Introduction

Carnivorous plants have adapted to grow in nutrient poor habitats and derive lacking nutrients from trapping and consuming animals1. Plants from the genus Nepenthes rely on the capturing success of their passive trapping organs called pitchers. Insects with both smooth attachment pads consisting of compliant material and hairy attachment pads bearing fibrillar surface structures2,3,4 can be trapped by pitchers, however, ants (smooth pads) comprise the main prey of most Nepenthes species5,6,7.

Pitchers of Nepenthes plants are composed of several zones, such as the leaf-like lid, ribbed peristome, waxy zone, transitional zone lacking specialized surface structures and digestive zone with numerous digestive glands8,9. These zones show different macro-morphologies as well as surface micro-architectures and play different roles in attracting, trapping, retaining and digesting of animal prey (see the review by Moran and Clarke10).

Nepenthes pitchers employ different trapping and retention mechanisms. The contribution of the waxy zone for trapping and retaining functions has been repeatedly reported1,9,11,12,13,14,15,16,17,18,19,20. During the past decade, the role of water-lubricated peristome surface21,22 and highly viscous trapping fluid23,24 for prey capturing was discovered. For a number of species, the relative contributions of the most widespread trapping structures, such as the peristome, wax crystals and viscous fluids, were recently inverstigated20,25,26.

In many Nepenthes species, the waxy zone situated inside the pitcher, just below the peristome (Fig. 1A), bears lunate cells (downward directed modified stomata), decreasing in number towards the pitcher's base and a thick epicuticular wax coverage, which may consist of several layers of wax crystals1,9,11,12,13,14,15,16,17,18,19,20,27. Our studies on N. alata revealed two easily distinguishable, superimposed wax layers, the upper and lower layers, differing in microstructure, chemical composition and mechanical properties13,15,20,27. The lower layer has a foam-like structure and is composed of interconnected membranous platelets. Irregular cavities occur between these crystals, positioned at acute angles relative to the pitcher wall surface. The upper layer consists of densely placed, separate irregular platelets arranged perpendicularly to the subjacent layer. Platelets of the upper and lower wax layer are connected to one another by thin stalks.

Figure 1
figure 1

Micro-morphology of test surfaces.

(A) Longitudinally dissected pitcher of N. alata. (B) Fracture of the wax coverage in the waxy zone, consisting of two layers, the upper and the lower wax layer. (C) Top view of upper wax layer where separate platelet-shaped wax crystals bearing thin stalks are clearly visible. (D) Top view of the lower wax layer, composed of interconnected platelets with cavities in-between. (E) Micro-rough epoxy resin replica with an asperity size of 3 μm. (F) Smooth epoxy resin replica. CA, cavities; CU, cuticle; IP, interconnected membranous platelets; LL, lower layer; ST, stalks; SP, separate irregular platelets; UL, upper wax layer; WZ, waxy zone. (B–D) Cryo-SEM micrographs. (E,F) SEM micrographs.

Full size image

Attachment ability of insects with both smooth and hairy adhesive pads was examined using locomotion experiments, in which insects either were unable to move or walked very slowly and covered extremely small distances on the upper wax layer of the intact waxy zone9,28. No such experiments were performed on the lower wax layer. In addition, traction forces were measured for insects with both types of adhesive pads on the upper wax layer13,17,19, but only for insects with hairy pads on the lower layer13. In all of these experiments, traction forces were strongly reduced on the waxy layers as compared with those on wax free surfaces.

According to our previously proposed hypotheses29, a reduction of insect adhesion on a plant surface covered with a 3D crystalline epicuticular wax may be explained by (1) the decrease of real contact area between the insect adhesive pads and the plant substrate caused by surface micro-roughness (roughness-hypothesis); (2) the absorption of the pad fluid by the structured wax coverage (fluid-absorption-hypothesis); (3) the dissolving of the wax crystals by the pad secretion, resulting in substrate slipperiness and hydroplaning (wax-dissolving-hypothesis); (4) the contamination of pads by wax crystals (contamination-hypothesis). For the N. alata waxy pitcher surface, the decrease in insect attachment ability was explained by prevention of insect adhesion via reduction of the real contact area9,13,17,18 and/or contamination1,11,13,28,30. Also, due to the fragile and brittle nature of wax crystals and their small dimensions, insects are not able to apply their claws for interlocking with crystals in order to climb up the pitcher wall17.

Although a number of studies have reported on the anti-attachment effect of the N. alata wax coverage on insects and discussed different mechanisms contributing to the anti-adhesive function1,9,11,13,17,18,28,30, adhesion force measurements on these surfaces have not been performed. This was mainly due to the absence of a reliable and sensitive method for testing surfaces with very low adhesive capabilities. Using a newly developed experimental approach31 and applying polydimethylsiloxane (PDMS) half-spherical indenters with elastic properties similar to those of the smooth adhesive pads in insects, we carried out pull-off force measurements on the lower and upper wax layers in the waxy zone of N. alata pitchers. The aim of our study was to estimate the adhesion properties of these layers. This information is essential to understand details of the trapping mechanism of the upper wax layer.

Results

Micro-morphology of test surfaces

The fractured wax coverage in the intact waxy zone confirmed our previous findings13,15,20,27 about the composite structure of two superimposed wax layers on the top of the cuticle in N. alata pitchers (Fig. 1B). The top views of the wax layers' surfaces displayed further details of the wax structure. The lower wax layer resembled foam and was composed of membranous platelets with irregularly-shaped cavities in-between (Fig. 1D). The upper layer emerged as separate irregular platelet-shaped wax crystals without a regular pattern (Fig. 1C). Thin stalks connected the platelets of the upper layer with the lower wax layer. Surfaces of the smooth and micro-rough epoxy resin replicas are shown in Fig. 1E,F.

Surface roughness

The standard roughness parameters (mean roughness Ra, root mean square of roughness r.m.s. and the maximum height of the profile Rz) of the tested samples, apart from the smooth replica, were relatively similar in their values (Table 1). Further analysis of topographical data in the rough surfaces showed two types of roughness, influenced by either (1) one wavelength (micro-rough replica) (Fig. 2A,B) or (2) two wavelengths (both plant surfaces) (Fig. 2C–F). While the one-wavelength type was represented by coarse domains with a smooth surface on hills and in valleys, the two-wavelength type was composed of one wavelength with a high frequency and low amplitudes caused by the wax coverage and a second wavelength with a low frequency and high amplitudes due to the epidermal cells' topography (Fig. 2D,F). High-frequency filtering of roughness data (not shown) supported these two types of roughness characters.

Table 1 Geometrical characteristics of test surfaces and absolute pull-off force values obtained on different test surfaces. n, number of measurements; Ra, mean roughness; r.m.s., root mean square of roughness; Foff, pull-off force; Rz, maximum height of the profile. Data are presented as mean ± s.d
Full size table
Figure 2
figure 2

Images of rough test surfaces obtained with white light interferometer.

(A,B) One-wavelength roughness of the micro-rough replica represented by coarse domains with a smooth surface on hills and in valleys. (C–F) Two-wavelength roughness of the upper wax layer (C,D) and the lower wax layer (E,F) showing combined roughness patterns due to the presence of wax crystals and epidermal cells. (A,C,E) Height maps. (B,D,F) 3D oblique maps.

Full size image

Pull-off forces

Absolute pull-off force values obtained on different test surfaces in the first type of experiment (three single measurements with each probe on the smooth replica, micro-rough replica and one of the wax layer samples) are presented in Table 1. We obtained significantly different normalized pull-off forces on tested samples (normality test: failed, P < 0.050; Friedman repeated measures ANOVA on ranks: Χ2 = 39.6, d.f. = 3, P = 1.3482E-8; Table 2 and Fig. 3A). Among all surfaces, the upper wax layer showed the lowest force. Forces on both the upper wax layer and micro-rough replica were significantly lower than those on the smooth replica and lower wax layer. As for the latter two samples, pull-off force on the lower layer was slightly higher compared to the smooth surface.

Table 2 Results of multiple pairwise comparisons (Wilcoxon signed rank test with Bonferroni correction) of normalized pull-off forces obtained on different surfaces. LL, lower wax layer; rough, micro-rough epoxy resin replica; P, Bonferroni corrected probability value; smooth, smooth epoxy resin replica; UL, lower layer
Full size table
Figure 3
figure 3

Normalized pull-off forces obtained on different surfaces in the first (A) and second (B) type of experiment.

The ends of the boxes define the 25th and 75th percentiles, with a line at the median and error bars defining the 10th and 90th percentiles. Circles indicate outliers. 3, rough epoxy resin replica having 3 μm asperity size; LL, lower wax layer; S, smooth replica; S1, smooth sample before contact with the wax surface; S2, smooth sample after contact with the wax surface; UL, upper wax layer.

Full size image

In the second type of experiment (three single measurements with the same probe on the smooth sample, upper wax layer and again the smooth sample), the comparison of normalized forces obtained in the first and second tests on the smooth sample showed no significant difference (normality test: passed, P = 0.069; paired t-test: t = 0.570, d.f. = 14, P = 0.578; Fig. 3B).

The effective E-modulus of polymer probes amounted to a value of 49.47 ± 13.99 kPa at a mean applied force of 697.27 ± 129.59 μN (n = 40). Using this value in equations, we obtained pressure values of pmin = 138.07 Pa for 0.5 mN applied force and pmax = 173.96 Pa for 1 mN applied force.

Effects of adhesion experiments on surfaces of samples and probes

Scanning electron microscopy (SEM) investigations of samples and probes performed after pull-off force measurements demonstrated that, with the exception of the upper wax layer, changes neither of the test surface nor of the probe topography occurred during experiments (Fig. 4B). In the case of the upper wax layer, a strong contamination of probes used in measurements was shown (Fig. 4A). Contaminated areas were covered with wax particles arranged in clusters corresponding to the pattern of epidermal cells (Fig. 4D). Further magnification showed a conspicuous shape of adhering wax crystals (Fig. 4F) similar to that observed in the intact upper layer (Fig. 1C). The wax structures on the contaminated probes were composed of platelets with regular margins and rather long, thin shafts, reaching from half to full size of the crystal platelets. The surface of these platelets was not completely smooth, but had a particular roughness due to an apparent folding. Apart from distinct wax crystals, smaller wax particles were found in the contact area. This indicates the fragile nature of crystals in the upper wax layer. We also observed some changes in the plant surface after the experiments (Fig. 4C,E). The contact zone lacked the upper wax layer and displayed a wax structure with irregular membranous platelets and cavities resembling the lower wax layer (Fig. 4G).

Figure 4
figure 4

Cryo-SEM micrographs of polymer probes (A,B,D,F) and plant samples (C,E,G) after adhesion experiments.

(A) Surface of the probe highly contaminated with wax crystals after being in contact with the upper wax layer. (B) Clean (non-contaminated) probe surface after contact with the lower wax layer. (C) The area of the waxy zone after being in contact with the sticky probe in an adhesion experiment. The contact zone has a whitish appearance due to changes in the wax coverage. (D) On the contaminated surface of the probe, wax crystals are arranged in clusters corresponding to the pattern of epidermal cells. (E) The border between the intact upper layer and the contact zone showing the remaining lower layer of wax. (F) The probe surface shown in (D) at high magnification. Adhered wax crystals are similar to those from the upper wax layer: platelets having a fine ridged structure and bearing thin stalks. (G) The contact zone with the lower wax layer after detachment from the probe. CA, cavities; IP, interconnected membranous platelets; LL, lower layer; S, stalks; SP, separate irregular platelets; UL, upper layer; WC, wax crystals.

Full size image

Discussion

Using a newly developed measuring method31, we characterized adhesive properties of the two wax layers in the waxy zone of the N. alata pitcher. It is well known that adhesive properties of biological surfaces are greatly affected by a number of factors, above all by surface topography17,18,29,32,33. Therefore, in addition to pull-off force measurements, we performed a detailed characterization of surface samples, including structural analysis and roughness measurements. Also, mechanical properties of counterparts strongly influence adhesion. Here, by using tacky half-spherical PDMS probes31 having an elasticity modulus of E = 50 kPa, we simulated the material of smooth adhesive pads in insects. For this type of attachment device, elasticity moduli of 27 kPa in Tettigonia viridissima (Orthoptera, Tettigoniidae)34 and 12–625 kPa in Carausius morosus (Phasmatodea)35 have been previously reported. Moreover, we applied a normal force ranging from 500 to 1000 μN resulting in a pressure of pmin = 138.07 Pa and pmax = 173.96 Pa, respectively. According to data in the literature17, in a typical prey such as an ant, the contact area of a single adhesive organ (arolium) is about 10000 μm2. For a middle-sized ant, weighing 3 mg and having from three to six feet in contact, we obtain a pressure of p6feet = 83 Pa to p3feet = 327 Pa. Thus, the pressure applied in the present study was within the range of that of a typical Nepenthes pitcher's prey.

We are aware that soft PDMS half-spheres lack several of the complex features of insect attachment devices, such as the presence of a secretion, a complex ultra-structure, a gradient in material properties and a pair of claws. However, the strength of our approach lies in the testing of the interaction between an adhesive material and Nepenthes wax layers under exactly defined conditions, such as the precise contact geometry and material properties of the adhesive counterpart as well as biologically relevant contact pressure. The following discussion is devoted to the interpretation of our experimental results from the biological perspective.

Lower wax layer

The observations of other authors and our own previous experimental data showed the ability of platelets from the upper wax layer to contaminate the feet of insects during their contact with the waxy zone11,13,28. Removal of the upper wax layer platelets from the pitcher surface should partially expose the lower layer, which can then contribute to the trapping mechanism. Moreover, since the upper wax layer is unable to regenerate27, the lower wax layer might be a kind of supplementary system supporting the trapping function of thus damaged areas of the waxy zone. Another reason for investigating the adhesive properties of the lower wax layer was to evaluate the biological significance of the upper wax layer (i.e. if the upper wax layer is missing, would the pitcher still be able to trap insects).

Pull-off forces measured on the lower wax layer were non-significantly higher than those on the smooth surface. This result might be explained by the general presumption that adhesion forces are proportional to the area of real contact between surfaces36, as has already been supported by various authors37,38,39. Similar force values obtained on the lower wax layer and smooth surface suggest a similar real contact area between probes and the two surfaces. On the lower wax layer in contact with the probe, a real contact area comparable to that on the smooth surface is probably achieved by a partial inflow of the probe material, having a low E-modulus, into the cavities between wax crystals (Fig. 5D). A slight trend for higher forces on the lower wax layer might be also due to pinning of probes by stiff, pointed and irregular crystal margins and/or some additional friction and interlocking effects. Also, a possible chemical effect of the fluid polyvinylsiloxane used to expose the lower wax layer cannot be entirely excluded.

Figure 5
figure 5

Scheme explaining hypothetical surface interactions between different samples and soft adhesive probes.

(A) On a smooth replica, full contact occurs. (B) On a micro-rough sample, the real contact area is reduced due to the formation of partial contacts. (C) A complex interaction of a probe with the intact waxy zone containing both upper and lower wax layers: the surface of the probe is contaminated with upper wax layer crystals, serving as a separation layer. (D) The probe follows the surface's profile of the lower wax layer resulting in a real contact area comparable to that on the smooth surface.

Full size image

High adhesion forces achieved on the lower layer are not in line with previous data13, which described an anti-attachment effect not only for the upper, but also for the lower wax layer. These experiments, however, were performed with tethered ladybird beetles bearing hairy attachment pads. A recent atomic force microscopy study of ladybird beetle setae showed the lowest E-modulus of 1 MPa at the setal tip40. Thus, the material in these hairy adhesive pads is much stiffer than that in smooth pads. For this reason, the inflow effect of pad material into the cavities between wax crystals may not occur for hairy systems, but might happen in the case of smooth systems. Unfortunately, no attachment experiments on the lower wax layer have been done with insects equipped with smooth attachment devices. We hypothesize that some insects with smooth adhesive pads would be able to walk on the lower wax layer of N. alata. To test this hypothesis, further experiments have to be performed.

Apparently, the standard roughness parameters (Ra, r.m.s., Rz) were not sufficient to describe peculiarities of the surface topography in detail. While the measured roughness parameters of the lower wax layer and micro-rough replica were very similar, pull-off forces differed significantly. This effect might be explained by the different character of recorded roughness, displayed in white light interferometer (WLI) oblique 3D plots (see Fig. 2B,F). The soft probe material might adapt well to both wavelengths of the lower wax layer topography, such as the short wavelength with low amplitudes caused by wax crystals and the longer wavelength with high amplitudes caused by epidermis cells (Fig. 5D). The wavelength of the micro-rough replica has a middle range amplitude compared to that of the lower wax layer. At this range, the probe might be not able to follow the topography of the surface. As a result, a number of partial contacts would occur and this would lead to a lower real contact area (Fig. 5B). The significantly lower pull-off force obtained on the micro-rough surface compared to the lower layer sample might indicate that insects bearing smooth attachment pads are less effective on this particular domain dimension of the surface structure.

Upper wax layer

The strong reduction of pull-off force on the upper wax layer compared to all other tested surfaces might be explained primarily by the breaking off of wax crystals from the upper wax layer that acts as a separation layer and also contaminate the sticky half-spheres (Fig. 5C). The contaminating effect of N. alata wax crystals has also been previously reported by other authors1,11,13,28, but was recently rejected by Scholz et al.17. We discovered here that the contaminating material on the probes and the remaining wax structures found at the contact area on the plant surface after adhesion experiments clearly showed certain similarities to the wax crystals from the upper and lower layers, respectively. Furthermore, these detached platelets had rather long stalks. This supports the previous conclusion that the connection between the superimposed upper wax layer and the underlying surface structure might be due to thin breakable stalks13,30. Analysis of the fracture behavior of epicuticular wax crystals, performed using a mathematical approach, was recently undertaken by Borodich et al.41. The authors showed the possibility of wax crystals breaking under the weight of an insect. We have observed a rather small amount of fractured crystal parts on the contaminated probes. This might be an indication that the stalks act as predeterminated breaking points.

We suggest that the breaking off of wax crystals from the upper wax layer is important primarily for decreasing attachment forces during the very first contact with the wax coverage (Fig. 5C). In this case, detached crystals serve as a separation layer between counterparts (the pitcher surface and insect pad), thus causing the adhesion reduction and slipperiness of the surface. Such an effect of a separation layer is well known in surface physics42,43.

We also hypothesized that the contamination might also impede insect pad adhesion during further contacts. This means that after detached platelets from the upper layer covered a rather large area on the pad and formed a micro-rough surface there, contact would occur between the micro-rough pad surface and stiff micro-rough lower wax layer of the plant. As the adhesion force directly depends on the real contact area, it would be low in this case. However, our results obtained in the second type of experiment showed that the amount of contamination after the first contact of the PDMS probe with the upper layer was insufficient to cause a significant reduction of the forces in subsequent tests. Further experiments are needed to explore this hypothesis.

Thus, a decrease in insect adhesion on the N. alata wax coverage might not be solely explained by roughness exposed to the insect pads, as has been suggested by Scholz et al.17. It is possibly caused primarily by the breaking off of wax crystals from the upper wax layer that acts as a separation layer between the pad and the pitcher surface.

Methods

Preparation of test surfaces

The carnivorous plant species Nepenthes alata Blanco (Nepenthaceae), endemic to the Philippines, was used in our study44. Pitchers (Fig. 1A) were obtained from plants grown in the green house at the Botanical Garden of the Kiel University (Kiel, Germany). Fresh pitchers were harvested from four plants having the same genetic origin. The wax crystal structure in N. alata, which has been previously examined and documented extensively, has been shown to be very similar for every plant of this species studied so far13,15,19,20. This justifies the use of genetically similar plants in our study.

In order to minimize the effect of lunate cells, the lower part of the waxy zone located just above the transitional layer and bearing almost no lunate cells was used in all experiments. The untreated plant surface was used to investigate the upper wax layer (UL). To expose the lower wax layer (LL), the UL was mechanically removed by treating the waxy zone with a two-component polyvinylsiloxane (Coltène Whaledent Dentalvertriebs GmbH., Konstanz, Germany). Fluid polyvinylsiloxane was applied to the pitcher surface and then peeled off after 5 min of polymerization.

As reference surfaces, two epoxy resin45 samples, smooth and micro-rough, were prepared. They were obtained from a smooth clean glass slide and micro-rough polishing paper with a defined asperity size of 3 μm (Serva Electrophoresis GmbH, Heidelberg, Germany) by applying a two-step molding method according to Gorb46. Data obtained on the smooth sample were used for further normalization of adhesion force values obtained on rough surfaces (see subsection Adhesion tests for details). Since the micro-rough epoxy resin sample exhibited roughness parameters similar to those of the intact wax coverage in N. alata (unpublished data), it was used as a solid micro-rough reference surface.

Characterisation of test surfaces and probes

Microscopy

The surface of plant and polymer samples as well as probes used in adhesion tests (see the subsection Adhesion tests) were examined before and after the tests in a scanning electron microscope (Hitachi S-4800, Hitachi High-Technologies Corporation, Tokyo, Japan). Pitcher surfaces and probes were investigated in a cryo-mode using Gatan ALTO 2500 cryo-preparation system (Gatan Inc., Abingdon, UK). A conventional SEM method was applied for examination of polymer replicas.

Small pieces of the plant material (0.5 cm × 0.5 cm) were cut out of the intact or treated waxy zone in N. alata pitchers (see Preparation of test surfaces) using a razor blade and then glued onto holders by means of polyvinyl alcohol Tissue-Tek O.C.T.TM Compound (Sakura Finetek Europe B.V., Zoeterwoude, the Netherlands). In order to obtain fractures of the wax coverage, plant samples were placed vertically on specific holders and mechanically fixed. Afterwards, samples were frozen in liquid nitrogen and quickly transferred into a cryo-preparation chamber (133°K). For fractures, an integrated metal knife was applied. PDMS probes were attached to holders using conductive carbon double-sided adhesive tape and transferred, without previous treatment with liquid nitrogen, directly into the cryo-preparation chamber. All plant samples and polymer probes were sputter coated under frozen conditions with gold-palladium (1:9, 10 nm thick) and studied in SEM at 2 kV accelerating voltage. Small pieces of replicas were mounted on holders in the same way as the probes, sputter coated with gold–palladium (1:9, 10 nm thick) and examined in SEM at 7 kV accelerating voltage.

Roughness

The roughness of the test substrates was characterized by applying a white light interferometer (Zygo NewView 6000, Zygo Corporation, Middlefield, Connecticut, USA) at a 50× magnification (scanned area of 140 μm × 105 μm). Ra, r.m.s. and Rz were extracted from topographical height maps obtained. For this purpose, six measurements at different places on each surface were performed and the mean value and standard deviation were calculated for each roughness parameter.

Adhesion tests

Adhesion experiments were carried out with the micro-tribometer Basalt 01 (Tetra GmbH, Ilmenau, Germany) consisting of (1) a platform, (2) a metal spring with an attached probe and (3) a fiber-optic sensor (Fig. 6) (described in detail by Scherge and Gorb47). Tacky and deformable half-spherical indenters (probes) made out of PDMS according to Purtov et al.31 were brought into contact at normal force ranging from 500 μN to 1000 μN with the test surface attached to the platform and then retracted at a constant speed of 37.6 ± 1 μm s−1. The contact time of probes with surface, i.e. the time between the first contact and detachment, ranged from 4 s to 10 s. Using force - distance curves obtained, pull-off forces were measured (see Purtov et al.31). All experiments were carried out at room temperature (22–24°C) and a relative ambient humidity of 35–42%.

Figure 6
figure 6

Experimental set-up for adhesion tests.

A test sample (SA) was mounted on a platform (PL). A soft indenter (PH), the sticky half-spherical probe made out of polydimethylsiloxane, was attached to a metal spring (MS). Driven by a motor, the sample and indenter were brought into contact and then retracted. Forces were detected using a fiber optic sensor (FOS) and a mirror (MR).

Full size image

We carried out two types of experiments. First, three single measurements were performed with each probe on (1) the smooth replica, (2) the micro-rough replica and (3) one of the wax layer samples. The order of measurements was not randomized. Plant surfaces were prepared just before tests and immediately used for adhesion measurements. In total, 120 tests with 40 PDMS probes were carried out: respectively, 40 on the smooth and micro-rough replicas, 18 on the UL and 22 on the LL. We used normalized data for statistical comparison of the forces obtained on different test substrates. Using this approach, we avoided the influences of possible differences in tackiness of individual probes. In order to gain normalization, the force value of rough samples (micro-rough replica, UL and LL) was divided by that of the smooth reference surface measured with the same probe. In the case of the smooth reference, we normalized individual measurements on glass by the mean value of all forces measured on glass.

In order to test the effect of the contamination by wax crystals from the UL on adhesion forces, we carried out the second type of experiment. With the same probe, pull-off measurements were performed first on the smooth sample, then on the UL and again on the smooth sample. We used 15 probes and carried out 45 tests in total. Force values were normalized to those obtained in the first test on the smooth sample.

Data obtained in the first type of experiment were statistically analyzed with Friedman repeated measures ANOVA on ranks. Multiple pairwise comparisons were performed using Wilcoxon signed rank test with Bonferroni correction. Force values measured in two tests on smooth sample in the second type of experiment were compared using paired t-test.

To prove the biological relevance of our experimental approach, we compared elastic properties of PDMS probes with material properties of insect adhesive pads and approximate pressure applied by probes to the plant surfaces with contact pressure of insects adhering to these surfaces. The elasticity modulus of the PDMS probes was calculated using the force - distance data obtained on smooth replica according to the 2-point method proposed by Ebenstein and Wahl48, which is based on the Johnson-Kendall-Roberts theory49. In order to obtain the approximate pressure applied by the PDMS probe on the plant surfaces, we calculated the pressure (p) for a maximum and minimum load force according to Popov50:

where F is the applied normal force (Fmin = 0.5 mN and Fmax = 1 mN), R is the radius of the polymer probe (1.5 mm) and E* is the reduced E-modulus describing the elastic behavior of the contact system. The reduced E-modulus is related to Young's modulus (Ei) of solids as follows:

For the E-modulus of polymer probe (E1), the value calculated as described above was used. For the plant surface, literature data on the digestive zone of the N. alata pitcher (E2 = 637.19 kPa)32 was entered into equation (2). For both the probe and plant surface, a Poisson ratio (ν) of 0.5 was assumed51,52.