Acta Chim. Slov. 2002, 49, 537-544. 537 INFLUENCE OF VARIETY AND GEOGRAPHICAL REGION ON MINERAL CONTENTS OF HAZELNUT (CORYLUS AVELLANA L.) VARIETIES Mustafa S. Dundar,* Emel Bahçivanci, Canan Muslu Sakarya University, Fen-Edebiyat Fak. Kimya Bölümü, 54100 Adapazari, TURKEY. (Email : dundar@sakarya.edu.tr) Received 01-02-2002 Abstract The mineral contents of three different varieties of hazelnut, namely Karafindik, Tombul, and Delisava, which are cultivated in Sakarya-Turkey, were studied for variety and geographical region and analysed by using flame atomic absorption spectrometry. A total of four elements, iron, copper, manganese, zinc, were determined in hazelnut kernels in hulled and dehulled forms. The mean levels of iron, copper, manganese, and zinc are 3.45±1.3, 1.61±1.1, 12.72±0.6, 2.63±0.6 in hulled kernels, 3.24±1.9, 1.72±0.9, 11.54±0.9, 2.76±0.4 in dehulled kernels, respectively. Mineral compositions of varieties showed significant differences from each other by means of geographical region and hulled or dehulled state of hazelnut kernels. It was found that the highest mineral levels were measured in Hendek district whereas in Akyazi district lowest mineral contents were obtained. Introduction Wild species of hazelnut, genus Corylus, are distributed in nearly all parts of temperate zones of the northern hemisphere while the major producers are Italy, Spain, USA, and Turkey which produces about 80% of the total hazelnut production of the world exports.1,2 Tombul is among the top, Kara Findik is among the very high quality hazelnut cultivars of the world. The Turkish variety of hazelnut, Delisava, is mainly consumed in-shell in the domestic market due to their moderate texture, taste, and other quality factors. Essential minerals improve the nutritional quality of Turkish hazelnut cultivars. They are not inferior cultivars but the demand for it is less due to overproduction of the better varieties.1,2,3 Parcerisa4 stated that the hazelnuts from Spain are influenced strongly by environmental and growing conditions. Studies indicate that mineral compositions of hazelnuts are affected by variety, geographical origin, harvest year, climate and the methods of cultivation.4,5 There is no researches on mineral composition of hazelnut varieties cultivated in Sakarya-Turkey based on hulled or dehulled forms of hazelnut kernels. Mineral compositions are of interest due to their M. S. Dundar, E. Bahçivanci, C. Muslu: Influence of variety and geographical region on mineral. 538 Acta Chim. Slov. 2002, 49, 537-544. prooxidant activity and health benefits.4,5,6 Therefore, this study was planned and carried out to determine the effects of geographical region, variety and kernel hull on mineral composition of Turkish hazelnut varieties. Experimental Reagents and Standards All chemicals used were of analytical-reagent grade (Merck, Darmstadt, Germany) and ultra high purity water (18 M? cm-1) was used throughout the experiments. Working standard solutions for calibration purposes containing 0.05-0.80 µg ml-1 of iron and zinc, 0.01-0.64 µg ml-1 of copper, and 0.10-0.80 µg ml-1 of manganese were prepared from spectroscopic grade stock standard solutions (1000 mg L-1) by serial dilution with 0.2% (v/v) HNO3 prior to use. Peak height was used in all measurements because better precision was obtained. The reproducibilites of the method expressed as coefficients of variation for Fe, Cu, Zn and Mn contents were: 1.10%, 0.60%, 0.20% and 0.10%, respectively. The sensitivities of the method for Fe, Cu, Zn, and Mn contents were; 0.48 mg/kg, 0.32 mg/kg, 0.21 mg/kg, 0.26 mg/kg, respectively. Cleaning and Storage Material All glassware and polyethylene bottles were kept overnight and cleaned by soaking in 10% HNO3, rinsing five times with ultra high purity water prior to use. Sampling Three samples corresponding to three commonly grown varieties of hazelnut (Corylus avellana L.) were collected from trees cultivated in five geographical districts of Sakarya province of Turkey (The varieties are Kara Findik, Tombul and Delisava from Kocaali, Karasu, Akyazi, Hendek, Ferizli districts). Samples in hulled and dehulled states were stored after wait at 100 °C for twelve hours. The hazelnut kernels were separated from hulls and stored in polyethylene bags as hulled and dehulled forms until analysis were performed. All analyses were conducted in duplicate and each value measured is a mean of three replicate determinations. M. S. Dundar, E. Bahçivanci, C. Muslu: Influence of variety and geographical region on mineral. Acta Chim. Slov. 2002, 49, 537-544. 539 Chemical analysis Mineral contents (iron, copper, zinc and manganese) of hazelnut varieties were determined according to standard method described in AOAC Official Methods of Analysis (AOAC Official Method 975.037) by wet ashing of hazelnut kernels and Shimadzu (Tokyo, Japan) AA6701F model flame atomic absorption spectrometer equipped with an autosampler was used in all measurements. Background absorption was corrected by using a deuterium lamp and Koto brand hollow-cathode lamps were used in all measurements. Statistical analysis One way analysis of variance (ANOVA), multiple range least significant difference (LSD) test and correlation analyses were carried out by using a statistical package program (AXUM ver. 4.1) for p<0.05 significance level. Means, standard error of mean and significance level for every parameter are shown in table 1. Further, percentage of recoveries of each element were determined in hazelnut kernels both hulled and dehulled forms using dry and wet ashing digestion methods (table 2). In these digestion methods, spiked samples were used to calculate percentage of recoveries to evaluate the reliability of the method used. Table 1. Statistical data:mean (×), standart error of mean (SE) and significance level for mineral composition. Kocaali Karasu Akyazi Hendek Ferizli ×××××SE P Iron 3.49 2.66 3.35 3.89 3.88 0.0067 NS Copper 1.60 1.49 1.51 1.81 1.64 0.0026 NS Manganese 15.21 13.36 6.40 16.07 12.54 0.0114 NS Zinc 2.68 2.59 2.53 2.79 2.57 0.0024 NS Iron 3.17 2.31 2.82 3.60 4.30 0.0092 NS Copper 1.72 1.70 1.60 1.93 1.67 0.0023 NS Manganese 13.37 17.54 3.62 14.33 8.83 0.0155 NS Zinc 2.82 2.97 2.62 2.77 2.64 0.0016 NS M. S. Dundar, E. Bahçivanci, C. Muslu: Influence of variety and geographical region on mineral. 540 Acta Chim. Slov. 2002, 49, 537-544. Table 2. Recovery test for Fe, Cu, Mn, and Zn in hulled and dehulled Kara Findik variety of Turkish hazelnut kernels from Kocaali district. Results are averages of three replicates. (mg/100 g). Percentages of Recoveries ±RSD% Iron Copper Manganese Zinc Hulled Kara Findik 83±3.2 76±1.0 74±0.9 83±0.2 Dehulled Kara Findik 86±7.6 85±1.7 98±0.7 84±0.2 Hulled Kara Findik 102±1.0 108±0.7 97±0.2 102+0.6 Dehulled Kara Findik 99±0.7 105±1.1 100±0.2 98±0.5 Results and Discussion Three different hazelnut varieties (Kara Findik, Tombul, and Delisava) collected from Kocaali, Karasu, Akyazi, Hendek, and Ferizli districts were divided into two different categories as hulled and dehulled forms and analysed for iron, copper, zinc, and manganese. The results in general showed that iron and manganese levels determined were higher in hulled kernels than dehulled kernels while in dehulled kernels higher amounts of copper and zinc were measured. The results obtained can be seen in tables 3 and 4 as hulled and dehulled kernels, respectively. The range of iron in all districts studied are between 1.61-5.64 mg/100 g. The highest level were obtained in dehulled Delisava kernels of Ferizli district at the amount of 5.64 mg/100g and the lowest level were dehulled Kara Findik kernels of Karasu at the amount of 1.61 mg/100 g. Copper concentrations were determined in the range of 1.38-1.97 mg/100g. The highest levels were found in dehulled Delisava kernels of Kocaali and Hendek districts at 1.97 mg/100 g. The lowest levels were obtained in hulled Delisava kernels of Akyazi and Karasu districts at the amount of 1.38 mg/100g. The range of manganese changes between 2.41-21.77 mg/100 g. The highest amounts were obtained from dehulled Kara Findik of Karasu district at the level of 21.77 mg/100 g while the lowest level was observed in dehulled Delisava kernels of Akyazi district at the rate of 2.41 mg/100 g. Zinc levels were determined between 2.15-3.77 mg/100g based on sample types and districts. In Karasu district, dehulled Delisava hazelnut M. S. Dundar, E. Bahçivanci, C. Muslu: Influence of variety and geographical region on mineral.. Acta Chim. Slov. 2002, 49, 537-544. 541 showed the highest level at 3.77 mg/100 g. However, the hulled Tombul kernels of Akyazi district showed the lowest level at 2.15 mg/100 g. In general, hazelnut samples picked up from Hendek district indicated the highest mineral compositions whereas Akyazi district showed lowest levels. Delisava variety showed the richest mineral composition among the three hazelnut varieties analysed. The lowest mineral contents were obtained in Tombul variety of all regions except for Hendek. On the other hand, Delisava variety showed highest mineral contents in all regions except for Ferizli and Akyazi districts (tables 3 and 4). Besides the varieties, geographical region differences affect the mineral compositions of varieties. These differences observed could be due to the geological formation of soil, the fertilisers and pesticides used, climate conditions, etc. Mineral composition of hazelnuts shown in tables 3 and 4, with respect to variety and geographical region, the mean values of iron, copper, manganese and zinc content of the samples were 3.45, 1.61, 12.72, 2.63 mg/100 g Table 3: Mineral contents of hulled Turkish hazelnut kernels in mg/100 g (n=3)* Iron Copper Manganese Zinc Hazelnut______________________________________________________________________________ Variety |y|ean Range Mean Range Mean Range Mean Range ±RSD% (Min-Max) ±RSD% (Min-Max) ±RSD% (Min-Max) ±RSD% (Min-Max) Kara Findik 3.29 ± 1.0 3.15 - 3.41 1.42 ± 0.7 1.39 - 1.44 16.48 ± 0.2 16.45 - 16.51 2.63 ± 0.6 2.50 - 2.73 Tombul 2.82 ± 0.6 2.79 - 2.84 1.58 ± 0.8 1.56 - 1.59 12.18 ± 0.6 12.06 - 12.29 2.17 ± 0.7 2.15 - 2.18 Delisava 4.37 ± 0.3 4.30 - 4.42 1.80 ± 1.3 1.78 - 1.82 16.98 ± 0.8 16.95 - 17.00 3.25 ± 0.1 3.24 - 3.26 Kara Findik 2.74 ± 1.5 2.69 - 2.76 1.70 ± 1.8 1.67 - 1.72 11.35 ± 0.3 11.34 - 11.36 2.53 ± 0.4 2.49 - 2.57 Tombul 2.45 ± 1.7 2.43 - 2.46 1.40 ± 1.3 1.39 - 1.41 8.70 ± 0.4 8.67 - 8.72 2.29 ± 0.2 2.19 - 2.37 Delisava 2.78 ± 0.3 2.77 - 2.79 1.38 ± 0.8 1.35 - 1.40 20.03 ± 0.3 19.91 - 20.13 2.94 ± 0.4 2.82 - 3.04 Kara Findik 3.69 ± 1.3 3.18 - 4.19 1.61 ± 1.0 1.56 - 1.65 3.87 ± 0.8 3.86 - 3.88 2.58 ± 0.7 2.52 - 2.63 Tombul 2.77 ± 0.7 2.75 - 2.78 1.55 ± 1.7 1.52 - 1.57 4.23 ± 1.7 3.98 - 4.47 2.15 ± 0.3 2.10 - 2.19 Delisava 3.59 ± 0.5 3.48 - 3.69 1.38 ± 1.2 1.35 - 1.39 11.09 ± 0.4 11.00 - 11.17 2.86 ± 0.7 2.51 - 3.21 Kara Findik 4.65 ± 1.5 2.46 - 4.83 1.77 ± 1.9 1.76 - 1.78 11.58 ± 0.5 11.23 - 11.93 2.67 ± 1.2 2.57 - 2.76 Tombul 2.45 ± 2.9 2.36 - 2.51 1.84 ± 0.7 1.68 - 1.97 16.68 ± 0.4 16.63 - 16.72 2.27 ± 0.6 2.26 - 2.29 Delisava 4.58 ± 0.3 4.57 - 4.59 1.82 ± 0.4 1.80 - 1.83 19.96 ± 0.5 19.84 - 20.06 3.44 ± 0.4 2.94 - 3.92 Kara Findik 4.20 ± 1.8 4.01 - 4.40 1.66 ± 1.4 1.64 - 1.67 14.54 ± 0.3 14.43 - 14.65 2.41 ± 1.3 2.30 - 2.51 Tombul 3.20 ± 2.6 3.13 - 3.23 1.79 ± 1.3 1.65 - 1.92 14.34 ± 0.2 13.72 - 14.96 2.45 ± 0.6 2.25 - 2.65 Delisava 4.24 ± 2.1 4.10 - 4.39 1.46 ± 0.9 1.45 - 1.47 8.75 ± 1.1 8.69 - 8.82 2.85 ± 0.6 2.84 - 2.86 Mean 3.45 ± 1.3 3.21-3.56 1.61 ± 1.1 1.57-1.64 12.72 ± 0.6 12.58-12.84 2.63 ± 0.6 2.51-2.74 * Each value is a mean ± relative standard deviation of three determinations. M. S. Dundar, E. Bahçivanci, C. Muslu: Influence of variety and geographical region on mineral. 542 Acta Chim. Slov. 2002, 49, 537-544. Table 4: Mineral contents of dehulled Turkish hazelnut kernels in mg/100 g (n=3)* Hazelnut Iron Copper Manganese Zinc Variety Mean Range Mean Range Mean Range Mean Range ±RSD% (Min-Max) ±RSD% (Min-Max) ±RSD% (Min-Max) ±RSD% (Min-Max) __ Kara Findik 2.78 ±0.7 2.62-2.92 1.48±1.1 1.43-1.50 16.18±0.2 15.93-16.53 2.48±0.5 2.44-2.50 Tombul 2.76 ±4.2 2.58-2.93 1.70±0.5 1.69-1.70 9.85±0.7 9.79-9.89 2.36±0.3 2.31-2.40 Delisava 3.96 ±0.9 3.87-4.04 1.97±0.2 1.95-1.96 14.07±0.3 13.98-14.15 3.61±0.1 3.60-3.62 Kara Findik 1.61 ±3.2 1.55-1.66 1.85±1.4 1.68-2.00 21.77±0.5 21.63-21.88 2.96±0.5 2.88-3.03 Tombul 2.14 ±3.7 2.06-2.21 1.55±0.7 1.83-1.55 9.14±0.9 9.09-9.18 2.18±0.1 2.13-2.23 Delisava 3.18 ±2.0 2.75-3.60 1.69±0.8 1.64-1.73 21.70±0.8 21.59-21.80 3.77±0.8 3.14-4.38 Kara Findik 2.37 ±3.1 2.23-2.49 1.67±0.4 1.66-1.68 3.58±2.1 3.56-3.60 2.73±0.6 2.62-2.82 Tombul 1.83 ±0.2 1.82-1.84 1.65±1.3 1.63-1.66 4.88±0.7 4.76-4.98 2.32±1.0 2.30-2.34 Delisava 4.26 ±2.7 2.03-6.48 1.48±2.7 1.44-1.50 2.41±2.3 2.39-2.42 2.82±0.3 2.81-2.83 Kara Findik 2.37 ±0.7 2.22-2.49 1.94±0.5 1.93-1.95 10.24±0.5 10.14-10.33 2.73±0.4 2.60-2.85 Tombul 3.91 ±1.9 3.19-4.61 1.88±1.4 1.86-1.89 14.02±1.3 13.69-14.34 2.60+0.2 2.15-3.04 Delisava 4.53 ±1.8 4.38-4.67 1.97±0.9 1.96-1.98 18.72±0.8 18.61-18.83 2.98±0.3 2.92-3.02 Kara Findik 3.60 ±0.3 3.49-3.71 1.85±0.3 1.82-1.88 13.24±0.4 13.14-13.34 2.47±0.9 2.40-2.54 Tombul 3.67 ±1.9 3.62-3.72 1.56±1.2 1.45-1.68 5.15±1.0 5.01-5.28 2.40+0.7 2.38-2.43 Delisava 5.64 ±0.7 5.50-5.78 1.60±0.6 1.56-1.64 8.10±1.2 7.74-8.46 3.04±0.1 2.98-3.09 Mean 3.24 ±1.9 2.93-3.54 1.72±0.9 1.68-1.76 11.54±0.9 11.40-11.67 2.76±0.4 2.64-2.87 * Each value is a mean ± relative standard deviation of three determinations. in hulled kernels and 3.24, 1.72, 11.54, 2.76 mg/100 g in dehulled kernels, respectively. Significant differences between mineral contents of varieties were observed (table 3 and 4) based upon hulled and dehulled forms of hazelnut kernels. The highest levels of manganese were determined in the samples of Delisava and Kara Findik. The results are comparable with those given by Açkurt et al.8 Ayfer et al.2 Baº et al.9, Mehlenbacher10 and by Pala et al.11 for Turkish hazelnut varieties. It can be seen from tables 2 and 3 that copper and zinc levels of hazelnut varieties were not affected by geographical region. However, there were significant differences observed in the levels of iron between the geographical regions. Mineral compositions of our results were also in agreement by previous studies carried out by Açkurt et al.8, Ayfer et al.2, Baº et al.9, Pala et al.11 and Özdemir1 for Turkish hazelnut varieties. The results reported by Parcerisa et al.4 for Spanish hazelnuts that significant changes were observed for manganese and copper according to geographical regions. M. S. Dundar, E. Bahçivanci, C. Muslu: Influence of variety and geographical region on mineral. Acta Chim. Slov. 2002, 49, 537-544. 543 In addition, a correlation study was applied to mineral compositions included in this work. Strong correlations observed in table 5 suggest that soil composition and uses of fertilizers influence mineral composition, which consequently contributes stability and quality of hazelnuts. Parcerisa et al.4 stated that composition of soil, uses of fertilizers and irrigation affect the mineral compositions of hazelnuts and consequently influence the stability and quality of the product. On the other hand, variety has a minor affect on the mineral composition of hazelnuts. A correlation study was performed between iron, copper, manganese and zinc contents. Results for the correlation coefficient (r) and significance level (p) are shown in table 5. Table 5. Correlation between mineral compositions and kernel forms. Iron Copper Manganese Iron 0.3180 NS Copper Manganese Copper r1 p2 0.0283 NS3 Manganese r p -0.0557 NS 0.4850 NS 0.1426 NS 0.2732 NS Zinc r p 0.3500 NS 0.3584 NS 0.4965 NS 0.6649 NS 0.1324 NS 0.5144 NS 1 r, correlation coefficient. 2 p, probability of significance. 3 NS, not significant. Significant differences were found for mineral contents between hulled and dehulled hazelnut kernels. In relation to the metal contents determined, it can be concluded that only iron and manganese contents in hulled kernels showed a negative correlation. So, other mineral contents are positively correlated. Among these four elements, copper content is the only one that shows significant differences between the varieties. The accuracy of the method for hazelnut samples was estimated with the analysis of spiked samples. In spiked samples examined, the recovery tests showed good agreement of the results with the calculated values (table 2) which indicating that present procedure can reliably be used for analysis of minerals in hazelnut samples. M. S. Dundar, E. Bahçivanci, C. Muslu: Influence of variety and geographical region on mineral. 544 Acta Chim. Slov. 2002, 49, 537-544. References 1. Ozdemir, M. Z. Lebensmittel Untersuchung und Forschung, 1985, 180, 289-292. 2. Ayfer, M., Uzun, A., Baº, F. Türk findik çeºitleri. Ankara: Black Sea Hazelnut Exporters Union, 1986. 3. Diaz, A., Torija, M.E., Valls, C. Anal. Bromatol, 1980, 32, 41-48. 4. Parcerisa, J., Rafeces, M., Castellote, A.I., Codony, R., Farran, A., Garcia, J., Lopez, A., Romero, A., Boatella, J. Food Chemistry, 1985, 53, 71-74. 5. Pershem, A.S., Breene, W.M., Lulai, E.C. Journal of Food Processing and Preservation,1995, 19, 9-25. 6. Alphan, E., Pala, M., Açkurt, F., Yilmaz, T. Acta Horticulturae, 1996, 445, 305-310. 7. AOAC, Official Methods of Analysis. 16th edn. Association of Official Analytical Chemists, Washington DC, 1996, pp.3-4. 8. Açkurt, F., Ozdemir, M., Biringen, G., Löker, M. Food Chemistry,1999, 65, 309-313. 9. Baº, F., Ömeroglu, S., Türdü, Aktaº, S. Gida, 1986, 11, 195-203. 10. Mehlenbacher, S.A. Acta Horticulture,1991, 290, 791-836. 11. Pala, M., Açkurt, F., Löker, M., Yildiz, M., Ömeroglu, S. Turkish Journal of Agriculture & Forestry,1996, 20, 43-48. Povzetek Prouèevana je bila vsebnost mineralov v treh vrstah lešnikov, Karafindik, Tombul in Delisava, ki jih gojijo v Turški pokrajini Sakarya. Študiran je bil vpliv geografskega podroèja. Vzorci so bili analizirani s plamensko atomsko absorpcijsko spektrometrijo. M. S. Dundar, E. Bahçivanci, C. Muslu: Influence of variety and geographical region on mineral.