUC and TA levels.
The a* values of fruit fresh were positively related to TA and GLUC, while flesh h° values were inversely related to TA and GLUC in these fig accessions. Fruit weight, fruit size, and fruit neck length were not correlated with TA, TAC, and TP (data not shown). Moyer et al. (2002) indicated that within the same fruit species and genotypes, smaller fruits tend to have higher TP content and TAC, as compounds resulting in higher TP and TAC are usually richest in fruit skin, and genotypes with smaller fruits have a relatively larger skin area compared to that of larger genotypes. Surprisingly, pomological characteristics such as fruit weight, fruit size, and fruit neck length were not correlated with fruit skin and flesh color values, as well as TA, TP, and TAC. This result may be explaining the lack of correlation among these traits by the color of the fruit skin (data not shown). Thus, pomological characters of species such as figs, which occur in different fruit color groups, have limited effects on the phytochemical characteristics. In addition, Significance: ** p < 0.01, *p < 0.05; TA: total anthocyanins; TP: total phenolics; TAC: total antioxidant capacity; FRUC: fructose; GLUC: glucose; SUC: sucrose. Influence of genotype and harvest year on phytochemical and fruit quality of figs these results could be explained by the fact that the physical fruit growth and the chemical component accumulation and biosynthesis are different process. The genotypes with black fruit skin color, 'Bursa Siyahı', had the most excellent fruit quality and richest phytochemical characters. 'Sarılop', which is the most important dried fig genotype had as good fruit quality as fresh fig genotypes, but had lower phytochemical profiles associated with yellow fruit. 'Göklop' and 'Yediveren' with green fruit skin color, 31-IN-16 with yellow fruit skin color, and 31-IM-05 with brown fruit skin color are promising for export markets in terms of fruit quality characters; however their phytochemical compositions, except for 31-IM-05, were lower due to green or yellow fruit colors. 'Bursa Siyahı' may be useful in a breeding study as a donor genotype for both high phytochemical levels and excellent fruit quality characters. Phenotypic and heritability calculations in several fruit species showed that the improvement of TA and TP is possible through selective breeding (Connor et al., 2005;Abidi et al., 2011). Therefore, the favorable fruit characters defined by 'Bursa Siyahı' should be useful in hybridization studies.
As conclusions, this study provides the first data describing the phytochemical characters and sugar compositions of some important Turkish fig genotypes.
The results indicated that considerable variation in phytochemical and fruit quality characters, and sugar compositions occurs in fig genotypes based on genotype, harvest year, and genotype-by-year interactions. The genotype with black fruit skin, 'Bursa Siyahı', had the highest levels of TA, TP, and TAC. Genotype had a highly significant effect on TA, TP, and TAC properties in figs. Harvest year had no effect on fruit neck length, pH, and acidity; however, harvest year did have a significant effect on variation in fruit color and phytochemical characters.