JOURNAL OF MORPHOLOGY 270:1348–1355 (2009)

Surviving Submerged—Setal Tracheal Gills for Gas Exchange in Adult Rheophilic Diving

Siegfried Kehl* and Konrad Dettner

Department of Ecology II, University of Bayreuth, 95440 Bayreuth,

ABSTRACT The gas exchange in adult diving beetles 2002). Adult (predaceous diving beetles) (Coleoptera: Dytiscidae) relies on a subelytral air store, are often used as a prime example of aquatic which has to be renewed in regular intervals at the having a transportable air store for respi- water surface. The dive duration varies from a few ration. This air store in the subelytral cavity, must minutes to 24 h depending on the species, activity, and be renewed at regular intervals at the water sur- temperature. However, some species remain submerged for several weeks. Stygobiont species do not ascend to face (Alt, 1912; Wesenberg-Lund, 1912; Ege, 1915; the surface and gas exchange of these species remains Rahn and Paganelli, 1968; Gilbert, 1986; Wichard unclear, but it is assumed that they require air filled et al., 2002; Balke, 2005). The spiracles of the voids for respiration or they use cutaneous respiration. adult beetles open into the subelytral cavity (Gil- In this study, we investigate the gas exchange in the bert, 1986). For gas exchange, adult Dytiscidae running water diving Deronectes aubei, which break the water surface with the tip of the abdo- survive submerged for over 6 weeks. The diffusion dis- men and atmospheric air enters the subelytral cav- tance through the cuticle is too great for cutaneous res- ity and flows into the emptied tracheae. During piration. Therefore, the dissolved oxygen uptake of sub- diving most Dytiscidae often press an air bubble merged beetles was determined and an oxygen uptake out of the subelytral cavity which functions as a via the rich tracheated elytra was observed. Fine struc- ture analyses (SEM and TEM) of the beetles showed physical gill or, more precisely, as a compressible tracheated setae mainly on the elytral surface, which gas gill (Rahn and Paganelli, 1968). The dive dura- acts as tracheal gills. Prevention of the air bubble forma- tion varies from a few minutes to 24 h depending tion at the tip of the abdomen, which normally act as on species, activity, and water temperature physical gill in Dytiscidae, resulted in no effect in oxy- (Madsen, 1967; Calosi et al., 2007). For example, gen uptake in D. aubei, but this was the sole way for a palustris can remain three hours sub- submerged to get oxygen. The merged at 208C water temperature before it shows setal gas exchange technique explains the restriction of signs of oxygen deficiency (Gilbert, 1986). How- D. aubei to rivers and brooks with high oxygen concen- ever, some species remain submerged for very long tration and it may also be used by subterran living div- periods. Hydroglyphus hamulatus remains sub- ing beetles, which lack access to atmospheric oxygen. The existence of setal tracheal gills in species in running merged in an oxygen-rich lake for 10 weeks water which are often found in the hyporheic zone and (Meuche, 1937) or sanmarkii, Nebriopo- in stygobiont species supports the known evolution of rus depressus, and Stictotarsus duodecimpustula- stygobiont Dytiscidae from species of the hyporheic tus for more than 35 days (Madsen, 2007). Stygo- zone. For species in running water, setal tracheal gills biont species (species of groundwater) do not could be seen as an adaptation to avoid drifting down- ascend to the surface for gas exchange (Castro and stream by the current. J. Morphol. 270:1348–1355, Delgado, 2001) and the respiration of subterranean 2009. Ó 2009 Wiley-Liss, Inc. water beetles is still not understood. Some authors assume subterranean dytiscids use air filled voids KEY WORDS: Dytiscidae; elytra; oxygen uptake; for respiration (e.g., Spangler and Decu, 1999; respiration; tracheation Cooper et al., 2007) and others propose cutaneous respiration (Ue´no, 1957; Ordish, 1976; Smrzˇ,

INTRODUCTION

Various taxa returned from terrestrial to *Correspondence to: Siegfried Kehl, University of Bayreuth, Ani- aquatic environments secondarily and evolved var- mal Ecology II, Universita¨tsstr. 30, 95440 Bayreuth, Germany. ious adaptations for gas exchange independently E-mail: [email protected] from each other. The tracheal system of aquatic insects can be either closed or open and insects can Received 16 February 2009; Accepted 10 April 2009 use oxygen dissolved in water or atmospheric air Published online 28 May 2009 in for respiration (Resh and Solem, 1984; Pritchard Wiley InterScience (www.interscience.wiley.com) et al., 1993; Eriksen et al., 1996; Wichard et al., DOI: 10.1002/jmor.10762

Ó 2009 WILEY-LISS, INC. SETAL TRACHEAL GILLS IN DYTISCIDAE 1349 1981). Smrzˇ (1981) suggested gas exchange via the carefully spread over the elytra with a small brush, assuring intraelytral tracheae in subterranean species, that the legs and head were not covered. In the closed system, untreated specimens were first measured, then those with cov- because the elytra of stygobiont diving beetles are ered elytra and afterwards the synthetic resin was removed richly tracheated and the longitudinal tracheae and a control was measured again. Because of the low level of have a greater relative diameter compared to most oxygen uptake, two beetles were measured simultaneously. Af- epigeic species. Madsen (2007) described small air ter the beetles were placed in the respiration chamber, the oxy- bubbles on the elytral surface of submerged run- gen saturation was measured in intervals to avoid a down- wardly drifting signal due to oxygen consumption by the elec- ning water Dytiscidae and suggested an exchange trode. The oxygen saturation was measured for 60 min every 10 with the tracheal system of the elytra. The rheo- min for 1 min with an x-t chart recorder and the oxygen uptake philious diving beetle Deronectes aubei (Mulsant) was calculated over the last 30 min. A 1-mm gauze fixed with a was observed under laboratory conditions in oxy- rubber ring in the respiration chamber separated the beetles from the rotating stir bar. gen saturated water at 138C, it remained sub- In the flow through respirometer, oxygen uptake of D. aubei merged for more than 6 weeks without the access and H. palustris (Linnaeus) was measured under constant oxy- to atmospheric oxygen. This species normally lives gen saturation. The flow of the flow-through respirometer, a in fast running waters in mountainous and alpine glass tube (150 mm length and 16 mm diameter) connected to a regions under gravel and stones (Fery and Bran- water tank (modified from Hanke et al., 1980), was adjusted to 2 ml/min. Control and test water were collected in bottles and cucci, 1997) and sometimes deep in the streambed. the oxygen content was measured by Winkler titration accord- To investigate the gas exchange of D. aubei and a ing to the German standard protocol (DIN 38408, part 21, possible oxygen uptake via intraelytral tracheae, 1984). Oxygen uptake was calculated by the difference of oxy- oxygen uptake was measured with a flow through gen concentration of control and respiration chamber water. Because of the low levels of oxygen uptake groups of nine speci- and closed respirometer under laboratory condi- mens of D. aubei and 10 specimens of H. palustris were meas- tions. In addition, the fine structures of the elytra ured for  1 h. The mean oxygen uptake was calculated for a were examined with scanning electron microscopy single specimen in mg/h. For D. aubei, four series of oxygen (SEM) and transmission electron microscopy uptake measurements were conducted: oxygen uptake of (TEM) techniques. Comparison of the relative di- untreated specimens, oxygen uptake of specimens where the physical gill was prevented by covering the tip of the abdomen ameter of the intraelytral longitudinal tracheae of with synthetic resin, oxygen uptake of specimens with synthetic adephagan water beetles of different habitats pro- resin covered elytra, and a final control measurement where vided information about the occurrence of cutane- the synthetic resin was removed from the elytra. Individuals ous respiration in aquatic beetles. were placed in the respiration chamber for 1 h before respira- tion water was collected. This was not possible in H. palustris because the activity of the beetles and also the frequency of using the physical gill decrease with duration of the experi- MATERIALS AND METHODS ment, so that the respiration water was collected for 1 h, start- Electron Microscopy ing 10 min after placing the beetles in the respiration chamber. Therefore, for H. palustris, a first and second measurement of For SEM, elytra were dehydrated in ethanol series, air dried, untreated specimens was conducted without opening the respi- mounted on stubs, and sputter-coated with gold. The samples ration chamber. Afterwards, the beetles were measured with a were viewed and photographed using a Philips SEM (FEI XL covered pygidium and with covered elytra. 30 ESEM). For TEM, the elytra were fixed in 2.5% glutaraldehyde in 0.1 mol/l cacodylate buffer (pH 7.3) for 1 h, embedded in 2% aga- Relative Diameter of Longitudinal Tracheae rose and fixed again in 2.5% glutaraldehyde in 0.1 mol/l caco- The diameter of the intraelytral longitudinal tracheae was dylate buffer overnight. Elytra were washed three times in 0.1 measured using a microscope (4003 magnification) for 45 ade- mol/l cacodylate buffer for 20 min and osmicated in 2% osmium phagan water beetles of different habitats. In some species, the tetroxide for 2 h. Elytra were washed and stained in 2% uranyl elytra were bleached with potassium hydroxide. The maximum acetate for 90 min, dehydrated in an ethanol series (30, 50, 70, determined diameter of longitudinal tracheae and the length of 95, and 3 3 100%), transferred to propylene oxide and embed- the elytra of the respective species were used to calculate the ded in Epon 812 (Serva, Heidelberg, Germany). Ultrathin sec- relative diameter of the longitudinal trachea (Drel 5 maximum tions (70 nm) were cut with a diamond knife (MicroStar, Hunts- diameter of longitudinal trachea/length of elytra). The type of ville, TX) on a Leica Ultracut UCT microtome (Leica Microsys- tracheation was grouped into three categories. Type 1: small tems, Vienna, ). Thin sections were mounted on longitudinal tracheae with few or no visible branching. Type 2: Pioloform-coated copper grids (Plano, Wetzlar, Germany) and large longitudinal tracheae with strong branching. Type 3: com- stained with saturated uranyl acetate, followed by lead citrate. pletely different tracheation. Sections were viewed with a Zeiss CEM 902 A TEM (Carl Zeiss, Oberkochen, Germany) at 80 kV. Micrographs were taken using SO-163 EM film (Eastman Kodak, Rochester, NY). RESULTS To test oxygen uptake via the elytra, we meas- Oxygen Uptake Measurements ured the dissolved oxygen uptake of D. aubei in a Oxygen uptake was determined in a Rank Digital Oxygen repeated measurements design (n 5 5) with and System, Model 10 (Rank Brothers, Bottisham, Cambridge, UK) without resin covered elytra of the beetles in a and in a flow through respirometer (Hanke et al., 1980). All ox- closed system, where the beetles had no access to ygen uptake measurements were conducted in a climate cham- atmospheric oxygen. The oxygen uptake of D. ber at 138C. To test oxygen uptake via the intraelytral tracheae, the elytra were covered with a film of synthetic resin dispersion aubei was significant higher in untreated beetles (Liquid Frisket, Rubbelkrepp, Schmincke, Erkrath, Germany). than in beetles with synthetic resin covered elytra. To enable the mobility of the beetles, the synthetic resin was In specimens where the synthetic resin had been

Journal of Morphology 1350 S. KEHL AND K. DETTNER For a constant oxygen saturation of the water, a flow through respirometer was used. The oxygen uptake of D. aubei of untreated beetles and control group was also significantly higher than in beetles with covered elytra (Fig. 2a). Furthermore, no dif- ferences was found between beetles with a covered pygidium and the untreated and control group. In H. palustris, a species from stagnant water, the resin cover of the pygidium reduced oxygen uptake significantly (Fig. 2b), but covering the elytra had no significant effect. Oxygen uptake for a sub- merged H. palustris is solely a function of the physical gill. The surface of the elytra of D. aubei is covered with three different types of setae (Fig. 3a–d) as described by Wolfe and Zimmermann (1984) for elytra of : long sensory setae in coarse punctures encircled by concentric ridges (sensillum trichoideum Type 1, Fig. 3b), rod-like setae with an enlarged base in compound punctu- res, in which the setae opening is recessed in a Fig. 1. Dissolved oxygen uptake of adult D. aubei at 138Cin a closed respiration chamber. In a repeated measure design, shallow crater (sensillum trichoideum Type 2, Fig. oxygen uptake was measured in untreated beetles, in beetles 3d), and spoon-shaped setae with an enlarged base with the elytra covered with a synthetic resin and of beetles in simple punctures (sensillum trichoideum Type with the synthetic resin removed (control). Different letters rep- 2, Figs. 3c and 4a). In TEM sections, these spoon- resent significant differences (repeated measures ANOVA, F(2,8) shaped setae are richly tracheated in the enlarged 5 9.3054, P < 0.05, Scheffe´-test for post hoc comparison < 0.05). basal part. Each bulbous setal base with its cuticu- lar funnel is associated with a narrow hair chan- removed (control), oxygen uptake was not signifi- nel, which traverse the cuticle (Figs. 3e and 4). cantly different to the untreated group (refer Each hair channel contains two sometimes heli- Fig. 1) cally arranged tracheoles. In the epidermal cells,

Fig. 2. Oxygen uptake of adult D. aubei (a) and adult H. palustris (b) measured in a flow through respirometer. For homogene- ity of variances data are natural logarithm transformed. Because of the low oxygen uptake, nine beetles were measured simultane- ously and the mean uptake per animal was calculated. D. aubei were measured as untreated (n 5 8), with synthetic resin covered pygidum (n 5 5), covered elytra (n 5 5), and a control measurement with removed synthetic resin (n 5 5). Significant differences are marked by letters (one-way ANOVA; F(2,15) 5 10.61, P 5 0.0013 and Scheffe´ test). For H. palustris, two consecutive measure- ments of untreated beetles were performed (n 5 4), followed by one with the pygidium covered (n 5 3) and one measurement with the elytra covered with synthetic resin (n 5 4). Significant differences marked by letters (one-way ANOVA: F(3,10) 5 14.23, P < 0.001 and Scheffe´ test).

Journal of Morphology SETAL TRACHEAL GILLS IN DYTISCIDAE 1351

Fig. 3. SEM micrographs of the elytral surface of D. aubei.(a) Surface of the elytron. Numerous spoon-shaped setae are broken off behind the bend in nature (middle). (b) Long sensory seta in coarse puncture encircled by concentric ridges (sensillum trichoi- deum Type 1) (c) Setal tracheal gills, which are spoon-shaped at the distal end and tracheated in the enlarged basal part (sen- sillum trichodeum Type 2). (d) Rod-like setae with enlarged bases in shallow craters on the elytral surface (sensillum trichodeum Type 2). (e) Semi-schematic diagram of the setal tracheal gills of D. aubei. The highly branched tracheoles in the basal part of the seta are connected via the hair channel with one of the four longitudinal main tracheae on the lower side of the elytra. (f) Elytral surface of O. sanmarkii with mechanoreceptors and modified setae (sensillum placoideum Type 3, white arrowhead). These modi- fied setae are richly tracheated. the tracheoles increase in diameter and connect significantly from species from stagnant water (n via tracheae to one of the four longitudinal tra- 5 17, median 5 0.0038) and bogs (n 5 9, median cheal trunks of the elytra (Figs. 3e and 4a). Tra- 5 0.0033). There was no significant difference in cheoles in the hair channel enter the bulbous base Drel between species from running water and spe- and become thinner, and finally assume a circular cies which occur in both running and stagnant arrangement within peripheral area (refer Fig. 4). water (n 5 3, median 5 0.079) as well as species In the rheophilious species O. sanmarkii (Sahl- from springs (n 5 4, median 5 0.0045; refer Fig. berg), we found flat modified setae (sensillum pla- 5). Although in four species from stagnant water coideum Type 3, Fig. 3f) which were also richly (Hygrotus impressopunctatus, Hygrotus confluens, tracheated via the hair channel (Fig. 4e). Graptodytes pictus, and Nebrioporus canaliculatus) The relative diameter of longitudinal intraelytral an increased Drel was found, there was a differ- tracheae (Drel) of the investigated species from ence in tracheation type to the investigated species running water (n 5 12, median 5 0.0108) differ from running water. We found differences in the

Journal of Morphology 1352 S. KEHL AND K. DETTNER

Fig. 4. Transmission electron micrographs of the elytra and setae of D. aubei (a–d) and O. sanmarkii (e). (a) Transverse section of the elytron with tracheated setae. (b) Detail view of basal seta with longitudinal cut tracheoles. (c) Longitudinal section of the elytron and seta. (d) Detail view of tracheated seta. Note branching of the transverse cut tracheoles (white arrow). (e) Transverse section of the elytron of O. sanmarkii with richly tracheated setae. T, trachea; TR, tracheoles; EXO, exocuticle; END, endocuticle; HC, ‘‘hair channel’’; S, seta; PC, pore canal.

Drel also between some species from running DISCUSSION water and species from stagnant water in the Our results show an oxygen uptake in the rheo- same genus. For example, in Bidessus minutissi- mus (running water) and B. grossepunctatus (bogs) philious species D. aubei via tracheated setae on or Hydroporus marginatus (running and stagnant the elytra. An oxygen uptake via the elytra was water) and all other examined Hydroporus-species assumed by Smrzˇ (1981) in stygobiont water bee- (standing water and bogs). All species with trac- tles based on their rich intraelytral tracheation. heated setae showed an increased Drel. However, the cuticle of the elytra seemed to be too

Journal of Morphology SETAL TRACHEAL GILLS IN DYTISCIDAE 1353

Fig. 5. Maximal relative diameter of the longitudinal trachea of 45 Hydradephaga from different habitats in relation to the length of their elytra. The respective species are marked by numbers and the tracheation type is given in brackets: Type 1: Small longitudinal tracheae with little or not visible branching. Type 2: Large longitudinal tracheae with strong branching. Type 3: Com- pletely different tracheation from Type 1 and Type 2. For full species name, refer Nilsson (2001). The habitat preferences are for Central according to Hess et al. (1999), except for Graptodytes crux, D. aubei sanfilippoi, and D. semirufus. The Drel in running water species are significantly different from the species from stagnant water and bogs (Kruskal-Wallis-ANOVA: H(4,45) 5 27.33; P < 0.001, HSD post hoc procedure for non-parametrics). thick ( 50 lminD. aubei) and compact for cuta- oxygen with the help of the physical gill. Oxygen neous respiration. The TEM pictures clearly show consumption of epizooic ciliates could not be tracheoles (according to the definition of Mill, excluded in all oxygen uptake measurements, but 1998) in the basal parts of the setae of the elytra. we only selected beetles with no visible growth of Whether tracheoles enter also the spoon-shaped sessile ciliates. Furthermore, the ciliates typically part of the setae remains unclear, because the settle on ventral side or on the sides of the prono- setae did not optimally absorbed the fixation and tum and did not affect the elytra. The activity of embedding chemicals and the internal structure of the beetles with covered elytra was not reduced the spoon-shaped part of the setae collapsed. How- but instead were more active in their efforts to rid ever, we assume that only the basal parts of the themselves of the synthetic resin. setae are tracheated and that these setae function Intraelytral tracheae also were found in Elmidae as tracheal gills and are able to take up dissolved and Chrysomelidae with a connection to an intrae- oxygen. Madsen (2007) described air bubbles on lytral air-filled cavity (Messner and Langer, 1984). the elytra of running water dytiscids. He proposed The function of that system remains unclear, but an exchange of oxygen between the bubbles and capturing air bubbles and/or role in hydrostatic the tracheal system. We never saw air bubbles on pressure control was proposed. the elytral surface at lower water temperatures Larvae of Haliplidae uses microtubular gills for (7–138C), but at higher temperatures, in small gas exchange (Seeger, 1971). The morphology and aquaria, this phenomenon was observed. We also finestructure of these tubular gills are completely dif- never saw D. aubei using a compressible gas gill ferent to the setal tracheal gills of adult Dytiscidae. at lower water temperatures in the laboratory or The diffusion rate for oxygen through the cuticle in the field. A gas gill is described by Gilbert is described by Fick’s law [Diffusion rate 5 (diffu- (1986) for H. palustris and also for many other sion coefficient*surface area*concentration gradi- Dytiscidae (Rahn and Paganelli, 1968; Di Giovanni ent)/diffusion distance]. So, the diffusion rate is et al., 1999). Blocking the physical gill in D. aubei optimized by an increase of the respiratory sur- did not significantly reduce oxygen uptake, face, a high concentration gradient, and a short whereas H. palustris could only take up dissolved diffusion distance (Eriksen et al., 1996). D. aubei

Journal of Morphology 1354 S. KEHL AND K. DETTNER is restricted to cold mountainous and subalpine larly when the SEM pictures of Phreatodessus ha- streams (Fery and Brancucci, 1997) with high dis- des and Kuschelydrus phreaticus (Ordish, 1976) solved oxygen concentrations, which assure a high are compared to the elytral surface of O. sanmar- oxygen concentration gradient. The respiratory kii. Furthermore, stygobiont species show an surface is made up of numerous setae (on average increased diameter and number of the longitudinal 5900/mm2), which cover more than 30% of the ely- tracheae (Smrzˇ, 1981). Thus, setal gills may be the tral surface. Assuming the tracheation only in the answer to the extensively discussed respiration basal parts of the setae, the respiration surface is question of stygobiont Dytiscidae. We collected D.  5% of the elytral surface, not considering the aubei often deep in the hyporheic zone, especially three-dimensional structure of the setae. Further- after flooding in Germany (Black forest) or during more, not only the elytra are equipped with trac- drought in the Southern Alps. This occurrence and heated setae, but also we found the same type of assuming the stygobiont species have setal tra- setae on the pronotum and in lower density on the cheal gills, supports the evolution scenario of sty- ventral side. This explains the remaining oxygen gobiont species, in which species of the hyporheic uptake of beetles with covered elytra (Figs. 1 and zone may be driven into a subterranean existence 2). To reduce the diffusion distance, the tracheal (Leys et al., 2003; Balke et al., 2004; Leys and system traverses through the 50 lm thick cuticle Watts, 2008). of the elytra via the hair channel into the base of Lotic species are displaced downstream by the the spoon-shaped setae. So, the diffusion distance current. This drift could be avoided by morphologi- is reduced to the cuticle of the setae and is less cal adaptations such as adhesive structures or flat- than 1 lm (0.59 lm on average). In tracheal gills tened, streamlined bodies or by ecological adapta- of Trichoptera, the diffusion distance is  0.77 lm tions such as compensatory flight, upstream mov- (Wichard, 1973). ing or a special ecological niche (Wallace and The diffusion distance not only includes the cuti- Anderson, 1996). Respiration at the water surface cle thickness, but also the so called boundary in fast flowing rivers obviously will increase the layer, which could be described as a stagnant or risk of drifting. Especially, the rivers and streams unmixed ‘‘halo of O2 depletion’’ around an aquatic where D. aubei occurs are characterized by regular respiratory surface (Graham, 1990). The thickness fast and heavy flooding during which gas exchange of the boundary layer decreases as the flow rate in is not possible at the water surface without drift- the adjacent water layer increases or the layer will ing. Thus, for rheophilious species, the elytral be mixed by turbulence (Eriksen et al., 1996). Per- setal tracheal gills help to avoid drifting, which is haps the typical appearance of the elytral setae essential in the case of Deronectes, because all spe- (spoon-shape) causes such turbulence, disturbing cies of Deronectes are poor swimmers (Ribera the laminar flow and reducing the thickness of the et al., 1997) and unable to fly (Fery and Brancucci, boundary layer. For effective cuticular diffusion of 1997; Kehl and Dettner, 2007). oxygen it is also necessary that the cuticle be membranous and without hydrophobic wax layers (Wasserthal, 2003). ACKNOWLEDGMENTS We found setal tracheal gills also in the other The authors are grateful to Stefan Do¨tterl (Uni- species of the D. aubei-group and also in the phylo- versity of Bayreuth) for comments on statistical genetically basal species Deronectes latus, indicat- analyses, Stefan Ku¨ chler, Sabrina Herter, Kai ing that all members of the genus Deronectes are Drilling (all University of Bayreuth), and Stefan equipped with tracheated setae. Richly tracheated Geimer and Rita Grotjahn (Electron Microscopy modified setae were found also in the rheophilic Facility of the University of Bayreuth), Ingo species O. sanmarkii. But in the genus Hydropo- Schmidt and Heiko Ro¨del (both University of rus, which is mainly distributed in bogs, fens, and Bayreuth) for providing the oxygen measurement stagnant water no setae with tracheation were equipment. They also thank Joseph Woodring found. Comparison of the longitudinal tracheae of (University of Bayreuth) for revision and improve- running water species and species with other habi- ment of the manuscript and Hans Fery (Berlin) for tat preferences showed a significant increased Drel helpful comments. and great branching of the intraelytral tracheation in running water species. This indicates setal tra- cheal gills in other running water species. Current LITERATURE CITED studies and comparisons with SEM pictures of ely- Alt W. 1912. U¨ ber das Respirationssystem von Dytiscus margin- tra of the Dytiscidae in literature (Wolfe and Zim- alis. Z Wiss Zool 99:357–413. mermann, 1984) suggest that gas exchange via Balke M. 2005. Dytiscidae. In: Beutel RG, Leschen RAB, edi- setal tracheal gills are also common in other tors. Handbuch der Zoologie, Band IV Arthropoda: Insecta, Coleoptera, Beetles. Berlin: Walter de Gruyter. Hydroporinae, especially in species occurring in Balke M, Watts CHS, Cooper SJB, Humphreys WF, Vogler AP. running water. We are confident that setal tra- 2004. A highly modified stygobiont diving beetle of the genus cheal gills also occur in stygobiont species, particu- Copelatus (Coleoptera. Dytiscidae): and cladistic

Journal of Morphology SETAL TRACHEAL GILLS IN DYTISCIDAE 1355

analysis based on mitochondrial DNA sequences. Syst Ento- Messner B, Langer C. 1984. Atmungsaktive Deckflu¨ gel als mol 29:59–67. Anpassung an submerse Lebensweise bei Ka¨fern. Zool Jb Calosi P, Bilton DT, Spicer JI. 2007. The diving response of a Anat 111:469–484. diving beetle: Effects of temperature and acidification. J Zool Meuche A. 1937. Der Ka¨fer Bidessus hamulatus Gyll. in Osthol- 273:289–297. stein. Ent Bl 33:427–438. Castro A, Delgado JA. 2001. Iberoporus cermenius, a new genus Mill PJ. 1998. Tracheae and tracheoles. In: Harrison FW, Locke and species of subterranean water beetle (Coleoptera: Dytisci- M, editors. Microscopic Anatomy of Invertebrates. New York: dae) from . Aquat Insects 23:33–43. Wiley-Liss. pp 303–336. Cooper SJB, Bradbury JH, Saint KM, Leys R, Austin AD, Hum- Nilsson AN. 2001. World catalogue of insects. Dytiscidae (Cole- phreys WF. 2007. Subterranean archipelago in the Australian optera). Stenstrup: Apollo Books. p 395. arid zone: Mitochondrial DNA phylogeography of amphipods Ordish RG. 1976. Two new genera and species of subterranean from central Western Australia. Mol Ecol 16:1533–1544. water beetle from New Zealand (Coleoptera: Dytiscidae). N Z Di Giovanni MV, Pirisinu Q, Giangiuliani G, Goretti E, Pampa- J Zool 3:1–10. nella L. 1999. Oxygen consumption in two aquatic coleoptera Pritchard G, McKee MH, Pike EM, Scrimgeour GJ, Sloty J. species: Hydrous piceus and Dytiscus marginalis. Ital J Zool 1993. Did the first insects live in water or in air? Biol J Lin- 66:329–332. nean Soc 49:31–44. Ege R. 1915. On the respiratory function of the airstores carried Rahn H, Paganelli CV. 1968. Gas exchange in gas gills of diving by some aquatic insects (Corixidae, Dytiscidae, Notonecta). Z insects. Resp Physiol 5:145–164. Allg Physiol 17:81–124. Resh VH, Solem JO. 1984. Phylogenetic relationships and Eriksen CH, Lamberti GH, Resh VH. 1996. Aquatic insect res- evolutionary adaptations of aquatic insects. In: Merritt RW, piration. In: Merritt RW, Cummins KW, editors. An Intro- Cummins KW, editors. An Introduction to the Aquatic Insects duction to the Aquatic Insects of North America. Dubuque, of North America. Dubuque, Iowa: Kendall/Hunt. pp 66–75. Iowa: Kendall/Hunt. pp 29–40. Ribera I, Foster GN, Holt WV. 1997. Functional types of diving Fery H, Brancucci M. 1997. A taxonomic revision of Deronectes beetle (Coleoptera: Hygrobiidae and Dytiscidae), as identified SHARP, 1882 (Insecta: Coleoptera: Dytiscidae) (part I). Ann by comparative swimming behaviour. Biol J Linnean Soc Naturhist Mus Wien 99B:217–302. 61:537–558. Gilbert M. 1986. The respiratory system and respiratory tech- Seeger W. 1971. Morphology, bionomics, and ethology of Hal- nique of Hydroporus palustris (L.) (Coleoptera. Dytiscidae). iplids with special consideration of function and morphology Entomol Basiliensia 11:43–65. (Haliplidae; Coleoptera). Arch Hydrobiol 68:400–435. Graham JB. 1990. Ecological, evolutionary, and physical factors Smrzˇ J. 1981. Respiration—a new function of some hydroporine influencing aquatic animal respiration. Am Zool 30:137–146. elytra (Coleoptera. Dytiscidae, Hydroporinae). Acta Entomol Gruppe G. 1984. DIN 38408 Teil 21 (1984) Gasfo¨rmige Bestand- Bohemoslov 78:209–215. teile (Gruppe G), Bestimmung des in Wasser gelo¨sten Sauer- Spangler PJ, Decu V. 1999. Coleoptera aquatica. In: Juberthie stoffes, iodometrisches Verfahren nach Winkler (G21). C, Decu V, editors. Encyclopaedia Biospeologica. Bucharest: Hanke W, Hamdorf K, Horn E, Schlieper C. 1980. Praktikum Socie´te´ de Biospe´ologie, Academie Roumaine. pp 1031–1046. der Zoophysiologie. Stuttgart: Fischer. Ue´no S-I. 1957. Blind aquatic beetles of Japan, with some Hess M, Spitzenberg D, Bellstedt R, Heckes L, Hendrich L, accounts of the fauna of Japanese subterranean waters. Fun- Sondermann W. 1999. Artenbestand und Gefa¨hrdungssitua- dam Appl Limnol 53:250–296. tion der Wasserka¨fer Deutschlands. Naturschutz und Land- Wallace JB, Anderson NH. 1996. Habitat, life history, and schaftsplanung 31:197–211. behavioral adaptations of aquatic insects. In: Merritt RW, Kehl S, Dettner K. 2007. Flugfa¨higkeit der in Deutschland vor- Cummins KW, editors. An Introduction to the Aquatic kommenden adephagen Wasserka¨fer (Coleoptera. Hydrade- Insects of North America. Dubuque, Iowa: Kendall/Hunt. pp phaga). Entomol heute 19:141–161. 29–40. Leys R, Watts CHS, Cooper SJB, Humphreys WF. 2003. Evolu- Wasserthal LT. 2003. 6 Atemsystem. In: Dettner K, Peters W, tion of subterranean water beetles (Coleoptera: Dytiscidae: editors. Lehrbuch der Entomologie. Heidelberg: Spektrum. : Bidessini) in the arid zone of Australia. Evolu- pp 165–183. tion 57:2819–2834. Wesenberg-Lund C. 1912. Biologische Studien u¨ ber Dytisciden. Leys R, Watts CH. 2008. Systematics and evolution of the Aus- Intern Rev Hydrobiol u Hydrogr Biol Suppl 5:1–129. tralian subterranean hydroporine diving beetles (Dytiscidae), Wichard W. 1973. Zur Morphogenese des respiratorischen Epi- with notes on Carabhydrus. Invertebr Syst 22:217–225. thels der Tracheenkiemen bei Larven der Limnephilini KOL. Madsen BL. 1967. Økologiske undersøgelser i nogle østjyske (Insecta. Trichoptera). Cell Tissue Res 144:585–592. vandløb. Part 3: Vandløbsbillers udbredelse og bioligi. Flora Wichard W, Arens W, Eisenbeis G. 2002. Biological atlas of og Fauna 73:21–36. aquatic insects. Stenstrup: Apollo Books. p 339. Madsen BL. 2007. Vandløbsbillers (Haliplidae. Dytiscidae; Cole- Wolfe GW, Zimmerman JR. 1984. Sensilla, punctation, reticula- optera) der a˚nder gennem dækvingerne. Flora og Fauna 113: tion, and body shape in the hydroporinae (Coleoptera, Dytisci- 71–81. dae). Int J Insect Morphol Embryol 13:373–387.

Journal of Morphology