Plant Seed Mucilage as a Glue: Adhesive Properties of Hydrated and Dried-in-Contact Seed Mucilage of Five Plant Species

Seed and fruit mucilage is composed of three types of polysaccharides—pectins, cellulose, and hemicelluloses—and demonstrates adhesive properties after hydration. One of the important functions of the mucilage is to enable seeds to attach to diverse natural surfaces. Due to its adhesive properties, which increase during dehydration, the diaspore can be anchored to the substrate (soil) or attached to an animal’s body and dispersed over varied distances. After complete desiccation, the mucilage envelope forms a thin transparent layer around the diaspore creating a strong bond to the substrate. In the present study, we examined the mucilaginous seeds of six different plant taxa (from genera Linum, Lepidium, Ocimum, Salvia and Plantago) and addressed two main questions: (1) How strong is the adhesive bond of the dried mucilage envelope? and (2) What are the differences in adhesion between different mucilage types? Generally, the dried mucilage envelope revealed strong adhesive properties. Some differences between mucilage types were observed, particularly in relation to adhesive force (Fad) whose maximal values varied from 0.58 to 6.22 N. The highest adhesion force was revealed in the cellulose mucilage of Ocimum basilicum. However, mucilage lacking cellulose fibrils, such as that of Plantago ovata, also demonstrated high values of adhesion force with a maximum close to 5.74 N. The adhesion strength, calculated as force per unit contact area (Fad/A0), was comparable between studied taxa. Obtained results demonstrated (1) that the strength of mucilage adhesive bonds strongly surpasses the requirements necessary for epizoochory and (2) that seed mucilage has a high potential as a nontoxic, natural substance that can be used in water-based glues.


Introduction
Mucilage of plant seeds and fruits (diaspores) is a polysaccharide complex consisting mainly of pectins and hemicelluloses, but also containing cellulose fibrils, which interact with each other by various types of bonds, forming a spatial network [1][2][3][4]. Diaspore mucilage can be classified based on its chemical composition into three main types: pectic, cellulose and hemicellulose. Pectic mucilage is typical of plant representatives from the genus Linum. Cellulose mucilage contains cellulose fibrils acting as a micro-/nanoscale "skeleton", which anchor the mucilage to the seed surface and is characteristic of representatives from families such as, e.g., Asteraceae, Brassicaceae or Lamiaceae [5][6][7][8][9]. Hemicelluloserich mucilage was recently defined in some species of the genus Plantago (P. ovata and P. psyllium), in which the mucilage envelope can be dominated by hemicelluloses such as, e.g., arabinoxylan [10].
Mucilage production by diaspores is known as myxodiaspory and is particularly common and well developed in plants inhabiting steppes, semi-deserts or disturbed habitats,

Morphology of the Mucilage Envelope
To characterize the mucilage envelope and to detect its basic components, we performed following staining reactions. Staining with 0.1% aqueous solution of ruthenium red (Sigma-Aldrich) aided in detection of pectins [34]. The presence of crystalline cellulose was tested in a polarized light microscope (Zeiss LSM 700) (Zeiss, Jena, Germany) [35]. Images were taken using a light microscope equipped with digital camera. To detect the presence of starch in the mucilage of Ocimum basilicum, the solution of potassium iodide with iodine (IKI) in water was used: it stained starch grains violet to dark violet [34].

The Pull-Off Force Measurements of Dehydrated Mucilage Envelope
The material tester Zwick Roell z 0.5 (Zwick GmbH & Co. KG, Ulm, Germany, Type: BT1-FR0.5TN.D14) was used in the experiment. It allowed us to conduct experiments in established, constant conditions. The results of the measurements were automatically recorded and processed by Zwick Roell software.
For every taxon, 15 individual seeds were hydrated for 30 min, in order to enable formation of the mucilage envelope. The seeds were then put on a glass slide and left to dry out for at least 24 h. After that time, the seeds had strongly adhered to the glass by the dried-out mucilage envelope. The contact area between the mucilage envelope and the glass substrate was measured using a stereoscopic microscope, Leica M205 A (Leica Mikrosysteme, Wetzlar, Germany). The contact area was used for the calculation of adhesion strength. For the adhesion measurements, two surfaces were used: a glass slide with an attached (by mucilage) seed and a SEM stub with a round cover glass bonded with commercial glue (Cyanoacrylate adhesives: 5925 ergo(R) Elastomer-Kisling GmbH, Bad Mergentheim, Germany). The glass slide with the attached seed was fastened to the lower pincer grip, and the SEM stub was fastened to the upper pincer grip (Figure 1) in the Zwick material tester. The upper (free) seed surface was attached to the SEM stub with a commercial glue (BONDIC GmbH, Köln, Germany) ( Figure 1). The upper grip was moved up with the speed of 2 mm/s, until the seed was detached from the glass slide. The force-distance curves were continuously recorded during the platform motion.

Morphology of the Mucilage Envelope
To characterize the mucilage envelope and to detect its basic components, we performed following staining reactions. Staining with 0.1% aqueous solution of ruthenium red (Sigma-Aldrich) aided in detection of pectins [34]. The presence of crystalline cellulose was tested in a polarized light microscope (Zeiss LSM 700) (Zeiss, Jena, Germany) [35]. Images were taken using a light microscope equipped with digital camera. To detect the presence of starch in the mucilage of Ocimum basilicum, the solution of potassium iodide with iodine (IKI) in water was used: it stained starch grains violet to dark violet [34].

The Pull-Off Force Measurements of Dehydrated Mucilage Envelope
The material tester Zwick Roell z 0.5 (Zwick GmbH & Co. KG, Ulm, Germany, Type: BT1-FR0.5TN.D14) was used in the experiment. It allowed us to conduct experiments in established, constant conditions. The results of the measurements were automatically recorded and processed by Zwick Roell software.
For every taxon, 15 individual seeds were hydrated for 30 min, in order to enable formation of the mucilage envelope. The seeds were then put on a glass slide and left to dry out for at least 24 h. After that time, the seeds had strongly adhered to the glass by the dried-out mucilage envelope. The contact area between the mucilage envelope and the glass substrate was measured using a stereoscopic microscope, Leica M205 A (Leica Mikrosysteme, Wetzlar, Germany). The contact area was used for the calculation of adhesion strength. For the adhesion measurements, two surfaces were used: a glass slide with an attached (by mucilage) seed and a SEM stub with a round cover glass bonded with commercial glue (Cyanoacrylate adhesives: 5925 ergo(R) Elastomer-Kisling GmbH, Bad Mergentheim, Germany). The glass slide with the attached seed was fastened to the lower pincer grip, and the SEM stub was fastened to the upper pincer grip (Figure 1) in the Zwick material tester. The upper (free) seed surface was attached to the SEM stub with a commercial glue (BONDIC GmbH, Köln, Germany) ( Figure 1). The upper grip was moved up with the speed of 2 mm/s, until the seed was detached from the glass slide. The forcedistance curves were continuously recorded during the platform motion. The sample with the seed was placed between two grips: the lower one was motionless and the upper one moved upwards. The lower seed surface was attached to the aluminum SEM stub by the dried mucilage envelope. The upper (mucilage free) seed surface was glued to the SEM stub by the commercial glue. The upper grip moved up and stopped when the seed was detached from the glass slide. The sample with the seed was placed between two grips: the lower one was motionless and the upper one moved upwards. The lower seed surface was attached to the aluminum SEM stub by the dried mucilage envelope. The upper (mucilage free) seed surface was glued to the SEM stub by the commercial glue. The upper grip moved up and stopped when the seed was detached from the glass slide.
The average time needed to perform biophysical experiment and measurement with Zwick (including the following steps: sample fixation in Zwick, setting up/checking the probe parameters, measurement and saving the data) ranged between 20-25 min. The measurement of the pull-off force for the individual sample took from few to several seconds (9-max 120 s).
As a control, 15 dry seeds of each studied taxon were fixed to the glass slide with a drop of UHU glue (all-purpose adhesive, liquid-UHU GmbH & Co. KG, Bühl, Baden, Germany). Prior to adhesion force measurements, images of the contact area between the seed and the glass substrate were taken using a stereoscopic microscope as mentioned above. After the measurements, a second image was taken, to visualize the area where the glue was peeled away from the glass. The area was estimated with AxioVison SE64 Rel. 4.8 software (Zeiss, Jena, Germany).
For the adhesion force measurements of Plantago ovata husk mucilage, 15 samples were also used. A round cover glass (12.5 mm in diameter) was adhered to a SEM stub with the commercial glue (Cyanoacrylate adhesives: 5925 ergo(R) Elastomer-Kisling GmbH, Bad Mergentheim, Germany). The seed husks were hydrated for 30 min to create the mucilage. Then, a 60 µL drop of mucilage material was put on the stub with the attached round cover slip and pressed with a second stub with an attached cover slip. The samples were air-dried for 24 h. As a control sample, two SEM stubs with cover slips were connected using a 60 µL droplet of commercial UHU glue (all-purpose adhesive, liquid-UHU GmbH & Co. KG, Bühl, Baden, Germany; contains-ethanol, methyl acetate) and air-dried for 24 h. For the measurement, a sample was put between two grippers of the Zwick tester, and the measurement was carried out as described above. The time needed for the measurements of individual samples using Zwick was ranging between 20-25 min.
Scanning electron microscopy (SEM) images were taken of all seed types of studied plant taxa, P. ovata husk and UHU after force measurements, in order to analyze the mucilage and UHU glue structure at the fracture zone. The samples were coated with gold palladium (film thickness 10 nm) using a Leica EM SCD 500 High Vacuum Sputter Coater (Leica Microsystems GmbH, Wetzlar, Germany), and visualized in SEM Hitachi S-4800, Hitachi High-Tech. Corp., Tokyo, Japan).

Young's Modulus Determination
Young's moduli of both UHU glue and husk mucilage from the seeds of P. ovata was determined. P. ovata was chosen, as we knew we were able to collect a sufficient amount of mucilage from this species for the experiment. UHU glue was used as a control. Defect-free stripes were prepared from the two materials and pulled with 2 mm/min velocity in the Zwick tester until 5% stretched. For calculation of the Young's moduli, the values of maximum stiffness and cross-sectional area were used. The area was measured using a VR-3000 3D Measurement Microscope.

Statistics
The statistical analysis was done using Sigma-Plot (Systat Software Inc, San Jose, CA, USA). The results were analyzed using the non-parametrical Kruskal-Wallis one-way ANOVA by ranks test (p < 0.05) and multiple comparison procedures (Dunn's Method, p < 0.05), because the normality test failed.

Mucilage Morphology and Its Basic Components
In studied taxa, the mucilage envelope forms a gel-like capsule around the seed within a few minutes after hydration. Based on our previous studies [18,19] and staining reactions applied in the present study, we were able to describe the basic components and the main types of the mucilage (  The presence of cellulose fibrils with crystalline regions was verified and visualized using polarized light microscope ( Figure 2J-M). The crystalline cellulose was present in the mucilage of L. sativum ( Figure 2J), O. basilicum ( Figure 2K) and S. hispanica ( Figure 2L) and in a very narrow region close to the seed surface in P. lanceolate ( Figure 2M). The mucilage of P. ovata represents a hemicellulose-rich type without cellulose fibrils (see also [36]). The presence of starch grains in the mucilage of O. basilicum was detected ( Figure 2E).

Measurements of Pull-Off Forces of the Dehydrated Mucilage Envelope
The studied plants can be divided into two groups with statistically significant differences ( Figure 3A,A'). The first group contains L. usitatissimum, P. lanceolata and S. hispanica, and the remaining species, L. sativum, O. basilicum and P. ovata, were allocated to the second group. The mean F ad (adhesion force) values for group one ranged from 1.01 to 1.59 N. In the second group, the mean F ad values were more than three times higher (e.g., O. basilicum with F ad = 4.44 N) ( Table 1).

Measurements of Pull-off Forces of the Dehydrated Mucilage Envelope
The studied plants can be divided into two groups with statistically significant differences ( Figure 3A,A'). The first group contains L. usitatissimum, P. lanceolata and S. hispanica, and the remaining species, L. sativum, O. basilicum and P. ovata, were allocated to the second group. The mean Fad (adhesion force) values for group one ranged from 1.01 to 1.59 N. In the second group, the mean Fad values were more than three times higher (e.g., O. basilicum with Fad = 4.44 N) ( Table 1).   The same situation was observed for the mean A 0 (contact area) values ( Figure 3B,B'). The mean contact area was relatively small in the first group ranging from 0.98 to 1.57 mm 2 . In the second group, the values were three to four times higher (Table 1).
In the case of adhesive strength (F ad /A 0 ), the mean results were similar in representatives of both groups ( Figure 3C,C', Table 1). The statistically significant differences were observed only for L. sativum, which differed from the other three (P. lanceolata, O. basilicum, S. hispanica).
It was observed that during individual measurements, the pulling force increased gradually until it reached the maximum-the point when the seed with some mucilage residue was detached from the glass substrate. The pull-off force (F ad ) for dried mucilage in all seeds varied from 0.58 to 6.22 N. The highest maximal value of F ad was measured in the cellulose mucilage of O. basilicum-6.22 N-which is considerably higher than the maximal F ad value of 2.8 N measured for the O. basilicum seeds embedded in the UHU glue. The lowest maximal value of F ad was that of P. lanceolata-2.03 N.
The contact area varied from 0.11 mm 2 for S. hispanica to 6.73 mm 2 for P. ovata. The mean contact area for seeds glued with UHU varied from 0.86 mm 2 for S. hispanica to 5.49 mm 2 for L. usitatissimum. The highest maximal value of adhesive strength, σ c = F ad /A 0 , was observed for S. hispanica (8.15 N/mm 2 ), and the smallest for L. sativum (1.02 N/mm 2 ) ( Table 1).

Adhesion Strength of P. ovata Husk Mucilage
Because the contact area of the tested samples was constant, 122.65 mm 2 for this experiment, the adhesion strength values were simple to obtain. The adhesion strength of dried-out P. ovata husk mucilage ranged between 0.04 and 0.31 N/mm 2 . The UHU glue control samples had adhesion strengths that varied from 0.14 to 0.19 N/mm 2 . However, the maximal adhesion force of UHU (30.4 N) was lower than the P. ovata (37.4 N) husk mucilage (Figure 4). The same situation was observed for the mean A0 (contact area) values ( Figure 3B,B'). The mean contact area was relatively small in the first group ranging from 0.98 to 1.57 mm 2 . In the second group, the values were three to four times higher (Table 1).
In the case of adhesive strength (Fad/A0), the mean results were similar in representatives of both groups ( Figure 3C,C', Table 1). The statistically significant differences were observed only for L. sativum, which differed from the other three (P. lanceolata, O. basilicum, S. hispanica).
It was observed that during individual measurements, the pulling force increased gradually until it reached the maximum-the point when the seed with some mucilage residue was detached from the glass substrate. The pull-off force (Fad) for dried mucilage in all seeds varied from 0.58 to 6.22 N. The highest maximal value of Fad was measured in the cellulose mucilage of O. basilicum-6.22 N-which is considerably higher than the maximal Fad value of 2.8 N measured for the O. basilicum seeds embedded in the UHU glue. The lowest maximal value of Fad was that of P. lanceolata-2.03 N.
The contact area varied from 0.11 mm 2 for S. hispanica to 6.73 mm 2 for P. ovata. The mean contact area for seeds glued with UHU varied from 0.86 mm 2 for S. hispanica to 5.49 mm 2 for L. usitatissimum. The highest maximal value of adhesive strength, σc = Fad/A0, was observed for S. hispanica (8.15 N/mm 2 ), and the smallest for L. sativum (1.02 N/mm 2 ) ( Table  1).

Adhesion Strength of P. ovata Husk Mucilage
Because the contact area of the tested samples was constant, 122.65 mm 2 for this experiment, the adhesion strength values were simple to obtain. The adhesion strength of dried-out P. ovata husk mucilage ranged between 0.04 and 0.31 N/mm 2 . The UHU glue control samples had adhesion strengths that varied from 0.14 to 0.19 N/mm 2 . However, the maximal adhesion force of UHU (30.4 N) was lower than the P. ovata (37.4 N) husk mucilage (Figure 4).

Mucilage Structure in Fractures after the Pull-off Measurements
After hydration, mucilage envelope formation and drying out, the mucilage envelope formed a thin, transparent layer around the seed, which firmly attached the seed to the glass surface. The seed was then detached from the glass, together with the remaining mucilage envelope fragments ( Figure 5). In L. usitatissimum and P. lanceolata, only small pieces of mucilage were found attached to the seed surface ( Figure 5A,E). Interestingly,

Mucilage Structure in Fractures after the Pull-Off Measurements
After hydration, mucilage envelope formation and drying out, the mucilage envelope formed a thin, transparent layer around the seed, which firmly attached the seed to the glass surface. The seed was then detached from the glass, together with the remaining mucilage envelope fragments ( Figure 5). In L. usitatissimum and P. lanceolata, only small pieces of mucilage were found attached to the seed surface ( Figure 5A,E). Interestingly, these two taxa also produced smaller amounts of mucilage in comparison to the other seeds studied, and in the other taxa, larger fragments remained attached to the seed surface ( Figure 5B-D,F). The main part of the mucilage envelope remained attached to the glass surface after the pull-off experiment in all cases. these two taxa also produced smaller amounts of mucilage in comparison to the other seeds studied, and in the other taxa, larger fragments remained attached to the seed surface ( Figure 5B-D,F). The main part of the mucilage envelope remained attached to the glass surface after the pull-off experiment in all cases. The segments of the mucilage envelope differed in structure ( Figure 6A-G) in comparison to the commercial glue sample (UHU), which resembled the fracture of an elastic polymer ( Figure 6H). Mucilage fractures of L. usitatissimum, L. sativum and Plantago seeds demonstrated a homogeneous structure ( Figure 6A-B,E-G), as opposed to the pieces of cellulose mucilage of O. basilicum, which showed heterogeneous structures with many

Elasticity Modulus and Strain Energy Density of P. ovata Dried Husk Mucilage
The Young's modulus of the dried-out husk mucilage of P. ovata was 423 ± 95 MPa, which is comparable to many stiff polymeric materials, e.g., Teflon [37], and one order of magnitude higher than the Young's modulus of polymerized UHU glue, 45.4 ± 21.5 MPa. For the known Young's moduli, E, and adhesion strength, σ c , the strain energy density, W, can be calculated as [38]: where h is the elastic film layer thickness, which connects the two rigid stubs. In our pull-off experiments, with a layer of dried-out husk mucilage of P. ovata or a layer of UHU glue between two glass surfaces, the failure of separation always took place at the interface between the elastic film and the glass. The strain energy density for the adhesive failure of the husk mucilage on the glass was 2.74 ± 0.64 mJ/m 2 , whereas the strain energy density for the adhesive failure of UHU glue on the glass was 36.5 ± 6.1 mJ/m 2 . A high correlation coefficient between adhesion strength and strain energy density for individual measurements on husk mucilage, R = 0.89 (N = 16), allows us to assign the variation of pull-off force to strain energy density in individual samples.

Discussion
In the present work, we demonstrated the strong adhesive properties of the dried seed mucilaginous envelope. The maximal pull-off force (F ad ) measured for different types of mucilage varied from 2.03 to 6.22 N and exceeded the adhesive properties of the commercial glue used as a control sample. The results of adhesive strength (F ad /A 0 ) measurements were similar between different mucilage types, leading us to conclude that, in general, the mucilage envelope possesses strong adhesive properties.
The adhesive properties of the seed/fruit mucilage envelope are known to be beneficial for plants by aiding in diaspore transport [13]. The ability of seeds/fruits to travel by means of adhesion can be crucial in the dispersion success of many plants. It allows the plant to be introduced into new habitats, which leads to an increase in its area of distribution. The diaspore adhesion to the ground can also increase its survival chance, since after hydration, mucilage formation and anchorage to the soil, the diaspore can begin germinating [7,39].
Furthermore, due to its adhesive properties, the mucilage has also been utilized in pharmaceuticals as a thickening, binding, disintegration, suspension, emulsifying, stabilizing and gelling agent [33].

Polysaccharide Structure Can Influence Adhesive Properties of the Mucilage Envelope
The adhesive properties of polysaccharides are generally accepted; they depend on the polysaccharide type, as well as on the water content [18,19,40]. For adhesion, different types of bonds across the interface are also important, e.g., ionic, covalent or hydrogen bonds. Together, they contribute to "wet" adhesion, meaning they are activated by moisture, and can, therefore, adhere non-specifically to many surfaces [41]. After activation and subsequent dehydration, they show a strong adhesion to dry surfaces. Polysaccharides (pectins, hemicelluloses) in the mucilage envelope have characteristic branched structures containing hydroxyl groups that are able to form hydrogen bonds [4,9,42,43]. This can enable them to interact with the surface of the substrate and promote a strong bond. For example, the seeds of Plantago ovata produce copious mucilage with the presence of both a highly branched structure of arabinoxylan and side-chains rich in sites with potential hydrogen-bond formation that can influence the rheological properties of the mucilage [10]. After hydration, arabinoxylans solubilize themselves and form a viscoelastic liquid outer layer that undergoes gelation in water. This layer in turn provides a reliable protective viscoelastic shell around the seed [10] but is soft/fluid enough to establish strong contact with the substrate when it dries. In contrast, the mucilage of P. lanceolata contains the highest proportion of unsubstituted xylan backbone [36] and, therefore, has lower adhesion values when compared with P. ovata.
The specific composition of L. usitatissimum mucilage can also be responsible for its strong adhesive properties. Important components of flax mucilage are branched rhamnogalacturonan (RG I) and arabinoxylan; the mixture of these two polysaccharides is responsible for the viscous properties of the mucilage in this plant species [43].

Cellulose Fibrils as an Additional Strengthening Component of the Mucilage Structure
The presence of cellulose fibrils can also be a factor in the maintenance of the seed envelope under mechanical loads. Cellulose fibrils allow for a more uniform distribution of mucilage around the seed and a stronger attachment to the seed surface -compare with data in [4]. The cellulose chains, composed of glucose residues, form elementary unbranched microfibrils [44][45][46][47][48], which can contain cellulose molecules that are highly ordered, and form crystal-like structures, in contrast to the unstructured, amorphous regions commonly seen. Hydroxyl groups in the crystalline phase form a large number of hydrogen bonds, which are responsible for the compact crystal structure of this region of the fibril. The free hydroxyls in the cellulose have a strong affinity to water, which can only be absorbed in the amorphous region of the fibril [47,49]. We observed in this study that cellulose fibrils can enhance the mechanical properties of mucilage. The maximal F ad measured during the detachment of the affixed seed from the glass surface was very high for the cellulose type of mucilage (Table 1). During the studies of the frictional properties of the critical-point dried cellulosic mucilage, we also detected adhesive forces caused by close contact of cellulose fibrils with the probe surface. We supposed that observed adhesive properties of mucilage can be based on van der Waals interactions between the cellulose fibers and the surface [30].
Cellulose mucilage of O. basilicum and S. hispanica demonstrated a specific lamellar structure. Such organization was observed in cellulose nanocomposite films obtained from microfibrillated cellulose (MFC) and amylopectin [50]. High tensile strength is combined with both high elasticity modulus and works of fracture in this nanocomposite. An elasticity modulus of 6.2 GPa, a tensile strength of 160 MPa and a strain-to-failure of 8.1% were observed at 70 wt. % of MFC reinforcement. The starch matrix used in this nanocomposite showed high compatibility with MFC, which aided in achieving the uniquely high MFC content. The amylopectin matrix of this nanocomposite is highly plasticized, which enables it to mimic the viscous properties in the cellulose "matrix" [50]. Only O. basilicum possesses starch grains in the mucilage. We can, therefore, conclude that the interaction between starch grains and cellulose (or other polysaccharide) fibrils can increase the adhesive properties measured in this mucilage.

Influence of the Contact Area and Mucilage Thickness on the Adhesive Properties
The differences between contact area measured for different seeds can result from the morphological features of the seeds and mucilage, such as the mass, size, and shape of the seed or the size (amount) of the mucilage envelope. The largest and most flattened seeds were characteristic of L. usitatissimum, but the contact area on the smooth glass surface was the smallest one in this species. High contact area was characteristic of O. basilicum and P. ovata, because these two taxa produce copious mucilage envelopes. In the natural habitat, all morphological seed features, including those of the mucilage, can play important roles in the seed attachment to the porous ground. It was observed that seed losses decreased with the increase in seed mass, and that the flat seeds had a higher resistance when removed from the soil surface. The presence of mucilage, which reduces susceptibility of seed removal by soil erosion, is an important mechanism for preventing further seed dispersal and maintaining the plant in its habitat [16].
Adhesion strength depends on the internal structure of the adhesive near the interface. The inclusion of other material in the glue matrix drastically influences adhesive strain energy density. While the adhesive strain energy density of the UHU glue was 36.5 mJ/m 2 , the adhesive strain energy density of husk mucilage of P. ovata seeds was just 2.74 mJ/m 2 . Such elastic properties of P. ovata husk mucilage can provide sufficient strength of the bond, preventing easy cohesive failure. The husk remnants might be responsible for the low strain energy density value. It is also possible to estimate the strain energy density value in a pulloff experiment on individual seeds. Assuming the same value of Young's modulus for the mucilage in seeds as for the husk mucilage and taking the dry mucilage thickness of 3 µm (Figure 6G), the strain energy density value of a pure mucilage is 3 mJ/m 2 . This value is very close to that of the husk mucilage. We can estimate the strain energy density value in a pulloff experiment on seeds, assuming the same value of Young's modulus and mucilage layer thickness. The strain energy density value for an individual seed, 21.6 mJ/m 2 , is one order of magnitude higher than that for the husk mucilage and close to the value for the UHU glue. Obviously, such elastic properties of P. ovata mucilage can provide sufficient strength of the bond, simultaneously preventing seed damage. Since the mucilage Young's moduli for other species are unknown, we can compare the W·E values, which is an important parameter to characterize the adhesion efficiency. After normalization of the W·E value for P. ovata mucilage, it is possible to distinguish three groups with similar values. In the first group (L. sativum, P. lanceolata, husk mucilage and UHU glue), the W·E values are relatively small: 0.13-0.18. In the second group (L. usitatissimum and S. hispanica), the W·E values are higher: 0.57-0.64. Finally, O. basilicum has a W·E of 0.35, which is an intermediary between the first and second groups.

Adhesive Properties and Their Biological Significance
Diverse functional significances of adhesive or viscous properties of the diaspores are well recognized but studied only at the morphological level [13]. Nevertheless, the studies characterizing adhesive properties of the mucilage envelope with reference to its basic chemical composition, morphology and biological function are lacking.
From the biological point of view, strong adhesion can be extremely important, particularly in relation to long-distance diaspore dispersal. Mucilaginous seeds can adhere to an animal's fur or a bird's feathers. Fixed with the dried mucilage, they can then be transported over long distances [11,[13][14][15]. Such a way of dispersal has been reported for mucilaginous seeds of several Euphorbia species, which were carried by birds between the Hawaiian Islands [11]. Similarly, the diaspores of Lepidium sativum, were dispersed, most likely by epizoochory, as was observed in New Zealand. Seabirds were likely the main dispersal vector to which the mucilaginous diaspores of diverse Lepidium taxa adhered, allowing them to be transported along various distances [51]. The example of the mucilaginous seeds of Plantago media, which were introduced from Europe to North America by traveling on human shoes and/or animal feet [5], demonstrates that the strong adhesive properties of the mucilage were the driving force in spreading this plant over the entire continent. Contrastingly, antitelochory has the opposite effect to diaspore dispersal, but adhesion also plays an important role here. Strong adhesion of the mucilaginous seeds to the soil protects them from being moved to less favorable habitats and enables germination in the surroundings of the mother plant [7].
Comparing our previous results on adhesion measurements of the hydrated mucilage envelope of L. usitatissimum (91 mN) and P. lanceolata (32 mN) [18,19] to the results obtained in the present study, we can differentiate between hydrated and dried mucilage adhesive properties. In the first case, low frictional and adhesive properties can be important for hydrated mucilage, in order to form the first contact between the seed and the substrate surface (ground, animal). In the second phase, dried-out mucilage, when the seed is strongly cemented to the surface, can be responsible for the long-distance dispersal of diaspores.
In conclusion, the results obtained demonstrated strong adhesive properties of the dried seed mucilage. The differences in measured variables depended on the mucilage type and, particularly, on the presence of a complex cellulose skeleton and/or highly branched polysaccharides (pectins, hemicelluloses), which are able to form adhesion-supporting hydrogen bonds. These results provide important information for the fields of seed biology and plant ecology.
Since the adhesive strength of the mucilage envelope can reach significantly higher values in comparison with the standard type of commercial glues, it has significant potential for the development of novel environment-friendly, non-toxic adhesives with strong bonding properties.