Experimental Analysis of Adhesion Of Chrysolina Polita (Chrysomelidae: Coleoptera) On A Variety of Surfaces

The phenomenon of the ability of some insects and other animals to adhere to smooth surfaces has always fascinated naturalists. The fundamental issue to the early microscopists was how the ‘fly’ could hold on to glass against the force of gravity. In the mid-nineteenth century there was considerable discussion on this subject. Indeed, today well over a hundred years later, there is still controversy about the actual mode(s) of adhesion employed. The major problem in understanding the derivation of the adhesive force has been the dearth of accurate information on the detailed structure of the adhesive setae. Stork & Evans (1976) used the SEM to reveal the diversity of structure in coleopteran adhesive setae which had been hinted at by West (1862), Simmermacher (1884), Dahl (1884) and Jeannel (1941). Recently, Stork (1980) has studied the morphology of adhesive setae in the Coleoptera more fully, suggesting reasons for their diversity.

Apart from Plateau (1872), who measured the pulling force in a few caraboids, no one has made a serious attempt to estimate the adhesive forces involved in traction of adhesive setae on smooth surfaces. This paper first examines ‘pulling force’ attained on several surfaces in Chrysolina polita (L.) in relation to body weight and sex, and secondly, investigates the mode(s) of adhesion involved in the use of adhesive setae.

A George Washington Ltd 50 g force transducer Type Di was coupled to a George Washington Ltd ink-writing oscillograph MD/2. The transducer was clamped to a Sonnenblik screw stand so that the optimum sensitivity to pulling was in the vertical plane. A sheet of glass, previously cleaned with acetone, was clamped behind the force transducer. Sixty-two specimens of Chrysolina polita (L.) (Chrysomelidae; Coleoptera) were individually weighed and a small loop of fine electrical wire attached to each of their elytra. This was accomplished by anaesthetizing them with carbon dioxide and gluing on the wire loops with a small amount of a mixture comprising 7 parts beeswax to 4 parts pale amber resin (BDH Chemicals Ltd.). The beetles were then reweighed and left in a cold room (approx. 15 °C) with an adequate supply of their food source (water mint, Mentha aquatica L.). Before being attached to the force transducer the beetles were allowed to warm up to room temperature (approx. 20–25 °C) for about 30 min. The beetles were hooked on to the force transducer and 10 min recordings were taken on a suitable attenuation for each beetle walking on the glass substratum. Similar recordings were taken on perspex (also previously cleaned with acetone) and rough-woven linen cloth (the latter was stretched over a piece of glass). The force transducer was calibrated before and after each trial. A gap of several hours was left between recordings for an individual beetle on the different substrates. Recordings other than those made when the beetle was pulling in the direction of optimum sensitivity of the force transducer (i.e. vertically) were discarded.

Chrysolina polita (L.) is a moderately large (6·5–8·5 mm) ‘leaf beetle’ found in waterside situations throughout much of the Palaearctic region. The larval host-plants are species of Mentha and other Labiatae. The adult beetles climb these and a variety of other waterside plants.

When the recordings had been taken for each beetle it was placed in a dry atmosphere to die. The tarsi of 20 individuals were mounted on SEM stubs and gold coated with an Edwards S150 sputter coater. The specimens were examined with a Cambridge S140 SEM and counts of the numbers of adhesive setae were made from electro micrographs.

The pulling forces of 15 specimens of C. polita were recorded on glass, perspex and cloth. The beetles were anaesthetized with carbon dioxide and their tarsal claws amputated distal to the third and fourth tarsomeres, using micro-scissors. The beetles were allowed several hours to recover before their pulling forces were recorded again on the same substrates. The results from the two treatments were compared using multiple regression analysis.

A smaller but identical version of the pulling apparatus (with the exception of the pen-recorder) was placed under an inverted bell jar. A beetle was attached to the force transducer and recordings were taken using clean glass as the substratum. After several minutes the pressure of the atmosphere in the bell jar was slowly reduced in stages of 100 mm of Hg down by a maximum of 600 mm of Hg reduced pressure, using a vacuum pump. The pressure was held for about a minute at each stage. Ten specimens (five of each sex) of C. polita were used in this experiment.

Pulling force recordings were taken for ten individuals of C. polita on glass, perspex and cloth. After each beetle had started to pull near the maximum of its ability on each substrate a Zerostat Antistatic Gun was held 12 in. from the beetle and slowly fired (as described in the instructions for the gun) four times in succession. The gun, when fired, emits ions from a peizo-electric crystal neutralizing the electrostatic forces. The beetles’ pulling forces were also recorded on a sheet of glass that had been coated with a fine layer of gold (using an Edwards S150 sputter coater). The gold-coated glass was earthed by a length of electrical wire attached at the other end to the screw stand. The pulling forces from both sets of experiments were analysed by multiple regression.

Pulling forces were recorded for C. polita on a sheet of glass that had been handled and not cleaned before use. The sheet of glass was then thoroughly cleaned with acetone and absolute alcohol, and pulling force recordings were taken again for the same beetles. The results were analysed by multiple regression.

The whole pulling apparatus, with the exception of the pen recorder, was placed in an air-tight Bassaire DB/H glove box, A number of beetles in closed containers and an open-topped beaker containing about 50–100 g of phosphorus pentoxide (to dry the atmosphere of the glove box) powder were also placed in the glove box. The apparatus was then left for 24 h to allow the phosphorus pentoxide to absorb the water vapour in the air. Pulling force recordings for individuals of C. polita were made on glass that had been cleaned in acetone and alcohol before being placed in the glove box. Before being placed on the substrate the beetles were allowed to walk on filter paper to remove some of the fluid (if present) from their tarsi. These pulling-force recordings have been compared by multiple regression analysis with normal pulling force recordings of other individuals taken in the same week.

In this study the maximum recorded force a beetle achieves on a particular substratum is called its pulling force. The pulling force for each beetle was found by calibrating the oscillograph recordings with weights and adding to the maximum recorded force, the weight of the beetle, its individual hook, and that of the hook on the force transducer.

It is fortunate that the beetles used in these experiments nearly always walked vertically up the substrates and therefore pulled in the direction of optimum sensitivity of the force transducer. In most cases the beetles achieved a maximum pulling force that was very little increased after the 10 min recording periods used in this experiment. Fig. 1 shows several photographs of individuals attached to the force transducer whilst adhering to glass. In Fig. 1 (c) an unrestrained beetle is walking up the sheet of glass and its tarsi are unevenly spaced with respect to each other and to the rest of the body. In contrast, the beetles in Fig. 1 (a) and (b) are pulling against the force transducer and the tarsi are more aligned and closer together. It is noticeable in both these photographs that only one tarsus is moving (slightly out of focus) at any one time. In Fig. 1 (d) the beetle has been pulled down the glass (by winding down the screw stand to which the force transducer is attached) and the adhering tarsi are being dragged behind the descending beetle. An example of part of a recorded trace of a beetle pulling (as in Fig. 1 a, b) is shown in Fig. 2. After calibration of the force transducer with weights the maximum pull on this trace is equivalent to 33·8 mN.

The pulling forces of 62 specimens of C. polita on glass, perspex and cloth are plotted against body weight in Figs. 3–5 and mean values are shown in Table 1. Although the mean body weight for females is almost twice that of the males, there appears to be little difference in their mean pulling forces on all three surfaces. Only the regression coefficients (S) for the pulling forces of the males against body weight on all three surfaces, and that on cloth for the males and females (combined), are significantly different from zero. Therefore, only the four regression lines are shown in Figs. 3–5. The correlation coefficients (R) for males (R = 0·83), females (R = 0·68) and males and females (combined) (R = 0·76) for the pulling forces on glass and perspex are all significantly different from zero at the 0·1 % level. These contrast with the correlation coefficients of all the groups for pulling forces on glass and cloth, and perspex and cloth, all of which (except that of the males and females (combined) pulling forces on glass and cloth (R = 0·29)) are not significantly different from zero at the 5 % level. The strong correlation between the pulling forces on glass and perspex can be seen in Fig. 6.

Some of the specimens viewed with the SEM were covered with debris and their tarsal adhesive setae were too obscured to count. The numbers of adhesive setae on the left-hand tarsi of 14 of the 20 specimens viewed were counted and these are plotted against body weight in Fig. 7. Two types of adhesive setae are present in C. polita first, climbing setae, which are found in both sexes, and secondly, male setae, which are only found in the males and are used to aid pairing. In Fig. 7 the numbers of male setae and climbing setae are shown separately and combined. Regression lines are shown for the numbers of (i) male setae, (ii) male and climbing setae (combined) in males, (iii) climbing setae in females, and (iv) climbing setae in males and females (combined), since their regression coefficients are significantly different from zero at, at least, the 5 % level. In all these groups the numbers of adhesive setae increase with body weight. In Fig. 8 the pulling forces on glass are plotted against number of adhesive setae (male and climbing (combined) in males). Regression lines are fitted for females and males and females (combined) since their regression coefficients are significantly different from zero at the 5 % level. In females and males and females (combined) the pulling force on glass increases with number of adhesive setae.

Removal of the tarsal claws appeared to have little obvious detrimental effect on the beetles. Tables 1 and 2 show that claw removal has no significant effect on the pulli forces achieved on glass and perspex but results in a significant (at the 0·1 level) reduction in pulling force on cloth.

There was no obvious reduction in pulling force on glass for any of the ten individuals used when the pressure of the surrounding atmosphere was reduced. In a few cases the pressure reduction caused an increase in the beetle’s activity and thus slightly greater pulling forces were recorded than at normal atmospheric pressure. On no occasion did the beetles fall off the glass. An example of an oscillograph trace recorded in this experiment is shown in Fig. 9.

The means and standard errors of the pulling forces on glass finely coated with gold, and glass, perspex and cloth, all with electrostatic forces reduced by use of the Zerostat antistatic gun, are presented in Table 1. Comparison by multiple regression of these results with the pulling forces achieved under normal conditions on the three substrates (Table 2) shows that the reduction of electrostatic forces does not have a significant effect (at the 5 % level) on pulling force.

The means and standard errors, of the pulling forces, before and after cleaning the sheet of glass, and a multiple regression analysis of these results are shown in Tables 1 and 2, respectively. The multiple regression analysis shows that no significant (at the 5 % level) reduction or increase in pulling force is produced by cleaning the sheet of glass with acetone and absolute alcohol.

The means and standard errors of the pulling forces on glass recorded in the dry atmosphere and similar data of normal pulling forces recorded in the same week, are presented in Table 1. Comparison of this data by multiple regression analysis, presented in Table 2, shows that there is no significant effect (at the 5% level) on pulling force produced by using a dry atmosphere.

The pulling force on glass increases significantly with the total number of adhesive setae in females and males and females (combined) ; (Fig. 8) with more data it is probable that this would prove to be the case for males. The multiple regression analysis (Table 2) confirms the view that pulling force on glass is dependent on the number of setae and that body weight and sex have no significant effect. When the numbers of adhesive setae are plotted against body weight, both males and females show a significant increase in total number of setae with body weight but not when combined (Fig. 7). In addition, the general range of the total numbers of adhesive setae are similar in both sexes, as are the means and ranges of pulling forces on smooth surfaces (Table 1, Figs. 3 and 4). However, the females are considerably heavier than the males (females mean body weight = 57·6 × 10⁻⁵ N, range 31·9–82·7 ×10⁻⁵ N, males 35·9 × 10⁻⁵ N, 23·4–51·6 × 10⁻⁵ N) and one might expect them to require greater pulling forces to counter the effect of increased detachment forces. Climbing setae have long ‘semi-rigid’ shafts but very flexible plates (the distal parts of the setae used in adhesion), often bearing small projections on the dorsal, non-adhesive surface, whereas male setae have more rigid shafts and larger discoidal plates (Stork, 1980). The male setae are probably specialized for adhering to certain smooth areas of the female’s hind body and almost certainly adapt less well to the irregular topography of plant cuticles than the more flexible climbing setae. A more accurate measure of the beetle’s possible resistance to detachment forces whilst on plant surfaces may be obtained by observing the relationship of numbers of climbing setae (alone) with body weight (Fig. 7). It is evident that the numbers of climbing setae in females are greater than those in males and also that (in spite of the effect of the extra weight of the eggs in the gravid females) the number of climbing setae increase significantly with body weight in males and females (combined). It is worth noting that during pairing the female carries the male and as a result may have to resist proportionally greater detachment forces.

Ignoring differences in relative size and shape of the setae (Stork, 1980) one can use the regression coefficient in Fig. 7 to estimate that the number of adhesive setae in for example, a male Tomarcha tenebricosa (Fabricus) (Chrysomelidae) weighing above 3·9 mN is about 106500. Stork (1977) visually estimated the number of adhesive setae for a specimen of this size as about 100000 using the SEM. The fact that these two estimates are so close suggests that the number of adhesive setae is related to body weight not only in C. polita but also in other chrysomelids. Similarly, a pulling force on glass of 1·73 N can be extrapolated for the same specimen of T. tenebricosa using the regression coefficient from Fig. 3. In fact, this figure is much greater than several estimates, using this species, recorded in some preliminary pulling experiments. It is probable that larger beetles such as this species have problems in manipulating their tarsi so that all the setae can come into contact with the substratum.

Most authors who have previously examined adhesion using adhesive setae have tended to support a single mechanism, virtually ignoring the several other possibilities, (e.g. Gillett & Wigglesworth, 1932). The figures for pulling forces in this present paper, if these are accepted as approximations or indicators of adhesive forces, can be used in conjunction with information on the structure of the adhesive setae (Stork, 1980) for probably the first detailed theoretical and practical approach to most of the previously proposed modes of adhesion.

Hooke (1667) and Leeuwenhoek (1690) proposed that the adhesive setae of insects could catch on to the minute irregularities of apparently smooth surfaces. Both glass and perspex when viewed with the SEM are remarkably smooth, even at the magnification of the adhesive setae, and SEM observations of setae in contact with glass (Stork, in prep.) show that neither the setae nor the setules they sometimes bear (on the dorsal non-adhesive surface) hook onto the surface. The results of the multiple regression analysis from the claw amputation experiment (Table 2) show that claw removal does not have a significant effect on adhesion to glass or perspex, and indicate that claws play only a minor role, if any, in adhesion to these surfaces. In contrast, the analysis shows, that the claws play a significant role in holding on to cloth, and that their removal drastically affects the beetles’ pulling forces. Some preliminary experiments involving the removal of adhesive setae by cauterization support the view that they are essential for climbing smooth surfaces but are not important in climbing cloth. Since adhesion on smooth and rough surfaces is achieved by different methods, it is not surprising that there is strong correlation between the pulling forces on glass and perspex (as shown in Fig. 6) and poor correlation between the pulling forces on these surfaces and those on cloth.

Derham (1798) was probably the first to suggest that the large cup-shaped adhesive structures on the front tarsi of the male Dytiscus hold onto the female’s hind body by the exclusion of air. In other words, air between the seta and the substrate is excluded, thus causing an internal pressure deficit which is maintained by the force of the surrounding atmosphere. This method requires that the edges of the cup-shaped structures be well sealed to prevent the intrusion of atmospheric air. Derham and others Suggested that each of the adhesive setae in other animals acted as a minute sucker. If these animals hold onto smooth surfaces by suction then the reduction of atmospheric pressure should result in a drop in their adhesive force. The oscillograph recordings of pulling forces for C. polita, as illustrated in Fig. i, show no such reduction under conditions of reduced pressure.

Nachtigall (1974) notes that to produce a reduced pressure a force must be applied to pull the centre of the suction cup away from the surface. The adhesive setae in most beetles are normally ‘hinged’ proximally and are clearly not designed for this purpose. However, a slight reduction in pressure may result from the setal plate (which is probably composed of ‘elastic’ cuticle), tending to return to its original shape on being applied to the substratum.

Male dytiscids live in an aquatic environment and use adhesive setae to hold on to the females. Under these conditions, adhesion cannot be accounted for by surface tension, and difficulties in excluding the fluid layer between the setae and the female’s elytra will almost certainly severely restrict the close contact required for direct molecular cohesion. In this case, the similarity of the dytiscid male setae to suction cups and the fact that they adhere in an aquatic environment makes suction the most likely mode of adhesion. A muscular mechanism need not necessarily be invoked for the production of the negative pressure. Nachtigall (1974) reports that the male ‘strikes’ the female with its tarsi. The possible elasticity of these large cuticular structures may cause a tendency to return to their original shape on attachment, and this may produce a negative pressure. The secretion produced at the base of the giant adhesive setae, as noted by Korschelt (1923) and Blunck (1912), probably acts as a sealant around the edges of the setae.

When discussing the possible modes of adhesion in gekkos, Maderson (1964) suggested that electrostatic forces between the adhesive hairs and the substratum must be considered as a possibility. The results of the multiple regression analysis (Table 2) show that the reduction of electrostatic forces by use of an antistatic gun or use of earthed gold-coated glass has no significant effect on the pulling forces of C. polita. It has been demonstrated that the Zerostat antistatic gun removes about 70–90% of charge on gramophone records and similar amounts of charge, or less, are probably present on the glass and perspex substrates. Electrostatic forces, therefore play a small role, if any, in the production of the adhesive force by adhesive setae.

The results from the multiple regression analysis show that cleaning the sheet of glass used in the pulling experiments with acetone, and attempting to remove as much water vapour as possible by measuring pulling forces in a dry atmosphere, does not have a significant effect on the pulling forces of C. polita. However, it is extremely difficult to remove absorbed water vapour from surfaces such as glass and perspex and these experiments in no way rule out the possibility of surface tension producing the adhesive force. There may even be sufficient fluid present on the setal plates for surface tension effects to occur.

Gillett & Wigglesworth (1932) suggested that adult Rhodnius prolixus Stal (Hemip- tera: Reduviidae) hold on to smooth surfaces by means of ‘seizure’ between the adhesive setae on their front tibiae and the substratum. They proposed that the dorsal surface of each seta made contact at an angle with the substratum and that a droplet of fluid was secreted in this angle. On pulling the tibiae, and thus the setae, towards the body, thinning of the fluid at the tip of each seta results in these surfaces coming into very close contact. Molecular adhesion (see below) at these points causes seizure to occur. However, SEM observations of setae adhering to glass in C. polita (in prep.) indicate that it is the ventral surface of the seta that comes into contact with the substratum, and seizure cannot occur. It is improbable that seizure accounts for adhesion in any animal using adhesive setae. Indeed, the dorsal surface of the adhesive setae in R. prolixus and in some flies have been observed to bear small projections similar to the setules in some beetles (Stork, unpublished).

Ruibal & Ernst (1965) noted that Gekko gecko may bear 10⁸–10⁹ spatulae on the hairs on their tarsi. They suggested,, as did Edwards & Tarkanian (1970) for R. prolixus, that direct ‘frictional’ forces between the spatulae and the substratum provided the adhesive force. Various forms of molecular interactions at the points of contact cause the molecules of two surfaces to be loosely or strongly bonded together. One example, which is bound to exist even if no others are present, is a group of relatively weak long range forces called van der Waal’s forces. Molecules are weakly attracted to each other over distances of a few Angstroms by the van der Waal’s forces, but at smaller distances a repulsive force occurs. If sufficiently large numbers of molecules are effectively in contact (i.e. separated by a few Angstroms) the attractive van der Waals’ forces may be considerable and molecular adhesion may occur. Without knowing the number of points of contact and the separation of the setae from the substratum a mathematical analysis of the feasibility of these forces as the mode of adhesion is not possible. However, the fact that there are so many individual adhesive units in beetles and in gekkos, and that the separation of the setae from the substratum has been demonstrated to be less than 100 nm (Stork, unpubl.), is perhaps strong evidence to support molecular adhesion as the adhesive force.

Histological investigations of the tarsi of C. polita by the present author have failed to show any obvious secretory cells. Although several genera of chrysomelids are noted for their copious production of a tarsal secretion there appears to be little evidence of similar secretions in most beetles. Recently, Hasenfuss (1978) and Bauchhenb & Renner (1977) have supported the views of many 19th century authors that flies produce an adhesive secretion. Clearly, further electron microscope studies are required to determine whether secretions are produced.

In conclusion, the most probable mode of adhesion on smooth surfaces in Chrysolina polita, and perhaps in most animals using adhesive setae, is direct molecular adhesion of the setae with the substratum. It is probable that the cohesive forces of a thin fluid layer would increase this adhesion. If a thicker layer of fluid is present then adhesion by direct molecular adhesion and surface tension will be considerably reduced. Whether most beetles produce secretions to aid adhesion on smooth surfaces is uncertain but perhaps this may occur in some genera and other groups such as flies. Finally, the unique structure of some male adhesive organs in aquatic caraboids, and the fact that these are used in an aquatic environment, makes adhesion by suction a distinct possibility in this group.

Using the above equation for an average individual C. polita (body weight about 45 ×10⁻⁵ N, cross-sectional area approximately 16 × 10⁻⁶m²), the estimated wind speeds required to remove the beetle solely by drag, equivalent to the pulling forces of 24 mN on glass and 40 mN on cloth, are about 48 m/s. (107 m.p.h.) and 62 m/s. (137 m.p.h.) respectively.

I would like to thank Dr R. F. Ker and Prof. R. McN. Alexander (University of Leeds) for their helpful discussion and criticism of an earlier draft of this manuscript. I am also grateful to Dr M. E. G. Evans and Dr I. Grant (University of Manchester) for their advice, to Mr L. Lockey for photographic assistance, and to Prof. E. R. Trueman for the use of departmental research facilities. This research was carried out under an S.R.C. award.