Adaptive Evolution of the Ventral Scale Microornamentations among Three Snake Species

Kexin WANG,Yutian ZHAO,Chaochao HU,Hong LI,Yanfu QU and Xiang JI

Jiangsu Key Laboratory for Biodiversity and Biotechnology,College of Life Sciences,Nanjing Normal University,Nanjing 210046,Jiangsu,China

Abstract Research on the ecological effect of microronamentations on the scale surface in reptiles has been carried out over the past few decades.It is found that the microornamentation pattern in reptiles is related to their habitats.This study examined the wettability on scale surface,as well as the differences in microornamentation on ventral scales from the mid-body region in three snake species,Hypsiscopus plumbea (aquatic),Oocatochus rufodorsata (semi-aquatic) and Elaphe carinata (terricolous).Moreover,the scale specimens were metallized and analyzed using scanning electron microscopy.Our results showed that there are microornamentations on the ventral scale surfaces of the tested species,which showed interspecific differences.To be specific,the aquatic snake shows the narrow,fine and regular denticulations which are connected to reduce friction and dirt shedding.By contrast,the terrestrial snake acquired the wider and shorter denticulation which would render more friction during locomotion but it shows greater water resistance to improve the capacity of dirt shedding and compensate for the disadvantage of short and wide denticulations.Additionally,the denticulation characteristics of the semi-aquatic snake fell in between those of aquatic and terrestrial snakes.Therefore,it is deduced in this study that the ventral scale microornamentations in snakes contribute to ecological adaptation to their preferential microhabitats.

Keywords Elaphe carinata, Hypsiscopus plumbea, Oocatochus rufodorsata,scale microornamentation,snake,wettabiligy

1.Introduction

The scale is the most typical character distinguishing squamates from other amniotes,birds and mammals (Abo-Eleneen and Allam,2011).In squamates,the main function of scales is to assist animals in locomotion and to protect them from mechanical abrasion,solar radiation and dehydration (Abo-Eleneen and Allam,2011; Wegeneret al.,2014).In snakes,scales are composed of the outermost epidermal layer and inner corium layer;of them,the epidermal layer is hard,robust,and inflexible,which is composed of β-keratin,while the outermost structure of scales is called oberhautchen (Baden and Maderson,1970;Kleinet al.,2010; Krey and Farajallah,2013).All oberhautchen cells can directly contact the environment,and they are associated with various shapes and sizes in different taxa,such as the surface profile,imbricate structure,level of cell borders,presence or absence of pores/pits,and the occurrence and size of denticulation on posterior cell margins (Arnold,2002).Typically,the arrangement patterns of these cells and their surface microstructures are defined as the microornamentation(Gower,2003; Ruibal,1968).So far,a large number of studies on the microornamentation in squamates have been carried out using scanning electron microscopy.

The earliest studies mainly focused on the relationship of microornamentation with taxa (Price,1982; Stewart and Daniel,1975).For instance,Price (1982) had found that the dorsal scale microornamentation patterns of snakes from aquatic,arboreal,burrowing and desert habitats were similar,and concluded that the microornamentation patterns had reflected the phylogenetic relationships instead of the ecological or habitat factors.Thereafter,numerous studies on scale microornamentations in squamates suggested that,the interspecific differences were increasingly correlated with ontogeny,body region,scale size and structure,as well as the taxon (Allam and Abo-Eleneen,2012; Gower,2003; Harvey,1993; Price and Kelly,1989).On the other hand,Renous and Gasc (1989) inferred that parts of the microornamentation patterns were related to the locomotor behavior.Over the past few decades,numerous studies have been carried out to examine the relationship of microornamentation with ecological adaptation (Allamet al.,2017; Arnold,2002; Gower,2003).For instance,lizards from dry habitats or vegetation away from the ground have possessed larger scales to suppress the shine generated by the smooth microornamentation(Arnold,2002).However,Arrigoet al.(2019) used 353 snake species to examine the relationship between dorsal scale microstructures between habitats and found the diversity and complexity of snake scale surface microstructures are related to the phylogeny,rather than their habitats.In addition,Arrigoet al.(2019) found that the scale surface microstructures could provide some level of hydrophobicity but it is still unclear which types of microornamentations are related to being higher hydrophobicity.Nevertheless,Gower (2003) predicts the narrow and regular microornamentation ridges,together with the possible pores/pits,can provide better hydrophobicity and dirt shedding capacity in burrowing snakes.

The ecological functions of microornamentations are mainly to strengthen the dirt shedding capacity and to reduce the friction during locomotion and shine (Arnold,2002; Gower,2003; Rocha-Barbosa and Moraes-e-Silva,2009).Currently,two competing scientific hypotheses have been proposed to interpret the relationship of ecological adaptation with the microornamentation patterns.For one thing,the demands for friction reduction and dirt shedding may require that smooth microornamentations are favored by the natural selection (Gower,2003).Notably,a more regular surface can effectively reduce the friction produced when squamates are passing through the grassland or narrow burrows (Arnold,2002; Gower,2003; Rocha-Barbosa and Moraes-e-Silva,2009).Such a phenomenon is referred to as the “Lotus-Effect”,which suggests that the regular microstructure is becoming superior in dirt shedding on the surfaces of plant leaves and insect wings(Barthlott and Neinhuis,1997; Wagneret al.,1996).For another,the microornamentation should possess more complex threedimensional (3D) structures to reduce shine (Arnold,2002).For instance,the uneven microornamentation is favorable to reduce shine,thus strengthening the camouflage ability of animals(Arnold,2002).In brief,the microornamentation patterns are controlled by multiple conflicting selection pressures.

Most previous studies have qualitatively analyzed microornamentations but most of them used a single sample and only a small proportion of them have involved quantitative analysis.For example,the data eigenvalues of the denticulation length are used to phylogenetically analyze the history of scale microornamentation in Lacertid lizards (Arnold,2002).Snake scales,especially for the ventral scales,are frequently,direct and continuous contact with their surrounding environment.They need to provide high friction for forward motion or low friction in sliding along the contact surface (Baumet al.,2014).Therefore,the microornamentation patterns,which display important functions,are related to specific adaptation to the ecological factors like habitat,substrate and locomotion (Klein and Gorb,2012; Krey and Farajallah,2013).

Scale microornamentations can be regenerated repeatedly,which are similar to those present in the process of embryogenesis (Alibardi,2000).Meanwhile,they can also effectively reduce the wear resistance on natural substrates and help to adapt to the crawling environment (Klein and Gorb,2012).Taken together,these characteristics have made it possible to carry out quantitative study on microornamentation to explain the influence of habitats on snakes.Noteworthily,the arrangement,distribution and shape of microornamentation patterns may probably affect the macrostructures and function of scales in different species,even though they are similar.

In this study,three most common snakes in China,includingHypsiscopus plumbea,Oocatochus rufodorsataandElaphe carinatawere selected.We use the microornamentation of ventral scales and wettability data for quantitative analysis to verify whether variations in microornamentation patterns on the ventral scales were related to their microhabitats.In these three species,H.plumbeaoccupies pond,river and paddy field,with the distribution scope ranging from south China to Indonesia(Jiang,1998).Meanwhile,O.rufodorsatais a semi-aquatic species,which can be extensively discovered in ditch,as well as the field along river banks from central China to Korea (Zhao,1998).E.carinatais a terrestrial snake species,which can be mainly found in most mountainous,hilly and plains in China and Vietnam(Huang,1998).

2.Material and methods

Seven male specimens for each species were collected in this study,among which,H.plumbeawas collected from Fujian Province of China,with the average snout-vent length of(42.90±0.99) cm (range:39.50-46.10 cm).Moreover,O.rufodorsataandE.carinatawere obtained from Zhejiang Province of China,with the average snout-vent lengths of (124.44±5.80) cm(range:108.60-149.30 cm) inO.rufodorsataand (125.49±7.40) cm(range:86.70-144.60 cm) inE.carinata,respectively.The ventral scales of these three snakes were shine and iridescence,but their abdomen colors were different from one to another.Specifically,the abdomen ofH.plumbeawas pale yellow,with a black vertical line in the center,while that ofO.rufodorsatawas yellow with black spots,and that ofElaphe carinatawas almost yellow,with dark spots on the posterior margin.All samples were then preserved in the refrigerator at-20 ℃ in the Laboratory of Nanjing Normal University.

2.1.WettabiligyThe wettability on material surface is one of the important performance indicators,which can be determined based on the balance of surface energy in the interfaces between air,liquid,and solid materials (Mao,2016).In addition,it can also be evaluated experimentally through surveying the contact angle of one drop of water on the skin surface.Typically,the larger contact angles can be measured on the hydrophobic surfaces.In this study,skin from the center of venter was cut into 1cm×1cm pieces and rinsed with the ultrasonic bath (Bandelin Sonorex RK 100) for 5 min before drying for 24 h.Afterwards,a pipette was used to add 10 μl distilled water droplets onto the abdominal scales placed onto the microscope slides at room temperature.Subsequently,the camera was fixed in parallel with the slide,and the sample was photographed and labeled.Later,all photographs were measured for contact angle using the ImageJ version 1.43u.

2.2.MicroornamentationAll ventral scales were selected from the midbody regions of the specimens,and cleaned before examination.Similarly,skin collected from the center of venter scales was cut into 5 mm×5 mm pieces and mounted on the microscope slides with string.Then,the samples were rinsed with distilled water for 20 min,fixed in 5% formalin for 24 h,rinsed for 1 h again,and cleaned in the ultrasonic bath for 5 min before they were immersed into a consecutive series of ethanol solutions (30%,50%,70%,80%,95% and 100%).Afterwards,the samples were maintained in each solution for 15 min at room temperature,followed by air-drying for 24 h and then affixing with the assistant double-sided adhesive tape to mount onto the aluminum scanning electron microscope(SEM) stubs.Subsequently,the mounted samples were coated with gold-palladium and observed under the SEM (JSM-5610LV) at the accelerating voltage of 8-10 kV,so as to select the appropriate magnifications for characterization and further analysis.Additionally,the dimensions of microornamentations were also measured in the SEM images using the ImageJ version 1.43u.Specifically,three pictures were obtained for each sample,and the average of every trait was calculated according to the 150 denticulations randomly selected in three pictures.Moreover,to quantitatively describe the microscopic morphology of the ventral scale,the length,width,spacing,area and row spacing of denticulations were measured,respectively.Additionally,the height of scales (from the base to the edge of scales) was also determined since the size of scale could affect that of denticulations.Figure 1 has depicted the measurement methods of these denticulation traits.

Figure 1 Schematic diagram of the characteristic parameters of denticulation.The width of denticulation refers to the upper width of denticulation,as denoted by “W”.The length of denticulation represents the vertical distance from top to bottom of denticulation,as indicated by “L”.The center distance of two denticulations stands for the distance between two denticulations,as suggested by “D”.The area of denticulation is the area enclosed by “abc”.The row spacing suggests the distance between the two rows of denticulation.W:Width; L:Length; D:Distance.

2.3.Data analysisThe STATISTICA 10.0 (StatSoft Inc.,Tulsa,Oklahoma,USA) was adopted for all statistical analyses,and the significance level was set at 0.05.Prior to parametric analyses,all data were tested for the homogeneity of variance using the Bartlett’s test,and the normally distributed data were analyzed using the Kolmogorov-Smirnov test.Meanwhile,the oneway analysis of variance (ANOVA) was utilized to examine the differences in the contact angle,the aspect ratio and the area of denticulations among these three species,as well as the difference in the length of denticulations (in which the height of denticulations was used as the covariate,while species as a factor),and the differences in the width,spacing and row spacing of denticulations (in which the length of denticulations was used as the covariate,whereas species as a factor).Besides,Tukey’s test was employed to test the post hoc comparisons for variables that were different among these three species.Throughout this article,all values are presented as the mean ±standard error (SE),and the significance level is set at a=0.05.

3.Results

3.1.WettabilityOur results suggested that contact angles were (91.20±2.03)° (80.88-96.62),(78.68±2.63)° (70.89-90.13) and(68.63±2.56)° (60.82-77.01) forE.carinata,O.rufodorsataandH.plumbea,respectively.Obviously,the terrestrial snake (E.carinata)had the largest contact angle,while the aquatic snake (H.plumbea) had the lowest one (F2,18=21.79,P＜0.0001).

3.2.Qualitative traits of scale microstructureNotably,the microornamentation on the scale surfaces of snakes seemed to be related to their living environments (Rocha-Barbosa and Moraes-e-Silva,2009).Therefore,we investigated whether the significant difference in scale microornamentations was found in three kinds of snakes.All ventral scales had no sensilla on the surface and the keel only existed on scales ofO.rufodorsataandH.plumbea.Under high magnification,the surface microstructures of three kinds of snakes displayed similar characteristics.As for the microscopic surface morphology in three kinds of snakes,a regularly arranged surface structure could be observed,which was composed of the small denticulations and micropores (Figure 2).Typically,the denticulations were caudally oriented and paralleled to the body axis,while the pores and pore-like structures were found on the scales in the three kinds of snakes,which were located at the junction of two adjacent denticulations(Figures 2B,2D and 2F).

Additionally,the denticulation ofH.plumbeashowed a triangular structure,along with a long and thin tip (Figure 2B),and the distance between adjacent denticulations was close to each other,with almost no plate structure (Figure 2B).Similarly,the denticulation ofO.rufodorsataalso exhibited a triangular structure,but its length was shorter than the row spacing,which had resulted in a “plate” structure (Figure 2D).By contrast,the denticulation ofE.carinataappeared to be a finger-like structure,which was shorter and thicker,with a more obvious plate structure (Figure 2F).

Figure 2 Ventral scales microstructure of ventral scales.A and B display the different magnifications of the ventral scales of Hypsiscopus plumbea.C and D exhibit the different magnifications of the ventral scales of Oocatochus rufodorsata.E and F present the different magnifications of the ventral scales of Elaphe carinata.Ke:Keel; De:Denticulation; Po:Pore; VS:Vertical stripe.

Table 1 Values of surface microstructure characteristics for the ventral scale.

3.3.Quantitative analysis of scale denticulationThe microscopic morphology of denticulations is depicted in Table 1,and the statistical analysis results were determined based on the measurement data.Using the height of scale as a covariate,the one-way ANCOVA revealed that there were significant differences in the length of denticulations among the three kinds of species (F2,17=5.17,P=0.02).Specifically,the denticulations inH.plumbea,andO.rufodorsatawere longer than that inE.carinata(Table 1).Besides,the width (F2,17=5.16,P=0.02) and row spacing (F2,17=10.90,P＜0.001) of denticulations successively increased withH.plumbea,O.rufodorsata and E.carinata(Table 1).Obviously,the denticulation spacing ofH.plumbeawas shorter than those of the other two kinds of snakes (F2,17=4.79,P=0.02) but the difference of the denticulation spacing betweenO.rufodorsata and E.carinatawas not statistically significant (Table 1).In addition,there was not the difference in the area of denticulations among the three kinds of snakes(F2,18=3.53,P=0.05; Table 1).On the other hand,the width-tolength ratio of denticulations successively elevated withH.plumbea,O.rufodorsata and E.carinata(F2,18=46.13,P＜ 0.001;Table 1).

4.Discussion

Snakes live in a wide spectrum of habitats,and their limbless locomotor and/or behavioral specializations have reflected the differences in epidermal architecture,which are correlated with their predatory strategies and properties of the contact surface where they move (Klein and Gorb,2014; Rocha-Barbosa and Moraes-e-Silva,2009).The shape and arrangement of microornamentation may affect the friction and wear on the scale surface when the animal moves.Therefore,the length of denticulations,the number of spinule rows,and the presence or absence of pits/pores may be correlated with the preferred habitats of snakes (Gower,2003; Rocha-Barbosa and Moraese-Silva,2009).In addition,Klein and Gorb (2012) discovered that the epidermal thickness might be related to the specific adaptation to animal habitat and locomotion.On this account,scales and microornamentations of their surface not only represented the microstructure of animal,but also stood for its macrostructure (Rocha-Barbosa and Moraes-e-Silva,2009;Smithet al.,1982).

Furthermore,the variations in microornamentation patterns of squamates have been extensively observed in intra-species and inter-species (Allam and Abo-Eleneen,2012;Allamet al.,2017; Gower,2003; Rocha-Barbosa and Moraes-e-Silva,2009).Such differences in microornamentations allow to explain the functional adaptation of microornamentations to their surrounding environment (Stewart and Daniel,1973).On the other hand,Williams and Peterson (1982) had proposed a novel evolution adaptation complex between adaptation and the morphological divergences of scales.In addition,Gower(2003) suggested that the variations in microornamentations could be ascribed to the selection pressure of various habitats.Besides,other studies also indicate that microornamentation variations on the scale surfaces are related to the natural selection pressures of their habitats (Allam and Abo-Eleneen,2012; Allamet al.,2017; Rocha-Barbosa and Moraes-e-Silva,2009). After analyzing surface nanostructures in 353 species spanning 19 of the 26 families of snakes and performing phylogenetic mapping on the snake phylogeny,Arrigoet al.(2019) did not find a correlation between nanomorphological characters and a simple life habit classification.Hence,these authors suggested that the diversity and complexity of snake skin surface nano-morphology are dominated by phylogenetic rather than habitat-specific functional constraints.

It was found in this study that the ventral scales of the three investigated snakes had attractive iridescence and faint reflections but these colors were almost indistinguishable under the naked eye.Notably,the contact angle of one water droplet on the material surface is indicative of the wettability of the contact surface (Spinneret al.,2013).In this study,the contact angles of the ventral scales of three kinds of snakes were compared,which suggested that the terrestrial snake (E.carinata) had the greatest contact angle whereas the aquatic snake (H.plumbea) had the lowest one.Additionally,the microornamention patterns were similar among the three snake species,which were made up of digitations and pits(Figure 2).The presence of cell border digitations is the ancestral state for snake skin microstructures which were subsequently and independently lost in multiple lineages (Arrigoet al.,2019).The following phenomena could be found after comparing the microornamentation data of ventral scales among the three species.(1) terrestrial snakes (E.carinata) had the shortest denticulation length,while aquatic snakes (H.plumbea) had the longest one (Table 1 and Figure 2); (2) the denticulation width of terrestrial snakes was remarkably larger than those of aquatic and semi-aquatic snakes when the denticulation length was set at the same length (Table 1 and Figure 2); (3) the column space between the denticulations of terrestrial and aquatic snakes was greater than that of aquatic snakes (Table 1 and Figure 2);and (4) the row space between the denticulations of terrestrial snakes was the greatest,while that of aquatic snakes was the smallest (Table 1 and Figure 2).

The big scale,especially for the ventral scale of snake,can generate shine (Arnold,2002),while the microornamentation contributes to reducing shine and generating iridescence,since it can form a reflective surface to reflect light rays in different ways or to scatter light,thus restricting or eliminating shine(Arnold,2002).Noteworthily,various mechanisms have been proposed to explain the iridescence in nature,such as the thinfilm interference,multilayer reflectors,photonic crystals,and diffraction gratings (Kinoshitaet al.,2008; Kroisset al.,2009;Seagoet al.,2009).On the other hand,the iridescent colors mainly allow the animals to communicate with conspecifics and avoid predators (Doucet and Meadows,2009).In squamates,a fine arrangement of microornamentation can reduce the shine on the scale surface and strengthen the dirt shedding capacity (Arnold,2002; Gower,2003).However,many researchers have supposed that iridescence is only regarded as a by-product of microornamentations,and there is no plausible explanation for the adaptive evolution.For instance,the fossorial snake spends most of their life time under the ground (Gower,2003).In addition,iridescent color has not been observed inT yphlops mirus(Gower,2003).Similarly,our results suggested no significant differences in shine and iridescence among these three kinds of snakes,although the microornamentation patterns among them showed significant differences in our study.

The wettability of scale surface can be determined by the scale-liquid contact angle,which represents a quantitative measure of the wettability of a solid surface by a liquid.Differences in the wettability between animal skins can be ascribed to the chemical composition properties and the geometric structure on the skin surface (Spinneret al.,2013;Arrigoet al.,2019).In squamates,the outermost structure of scales is referred to as oberhautchen (Irishet al.,1988),and microornamentation includes the surface features of oberhautchen and the epidermal folding (Arnold,2002).Hence,the 3D structure and arrangement of microornamentation will affect the wettability of the scale surface.Specifically,the microornamentation,especially for the denticulations and ridges,can improve the wettability of scale surface,as previously reported in numerous publications (Autumn and Hansen,2006; Gower,2003; Spinneret al.,2014; Spinneret al.,2013; Arrigoet al.,2019).

Generally,a contact angle of ＜ 90° (low contact angle)has indicated the quite favorable wetting of the surface,and the fluid can spread over a large surface area.By contrast,a contact angle of ＞ 90° (high contact angle) is indicative of the unfavorable wetting of the surface.Therefore,the fluid has minimal contact with the surface and will form a compact liquid droplet (Shiet al.,2016).Nowadays,the contact angle of scale surface has been measured in some species (Gower,2003;Spinneret al.,2013; Arrigoet al.,2019).According to our results,terrestrial snakes had the largest contact angle of ＞ 90°,while aquatic snakes had the smallest one,and the contact angle of terrestrial and semi-aquatic snakes was ＜ 90°.Therefore,it could be deduced that the scales of terrestrial snakes had higher water resistance.Besides,Gower (2003) suggested that the contact angles were ＞ 90° forMelanophidium wynaudense,Uropeltis spp.,Rhinophis philippinus,andCylindrophis maculatus,while those were ＜ 90° forTyphlops mirusandT.reticulatus.

The microornamentation has possessed the narrow and regular ridges,which may be more advantageous in reducing friction and dirt shedding.Meanwhile,these microornamentation patterns may be correlated with the passively reduced adhesion of moist substrate onto the scales,since lower energy is required on the scale surface (Gower,2003).For example,the microornamentation with narrow,finely and regularly spaced ridges is correlated with the less wettable scales in Uropeltid snakes (Gower,2003); on this account,Uropeltid snakes had superior dirt shedding capacity under moist conditions (Gower,2003).In this study,aquatic snakes had the longest,while the finest and narrowest ridges,thus,they might perform better in dirt shedding (Table 1 and Figure 2).

Scales of snakes,especially for terrestrial snakes,are involved in locomotion to promote movement through muscle activity(Wegeneret al.,2014).Specifically,the number and geometry structure of microornamentation will change the friction between scales and substrate,in the meantime of protecting their bodies from abrasion (Alibardi,2003; Oufieroet al.,2011).To enable the undulating locomotion,the snake skin has to offer high friction to support forward motion whereas low enough friction to enable the sliding along the substrate simultaneously (Baumet al.,2014).Many researchers have discovered that the greater frictional resistance of scales from the posterior to the anterior direction is expected when the long denticulation has overlapped the posterior borders of each spinule row (Gower,2003; Rocha-Barbosa and Moraes-e-Silva,2009).Different from previous results,we found that terrestrial snakes (E.carinata) owned the greater denticulation row space,along with the largest denticulation width.Therefore,it was supposed that increasing the denticulation width contributed to increasing the friction,which could in turn provide the snakes with the strong power to glide forward.Additionally,our results indicated that,terrestrial snakes could acquire friction during locomotor,however,they had to sacrifice their dirt shedding ability.It could be discovered based on the contact angle data that,it might be possible to change the scale properties and improve the water-resistance to compensate for the lack of dirt shedding.

In conclusion,these three snake species investigated in this study have exhibited significant differences in the wettability and microornamentation pattern on the ventral scale surface,suggesting that these differences may actually be ascribed to the natural selection pressures and the tradeoff between wettability and friction,which may ultimately result in changes of the skin properties for terrestrial snakes.Besides,such differences are conducive to the survival of snakes,which help them to occupy more ecological niches.This study may potentially provide important quantitative data to explore a succession of questions about the relationship between the diversity and ecological adaptation of habitats among snakes.Despite this,we are still unable to draw definitive conclusions indicating that the microstructures of the three snakes are related to their habitats,the current results may also be related to the small number of species.Therefore,it is also necessary to increase the species and the exact physiological,mechanical and behavioral studies to confirm whether this result can be verified in a larger species range.

AcknowledgementsThis study was carried out in compliance with the current laws of China,and was supported by grants from Natural Science Foundation of China(31200283 and 31770443),and the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD)of Jiangsu Higher Education Institutions.The authors would like to thank Kun GUO and Li MA for help in sample collection during the research.