Background: The essential oil is an important compound of the root and rhizome of medicinally used valerian (Valeriana officinalis L. s.l.), with a stated minimum content in the European pharmacopoeia. The essential oil is located in droplets, of which the position and distribution in the total root cross-section of different valerian varieties, root thicknesses and root horizons are determined in this study using an adapted fluorescence-microscopy and automatic imaging analysis method. The study was initiated by the following facts: • A probable negative correlation between essential oil content and root thickness in selected single plants (elites), observed during the breeding of coarsely rooted valerian with high oil content. • Higher essential oil content after careful hand-harvest and processing of the roots. Results: In preliminary tests, the existence of oil containing droplets in the outer and inner regions of the valerian roots was confirmed by histological techniques and light-microscopy, as well as Fourier-transform infrared spectros- copy. Based on this, fluorescence-microscopy followed by image analysis of entire root cross-sections, showed that a large number of oil droplets (on average 43% of total oil droplets) are located close to the root surface. The remaining oil droplets are located in the inner regions (parenchyma) and showed varying density gradients from the inner to the outer regions depending on genotype, root thickness and harvesting depth. Conclusions: Fluorescence-microscopy is suitable to evaluate prevalence and distribution of essential oil droplets of valerian in entire root cross-sections. The oil droplet density gradient varies among genotypes. Genotypes with a lin- ear rather than an exponential increase of oil droplet density from the inner to the outer parenchyma can be chosen for better stability during post-harvest processing. The negative correlation of essential oil content and root thickness as observed in our breeding material can be counteracted through a selection towards generally high oil droplet density levels, and large oil droplet sizes independent of root thickness. Keywords: Valerian, Medicinal plant, Root slice, Thin-section, Oil droplet, Fluorescence-microscopy, Fourier-transform infrared (FTIR) spectroscopy, Nile Blue A, Sudan-III *Correspondence: Michael.Penzkofer@LfL.bayern.de Institute for Crop Science and Plant Breeding, Research Group Medicinal and Spice Plants, Bavarian State Research Center for Agriculture (LfL), Am Gereuth 2, 85354 Freising, Germany Full list of author information is available at the end of the article © The Author(s) 2018. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creat iveco mmons .org/licen ses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creat iveco mmons .org/ publi cdoma in/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Penzkofer et al. Plant Methods (2018) 14:41 Page 2 of 15 oil contents were achieved . Due to careful handling, Background the surface was not damaged and the essential oil, close Valerian (Valeriana officinalis L. s.l.) is an herbaceous to the root surface, still present. The presence of essential perennial plant with a huge variability regarding habitus, oil close to the root surface was confirmed by Holzner- composition of ingredients, and agro-economic traits. Lendbrandl  und Fridvalszky , who additionally The leaves usually are imparipinnate, and the leaflets, recognized small round bodies named ‘oil sacs’ in the weakly to strongly serrated. For blooming, a vernaliza- parenchyma of the roots. These ‘oil sacs’ were found pre - tion is necessary and hence, the first inflorescence usu - dominantly in the outer parenchyma. Violon et al.  ally develops in the second year of cultivation. Valerian identified ‘oil droplets’ also in the inner parenchyma. occurs on sporadically wet habitats in the temperate zone All previous investigations remain vague about the oil of the northern hemisphere. This indicates that a secure droplet identification and distribution across the cross- water supply is necessary for cultivation. Usually, the section. In addition, they do not give information con- rootstock forms a dense meshwork of thin roots . cerning oil droplets among different varieties, at different For medicinal purposes, the entire root system includ- root diameters on the same plant, or at different posi - ing the rhizome is used . Preparations based on tions along the roots. valerian roots are used against restlessness and sleep dis- The application of various vibrational spectroscopy turbances . In Germany, the dried root of valerian is a methods for visualizing secondary metabolites in dif- component of about 86 phytopharmaceutical and home- ferent plant tissue is already described for e.g. polya- opathic preparations. In North America (USA, Canada, cetylenes and carotenoids in carrots, or essential oil Mexico), due to other admission procedures, more than components in fennel, chamomile and curcuma [23–28]. 1000 products with valerian root are obtainable. In Ger- The Fourier-transform infrared FITR imaging method many alone, the demand for dried roots amounts to app. allows one to study the occurrence and distribution of a 1000 tons, equal to a market size of app. 4 Mio.€ [4–8]. wide range of molecules in cell tissues. However, it has To counteract the losses of root mass and secondary not yet been applied for the essential oil in valerian. The compounds during harvesting, cleaning and the further fluorescence-microscopic method is suitable for the visu - production process of dried valerian roots, breeding was alization and localization of secondary compounds in started in 2008 to develop new varieties of valerian with plant roots, and was used with sunflowers and mountain a coarser root-system (thicker adventitious roots) and arnica [29–31]. Furthermore, a spectral-sensitive camera with high contents of secondary compounds. A coarser could make more oil droplet structures visible, or make root system would probably preserve the secondary com- chemical differentiation possible, respectively [32, 33]. pounds, essential oil and valerenic acid . According Our intention was to give a more detailed histochemi- to the European Pharmacopoeia, the minimum content −1 cal description within the valerian roots. The develop - of essential oil must be 4 ml kg and of valerenic acid ment of an appropriate method to visualize and clearly at 0.17% (m/m) . The most frequent major constitu - identify the essential oil droplets required several consec- ents of essential oil of Valeriana officinalis L. s.l. are the utive steps grouped into two fields: (k) Verification of oil monoterpenes borneol and its esters, bornyl acetate and droplets and (kk) generation of an essential oil distribu- bornyl isovalerate [10–15]. tion map. Verification of the oil droplets (k) was done by In contrast to the abundant analyses of pharmaceuti- light-microscopic imaging and subsequent confirmation cal secondary compounds and their medicinal values, of the essential oil in the found oil droplets by Fourier- there are relatively few studies related to the physiology transform infrared (FTIR) imaging. Based on the results and localization within the root. Zacharias  described of these investigations, fluorescence-microscopic imag - essential oil to be located in ‘‘[…] the outer exodermis ing for generating oil droplet maps (kk) could be applied. […]’’, whereas Tschirch and Oesterle  found it in the Based on this, the study at hand was carried out to bet- ‘‘[…] single-row hypodermis […]’’. Both authors described ter understand the histochemical background of the two one ‘oil droplet’ in a single exodermis cell. Localization observations (i) and (ii). Observation (i) must be well of essential oil only in the outer cell layers of the vale- interpreted to assess the achievability of the breeding tar- rian roots would support the two following observations get of a thick root-system with good essential oil content. made during the breeding of coarse valerian: (i) Consid- Understanding the histological background of observa- ering 200 selected plants (elites), the essential oil content tion (ii) may explain why the entire essential oil is not all decreased with the increase of root thickness . This lost during a more robust, mechanized root harvesting behavior is explainable, because with increasing root and processing in large-scale valerian field production. diameter, the root surface area decreases in relation to We postulated that a great part of the essential oil drop- the root volume (calculated as cylinder). (ii) After care- lets occur in the inner parts of the valerian root. ful hand-harvesting and hand-processing, high essential Penzkofer et al. Plant Methods (2018) 14:41 Page 3 of 15 elite was dug out carefully and the adventitious roots Methods were separated into four diameter fractions (< 2, < 3, < 4, Plant material > 4 mm; Table 1). The fresh roots were stored in air-tight In 2010, three elite plants were selected from the variety containers at 5–6 °C to prevent dehydration of the roots ‘Anton’ (seed source: N.L. Chrestensen Erfurter Samen- and a loss of essential oil. Prior to preparation for micros- und Pflanzenzucht GmbH, Erfurt, Germany, 2008) and copy, the roots were washed carefully with water. one elite plant from the variety ‘Lubelski’ (seed source: PHARMASAAT Arznei- und Gewürzpflanzensaatzucht Imaging methods GmbH, Artern, Germany, 2009) based on their differ - Verification of oil droplets (k) ing root morphology and essential oil contents (Table 1). Classic light‑microscopic imaging To confirm literature Two elites were characterized as thin-rooted, meaning observations we used the classic method for microscopic that they predominantly formed a highly branched and imaging by fixing the cell components with a formalde - felted rootstock with thin adventitious roots. The other hyde-propionic acid–ethanol-solution (5–5–90% FPA) two elites predominantly formed a chunkier and less and embedding the fixed roots in historesin (2-hydroxy - felted rootstock with thicker adventitious roots; these ethyl methacrylate). The starch was colored with a Lugol’s were called thick-rooted. solution (potassium tri-iodide) and washed out with In contrast to common cultivation, in the current study, potassium hydroxide (KOH). clones were used in order to have plants with known Thin slices of embedded roots, as well as thin longitu - analytical and identical genetic backgrounds. Plants of dinal-section slices of root parenchyma and exodermis seed propagated valerian populations would probably from fresh valerian roots, were colored with sudan-III- vary too much in the contents of secondary compounds solution to make the lipids visible . The sections were . Therefore, the four elite plants were cloned by ster - evaluated with a light microscope (ZEISS Axiostar plus, ile micropropagation using inflorescences at an early Carl Zeiss AG, Oberkochen, Germany) at magnification bud stage as starting tissue and side shoots as propaga- of 200. tion parts. Rooted plantlets were cultivated for 2 years in the field of the experimental station Baumannshof Fourier‑transform infrared (FTIR) imaging Thin-sec - of the Bavarian State Research Center for Agriculture tion slices of 0.01 mm thickness were made by a freezing (48°42′N /11°32′E, 360 m MSL). A detailed description microtome (Leica CM 1100, Leica Biosystems Nussloch of the plant material, the cloning by in vitro propagation GmbH, Nussloch, Germany). Each slice was screened and the cultivation conditions are described in Penzkofer for colorless round bodies by light microscopy that was et al. , where the same plant material was used. integrated in the FTIR spectrometer. Sections with round Due to the age of the plants, flowering was induced and bodies were subsequently analyzed applying FTIR imag- shoots started to develop in spring of the harvest year. ing. FTIR transmission spectra of the thin-section slices The inflorescence development influences the essen - were recorded with the FTIR spectrometer Varian 4100 tial oil content and causes a decrease of the essential oil FTIR (Agilent, Waldbronn, Germany) combined with content from summer to autumn . Therefore, the the IR-microscope Varian UMA 600 (Agilent, Wald- inflorescences were cut off in an early stage of develop - bronn, Germany). The thin-section slices were placed on ment to counteract the decline . Our plant material a ZnSe window and FTIR images were produced with a was harvested in autumn (2014), the usual harvest time 32 × 32 focal plane array detector (FPA). The spectra were for valerian cultivation. At least one cloned plant of each Table 1 Root morphology types and content of essential oil in the root drug of the cloned elite plants (CE) Cloned elites Varieties Morphological Essential oil Root fraction (diameter) root structure −1 b b b b Classification ml kg RF1 (< 2 mm) RF2 (< 3 mm) RF3 (< 4 mm) RF4 (> 4 mm) CE1 ‘Anton’ Thin-rooted 8.9 (14.1) ● ● ● — CE2 ‘Anton’ Thin-rooted 10.4 (19.9) ● ● ● CE3 ‘Anton’ Thick-rooted 5.0 (12.9) ● ● ● ● CE4 ‘Lubelski’ Thick-rooted 6.9 (11.3) ● ● ● ● Prevalence of root thickness fractions (RF) in the clone plants Essential oil content of the mother plant (elites) and in brackets the cloned elites, determined in 2010 and 2012, respectively Diameter at the base (Horizon 1, see Fig. 1) is decisive for the classification to a fraction. The ● indicates the formed and the — indicates the not formed fractions Penzkofer et al. Plant Methods (2018) 14:41 Page 4 of 15 −1 With a moderate pressure, a constant velocity and with- recorded over the wavelength range of 4000–850 cm out displacement, a straight metallic blade was moved and 128 scans per spectrum were accumulated. −1 through the valerian root tissue. The cut was performed The absorption band at 998 cm represents mainly −1 in a constant angle of 20°. Through the moveable and cellulose, whereas the signal at 1737 cm was assigned height adjustable hanger, thin root-slices of a uniform to C=O vibration of bornyl acetate. Both signals were thickness of 0.2 mm were able to be produced. These used to display the distribution in pseudo-color images. root-slices were then placed on top of a drop of water on a microscope slide. From each horizon, 15–20 thin-sec- Fluorescence‑microscopic imaging and mapping of differing tions were cut, but just a low number of slices (seven on root material (kk) average) were suitable for the following image processing, Sample preparation and producing of thin analysis and evaluation. roots ‑ lices From each cloned elite (CE) and root fraction (RF), two well-developed fresh roots with the typical root appearance were chosen and classified into three approxi - Staining and uore fl scence‑microscopy After the root- mately 60 mm long parts, called horizons (HZ1 to HZ3, slices were cut, the water was removed and, based on the Fig. 1). HZ1 represents the rhizome-near part, which experiences of Fridvalszky , one to three drops of 1% would certainly be harvested after cultivation, and HZ3 aqueous solution of Nile Blue A (Carl Roth GmbH und represents the rhizome-far part, which probably remains Co. KG, Karlsruhe, Germany) were applied. After incu- in the ground. Along the whole length of the horizons, bating the root-slices for one minute at room tempera- several thin-sections were taken and prepared. ture, the solution was carefully removed. A drop of water The cutting of thin root-slices was done by hand with was applied and the root-slices were covered with a cover a height adjustable cylinder-microtome. A segment of glass. A further incubation period of at least 30 min fol- the respective root diameter fraction and horizon was lowed. clamped between a buffer-material, cut out from a car - The fluorescence-microscopy was done with a mag - rot root parenchyma. This material has a comparable nification of 100 (ocular 10× , objective 10×, ZEISS structure and consistency to the examined valerian roots. Axiostar plus, Carl Zeiss AG, Oberkochen, Germany). Thereby, the fresh valerian roots were well enclosed. The initial light generated by a HBO50 high pressure Fig. 1 The valerian root system components and visualization of the investigated horizons Penzkofer et al. Plant Methods (2018) 14:41 Page 5 of 15 mercury arc lamp was filtered for excitation at 430– Data evaluation and statistical analysis For data evalu- 510 nm and for emission at 475–575 nm. The images ation, the determined x–y-coordinates were adapted and were taken with the connected digital camera ZEISS related to each other. The coordinates of the centers were AxioCam ERc 5 s (a hyperspectral camera was not used as new coordinate origins (formula Ia and Ib) and available) and instantly transferred to the connected the distances of the oil droplets and the edges related to computer. Due to the limited lens coverage, the final the center were calculated with Formula II. This was only picture of a complete root-slice had to be composed of achievable when the center, the oil droplet and the corre- several partial images. sponding point of the root edge all lay on an imaginary line. The green fluorescence oil droplets were imaged with An example is shown in Fig. 3A. The corresponding point a high-contrast against the black background. Despite of the root edge was calculated with help of the polar angle, all precautions taken, the root-slices did not always have which must be the same for the distance between the center exactly the same thickness and the cytoplasm of the cut and the oil droplet (CD in Fig. 3A) and for the distance cells was more or less leaked, so that the light transmit- between the center and the corresponding point of the edge tance of the cell layers varied among and within the root- (CE in Fig. 3A). The polar angle of both distances was cal - slices. In order to make the oil droplets clearly visible in culated with Formula III. At last, the relative distances of all areas of the root slice, brightness and contrast were the oil droplets and the root edges to the center were deter- adjusted through the camera software (ZEISS AxioVison, mined (Formula IV). This data was used for the evaluation. Carl Zeiss AG, Oberkochen, Germany) by one to two The relative distance data was assigned to one of nine units, upwards or downwards, for each image. The depth classes with a class width of 11.11%; this led to the interval of focus on the microscope was not changed so that the limits for class 1 = [0–11.10%); class 2 = [11.11–22.21%); same cell layer was shown on each image. class 3 = [22.22–33.32%); class 4 = [33.33–44.43%); class 5 = [44.44–55.54%); class 6 = [55.55–66.65%); class Image processing All partial images of one root slice were 7 = [66.66–77.76%); class 8 = [77.77–88.87%); class converted from the camera software’s own file format to 9 = [88.88–100%]. Figure 3B shows a generalized illus- the compressed free TIFF file format and then manually tration of the nine classes. Class 1 was always within the merged to one complete root-slice image with the image central cylinder; class 2 delineated mostly the border of editing program Adobe Photoshop CS6 (Adobe Systems the central cylinder. The area of both classes together Software, Dublin, Republic of Ireland). was addressed as central cylinder. The classes 3–8 com - The composed root-slice images were analyzed with prised the parenchyma, class 9 the outer cell layers. The the image analysis software ImageJ . Software mac- area of the classes increased from the center to the edge. ros were developed to generate black-white-masks of the Therefore, both the number of oil droplets in each class oil droplets, the root-slice center and the root-slice edge and the oil droplet density were determined to compare from each composed root-slice image (Fig. 2B–D). More the classes. The density was calculated as the quotient of information on the functionality is given in the Addi- number of oil droplets over the class area. tional file 1. For statistical analysis, in each class, the mean number The principle steps were to convert the composed of oil droplets of the root-slices was determined. The dis - image of the root slice into 8 bits grayscale image, to tribution of these oil droplets as affected by the horizons, reduce the background noise using ImageJ image filter - the root diameter fractions, the root classification and the ing operators (median, dilate), and segment the oil drop- genotypes, were compared with the Friedman-Test and let by using the Huang Threshold method implemented Wilcox-Test. To compare the different factor levels, such in ImageJ. as different genotypes (CE), root diameter fractions (RF) Due to the variable position and the inconsistently or horizons (HZ), the oil droplet density for each root- round shape of the central cylinder, as well as the edge slice was calculated. An analysis of variance (ANOVA) of the root-slice, the center of the root-slice was manu- and t Test was performed. For all statistical analyses, sig- ally marked in the original composed root-slice image. nificance was given at p < 0.05. After treating images that way, they were converted into Unless otherwise described, the following data repre- a binary image (black and white) and x–y-coordinates sent the mean values over the other factors and their fac- of the now black particles were determined by using fil - tor levels. ters for particle size and particle circularity. The deter - Formula: mination of the x–y-coordinates of the centers of the x = x −x i C mod (Ia) root slices and each point of the edges were done in a similar manner. Each image was treated with the same y = y −y mod i C (Ib) adjustments. Penzkofer et al. Plant Methods (2018) 14:41 Page 6 of 15 Fig. 2 Different black-white-masks derived from the original composed root-slice image. A Root-slice image with green shining oil droplets. B–D Black-white masks of B: the oil droplets, C the center of the root slice, D the edge of the root-slice. The x–y-coordinates of the oil droplets and the center (black particles in B and C) and each point of the edge in D, and the size of the grey area (seen in B–D) were determined automatically by the image analysis software oil droplets and root edges, respectively, x y = x- and C, C 2 2 r = x + y (II) y-coordinates of centers, r = distance (radius) in pixel mod mod points, ϕ = polar angle, pp = relative distance, r = dis- CD tance (radius) from center to oil droplet, r = distance CE (radius) from center to root edge mod ϕ = arccos (III) Results Classic light‑microscopic imaging Figure 4A shows the colored cross-sections of the root (r ∗ 100) CD pp = (IV) parenchyma of a randomly chosen valerian root. Between CE the colored grains of starch, colorless round bodies are visible. It is assumed that these colorless round bodies x y = according to the new origin coordinate mod- contained essential oil (Fig. 4A). mod, mod ified x- and y-coordinate, x y = x- and y-coordinates of i, i Penzkofer et al. Plant Methods (2018) 14:41 Page 7 of 15 Fig. 3 A Schematic illustration of a root-slice segment with the identified elements center (C), oil droplets (D, labeled is one oil droplet) and the root edge (E). φ indicates the polar angle with C as pole and the x-axis (horizontal line) as polar axis. B Generalized illustration of the nine classes (1–9), to which the oil droplets (pp, Formula IV ) were assigned based on their relative distance to the center Fig. 4 Microscopic images of thin-section slices from fixed and embedded (A), and fresh (B + C) valerian roots. A Cross-section through the root parenchyma with stained starch (Lugol’s solution) and intermediary colorless round bodies (black arrow pointer). B + C Longitudinal-section of the root parenchyma and exodermis. In addition to the colored starch, the red-colored bodies, which were stained by use of a sudan-III-solution, are visible (black arrow pointer) In thin cross-sections and longitudinal-section slices of microscopic picture, two intact oil bodies might be iden- −1 fresh roots, only few colorless round bodies were stained tified due to the intensive absorption at 1737 cm . The red by the application of the sudan-III-solution (Fig. 4B, red colored part in the right upper corner of that image C). might be the result of destroyed oil bodies and smeared essential oil due to microtome preparation of the root Fourier‑transform infrared (FTIR) imaging slide. Generally, in the sections used for FTIR imaging, Figure 5 presents the light-microscopic picture of a thin it was very rare to find intact colorless bodies for which −1 valerian root slice and the corresponding FTIR images high absorbance around 1737 cm was observed. Unfor- −1 obtained by integration of the absorbance at 1737 cm tunately, the presence of oil bodies could not be con- −1 and 998 cm (top, from left to right). Whereas the sig- firmed by adjacent staining experiments. −1 nal at 1737 cm can be tentatively assigned to bornyl −1 acetate, the absorption at 998 cm mainly represents Fluorescence‑microscopic imaging and mapping cellulose matrix. The spectrum presented in Fig. 5D was of different root material taken at the crossed lines shown in the integration map A total of 678 root-slices of valerian were evaluated. The −1 for 998 cm (Fig. 5C). As indicated by arrows in the light number of root-slices was distributed quite evenly among Penzkofer et al. Plant Methods (2018) 14:41 Page 8 of 15 −1 Fig. 5 Light-microscopic (A) and FTIR images, of valerian root cross section showing high absorbance for mainly bornyl acetate at 1737 cm (B) −1 and cellulose at 998 cm (C). The spectrum was taken from the cross mark in the appropriate FTIR images for cellulose (D). Colors blue, green, yellow, and red represent increasing content—the warmer the color, the higher the spectral intensity the different levels of the factors: cloned elites (CE), root significant effect on the density of oil droplets (Fried - diameter and classification, root horizon (HZ). However, man: CE p = 0.014; RF p < 0.001; HZ p = 0.016; Wilcox : fewer root-slices were analyzed in regard to the root root classification p = 0.012). diameter fraction RF4, because the thin-rooted cloned The root diameter of each factor level affected the elites did not form roots with a diameter greater than number of oil droplets. Thicker root-slices contained 4 mm (Tables 1, 2). more oil droplets, as presented in Table 2. This meant, The distribution of the mean number of oil droplets ultimately, that the number of oil droplets increased in each factor level was approximately constant (Fig. 6). with root thickness, root diameter and in the upper Significant differences between the distributions of the located root horizons as compared to the lower hori- factor levels of root classification (Wilcox: p = 0.012), zon. As shown for the class comparison, the density of root diameter fraction and horizon (Friedman: p < 0.001 oil droplets allowed for a better comparison between and p = 0.005, resp.) were especially visible in class nine factor levels. Among the cloned elites, CE2 showed the (Fig. 6B, C). In general, the mean number of oil droplets highest oil droplet density (ANOVA p < 0.001; Fig . 8C). increased from the center to the outer cell layers. The RF1 was the root diameter fraction with the highest oil central cylinder was almost free of oil droplets, whereas, droplet density (ANOVA p < 0.001; Fig . 8C), whereas the parenchyma included 57% (42–64%) of the mean each horizon showed a significantly different oil droplet number of oil droplets on average. In the outer cell layer, density (ANOVA p < 0.001; Fig . 8D), with an increasing completely covered by class nine, 43% (36–56%) of the oil droplet density from HZ1 to HZ3. mean number of oil droplets were present, on average. Figure 8A shows fluorescent oil droplets in a repre - To consider the varying areas of the classes, the oil sentative root cross-section. Based on the image area droplet density was a more suitable parameter to com - of the oil droplets, the size of oil droplets allocated to pare the classes than the number of oil droplets. A classes 3–7 was approximately 72% larger than the size constant density was not found. Similar to the mean of the oil droplets allocated to classes 1 and 2. This was numbers, the density of oil droplets also rose up from more or less recognizable for all root-slices, but was not class one to class nine (Fig. 7). Each factor showed a Penzkofer et al. Plant Methods (2018) 14:41 Page 9 of 15 Table 2 Number of slices, sum of oil droplets and mean Violon et al. , who concluded hereupon that the lipid number of oil droplets in the root-slices of four cloned bodies contained ‘volatile oil’. elites (CE1, CE2, CE3, CE4), of two root classifications (thin- and thick-rooted), of four root diameter fractions (RF1, The colorless round bodies are filled with essential oil RF2, RF3, RF4) and of three horizons (HZ1,HZ2, HZ3) Generally, FTIR imaging can provide information about Number Sum of oil droplets Mean number ± standard the distribution of different plant constituents without of root‑ deviation of oil droplets destructing the plant constituent containing cell struc- slices per root‑slice ture. These constituents include lipids, carbohydrates, Cloned elites and lignin, as well as secondary metabolites, e.g. terpe- CE1 176 34,665 197 ± 84 noids . CE2 162 43,466 268 ± 149 Nevertheless, the local accumulation of the strong CE3 171 61,419 359 ± 255 −1 absorbance at 1737 cm tentatively assignable to bornyl CE4 178 53,666 301 ± 168 acetate indicates an essential oil distribution in distinct Root classification cellular structures. As bornyl acetate is a principal com- Thin 338 78,131 231 ± 125 ponent of the essential oil of valerian, the results from Thick 349 115,085 330 ± 217 FTIR imaging confirms the theory that essential oil is Root diameter fractions located within the colorless bodies. However, only very RF1 211 37,422 177 ± 87 few intact oil droplets were found in the thin cryo-sec- RF2 195 42,647 219 ± 98 tions. The preparation of thin root slides of a thickness RF3 210 77,602 370 ± 192 below 0.01 mm usually resulted in disruption of the oil RF4 71 35,545 501 ± 239 bodies and smearing of the essential oil. Therefore, the Horizons results of FTIR imaging have to be seen as basic stud- HZ1 206 68,904 334 ± 225 ies combining visual images of the sample with chemi- HZ2 246 74,994 305 ± 184 cal information. The authors’ own previous studies on HZ3 235 49,318 210 ± 107 valerian root sections performed with FT-Raman spec- The root diameter fraction was only formed by CE3 and CE4 (Table 1) troscopy did not deliver additional information about essential oil distribution. Only the differentiation of vari - ous root tissue mainly based on carbohydrate profile was examined in detail. In the outer cell layer of young lat- obtained and with the instrument used, the local resolu- eral roots we could already find oil droplets (Fig. 8 B). tion was limited to around 150 µm (data not shown). Modern Raman microscopes achieve a local resolution below 1 µm depending on the laser and aperture used. To Discussion gather suitable signal intensity, the Raman laser needs to Histochemical structures and methods (k, kk) be exactly focused, which demands for the extremely thin Comparable histological structures were found as reported root slices. With application of the necessary laser power, for valerian in literature the brownish root material often started burning and the We compared our breeding plant material with the exist- spectra showed high fluorescence appearance. A reduc - ing literature information by using classic light-micro- tion in laser power resulted in a lack of signals. There - scopic imaging techniques. Many grains of starch are fore, if the analytes don’t contain Raman active bonds densely arranged in the cells and distributed over the (e.g. C–C double bonds in carotenoids) and the sam- entire parenchyma [20–22, 39]. In longitudinal-section ple is colored (as the majority of plant derived material slices from fresh roots, oil droplets were visible, who is), Raman spectroscopy investigations are technically are located near the cell-wall [20, 40]. Large-lumen cells challenging. of the exodermis were observed, however, the filling In recent years, another vibrational spectroscopy of the cell with one oil droplet [20, 21, 39] could not be method gained attention in plant analysis. Hyperspectral detected. (near infrared) spectroscopy imaging can be used for var- The staining of the round bodies with sudan-III-solu - ious analytical purposes as species identification, disease tion worked well in the fresh roots, but not in the embed- detection or nutrient quantification, but is focused on ded roots. Probably, the essential oil volatilized due to macroscopic samples due to the relatively coarse spatial either the use of ethanol during the embedding process resolution of several 100 µm [42–45]. or due to the ethanol-containing sudan-III-solution itself. Taking into account the above described characteris- Similar effects were observed by Fridvalszky  and tics of each method, FTIR imaging in combination with subsequent staining to affirm the preliminary results, Penzkofer et al. Plant Methods (2018) 14:41 Page 10 of 15 Fig. 6 Mean number (columns) and density (points; mean number/class area) of oil droplets, determined from root thin section slices of valerian and allocated in nine classes (1–9). Shown are four cloned elites (CE1, CE2, CE3, CE4) (A), two root classifications (thin- and thick-rooted) (B), four root diameter fractions (RF1, RF2, RF3, RF4) (C) and three root horizons (HZ1, HZ2, HZ3) (D). Vertical lines: standard deviation. Significant codes: *< 0.05, **< 0.01, ***< 0.001. The same capital and small form letters mark that the oil droplet mean numbers and the densities are equal (p < 0.05, SNK), respectively. The colored background marks the classes of the parenchyma. Further details of the cloned elites, the root classification, the root diameter fractions, the horizons and the parenchyma are shown in the text Relative distances and class width are suitable to compare seems to be the most favorable analytical strategy. The the root diameter fractions and root horizons resulting FTIR distribution maps showing the strong −1 The oil droplet position was calculated as relative dis - C = O vibration at 1737 cm confirm the results of tance from the root center to the root edge. The relative staining experiments and fluorescence-microscopy. distances were allocated to nine width classes, which Due to the failed repeatability of the FTIR imaging of allowed a comparison between the root diameter frac- oil droplets, the more robust method of fluorescence- tions and root depth (horizons), regardless of the real microscopy was chosen to visualize the oil droplets. Penzkofer et al. Plant Methods (2018) 14:41 Page 11 of 15 Fig. 7 Mean oil droplet density (number of oil droplets/mm ), determined from root thin section slices of valerian. Shown are four cloned elites (CE1, CE2, CE3, CE4) (A), two root classifications (thin- and thick-rooted) (B), four root diameter fractions (RF1, RF2, RF3, RF4) (C) and three horizons (HZ1, HZ2, HZ3) (D). Further details of the cloned elites, the root classification, the root diameter fractions and horizons are shown in the text. Vertical lines: standard deviation. Significant codes: *< 0.05, **< 0.01, ***< 0.001. Different small form letters mark significant differences between the oil droplet densities (p < 0.05, SNK). Number of root slices given in Table 2 root diameter of the considered root slice. The root congruent with the edge of the root slice. Further infor- slices seldom showed a circular shape. Often, the root mation concerning the oil droplet distribution would be slices were oval or had coves. An absolute and constant achieved if densities of the classes are calculated based class width would not have been appropriate, because on absolute diameters and cross-section areas. For our the border of class nine would not necessarily be purposes, the absolute oil droplet densities of the total cross-section areas were sufficient. Penzkofer et al. Plant Methods (2018) 14:41 Page 12 of 15 Fig. 8 Fluorescent oil droplets in cross-sections of valerian roots: adventitious root (A), growing lateral root breaking through the exodermis (B, white arrow pointer) Meaning of oil droplet distribution and density for valerian u Th s, the number of oil droplets was not a suitable iden - breeding and production (i, ii) tifier for the essential oil content. This may be due to the With the current investigations, we were able to show varying oil droplet sizes in the inner parenchyma. that oil droplets can be found in the whole parenchyma. In contrast, the essential oil droplet density based on These results are in good agreement with previous stud - total root cross-section area is more informative. Clone ies of Violon et al.  and Szentpetery et al. . Besides CE2 had the highest oil droplet number/mm (Fig. 7) and the identification of the oil droplets, the number and it showed the highest oil content in the study by Penz- density of oil droplets are also now available for different kofer et al. . However, to evaluate the relationship varieties, different root diameters on the same plant or between essential oil droplet density and essential oil for different positions along the roots. content, more data is needed. Conclusions about the relationship of oil droplet occurrence The number and density of oil droplets are not negatively and essential oil content of genotypes are limited related to the root thickness Chemical-analytical data does not exist for the currently The mapping results showed an expected increase of the studied plant tissue, but genetically identical plant mate- number of oil droplets per cross-section from thinner to rial has been investigated analytically by Penzkofer et al. thicker roots (Table 2). For the fractions with low to high . We tried to compare and re-estimate the results of root diameters (RF1-RF3), the number of oil droplets both studies. was in accordance with the essential oil content results The comparison of the four clones gives only limited of Penzkofer et al. . The fraction with the largest root information for the relationship between oil droplet diameter (RF4) showed the highest number of oil drop- occurrence and oil content, due to the low number of lets in our study, but the lowest essential oil content in clones, the limited genetic variability among clones and their study. It has to be noted that the essential oil con- the environmental and year’s effect, which all potentially tent was determined by a very small number of data influence the essential oil content [46, 47]. Even Penz- points and comprised only the clone CE4, because the kofer et al.  could only detect trends in this regard, other clones did not develop roots of this diameter. Con- but identified clone CE2 as the one with the high - sidering CE4 only, the oil droplet number was the same est essential oil content. This did not coincide with our for thick and very thick roots (RF3 and RF4, respectively). essential oil mapping data. Concerning the number of The oil droplet density of total root cross-sections was oil droplets per root cross-section, clone CE3 shows the highest in the thin roots (RF1) and lower in medium to highest values, followed by clone CE4 and CE2 (Table 2). thick roots (RF2-4). Thus, no linear relationship between Penzkofer et al. Plant Methods (2018) 14:41 Page 13 of 15 oil drop density and root thickness can be derived. The The comparison of horizons provides further infor - high oil droplet density in thin roots may partially be mation about the reliability of the production process. an effect of root age, assuming that a higher portion of The horizons represent harvest depths: The greater the young roots occurred in this root fraction, and that oil depth, the thinner the formed roots, and the greater the droplets are formed at a very early stage of development, likelihood that they will be more easily lost during the a phenomenon we were able to see with our image of a harvesting process. young lateral root. However, root age may not entirely These horizons contain a lower number of oil drop - explain the difference in oil droplet density between RF1 lets. Thus, the harvest of the root parts in deeper soil and RF2 as there was no difference between RF2 and RF3. layers does not have to be exhaustive in order to obtain Concerning the breeding target of thicker roots, we high essential oil contents in the root drug, especially found a positive relationship between root thickness and when 96% of the root mass is located in the topmost number of oil droplets per root cross-section. We did not 10 cm of the ground . find a negative relationship between root thickness and oil droplet density, especially when medium to very thick roots were considered. Consequently, there is a poten- Conclusion tial for selection of thick roots that have high oil drop- For the first time, the essential oil distribution in entire let densities and, therefore, also likely high essential oil valerian root cross-sections of different varieties, root content. This is supported by the experience that many thicknesses and root horizons were visualized, apply- inbred lines with coarse roots and high essential oil con- ing an imaging and mapping fluorescence-microscopy tent could be derived from clone CE4. technique. The applied methods of FTIR spectroscopy and fluorescence-microscopy allowed for the determi - nation of oil droplets as clearly differentiated structures. Oil droplet distribution should be considered in cultivation, Our results give insight into the cross-sectional harvesting and breeding essential oil distribution at valerian harvest time. Considering that 43% of the oil droplets were localized Although the existing natural variability and the limited in the outer section of the roots (class nine), it is evi- plant material investigated in this study only allow for dent that a careful root harvest and processing, without a brief overview, some aspects derived from the results damaging the root surface, is important. This also means could be used for further investigations and for future that a high portion of oil droplets are located in the inner breeding work: parenchyma, information which is important for valerian cultivation, as well as for breeding selection. 1. The number, density and distribution of essential oil Heindl and Hoppe  reported that intense and long droplets of genetically different plant material vary washing of valerian roots leads to a loss of secondary and allow for the selection of suitable plant material compounds. Therefore, the more essential oil is stored for breeding purposes, independent of root thick- in the inner parenchyma, the lower should be the losses ness. during the processing, and the higher is the processing 2. The breeding plant material should generally exhibit stability of the roots. a high oil droplet density level as this is one of the One of the main breeding targets in the past was factors for the essential oil content. increasing the essential oil content . With this target 3. A high and homogeneous oil droplet density should in mind, the losses of oil during processing and storage be aspired in the inner parenchyma in order to avoid are minor. We now show that in different single plants, essential oil losses during harvesting and post-har- represented by the clones of elites, different essential vest processes. oil distributions exist. Choosing plant types with stead- ily increasing oil droplet density curves instead of plants It still remains unclear which oil droplet character- with exponentially increasing oil droplet densities istic is preferred as a selection trait. In consequence, a towards the root edge implies that a higher portion of the compromise between the absolute oil droplet density oil will be located in the inner parenchyma. Moreover, and the oil droplet density curve must be found, while processing stability and possibly, also, storage stability of at the same time, considering root thickness and indi- valerian roots can be increased. Our imaging and map- vidual genetic background. ping method, in combination with the calculated density Finally, a careful, but not necessarily exhaustive curves, can serve as a selection tool for identifying suit- harvest, as well as careful post-harvest procedures, able plants. Penzkofer et al. Plant Methods (2018) 14:41 Page 14 of 15 References confirmed as best methods for conserving the essen - 1. Heuberger H, Lohwasser U, Schmatz R, Tegtmeier M. Baldrian (Valeriana tial oil and obtaining a high quality of the valerian root officinalis L.). In: Hoppe B, editor. Handbuch des Arznei- und Gewürzp - drug. flanzenanbaus. Band 4 Arznei- und Gewürzpflanzen A-K, vol. 4., Arznei- und Gewürzpflanzen A-KBernburg: Verein für Arznei- und Gewürzp - Additional file flanzen SALUPLANTA e.V.; 2012. p. 164–83. 2. European Pharmacopoeia. Baldrianwurzel Valerianae radix s.l. 9th ed. Stuttgard, Eschborn: Deutscher Apotheker Verlag, Govi-Verlag- Additional file 1. Analyzation steps and software macros for localiza- Pharmazeutischer Verlag GmbH; 2017. tion of the oil droplets, root-slice area and edges, as well as the root-slice 3. European Scientific Cooperative on Phytotherapy (ESCOP). Monograph center. Valerianae radix, Valerian Root. 2nd ed. supplement. Stuttgart, New York: Thieme; 2009. 4. FNR, Fachagentur Nachwachsende Rohstoffe e.V. Arzneipflanzen Anbau und Nutzen. https ://media thek.fnr.de/arzne ipfla nzen-anbau -und-nutze Abbreviations n.html (2013). Accessed 12 Mar 2018. CE: cloned elite plant; RF: root thickness fraction; HZ: root horizon; FTIR: 5. FNR, Fachagentur Nachwachsende Rohstoffe e.V. Marktanalyse Fourier-transform infrared. Nachwachsende Rohstoffe.https ://media thek.fnr.de/brosc huren /samml ungen /schri ftenr eihe-nr/markt analy se.html (2014). Accessed 12 Mar Authors’ contributions MP performed the preliminary investigations, adapted the fluorescence- 6. Licensed Natural Health Products Database (LNHPD): Health Canada, microscopy technique, designed the experiment and analyzed the data. AB Ottawa. https ://healt h-produ cts.canad a.ca/lnhpd -bdpsn h/index -eng.jsp performed the fluorescence-microscopic works, including sample preparation, (2018). Accessed 12 Mar 2018. staining and image processing, and was a major contributor in writing the 7. Database of registered medicines containing valerian in Mexico. manuscript. AN and AK assisted the work and the interpretation of the Fourier- Comisión Federal para la Protección Contra Riesgos Sanitarios (COFEPRIS), transform infrared (FTIR) imaging and co-wrote the sections, which included Mexico City. http://www.cofep ris.gob.mx/AS/Pagin as/Regis tros%20San the FTIR. HS and HH enabled the investigation in their working groups. All itari os/Regis troSa nitar ioMed icame ntos.aspx (2016). Accessed 12 Mar authors read and approved the final manuscript. 8. Dietary Supplement Label Database (DSLD). National Institutes of Health Author details (NIH), Bethesda, Rockville. https ://ods.od.nih.gov/Resea rch/Dieta ry_Suppl Institute for Crop Science and Plant Breeding, Research Group Medicinal ement _Label _Datab ase.aspx (2018). Accessed 12 Mar 2018. and Spice Plants, Bavarian State Research Center for Agriculture (LfL), Am 9. Heuberger H, Penzkofer M. Demonstrationsprojekt Arzneipflanzen Gereuth 2, 85354 Freising, Germany. Federal Research Centre for Cultivated (KAMEL): Züchterische Verbesserung von Baldrian zur Erhöhung der Plants, Institute for Ecological Chemistry, Plant Analysis and Stored Product Rentabilität und Drogenqualität. http://www.lfl.bayer n.de/ipz/heilp flanz Protection, Julius Kühn-Institute (JKI), Königin-Luise-Straße 19, 14195 Berlin, en/16895 7/index .php (2017). Accessed 5 Sept 2017. Germany. 10. Stoll A, Seebeck E. 1. Mitteilung über Valeriana. Die Isolierung von Hes- peritinsäure, Behensäure und von zwei unbekannten Säuren aus Baldrian. Acknowledgements Liebigs Ann Chem. 1957;603:158–68. Special thanks to Dr. Elisabeth Seigner, Dr. Martin Müller and Stefanie Gellan 11. Stoll A, Seebeck E, Stauffacher D. Isolierung und charakterisierung von of the Research Groups IPZ5c, IPZ1a and IPZ4a at the Bavarian State Research bisher unbekannten Inhaltsstoffen aus dem Neutralteil des frischen Center for Agriculture (LfL) for supply, guidance and assistance in the Baldrians. Helv Chim Acta. 1957;40:1205–29. fluorescence-microscopy, to Dr. Dieter Treutter† and Ina Tittel of the Technical 12. Reichling J, Horz KH, Bodesheim UB. Valeriana. In: Hänsel R, Keller K, University of Munich, TUM School of Life Sciences Weihenstephan, depart- Rimpler H, Schneider G, editors. Hagers Handbuch der Pharmazeutischen ment of fruit science for supply, guidance and assistance in the histological Praxis, vol. 6. 5th ed. Berlin: Springer; 1994. p. 1067–95. techniques and light-microscopy. 13. Houghton P. The chemistry of valeriana. In: Houghton P, editor. Valerian— the genus valeriana. Amsterdam: Harwood Academic Publisher; 1997. p. Competing interests 21–54. The authors declare that they have no competing interests. 14. Bos R. Analytical and phytochemical studies on valerian and valerian based preperations. Dissertation, Rijksuniversiteit Groningen; 1997. Availability of data and materials 15. World Health Organization ( WHO). WHO monographs on selected The datasets used and/or analysed during the current study are available from medicinal plants: Radix Valerianae. 1st ed. Geneva: World Health Organiza- the corresponding author on reasonable request. tion; 1999. p. 267–76. 16. Zacharias E. Ueber Secret-Behälter mit verkorkten Membranen. Botanis- Consent for publication che Zeitung. 1879;40:635–47. Not applicable. 17. Tschirch A, Oesterle O. Anatomischer Atlas der Pharmakognosie und Nahrungsmittelkunde. Leipzig: Tauchnitz; 1900. Ethics approval and consent to participate 18. Heuberger H, Penzkofer M. Demonstrationsprojekt Arzneipflanzen: Not applicable. Züchterische Verbesserung von Baldrian. In: 2. Kongress – Mit Pflanzen- züchtung zum Erfolg. 2017. https ://fnr.de/filea dmin/allge mein/pdf/veran Funding stalt ungen /Pflan zenzu echtu ng_2017/GFG_Band_55_2._Kongr ess_Pflan The valerian breeding project is part of the Demonstration Project Medicinal zenzu echtu ng_web.pdf. Accessed 12 Mar 2018. Plants (KAMEL) and is being supported by Fachagentur Nachwachsende 19. Penzkofer M, Ziegler E, Heuberger H. Contents of essential oil, valerenic Rohstoffe e.V. (FNR) on behalf of the German Federal Ministry of Food and acids and extractives in different parts of the rootstock of medicinal vale - Agriculture based on a decision of the German Bundestag. rian (Valeriana officinalis L. s.l.). JARMAP. 2014. https ://doi.org/10.1016/j. jarma p.2014.08.002. Publisher’s Note 20. Holzner-Lendbrandl J. Beiträge zur Kenntinis der Histogenese von Baldri- Springer Nature remains neutral with regard to jurisdictional claims in pub- anwurzeln unter besonderer Berücksichtigung der ölführenden Gewebe. lished maps and institutional affiliations. In: Egle K, Wilhelm T, editors. Beiträge zur Biologie der Pflanzen Band 39. Berlin: Duncker and Humblot; 1963. p. 323–66. Received: 6 September 2017 Accepted: 24 May 2018 21. Fridvalszky L. Über die mikroskopische Untersuchung der ätherischen Öle in der Wurzel von Valeriana officinalis L. Acta Biol Hung. 1957;8:81–9. Penzkofer et al. Plant Methods (2018) 14:41 Page 15 of 15 22. Violon C, Dekegel D, Vercruysse A. Microscopical study of valepotriates 35. Strazewicz WJ. Eec ff t of time of harvesting on the quality and oil content in lipid droplets of various tissues from valerian plants. Plant Cell Rep. of valerian root. In: Schaette R, editor Zur Wertbestimmung, Pharma- 1983;2:300–3. kologie und umweltbedingten Wirkstoffbildung von Radix Valerianae. 23. Özparpucu M, Rüggeberg M, Gierlinger N, Cesarino I, Vanholme R, Boer- Dissertation, Ludwig-Maximilian-Universität München; 1971. jan W, Burgert I. Unravelling the impact of lignin on cell wall mechan- 36. Eisenhuth F. Untersuchungen über die Modifikation der Leistung und ics: a comprehensive study on young poplar trees downregulated for der Qualität bei Valeriana officinalis L. durch Anbautechnik und Standort. CINNAMYL ALCOHOL DEHYDROGENASE (CAD). Plant J. 2017. https ://doi. Pharmaize. 1955;10:501–6. org/10.1111/tpj.13584 . 37. Johansen DA. Plant microtechnique. New York: McGraw-Hill; 1940. 24. Baranska M, Schulz H, Baranski R, Nothnagel T, Christensen LP. In situ 38. Schneider CA, Rasband WS, Eliceiri KW. NIH Image to ImageJ: 25 years of simultaneous analysis of polyacetylenes, carotenoids and polysaccha- image analysis. Nat Methods. 2012;9(Suppl 7):671–5. rides in carrot roots. J Agric Food Chem. 2005. https ://doi.org/10.1021/ 39. Kutschera L. Wurzelatlas mitteleuropäischer Grünlandpflanzen, Valeri- jf051 0440. anaceae, Baldriangewächse. Volume 2, part 2. Stuttgart: Gustav Fischer 25. Victor T, Delpratt N, Cseke SB, Miller LM, Cseke LJ. Imaging nutrient distri- Verlag; 1992. bution in the rhizosphere using FTIR imaging. Anal Chem. 2017. https :// 40. Szentpetery RG, Sarkany S, Fridvalszky L, Nagy J. Light-microscopic stud- doi.org/10.1021/acs.analc hem.6b043 76. ies on volatile oil excretion in Valeriana collina WALLR. Acta Agron Hung. 26. Gonzalez-Torres A, Rich AM, Marjo CE, Henderson RK. Evaluation of bio- 1966;15:321–34. chemical algal floc properties using reflectance Fourier-transform infrared 41. Schulz H, Krähmer A, Naumann A, Gudi G. Infrared and Raman spectro- imaging. Algal Res. 2017. https ://doi.org/10.1016/j.algal .2017.09.017. scopic mapping and imaging of plant materials. In: Salzer R, Siesler HW, 27. Kumar S, Lahlali R, Liu X, Karunakaran C. Infrared spectroscopy combined editors. Infrared and Raman spectroscopic imaging. Weinheim: Wiley-VCL; with imaging: a new developing analytical tool in health and plant 2014. p. 227–93. science. Appl Spectrosc Rev. 2016. https ://doi.org/10.1080/05704 42. Manley M, Williams P, Nilsson D, Gelad P. Near infrared hyperspectral 928.2016.11578 08. imaging for the evaluation of endosperm texture in whole yellow maize 28. Baranska M, Schulz H, Rösch P, Strehle MA, Popp J. Identification of (Zea maize L.) Kernels. J Agric Food Chem. 2009. https ://doi.org/10.1021/ secondary metabolites in medicinal and spice plants by NIR-FT-Raman jf901 8323. microspectroscopic mapping. Analyst. 2004. https ://doi.org/10.1039/ 43. Vermaak I, Viljoen A, Wiklund S, Wiklund Lindström S. Hyperspectral imag- b4089 33m. ing in the quality control of herbal medicines—the case of neurotoxic 29. Li Z, Tang T, Liang S, Ning X, Bai M, Wu H. the synthesis and storage sites Japanese star anise. J Pharm Biomed Anal. 2013. https ://doi.org/10.1016/j. of phenolic compounds in the root and rhizome of Echinacea purpurea. jpba.2012.11.039. Am J Plant Sci. 2012. https ://doi.org/10.4236/ajps.2012.34066 . 44. Türker-Kaya S, Huck CW. A review of mid-infrared and near-infrared 30. Pljevljakušić D, Rančić D, Ristić M, Vujisić L, Radanović D, Dajić-Stevanović imaging: principles, concepts and applications in plant tissue analysis. Z. Rhizome and root yield of the cultivated Arnica montana L., chemical Molecules. 2017. https ://doi.org/10.3390/molec ules2 20101 68. composition and histochemical localization of essential oil. Ind Crops 45. Liu C, Liu W, Chen W, Yang J, Zheng L. Feasibility in multispectral imaging Prod. 2012. https ://doi.org/10.1016/j.indcr op.2012.02.030. for predicting the content of bioactive compounds in intact tomato fruit. 31. Kromer K, Kreitschitz A, Kleinteich T, Gorb SN, Szumny A. Oil secretory Food Chem. 2015. https ://doi.org/10.1016/j.foodc hem.2014.10.052. system in vegetative organs of three arnica taxa: essential oil synthesis, 46. Heeger EF. Handbuch des Arznei- und Gewürzpflanzenbaues. Berlin: distribution and accumulation. Plant Cell Physiol. 2016. https ://doi. Deutscher Bauernverlag; 1956. org/10.1093/pcp/pcw04 0.32. 47. Bernath J. Cultivation of valerian. In: Houghton P, editor. Valerian—the 32. Schultz RA, Nielsen T, Zavaleta JR, Ruch R, Wyatt R, Garner HR. Hyper- genus valeriana. Amsterdam: Harwood Academic Publisher; 1997. p. spectral imaging: a novel approach for microscopic analysis. Cytom A. 77–100. 2001;43:239–47. 48. Heindl A, Hoppe B. Aufbereitung vor der Trocknung. In: Hoppe B, editor. 33. Kuska M, Wahabzada M, Leucker M, Dehne H-W, Kersting K, Oerke E-C, Handbuch des Arznei- und Gewürzpflanzenanbaus, vol. 2., Grundlagen Steiner U, Mahlein A-K. Hyperspectral phenotyping on the microscopic des Arznei- und Gewürzpflanzenanbaus IIBernburg: Verein für Arznei- scale: towards automated characterization of plant-pathogen interac- und Gewürzpflanzen SALUPLANTA e.V; 2010. p. 235–9. tions. Plant Methods. 2015. https ://doi.org/10.1186/s1300 7-015-0073-7. 49. Neumaier G. Verfahrenstechnische analyse und Entwicklung von Ernte- 34. Eisenhuth F. Qualitätsforschung und Leistungsfragen bei Valeriana offici- und Reinigunssystemen für Baldrianwurzeln. Dissertation, Technische nalis. Pharmaize. 1956;11:271–86. Universität München; 2017. Ready to submit your research ? Choose BMC and benefit from: fast, convenient online submission thorough peer review by experienced researchers in your ﬁeld rapid publication on acceptance support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year At BMC, research is always in progress. Learn more biomedcentral.com/submissions
Plant Methods – Springer Journals
Published: Jun 4, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera