Acta Botanica 2-2016 - za web.indd 266 ACTA BOT. CROAT. 75 (2), 2016 Acta Bot. Croat. 75 (2), 266–271, 2016 CODEN: ABCRA 25 DOI: 10.1515/botcro-2016-0026 ISSN 0365-0588 eISSN 1847-8476 Short communication Infl uence of soil traits on polyphenols level in Moltkia petraea (Tratt.) Griseb. (Boraginaceae) Dario Kremer1, Renata Jurišić Grubešić1, Dalibor Ballian2, Danijela Stešević3, Ivan Kosalec1, Jadranka Vuković Rodríguez1, Marija Vukobratović4, Siniša Srečec4* 1 Faculty of Pharmacy and Biochemistry, University of Zagreb, A. Kovačića 1, Zagreb HR-10000, Croatia 2 Faculty of Forestry, University of Sarajevo, Zagrebačka 20, BIH-71000, Bosnia and Herzegovina 3 Faculty of Natural Sciences and Mathematics, University of Montenegro, Džordža Vašingtona bb, 81000 Podgorica, Montenegro 4 Križevci College of Agriculture, M. Demerca 1, HR-45260 Križevci, Croatia Abstract – The Illyric–Balkan endemic species Moltkia petraea (Tratt.) Griseb. is very interesting as a poten- tial horticultural and medicinal plant. The aim of this study was to investigate soil conditions of M. petraea habitats, the phenolic content in plant parts, and the infl uence of soil properties on the phenolic contents. The results were evaluated using Spearman rank order correlations. Analyzed soil samples contained very low to intermediate levels of physiologically active phosphorus, but were very rich in potash. Organic matter content of soil was high. Phenolic compound content was higher in leaves than in fl owers or stems. The analyses showed that M. petraea possesses considerable quantities of phenolic compounds and has no specifi c de- mands for particular soil conditions. A negative correlation was found between soil phosphorus content and total phenols content in leaves and stems, and with the total phenolic acids content in fl owers. Organic matter in soil also found to have a negative infl uence on total tannins content in stems. Among the tested geographi- cal locations, the Mljet population showed a higher degree of separation from the remaining locations. Keywords: endemic, Moltkia petraea, phenolic compounds, plant habitats, soil traits * Corresponding author, e-mail: ssrecec@vguk.hr Introduction The north-western Balkan Peninsula, particularly the area of Dinaric Alps, though poorly investigated, is known to be very rich in endemic plant species. From the biogeo- graphical perspective, the Dinaric Alps mountain complex towards the north and the Mediterranean region towards the south are hotspots of endemicity. This is the meeting point of two large phytogeographic regions: the Euro Siberian – North American and the Alpine – high Nordic regions (Ređić et al. 2011). Moltkia petraea (Tratt.) Griseb. is a typical representa- tive of an endemic plant with horticultural and medicinal potential. It is an endemic, lithophytic, xerothermic Illyric– Balkan species distributed along the Adriatic Coast in Croa- tia, Bosnia and Herzegovina, Montenegro, Albania, and Greece, mostly in Mediterranean and sub-Mediterranean regions, at altitudes between near sea level to 2000 m. It is a dense, dwarf shrub that grows up to 40 cm and blooms from May to July with very decorative, deep violet-blue, tubular fl owers (Šilić 2005). M. petraea is a strictly protect- ed and threatened species in Croatia (Anonymous 2013). Among biologically active compounds, phenolic com- pounds have attracted a great deal of public and scientifi c interest due to their health-promoting effects as antioxi- dants. The content of biologically active compounds varies greatly among species, including closely related species, making them useful chemotaxonomic markers. The differ- ences between populations of a species could very often be signifi cant (Dunkić et al. 2012). These differences could be associated with the region of origin, growth phase, habitat condition, and seasonal environmental variability, which encompass biotic and abiotic factors (Buchwald et al. 2015, Ramegowda and Senthil-Kumar 2015). According to Dap- kevicius et al. (2002), it is possible to increase the content of pharmacologically desirable compounds through agri- culture techniques, such as irrigation or using photo biore- actor systems. On the other hand, Buchwald et al. (2015) did not observe a statistically signifi cant infl uence of min- eral fertilization on the level of main active compounds. To date, chemical compound contents in M. petraea have only been investigated by Zovko Končić et al. (2010). The aim of this study was to investigate the soil traits of M. petraea habitats and to evaluate the infl uence of soil PHENOLIC SUBSTANCES IN MOLTKIA PETRAEA ACTA BOT. CROAT. 75 (2), 2016 267 properties on phenolic compound accumulation. The evalu- ation of soil traits in native populations of M. petraea is the fi rst step towards the determination of M. petraea as a plant with possible horticultural signifi cance. Materials and methods Plant material Samples of Moltkia petraea (Tratt.) Griseb. were col- lected during the blooming period in June and July of 2011 at ten locations along Croatian Adriatic coast, and in the Di- narides and Durmitor mountain ranges of Montenegro and Bosnia and Herzegovina (On-line Suppl. Fig. 1). Altitude and latitude of each habitat locality were determined by a GPS locator (Tab. 1). Voucher specimens of herbal material were deposited in the Fran Kušan Herbarium of the Depart- ment of Pharmaceutical Botany with Fran Kušan Pharma- ceutical Botanical Garden at the Faculty of Pharmacy and Biochemistry, University of Zagreb, Croatia. Above-ground parts of several dozen randomly selected plants were harvested from mature plants on a dry day and mixed to obtain the randomly selected sample. Samples were air-dried for three weeks in a well-ventilated room at 60% relative humidity and room temperature (22 °C), sin- gle-layered and protected from direct sunlight. Air-dried samples were placed in double paper bags labelled with the sample number, and stored in a dry place at room tempera- ture (22 °C; 60% of humidity) protected from light for fi ve months until analysis. Soil sampling Simultaneously with the sampling of plant material, soil was also collected in situ from all localities. Soil sampling was provided with a soil probe, from depths of 0–15 cm, due to the very shallow depth of the undeveloped karst soil. Soil samples were prepared for chemical analyses accord- ing to the ISO 11464:1994 method (Pernar et al. 2013). Soil analyses The pH of soil in H2O and 1 M solution of KCl were analyzed according to the ISO 10390:1994 method (Pernar et al. 2013), while organic matter content (%) was analyzed according to Tjurin’s method (Lal et al. 2002). Total nitro- gen (%) was analyzed according to the ISO 11261:1995 method (Pernar et al. 2013), total content of CaCO3 was an- alyzed according to the Scheibler method (Tatzber et al. 2007), and contents of physiological active phosphorus and potash, calculated in mg of P2O5 and K2O per 100 grams of soil, were analyzed according to the Egner-Riehm-Domin- go’s ammonium lactate method (Page 1982). All soil analy- ses were performed in triplicate. Phenolic compounds analyses Total polyphenols and tannins contents were determined according to Schneider (1976). This procedure is based on a reaction with Folin-Ciocalteu’s phenol reagent (FCR) and spectrophotometric determination of total polyphenols and tannins (indirectly, after precipitation with casein) at 720 nm. Tannin was used as the standard substance. Tab. 1. Habitats of Moltkia petraea and collection data of researched plant and soil samples. States and locality of sampling Voucher No. Latitude; Longitude Altitude (m) Abbreviation Croatia Omiš HFK-HR-113-2011 42°26’17’’ N; 16°42’01’’ E 30 Om Vošac HFK-HR-117-2011 43°18’46’’ N; 17°03’07’’ E 1295 Vo Mljet HFK-HR-122-2011 42°42’35’’ N; 17°40’36’’ E 225 Ml Sniježnica HFK-HR-133-2011 42°34’08’’ N; 18°21’28’’ E 1152 Sn Bosnia and Herzegovina Diva Grabovica HFK-HR-124-2011 43°35’59’’ N; 17°41’04’’ E 251 DG Drežnica HFK-HR-125-2011 43°31’48’’ N; 17°42’22’’ E 208 Dr Rujišta HFK-HR-123-2011 43°27’30’’ N; 17°57’02’’ E 964 Ru Rakitnica HFK-HR-127-2011 43°34’39’’ N; 18°05’20’’ E 768 Ra Montenegro Orjen HFK-HR-142-2011 42°33’45’’ N; 18°47’47’’ E 760 Or Lovćen HFK-HR-146-2011 42°23’44’’ N; 18°48’28’’ E 1365 Lo KREMER D., JURIŠIĆ GRUBEŠIĆ R., BALLIAN D., STEŠEVIĆ D., KOSALEC I., ET AL. 268 ACTA BOT. CROAT. 75 (2), 2016 The total fl avonoid content (quercetin type) was deter- mined using the method according to Christ and Müller (1960). This procedure includes hydrolysis of glycosides, ex- traction of total fl avonoid aglycones with ethyl acetate and complex formation with AlCl3 at 425 nm. The yield was cal- culated as quercetin according to the following expression: Total fl avonoids (%) = A × 0.772 / b A = absorbance; 0.772 = conversion factor related to spe- cifi c absorbance of quercetin at 425 nm; b = mass of dry herbal material (g) Total phenolic acids content was determined according to the monograph of Rosmarini folium in European Pharmaco- poeia (2007). Phenolic acids in the extracts were measured spectrophotometrically at 505 nm (three independent analy- ses), using the nitrite-molybdate reagent of Arnow, in a sodium hydroxide medium, and the percent of their content, expressed as rosmarinic acid, was calculated from the expression: Total phenolic acids (%) = A × 2.5 / m A = absorbance; 2.5 = conversion factor related to specifi c absorbance of rosmarinic acid at 505 nm; m = mass of the substance to be examined (g), taking the specifi c absor- bance of rosmarinic acid to be 400. The contents of total polyphenols, tannins, total fl avo- noids, and total phenolic acids were evaluated from three independent analyses and were expressed as the percentag- es of dry mass of herbal material. A UV/Vis spectropho- tometer Agilent 8453 (Agilent, Germany) with PC-HP 845x UV-Visible System (Agilent, Germany) and 1 cm quartz cells was used for all absorbance measurements. Statistical analysis Statistical comparisons of phenolic compound contents among investigated populations and between plant organs were conducted using one-way ANOVA followed by Schef- fe’s post-hoc test at the P ≤ 0.05 level. Prior to ANOVA, data was transformed using angular (i.e. arcsin) transformation. The results were evaluated using multivariate analysis. Principal component analysis (PCA) calculation was based on the correlation matrix between the values of the charac- teristics, meaning that the contribution of each variable was independent of the range of its values. To confi rm the re- sults of the PCA, the unweighted pair-group method with arithmetic mean (UPGMA) with Euclidean distance (DE) was conducted. UPGMA generally yields results that are the most accurate for classifi cation purposes. Interactions between soil traits and the content of differ- ent biologically active compounds were analyzed using the Spearman rank order correlation matrices. Prior to analysis, data were transformed using angular transformation. Statis- tical analyses were performed using the Statistica 7 soft- ware package (StatSoft Inc., Tulsa, OK, USA). Results and discussion Soil properties Chemical properties of soil where native populations of M. petraea grow are presented in Tab. 2. Soil samples mea- sured in a 1 M solution of KCl showed a neutral to slightly alkaline reaction. Soils from all sites contained a very low to intermediate amount of physiologically active phospho- rus (P2O5), but were very rich in potash (K2O). Soil organic matter content was very high at all localities. Due to the very shallow depth of the undeveloped karst soils, the high organic matter content was related to mulch not humus (Tab. 2). According to descriptive statistics, very high vari- ability of CaCO3, organic matter and potash contents was obtained in soil samples from different sites, indicating that Tab. 2. Variability of chemical properties of soil in different habitats of Moltkia petraea. AL – ammonium-lactate, BLQ – below limit of quantifi cation. Habitat (locality abbreviation) pH CaCO3 (%) Organic matter (%) Nitrogen (%) AL-method (mg/100 g) H2O 1 M KCl P2O5 K2O Omiš 7.78 7.33 29.93 5.04 0.41 8.50 26.61 Vošac 7.70 7.32 39.84 7.89 0.64 3.25 68.14 Mljet 7.68 7.25 27.10 12.85 0.48 2.79 61.36 Sniježnica 7.45 6.96 2.63 19.89 0.57 3.25 45.42 Diva Grabovica 7.33 6.89 BLQ 35.69 0.68 4.68 46.10 Drežnica 7.66 7.31 50.59 8.35 0.68 2.43 38.72 Rujišta 7.76 7.21 34.10 4.19 0.33 1.22 26.88 Rakitnica 7.04 6.59 BLQ 36.26 0.70 5.29 36.45 Orjen 7.24 6.97 6.45 13.76 0.41 11.09 125.00 Lovćen 7.71 7.17 17.97 9.67 0.40 1.55 27.81 Mean 7.54 7.10 26.08 15.36 0.53 4.41 50.25 Stand. dev. 0.26 0.24 16.35 11.78 0.14 3.16 29.82 Var. 0.07 0.06 267.32 138.68 0.02 9.98 889.28 Coef. of var. 3.39 3.40 62.70 76.67 26.37 71.70 59.35 Stand. error 0.08 0.08 5.78 3.72 0.04 0.10 9.43 PHENOLIC SUBSTANCES IN MOLTKIA PETRAEA ACTA BOT. CROAT. 75 (2), 2016 269 M. petraea has no specifi c demands regarding particular soil conditions. Phenolic compounds The contents of total polyphenols (TP), tannins (T), to- tal fl avonoids (TF), and total phenolic acids (TPA) in leaves, fl owers, and stems of the investigated M. petraea populations are presented in Tab. 3. TP, TF and TPA con- tents were highest in leaves, while T content was highest in fl owers in most populations. In general, the concentrations of the analyzed bioactive compounds were lowest in stems. The smallest differences between populations were ob- served in TF content. It can be concluded that TF content in M. petraea is under a lesser infl uence of habitat conditions than the contents of TP, T and TPA. The results confi rmed that the content of biological ac- tive compounds in analyzed plants varied between popula- tions. According to Dunkić et al. (2012), the contents of TP, T and TF in aerial parts of the investigated populations of Satureja montana L. and S. subspicata Vis. (Lamiaceae) re- vealed a statistically signifi cant within-species difference, depending on the locality and plant organ used for determi- nation. Variations between locations could be ascribed to biotic (vermin, alleopathy, diseases) and abiotic (climate, soil, fertilization) factors (Young et al. 2005). The content of biologically active compounds also depends on plant age and harvesting time (Kołodziej and Sugier 2013). With regard to the analyzed phenolic compounds, the PCA and UPGMA separated the investigated M. petraea populations as presented in Fig. 1. The fi rst principal com- ponent (PC 1) explained 46.1% of the total variance, the second 25.9%, and the third component 13.9%. Thus, the fi rst three components accounted for 85.9% of the variance, emphasizing the usefulness of the PCA. The most similar populations were Vo and Lo, Om and DG, and Sn and Or, respectively (Fig. 1A). The population Ml showed a higher Tab. 3. Content of total polyphenols, tannins, total fl avonoids and total phenolic acids in leaves, fl owers, and stems of Moltkia petraea expressed as mean ± standard deviation of the three independent analyses. Capital letters and symbols in superscript denote difference between populations for leaves (A, B, D, E, F, G, H, I, J), fl owers (K, L, M, N, O, P, R, S, T, U) and stems (V, W, X, Y, Z, “, Γ, Λ, Π, Σ) related to certain investigated trait. Sample Plant part Total polyphenols (%) Tannins (%) Total fl avonoids (%) Total phenolic acids (%) Omiš leaves 4.29±0.10A 1.04±0.05A 0.42±0.00A 2.06±0.06A Omiš fl owers 3.91±0.02K 1.06±0.02K 0.19±0.01K 1.72±0.02K Omiš stems 3.33±0.09V 1.01±0.05V 0.04±0.01V 1.76±0.03V Vošac leaves 5.78±0.09AB 1.95±0.01AB 0.44±0.00AB 3.43±0.02AB Vošac fl owers 6.07±0.07KL 2.45±0.05KL 0.28±0.00KL 3.12±0.08KL Vošac stems 4.44±0.03VW 1.53±0.07VW 0.15±0.00VW 2.20±0.02VW Mljet leaves 5.88±0.11AD 0.39±0.09ABD 0.37±0.01ABD 3.00±0.02ABD Mljet fl owers 5.53±0.07KLN 1.17±0.02LMN 0.21±0.00LMN 2.58±0.04KLMN Mljet stems 5.53±0.06VWXY 0.16±0.02VWXY 0.12±0.00VWXY 2.92±0.04VWXY Sniježnica leaves 4.62±0.02BDE 1.07±0.02BDE 0.39±0.00ABDE 1.89±0.06ABDE Sniježnica fl owers 4.64±0.07KMNO 1.20±0.02LMO 0.18±0.00LMO 1.71±0.06LMNO Sniježnica stems 3.26±0.03WXYZ 0.53±0.04VXYZ 0.31±0.00VWXYZ 1.56±0.02VWXYZ Diva Grabovica leaves 3.97±0.02BDF 0.14±0.03ABEF 0.84±0.01ABDEF 2.01±0.05BDF Diva Grabovica fl owers 4.46±0.02KLMNP 0.81±0.09LMNOP 0.43±0.02KLMNOP 2.20±0.02KLMNOP Diva Grabovica stems 3.37±0.05WX” 0.88±0.05WXYZ” 0.09±0.02VWXYZ” 1.70±0.06WY” Drežnica leaves 6.01±0.05AEFG 1.93±0.03ADEFG 0.81±0.01ABDEFG 2.59±0.02ABDEFG Drežnica fl owers 5.19±0.03KLMOPR 1.48±0.02KLPR 0.45±0.01LMNOR 2.49±0.03KLMOPR Drežnica stems 5.26±0.00VWXZ”Γ 1.83±0.00VXYZ”Γ 0.16±0.01VYZ”Γ 2.15±0.09VXYZ”Γ Rujišta leaves 6.28±0.45AEFGH 3.00±0.43ABDEFGH 0.91±0.00ABDEFGH 1.98±0.06BDGH Rujišta fl owers 4.30±0.01LMNRS 1.48±0.06KLPS 0.52±0.01KLMNOPRS 2.12±0.03KLMNORS Rujišta stems 5.09±0.22VWXZ”Λ 2.55±0.20VWYZ” ΓΛ 0.18±0.01VYZ” 1.72±0.03WYZΓΛ Rakitnica leaves 5.31±0.28AFGHI 1.61±0.26DFH 1.13±0.01ABDEFGHI 2.57±0.05ABEFHI Rakitnica fl owers 4.65±0.21KLMNRT 1.11±0.12LMPRST 0.48±0.01KLMNOPRST 1.92±0.05KLMNOPRST Rakitnica stems 3.87±0.15VWYZ”ΓΛΠ 1.00±0.11WXYZΓΛΠ 0.19±0.01VWYZ”Γ 2.20±0.05VXYZ”ΛΠ Orjen leaves 5.42±0.10AFGJ 1.24±0.08BDFGH 0.50±0.00ABDEFGHIJ 1.58±0.03ABDEFGHIJ Orjen fl owers 5.55±0.04KLMOPSTU 1.51±0.02KLPT 0.27±0.00KMNOPRSTU 2.00±0.05KLMNOPRU Orjen stems 3.88±0.04VWXZ”ΓΛΣ 0.30±0.02VWXZ”ΓΛΠΣ 0.20±0.00VWYZ”Γ 1.50±0.02VWXY”ΓΛΠΣ Lovćen leaves 6.61±0.04AEFGIJ 1.71±0.02ADFH 0.37±0.00ABDEFGHIJ 2.53±0.03ABDEFHJ Lovćen fl owers 5.90±0.06KLMOPRST 1.77±0.01KLNOPT 0.21±0.00LMOPRSTU 3.26±0.03KMNOPRSTU Lovćen stems 5.32±0.07VWXYZ”ΠΣ 1.19±0.04XYZΓΛΠΣ 0.19±0.00VWYZ”Γ 2.44±0.04VWXYZ”ΓΛΠΣ KREMER D., JURIŠIĆ GRUBEŠIĆ R., BALLIAN D., STEŠEVIĆ D., KOSALEC I., ET AL. 270 ACTA BOT. CROAT. 75 (2), 2016 degree of separation. This population is situated on an is- land in the Adriatic Sea and it is under the stronger infl u- ence of the Mediterranean climate than the remaining popu- lations studied. The populations Vo and Lo are situated at similar altitudes on Mt Biokovo and Mt Lovćen, respec- tively. Both mountains are near the Adriatic Sea and under similar climatic conditions. Consequently, the geographical position of the populations Vo and Lo could explain the similarities obtained in the multivariate analysis. The popu- lations Sn and Or are geographically close and their simi- larities in the multivariate analysis were expected. The sim- ilarity between the populations Om and DG is more diffi cult to explain. Both populations are found in canyons, Om in the Cetina River Valley, and DG in the Neretva River Val- ley. It is possible that this environmental factor also played a signifi cant role in the accumulation of phenolic com- pounds. Although some studies of plant species such as Tanacetum cinerariifolium (Trevir.) Sch. Bip. (Grdiša et al. 2014), Campanula pyramidalis L. (Lakušić et al. 2013), Edraianthus tenuifolius (Waldst. et Kit.) A. DC. (Surina et al. 2011), and Cardamine maritima DC. (Kučera et al. 2008) showed a greater or lesser phylogeographical or tax- onomical split in the area of the Neretva River Valley, such a split was not confi rmed here. Similar results to the PCA were obtained using UPGMA, which separated three groups of populations (Fig. 1B). Similar populations were Vo (Vošac), Lo (Lovćen) and Ml (Mljet) which formed one large group. The remaining populations formed the second group containing two subgroups. The most similar popula- tions were Sn and Or, which formed a single cluster con- nected at a Euclidean distance of 2.64. Correlation between phenolic substances and chemical properties of soil Spearman rank order correlations between soil reaction (pH in H2O and in 1 M KCl), nitrogen content, and potash content (expressed as content of K2O in mg per 100 grams of soil) in soils of all habitats (as independent variables) and biologically active substances (i.e. total polyphenols, tannins, total fl avonoids, and total phenolic acids in leaves, fl owers and stems as dependent variables) in all plant sam- ples of M. petraea were not signifi cant (On-line Suppl. Tab. 1). However, correlations between the content of phospho- rus and content of total polyphenols in leaves and stems, as well as the content of total phenolic acids in fl owers of M. petraea were strongly negative. Spearman rank order cor- relations between the content of CaCO3 in soil of M. pe- traea habitats and the content of biologically active sub- stances were calculated only for eight (of ten) habitats of M. petraea, due to the very low CaCO3 content (below minimal quantities) in soil samples from the two remaining loca- tions. However, only one comparison of CaCO3 content in the soil and tannin content in the stem of M. petraea indi- cated a strong positive correlation. Comparison of the or- ganic matter content in soil of habitats (independent vari- able) and the content of biologically active substances (dependent variable) showed a moderately negative corre- lation in comparison with the total tannin content in the stems of M. petraea. No other comparisons gave signifi cant correlations (On-line Suppl. Tab. 1). The results of this study showed negative correlations between the phosphorus content in soil and TP content in leaves and stems. Also, a negative correlation was found between the phosphorus content in soil and TPA content in fl owers. These results are in line with results of Gerschen- zon (1983) who found that soil defi ciencies in phosphorus, sulphur, iron, calcium, and magnesium stimulates the pro- duction of phenolic compounds in plant tissues. Higher phenolic contents have also been reported as a response to phosphate starvation in Phaseolus vulgaris L. (Juszczuk et al. 2004). According to Tavarini et al. (2015) nitrogen fertil- ization of soil in the amount of 150 kg per ha will optimize the content of total phenols and fl avonoids in leaves of Ste- via rebaudiana Bertoloni (Asteraceae). Buchwald et al. (2015) showed that mineral fertilization with nitrogen, phosphorus and potassium did not substantially affect the level of phenolic acids in the raw material of Rhodiola ro- sea L. It was also found that using bio-fertilizer signifi cant- ly increased total fl avonoid contents in Anethum graveolens L. (Apiaceae) (Said-Al Ahl et al. 2015). Accordingly, it can be concluded that M. petraea has no specifi c demands regarding particular soil conditions. Two locations (Sn and Or) showing similar phenolic compounds levels, had quite different soil chemical properties. This Fig. 1. Principal component analysis (A) and the unweighted pair- group method with arithmetic mean (UPGMA) (B) of the pheno- lic compounds content in Moltkia petraea populations. Om – Omiš, Vo – Vošac, Ml – Mljet, Sn – Sniježnica, DG – Diva Grabovica, Dr – Drežnica, Ru – Rujišta, Ra – Rakitnica, Or – Or- jen, Lo – Lovćen. PHENOLIC SUBSTANCES IN MOLTKIA PETRAEA ACTA BOT. CROAT. 75 (2), 2016 271 suggests that biosynthesis of phenolic compounds in M. pe- traea is largely affected by other factors (genetic factors, topography, and exposition of plant site in relief, possible infl uence of Aeolian deposits in site, or other soil proper- ties) rather than the investigated soil properties. However, due to the very nice habitus (particularly because of the shape and colour of fl owers) and low soil and water de- mands, M. petraea may be considered a potentially valu- able horticultural plant in landscape architecture. Acknowledgements These researches were supported by the Ministry of Sci- ence, Education and Sports of the Republic of Croatia (project grant no. 006–0000000–3178). Authors would like to thank to Dr Valentina Papić Bogadi and Linda Zanella, MSc for helpful comments on the manuscript and for cor- recting the English style. References Anonymous, 2013: Ordinance of strictly protected species (In Croatian). Narodne novine 144, 7–84. Buchwald, W., Mordalski, R., Kucharski, W. A., Gryszczyńska, A., Adamczak, A., 2015: Effect of fertilization on roseroot (Rhodiola rosea L.) yield and content of active compounds. Acta Scientiarum Polonorum Hortorum Cultus 14, 109–121. Christ, B., Müller, K. H., 1960: Zur serienmäßigen bestimmung des gehaltes an fl avanol-derivaten in drogen. Archiv der Phar- mazie 293, 1033–1042. Dapkevicius, A., Van Beek, T. A., Lelyveld, G. P., Van Veldhuizen, A., De Groot, A., Linssen, J. P., Venskutonis, R., 2002: Isola- tion and structure elucidation of radical scavengers from Thy- mus vulgaris leaves. Journal of Natural Products 65, 892–896. Dunkić, V., Kremer, D., Dragojević Müller, I., Stabentheiner, E., Kuzmić, S., Jurišić Grubešić, R., Vujić, L., Kosalec, I., Randić, M., Srečec, S., Bezić, N., 2012: Chemotaxonomic and micromorphological traits of Satureja montana L. and S. subs- picata Vis. (Lamiaceae). Chemistry & Biodiversity 9, 2825– 2842. European Pharmacopoeia, 2007: Rosmarini folium. In: European pharmacopoeia commission 6th ed. 2, 2377–2378. Council of Europe, Strasbourg. Gerschenzon, J., 1983: Changes in the level of plant secondary metabolites under water and nutrient stress. In: Timmermann, B. N., Steelink, C., Loewus, F. A. (eds.), Phytochemical adap- tation to stress, 273–320. Plenum Press, New York. Grdiša, M., Liber, Z., Radosavljević, I., Carović-Stanko, K., Ko- lak, I., Satović, Z., 2014: Genetic diversity and structure of Dalmatian pyrethrum (Tanacetum cinerariifolium Trevir. / Sch./ Bip., Asteraceae) within the Balkan Refugium. PLoS ONE 9, e105265. Juszczuk, I. M., Wiktorowska, A., Malusa, E., Rychter, A. M., 2004: Changes in the concentration of phenolic compounds and exudation induced by phosphate defi ciency in bean plants (Phaseolus vulgaris L.). Plant and Soil 267, 41–49. Kołodziej, B., Sugier, D., 2013: Infl uence of plants age on the chemical composition of roseroot (Rhodiola rosea L.). Acta Scientiarum Polonorum Hortorum Cultus 12, 147–160. Kučera, J., Tremetsberger, V. J., Marhold, K., 2008: Molecular study of the Cardamine maritima group (Brassicaceae) from the Balkan and Apennine peninsulas based on amplifi ed frag- ment length polymorphism. Plant Systematics and Evolution 275, 193–207. Lakušić, D., Liber, Z., Nikolić, T., Surina, B., Kovačić, S., Bogdanović, S., Stefanović, S., 2013: Molecular phylogeny of the Campanula pyramidalis species complex (Campanulace- ae) inferred from chloroplast and nuclear non-coding sequenc- es and its taxonomic implications. Taxon 62: 505–524. Lal, R., Kimble, M., Follett, R. F., 2002: Agriculture practices and policies for carbon sequestration in soil. CRC Press LLC, Co- lumbus, Ohio. Page, A. L., 1982: Methods in soil analysis, Part 2: Chemical and microbiological properties. American Society of Agronomy, Wisconsin, Madison. Pernar, N., Bakšić, D., Perković, I., 2013: Soil survey fi eld and laboratory methods; manual for sampling and analysis (in Croatian). University of Zagreb, Faculty of Forestry and Croa- tian Forests, Zagreb. Ramegowda, V., Senthil-Kumar, M., 2015: The interactive effects of simultaneous biotic and abiotic stresses on plants: mecha- nistic understanding from drought and pathogen combination. Journal of Plant Physiology 176, 47–54. Ređić, S., Barudanović, S., Trakić, S., Kulijer, D., 2011: Vascular plant biodiversity richness and endemo-relictness of the karst mountains Prenj, Čvrsnica and Čabulja in Bosnia and Herze- govina (w. Balkan). Acta Carsologica 40, 527–555. Said-Al Ahl, H. A. H., Sarhan, A. M. Z., Abou Dahab, A. D. M., Abou-Zeid, E-S. N., Ali, M. S., Naguib, N. Y., 2015: Bio-fer- tilizer and gamma radiation infl uencing fl avonoids content at different parts of dill herb. International Journal of Life Sci- ence and Engineering 1, 145–149. Schneider, G., 1976: Zur bestimmung der gerbstoffe mit casein. Archiv der Pharmazie 309, 38–44. Surina, B., Schönswetter, P., Schneeweiss, G. M., 2011: Quaterna- ry range dynamics of ecologically divergent species (Edraian- thus serpyllifolius and E. tenuifolius, Campanulaceae) within the Balkan refugium. Journal of Biogeography 38, 1381–1393. Šilić, Č., 2005: Atlas of Dendrofl ora (trees and shrubs) in Bosnia and Herzegovina [in Bosnian]. Matica Hrvatska Čitluk, Čitluk. Tatzber, M., Stemmer, M., Spiegel, H., Katzlberger, C., Haberhau- er, G., Gerzabek, M. H., 2007: An alternative method to mea- sure carbonate in soils by FT-IR spectroscopy. Environmental Chemistry Letters 5, 9–12. Tavarini, S., Sgherri, C., Ranieri, A. M., Angelini, L. G., 2015: Effect of nitrogen fertilization and harvest time on steviol gly- cosides, fl avonoid composition and antioxidant properties in Stevia rebaudiana Bertoni. Journal of Agriculture and Food Chem istry 63, 7041–7050. Young, J. E., Zhao, X., Carey, E. E., Welti, R., Yang, S. S., Wang, W., 2005: Phytochemical phenolic in organically grown vege- tables. Molecular Nutrition and Food Research 49, 1136– 1142. Zovko Končić, M., Kremer D., Gruz, J., Strnad, M., Biševac, G., Kosalec, I., Šamec, D., Piljac Žegarac, J., Karlović, K., 2010: Antioxidant and antimicrobial properties of Moltkia petraea (Tratt.) Griseb. fl ower, leaf and stem infusions. Food and Chemical Toxicology 48, 1537–1542.