2009. The Journal ofArachnology 37:68-71 Characterization of the green iridescence on the chelicerae of the tube web spider, Segestriaflorentima (Rossi 1790) (Araneae, Segestriidae) A. L. Ingram, A. D. Ball and A. R. Parker"^: Department ofZoology, Natural History Museum, Cromwell Road, South Kensington, London SW7 5BD, UK. E-mail; [email protected] O. Deparis: Facultes Universitaires Notre-Dame de la Paix, 61 Rue de Bruxelles, B-5000 Namur, Belgium J. Boulenguez and S. Berthier: Laboratoire d’Optique des Solides, CNRS Universite Paris 6, France Abstract. Segestria florentina (Rossi 1790) (Segestriidae) displays iridescent green coloration on the paturons of the chelicerae. This was confirmed by reflectance measurements, which gave a spectral peak at 505 nm. Scanning electron microscopy did not identify cuticular surface scales or sculpturing, suggesting that the cause of the iridescence was subsurface. Transmission electron microscopy revealed 86 alternate dark and light layers in the exocuticle, the mean dimensions of which were 126 nm ± 28 nm and 88 nm ± 55 nm respectively. The identity of each layer was initially unclear. However, by using a combination of materials with different refractive indices in calculations of theoretical reflectance spectra, weconcluded that they were most likely to be composed ofchitin and air, since a peak of480 nm was obtained, which most closely matched that which was recorded. The function of the green color is not clear, since S. florentina has relatively poor vision and relies predominantly on vibratory and acousticsignals. The study provides useful information relevant to research into the evolution ofstructural colors in spiders and, more generally, in nature. Keywords: Structural color, photonic, multilayer reflector Structural colors are the result of the interaction of light Spectral measurements were taken from a specimen of S. with physical structures, termed photonic crystals, which are florentina, which was provided in 70% ethanol byThe Natural in oron the surface ofa substratum. They have been identified History Museum (London). Achelicerawasdissected fromthe from a diverse range of taxa, most notably butterflies specimen and illuminated with a halogen source incident at 0° (Ghiradella 1991; Kinoshita et al. 2002; Biro et al. 2003; to the surface (“normal incidence”). Reflected light returning Vukusic et al. 2004; Ingram 2008) and birds (Prum et al. 1999; along the same illumination pathway (“backscatter”) was Zi et al. 2003; Li et al. 2005; Vigneron et al. 2006), which have measured using an Avantes Avaspec 2048/2 spectrometer. The long been the focus ofphotonics research since theirstructural reflection was normalized using a white diffusive standard. colors are very obvious. There are, however, many more Due to the difficulty ofmeasuring spectra from such a small instances of structural coloration, which are yet to be specimen and also the curvature of the sample, a second described - for example, in spiders. Of the few studies measurement was taken from the second chelicera ofthe same conducted, almost all have concentrated on jumping spiders individual, for confirmation. A 2 mm^ section of the dorsal (Salticidae), as structural colors occur mainly in this family. paturon was dissected from one sample and glued with Multilayer reflectors have been the most commonly occurring Araldite'^^ to a glass slide. This provided a sufficiently flat photonic crystal, identified from modified setae or cuticular surface from which to measure reflectance spectra with a scales (Cutler & Richards 1972; Hill 1979; Holl 1987; Land et Roper Spectral-DV"^^ spectrometer mounted on an Olympus al. 2007) and epicuticular surface sculpturing (Parker & microscope, equipped with a rotating slide holder. This Hegedus 2003) on the abdominal and cephalothoracic regions, experimental set-up enabled spectra to be recorded from on which these colors are typically located. Diffraction multiplepoints onthe specimen. Datawerethencollated using gratings have also been discovered, individually or in Melange image processing software. The specimen was combination with a multilayer reflector (Kochalka 1980; illuminated, as before, with a halogen source at close to Parker & Hegedus 2003). Here, we extend the literature on normal incidence and backscattered light (375-700 nm) was spider photonics by describing the structural origin of the collected from a narrow cone around the same angle. Spectra green iridescence on the chelicerae of the tube-web spider, were acquired at 20X magnification. Segestria florentina (Rossi 1790), using a combination of Previous studies have identified cuticular surface and electron microscopy, spectroscopy and optical modelling. subsurface structures as the cause of interference colors in Structural colors have not previously been examined from spiders. A sample was therefore prepared for scanning and this haplogyne family. It is therefore hoped that the results will transmission electron microscopy (SEM and TEM respective- provide useful information, which when considered with those ly). For the former, a chelicera was transferred from 70% describing iridescence in the highly-derived Salticidae, will ethanol to 70% acetoneand thendehydrated throughagraded contribute to our understanding ofthe evolution ofstructural acetone series before being critical point dried. It was then color in spiders. coated with gold, mounted on a metal stub using Araldite"^^ and viewed with a Philips XL30 Field Emission SEM. Images ••Former address: School of Biological Sciences, University of were recorded digitally. A further sample was prepared for Sydney, New South Wales 2006, Australia. TEM, which was immersed in 70% acetone and then 68 INGRAM ET AL.—GREEN IRIDESCENT CHELICERAE IN SEGESTRIA 69 progressively dehydrated through increasingly concentrated images, such a calculation did not lead to a clear peak in the acetone to 100% dried acetone. Samples remained here for spectrum because variations of the layer thickness across the 48 h before beginning the resin infiltration process using stack tended to destroy light interference. For this reason, we TAAB medium grade resin. Previousexperience (Kennaway et chose to model the actual structure by a periodic multilayer al. 2004) has shown that the most important factor in stack with the same numberoflayers but constant thicknesses. successfully infiltrating terrestrial arthropods is first to ensure Initial observations of the chelicerae showed that under that there is no residual water in the specimens (through most angles ofillumination and observation, the bright green extended dehydration times) and secondly to ensure that the iridescence originated from the dorsal surface of the paturon resin has penetrated thespecimens properly(by useofthinned, (Fig. 1). This was confirmed by the two measured spectra or low viscosity resin). The resin infiltration process consisted taken from this region, which indicated two similar reflectance ofimmersing the sample in a mixture ofresin and acetone and peaks around 505 nm (green) (Fig. 2). Scanning electron progressively increasing the concentration of resin in the microscopy revealed that the dorsal surface of the paturon mixture. This was achieved over four days starting with a was predominantly smooth, devoid of scales or surface mixture of approximately 20% resin, 80% dried absolute sculpting, suggesting that the origin of the iridescence lay acetone. Each mixture was changed after about 12 h and each below the cuticular surface. Sections of the paturon were, step was carried out for 24 h. Finally, the infiltrated sample therefore, examined with the TEM and showed that the was placed in a resin-filled, flat-ended BEEM capsule and exocuticle is composed of an outer region extending 7-8 pm polymerized for 8 h at 70° C. Sections for TEM were cut at towards the interior, in which around 86 alternate dark and 70-90 nm thickness, collected onto grids and counterstained light layers were observed (Fig. 3). Measurements of the usingalcoholicuranyl acetateand Reynold’s leadcitrate, prior thickness ofeach layer showed that they were highly variable: to examination in a Hitachi H7100 TEM. Images were 126 nm ± 28 nm and 88 nm ± 55 nm, respectively. For this recorded onto film and then scanned from prints. reason, the mean values were used for the theoretical Theoretical reflectance spectra were calculated by the reflectance spectra calculation. continued-fraction method, which relies on exactly solving The material of the high density layers was assumed to be Maxwell’s equations for an arbitrarily stratified medium chitin, based on previous literature (Richards 1951; Parker & (Dereux et al. 1988). In this method, the reflection coefficients Hegedus 2003). The resultsoftheacetone test showed nocolor for transverse-electric and transverse-magnetic polarized light change, suggesting that the less dense layers were not are expressed in terms of the surface impedances, which take composed of air. Based on this, we modeled spectra using a the form of continued fractions. These continued fractions combination ofchitin and a lipid with a refractive index (RI) terminate for a finite number of layers deposited onto an of 1.46 (Bausch & Lomb), since cuticle contains lipid infinitely thick substrate (of known refractive index). Their (Richards 1951). Also chitin and a low RI chitin-based values are determined as soon as the thickness and the material {n = 1.40) (Bernard & Miller 1968), and finally refractive index ofall the layers are known and the incidence chitin and air, to give weight to the acetone test. Assuming a conditions are given. The incidence angle was set to 0° refractive index ofchitin equal to 1.56 (real part) (Land 1972), (incidence medium: air) so that the specular reflection the optical path length across a chitin/low density bi-layer at occurred exactly in the backward direction (“backscatter”). normal incidence is equal to ^ = (r/^./,,-,,,, X «,./„„„) + (r//o„. X TEM Layer thickness values were determined from cross- where and diow are the average actual thicknesses sectional images and used for calculation. The stratified of the chitin layer (126 nm) and low density layer (88 nm), medium was assumed to be formed by the alternation of low 1-56 (chitin) and «/„„= 10 1.40 or 1.46 (air, lipid, or , and high density materials (the reason why we chose to model low RI chitin-based material) are the refractive indices the stack of layers by a periodic multilayer will be explained respectively. The condition of constructive interference upon later on). The precise identity ofthe two constituent materials reflection at normal incidence in a quarter wavelength was, however, unclear in the absence of data concerning the multilayer stack, S = m X 2.12, predicts that the Bragg refractive indices of materials comprising spider reflectors. wavelength (m = 1) is located in the visible range: 567 nm Since these values are hard to determine experimentally, we (air), 637 nm (lipid) and 648 nm (low RI chitin-based used the results from a number of different approaches: 1) a material) from the green end of the spectrum to the red basic infiltration test (usingacetone) fordetecting the presence (Fig. 4), suggesting that the low density layers are most likely of empty (air-filled) layers (see Parker 2000); 2) previous to be composed of air. The failure of the acetone test to studies, which have reported spider reflectors from the identify air as a constituent is likely due to the fact that the abdominal and cephalothoracic regions as being comprised structure is sealed. The reflectance of a stack made of 50 of air and chitin layers (Parker & Hegedus 2003; Land et al. alternating layers ofchitin and 1) air 2) modified chitin and 3) 2007); 3) using the aforementioned information, we alternated lipid was calculated at normal incidence. We chose to consider the refractive indices of possible constituent materials to the first 50 alternating layers (over 86 in total) due to the large obtain the best fit between the empirical and theoretical fluctuations observed in layer thicknesses. The averaged spectra. The refractive index ofthe organic material was taken thickness values of these layers, = 115 nm and cl/,,,,. = to be constant over the whole spectral range. The refractive 62, were found to be slightly lower than the ones cited above, index of the substrate (beneath the stratified medium) was leading to a shift of the Bragg wavelengths to shorter taken to be equal to that of the bulk organic material. wavelengths: 481 nm (air), 530 nm (lipid), and 538 nm (low Although the continued-fraction method allowed us to RI chitin-based material). An imaginary part (i) was added to calculate the reflectance for all layers appearing on the TEM the refractive index ofchitin and modified chitin to take into 70 THE JOURNAL OF ARACHNOLOGY — Figures 1^. SegestriaJhrentina (Rossi 1790) (Segestriidae). 1. Frontal view of an adult female, indicating green iridescent chelicera. 2. Reflectancespectra (au- arbitrary units) recorded from thepaturon usingtwo types ofspectrometer(Avantes/Olympusmicroscope-mounted). Both illuminationand measurementanglesweresetto0°(normalincidence)ineachcase. 3. Scanningelectronmicrographofatransversesection through apaturon. 4. Theoretical rellectance spectra(%) from thepaturon, bythecontinued-fraction technique. Incidenceand reflection angles set to 0°. Threecombinations ofmaterials were used: Chitin/air, chitin/lipid andchitin/modified chitin with the refractive indicesset to 1.0 (air), 1.56 (chitin), 1.46 (modified chitin), and 1.40 (lipid). account optical absorption in the chitin material («(•/„„„ = 1-56 have been aware of their existence for some time and yet the -I- i X 0.05, =1.40 -I- i X 0.05). This complex refractive cause of the color has remained unexplained. In the current index was assumed to beconstant across thewhole wavelength study wechose to examine the iridescent green chelicerae ofS. range (350-710 nm), in a first approximation. The results Jhrentina, fromwhich thecause ofthecolorwasidentifiedasa showed that the best agreement with the measured reflectance constructively-interfering multilayer reflector. Studies have spectra was obtained using chitin and air (Fig. 4). identified this type of reflector elsewhere in spiders (Cutler & The chelicerae display some of the most striking structural Richards 1972; Hill 1979; Holl 1987; Parker & Hegedus 2003; colors found in spiders (Jackson 1982). They are probably Land et al. 2007) and it is emerging as the most common type most well known from the Salticidae, however, they also occur ofphotonic crystal in this group, as it is in butterflies (Ingram in the Segestriidae. In both cases, naturalists and scientists 2008) and structurally colored animals in general (Parker . INGRAM ET AL.—GREEN IRIDESCENT CHELICERAE IN SEGESTRIA 71 2006). This convergently evolved optical structure typically Ghiradella, H. 1991. Lightandcoloron thewing-structuralcolorsin provides an effective method of displaying bright color to butterflies and moths. Applied Optics 30:3492-3500. conspecifics without attracting the unwanted attention of Hill, D.E. 1979. The scales ofsalticid spiders. Zoological Journal of predators (Parker 1998). However, it is unclear if this is the the Linnean Society 65:193-218. intended function of the color in S. florentina. Unconfirmed Holl, A. 1987. Coloration and chromes. Pp. 16-24. In Ecophysiology ofSpiders. (W. Nentwig, ed.). Springcr-Verlag, Berlin. reports suggest that the similarly green iridescent chelicerae of Ingram, A.L. & A.R. Parker. 2008. A review of the diversity and the unrelated genus, Phidippus (Salticidae), are employed in evolution of photonic structures in butterflies, incorporating the conspecific (mate) recognition (Irene Lindsey, pers. comm.). work of John Huxley (The Natural History Museum, London Additional evidence for this originates from studies of vision from 1961 to 1990). Philosophical Transactions of the Royal in the related species, P. regius Koch 1846, which showed that Society B 1502:2465-2480. the eyes have a corresponding peak spectral sensitivity in the Jackson, R.R. 1982. The behavior of communicating in jumping green (de Voe 1975). It is unlikely that the chelicerae of S. spiders. Pp. 213-245. In Spider Communication: Mechanisms and florentina function in mate recognition, since the family is Ecological Significance. (P.N. Witt & J.S. Rovner, eds.). Princeton known to have relatively poor vision, relying predominantly University Press, Princeton, New Jersey. on vibratory (Barth 1982) and acoustic signals (Gertsch 1979). Kennaway, G.M., A.S. Baker & A.D. Ball. 2004. A method for To determine the role ofthe color, behavioral and visual data preparinglightlysclerotized mites forexamination bytransmission are required to determine if the green has some adaptive Kineolsehcittrao,n mMi.c,rosSc.opyY.osShyisotkeam,atiYc.anFdujAiippl&iedN.AcaOrkoalmoogyto9.:3-290.02. function in signaling. Photophysicsofstructuralcolourin the Morphobutterflies. Forma ACKNOWLEDGMENTS 17:103-121. We would like to acknowledge funding provided by the Kochalka, J.A. 1980. The cuticular reflectors of certain spiders. American Arachnology 22:11. nEou.ro1p2e91a5n. UAnRiPonwaBsiofPuhnodted(NbEySTT)heprRoojyeactl, Suoncideetry caonndtratchte Lanudl,traMv.iFo.le,tJ.reHfloercwtionogds,caMle.sL.ofLaijmum&piLn.gDasipiqdienr..2P0r0o7c.eeOdpitnigcss oofftthhee Australian Research Council. Thanks go to Paul Hillyard and Royal Society B 274:1583-1589. Janet Beccaloni (Natural History Museum, London) for Li, Y.H., Z.H. Lu, H.W. Yin, X.D. Yu, X.H. Liu & J. Zi. 2005. specimens and helpful discussions of the manuscript. F. Structural origin ofthe brown color of barbules in male peacock Farr-Cox is thanked for Figure 1 tail feathers. Physical Review E 72:010902-010905. Parker,A.R. 1998.Thediversityand implicationsofanimalstructural LITERATURE CITED colours. Journal ofExperimental Biology 201:2343-2347. Barth, F.G. 1982. Spiders and vibratory signals: sensory reception Parker, A.R. 2000. 515 million years ofstructural colour. Journal of and behavioural significance. Pp. 67-122. In Spider Communica- Optics A: Pure and Applied Optics 2:R15-R28. tion: Mechanisms and Ecological Significance. (P.N. Witt & J.S. Parker, A.R. 2006. Conservative photonic crystals imply indirect Rovner, eds.). Princeton University Press, Princeton, New Jersey. transcription from genotype to phenotype. Recent Research Bausch & Lomb. Report S5195, 0675, Refractive index and percent Developments in Entomology 5:1-10. dissolved solids scale, Analytical Systems Division. Parker, A.R. & Z. Hegedus. 2003. Diffractive optics in spiders. Bernard,G.D. &W.H. Miller. 1968. Interferencefiltersinthecorneas Journal ofOptics A: Pure and Applied Optics 5:S111-Sl16. of Diptera. Investigative Ophthalmology and Visual Science Prum, R.O., R. Torres, S. Williamson & J. Dyck. 1999. Two- 7:416-434. dimensional Fourier analysis of the spongy medullary keratin of Biro, L.P., Z. Balint, K. Kertesz, Z. Vertsey, G.I. Mark, E. Horvath, structurally coloured feather barbs. Proceedings of the Royal J. Balazs, D. Mehn, I. Kiricsi, V. Lousse & J.-P. Vigneron. 2003. Society ofLondon Series B-Biological Sciences 266:13-22. Role ofphotonic crystal-type structures in the thermal regulation Richards, A.G. 1951. The Integument ofArthropods. University of of a lycaenid butterfly sister species pair. Physical Review E Minnesota Press, Minneapolis, Minnesota. 411 pp. 67:0219071-0219077. Vigneron,J.P.,J.-F. Colomer, M. Rassart, A.L. Ingram & V. Lousse. Cutler, B. & A.G. Richards. 1972. Sclerotization and the localisation 2006. Structural origin of the colored reflections from the black- of brown and black colors in chelicerates (Arthropoda). Zoolo- billed magpie feathers. Physical Review E 73:021914-021921. gische Jahrbuecher: Anatomic und Ontogenie der Tiere 89: Vukusic, P., J.R. Sambles & C.R. Lawrence. 2004. Structurally 404-421. assisted blackness in butterfly scales. Proceedings of the Royal de Voe, R. 1975. Ultraviolet and green receptors in principal eyes of Society ofLondon Series B - Biological Sciences 271:S237-S239. Derjeuumxp,inA.g,sJp.iPd.erVsi.gnJeoruornn,alP.ofLaGmebneirnal&PAh.yAs.ioLluocgays.661:918983.-P2o0l7a.ritons Zi,CJo.l,oXr.atYiuo,nY.strLait,egXi.esHui,nC.peXauc,ocXk. Wfeaanthge,rsX..PLruoice&edMi.ngsFu.of200t3h.e in semiconductor multilayered materials. Physical Review B National Academy ofSciences USA 100:12576-12578. 38:5438-5452. Gertsch, W.J. 1979. American Spiders. Van Nostrand Reinhold, New York. 274 pp. Manuscript received22 November2007, revised28 July2008.