Introduction

Quercus is one of the most important woody genera with large number of species including trees and shrubs in the Northern Hemisphere. Major distribution centers in the world are North America, Europe and Eastern Asia. Eastern Asia has the highest diversity with about 250 species. Oaks such as Turkish oak, sessile oak, pubescent oak and English oak have been an important source of fuel, fodder, and building materials throughout history. Tannins, dyes and valuable nutrient are important coproducts from acorns. These trees are tolerant to different environmental and climatic conditions and also contribute to erosion control. Eighteen of these species are native to Turkey. Related subspecies, varieties and natural hybrids cover about 6.5 million ha in Thrace and Anatolia [1, 2]. However, the classification of individual trees to a given species is quite challenging. Widespread hybrid combinations are known to occur only in the wild within the members of the same sub-genus. For example, no hybrids are reported between Lepidebalanus (white oaks) and Erythrobalanus (black and red oaks) [3]. Still, intermediate types, intraspecific variation and environmental influences may have obscured detection of hybridization.

Extreme variability occurs especially in the populations of fairly broad geographic distributions. Several problems of nomenclature and typification particularly in the wide-ranging groups are still unresolved [4] and modern monographic treatment is needed to improve phylogenic analysis. Some molecular genetic studies have been conducted on the diversity and phylogeny of Quercus. Phylogenetic relationships within Quercus subgenus Quercus using chloroplast DNA (cpDNA) restriction sites, nucleotide sequences of the internal transcribed spacers (ITS) and nuclear ribosomal DNA repeats were analyzed and individual gene trees were reported to be congruent and often complementary in supporting clades that generally correspond to previously recognized taxonomic groups [5]. Good resolution among most species groups and distribution patterns were obtained in the phylogeny of Asian Quercus species based on the non-coding region sequences of chloroplast genome and ITS sequences of nuclear rDNA. Yet, ITS and chloroplast DNA phylogeny were largely congruent with regard to the infrageneric classification system, thus, the species relationships among very closely related species were unresolved [6]. However, chloroplast DNA variation at the intraspecific level were detected in Quercus and the geographical distribution of cpDNA variants was reported to be largely dependent on species migration [7, 8].

On the other hand, fruits of oaks are a stable and generative organ, and phenotypic differences in acorn composition are an important diagnostic element [9]. Chemical content of seeds is determined ultimately by genetic factors and varies widely among species and their varieties and cultivars [10]. Fatty acid profiles of the seed oils have great taxonomic value in the plant kingdom [11]. Biochemical systematics has been frequently used as a tool in studies of some plant groups at different taxonomic levels [12–16]. Chemosystematic differentiation based on differences in acorn fatty acid composition between Italian and Spanish populations of Q. ilex and Q. rotundifolia yielded partial separations of the individual populations [17]. Such separation using acorn fatty acids for native and hybrid populations of Q. agrifolia and Q. wislizenii was also achieved [18]. French Mediterranean evergreen oak populations were reported to be intermediate and heterogeneous for the fatty acid profiles between Spanish and Italian oak, suggesting a zone of hybridization [19]. In addition to genetic factors, some environmental conditions may also influence variations of this parameter in Quercus. Acorns of Q. brantii from different climatic conditions in the Zagross mountains produced different contents of crude fat, although fatty acid values were not significantly different [20]. A few studies have examined the chemical contents of Turkish Quercus acorns from a taxonomical point of view [21, 22]. This investigation examined oaks that were very polymorphic, taxonomically problematic and important woody groups in the "Flora of Turkey". Total oil contents and compositional fatty acid levels in the mature acorns of 16 Quercus taxa were associated with taxonomical and diagnostic parameters at the infrageneric levels. These data have provided biochemical markers that help establish phylogenic associations of this genus, and also reveal potentially account as an alternative source for dietary nutrition.

Materials and Methods

Mature acorn specimens were collected at random from ten individual trees of native populations of Q. pontica C. Koch (pontine oak), Q. robur L. subsp. robur (pedunculate oak) Q. hartwissiana Steven (strandzha oak), Q. frainetto Ten. (Hungarian oak), Q. petraea (Mattuschka) Liebl. subsp. petraea (sessile oak), Q. vulcanica (Boiss. & Heldr. ex) Kotschy (kasnak oak), Q. pubescens Willd. (pubescent oak), both variety of Turkish oaks; Q. cerris L. var. cerris and Q. cerris L. var. austriaca (Willd.) Loudon, Q. ithaburensis Decne. subsp. macrolepis (Kotschy) Hedge & Yalt. (tabor oak), Q. brantii Lindl. (Brant’s oak), Q. libani Olivier (Lebanon oak), Q. trojana P.B. Webb (Macedonian oak), Q. ilex L. (holm oak), Q. aucheri Jaub. & Spach (boz-pirnal oak), Q. coccifera L. (kermes oak). In addition, acorn samples of two varieties of Q. cerris were collected from three individual populations in the geographically different Marmara and Eagean regions. Three values of total oil and fatty acid compositions were presented for each variety of this species to show potential environmental influence on these traits. Each species was verified by the classification criteria in “Flora of Turkey” [4] and by comparison with identified specimens from the ISTF (Herbarium of Faculty of Science of Istanbul University) and the ISTO herbarium (Herbarium of Faculty of Forestry of Istanbul University). Voucher specimens were deposited with the Division of Botany, Istanbul University.

Sample Preparation and Analysis

Air dried mature acorn samples of each Quercus taxon were dehulled, and the kernels (dicotyledons) were ground into meal and homogenized with pestle and mortar. Total oil content was detected with “Tecator Soxtec System HT”. Powdered material (3 g) from each sample was added to oil in cartridge (W1) with 25–50 mL ether into a weighted extraction pot (W2). Extraction was carried out for 15 min with rinsing for 30–45 min. The extracted seed meals were air dried to remove traces of solvent and oven dried at 100 °C. The pots were cooled in a desiccator and weighed (W3). The following equation was used to calculate percentages of the oil;

$$ {\text{Oil }}\% = {\left( {{{\left( {W3 - W2} \right)}} \mathord{\left/ {\vphantom {{{\left( {W3 - W2} \right)}} {W1}}} \right. \kern-\nulldelimiterspace} {W1}} \right)} \times 100 $$

The oil was transferred into glass sealed amber dark bottles, capped and stored at −18 °C until analyzed. The IUPAC standard method for the preparation of the fatty acid methyl esters were used [23]. Approximately 0.150 g. of powdered material from each sample was added to 5 mL 0.5 N NaOH in 50% aqueous methanol with the glass. The solution was incubated at 100 °C for 15 min in a water bath for saponification, then boiled for 5 min with 5 mL BF3 and to the addition of 2–5 mL n-heptane. This mixture (25 mL) was washed with saturated NaCl. An aliquot (1–2 mL) of the heptane phase was dehydrated with a few crystals of anhydrous Na2SO4. The methyl esters of 33 fatty acids were quantified by thermoquest trace GC equipped with a SP-2330 fused silica capillary column (30 m, 0.25 mm, ID-0.20 μm). The oven temperature was held at 120 °C for 2 min and increased at a rate of 5 °C/min and held at 220 °C for 8 min. Injector and detector temperatures were 240 and 250 °C, respectively. Hydrogen was used as carrier gas at a flow rate of 35 mL/min. Split flow, split ratio and sample injection was 75 mL/min, 1/150 and 0.5 μL. Identification and quantification of fatty acid methyl esters was accomplished by comparing the retention times of the peaks with authentic standards (Sigma).

Data Evaluation

Statistical analysis of the experimental results at the p < 0.05 significance level (SPSS 10.0). Principle component analysis of the ratios of total percentages of the fatty acids was conducted with statistiXL.

Results and Discussion

The range of total oil concentration among all taxa was between 0.7% (w/w) (Q. vulcanica) and 7.4% (w/w) (Q. brantii). The average total oil concentration in section Cerris, Quercus and Ilex was 3.91, 1.35, 0.85% (w/w), respectively. Minor variations in total oil content were detected between samples of two varieties of Q. cerris collected from geographically different populations, with relatively higher values in Q. cerris var. austriaca. Percentages of the fatty acids in the examined species were documented in Tables 1, 2, 3, 4. The major fatty acids were oleic (10.2–54.4%), linoleic (24.2–49.1%), palmitic (13.6–30.4%), alpha linolenic (1.5–8.6%) and stearic acid (1.5–4.5%). The values for palmitic, stearic and oleic acid at the section level were significantly different (p < 0,05). Total percent of saturated fatty acids were between 17.01 and 38.63%. The concentration of mono, poly and unsaturated fatty acids in total oil were between 11.05 and 55.47%, 26.15–54.72% and 57.44–81.62%.

Table 1 Total oil percentages, fatty acid compositions and some of their ratios for the acorns of examined taxa in section Quercus
Full size table
Table 2 Total oil percentages, fatty acid compositions and some of their ratios for the acorns of examined taxa in section Cerris Loudon
Full size table
Table 3 Total oil percentages, fatty acid compositions and some of their ratios for the acorns of two varieties of Q. cerris (section Cerris Loudon) collected from three different populations
Full size table
Table 4 Total oil percentages, fatty acid compositions and some of their ratios for the acorns of examined taxa in section Ilex Loudon
Full size table

The concentration of total saturated fatty acids among species compared to mono, poly and total unsaturated fatty acids at the section level was significantly different (p < 0.05). The highest ratios of mono:poly unsaturated fatty acid were found in Q. brantii and Q. trojana (section Cerris). Whereas, Q. aucheri and Q. coccifera (section Ilex) exhibited lower ratios. The lowest ratios of total saturated:total unsaturated fatty acids were found in Q. brantii and Q. libani (section Cerris), whereas, the species from section Ilex exhibited a high value. Similar values in related species and infraspecific hybrids were observed in both varieties of Q. cerris. Both varieties of Q. cerris could be delineated from the other species in section Cerris for almost all parameters examined, except for stearic acid (p < 0.05). In addition, significant variation in palmitic acid values was the only difference between both varieties of Q. cerris (p < 0.05).

In contrast, the highest oleic acid concentrations obtained are in Q. brantii and Q. trojana from that section. A remarkably high value for this fatty acid also was detected in Q. vulcanica from section Quercus (Table 1). Considerably lower levels for oleic acid were observed in both varieties of Q. cerris, but, the lowest levels are in the species from section Ilex. The highest value of linoleic acid was observed in Q. hartwissiana (section Quercus) (49.1) and the lowest concentrations were obtained in the species of section Cerris except for Q. cerris, in addition to Q. vulcanica from section Quercus. Both varieties of Q. cerris expressed high levels for this fatty acid (Table 3). The levels of α-linolenic acid as the other poly unsaturated fatty acid are generally higher in section Ilex (Table 4), but lower in section Cerris. As observed for linoleic acid, α-linolenic acid levels in both varieties of Q. cerris were greater than the section averages. Different ratios of linoleic:α-linolenic acids were also compared at the sectional and specific level. Significant differences were found only between the two varieties of Q. cerris compared to other species from its section (p < 0.01).

Principle component analysis based on five different ratios of fatty acids applied to the whole series of Quercus species show that the first two axes of this analysis accounted for 88.7% of the variation in the data. Principle components 1 and 2 explain 65.9 and 22.8% of total variation, respectively. This analysis defines these relatively distinct groups. Section Ilex is clearly segregated, however two varieties of Q. cerris (Table 3) appear to be aligned with the section Quercus gene pool, and Q. vulcanica from section Quercus appear to be more closely related to section Cerris. In addition, Q. petraea subsp. petraea appear to be aligned with Q. vulcanica in section Cerris group (Fig. 1).

Fig. 1
figure 1

Principle component analysis of 16 Quercus species based on five calculated ratios of the fatty acid percentages (total saturated/mono unsaturated, total saturated/poly unsaturated, total saturated/total unsaturated, mono/poly unsaturated, linoleic/α-linolenic acid). Filled diamonds section Quercus, filled circles section Cerris, filled squares section Ilex

Full size image

Besides distinctive morphological features, fatty acid profiles in acorns may reflect the delineation of Quercus at the infrageneric level in the absence of ecological factors, climatic and seasonal variations. For example, remarkable variation in linolenic acid contents were experienced in the different genotypes, cultivars and mutant lines in the seed samples of Linum species [24]. Fatty acid patterns revealed lower intraspecific variability and higher taxonomic resolution [25]. Thus, characteristic fatty acid profiles of each taxa may reflect different pathways involved in oil biosynthesis and accumulation. The percentage of palmitic, stearic, oleic, linoleic and α-linolenic acid, total saturated and unsaturated fatty acids and their ratios in acorns appear to be useful parameters in the delimitation of Quercus taxa, especially for Quercus species [17, 18]. Our observations support that concept. Palmitic, oleic, α-linolenic acid concentrations and total percent of saturated and mono-unsaturated fatty acids are distinctive characteristics between Q. brantii and Q. trojana. Quercus brantii is a polymorphic species that hybridizes with Q. libani which is one of the most characteristic oak-trees of east Anatolia and Iraq. But, hybridization between Q. brantii and Q. trojana has not been reported [4, 9]. Our results for crude fat content are in agreement with the findings on Q. brantii from Mediterranean climate of Iran, but, palmitic and oleic acid show relatively lower, and linoleic acid higher levels [20]. In addition, α-linolenic acid and total percent of saturated and poly unsaturated fatty acids have considerably different levels between Q. libani and Q. trojana. The latter species is reported to vary little in its characters and is closely related to Q. libani. The distribution areas of these two oaks touch in the mountains of the central Taurus [4, 9]. However, Q. libani is completely different from Q. brantii for all examined parameters. This species, having more limited distributional area between Q. brantii and Q. trojana, may be better classified as a subspecies of Q. trojana on the basis of some morphological features [9]. However, specific separation was expressed as being merited in the "Flora of Turkey"; because of has fewer distinctive variants than most Turkish species of oak and is clearly related to the West Anatolian Q. trojana [4].

Although there is considerable similarity among fatty acid profiles between these species, the most distinctive parameters are the ratio of mono:poly unsaturated fatty acids in total and linoleic:α-linolenic acid concentrations. These species are close to Q. ithaburensis subsp. macrolepis for many morphological features and hybridize frequently with it. They also express considerably different saturated, monounsaturated, total unsaturated fatty acid concentrations and ratios of saturated and unsaturated fatty acids. In addition, hybridizations are frequent among Q. cerris from the same section [4]. Palmitic and stearic acid levels in both taxon present similar values. Quercus cerris is widespread, variable and moderately mesophyllic species [9], and hybridizes with Q. pubescens and Q. libani [4]. Quercus cerris and Q. pubescens, from different sections show some similarity for general fatty acid profiles. Both varieties of Q. cerris exhibit highly stable concentrations within variety for palmitic acid, even though samples were collected from three different populations. Significant differences for this trait may then be a useful tool at the variety identification level in this species.

Total saturated, total unsaturated fatty acid concentrations and their ratio also show considerable stability between two varieties of Q. cerris. On the contrary, remarkable fluctuations within each variety for individual unsaturated fatty acid were observed according to the populations from different regions. Unsaturated fatty acid levels may be more susceptible to environmental conditions or population characteristics in Q. cerris. Significant difference of fatty acid profiles in both varieties of Q. cerris apart from general characteristics of its section may account for some of the genetic distance. This species covers the largest distributional areas in Anatolia compared to other members of the section. It was reported that Q. hartwissiana could be placed between Q. robur and Q. petraea in its morphological features and quite often there is no distinction between them [9]. In our observations, Q. petraea ssp. petraea differs from the others with higher mono-unsaturated and lower poly unsaturated fatty acid contents resulting in the high ratio of mono:poly unsaturated fatty acids. Additionally, the ratio of linoleic:α-linolenic acid shows a considerably higher value in this subspecies. This taxon has differences in fatty acid composition compared to all taxa from its section (section Quercus), except for Q. vulcanica endemic to Turkey. However, from the latter it differs in higher palmitic and lower stearic acid levels. The ratios of mono:poly unsaturated and linoleic α-linolenic acid are also distinctive parameters. All other traits examined in both taxa show considerable similarity in general. Quercus vulcanica is reported to be similar to Q. petraea ssp. pinnatiloba, but also to Q. frainetto. It differs from Q. petraea ssp. pinnatiloba in the flat scales of the cupule, secondary leaf lobes and intercalary veins. From Q. frainetto it differs in the longer petioles and the leaves evenly distributed over the shoots [4]. Similarity in higher stearic acid levels and considerably differences for almost all parameters were observed between Q. vulcanica and Q. frainetto, a typical species of forest tree growing fast and characterized by a great resistance to drought [9]. A high level of saturated fatty acids were also examined in Q. frainetto and Q. pubescens apart from other members of the section. Quercus pubescens belongs to the most xeric oak species in west Anatolia growing as xerothermic scrubs in anthropogenic steppe or semi-steppe terrain; rarely in macchie [4, 9]. On the other hand, Q. coccifera is thought to be a dominant member of the phryagana and macchie flora of Turkey [26]. This species is related to Q. ilex and Q. aucheri which are endemic to Turkey, and it exhibited very similar values within its section and the highest saturated fatty acid concentrations among all examined taxa. But, high oleic and low linoleic acid levels were found in Q. ilex, a typical component of evergreen forests and maquis. All examined taxa having xerophytic nature mainly contained higher concentrations of saturated fatty acids. But, the information on the effect of environmental variation of concentrations for the fatty acids growing in different regions is needed to evaluate the utility of this characteristic. The lowest oil contents in species from section Ilex were associated with partly xerophytic characteristic. Different acorn samples of Q. brantii collected from different climatic zones in the Zagrossian region of Iran were reported to show no significant difference in fatty acid composition, but, crude fat contents varied significantly [20]. In addition, the acorn specimen of Q. cerris var. cerris from a Mediterranean climate resulted the lowest amount of total oil in our study. Similar total oil contents were detected for both geographically close populations of Q. cerris var. austriaca having similar ecological and climatic conditions. But, some differences for total oil contents were obtained in the acorn samples collected from individual population of Q. cerris. On the other hand, the relative percentages of fatty acids may be more stable and determinative in order to understand taxonomic and phylogenetic relations in Quercus. Many studies have reported phylogenetic relationships are associated with differences in the fatty profile of the seed oils [27–29]. In this study, similar fatty acid compositions for related species plus their morphology may account for its common ancestral stock. In that regard, each close related taxa examined here shows typical fatty acid composition. It is also possible to delineate taxa for fatty acid profile at the section level. Some ratios between saturated and unsaturated fatty acids have significantly difference implying its reliable characteristic for sectional and specific delineations in Quercus. Segregation of the three sections could be accomplished with using all ratios as a marker set. These results also provide an insight to the source potential of Turkish Quercus acorns based on fatty acid analysis.

Acorn oils are reported to have good nutritional quality with a flavor comparable to olive oil [30, 31]. Crude fat content of the acorns can be compared generally with the values of some grains [10, 20]. The highest levels of total oil are in Q. brantii and Q. trojana. Four species examined here from section Cerris (except for Q. cerris) show similar composition with olive, canola and hazelnut oils. Both varieties of Q. cerris correspond with cotton oil composition generally. The oil of Q. hartwissiana especially from section Quercus has very similar compositional fatty acid characteristics to sunflower oil. In addition, Q. pubescens has similarity with maize oil. Quercus hartwissiana and Q. robur ssp. robur show also common characteristics in fatty acid composition compared with soyabean oil. It is possible to say that sesame oil may be compared in general with Q. petraea ssp petraea and Q. vulcanica. Acorns of Quercus may be used as potential alternative food reserves for crude vegetable oil.

Mature acorn samples of examined taxa growing in their natural habitats showed a characteristic fatty acid composition. Our results suggest that these parameters may be valuable tools to segregate taxa at the sectional and specific level. Significantly different concentrations, critical values, total percentages and the relative ratios especially of saturated and unsaturated fatty acids seem to be useful for the characterization of Quercus at the infrageneric level in accord with established phylogenic associations. But, expanded evaluation is needed to understand how much of the variation of these parameters in the genotypes is valid for any examined species collected from different regional localities and climatic conditions.