Arthropod Structure & Development 43 (2014) 97e102
Contents lists available at ScienceDirect
Arthropod Structure & Development journal homepage: www.elsevier.com/locate/asd
Mouthpart separation does not impede butterfly feeding Matthew S. Lehnert*, Catherine P. Mulvane, Aubrey Brothers Department of Biological Sciences, Kent State University at Stark, 6000 Frank Ave. NW, North Canton, OH 44720, USA
a r t i c l e i n f o
a b s t r a c t
Article history: Received 15 November 2013 Accepted 13 December 2013
The functionality of butterfly mouthparts (proboscis) plays an important role in pollination systems, which is driven by the reward of nectar. Proboscis functionality has been assumed to require action of the sucking pump in the butterfly’s head coupled with the straw-like structure. Proper proboscis functionality, however, also is dependent on capillarity and wettability dynamics that facilitate acquisition of liquid films from porous substrates. Due to the importance of wettability dynamics in proboscis functionality, we hypothesized that proboscides of eastern black swallowtail (Papilio polyxenes asterius Stoll) (Papilionidae) and cabbage butterflies (Pieris rapae Linnaeus) (Pieridae) that were experimentally split (i.e., proboscides no longer resembling a sealed straw-like tube) would retain the ability to feed. Proboscides were split either in the drinking region (distal 6e10% of proboscis length) or approximately 50% of the proboscis length 24 h before feeding trials when butterflies were fed a red food-coloring solution. Approximately 67% of the butterflies with proboscides split reassembled prior to the feeding trials and all of these butterflies displayed evidence of proboscis functionality. Butterflies with proboscides that did not reassemble also demonstrated fluid uptake capabilities, thus suggesting that wild butterflies might retain fluid uptake capabilities, even when the proboscis is partially injured. Ó 2013 Elsevier Ltd. All rights reserved.
Keywords: Proboscis Functionality Lepidoptera Wettability Pollination Fluid uptake
1. Introduction Mouthpart functionality of fluid-feeding insects e more than half of all known animal species (Foottit and Adler, 2009) e is an important component of disease transmission and the stability of insect-pollination systems (Kingsolver and Daniel, 1995). Mouthparts of fluid-feeding insects, such as butterflies and moths (Lepidoptera), might be subjected to damage while seeking mates, searching for food, or from predator encounters. Mouthparts rendered nonfunctional, therefore, could affect fitness (Krenn, 1997) and impact insecteflower interactions. Most Lepidoptera have a coilable, tube-like proboscis that transports fluids, such as nectar, sap, fruit juices, and blood (Adler, 1982) to the insect’s gut. The lepidopteran proboscis is composed of two elongated maxillary galeae that are connected by overlapping dorsal and interlinking ventral structures (i.e., legulae) to form a food canal (Eastham and Eassa, 1955; Krenn et al., 2005; Krenn, 2010). The distal 5e20% of the proboscis has dorsal legulae that are elongated and more widely spaced (Krenn et al., 2001) (i.e., the drinking region), which facilitates fluid uptake (Lehnert et al., 2013). The merging of the galeae into a functional proboscis takes place after adult eclosion from the pupa, and consists of coiling and * Corresponding author. Tel.: þ1 (330) 244 3349. E-mail addresses: [email protected] (M.S. Lehnert), [email protected] (C. P. Mulvane), [email protected] (A. Brothers). 1467-8039/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.asd.2013.12.005
uncoiling actions of the proboscis accompanied by the presence of saliva droplets (Krenn, 1997). Proboscis assembly must occur before sclerotization of the legular cuticle, otherwise the proboscis is putatively nonfunctional and reassembly cannot occur (Krenn, 1997). A functional proboscis is widely considered a sealed tube that operates similar to a drinking straw (Krenn, 2010; Bauder et al., 2013), solely relying on the sucking pump in the head for fluid uptake (Kingsolver and Daniel, 1995; Eberhard and Krenn, 2003); however, recent experiments have demonstrated that aqueous solutions can enter between dorsal interlegular spaces along the proboscis (i.e., not a sealed tube) (Monaenkova et al., 2012) and that a straw-like structure is not necessary for functionality (Grant et al., 2012). The proboscis employs capillarity via interlegular spaces to build liquid bridges in the food canal for the sucking pump to act on when feeding from liquid films and porous substrates (Monaenkova et al., 2012), such as rotting fruit. Fluid uptake is further regulated by wettability dynamics (i.e., hydrophilicity and hydrophobicity) of proboscis structures (e.g., hydrophilic dorsal legulae, chemosensilla, and the food canal) and surface roughness (e.g. microbumps that create an overall hydrophobic surface, explained using the Cassie-Baxter model, Cassie and Baxter, 1944; Lehnert et al., 2013). Based on our current understanding of the multifaceted fluid uptake system of butterfly proboscides we hypothesized that previously assembled proboscides of two distantly related nectar-feeding butterfly species, the eastern black swallowtail, Papilio polyxenes asterius (Papilionidae), and cabbage
98
M.S. Lehnert et al. / Arthropod Structure & Development 43 (2014) 97e102
butterfly, Pieris rapae (Pieridae), will maintain functionality following experimental splitting of the galeae as long as both galeae are subsequently placed in a feeding solution. 2. Materials and methods 2.1. Butterfly rearing and proboscis measurements Eggs of P. p. asterius were obtained from a female captured in North Canton, OH. The larvae were reared on parsley (Petroselinum crispum) and kept in RubbermaidÒ Takealong containers. Larvae of P. rapae were obtained from Carolina Biological Supply Company (Burlington, NC, USA) and reared on artificial diet. Larvae and pupae of both species of butterflies were maintained at 22 C, 61% relative humidity (r.h.), and an L18:D6 photoperiod in an environmental chamber (Percival Scientific, Inc., Perry, IN, USA). Adults from an F2 generation of P. p. asterius also were used for experiments. All adults were fed a dilute honey:water solution (1:5) daily for at least three days before feeding experiments and kept in glassine envelopes in a refrigerator (4 C) between feeding times. Proboscis lengths and drinking region lengths were measured to determine possible effects on functionality between treatments. In order to acquire proboscis measurements, butterflies were stabilized on a piece of Styrofoam and proboscides were uncoiled using insect pins. Images of the total length of proboscides (0.78 magnification) and drinking regions (4.0 magnification) were acquired for each butterfly with a Leica M205 C stereomicroscope and an IC 80HD camera (Heerbrugg, Switzerland) and measured using ImageJ software (http://rsbweb.nih.gov/ij/). The drinking
region was measured from the tip of the proboscis to a transition point where the dorsal legulae narrow and remain similar in width for the remainder of the proboscis length (Fig. 1A). Although wettability dynamics of proboscides have been reported for other butterfly species (Monaenkova et al., 2012; Lehnert et al., 2013), we demonstrated these dynamics using the proboscis of an individual P. p. asterius. The galeae were split and one galea (unstained) was placed on a slide in dH2O with a coverslip and imaged using an Olympus Confocal Microscope IX81 with DSU (Center Valley, PA, USA) (999.6 ms exposure, 20 magnification, 30 slices, 1.60 average depth slice, CY3 channel). The other galea was stained with Nile red for 24 h and imaged similarly to distinguish hydrophilic and hydrophobic structures. Proboscides of P. rapae were dehydrated in an ethanol series (80%, 90%, 100%, 24 h each), air-dried with hexamethyldisalizane, platinum sputter-coated for approximately 1 min, and imaged with a Hitachi TM3000 scanning electron microscope (Hitachi High Technologies America, Inc., Dallas, TX, USA). 2.2. Experimental feeding trials All butterflies were fed a 20% sucrose solution and kept at room temperature (24 C, 61% r.h.) in a netted Bug Dorm (BioQuip Products, Rancho Dominguez, CA, USA) 24 h prior to feeding experiments. Randomly selected butterflies were prepped for the experimental feeding trials by separating the two galeae either in the drinking region using an insect pin or had approximately 50% of their proboscis separated proximally starting at the tip (inset in Fig. 1A) immediately after being fed the 20% sucrose solution. Before proboscides were split, all butterflies had their proboscides
Fig. 1. Proboscis assembly and fluid uptake of split proboscides of butterflies. (A) Stereomicroscope image of an uncoiled proboscis of P. p. asterius displaying the galeae (ga) and overlapping dorsal legulae (dl). The dorsal legulae are larger and more widely spaced in the drinking region; the remainder of the proboscis represents the nondrinking region. The inset shows a proboscis of P. p. asterius split with insect pins (ip) for the red-50 treatment. (B) Photograph of a P. p. asterius obtained shortly after emergence showing the partially assembled proboscis (separated galeae) during the assembly process. (C) SEM image of a single galea of P. rapae showing the food canal (fc) and dorsal (dl) and ventral legulae (vl) that interlink during proboscis assembly. (D) Stereomicroscope image of the dorsal legulae of P. p. asterius in the nondrinking region; there is little overlap of the dorsal legulae. (E) SEM image of a proboscis of P. rapae showing the overlapping dorsal legulae in the nondrinking region and microbumps (mb). The arrangement of the dorsal legulae differs between P. p. asterius and P. rapae.
M.S. Lehnert et al. / Arthropod Structure & Development 43 (2014) 97e102
investigated by prodding the proboscis with an insect pin to ensure the galeae were assembled (Fig. 1B shows a proboscis during the assembly process after eclosion from the pupa). Butterflies were assigned into a control group or one of three experimental treatments and then fed solutions for 10e13 min by manually placing only the drinking region of both galeae into a 100 ml capillary tube 24 h after experimental splitting. The diameter of the capillary tube was larger than the width of both galeae in the drinking region, therefore, the tube had little effect on pressing the galeae together, thus avoiding simulating a sealed proboscis. The control group consisted of butterflies with an unsplit proboscis fed an aqueous 20% sucrose solution. The three treatments consisted of: 1) butterflies with unsplit proboscides fed a 20% sucrose and redfood coloring mixture, referred to hereafter as red-sucrose solution (i.e., red-unsplit treatment), 2) butterflies with proboscides split at the drinking region fed the red-sucrose solution (i.e., red-dr treatment), and 3) butterflies with approximately 50% of the proboscis split and fed the red-sucrose solution (i.e., red-50 treatment). Due to the dark pigmentation of the proboscis of P. p. asterius, feeding ability was determined by dissecting the butterflies in PBS physiological saline solution (7.2 pH) using insect pins and dissecting scissors within 1 h after the feeding trials (butterflies were temporarily stored in 4 C refrigerator before dissections). The crop of the alimentary canal was isolated and the average a* color values (a* ¼ green to red, scale 120 to 120, L*, a*, and b* color values) were acquired using LensEye color analysis software (Lehnert et al., 2011). Dissections of P. rapae specimens were not necessary due to transparency of the dorsal legulae of the proboscis and the ability to visualize fluid flow in the food canal. 2.3. Data analysis Significant differences (p < 0.05) in proboscis length, drinking region length, percentage of proboscis length represented by the drinking region (i.e., percentage drinking length), and a* color values were tested between sexes (t-test) and between treatments (analysis of variance) for P. p. asterius. Significant differences in mean a* color values were ranked between treatments using Fisher’s least significant difference (LSD) test. Due to differences in sample sizes between treatments with P. rapae, an independent sample Kruskale Wallis test was used to determine significant differences (p < 0.05) in proboscis length, drinking region length, and percentage drinking length. All statistical analyses were processed in SPSS Statistic 21.0. 3. Results 3.1. Evidence for proboscis reassembly All butterflies with experimentally split proboscides were checked for reassembly by applying pressure to the dorsal legulae
99
in previously split regions with the tip of an insect pin before being fed. All butterflies that had the galeae separated in the drinking region (i.e., red-dr treatment, Table 1) had proboscides that appeared reassembled before feeding, except for one individual of P. p. asterius, which was observed to have a fluid, likely saliva, present between the separated galeae of the drinking region. An individual of P. rapae died before we could check for proboscis functionality; however, inspection revealed a reassembled drinking region. It is unknown if reassembly of the drinking region involved complete interlinking of the ventral legulae; only the dorsal legulae were inspected. Based on visual inspection, all P. p. asterius in the red-50 treatment (n ¼ 4) had reassembled proboscides however, when prodded with an insect pin only one remained assembled, the other three proboscides had galeae that partially separated where pressure was applied. Three of the four P. rapae in the red-50 treatment remained reassembled after pressure was applied with an insect pin (Table 1). The individual without a reassembled proboscis had the galeae almost completely separated (over 90%), indicating further separation after our experimental splitting, and one of the galeae was coiled and appeared dry, which prevented us from being able to insert both galeae into the capillary tube to test proboscis functionality. There were no significant differences in proboscis length (mean s.e.m., 17.92 0.34 mm), drinking region length (1.64 0.03 mm), or percentage drinking region (9.21 0.23%) between treatments or sexes of P. p. asterius (n ¼ 4 per treatment) or P. rapae (proboscis length 10.84 0.14 mm, drinking region length 0.77 0.02 mm, percentage drinking region 7.11 0.18%; n ¼ 12), suggesting that the differences observed between treatments were not influenced by these structural measurements (Table 1).
3.2. Feeding ability of split proboscides All individuals that had the drinking region of both galeae inserted into the capillary tube were observed to feed (suggested by the depletion of solution in the tube), except one individual of P. rapae from the unsplit-red treatment. All individuals of P. rapae with proboscides that reassembled and fed the red-sucrose solution were observed to have a red fluid traveling through the food canal, which was visible through the semi-transparent dorsal legulae (Fig. 2A); butterflies in the control group were observed to have clear fluids traveling through the proboscis. Three individuals of P. p. asterius from each of the red-unsplit and red-50 treatment had a red droplet appear on the dorsum of the dorsal legulae at proximal regions of the proboscis (i.e., nondrinking region) while feeding (two simultaneous droplets on one individual); the droplets exhibited a pulsing-like motion. In addition, anti-parallel movements of the galeae were observed on an individual of P. p.
Table 1 Measurements (mean s.e.m.) of proboscides and confirmation of mouthpart functionality of butterfly species. All butterflies in each treatment were checked for proboscis reassembly and feeding ability after experimental splitting. The numbers of individuals that reassembled their proboscis and fed on solutions are followed by the total sample size for that treatment in parentheses. Mouthpart functionality was determined by color quantification of the crop (a* values) or visual observation of fluids in the proboscis. Species
Treatment
Reassembled proboscides
Individuals that fed
Mouthpart functionality (* ¼ a* color values)
P. p. asterius
Control Red-unsplit Red-dr Red-50
4(4) 4(4) 3(4) 1(4)
4(4) 4(4) 3(4) 1(4)
31.83 53.46 48.20 48.87
P. rapae
Control Red-unsplit Red-dr Red-50
2(2) 3(3) 3(3) 3(4)
2(2) 2(2) 2(2) 3(4)
Clear fluid present Red fluid present Red fluid present Red fluid present
**Significant differences (p < 0.0001) between a* color values of dissected crops.
0.82** 4.16* 1.81* 2.04*
Proboscis length (mm)
Drinking region length (mm)
18.19 18.61 17.98 16.90
0.89 0.25 1.00 0.21
1.59 1.72 1.65 1.62
0.08 0.05 0.03 0.09
10.75 11.21 11.02 10.47
0.49 0.24 0.14 0.23
0.72 0.82 0.82 0.73
0.01 0.04 0.03 0.04
100
M.S. Lehnert et al. / Arthropod Structure & Development 43 (2014) 97e102
Fig. 2. Evidence for functionality of previously split proboscides. (A) Proboscis (pr) of a P. rapae fed the red-sucrose solution, which was visible in the food canal (fc) through the transparent dorsal legulae. (B) Image of a section of the alimentary canal in an individual of P. p. asterius from the red-50 treatment obtained after a dissection that revealed the engorged red fluid-filled crop (cr), midgut (mg), and hindgut (hg).
asterius from the red-50 treatment that had a reassembled proboscis. All P. p. asterius fed the red-sucrose solution had an enlarged crop filled with red fluid (Fig. 2 B). Butterflies in the control treatment also had an enlarged fluid-filled crop, but it lacked a red fluid and was clear. Color quantification of the crops revealed no significant differences in a* (red) color values between treatments where butterflies were fed red-sucrose solution, indicating that all butterflies retained the ability to feed. These treatments, however, had significantly higher a* color values (i.e., were more red) (p < 0.0001) compared to butterflies in the control group (fed sucrose solution) (Table 1). Other components of the alimentary canal, such as the midgut and hindgut also appeared red. One individual defecated red fluid during the dissecting process. 3.3. Wettability of P. p. asterius proboscis Confocal microscopy of a proboscis of P. p. asterius revealed hydrophilic and hydrophobic structures (Fig. 3). The galea stained with Nile red exposed an overall hydrophobic galea with a microbump pattern. Only the dorsal legulae and chemo- and mechanosensilla were observed on the unstained galea, likely due to autofluorescence. 4. Discussion This study revealed two unreported components of proboscis functionality: 1) butterflies with previously split proboscides can retain the ability to feed, at least under laboratory conditions, and 2) butterflies might be able to partly reassemble their proboscis when split. We propose that proboscides of P. p. asterius that remained partially split after experimental splitting were able to still feed due to the wettability dynamics (Fig. 3, Lehnert et al.,
2013). The hydrophobic galeae might assist in channeling fluids to the hydrophilic food canal, which combined with the horizontal positioning of the proboscides in these studies, could have facilitated the movement of liquids to areas of the proboscis that remained together where stable liquid bridges could be formed for the sucking pump to act on (Monaenkova et al., 2012). Butterflies with proboscides that reassembled might retain the ability to form stable liquid bridges for fluid uptake, including the regions where splitting had occurred. These findings could be of importance in studies that use the butterfly proboscis as a model for the development of microfluidic devices (Tsai et al., 2011). This study indicated that proboscides of P. rapae are more likely to reassemble than those of P. p. asterius. Given the structural differences of the dorsal legulae between species, and their method of overlapping, we propose that the dorsal legulae might play an important role in proboscis reassembly (Fig. 1D,E). The zipper-like arrangement of dorsal legulae of P. rapae, which is lacking in the arrangement of dorsal legulae of P. p. asterius, could facilitate reassembly. Complete reassembly of the proboscis would require the hook-shaped ventral legulae (Fig. 1C) to interlink (Krenn, 1997), which was not examined in this study. Various factors might influence the partial reassembly of proboscides including the presence of saliva, which contains lubricative and enzymatic properties in insects (Terra, 1990), but requires further analysis in butterflies (Tokarev et al., 2013). Our observations here warrant further study of reassembly mechanisms of butterfly proboscides. Ancient lepidopteran lineages, such as the Eriocraniidae (Grimaldi and Engel, 2005; Krenn, 2010), possess a short proboscis with a structural arrangement that would support capillarity (e.g., interlegular spaces) (Monaenkova et al., 2012). Observations of Eriocraniidae have indicated that the galeae become separated while feeding on water (Kristensen, 1968), which suggests that capillarity and wettability might play an essential role in fluid
M.S. Lehnert et al. / Arthropod Structure & Development 43 (2014) 97e102
101
Fig. 3. Confocal microscopy images of structures and overall wettability of proboscides of P. p. asterius. (A) An unstained galea showing the food canal (fc), dorsal legulae (dl), chemosensilla (cs), and ventral legulae (vl). (B) An unstained galea (ga) shows the autofluorescence of the dorsal legulae and mechanosensilla (ms). (C) The other galea of the same proboscis shown in (B), but stained with Nile red, reveals the microbump (mb) pattern and overall hydrophobic galea (chemosensilla, cs, also have autofluorescence). All confocal images were acquired of the drinking region of the proboscis.
uptake processes in these Lepidoptera. Capillary action and wettability have adaptive value when feeding from exposed liquid films (Monaenkova et al., 2012; Lehnert et al., 2013), which were likely present before the availability of pools of nectar associated with the radiation of the Angiosperms; the coilable proboscis of Lepidoptera originated during the mid-Mesozoic (Labandeira, 2010). Retaining these physical properties for fluid uptake in more derived Lepidoptera facilitate feeding on liquid films (e.g., rotting fruit) (Monaenkova et al., 2012; Lehnert et al., 2013), and are beneficial if the proboscis is injured, as indicated in this study. This study suggests that wild butterflies with split or injured proboscides might be able to retain some functionality. It is unclear if butterflies with split proboscides have fluid uptake rates that significantly deviate from those with fully assembled proboscides however, this is an aspect of proboscis functionality worth exploring in future studies. A split proboscis might be a common occurrence in wild individuals. Lab-reared butterflies, for instance, often do not properly assemble their proboscis, possibly due to dry ambient conditions or an obstruction, such as a leg or maxillary palp, preventing assembly (personal observations). Injured mouthparts could be a common occurrence in other fluid-feeding insects important in pollination systems, such as flies (Diptera) and bees and wasps (Hymenoptera) (Barth, 1991; Willmer, 2011), where capillarity and wettability also might play an intricate role in fluid uptake.
Acknowledgments We thank Meredith Jenkins, Eric Brown, and Valerie Kramer for assisting in colony rearing, and Peter H. Adler, Shorook Attar, and Richard Harper for reading an early version of the manuscript. We also thank the Clemson Microscope Facility for assistance with SEM imaging. This work was partially supported by a National Science Foundation [EFRI 0937985]. References Adler, P.H., 1982. Soil- and puddle-visiting habits of moths. J. Lepid Soc. 36, 161e173. Barth, F.G., 1991. Insects and Flowers. Princeton University Press, Princeton, NJ. Bauder, J.A.-S., Handschuh, S., Metscher, B.D., Krenn, H.W., 2013. Functional morphology of the feeding apparatus and evolution of proboscis length in metalmark butterflies (Lepidoptera: Riodinidae). Biol. J. Linn. Soc. 110, 291e304. Cassie, A.B.D., Baxter, S., 1944. Wettability of porous surfaces. Trans. Faraday Soc. 40, 546e551. Eastham, L.E.S., Eassa, Y.E.E., 1955. The feeding mechanism of the butterfly Pieris brassicae L. Philos. Trans. R. Soc. Lond. B 239, 1e43. Eberhard, S.H., Krenn, H.W., 2003. Salivary glands and salivary pumps in adult Nymphalidae (Lepidoptera). Zoomorphology 122, 161e167. Foottit, R.G., Adler, P.H. (Eds.), 2009. Insect Biodiversity: Science and Society. John Wiley and Sons. Grant, J.I., Djani, D.M., Lehnert, M.S., 2012. Functionality of a reduced proboscis: fluid uptake by Phigalia strigataria (Minot) (Geometridae: Ennominae). J. Lepid Soc. 66, 211e215. Grimaldi, D., Engel, M.S., 2005. Evolution of the Insects. Cambridge University Press, New York, NY.
102
M.S. Lehnert et al. / Arthropod Structure & Development 43 (2014) 97e102
Kingsolver, J.G., Daniel, T.L., 1995. Mechanics of food handling by fluid-feeding insects. In: Chapman, R.F., de Boer, G. (Eds.), Regulatory Mechanisms in Insect Feeding. Chapman and Hall, New York, pp. 32e73. Krenn, H.W., 1997. Proboscis assembly in butterflies (Lepidoptera) - a once in a lifetime sequence of events. Eur. J. Entomol. 94, 495e502. Krenn, H.W., 2010. Feeding mechanisms of adult Lepidoptera: structure, function, and evolution of the mouthparts. Annu. Rev. Entomol. 55, 307e327. Krenn, H.W., Zulka, K.P., Gatschnegg, T., 2001. Proboscis morphology and food preferences in nymphalid butterflies (Lepidoptera: Nymphalidae). J. Zool. 254, 17e26. Krenn, H.W., Plant, J., Szucsich, N.U., 2005. Mouthparts of flower-visiting insects. Arthropod Struct. Dev. 34, 1e40. Kristensen, N.P., 1968. The morphology and functional evolution of the mouthparts in adult Lepidoptera. Opusc. Entomol. 33, 1e2. Labandeira, C.C., 2010. The pollination of mid Mesozoic seed plants and the early history of long-proboscid insects. Ann. Missouri Bot. Gard. 97, 469e513. Lehnert, M.S., Monaenkova, D., Andrukh, T., Beard, C.E., Adler, P.H., Kornev, K.G., 2013. Hydrophobicehydrophilic dichotomy of the butterfly proboscis. J. R. Soc. Interf. 10, 20130336. http://dx.doi.org/10.1098/rsif.2013.0336.
Lehnert, M.S., Balaban, M.O., Emmel, T.C., 2011. A new method for quantifying color of insects. Fl. Entomol. 94, 201e207. Monaenkova, D., Lehnert, M.S., Andrukh, T., Beard, C.E., Rubin, B., Tokarev, A., Lee, W.-K., Adler, P.H., Kornev, K.G., 2012. Butterfly proboscis: combining a drinking straw with a nanosponge facilitated diversification of feeding habits. J. R. Soc. Interf. 69, 720e726. Terra, W.R., 1990. Evolution of digestive systems in insects. Annu. Rev. Entomol. 35, 181e200. Tokarev, A., Kaufman, B., Gu, Y., Andrukh, T., Adler, P.H., Kornev, K.G., 2013. Probing viscosity of nanoliter droplets of butterfly saliva by magnetic rotational spectroscopy. Appl. Phys. Lett. 102. http://dx.doi.org/10.1063/1.4788927. Tsai, C.-C., Mikes, P., Andrukh, T., White, E., Monaenkova, D., Burtovyy, O., Burtovyy, R., Rubin, B., Lukas, D., Luzinov, I., Owens, J.R., Kornev, K.G., 2011. Nanoporous artificial proboscis for probing minute amount of liquids. Nanoscale 3, 4685e4695. Willmer, P., 2011. Pollination and Floral Ecology. Princeton Univ. Press, New Jersey.
Contents lists available at ScienceDirect
Arthropod Structure & Development journal homepage: www.elsevier.com/locate/asd
Mouthpart separation does not impede butterfly feeding Matthew S. Lehnert*, Catherine P. Mulvane, Aubrey Brothers Department of Biological Sciences, Kent State University at Stark, 6000 Frank Ave. NW, North Canton, OH 44720, USA
a r t i c l e i n f o
a b s t r a c t
Article history: Received 15 November 2013 Accepted 13 December 2013
The functionality of butterfly mouthparts (proboscis) plays an important role in pollination systems, which is driven by the reward of nectar. Proboscis functionality has been assumed to require action of the sucking pump in the butterfly’s head coupled with the straw-like structure. Proper proboscis functionality, however, also is dependent on capillarity and wettability dynamics that facilitate acquisition of liquid films from porous substrates. Due to the importance of wettability dynamics in proboscis functionality, we hypothesized that proboscides of eastern black swallowtail (Papilio polyxenes asterius Stoll) (Papilionidae) and cabbage butterflies (Pieris rapae Linnaeus) (Pieridae) that were experimentally split (i.e., proboscides no longer resembling a sealed straw-like tube) would retain the ability to feed. Proboscides were split either in the drinking region (distal 6e10% of proboscis length) or approximately 50% of the proboscis length 24 h before feeding trials when butterflies were fed a red food-coloring solution. Approximately 67% of the butterflies with proboscides split reassembled prior to the feeding trials and all of these butterflies displayed evidence of proboscis functionality. Butterflies with proboscides that did not reassemble also demonstrated fluid uptake capabilities, thus suggesting that wild butterflies might retain fluid uptake capabilities, even when the proboscis is partially injured. Ó 2013 Elsevier Ltd. All rights reserved.
Keywords: Proboscis Functionality Lepidoptera Wettability Pollination Fluid uptake
1. Introduction Mouthpart functionality of fluid-feeding insects e more than half of all known animal species (Foottit and Adler, 2009) e is an important component of disease transmission and the stability of insect-pollination systems (Kingsolver and Daniel, 1995). Mouthparts of fluid-feeding insects, such as butterflies and moths (Lepidoptera), might be subjected to damage while seeking mates, searching for food, or from predator encounters. Mouthparts rendered nonfunctional, therefore, could affect fitness (Krenn, 1997) and impact insecteflower interactions. Most Lepidoptera have a coilable, tube-like proboscis that transports fluids, such as nectar, sap, fruit juices, and blood (Adler, 1982) to the insect’s gut. The lepidopteran proboscis is composed of two elongated maxillary galeae that are connected by overlapping dorsal and interlinking ventral structures (i.e., legulae) to form a food canal (Eastham and Eassa, 1955; Krenn et al., 2005; Krenn, 2010). The distal 5e20% of the proboscis has dorsal legulae that are elongated and more widely spaced (Krenn et al., 2001) (i.e., the drinking region), which facilitates fluid uptake (Lehnert et al., 2013). The merging of the galeae into a functional proboscis takes place after adult eclosion from the pupa, and consists of coiling and * Corresponding author. Tel.: þ1 (330) 244 3349. E-mail addresses: [email protected] (M.S. Lehnert), [email protected] (C. P. Mulvane), [email protected] (A. Brothers). 1467-8039/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.asd.2013.12.005
uncoiling actions of the proboscis accompanied by the presence of saliva droplets (Krenn, 1997). Proboscis assembly must occur before sclerotization of the legular cuticle, otherwise the proboscis is putatively nonfunctional and reassembly cannot occur (Krenn, 1997). A functional proboscis is widely considered a sealed tube that operates similar to a drinking straw (Krenn, 2010; Bauder et al., 2013), solely relying on the sucking pump in the head for fluid uptake (Kingsolver and Daniel, 1995; Eberhard and Krenn, 2003); however, recent experiments have demonstrated that aqueous solutions can enter between dorsal interlegular spaces along the proboscis (i.e., not a sealed tube) (Monaenkova et al., 2012) and that a straw-like structure is not necessary for functionality (Grant et al., 2012). The proboscis employs capillarity via interlegular spaces to build liquid bridges in the food canal for the sucking pump to act on when feeding from liquid films and porous substrates (Monaenkova et al., 2012), such as rotting fruit. Fluid uptake is further regulated by wettability dynamics (i.e., hydrophilicity and hydrophobicity) of proboscis structures (e.g., hydrophilic dorsal legulae, chemosensilla, and the food canal) and surface roughness (e.g. microbumps that create an overall hydrophobic surface, explained using the Cassie-Baxter model, Cassie and Baxter, 1944; Lehnert et al., 2013). Based on our current understanding of the multifaceted fluid uptake system of butterfly proboscides we hypothesized that previously assembled proboscides of two distantly related nectar-feeding butterfly species, the eastern black swallowtail, Papilio polyxenes asterius (Papilionidae), and cabbage
98
M.S. Lehnert et al. / Arthropod Structure & Development 43 (2014) 97e102
butterfly, Pieris rapae (Pieridae), will maintain functionality following experimental splitting of the galeae as long as both galeae are subsequently placed in a feeding solution. 2. Materials and methods 2.1. Butterfly rearing and proboscis measurements Eggs of P. p. asterius were obtained from a female captured in North Canton, OH. The larvae were reared on parsley (Petroselinum crispum) and kept in RubbermaidÒ Takealong containers. Larvae of P. rapae were obtained from Carolina Biological Supply Company (Burlington, NC, USA) and reared on artificial diet. Larvae and pupae of both species of butterflies were maintained at 22 C, 61% relative humidity (r.h.), and an L18:D6 photoperiod in an environmental chamber (Percival Scientific, Inc., Perry, IN, USA). Adults from an F2 generation of P. p. asterius also were used for experiments. All adults were fed a dilute honey:water solution (1:5) daily for at least three days before feeding experiments and kept in glassine envelopes in a refrigerator (4 C) between feeding times. Proboscis lengths and drinking region lengths were measured to determine possible effects on functionality between treatments. In order to acquire proboscis measurements, butterflies were stabilized on a piece of Styrofoam and proboscides were uncoiled using insect pins. Images of the total length of proboscides (0.78 magnification) and drinking regions (4.0 magnification) were acquired for each butterfly with a Leica M205 C stereomicroscope and an IC 80HD camera (Heerbrugg, Switzerland) and measured using ImageJ software (http://rsbweb.nih.gov/ij/). The drinking
region was measured from the tip of the proboscis to a transition point where the dorsal legulae narrow and remain similar in width for the remainder of the proboscis length (Fig. 1A). Although wettability dynamics of proboscides have been reported for other butterfly species (Monaenkova et al., 2012; Lehnert et al., 2013), we demonstrated these dynamics using the proboscis of an individual P. p. asterius. The galeae were split and one galea (unstained) was placed on a slide in dH2O with a coverslip and imaged using an Olympus Confocal Microscope IX81 with DSU (Center Valley, PA, USA) (999.6 ms exposure, 20 magnification, 30 slices, 1.60 average depth slice, CY3 channel). The other galea was stained with Nile red for 24 h and imaged similarly to distinguish hydrophilic and hydrophobic structures. Proboscides of P. rapae were dehydrated in an ethanol series (80%, 90%, 100%, 24 h each), air-dried with hexamethyldisalizane, platinum sputter-coated for approximately 1 min, and imaged with a Hitachi TM3000 scanning electron microscope (Hitachi High Technologies America, Inc., Dallas, TX, USA). 2.2. Experimental feeding trials All butterflies were fed a 20% sucrose solution and kept at room temperature (24 C, 61% r.h.) in a netted Bug Dorm (BioQuip Products, Rancho Dominguez, CA, USA) 24 h prior to feeding experiments. Randomly selected butterflies were prepped for the experimental feeding trials by separating the two galeae either in the drinking region using an insect pin or had approximately 50% of their proboscis separated proximally starting at the tip (inset in Fig. 1A) immediately after being fed the 20% sucrose solution. Before proboscides were split, all butterflies had their proboscides
Fig. 1. Proboscis assembly and fluid uptake of split proboscides of butterflies. (A) Stereomicroscope image of an uncoiled proboscis of P. p. asterius displaying the galeae (ga) and overlapping dorsal legulae (dl). The dorsal legulae are larger and more widely spaced in the drinking region; the remainder of the proboscis represents the nondrinking region. The inset shows a proboscis of P. p. asterius split with insect pins (ip) for the red-50 treatment. (B) Photograph of a P. p. asterius obtained shortly after emergence showing the partially assembled proboscis (separated galeae) during the assembly process. (C) SEM image of a single galea of P. rapae showing the food canal (fc) and dorsal (dl) and ventral legulae (vl) that interlink during proboscis assembly. (D) Stereomicroscope image of the dorsal legulae of P. p. asterius in the nondrinking region; there is little overlap of the dorsal legulae. (E) SEM image of a proboscis of P. rapae showing the overlapping dorsal legulae in the nondrinking region and microbumps (mb). The arrangement of the dorsal legulae differs between P. p. asterius and P. rapae.
M.S. Lehnert et al. / Arthropod Structure & Development 43 (2014) 97e102
investigated by prodding the proboscis with an insect pin to ensure the galeae were assembled (Fig. 1B shows a proboscis during the assembly process after eclosion from the pupa). Butterflies were assigned into a control group or one of three experimental treatments and then fed solutions for 10e13 min by manually placing only the drinking region of both galeae into a 100 ml capillary tube 24 h after experimental splitting. The diameter of the capillary tube was larger than the width of both galeae in the drinking region, therefore, the tube had little effect on pressing the galeae together, thus avoiding simulating a sealed proboscis. The control group consisted of butterflies with an unsplit proboscis fed an aqueous 20% sucrose solution. The three treatments consisted of: 1) butterflies with unsplit proboscides fed a 20% sucrose and redfood coloring mixture, referred to hereafter as red-sucrose solution (i.e., red-unsplit treatment), 2) butterflies with proboscides split at the drinking region fed the red-sucrose solution (i.e., red-dr treatment), and 3) butterflies with approximately 50% of the proboscis split and fed the red-sucrose solution (i.e., red-50 treatment). Due to the dark pigmentation of the proboscis of P. p. asterius, feeding ability was determined by dissecting the butterflies in PBS physiological saline solution (7.2 pH) using insect pins and dissecting scissors within 1 h after the feeding trials (butterflies were temporarily stored in 4 C refrigerator before dissections). The crop of the alimentary canal was isolated and the average a* color values (a* ¼ green to red, scale 120 to 120, L*, a*, and b* color values) were acquired using LensEye color analysis software (Lehnert et al., 2011). Dissections of P. rapae specimens were not necessary due to transparency of the dorsal legulae of the proboscis and the ability to visualize fluid flow in the food canal. 2.3. Data analysis Significant differences (p < 0.05) in proboscis length, drinking region length, percentage of proboscis length represented by the drinking region (i.e., percentage drinking length), and a* color values were tested between sexes (t-test) and between treatments (analysis of variance) for P. p. asterius. Significant differences in mean a* color values were ranked between treatments using Fisher’s least significant difference (LSD) test. Due to differences in sample sizes between treatments with P. rapae, an independent sample Kruskale Wallis test was used to determine significant differences (p < 0.05) in proboscis length, drinking region length, and percentage drinking length. All statistical analyses were processed in SPSS Statistic 21.0. 3. Results 3.1. Evidence for proboscis reassembly All butterflies with experimentally split proboscides were checked for reassembly by applying pressure to the dorsal legulae
99
in previously split regions with the tip of an insect pin before being fed. All butterflies that had the galeae separated in the drinking region (i.e., red-dr treatment, Table 1) had proboscides that appeared reassembled before feeding, except for one individual of P. p. asterius, which was observed to have a fluid, likely saliva, present between the separated galeae of the drinking region. An individual of P. rapae died before we could check for proboscis functionality; however, inspection revealed a reassembled drinking region. It is unknown if reassembly of the drinking region involved complete interlinking of the ventral legulae; only the dorsal legulae were inspected. Based on visual inspection, all P. p. asterius in the red-50 treatment (n ¼ 4) had reassembled proboscides however, when prodded with an insect pin only one remained assembled, the other three proboscides had galeae that partially separated where pressure was applied. Three of the four P. rapae in the red-50 treatment remained reassembled after pressure was applied with an insect pin (Table 1). The individual without a reassembled proboscis had the galeae almost completely separated (over 90%), indicating further separation after our experimental splitting, and one of the galeae was coiled and appeared dry, which prevented us from being able to insert both galeae into the capillary tube to test proboscis functionality. There were no significant differences in proboscis length (mean s.e.m., 17.92 0.34 mm), drinking region length (1.64 0.03 mm), or percentage drinking region (9.21 0.23%) between treatments or sexes of P. p. asterius (n ¼ 4 per treatment) or P. rapae (proboscis length 10.84 0.14 mm, drinking region length 0.77 0.02 mm, percentage drinking region 7.11 0.18%; n ¼ 12), suggesting that the differences observed between treatments were not influenced by these structural measurements (Table 1).
3.2. Feeding ability of split proboscides All individuals that had the drinking region of both galeae inserted into the capillary tube were observed to feed (suggested by the depletion of solution in the tube), except one individual of P. rapae from the unsplit-red treatment. All individuals of P. rapae with proboscides that reassembled and fed the red-sucrose solution were observed to have a red fluid traveling through the food canal, which was visible through the semi-transparent dorsal legulae (Fig. 2A); butterflies in the control group were observed to have clear fluids traveling through the proboscis. Three individuals of P. p. asterius from each of the red-unsplit and red-50 treatment had a red droplet appear on the dorsum of the dorsal legulae at proximal regions of the proboscis (i.e., nondrinking region) while feeding (two simultaneous droplets on one individual); the droplets exhibited a pulsing-like motion. In addition, anti-parallel movements of the galeae were observed on an individual of P. p.
Table 1 Measurements (mean s.e.m.) of proboscides and confirmation of mouthpart functionality of butterfly species. All butterflies in each treatment were checked for proboscis reassembly and feeding ability after experimental splitting. The numbers of individuals that reassembled their proboscis and fed on solutions are followed by the total sample size for that treatment in parentheses. Mouthpart functionality was determined by color quantification of the crop (a* values) or visual observation of fluids in the proboscis. Species
Treatment
Reassembled proboscides
Individuals that fed
Mouthpart functionality (* ¼ a* color values)
P. p. asterius
Control Red-unsplit Red-dr Red-50
4(4) 4(4) 3(4) 1(4)
4(4) 4(4) 3(4) 1(4)
31.83 53.46 48.20 48.87
P. rapae
Control Red-unsplit Red-dr Red-50
2(2) 3(3) 3(3) 3(4)
2(2) 2(2) 2(2) 3(4)
Clear fluid present Red fluid present Red fluid present Red fluid present
**Significant differences (p < 0.0001) between a* color values of dissected crops.
0.82** 4.16* 1.81* 2.04*
Proboscis length (mm)
Drinking region length (mm)
18.19 18.61 17.98 16.90
0.89 0.25 1.00 0.21
1.59 1.72 1.65 1.62
0.08 0.05 0.03 0.09
10.75 11.21 11.02 10.47
0.49 0.24 0.14 0.23
0.72 0.82 0.82 0.73
0.01 0.04 0.03 0.04
100
M.S. Lehnert et al. / Arthropod Structure & Development 43 (2014) 97e102
Fig. 2. Evidence for functionality of previously split proboscides. (A) Proboscis (pr) of a P. rapae fed the red-sucrose solution, which was visible in the food canal (fc) through the transparent dorsal legulae. (B) Image of a section of the alimentary canal in an individual of P. p. asterius from the red-50 treatment obtained after a dissection that revealed the engorged red fluid-filled crop (cr), midgut (mg), and hindgut (hg).
asterius from the red-50 treatment that had a reassembled proboscis. All P. p. asterius fed the red-sucrose solution had an enlarged crop filled with red fluid (Fig. 2 B). Butterflies in the control treatment also had an enlarged fluid-filled crop, but it lacked a red fluid and was clear. Color quantification of the crops revealed no significant differences in a* (red) color values between treatments where butterflies were fed red-sucrose solution, indicating that all butterflies retained the ability to feed. These treatments, however, had significantly higher a* color values (i.e., were more red) (p < 0.0001) compared to butterflies in the control group (fed sucrose solution) (Table 1). Other components of the alimentary canal, such as the midgut and hindgut also appeared red. One individual defecated red fluid during the dissecting process. 3.3. Wettability of P. p. asterius proboscis Confocal microscopy of a proboscis of P. p. asterius revealed hydrophilic and hydrophobic structures (Fig. 3). The galea stained with Nile red exposed an overall hydrophobic galea with a microbump pattern. Only the dorsal legulae and chemo- and mechanosensilla were observed on the unstained galea, likely due to autofluorescence. 4. Discussion This study revealed two unreported components of proboscis functionality: 1) butterflies with previously split proboscides can retain the ability to feed, at least under laboratory conditions, and 2) butterflies might be able to partly reassemble their proboscis when split. We propose that proboscides of P. p. asterius that remained partially split after experimental splitting were able to still feed due to the wettability dynamics (Fig. 3, Lehnert et al.,
2013). The hydrophobic galeae might assist in channeling fluids to the hydrophilic food canal, which combined with the horizontal positioning of the proboscides in these studies, could have facilitated the movement of liquids to areas of the proboscis that remained together where stable liquid bridges could be formed for the sucking pump to act on (Monaenkova et al., 2012). Butterflies with proboscides that reassembled might retain the ability to form stable liquid bridges for fluid uptake, including the regions where splitting had occurred. These findings could be of importance in studies that use the butterfly proboscis as a model for the development of microfluidic devices (Tsai et al., 2011). This study indicated that proboscides of P. rapae are more likely to reassemble than those of P. p. asterius. Given the structural differences of the dorsal legulae between species, and their method of overlapping, we propose that the dorsal legulae might play an important role in proboscis reassembly (Fig. 1D,E). The zipper-like arrangement of dorsal legulae of P. rapae, which is lacking in the arrangement of dorsal legulae of P. p. asterius, could facilitate reassembly. Complete reassembly of the proboscis would require the hook-shaped ventral legulae (Fig. 1C) to interlink (Krenn, 1997), which was not examined in this study. Various factors might influence the partial reassembly of proboscides including the presence of saliva, which contains lubricative and enzymatic properties in insects (Terra, 1990), but requires further analysis in butterflies (Tokarev et al., 2013). Our observations here warrant further study of reassembly mechanisms of butterfly proboscides. Ancient lepidopteran lineages, such as the Eriocraniidae (Grimaldi and Engel, 2005; Krenn, 2010), possess a short proboscis with a structural arrangement that would support capillarity (e.g., interlegular spaces) (Monaenkova et al., 2012). Observations of Eriocraniidae have indicated that the galeae become separated while feeding on water (Kristensen, 1968), which suggests that capillarity and wettability might play an essential role in fluid
M.S. Lehnert et al. / Arthropod Structure & Development 43 (2014) 97e102
101
Fig. 3. Confocal microscopy images of structures and overall wettability of proboscides of P. p. asterius. (A) An unstained galea showing the food canal (fc), dorsal legulae (dl), chemosensilla (cs), and ventral legulae (vl). (B) An unstained galea (ga) shows the autofluorescence of the dorsal legulae and mechanosensilla (ms). (C) The other galea of the same proboscis shown in (B), but stained with Nile red, reveals the microbump (mb) pattern and overall hydrophobic galea (chemosensilla, cs, also have autofluorescence). All confocal images were acquired of the drinking region of the proboscis.
uptake processes in these Lepidoptera. Capillary action and wettability have adaptive value when feeding from exposed liquid films (Monaenkova et al., 2012; Lehnert et al., 2013), which were likely present before the availability of pools of nectar associated with the radiation of the Angiosperms; the coilable proboscis of Lepidoptera originated during the mid-Mesozoic (Labandeira, 2010). Retaining these physical properties for fluid uptake in more derived Lepidoptera facilitate feeding on liquid films (e.g., rotting fruit) (Monaenkova et al., 2012; Lehnert et al., 2013), and are beneficial if the proboscis is injured, as indicated in this study. This study suggests that wild butterflies with split or injured proboscides might be able to retain some functionality. It is unclear if butterflies with split proboscides have fluid uptake rates that significantly deviate from those with fully assembled proboscides however, this is an aspect of proboscis functionality worth exploring in future studies. A split proboscis might be a common occurrence in wild individuals. Lab-reared butterflies, for instance, often do not properly assemble their proboscis, possibly due to dry ambient conditions or an obstruction, such as a leg or maxillary palp, preventing assembly (personal observations). Injured mouthparts could be a common occurrence in other fluid-feeding insects important in pollination systems, such as flies (Diptera) and bees and wasps (Hymenoptera) (Barth, 1991; Willmer, 2011), where capillarity and wettability also might play an intricate role in fluid uptake.
Acknowledgments We thank Meredith Jenkins, Eric Brown, and Valerie Kramer for assisting in colony rearing, and Peter H. Adler, Shorook Attar, and Richard Harper for reading an early version of the manuscript. We also thank the Clemson Microscope Facility for assistance with SEM imaging. This work was partially supported by a National Science Foundation [EFRI 0937985]. References Adler, P.H., 1982. Soil- and puddle-visiting habits of moths. J. Lepid Soc. 36, 161e173. Barth, F.G., 1991. Insects and Flowers. Princeton University Press, Princeton, NJ. Bauder, J.A.-S., Handschuh, S., Metscher, B.D., Krenn, H.W., 2013. Functional morphology of the feeding apparatus and evolution of proboscis length in metalmark butterflies (Lepidoptera: Riodinidae). Biol. J. Linn. Soc. 110, 291e304. Cassie, A.B.D., Baxter, S., 1944. Wettability of porous surfaces. Trans. Faraday Soc. 40, 546e551. Eastham, L.E.S., Eassa, Y.E.E., 1955. The feeding mechanism of the butterfly Pieris brassicae L. Philos. Trans. R. Soc. Lond. B 239, 1e43. Eberhard, S.H., Krenn, H.W., 2003. Salivary glands and salivary pumps in adult Nymphalidae (Lepidoptera). Zoomorphology 122, 161e167. Foottit, R.G., Adler, P.H. (Eds.), 2009. Insect Biodiversity: Science and Society. John Wiley and Sons. Grant, J.I., Djani, D.M., Lehnert, M.S., 2012. Functionality of a reduced proboscis: fluid uptake by Phigalia strigataria (Minot) (Geometridae: Ennominae). J. Lepid Soc. 66, 211e215. Grimaldi, D., Engel, M.S., 2005. Evolution of the Insects. Cambridge University Press, New York, NY.
102
M.S. Lehnert et al. / Arthropod Structure & Development 43 (2014) 97e102
Kingsolver, J.G., Daniel, T.L., 1995. Mechanics of food handling by fluid-feeding insects. In: Chapman, R.F., de Boer, G. (Eds.), Regulatory Mechanisms in Insect Feeding. Chapman and Hall, New York, pp. 32e73. Krenn, H.W., 1997. Proboscis assembly in butterflies (Lepidoptera) - a once in a lifetime sequence of events. Eur. J. Entomol. 94, 495e502. Krenn, H.W., 2010. Feeding mechanisms of adult Lepidoptera: structure, function, and evolution of the mouthparts. Annu. Rev. Entomol. 55, 307e327. Krenn, H.W., Zulka, K.P., Gatschnegg, T., 2001. Proboscis morphology and food preferences in nymphalid butterflies (Lepidoptera: Nymphalidae). J. Zool. 254, 17e26. Krenn, H.W., Plant, J., Szucsich, N.U., 2005. Mouthparts of flower-visiting insects. Arthropod Struct. Dev. 34, 1e40. Kristensen, N.P., 1968. The morphology and functional evolution of the mouthparts in adult Lepidoptera. Opusc. Entomol. 33, 1e2. Labandeira, C.C., 2010. The pollination of mid Mesozoic seed plants and the early history of long-proboscid insects. Ann. Missouri Bot. Gard. 97, 469e513. Lehnert, M.S., Monaenkova, D., Andrukh, T., Beard, C.E., Adler, P.H., Kornev, K.G., 2013. Hydrophobicehydrophilic dichotomy of the butterfly proboscis. J. R. Soc. Interf. 10, 20130336. http://dx.doi.org/10.1098/rsif.2013.0336.
Lehnert, M.S., Balaban, M.O., Emmel, T.C., 2011. A new method for quantifying color of insects. Fl. Entomol. 94, 201e207. Monaenkova, D., Lehnert, M.S., Andrukh, T., Beard, C.E., Rubin, B., Tokarev, A., Lee, W.-K., Adler, P.H., Kornev, K.G., 2012. Butterfly proboscis: combining a drinking straw with a nanosponge facilitated diversification of feeding habits. J. R. Soc. Interf. 69, 720e726. Terra, W.R., 1990. Evolution of digestive systems in insects. Annu. Rev. Entomol. 35, 181e200. Tokarev, A., Kaufman, B., Gu, Y., Andrukh, T., Adler, P.H., Kornev, K.G., 2013. Probing viscosity of nanoliter droplets of butterfly saliva by magnetic rotational spectroscopy. Appl. Phys. Lett. 102. http://dx.doi.org/10.1063/1.4788927. Tsai, C.-C., Mikes, P., Andrukh, T., White, E., Monaenkova, D., Burtovyy, O., Burtovyy, R., Rubin, B., Lukas, D., Luzinov, I., Owens, J.R., Kornev, K.G., 2011. Nanoporous artificial proboscis for probing minute amount of liquids. Nanoscale 3, 4685e4695. Willmer, P., 2011. Pollination and Floral Ecology. Princeton Univ. Press, New Jersey.
