Study of petal topography of Lysimachia arvensis grown under natural conditions

Abstract

The present study revealed that adaxial and abaxial petal epidermises of the blue-flowered Lysimachia arvensis consist of elongated, multi-micro-papillate cells, which may aid the rapid petal expansion. The epidermal cells are covered by a wrinkled relief, which is further ornamented by submicron features that increase in size the surface area of lobes; this may be a well-adapted mechanism of the small-sized flowers of L. arvensis with the short life span. The sculpturally increased surface area of adaxial epidermal cells of petals is expected to contribute to optical and adhesive properties, and wettability of the floral tissues. The adaxial and the abaxial petal surfaces of L. arvensis possess submicron cuticular folds, smaller than the sub-wavelength visible spectrum, which reflect radiation of shorter rather than longer wavelengths, whereas intense absorption was detected in the red spectral region. Also, three-celled capitate trichomes with a pigmented spherical head, which are densely distributed at the corolla margins of L. arvensis, may be involved in adhesive, defensive and functional properties of the floral tissues.

Keywords:

• epidermis
• flower
• Lysimachia arvensis
• petal
• properties
• surface
• topography

Introduction

Flowers are among the most spectacular products of nature. Blossoming of plants is regulated by mechanisms that act to ensure that flower emergence occurs in suitable environmental conditions (Elzinga et al. 2007; Tooke and Battey 2010). Major importance has been given to floral advertisement, which is related to multifunctional traits; for example, a three-dimensional cuticular relief, observed on petal surfaces, protects the floral tissues against physical, chemical and biological attack, it influences optical properties, it reduces the absorbance of UV radiation that reaches the cells and forms microsculptures (Kevan and Lane 1985; Jacobs, Koper, and Ursem 2007; Whitney and Glover 2007; Domínguez, Heredia-Guerrero, and Heredia 2011; Javelle et al. 2011; Glover et al. 2013). It is worth mentioning that multifunctional, submicron features of petal epidermises, which serve as interfaces between the tissues and their environment, provide a potential natural template for man-made applications, which have been transferred to biomimetic materials (Ma and Sun 2009; Bhushan and Her 2010; Shimomura 2010; Uchida et al. 2010; Qian et al. 2011; Telford et al. 2013).

Lysimachia arvensis (L.) U. Manns & Anderb. (Primulaceae), formerly Anagallis arvensis L. (Manns and Anderberg 2009), is a slender little plant, creeping on the ground with spreading stems in wild and cultivated fields, on road sides and in gardens, and considerable research has focused on it (Marsden-Jones and Weiss 1960; Verma 1972; Simpson and Neff 1983; Green, Havelange, and Bernier 1991; Hernández et al. 1991; Hao et al. 2004; Anderberg, Manns, and Källersjö 2007; Manns and Anderberg 2007; Jiménez-López et al. 2015). Lysimachia arvensis is probably native to the Mediterranean region that has been dispersed world-wide (Chrtek and Osbornová-Kosinová 1986; Brockett and Cooperrider 1983; Manns and Anderberg 2011; de Foucault, El Beyrouthy, and Arnold-Apostolides 2013). However, over the last few decades, the number and size of populations of this species declined due to modern land-use techniques, application of fertilizers, herbicide use and cleaning procedures (Brütting et al. 2012; Bergmeier and Strid 2014; Šilc, Lososová, and Vrbničanin 2014).

The objective of this study was to identify morphological features of petals of L. arvensis, to understand adaptations of floral tissues – expanded under field conditions – to the abiotic environment. Lysimachia arvensis blossoms by exhibiting bluish flowers under adequate water availability in the field, during spring, i.e. the main flowering period in the Mediterranean region (Harris 2007; Rhizopoulou et al. 2012; Farré-Armengol et al. 2014; Rhizopoulou and Pantazi 2015). Stereoscopy, scanning electron microscopy and atomic force microscopy were used to study the micromorphology of petals of L. arvensis. To the best of our knowledge, the topography of adaxial and abaxial petal surfaces of L. arvensis using high-resolution imaging, which enables detailed study of plant surfaces and understanding of relationships between floral tissues and environmental stimuli, has not hitherto been investigated.

Material and methods

Plant material and study area

Plants of the annual plant Lysimachia arvensis (L.) U. Manns & Anderb. (Primulaceae) grow under ambient conditions, creeping close to the ground, in the campus of the University of Athens (38°14.3′N, 23°47.8′E, altitude 250 m, above sea level). Flowering of plants of L. arvensis was observed in fifty individuals, on a regular basis, i.e. every day, between April and May. The mean monthly temperature ranged from 17 ± 2°C (April) to 22 ± 3°C (May), and the monthly precipitation varied from 26 ± 4 mm (April) to 15 ± 2 mm (May). Induction of flowers before the dry summer, in the Mediterranean region, may be interpreted as a predictably evolved response of plants to unfavourable water deficit conditions. Floral longevity was defined as the time from opening to the wilting of the whorls. Fifty expanded, turgid flowers were randomly harvested from L. arvensis plants growing in the field. The corolla diameter was measured on intact flowers of L. arvensis growing in open habitats using a ruler.

Microscopy and stereoscopy

The study was carried out on fully mature petal regions of L. arvensis (Figure 1). Petal lobes were viewed using a Zeiss Stemi 2000 C stereoscope equipped with a ProgRes C3 digital camera, and the pictures were recorded using a personal computer. Samples from the middle part of the petal blade were carefully cut into square pieces (4 mm²) and fixed in 3% glutaraldehyde in sodium phosphate buffer at pH 7, at room temperature, for 2 h. The samples were then washed three times by immersion in buffer for 30 min each time, and post-fixed in 1% OsO₄ in the same buffer at 4°C. Dehydration of the material was accomplished with a gradient of increasing concentrations of acetone solutions. Dehydrated tissues were embedded in SPURR (Serva) resin. Semi-thin sections of resin-embedded tissue (LKB Ultratome III microtome) were stained in Toluidine Blue ‘0’, in 1% borax solution, photographed and digitally recorded using a Zeiss Axioplan light microscope (Carl Zeiss Inc., Thornwood, NY, USA) equipped with a digital camera (Zeiss AxioCam MRc5). Also, dehydrated samples were dried at the critical point in a Bal-Tec CPD-030 dryer, mounted with double adhesive tape on stubs, and sputter coated with 20-nm gold in a Bal-Tec SCP-050. The adaxial and abaxial epidermises of petals were examined using a JEOL JSM 840 scanning electron microscope (JEOL Ltd, Tokyo, Japan); software associated with this microscope was used for the production of the digital images. Adaxial and abaxial petal areas (5 × 5 μm) were imaged by using a tap mapping atomic force microscope (Multimode SPM, Veeco, Santa Barbara, CA, USA). Several parameters were analysed and processed, using the software package Nanoscope III (Veeco, USA), to detect detailed information about the petal surfaces. The quantitative measurements include horizontal and vertical distances that represented the height of a step between nanofolds, surface roughness, surface distance, i.e. the length between the markers, and the surface area ratio that represents the density of nanofolds. Topological, quantitative components obtained from 40 different samples are presented in representative micrographs and the mean values given in Table 1 are followed by the standard deviation (± SD); height, amplitude modes and distances between the cursors of the presented lines in Figure 4(B, Figure 4E) correspond to the average profile of three lines crossing each one of the 40 samples drawn by the investigator, according to McNally and Borgens (2004).

Figure 1. Flowers of Lysimachia arvensis grown in the field (left, scale bar: 1 cm), the sampling region of fully expanded petals is indicated by an arrow and a transverse section through the petal expanded region is presented (right, scale bar: 10 μm).

Table 1. Nanometric estimates of trais of the adaxial and the abaxial petal surfaces of Lysimachia arvensis using atomic force microscopy.

Petal surface	Vertical distance (nm)	Horizontal distance (nm)	Surface distance (nm)	Roughness (Ra) (nm)	Surface area ratio (Sr)
Adaxial	150±2^a	1,445 ± 7^c	1,504 ± 8^c	39 ± 3^e	1.11 ± 0.04^f
Abaxial	106±5^b	430 ± 5^d	460 ± 5^d	37 ± 2^e	1.09 ± 0.05^f

Data are means of 40 samples (± SD); means followed by the same lower case letters are not statistically different at p ≤ 0.05.

In situ measurements of optical properties of fresh petal tissues

Reflectance, transmittance and absorbance were measured in situ on fresh, expanded flowers of L. arvensis adjacent to those sampled for morphological measurements, using a UV/VIS spectrophotometer (Perkin Elmer Lambda-950) equipped with an integrating sphere and glass-fibre tubes (Stratakis et al. 2009; Gkikas, Argiropoulos and Rhizopoulou 2015).

Spectrophotometric measurements of absorption of petal extracts

Fresh petal-tissues were collected from the same corollas sampled for the in situ measurements of reflectance of L. arvensis and were frozen in liquid nitrogen; then, they were crushed into fine powder and the pigments were extracted with 1% (volume/volume) HCl in methanol and 80% ethanol solutions. Spectrophotometric measurements of absorption bands of solutions of filtered and purified petal extracts were estimated using a Novaspec II spectrophotometer (Pharmacia Biotech, Cambridge, UK).

Flower photography

Black and white photographs of L. arvensis flowers were taken through monochromatic filters (Omega Optical, Brattleboro, VT, USA) in selected spectra, i.e. at 550 nm, 450 nm and UV, under ambient conditions and natural light, using a portable camera-system based on an 8-bit charge-coupled device sensor, which was developed at the Institute of Electronic Structure and Laser (Foundation for Research & Technology Hellas, Heraklion, Greece).

Statistical analysis

Statistically significant differences between measurements in the two petal surfaces were estimated using one-way analysis of variance and Welch & Brown-Forsythe tests. Statistical analysis was performed with the SPSS statistical program. The reported values are means ± SD.

Results

The corollas of L. arvensis are small (5 ± 1 mm long, Figure 1: left) and exhibit a 2-day lifespan. Lysimachia arvensis takes approximately 1 month to flower and the duration of blossoming in the field averaged a 2-week period extending from April until May.

Transverse sections of petals show papillate epidermal cells at the adaxial and abaxial sides (Figure 1; right) and a narrow mesophyll (Figure 2A). Adaxial and abaxial surfaces of petals of the small, short-lived flowers of L. arvensis possess multipapillate, elongated, epidermal cells (Figure 2B, Figure 2C, respectively), which exhibit approximately five (Figure 2C) to ten (Figure 2B) micro-papillae surfaces, arranged in a single row per cell. On the outer margin of petals, numerous (Figure 3A) capitate-stalked trichomes (approximately 15 hairs mm⁻¹) consist of a nearly spherical head cell, a stalk cell and a basal cell, which emerges from the epidermis (Figure 3B).

Figure 2. Scanning electron micrographs of petal of Lysimachia arvensis. (A) Transverse section (scale bar: 5 μm), (B) adaxial epidermis and (C) abaxial epidermis (scale bar: 10 μm) consisting of multi-micro-papillae arranged in elongated cells, as indicated by arrows at the edges of cells.

Figure 3. (A) Stereoscopic (scale bar: 0.5 mm) micrograph and (B) scanning electron micrograph (scale bar: 20 μm) showing the outer margin of petals of Lysimachia arvensis with the predominant three-celled capitate trichomes consisting of a nearly spherical head, a stalk cell and a basal cell.

The epidermal cells of petals are covered by a ridged, cuticular relief (Figure 4). High-resolution imaging of adaxial (Figure 4A) and abaxial (Figure 4D) petal epidermises of L. arvensis revealed cuticular folds at the nanometer scale (Table 1). The adaxial and abaxial epidermises possess cuticular folds with a vertical distance of 150 nm and 106 nm, respectively, and a horizontal distance of 1450 nm and 450 nm, respectively. Therefore, adaxial epidermal cells of petals of the small flowers of L. arvensis possess a horizontal distance that is 3.5-fold higher than that of the abaxial epidermal cells (Table 1); in other words, spacing between folds of the adaxial and abaxial epidermises is approximately 1.5 μm and 0.5 μm, respectively. The adaxial horizontal distance is approximately 9.5-fold higher than the adaxial vertical distance, and the abaxial horizontal distance is four-fold higher than the abaxial vertical distance (Table 1).

Figure 4. Atomic force micrographs of (A–C) adaxial and (D–F) abaxial petal surface of Lysimachia arvensis. (A, D) Three-dimensional profiles; (B, E) integrated line of measured points on plane profiles; and (C, F) profile views of the line section.

Absorption, reflectance and transmission measured on intact, expanded fresh petals of L. arvensis were found to be approximately 30% between 360 and 440 nm (Figure 5A). The absorption increased from 32% at 400 nm to 82% at 630 nm (Figure 5B), whereas reflectance and transmission fell to 10% between 580 and 630 nm (Figure 5A). In particular, reflectance was about 15% at 320 nm, 32% in the blue spectral range, 8% at 610 nm and 35% at 680 nm (Figure 5). Also, the absorption of petal extracts of L. arvensis increased from 500 to 540 nm (Figure 6A), but the absorption of fresh tissues was higher than that of petal extracts (Figures 5, Figure 6A). A remarkable difference in absorption spectra between intact petals and solutions of petal extracts occurred at 600 nm, being approximately 80% and 10%, respectively (Figures 5B, Figure 6A). The presence of UV-absorbing compounds (Figures 5, Figure 6) is also indicated and this may protect the short-lived tissues from harmful radiation. It appears that petals of L. arvensis are still absorptive at 550 nm (Figure 7A), whereas they are mostly reflective at 450 nm (Figure 7B). The central area of the corolla absorbs UV, contrasting against the UV-reflecting mid-part and periphery of the petal lobes (Figure 7C).

Figure 5. In situ spectral reflectance (Α), transmission (Α) and absorption (Α, Β) of petals of Lysimachia arvensis.

Figure 6. Absorption spectra of fresh petal extracts of Lysimachia arvensis in 1% HCl in (A) methanol and (B) 80% ethanol solutions, over the 250- to 700-nm wavebands.

Figure 7. Representative images of intact, small-sized flowers (5 ± 1 mm, long) of Lysimachia arvensis photographed under ambient conditions. (A) Absorptive fully expanded petals and reflective axile anthers at 550 nm. (B) Reflective petal areas contrasting against the absorptive central part of the corolla (as indicated by the arrow) and axile anthers at 450 nm. (C) A defined UV absorptive central area (i.e. the dark part of the corolla) contrasting against the UV-reflecting mid-part of the lobe and the periphery.

Discussion

The longitudinally elongated, multipapillate epidermal cells, found on both the adaxial and the abaxial surfaces of petals may aid the rapid petal expansion of the flowers of L. arvensis with the short lifespan (Barthlott 1981; Kay, Daoud, and Stirton 1981; Argiropoulos and Rhizopoulou 2012a).

The three-celled capitate trichomes, densely arranged at the corolla lobe margins (approximately 15 hairs mm⁻¹), consist of terminal spherical heads that appear purplish (Figure 1A), being filled with pigments (e.g. anthocyanins) that have most probably been transferred from epidermal cells through the basal and stalk cells (Fahn 1979; Manns and Anderberg 2007). Hence, capitate pigmented trichomes positioned on the periphery of petals form boundaries that provide contrast to the background and may play a role in visible distinction (Baird et al. 2013). It has been argued that if colour is to be used as a signal to be spotted by other organisms, then it must stand out from a green-grey-yellowish background (Lehrer, Srinivasan, and Zhang 1990; Miller, Owens, and Rørslett 2011). Trichomes found on petals and reproductive organs of plant species from other families are involved in defence, attractiveness, feeding deterrence and UV protection of the tissues (Vrachnakis 2002; Clegg and Durbin 2003; Martin and Glover 2007; Glas et al. 2012; Vitarelli et al. 2015). Also, glandular trichomes observed on floral tissues of other Lysimachia species secrete fluids, which are collected by pollinators (Simpson and Neff 1983; Schäffler, Balao, and Dötterl 2012). The brightly coloured, spherical heads of the capitate trichomes, probably mimicking shining nectar droplets, may facilitate pollen dispersal of the nectarless flowers of L. arvensis (Aldasoro, Aedo, and Navarro 2000; Possobom, Guimarães, and Machado 2015). In considering that pollen is the only reward for pollinators (Gibbs and Talavera 2001; Raine and Chittka 2007), the pigmented capitate trichomes on the margin of petals may attract insects by deceit. Also, the marginal, stalked, purplish capitate trichomes of the small flowers of L. arvensis, possess an amplified peripheral surface volume, which may play a role of floral ‘façade’, which may result, to some extent, in the protection of drifted pollen grains. It is worth mentioning that the size of the isopolar, tricolporate pollen grain of L. arvensis is the same order of magnitude as that of the head cells of the capitate trichomes prevailing at the corolla apex (Fernandez and Garcia-Martin 1985; Moneim, Atta, and Shehata 2003). The average pollen grain diameter is approximately 18 μm, the average polar and equatorial axis lengths are 18.57 μm and 18.25 μm, respectively (Moneim, Atta, and Shehata 2003), and the diameter of head cells of capitate trichomes is approximately 18 μm. It is probable that the pollen grains of L. arvensis adhered to the volume of capitate trichomes are retrieved and redistributed at the corolla lobes, facilitating self-fertility of the species during the closure of (autogamous) flowers at the end of the day (Imhoff et al. 1979; Simpson and Neff 1983; Gibbs 2014). Actually, flowers of L. arvensis open in the morning and close in the evening (Imhoff et al. 1979; Ortiz et al. 2015; personal observation). Interestingly, closed flowers of L. arvensis disappear within the greenish background, because the abaxial bluish petal surfaces are mostly covered by the abaxial, green sepal surfaces.

In addition, flowers of L. arvensis close in cloudy, misty and humid weather (Gibbs and Talavera 2001; personal observation); this closure may be coordinated to submicron traits of petals, because petal surface wettability has been found to be affected by cuticular, corrugated folds of the tissues (Wagner et al. 2003; Argiropoulos and Rhizopoulou 2012b; Taneda et al. 2015). High-resolution imaging revealed nanosculptures that may influence the adhesive properties of petals and affect the tissue water status (Barthlott 1981; Gale and Owens 1983; Miller, Owens, and Rørslett 2011; Whitney et al. 2011; Chimona et al. 2012; Koch et al. 2013; van Doorn and Kamdee 2014; Buschhaus, Hager, and Jetter 2015). It is expected that the vertical-to-horizontal distance ratio (0.10) of the adaxial (inner) petal surfaces, presumably leading to an elevated contact area between the folds of the petal surface and water, will affect both the firmness of the contact with water and the tissue water status, i.e. by retaining water during a misty, humid day, whereas, the vertical-to-horizontal distance ratio (0.25) of the abaxial surface indicates a decline in water retention on the outer petal surfaces.

It appears that cuticular microfolding of adaxial, epidermal site of the small corollas of L. arvensis increases in size the surface area of adaxial epidermal cells and aids optical and adhesive properties of petal surfaces of the short-lived flowers. The cuticular relief, representing an imprint of structural epidermal features at the nanoscale level, displays a higher horizontal-to-vertical distance ratio at the adaxial surface (9.63) than that at the abaxial surface (4.06). Vertical distances between folds estimated on the adaxial and the abaxial surfaces of petals of L. arvensis are comparable to the sub-wavelength regimen (Table 1). Nanometric surface structures much smaller than the visible wavelength spectrum can be much more effective in reflecting radiation of shorter rather than longer wavelengths and producing blue colours, which could favour attraction of insect pollinators (Papiorek, Rohde, and Lunau 2013). Floral tissues exhibit blue colours through modifications of pigments and structural traits (Lee 2007; Yoshida, Mori, and Kondo 2009; Vignolini et al. 2013).

The surface roughness of the adaxial petal surfaces of L. arvensis (39 nm) may aid the capture of low angle light by the tissues that enhances the brilliance of flowers (Brodersen and Vogelmann 2007); this may be advantageous for the proximate to the ground open corollas, under ambient conditions. Hence, petal nanosculpturing may result in a coloured appearance with a greater reflectance spectrum and colour stability for a (relatively) longer-lasting period, in comparison to the pigmented coloration that fades faster with petal desiccation and loss of turgor (Mamareli and Rhizopoulou 2009; Glover and Whitney 2010).

Spectral reflectance and transmission of incident light falling on petals of L. arvensis were inversely related to the absorption (Figure 5A). Absorption, reflectance and transmission in fresh, intact petals were found to be approximately 35% between 360 and 440 nm. The absorption increased to 80% from 580 to 640 nm, and a maximum was detected in the 620–630 nm range (Figure 5B), whereas transmission reduced to 10% (Figure 5A). Absorption that varies from 35% at 490 nm to 75% at 550 nm (Figure 5B), indicates anthocyanins confined to petal lobes of L. arvensis (Ishikura 1978; Freyre and Griesbach 2004). It is expected that convexly shaped, adaxial epidermal cells of petals enhance the proportion of incident light that enters the tissues and can be absorbed by pigments (Kay, Daoud, and Stirton 1981; Bone, Lee, and Norman 1985; Vogelmann 1993; Pfündel, Agati, and Zerovic 2006; Lee 2007; Argiropoulos and Rhizopoulou 2013; Gkikas, Argiropoulos and Rhizopoulou 2015); light passing through epidermal cells interacts with pigments, which are involved in defence, protection and signalling (Gorton and Vogelmann 1996; Miller, Owens, and Rørslett 2011). On one hand, absorption in the 520–640 nm range produces blue colours that are the preferred attractant to insect pollinators (Kay, Daoud, and Stirton 1981; Morawetz et al. 2013). On the other hand, reflectance was low (< 10%) in the yellow-orange spectral range (Figure 5A). Elevated reflectance was detected at 300–450 nm and 670–700 nm, which is in agreement with results from other Lysimachia species (Schäffler, Balao, and Dötterl 2012). The absorbance of pigmented solutions of petal extracts was found to be substantially lower than that of intact petals, in the 520–620 nm range, and this could be attributed to the effect of some structural elements of petals.

Petals of L. arvensis seem to be UV protected by means of high absorption (> 80%) in the UV wavelength range (Frohlich 1972; Ishikura 1978; Gronquist et al. 2001; Jacobs, Koper, and Ursem 2007; Koski and Ashman 2014; Llorens et al. 2015; Papiorek et al. 2015). The UV absorptive, dark patch in the central part of the flower contrasts against the mid-part of the lobe and the periphery; this is shown to be a feature of flower recognition and a guide to floral visitors, which concomitantly preserve pollen grains from injurious effects of UV radiation (Chittka et al. 1994; Lunau 2000; Moneim, Atta, and Shehata 2003; Garcia et al. 2014).

Notes on contributors

Sophia Rhizopoulou is associate Professor of Plant Environmental Physiology in the National and Kapodistrian University of Athens (Department of Biology) and has published numerous research articles and books related to plant life in the Mediterranean region [see: http://publicationslist.org/sophia.rhizopoulou].

Emmanuel Spanakis is researcher at the University of Crete (Department of Materials Science and Technology, and Affiliated member of the Institute of Electronic Structure and Laser in Crete.

Apostolos Argiropoulos is a post-PhD fellow in the National and Kapodistrian University of Athens.

Acknowledgments

The research was supported by the grant PENED 03ED174 to S.R., co-financed by European Social Fund-ESF and the Greek Ministry of Research and Innovation.