Soil parameters and bioclimatic characteristics affecting essential oil composition of leaves of Pistacia lentiscus L. from València (Spain)

The variability of the soils found in an area together with the diversity of the bioclimatic parameters will affect the chemical profile of plant species, in our case Pistacia lentiscus L. The aim of this work is to analyse the bioclimatic characteristics and soil parameters affecting the essential oil (EO) composition of the leaves of the Pistacia lentiscus L. growing in València (Spain). The EO compositions of dried leaves of Pistacia lentiscus L. from five sampling sites with noticeable differences in soil and climate features were analysed by GC(MS) and GD(FID). The main bioclimatic and soil parameters were also determined in order to relate them to oil composition. α -pinene (2.8-39.2%), myrcene (0.6-59.3%), β -phellandrene (0.0-42.8%), germacrene-D (3.9-20.4%) and terpinen-4-ol (0.2-19.0%) were found to be the major compounds showing a high variability. The five sampling sites presented well-defined bioclimatic and edaphic characteristics that could be related to specific EO profiles. The results suggested that the EO composition of P. lentiscus L. depends more on the edaphic factors interacting with bioclimate conditions than on the geographical origin itself. Some general trends could be established based on the results: the Calcaric Arenosol (Saler) with a high evapotranspiration index was associated with a high sesquiterpenic fraction, (germacrene-D and β -caryophyllene, mainly). The Luvisol (Borrell and L’Ull), with high amounts of α + β -pinene, were found to be related to sub-humid bioclimatic conditions and clayey and fertile soils (high CEC and OM levels). This last requirement was also observed to be related to the myrcene content


Introduction
Soil is a basic component for the development of plants and ecosystems.There is a partially bidirectional relationship between the vegetation and the characteristics of the soil; soils affect the vegetation and the vegetation itself conditions the soil properties.For this reason, the variability of the soils found in a given area will affect the proportion of both the inorganic and biochemical compounds of specific species.The bioclimatic variability that we find in a study area will also be another factor that will affect said composition.

PALABRAS CLAVE
The essential oil (EO) composition of its aerial parts (leaves and twigs, fruits and galls) has been studied extensively.Recent data from many studies carried out in different Mediterranean countries report the major components for each type of terpenic group (Llorens-Molina et al. 2015), showing an extensive chemical variability affecting both major compounds and total amounts of terpene types.This research has focused mainly on fresh or dried leaves and twigs.

Plant material
Samples of approximately 200 g of fresh material of leaves and twigs were harvested from all around 20 individual shrubs randomly selected growing in five different locations (Table 1) (4 individuals x location).A voucher specimen was placed in the Herbarium of Mediterranean Agroforestry Institute at the Universitat Politècnica de València (Spain) (VALA 9581-9585).After removing damaged and yellowish material, each whole sample was dried in a dry and shady place at room temperature for 20 days just before EO extraction.

Essential oil extraction
After homogenizing each sample, 50 g was subjected to hydrodistillation for 3 h in a Clevenger-type apparatus.EO was swept away with 2 mL of dichloromethane (DCM).After removing the aqueous phase, the EO extract was dried over anhydrous sodium sulphate.The solvent was removed under reduced pressure, using a rotary evaporator Laborota 4001 (Heidolph Instruments, Schwabach, Germany) at 20 ºC.Solutions containing 10 µL EO/1 mL of DCM were kept in the dark at -18 ºC in sealed vials until analysis.

GC analysis
The analysis was carried out by gas chromatography with flame ionization detector (GC-FID) in a Clarus 500 GC (Perkin-Elmer Inc.) chromatograph equipped with a capillary column ZB-5 (30 m x 0.25 mm i.d.x 0.25 mm film thickness; Phenomenex Inc.).The injection volume was 1 µL.The GC oven temperature was programmed from 50 ºC to 250 °C at a rate of 3 °C min -1 .Helium was the carrier gas (1.2 mL min -1 ).Detector temperature was set at 250 ºC.The percentage composition of the EO was computed from GC peak areas without correction factors by means of the software Total Chrom 6.2 (Perkin-Elmer Inc.).Analysis by GC-MS was carried out using a Clarus 500 GC-MS (Perkin-Elmer Inc.) equipped with the same column and programmed with the same oven temperatures mentioned above.Ionization source temperature was set at 200 ˚C and 70 eV electron impact mode was employed.MS spectra were obtained by means of total ion scan (TIC) mode (mass range m/z 45-500 uma).The total ion chromatograms and mass spectra were processed with the software Turbomass 5.4 (Perkin-Elmer Inc.).Kovats retention indices were determined by injection of C8-C25 nalkanes standard (Supelco®) under the same GC conditions.The essential oil components were identified by comparison of their mass spectra with those of computer library NIST MS Search 2.0 and available data in the literature (Adams 2007).The identification of the following compounds was confirmed by comparison of their experimental Kovats RI with those of authentic reference standards (Sigma-Aldrich®): α-pinene, β-pinene, camphene, myrcene, camphor, terpinolene, borneol, terpinen-4-ol, bornyl acetate and linalool.

Statistical analysis
The statistical analysis was carried out by means of one-way ANOVA, Principal Component

Results of soil parameters
Soil parameter analyses from the different sampled sites are displayed in As displayed in Figure 1, there were statistically significant differences (P < 0.05) between active lime, organic matter, cation exchange capacity,  exchangeable sodium percentage, sand, silt and clay when the one-way ANOVA was applied.
High individual variability means among the different within sampling site could be observed in every one of the analysed properties (Table 3).The most relevant ones were submitted to ANOVA analysis (Figure 1).Active lime (AL) showed statistically significant differences (P < 0.001), with the higher values in Lliria and L'Ull.With regard to the OM level, two groups of sampling sites were found: Saler and Segart accounted for a significantly lower level (P < 0.01) with respect to Borrell, Lliria and L'Ull.
The Cation Exchange Capacity (CEC) showed high variability, with Borrell, Lliria and L'Ull being significantly more fertile.In the same way, EC of saturation extracts presented statistically significant differences (P < 0.001).The Saler value was significantly higher than the other ones.
As for sand, silt and clay, significant differences were also found (P < 0.01, P < 0.01, P < 0.01, respectively).Obviously, the highest level of sand (and the lowest level of clay) was found in Saler (coastal site).By contrast, Borrell and Segart were found to be the richest in clay, whereas Lliria and L'Ull had a higher silt content.
A clear dichotomy was found between Saler and the rest of sampling points related to ESP, which showed a significantly higher content.

Essential oil composition
The essential oil composition of the 20 individuals sampled in the studied sites is displayed in Table 4.The major compounds were found to be similar to the ones reported in the literature, but great individual differences were noted even within the same location.Of the total compounds, hydrocarbon monoterpenes accounted for between 35.8% (El Saler, M1) and 74.9% (Segart, M1), oxygenated monoterpenes from 0.2% (Segart, M1) to 22.5% (El Saler, M1) and hydrocarbon sesquiterpenes from 12.3% (L'Ull, M4) to 43.9% (Lliria, M1).Oxygenated sesquiterpene and diterpene fractions accounted for no more than 3.1% (El Saler, M1) and 0.8% (Lliria, M4), respectively.
Despite the great individual variability, some statistically significant differences (P < 0.05) were found when one-way ANOVA was applied in order to compare the amounts of major and grouped compounds according to the individual's sampling points (Table 5).The hydrocarbon monoterpene fraction, as a whole, was significantly lower in Saler (and Lliria, without achieving statistical significance).Within this fraction, the amount of myrcene showed a different tendency, with a significant increase in Lliria and Segart (this last one with no significant differences).The content of α-pinene was significantly higher in Borrell and L'Ull compared to Lliria and Segart (β-pinene also showed a similar variation).Likewise, limonene was significantly lower in Lliria and Segart with respect to the rest of sampling sites, whereas β-phellandrene was not found in Lliria.The percentage of terpinen-4-ol was significantly higher in Saler with respect to Borrell.As for the sesquiterpenic fraction, Lliria and Saler presented a significant increase if compared to the rest of locations.  Constituents listed in order of increasing retention indices (RI) for each group of compounds.Unidentified components accounting for less than 0.5% are not reported. 2Kovats retention indices referred to n -alkanes, determined on a DB5 capillary column. 3Method of identification: all the reported compounds have been identified by comparison of GC-MS data of NIST computer mass library and Kovats RI with those reported by R. P. Adams (35).α-pinene, β-pinene, myrcene, α-terpinene, p-cymene, limonene, terpinolene, (Z)-thujone, (E)-thujone, camphor and terpinen-4-ol were identified by comparison of RI with those of authentic samples. 4Not detected. 5Percentage values less than 0.1% are denoted as tr (traces).
[ SOIL PARAMETERS AND BIOCLIMATIC CHARACTERISTICS AFFECTING ESSENTIAL OIL COMPOSITION OF LEAVES OF PISTACIA LENTISCUS L. FROM VALÈNCIA (SPAIN) ] Table 4. Essential oil composition for the 20 (cont.) interpretation connecting this trend with the oil composition was obtained by Principal Component Analysis (PCA) with compounds defining the main chemotypes of P. lentiscus L. EO reported in the literature.These results are graphically represented for clarity (Figure 3) and subjected to discussion.
The obtained plot according to axes 1 (24.9% of the variance) and 2 (21.0% of the variance) grouped the samples in such a way that more clear linkages between bioclimatic and soil parameters and EO chemical profiles were observed.Thus, Borrell, with the highest fertility and subhumid conditions was characterized by the pinene chemotype, whereas Lliria, with a higher rate of active lime, lower clay content and drier climate was characterized by the high proportion of myrcene, although it should be taken into account that the levels of this compound have shown great variability between samples, both in Lliria and in Segart.
Likewise, the Giacobbe index leads to the consideration that this location is susceptible to water stress.In this respect, it is worth mentioning the reported data from Aissi et al. (2016), in which the higher levels of myrcene were found in upper semi-arid climates and calcareous soils.

Discussion
In order to define the distinctive edaphic and bioclimatic features of the five sampling locations affecting EO composition, a discriminant analysis based on the main soil and climate parameters and the main oil compounds was performed (Figure 2).Three functions with P < 0.05 were defined with λ Wilks values: 2.9869 E-7, 7.4627 E-5 and 7.386 E-3, respectively.A first distinction can be pointed out according to the content of active lime.Soils of calcareous nature were found in three sampling sites (Saler, L'Ull, Lliria), while those of Segart and Borrell have lower carbonate contents.
Within each of these groups, axis 1 marks an increasing fertility owing to the higher levels of active lime, clay and CEC (cation exchange capacity).On the other hand, it is worth mentioning the salinity observed in Saler soils, evidenced by its conductivity value, as well as the high chloride and sodium content because of its proximity to the sea.From the bioclimatic point of view, axis 1 also marks the difference between three of the locations Segart and Saler), drier than Lliria, whereas axis 2 separates L'Ull and Borrell by the increment in organic matter, clay contents and the subhumid climate.Further   2016), α + β-pinene and myrcene were the predominant compounds instead of the sesquiterpene fraction.Since in this work, the detailed soil parameters are not considered, this fact leads to the suggestion that the major influence of soil fertility is a consequence of the high rates of CEC, OM and clay.
Samples from Segart were rich in the monoterpenic fraction other than α + β-pinene and terpinen-4-ol.This location was characterized by its lower ombrothermal index than Borrell and L'Ull, related to a lower average annual rainfall and a high summer drought.Average values of terpinen-4-ol were also found in a moderately high proportion (2.5-10.5%).These amounts were similar to those reported by Llorens-Molina et al. (2015), in which two sites coinciding with those studied in this work (Segart and L'Ull) were 5.1-6.7%compared to 7.2-8.3%,respectively.No significant differences were found between these sites for this compound in both studies.These values also agree with those referred by Aissi et al. (2016) (2.7-9.6%), with the higher values in upper semi-arid locations.
On the other hand, with respect to Saler, it also showed a major influence of evapotranspiration and EC.It is worth mentioning the specific features of sample 4 from Segart, which can be considered as a transition between siliceous and calcareous soils.Sample 4 has a greater content of active lime than the other three and is slightly more fertile (higher values of EC (0.8 dS/m) and has more organic matter (OM) (2.2%).According to the aforementioned tendency, it also showed a higher myrcene content than the Lliria and Segart oils.

Conclusions
The five sampling locations presented well defined bioclimatic and edaphic characteristics that could be related with specific EO profiles.These results suggest that the EO composition of P. lentiscus L. depends on bioclimatic and edaphic factors rather than on the geographical origin of the samples.Even more, the edaphic factor, seems to have more influence than the bioclimatic factor on EO composition.

Some general trends have been established
based on the obtained results: the Calcaric Arenosol (Saler) with a high evapotranspiration index could be associated with a high sesquiterpenic fraction, (germacrene-D and β-caryophyllene, mainly).For the Luvisol (Borrell and L'Ull) high amounts of α+β-pinene were found to be related with sub-humid bioclimatic conditions and clayey and fertile soils (high CEC and OM levels).This last requirement was also observed in relation to the myrcene content, but with drier climatic conditions and

Figure 1 .
Figure 1.Mean values and LSD Fisher intervals (95%) of active lime, organic matter, cation exchange capacity, exchangeable sodium percentage, sand, silt and clay of soil at different sampling sites where the Pistacia lenticus L-leaves were sampled.Differences among mean values with different letters are statistically significant (P < 0.05).

Table 1 .
Location of the sampling sites

Table 2 .
Bioclimatic indices of P. lentiscus L. corresponding to the sampling locations : yearly average temperature.m: average temperature of the minima of the coldest month of the year.M: average temperature of the maxima of the coldest month of the year.T max: average temperature of the warmest month.T min: average temperature of the coldest month.Tm: average temperature of each month.P: total yearly precipitation (mm).It a : Thermicity index, 10 × (T + M + m).Ic b : Continentality index, T max -T min.Io c : Ombrothermic index, (P/12) × 10/ΣTm.Ppv d : Summer precipitation in mm of the three consecutive warmest months in the year.Ttve: Value in tenths of a degree resulting from the sum of the monthly average temperatures of the three consecutive warmest months of the year.ETo f : Evapotranspiration of Pemman-Monteith.IG Tg (Giacobbe index): (Pjun + Pjul + Paug)/Tm of the warmest month.[LLINARES J. V., LLORENS-MOLINA J. A., MULET J. & VACAS S. ]

Table 3 .
The

Table 3 .
Soil parameters from the different sampling sites

Table 4 .
Essential oil composition for the 20 samples (cont.)