1662

The Journal of Experimental Biology 209, 1662-1677 Published by The Company of Biologists 2006 doi:10.1242/jeb.02203

The significance of spiracle conductance and spatial arrangement for flight muscle function and aerodynamic performance in flying Drosophila Nicole Heymann and Fritz-Olaf Lehmann* Department of Neurobiology, University of Ulm, Albert-Einstein-Allee 11, 89081 Ulm, Germany *Author for correspondence (e-mail: [email protected])

Accepted 7 March 2006

Summary During elevated locomotor activity such as flight, capacity for carbon dioxide in Drosophila amounts to Drosophila satisfies its increased respiratory demands by approximately 33.5·␮l·g–1·body·mass, estimated from the increasing the total spiracle opening area of the tracheal temporal integral of carbon dioxide release rate during gas exchange system. It has been assumed that in a the resting period after flight. By comparing flight diffusion-based system, each spiracle contributes to variables in unmanipulated and ‘spiracle-blocked’ flies at oxygen flux into and carbon dioxide flux out of the comparable flight forces, we found that (i) stroke tracheal system according to the size of its opening. We amplitude, stroke frequency and the chemo-mechanical evaluated this hypothesis by determining how a reduction conversion efficiency of the indirect flight musculature in size and interference with the spatial distribution of gas were broadly independent of the arrangement of spiracle exchange areas impair flight muscle function and conductance, while (ii) muscle mechanical power aerodynamic force production in the small fruit fly significantly increased, and (iii) mean lift coefficient and Drosophila melanogaster. This was done by selectively aerodynamic efficiency significantly decreased up to blocking thoracic spiracles of tethered flies flying inside a approximately 50% with an increasing number of blocked flight simulator. Flow-through respirometry and spiracles. The data suggest that Drosophila apparently simultaneous measurements of flight force production and maximizes the total efficiency of its locomotor system for wing kinematics revealed a negligible functional safety flight by allowing oxygen delivery to the flight margin for respiration. Maximum locomotor performance musculature through multiple spiracles of the thorax. was only achieved by unmanipulated flies, supporting the general assumption that at the animal’s maximum Key words: respiration, gas exchange, spiracle opening, aerodynamic locomotor capacity, maximum spiracle opening area force production, IFM, asynchronous flight muscle, flight power matches respiratory need. The maximum total buffer requirements, breathing, fruit fly, Drosophila melanogaster.

Introduction morphological architecture of the tracheal system, including Energetically demanding processes in animals greatly size and location of the spiracle openings (Weis-Fogh, 1964a; influence most physiological systems, and above all the Weis-Fogh, 1964b). For example, Drosophila may store respiratory system. In flying insects, in particular, metabolic oxygen by using haemoglobin in the tracheal walls to reinforce activity may increase 10- to 15-fold over resting metabolism, oxygen partial pressure gradients (de Sanctis et al., 2005; forcing the tracheal gas exchange system to increase the supply Hankeln et al., 2005), and bicarbonates in the haemolymph of oxygen and the release of carbon dioxide (Casey, 1989; may buffer tracheal carbon dioxide at the expense of changes Casey and Ellington, 1989; Ellington et al., 1990; Lehmann in pH (e.g. Gulinson and Harrison, 1996). and Dickinson, 1997). However, despite many years of In the past, several researchers have attempted to determine investigation, there is still uncertainty about how exactly the significance of respiratory currents inside the tracheal oxygen enters the tracheal system and distributes within the system by studying the role of individual spiracles in tracheal network, and carbon dioxide eventually leaves the respiration and energetics. Bailey was one of the first respiratory system through the open spiracles (Wigglesworth, physiologists to investigate the interplay between thoracic and 1972). In general, respiratory currents inside the tracheal abdominal spiracles by blocking individual thoracic spiracles system are difficult to predict because they depend not only on in the honey bee (Bailey, 1954). He found that under CO2 several physiological factors, such as the buffer capacities for stress, the resting animal produced a net tracheal air current respiratory gases and ventilatory strategies, but also on the from the thorax to the abdomen using abdominal pumping.

THE JOURNAL OF EXPERIMENTAL BIOLOGY Spiracle function in Drosophila 1663

This tracheal convection was, however, only present when the fact that the left side of the tracheal system is virtually isolated propodeal spiracle (=metathoracic spiracle) was blocked, and from the right side. In Drosophila, there are only a few tracheal ceased when air was prevented from entering through the first commissures (pro- and metathoracic dorsal commissure, and thoracic spiracle. Similar to Bernoulli ventilation, which was pro-, meso- and metathoracic ventral commissures) that allow later proposed for flight respiration in large beetles (Miller, exchange of respiratory gases between the body sides (Miller, 1966), Bailey also suggested that in flying bees air is inhaled 1950). In the locust Schistocerca gregaria the thoracic system by the first spiracles and exhaled via the propodeal spiracles. is, moreover, isolated from the abdomen (Weis-Fogh, 1964b), In locusts and cockroaches, inspired air is routed similarly while in Drosophila large tracheae connect the thoracic from the anterior spiracles to the segmental tracheae and leaves parenteric air sac with the abdominal air sac (Miller, 1950). the system through the abdominal spiracles (Miller, 1982). For geometric reasons, changes in the spatial distribution of the This type of breathing was also demonstrated in Lepidopteran total spiracle opening areas might directly hinder proper flight pupae (Schneiderman, 1960), which may even control the muscle function because the insect flight muscle has no activity of individual spiracles (Slama, 1999). A unidirectional anaerobic capacity and thus relies on instantaneous oxygen air flow was reported for tethered flying hawkmoths Manduca supply (Ziegler, 1985). By contrast, the extensive tracheal sexta (Wasserthal, 2001). In this insect, there is an air stream development in Drosophila, with broad latero-linear air sacs towards the posterior spiracles that results from a such as the parentic-, pleural- and lateroscutal sac, favours the subatmospheric (negative) pressure at the mesothoracic establishment of homogenous partial pressures on the spiracle and a positive pressure in the mesoscutellar air sacs. ipsilateral side even when gas exchange rates are not balanced By contrast, in resting wingless dung beetles the direction of between the ipsilateral meso- and metathoracic spiracle. air flow is inverted (retrograde convection) and tracheal gases Consequently, in this study we examine how reducing the flow forward from the posterior to the anterior body (Duncan size and deleteriously interfering with the spatial distribution and Byrne, 2002). These authors also reported a lateral of gas exchange areas impair flight muscle function and asymmetry of air flow inside the animal and demonstrated that aerodynamic force production in the small fruit fly Drosophila the right mesothoracic spiracle is the primary route for melanogaster. This was done by selectively blocking thoracic respiratory gas exchange. A study on running energetics in the spiracles in tethered animals flying inside a flight simulator. ant Camponotus hypothesised that asymmetries in gas Flow-through respirometry and simultaneous measurements of exchange rate might also occur during the discontinuous gas flight force production and wing kinematics allow the exchange cycle (DGC) (Lipp et al., 2005). This hypothesis determination of changes in vital flight parameters under a vast arose from the finding of various numbers of ‘O’-peaks within variety of breathing conditions. Incorporating energetic and a single gas exchange cycle, suggesting that multiple spiracles aerodynamic theory, we show how (i) maximum mechanical may be involved in tracheal gas exchange. Moreover, in the power output of the indirect flight muscle (IFM), (ii) the arid-adapted ant Cataglyphis bicolor, the thoracic spiracles act efficiency with which the muscle converts chemical energy as high-capacity gateways to the tracheal system and are into muscle mechanical power, (iii) lift and drag coefficients responsible for approximately 90% of the total gas exchange for flapping wing motion and (iv) aerodynamic efficiency, all rate during running activity (Lighton et al., 1993a). In this ant, change with changing arrangements of spiracle conductance. the abdominal spiracles combined only have approximately half the diffusive capacity of a single thoracic spiracle (Lighton et al., 1993a). Materials and methods Size constraints in small insects such as the fruit fly Animals Drosophila, make it difficult to assess tracheal air currents, The data for this study were collected from 3- to 5-day-old spiracle function or tracheal partial pressures. The situation is fruit fly females Canton S wild-type Drosophila melanogaster even more complex in flying animals in which the spiracles Meigen. The animals stemmed from an inbred laboratory strain may dynamically vary their opening areas according to and were reared at room temperature (22°C) on commercial changes in flight power requirements. For example, the results Drosophila food (Carolina Biological, Burlington, NY, USA). of a study on spiracle opening behaviour in flying Drosophila To estimate in vivo flight muscle mechanical power output and are consistent with at least three basic mechanisms of spiracle respiratory rate, we anaesthetized the animals on a cold stage control (Lehmann, 2001). The fly may achieve an average at 4°C and tethered them to small tungsten holders using UV- tracheal conductance by either (i) matching the area of each light activated glue (Clear Glass Adhesive, Henkel Loctite, thoracic spiracle to the respiratory needs, (ii) dynamically Düsseldorf, Germany). Curing time was 20·s using a 75·W closing and opening the spiracles over time, or (iii) closing tungsten halogen lamp. Similar to the testing procedure some spiracles while other spiracles remain open. All mentioned below, we tested the toxicity of both the non- mechanisms could give similar mean tracheal conductance polymerised and polymerised glue on 25 Drosophila that had when estimated within a certain time period, although the latter been exposed to the glue for approximately 24·h. The data mechanism, in particular, might produce temporal fluctuations show that none of the animals died within the observation in the local supply of oxygen. The importance of matching at period and superficially we did not observe any changes in least the bilateral control of spiracle activity arises from the their behaviours.

THE JOURNAL OF EXPERIMENTAL BIOLOGY 1664 N. Heymann and F.-O. Lehmann

We determined the functional significance of changes in negligible. We calculated mean lift and drag coefficients from respiratory gas exchange through the four major thoracic total flight force, wing velocity and wing size by employing spiracles of Drosophila by covering individual spiracle quasi-steady aerodynamic theory and assuming that the openings with small droplets of commercial two-component chordwise aerodynamic circulation is maximum close to a epoxy glue (5-min epoxy, R&G, Waldenbuch, Germany). The spanwise location of 65% wing length (Birch and Dickinson, resin component of the glue consisted of bisphenol A- and F- 2001; Lehmann et al., 2005; Ramamurti and Sandberg, 2001). epichlorhydrine, and the active component of the hardener was A detailed description of this procedure is given in a previous 2,4,6-tri(dimethylaminomethyl)phenol. The glue was selected study on Drosophila flight (Lehmann and Dickinson, 1998). according to its property to cure fast and in very small volumes. Body mass (means ± s.d.) of the tested animals was We tested the glue’s toxicity for Drosophila by exposing 25 0.92±0.08·mg (N=10 flies, all thoracic spiracles left open), flies in a standard Drosophila vial to the resin, the hardener 0.88±0.18·mg (N=26, 1 spiracle sealed), 0.94±0.20·mg (N=43, and the polymerised glue. Although the flies did physically 2 spiracles sealed), 0.96±0.17·mg (N=23 flies, 3 spiracles contact the polymerised glue and its components, none of sealed) and 0.92±0.08·mg (N=5 flies, 4 spiracles sealed). A the animals died or superficially changed behaviour within the statistical test on body mass showed no significant differences testing period of 24·h. Moreover, it seems unlikely that the between the tested groups (ANOVA, P>0.05). Mean small cuticle area covered by the epoxy droplets influenced temperature during the experiments was 23.5°C. the mechanical properties of the thoracic box during wing flapping, for two reasons: first, the glue droplet was only Tracheal gas exchange area marginally larger than the opening area of the meso- or To determine maximum spiracle opening area of the four metathoracic spiracles, and second, an experiment in which we thoracic spiracles, we prepared 10 females for scanning placed a glue droplet close to but not on the animal’s spiracle electron microscopy (SEM) using a standard laboratory opening did not produce any alterations in aerodynamic force protocol (Fig.·1A). In Drosophila, the prothoracic spiracle is production. Throughout the manuscript we use the term reduced while the mesothoracic spiracle has moved anteriorly ‘unmanipulated flies’ to mean tethered flying animals with (Fig.·1B). By contrast, the metathoracic spiracle is located none of their thoracic spiracles sealed. directly beneath the beating haltere. When at all possible, The flies were allowed to recover from the tethering and before taking pictures we oriented the thoracic spiracle sealing procedure for at least 30·min before being placed into openings perpendicular to the focal plane of the SEM, in order the flight arena. To derive muscle mass-specific power output, to avoid measurement errors due to image distortions. We we measured the mass of each fly after each experiment using estimated spiracle opening area from the images using self- a balance (sensitivity 0.01·mg; Sartorius MC210P, Göttingen, written software routines in Scion Image (Scion, Frederick, Germany) and assuming a flight muscle-to-body mass ratio of Maryland, USA; red areas, Fig.·1). Fig.·1C,D shows that the approximately 0.3 (Lehmann and Dickinson, 1997). For body metathoracic spiracle opening (~2745·␮m2) is approximately mass estimation, we did not remove the glue from the spiracle 20% larger than the mesothoracic opening (~2186·␮m2), openings because its contribution to the fly’s total mass was resulting in a total maximum area for thoracic respiratory gas

Fig.·1. Location and size of spiracle openings in the fruit fly Drosophila melanogaster. (A) sp1, mesothoracic spiracle; sp2, metathoracic spiracle; sp3–9, abdominal spiracles. (B) Scanning electron microscopic image of Drosophila shows the position of the anterior spiracle sp1 between propleura and mesopleura. The posterior spiracle sp2 is located between the basis of the haltere and the mesomera. In Drosophila the prothoracic spiracle is reduced. (C,D) Shape and size of the spiracle opening area of sp1 and sp2, respectively. Red shading approximately indicates measured spiracle opening area. The metathoracic spiracle opening area is approximately 26% larger than the mesothoracic spiracle opening area. Values are means ± s.d., N=10 flies.

THE JOURNAL OF EXPERIMENTAL BIOLOGY Spiracle function in Drosophila 1665 exchange of approximately 9862·␮m2. In comparison, a experiments in which gas exchange through all 18 spiracles previous study described the thoracic spiracles in Drosophila was completely blocked, the flies died within approximately melanogaster as oval openings of approximately 10·min after the treatment (N=5 animals). This result is similar 60·␮mϫ25·␮m at the surface (Manning and Krasnow, 1993). to a study on the regulation of carbon dioxide release from the From these values we calculated a total thoracic exchange area thoracic and abdominal spiracles in the ant Cataglyphis of 5675·␮m2, which is approximately 42% smaller than the (Lighton et al., 1993a). In this animal, the abdominal spiracles value measured in the present study. There could be several have half the diffusive capacity of a single thoracic spiracle reasons for this discrepancy. First, although the aspect ratio and can fully meet the ant’s oxygen uptake and carbon dioxide between 60 and 25·␮m fairly matches the elongated shape of release at resting metabolism. the mesothoracic spiracle, the metathoracic spiracle seems to be more circular with a diameter between 50 and 60·␮m, and Testing procedure for spiracle seals thus appears to be larger than measured by Manning and To evaluate the seal of the epoxy droplets used to restrict Krasnow. Second, since we measured total opening area from respiratory gas exchange through the spiracle openings, we the size of the thick sclerite that borders the spiracle opening developed the experimental setup shown in Fig.·2. The animals at the surface, we might have overestimated the inner opening were anaesthetized using carbon dioxide and prepared by size (Nikam and Khole, 1989). Third and alternatively, the ablating head and abdomen (N=15 flies). Subsequently, we females that were used in the present study were chosen sliced the thorax along the sagittal plane and carefully removed according to body size and represented the largest animals all tissues out of the flies’ body. One side of the thorax was from the population. Larger females typically show a better then mounted above a 0.5·mm hole in a 15.6·cm3 respirometric flight endurance than their smaller relatives. Consequently, chamber using epoxy glue. We permanently sealed the since larger fruit flies exhibit larger spiracle openings (F.- metathoracic spiracle and only modified the mesothoracic O.L., unpublished data), it might be that the difference spiracle opening for testing. To estimate the diffusivity of the between the two opening measurements is a real difference in spiracle opening for CO2, we aligned a bell-shaped gas outlet size rather than due to substantial differences in measurement (3.8·cm3 volume) above the fly thorax. Either ambient air or techniques. 100% CO2 was supplied via an electrically activated two-way Compared to thoracic spiracles, the area of the 14 abdominal valve. To avoid potential changes in cuticle diffusion spiracles was not measured directly, but was estimated by coefficient due to dry-out processes, both gases were routed assuming that it represents 5% of the fly’s total gas exchange through a plastic bottle containing purified water. Similar to area or approximately 493·␮m2 (Manning and Krasnow, the in vivo respirometric measurements described below, water 1993). A simple behavioural test showed that gas exchange and CO2 were removed from room air and pulled at a rate of mediated by abdominal spiracles is high enough to satisfy the 1000·ml·min–1 through the respirometric test chamber. oxygen needs during resting metabolism, whereas in Fig.·2B shows a typical trace of CO2 measurement, with

Fig.·2. Experimental apparatus for A 100% CO B evaluation of thoracic spiracle seals. 2 (A) The thorax of Drosophila is CO2 sliced into halves along the sagittal 200 p.p.m. plane and mounted on top of a flow- Ambient air Valve through respirometric chamber. Flight musculature is removed and Mesothoracic the metathoracic spiracle is spiracle 80 s permanently sealed by epoxy glue. A Metathoracic C 0.5·mm hole in the wall of the spiracle respirometric chamber permits Thorax ambient gas to be pulled through the half Glue open mesothoracic spiracle inside the D chamber. A bell-shaped gas outlet mounted above the chamber allows alterations in ambient CO2 Flow-through concentration by connecting the gas respirometry tubing either to pressurized room air Respirometric or to a CO2 reservoir using an electric chamber valve. (B) Example of how gas flux through the open mesothoracic spiracle varies while alternately connecting the gas tubing to room air (grey) and CO2 (blue). Measuring units are given in parts-per-million analysed air (p.p.m.). (C) The mesothoracic spiracle seal completely blocks CO2 flux into the respirometric chamber. (D) Removing the spiracle seal from the mesothoracic spiracle after testing restores spiracle conductance for carbon dioxide (same thorax half in B–D).

THE JOURNAL OF EXPERIMENTAL BIOLOGY 1666 N. Heymann and F.-O. Lehmann periodic switching between ambient air and the supply of CO2. Respiratory measurements and flight arena After this pre-check, the mesothoracic spiracle was sealed and The tethered fruit flies were flown in a virtual reality flight the testing procedure performed again. Despite the high partial arena in which stroke amplitude, stroke frequency, total force pressure of CO2 on the outside of the thorax, the small seal production and carbon dioxide release were measured typically blocked the entire inflow of CO2 into the simultaneously (Dickinson and Lighton, 1995; Fig.·3A). A respirometric chamber (Fig.·2C). Afterwards, in a control more detailed description of the experimental apparatus and the experiment, we carefully removed the seal that fully restored procedure is given elsewhere (Lehmann and Dickinson, 1997). the diffusivity of the mesothoracic spiracle for CO2 (Fig.·2D). Under closed-loop feedback conditions, fruit flies actively Note that the vibrating thorax of a flying fly might cause modulated the azimuth velocity of a vertical dark stripe leakage of the spiracle seal, and this effect is not simulated in displayed in the panorama using the relative difference in the simple testing procedure. Therefore, we covered the stroke amplitude between the two beating wings. The feedback mesothoracic spiracle of tethered flies and flew the animals for conditions were set according to previous experiments on at least 15·min. We subsequently dissected those animals in a tethered flying fruit flies (Lehmann and Dickinson, 2001). procedure similar to that described above and checked for any While flying in closed-loop, Drosophila changed both gas leakages. In none of the five tested flies did we measure kinematic and respirometric parameters in response to the any CO2 flux through the sealed spiracles, suggesting that vertical motion of an open-loop striped grating. We have thorax vibrations during flight do not harm the quality of the previously shown that under those conditions fruit flies spiracle seals. In sum, the outcome of the pre-tests convinced maximize their locomotor output, while trying to follow the us that the epoxy seal used in this study is able to sufficiently upward motion of the horizontal stripes (Lehmann and block respiratory gas exchange through the spiracle openings Dickinson, 1998). This procedure is equivalent to adding in a flying fruit fly. weight to the animal and elicits maximum flight forces similar

A IRD C BP 200 B 250

) 160 200

(Hz

de (deg.)

PSD 120 cy 150

amplitu 80 100

stroke 40 50

L Stroke frequen 0 0

Mean 03124 03 12 4 WSA

1.6 D 1.6 E 1.6 F 1.36 1.2 1.2 1.2 1.01

D 0.8 L 0.8 0.8 — —

C

0.74 C 0.48 0.4 0.4 0.4 0.29

Flight force/body mass 0 0 0 0 123 4 0312 4 0312 4 No. of open thoracic spiracles

Fig.·3. Virtual-reality arena and flight data plotted as a function of open spiracles in Drosophila. (A) Set-up as described (Lehmann and Dickinson, 1997). To elicit maximum locomotor performance of the animal, a 30° stripe drum (BP) displayed in the electronic flight arena was oscillated under open-loop conditions in a vertical direction around the tethered flying fly. IRD, infrared diode; PSD, position detector of flight force laser balance; L, laser; WSA, wing stroke analyser. (B) Wing stroke amplitude, (C) wing stroke frequency, (D) maximum normalized — — flight force production, (E) mean lift coefficient CL, and (F) mean drag coefficient CD, based on a quasi-steady aerodynamic approach, plotted against the number of open thoracic spiracles (grey). Abdominal spiracles remained unsealed in all experiments. Data represent mean values of all data points within a flight sequence that fell within the top 1% of flight force (equal to maximum locomotor capacity of the fly). Number of tested flies: N=5 (0), N=23 (1), N=43 (2), N=26 (3) and N=10 (4 open thoracic spiracles). To distinguish the changes resulting from the modifications of local spiracle gas conductance from those associated with alterations in total flight force production, we estimated kinematic and aerodynamic parameters in unmanipulated animals (see text) within a ±2% range of flight forces that match the maximum values shown in D (red). Red area in B,C, E,F indicates ±s.d. N=10 flies. See text for details.

THE JOURNAL OF EXPERIMENTAL BIOLOGY Spiracle function in Drosophila 1667 to those obtained in freely flying flies using a weight-lifting gas exchange to the abdominal spiracles. To estimate how total technique (Lehmann, 1999). spiracle opening area constrains maximum locomotor capacity Concurrently, we employed flow-through respirometry with in the fruit fly, we compared mean values of all data points a flow rate of 1000·ml·min–1 and used a Li-cor 7000 gas within a flight sequence that fell within the top 1% of flight analyser (Licor, Lincoln, Nebraska, USA) to measure the rate force. These means are plotted in grey (Figs·3–5). Surprisingly, of carbon dioxide release during flight. The internal filter wing kinematics were less affected by the restriction of frequency of the gas analyser was set to 0.2·s. Data sampling diffusive area than we expected, because a 95% decrease of frequency was 125·Hz and wash-out time constant ␶ of the diffusive area from approximately 10 355·␮m2 (unmanipulated 15·ml respirometric chamber was approximately 910·ms animal) to approximately 493·␮m2 (abdominal spiracle (wash-out timeϰe–x/␶; Lehmann and Heymann, 2005). exchange area) caused only an approximately 10.7% Respirometric data of each animal were corrected for both concomitant decrease in stroke amplitude (from 176±7.0 to temporal shift due to wash-out and the gas delay due to the 157±8.1°, Fig.·3B) and an approximately 20% decrease in connecting tubings, and were eventually normalized to stroke frequency (from 221±12.0 to 177±8.9·Hz, Fig.·3C). standard temperature and pressure (STP). Further analysis and These small changes in wing kinematics coincided with a calibration of kinematic and respiratory data were performed linear decrease in flight force production from 1.36 to 0.29 as described elsewhere (Barton et al., 2005). Metabolic data body-mass specific force with decreasing spiracular gas and power requirements for flight are given as flight-specific exchange area (linear regression fit, y=0.16+0.11x, R2=0.97, values by subtraction of resting rates, and if not stated P=0.002; Table·1, Fig.·3D). The data show that active flight in otherwise also expressed as indirect flight muscle (IFM) mass- Drosophila defined as the locomotor performance at which the specific units. animals produce flight forces equal to or greater than body weight, required gas exchange through at least three thoracic Experimental procedure spiracles including the abdominal spiracles (Fig.·3D). Note that To vary the size of total diffusive spiracle area in the flying despite their small total opening area (5% of 10 355·␮m2), the fruit fly, we tested different combinations of spiracle sealing in functional relevance of the 14 abdominal spiracles for oxygen random sequences and subsequently pooled the data subsets supply broadly compares with the significance of each thoracic (Figs·3–5). This means that in case of a single seal, we tested spiracle: together, they contributed approximately 20% to flight performance under four different experimental maximum aerodynamic performance of an unmanipulated conditions: a blockage of the left sp.ms (mesothoracic spiracle, flying fruit fly (Fig.·3D). N=6 flies), right sp.ms (N=7 flies), left sp.mt (metathoracic The different regression slopes of kinematic measures spiracle, N=7 flies) and right sp.mt (N=6 flies). In flies with two (amplitude: 1.90ϫ103·deg.·␮m–2, frequency: 4.70ϫ sealed spiracles we investigated flight and muscle performance 103·Hz·␮m–2·diffusive·area) and relative flight force production under six different experimental conditions (sealed left and (0.11ϫ103·relative·force·␮m–2·diffusive·area; Table·1) imply right sp.ms, N=8; left and right sp.mt, N=7; left sp.ms and that changes in wing velocity cannot exclusively explain the sp.mt, N=8; right sp.ms and sp.mt, N=6; left sp.ms and right changes in aerodynamic force production, because wing sp.mt, N=6; and the right sp.ms and left sp.mt, N=8 flies). velocity is directly proportional to the product between stroke Results obtained from flies with three sealed spiracles are mean amplitude and frequency (Lehmann and Dickinson, 1998). values of four experimental conditions (sealed left sp.ms, sp.mt Consequently, changes in total spiracular conductance also and right sp.ms, N=7; left sp.ms, sp.mt and right sp.mt, N=8; altered the lift coefficient for flapping wing motion that right sp.ms, sp.mt and left sp.ms, N=4; right sp.ms, sp.mt and decreased by approximately 69%, from 1.21 at a diffusive area left sp.mt, N=4 flies). Moreover, previous results on of 10355·␮m2 to 0.38 at 493·␮m2 (Fig.·3E). Moreover, since Drosophila flight energetics have shown that there is a flight force is the vector sum of lift and wing profile drag, the transient effect on flight performance after the initial take-off wing’s drag coefficient decreased likewise and similarly to the 2 during which flight force production and CO2 release rate peak lift coefficient by approximately 59%, from 0.65 at 10 355·␮m for a short time. This transient peak was present in most of our spiracle opening area to 0.38 during breathing through experiments except in flies with all four thoracic spiracles abdominal spiracles only (Fig.·3F). sealed. Consequently, to circumvent any transient phenomena, we excluded the first 5·s and the last 2·s of each flight sequence Power requirements and metabolic power from our analysis. Throughout the manuscript all values are As a consequence of the reductions in (i) wing kinematics, given as means ± s.d. (ii) aerodynamic lift production and (iii) drag coefficient, the maximum power requirements for flight, such as the flight muscle mass-specific induced- and profile requirements, Results decreased with the decreasing number of open thoracic Kinematic and aerodynamic attenuations spiracles (Fig.·4A,B). Superficially, mass-specific induced All flies tested in the flight arena were able to fly power seemed to be more affected by the respiratory continuously under tethered conditions for at least 14·s per restrictions than profile power and decreased by 91% from flight sequence, including those animals in which we restricted ~33.8·W·kg–1 in unmanipulated animals to ~3.2·W·kg–1, when

THE JOURNAL OF EXPERIMENTAL BIOLOGY 1668 N. Heymann and F.-O. Lehmann

Table·1. Parameters of linear regression fits between the size of the diffusive exchange area of thoracic and abdominal spiracle openings and various flight measures Diffusive area ( m2) (x) vs: y-intercept Slope ( 103) R2 P Figure (degrees) 155±1.23 1.90±0.19 0.97 0.002** 3B n (Hz) 180±7.05 4.70±1.08 0.87 0.022* 3C –1 Max FTMb 0.16±0.07 0.11±0.001 0.97 0.002** 3D CL 0.27±0.07 0.087±0.009 0.96 0.003** 3E CD 0.32±0.04 0.028±0.006 0.88 0.017* 3F –1 P*ind (W·kg ) –1.87±3.19 3.1±0.49 0.93 0.008** 4A –1 P*pro (W·kg ) 15.0±3.79 4.3±0.58 0.95 0.005** 4B –1 P*mech (W·kg ) 13.1±6.80 7.5±1.00 0.95 0.006** 4C –1 PMR* (W·kg ) 354±53.6 66.7±8.20 0.96 0.004** 4D M (%) 5.72±1.17 0.29±0.18 0.47 0.20 NS 5A A (%) 9.95±1.17 1.70±0.18 0.97 0.003** 5B T (%) 0.38±0.25 0.20±0.04 0.90 0.014* 5C

Parameters were determined at 1% maximum flight force production while the fly was varying power output in response to the motion of the visual open-loop lift stimulus. For statistical analysis the number of blocked spiracles (Figs·3–5) were converted into total gas exchange area using the data shown in Fig.·1C,D and assuming that abdominal spiracle area is approximately 5% of the total spiracle opening area. Total spiracle opening exchange area is: 10 355· m2 (unmanipulated animal), 7889· m2 (1), 5424· m2 (2), 3780· m2 (3) and 493· m2 (4 thoracic spiracles sealed). Significance level of slope: *P<0.05, **P<0.01; NS, not significant. For further abbreviations see List of symbols. gas exchange only occurred through the abdominal spiracles within the flight motor (Ellington, 1984b). Similar to induced (Fig.·4A, Table·1). We found significantly smaller changes for and profile power requirements, this measure decreased with mass-specific profile power, which decreased by ~69% with decreasing spiracle opening area by ~7.5±1.0·W·kg–1·␮m–2 decreasing gas exchange area from ~64.7 in unmanipulated spiracle area (linear regression fit, y=13.1+7.5x; Fig.·4C, flying Drosophila to 20.2·W·kg–1 when all thoracic spiracles Table·1). Flight-specific metabolic power was calculated from were sealed (91±22% vs 69±15%, t-test, P<0.05, N=8; Fig.·4B, the instantaneous measurements of CO2 release during flight Table·1). According to Ellington’s energetic theory for and was ~68% lower in flies that only breathed through flapping wing motion, mechanical power output of the abdominal spiracles (334·W·kg–1, Fig.·4D) compared to the asynchronous flight muscle is the sum of induced and profile unmanipulated control group (~1031·W·kg–1; Fig.·4D, power requirements, assuming 100% energy elastic storage Table·1).

Changes in muscle and flight efficiency 50 A 80 B Due to the changes in flight muscle mass-specific mechanical power output and metabolic power, muscle 40 )

) 60 –1 efficiency changed likewise. Fig.·5A shows that muscle –1 30 efficiency, defined as the ratio between metabolic and muscle

(W kg 40 mechanical power output, decreased from ~9.8±1.2% in (W kg 20

pro

*

ind

* P 20 P 10

0 0 Fig.·4. Changes in flight muscle mass-specific power requirements for 0 1234 0 1234 flight and metabolic power with increasing numbers of thoracic spiracles participating in respiratory gas exchange (grey bars, A–D). 120 C 1500 D Data were measured while the flies produced maximum aerodynamic flight forces (topmost 1% values of each flight sequence). Red data

)

)

–1 80 –1 1000 indicate power values estimated in—— unmanipulated flies (see text) at P* the corresponding flight force. ind, induced power requirements——

(W kg (W kg are the costs to generate an air downward momentum; P*pro, profile

MR power requirements are the costs to overcome drag on the beating 40 * 500

mech * ———

P P wings; P*mech, flight muscle mechanical power output that is equal to the sum of induced and profile power requirements assuming 100% 0 —— 0 elastic storage; and P*MR = metabolic power estimated from 0 1234 0 1234 measurements of CO2 release through the spiracle. For more No. of open thoracic spiracles explanations see Fig.·3 legend. Values are means ± s.d. (N=10 flies).

THE JOURNAL OF EXPERIMENTAL BIOLOGY Spiracle function in Drosophila 1669

A B C P*MR P*mech Aerodynamic P*MR IFM efficiency efficiency Total efficiency

P*mech P*RF P*mech P*RF Heat P*ind+P*pro Heat ()P*ind +P*pro 14 30 14

(%) 12 25 (%) 12

(%) 10 20 10

iciency 8 ciency 8 eff 15 6 6

ht effi

efficiency 10 4 4

flig

dynamic IFM 2 5 2

Total

0 Aero 0 0 0 1234 0 1234 0 1234 No. of open thoracic spiracles

Fig.·5. Flight efficiencies in Drosophila plotted against the number of open thoracic spiracles during flight. (A) Chemo-mechanical conversion— efficiency— of the indirect flight muscles (IFM). (B) Aerodynamic efficiency is the ratio between Rankine–Froude power estimate for flight, P*RF, and the sum of induced and profile power (Ellington, 1984b). (C) Total flight efficiency is equal to the product between muscle efficiency and aerodynamic efficiency. Results are calculated using grey and red data shown in Figs·3 and 4. See text and legend of Fig.·3 for explanations. Values and means ± s.d. (N=10 flies). unmanipulated flies to ~5.6±1.5% in flies that only breathed wing kinematics including muscle-mechanical and through a single thoracic spiracle. Interestingly, muscle aerodynamic efficiency co-vary with alterations in lift efficiency apparently recovered (7.2±1.8%) when Drosophila production – and most of the measured alterations may be only breathed through abdominal spiracles compared to gas attributed to this effect (Lehmann, 2002; Lehmann and exchange through a combined diffusive area of abdominal and Dickinson, 1997). Thus, we separated the changes due to 1 (5.6±1.5%) and 2 (6.3±1.6%) thoracic spiracles (Fig.·5A). variations in flight force production from those caused by However, this result might be partly ascribed to the delayed variations in the arrangement of spiracle conductance. This release of CO2 (see paragraph below) after flight initiation that was achieved by comparing the various measures (grey bars) produced a temporal mismatch between steady-state power in Figs·3–5 with measures from unmanipulated animals at requirements for flight and respiratory activity of the animal. flight forces that matched (within ±2% accuracy) the maximum In this sense, the apparent recovery of muscle efficiency during force of those flies whose spiracles had been manipulated. breathing through the abdominal spiracles does not reflect According to Fig.·3D, these body weight-specific forces are: changes in the physiological state of the flight musculature but 1.01 (1 thoracic spiracle blocked), 0.74 (2 thoracic spiracles results from the measuring method. blocked), 0.48 (3 thoracic spiracles blocked) and 0.29 relative In contrast, aerodynamic efficiency of wing motion, defined flight force (four thoracic spiracles blocked). In other words, as the ratio between minimum power requirements for flight while the grey bars in Figs·3–5 represent measures at 1% (i.e. Rankine–Froude power) and muscle mechanical power, maximum flight force production of manipulated flies (0–3 decreased linearly with decreasing gas exchange area from a open spiracles) and the unmanipulated control group (4 maximum of ~26.8±0.92% to ~11.2±5.3% during abdominal spiracles open), the red data (means ± s.d., Figs·3–5) were breathing (Fig.·5B, Table·1) (Ellington, 1984b). Total flight measured in unmanipulated animals at times when the flies efficiency is the product of muscle and aerodynamic efficiency produced flight forces equal to one of the five force values and a measure for the overall performance of the chemo- shown in Fig.·3D. As mentioned before in the Materials and aerodynamic conversion process of Drosophila’s flight methods, the fruit flies varied flight force production between apparatus. During flight of the tiny fruit fly, total flight maximum (1.36 force/weight) and minimum (0.29 efficiency decreased from ~2.61±0.35% in unmanipulated force/weight) values in response to the vertically oscillating animals to a value well below 1% (~0.77±0.30%) during pure horizontal background stripe grating displayed inside the abdominal breathing (Fig.·5C). virtual-reality background flight arena. To further highlight the effect of changes in local oxygen supply to the flight muscles, Spatial distribution of spiracle opening areas and flight we subtracted these data (red) from the results obtained during performance spiracle manipulation (grey data, Figs·3–5) and then plotted the Most of the changes in wing kinematics, flight power most essential differences as a function of spiracle opening requirements, metabolic rates and flight efficiencies shown in area in Fig.·6. Figs·3–5 (grey data) can be explained by the changes in flight We found quite complex dependencies between the various force production. It has previously been demonstrated that flight variables and diffusive spiracle area given by the number

THE JOURNAL OF EXPERIMENTAL BIOLOGY 1670 N. Heymann and F.-O. Lehmann

Fig.·6B). However, muscle efficiency appeared to be 150 240 significantly higher during pure abdominal breathing (5% of 180 100 total spiracle opening area) when compared with 120 unmanipulated animals (t-test, P<0.001, Fig.·6B). The

mech

*

P 50 tendency towards higher relative muscle efficiencies at smaller

Δ 60 respiratory exchange areas suggests that even the worst 0 Δ 0 condition for tracheal oxygen supply (abdominal breathing)

RelativeA (%) B does not impair the IFM chemo-mechanical conversion –50 –60 Relative IFM efficiency (%) ϫ –3 2 024681012 024681012 efficiency (linear regression fit, y=63.2–7.6 10 x, R =0.41, N=5, P=0.24). (3) Relative lift coefficient significantly 40 20 increased with increasing diffusive area (linear regression fit, ϫ –3 2 0 y=–68.6+6.84 10 x, R =0.97, N=5, P=0.002, Fig.·6C). 0 During pure abdominal breathing, the lift coefficient was –20

L

C ~54% of the lift coefficient determined in unmanipulated Δ –40 –40 animals producing similar aerodynamic force. (4) With

Δ –60 increasing spiracle opening area, aerodynamic efficiency –80 efficiency (%) –3 –2 Relative (%) ϫ ␮ Rel. Aerodynamic –80 increases significantly with a slope of 5.37 10 · m (linear –3 2 –120 C –100 D regression fit, y=–55.7+5.37ϫ10 x, R =0.97, N=5, P=0.0026, 024681012 024681012 Fig.·6D), suggesting that oxygen supply through all thoracic Diffusive area (1032μ m ) and abdominal spiracles produces the best aerial performance score in Drosophila (Fig.·6D). Aerodynamic efficiency during Fig.·6. Significance of the spatial distribution of spiracle exchange pure abdominal breathing was ~48% below the maximum areas for (A) IFM mechanical power output, (B) flight muscle value obtained in unmanipulated flies. In conclusion, the latter efficiency, (C) lift coefficient, and (D) aerodynamic efficiency. Data results suggest that manipulations of the arrangement of show the relative difference (⌬) in performance between spiracle conductance are more likely to cause subtle alterations unmanipulated flies and animals in which up to 4 thoracic spiracles have been sealed during flight. The differences are scaled to the in wing motion that alter aerodynamic efficiency, than performance of the unmanipulated control group. Performance scores significantly changing the chemo-mechanical conversion are plotted against total diffusive area of the animal’s abdominal and efficiency of the asynchronous flight musculature. thoracic spiracles that may participate in tracheal gas exchange. Due to the reduction of maximum flight force production with decreasing Difference between meso- and metathoracic oxygen supply total spiracle opening area, data are calculated at 0.29 (493·␮m2 total To evaluate the functional differences between meso- and spiracle area), 0.48 (3780·␮m2), 0.74 (5424·␮m2), 1.01 (7889·␮m2) metathoracic spiracle-mediated respiration, we compared and 1.36 (10 355·␮m2) relative flight force production for 0–4 thoracic kinematic, aerodynamic and energetic variables in experiments spiracles open, respectively. A value of 1.0 normalized force means in which we sealed either the two anterior meso- or the two that the fly produces a flight force equal to body weight. Grey areas caudal metathoracic spiracles. Fig.·7 shows the relative in the pictograms indicate maximum total spiracle opening area performance differences (‘meso- – meta-’) normalized to the available for respiratory gas exchange and red lines indicate linear regression fits. See legend of Fig.·3 for number of tested flies and text absolute performance measured in unmanipulated flies. The for more explanations. Values are means ± s.d. metathoracic spiracle opening area is ~26% larger than the mesothoracic spiracle opening area (Fig.·1), so we expected an ~26% higher contribution of the metathoracic spiracle to the of open spiracles. The major results can be summarized as overall flight performance score. Although the data in Fig.·7 follows: (1) Flight force- and body mass-specific mechanical confirmed this hypothesis, yielding a mean relative difference power increased when gas exchange between the tracheal of –23.6±12.6% (N=12 measures), the strength in reduction system and the ambient air was forced through a decreasing slightly differed between the various flight measures. Except number of thoracic spiracles (linear regression fit, y=60.0 for stroke amplitude, stroke frequency and metabolic power, –6.0ϫ10–3x, R2=0.82, N=5, P=0.03, Fig.·6A). Pure abdominal all values were significantly different from zero (t-test, P<0.01, breathing apparently permitted up to ~64% higher muscle Fig.·7). mechanical power output than breathing through all 18 spiracles in unmanipulated flies (at 0.29 relative flight force). CO2 buffer capacity The relative changes in muscle mechanical power output at The results in Figs·3–7 represent data measured 5·s after 493, 3780 and 5424·␮m2 spiracle opening areas were flight initiation and 2·s before flight stop of a flight sequence significantly different from those at 7889 and 10 355·␮m2 (t- (see Materials and methods). A previous study has shown that test, P<0.05, Fig.·6A). (2) Compared to unmanipulated during closed-loop conditions in a flight arena, tethered fruit animals, relative muscle efficiency did not change significantly flies produce a pronounced force peak immediately after flight when oxygen supply was restricted to 3 (7889·␮m2) and 2 initiation (Lehmann and Dickinson, 1997). In our experiments (5424·␮m2 opening area) thoracic spiracles (t-test, P>0.05, total flight force diminished after this initial peak to a steady

THE JOURNAL OF EXPERIMENTAL BIOLOGY Spiracle function in Drosophila 1671

h

c

R A B e C

M

m

ind

*

*

*

*

pro

⌽ P

P ␩␩␩

P Fig.·7. Relative differences (a–b) between 10 n CL CD FT/wb 20 P 10 MTA oxygen supply via (a) meso- and (b) 0 0 metathoracic spiracles of (A) wing kinematics, 0 –10 (B) power requirements for flight and metabolic –10 power, and (C) flight efficiency. Differences are –20 –20 –20 scaled to the flight performance of –30 –40 –30 unmanipulated animals that could breathe –40 –40 through all thoracic and abdominal spiracles. See –60 –50 –50

text, legends of Figs·3–5 and list of symbols and Performance change (%) abbreviations for more explanation. –60 –80 –60

state value that was typically equal to hovering flight force and characteristically found a transient steep decrease in CO2 thus equal to the animal’s body weight (Fig.·8A, black trace). release (Fig.·9D, arrow) followed by a more moderate decrease The force peak, including the corresponding peak in CO2 during which flies released the remaining tracheal CO2 over a release rate, ceased completely when respiratory flux was time period of up to 40·s (Fig.·10A). limited to ~50% total thoracic spiracle opening area (Fig.·8, red Carbon dioxide buffering and delayed release produced by trace). During pure abdominal breathing, we found a transient nine flight starts and stops of a single fly are shown in Fig.·9A,B. mismatch between force production and CO2 release rate, Averaged data for six flies are plotted in Fig.·9C,D. The temporal suggesting that CO2 is partly buffered in the tracheae and dynamics of CO2 release rate after flight stop suggests a first haemolymph at this early stage of the flight sequence (Fig.·8, order exponential decay with a mean time constant of 14.6·s –x/14.6 2 2 blue trace). At flight stop, the buffered CO2 was then (y=–2.3+e , R =0.92, ␹ /d.f.=1.49; Fig.·10A, red line). This apparently released during the subsequent resting period is consistent with the assumption of partial pressure gradient- (Figs·8, 9). Moreover, at the instant after flight stop, we driven diffusive respiratory processes in the fruit fly. For comparison, in unmanipulated flies CO2 release rate after a flight sequence returned back to 67% of the resting rate within A ~1.89±1.08·s (time constant of first order exponential fit, y=–0.11+e–x/1.89, R2=0.79±0.15, ␹2/d.f.=4.86±3.0, N=25 fits, 9 Hovering force unmanipulated animals). Approximately half of this time, however, is due to the wash-out time constant of the 10 N respirometric chamber (0.9·s; see Materials and methods). The temporal integral of post-flight CO2 release rate allowed us to estimate the combined in vivo haemolymph–tracheal buffer capacity for CO2 in Drosophila. We plotted these values B in Fig.·10B as a function of mean flight force that was measured within the last 2·s prior to rest (pre-resting force). According to our analysis, mean CO2 buffer capacity of Drosophila amounted to approximately 33.5±13.9·␮l·g–1·body·mass (N=35 sequences, blue line, Fig.·10B). Interestingly, this buffer 0.5 p.p.m. CO capacity did not significantly change with increasing pre-resting 2 flight force production or power requirements for flight as 60 s shown by linear regression analysis (linear regression fit on pre- resting force, P=0.46, N=35 flight sequences, Fig.·10B). Rest Flight* Flight *

Fig.·8. Flight force development and CO2 release dynamics during Discussion flight initiation in three different respiratory conditions. (A,B) Flight In the present study we have focused on the impact of force development (A) and respiratory gas release (B) in three single tracheal gas exchange on flight muscle function and fruit flies starting flight from rest. Black trace, unmanipulated flying aerodynamic performance in tethered flying fruit flies animal; red trace, metathoracic spiracles sealed on both body sides; Drosophila. The most prominent results may be summarized blue trace, all thoracic spiracles sealed. *Resting periods of as follows. (1) Maximum locomotor performance can only be Drosophila, in which gas exchange was restricted to abdominal achieved when oxygen is delivered through all thoracic and breathing (blue). CO2 release rate is given in parts per million (p.p.m.) analysed air. Arrow indicates stimulus artefact (under pressure peak) abdominal spiracles. Hovering performance requires gas resulting from the experimental procedure we used to elicit flight. exchange through at least three thoracic and all abdominal Vertical grey area indicates the time in which the unmanipulated fly spiracles. Linear regression analysis suggests that maximum transiently produced forces in excess of hovering force. spiracle opening area apparently matches the metabolic need

THE JOURNAL OF EXPERIMENTAL BIOLOGY 1672 N. Heymann and F.-O. Lehmann

A C Fig.·9. Dynamics of flight force 0.6 0.3 production and flight-specific CO2 Flight Rest release at flight initiation (A,B) and Rest Flight Rest Drosophila 0.4 0.2 flight end (C,D) of in which respiratory gas exchange was limited to gas flux through 0.2 0.1 abdominal spiracles. (A) Flight force production and (B) CO release, in a 0 2 0 single fly exhibiting nine successive Flight force/weight flight sequences (superimposed –0.2 –0.1 coloured lines). Note the time course of CO2 release after flight initiation B D and during the post-flight respiration 25 period. Time after x-scale break ) 40 Rest Stimulus Flight Rest Rest

–1 Flight artefact indicates mean time of all nine h 30 20 –1 sequences. (C,D) Dynamics of force production (C) and CO release (D) 15 2 20 at flight stop averaged over 65 flight 10 10 sequences derived from six flies. Mean values are plotted in black;

release (ml g 2 0 5 grey area indicates s.d. Flight time

CO was 49.9±65.7s (mean ± s.d.). Arrow –60 0 indicates a transient steep decrease in 0246810 52 54 56 58 60 62 024681012 respiratory CO2 release immediately Flight time (s) Flight time (s) after flight stop. of the animal at maximum locomotor performance (Figs·3–5). and aerodynamic efficiency (Figs·3E, 4B, 5B, 6). All those (2) The chemo-mechanical conversion efficiency of the flight measures significantly decrease with decreasing spatial musculature is broadly independent of the arrangement of homogeneity of tracheal oxygen supply. (4) The delay of CO2 thoracic spiracle conductance (Fig.·5A). (3) We found that the release after flight stop in flies in which breathing was spatial distribution of tracheal gas exchange areas strongly restricted to abdominal gas exchange revealed that the influences lift coefficient, profile power requirements for flight, combined CO2 buffer capacity of the tracheal dead space and the haemolymph’s bicarbonate level amounts to ~33.5·␮l·g–1·body·mass.

30 ) 80 ) Flight Rest –1 Spiracle conductance and tracheal partial pressure

–1

l g h for CO

␮ 2

–1 20 60 CO2 buffer Although tethered flying Drosophila sporadically capacity –x/␶ 10 y= ke 40 employ ventilation to manipulate tracheal gas flow, diffusion is still to be considered the main type of respiration in fruit flies, and diffusive theory may be

release (ml g 0 20

2

buffer capacity ( applied (Kestler, 1985; Lehmann, 2001; Lehmann and

2

CO –10 A 0 B Heymann, 2005; Weis-Fogh, 1964a). Together with

CO 0 10203040 0 0.1 0.2 0.3 our experimental data, the analytical framework Time (s) Pre-resting flight force/weight conveniently permits estimations of spiracle conductance and partial pressures for CO2 inside the Fig.·10. Combined tracheal–haemolymph buffer capacity for CO2 in tracheal system, and thus estimations of the animal’s Drosophila estimated from the amount of flight-specific CO2 released after physiological stress resistance to high tracheal CO2 flight stop. Respiration was limited to gas exchange through the abdominal concentration during flight. Tracheal partial pressure spiracles (thoracic spiracles sealed). (A) In a single fly, CO2 release decayed for a gas, PT, that diffuses through a spiracle opening exponentially after flight stop and approached zero after ~40·s. The area under can be expressed as: the curve = CO2 buffer capacity of the tracheal system and haemolymph (light –1 grey). Red line represents first-order exponential fit to data. (B) CO2 buffer PT = MG + PA·, (1) capacity per gram body mass (35 flight sequences, 6 flies) derived from estimations of the area under curve after flight stop (as shown in A). Data are in which M is the rate of gas flux, G is the conductance, plotted as a function of pre-resting flight force produced during the last 2·s and PA is the partial pressure of the gas in the ambient before the animals ceased to fly. Blue line indicates mean value ± s.d. (shaded air (Kestler, 1985). The conductance for gas flow in grey). turn depends on tracheal geometry, the diffusion

THE JOURNAL OF EXPERIMENTAL BIOLOGY Spiracle function in Drosophila 1673 coefficient D and the capacitance coefficient ␤ for that gas. partially cope with the reduction in spiracle opening area by This relationship can be expressed as: increasing the partial pressure gradient for CO2 between tracheal system and ambient air of up to 5–6·kPa. G = (A /L )D␤·, (2) T T In comparison, since ambient partial pressure for oxygen is in which AT is total spiracle exchange area and LT is the typical constant, oxygen uptake rate in Drosophila can only be tracheal length. Previous experiments on Drosophila have reinforced by ventilation or by actively lowering the tracheal shown that LT can be approximated as 111·␮m, and AT was partial pressure for oxygen (Wigglesworth, 1972). For calculated from our histological measurements in Fig.·1 example, it has recently been shown that during hovering flight (Lehmann, 2001). The diffusion coefficient for CO2 is force production, tethered Drosophila sporadically employ the 0.165·cm2·s–1, and the capacitance coefficient is proboscis as a pump to actively ventilate their tracheal system 410.5·nmol·cm–3·kPa–1 (Kestler, 1985). From the values and (Lehmann and Heymann, 2005). A possible mechanism for equations above, we calculated the maximum spiracle active buffering of tracheal oxygen is haemoglobin storage. conductance for our various experimental conditions as: 27.9 This phenomenon is discussed in the paragraph on ‘the (unmanipulated animal), 21.2 (1 thoracic spiracle sealed), 14.6 significance of CO2 buffer capacity’. (2 thoracic spiracles sealed), 10.2 (3 thoracic spiracles sealed) There are comparatively few data on tracheal partial and 1.32·ng·gas·kPa–1·s–1 (four thoracic spiracles sealed). pressure estimates of respiratory gases in flying insects, and Previous analyses of tracheal partial pressure in diffusion- most of the available data refer to the DGC. For example, based respiratory systems relied on the assumption that Harrison et al. (Harrison et al., 1995) reported for grasshoppers maximum spiracle opening area in an insect matches the that during the DGC interburst period, haemolypmph PCO2 instantaneous gas exchange rate at maximum locomotor rises from 1.8 to 2.26·kPa with minimal acidification of the performance (Lehmann, 2001). Originally, this assumption haemolymph. The authors concluded that spiracle opening is was reinforced by the finding that most insects have developed induced at internal threshold levels between 2 and 2.9·kPa strategies to avoid respiratory water loss through the open (Gulinson and Harrison, 1996). Moreover, it has previously spiracles, i.e. the discontinuous gas exchange cycle, DGC been shown that in many insects an endotracheal partial (Harrison and Roberts, 2000; Lighton, 1994; Lighton, 1996; pressure of ~4–6·kPa triggers the peripherally mediated Miller, 1981; Slama, 1994; Snyder et al., 1995). Gas exchange inactivation of the spiracle closer muscle (for reviews, see areas larger than needed would reinforce tracheal water loss at Lighton, 1996; Krogh, 1913). The highest values reported for maximum locomotor performance and thus increase the risk of PT,CO2 during the DGC was for Cecropia pupae of ~7·kPa that desiccation in xeric environments (Lighton, 1994). The data in did not fall below a minimum threshold of about 3.6·kPa Fig.·3D provide support for this hypothesis and our (Burkett and Schneiderman, 1974). The overall shift of PT,CO2 experiments thus permit derivation of maximum tracheal in Lepidopteran pupae towards higher tracheal partial partial pressure for CO2 (PT,CO2) under the various breathing pressures has been interpreted as a consequence of the high conditions. Employing the equations above and setting spiracle vulnerability of the pupae to desiccation (Harrison et al., 1995). conductance according to the various maximum spiracle If we consider that in an insect a high PT,CO2, due to opening areas, we estimated the following mean values for prolonged DGC interburst intervals, indicates a strategy to PT,CO2 inside the tracheal system during flight: 0.98±0.23 avoid respiratory water loss, the comparatively low PT,CO2 (unmanipulated animal), 0.83±0.17 (1 thoracic spiracle measures in unmanipulated flying Drosophila would in turn sealed), 1.20±0.25 (2 thoracic spiracles sealed), 1.47±0.24 (3 suggest that in this insect respiratory water loss is of minor thoracic spiracles sealed) and 5.52±0.36·kPa (4 thoracic importance for total water balance. This interpretation is driven spiracles sealed). A slightly higher value for tracheal partial by the finding that in several insects, respiratory water loss is pressure of CO2 (1.4±0.2·kPa) at maximum flight performance relatively small compared to cuticular transpiration so that the in unmanipulated Drosophila was calculated in a study on DGC does not predominantly influence the water balance of spiracle control strategies (Lehmann, 2001). the animal. For example, the ratio between cuticular and The above results demonstrate that according to total respiratory transpiration is 98.1:1.9 in Camponotus vicina and spiracle opening area, mean PT,CO2 may increase 92.0:8.0 in the ant Cataglyphis bicolor (Lighton, 1988), approximately 1.5-fold from ~0.98·kPa in unmanipulated flies 87.0:13.0 in the ant Pogonomyrmex rugosus (Lighton et al., (10 355·␮m2 opening area) to 1.47·kPa in flies in which total 1993b) and the cockroach Periplaneta americana (Machin gas exchange area was limited to 3780·␮m2. A gas exchange et al., 1991), 97.0:3.0 in the grasshopper Romalea guttata 2 area of 493·␮m even produces a PT,CO2 that is 6.6 times higher (Hadley and Quinlan, 1993) and 95.4:4.6 in the grasshopper than in unmanipulated flies. Although statistical analysis Taeniopoda eques (Quinlan and Hadley, 1993). However, in reveals that all estimations of PT,CO2 are significantly different flying Drosophila melanogaster this ratio appears to be almost from each other (t-test, P<0.05), a linear regression fit suggests inverted and amounts to ~17.4:82.6 (Lehmann, 2001). From that the slope between PT,CO2 and gas exchange area is not these values we conclude that the spiracles represent a significantly different from zero (linear regression fit, significant route for tracheal water loss in the fruit fly, and with y=4.38–4.26ϫ10–4x, R2=0.66, P=0.09, N=5). Nevertheless, the respect to the hypothesis above, this cannot easily account for small trend in the data set suggests that Drosophila is able to the low PT,CO2 in the unmanipulated flying animal.

THE JOURNAL OF EXPERIMENTAL BIOLOGY 1674 N. Heymann and F.-O. Lehmann

The significance of CO2 buffer capacity Nevertheless, we should keep in mind that this conclusion, The dead volume of the tracheal system serves as a buffer among the other results in this study, may critically depend on space for respiratory gases. In addition, Drosophila may store the simple assumption that respiratory gas exchange in oxygen using haemoglobin that is mainly synthesized in the Drosophila relies on diffusion alone and can be described by tracheal walls and the fat body of the animal (de Sanctis et al., the simple analytical model above. This approach neglects any 2005; Hankeln et al., 2005). During transient locomotor activity, dynamics resulting from both buffering of respiratory gases the indirect flight musculature might benefit from haemoglobin- and tracheal ventilation. In addition, there are potential errors mediated oxygen transport and storage that produces a temporal associated with our measurement technique including the mismatch between the uptake rate of oxygen through the difficulty to temporally match locomotor performance and CO2 spiracles and flight force production. To largely circumvent this release rate of the flying animal. The outcome of this study problem in our respiratory measurements, we excluded the first should thus be regarded with care and direct measurements of 5·s in each flight sequence from our analysis, expecting to tracheal partial pressures at the various experimental achieve rather steady-state flight metabolism of 10–15 times the conditions have still to verify our predictions. resting rate (Casey, 1989; Casey and Ellington, 1989; Lehmann and Dickinson, 1997). By contrast, tracheal CO2 concentration Significance of spatial distribution of tracheal gas exchange is correlated with the haemolymph bicarbonate level that buffers areas on muscle function and lift production CO2 at the expense of changes in pH (Harrison et al., 1995). One of the most unexpected results of the present study is Gulinson and Harrison investigated CO2 buffering in the the finding that muscle efficiency changes only slightly in grasshoppers Romalea guttata and Schistocerca americana by response to manipulations of the spatial distribution of spiracle injections of NaHCO3, HCl and NaOH into the haemolymph exchange areas. The pronounced relative difference in muscle (Gulinson and Harrison, 1996). In our experiments, by contrast, efficiency of 143±60.3% at 0.29 relative flight force we derived CO2 buffer capacity from the delay in gas release production appears to be an exception and probably is partly after flight stop (Fig.·10). Since the amount of total flight- due to the fly’s CO2 buffering capacity (Figs·6B, 10). specific CO2 released after flight was independent of pre-resting Nevertheless, we found a trend in the data set suggesting that flight force production and thus of metabolic activity, we suggest relative muscle efficiency is inversely correlated with spiracle –1 that the value of 33.5·␮l·CO2·g ·body·mass may represent a diffusive area (Fig.·6B). The same trend is also visible in the maximum estimation of Drosophila’s total CO2 buffer capacity second muscle physiological parameter, i.e. mechanical power (tracheal and haemolymph buffer). output (Fig.·6A). Note that these results run counter to the To assess the flight time during which an unmanipulated results on the relative lift coefficient (Fig.·6C), relative drag fruit fly may rely on CO2 buffering instead of CO2 release coefficient (data not shown) and relative aerodynamic through the spiracles, we converted the value of efficiency (Fig.·6D), which taken together show a significant –1 33.5·␮l·CO2·g ·body·mass into units of time that yielded decrease in magnitude with decreasing spiracle exchange area. 2.30±0.95·␮l·s–1·g–1·body·mass. Subsequently, we compared this In sum, the findings above suggest that changes in the spatial measure with the mean CO2 release rate produced during distribution of tracheal oxygen supply due to local blocking of hovering flight conditions, that is 2.8·␮l·s–1·g–1·body·mass individual spiracles negatively affect the ability of the animal (Lehmann et al., 2000). The ratio between both values (2.3/2.8) to produce flight force – but reinforce the production of suggests that CO2 buffer capacity ensures flight for only ~0.82·s mechanical power and the efficiency of the mechano-chemical and thus for a relatively short flight time. Moreover, this value conversion process of the indirect flight musculature. might explain why the large thoracic spiracles open immediately The results above are rather surprising because in after flight initiation in this insect (Fig.·8) (Lehmann, 2001). unmanipulated fruit flies, muscle efficiency decreases with In a resting fruit fly that exclusively breathes through the decreasing flight force production (Lehmann, 2002). It has been abdominal spiracles, a tracheal CO2 partial pressure threshold suggested that low muscle efficiency either reflects a decreased of 5.52·kPa (see previous paragraphs) would be reached crossbridge activation between actin and myosin filaments or within 3.1·s at the given CO2 buffer capacity (resting an unfavourable strain regime of the asynchronous flight muscle metabolism=0.74±0.33·␮l·s–1·g–1·body·mass) (Lehmann et al., (Josephson, 1999; Josephson et al., 2001; Lehmann and 2000). Ignoring all potential errors associated with these findings, Dickinson, 1997). Therefore, we originally hypothesized that at the latter result might explain why Drosophila melanogaster flight forces below maximum performance, an inhomogeneous rarely employs a clear DGC pattern during rest compared to supply of oxygen to the IFM would even reinforce this many other insects (Harrison et al., 1995; Lighton, 1994; Lighton, attenuation in crossbridge activation. Instead, our data 1996; Williams and Bradley, 1998). During discontinuous apparently show that the geometry of the tracheal system and breathing the spiracles open only sporadically with typical time the location of gas exchange areas (spiracles) are of minor periods in the range of several minutes. Within a DGC interburst importance for IFM overall efficiency. A possible explanation interval of several minutes, however, Drosophila’s PT,CO2 would for this finding might be the fact that the large air sacs of the exceed the critical threshold value of 4–6·kPa several-fold, dorsal and lateral tracheal system [pleural-, notopleural-, suggesting that the fruit fly must allow continuous gas exchange lateroscutal- and medioscutal sacs (Demerec, 1965)] through the thoracic spiracles even during rest. homogenize oxygen concentration within the fly body.

THE JOURNAL OF EXPERIMENTAL BIOLOGY Spiracle function in Drosophila 1675

However, if this explanation is true, it still remains puzzling could thus explain, for example, why the mean drag coefficient why muscle mechanical power output so strongly depends on may relatively increase while the mean lift coefficient decreases the changes in the arrangement of spiracle conductance. during flight (Lehmann, 2004). A detailed reconstruction of the We also unexpectedly found that relative mean lift wing kinematic pattern by means of high-speed video technique coefficient and aerodynamic efficiency differ up to should allow us to tackle this hypothesis for the underlying approximately 50% between flies facing alterations of oxygen aerodynamic mechanisms in the future. supply distribution and unmanipulated animals (Fig.·6C,D). By The clap-and-fling mechanism for lift enhancement contrast, kinematic variables such as stroke amplitude were supposedly may not contribute to the relative change of lift-to- affected relatively little, and stroke frequency was even widely drag ratio in Drosophila, because its occurrence changes the indistinguishable between unmanipulated and spiracle- lift-to-drag ratio only slightly, from 0.57 to 0.58 (Ellington, manipulated flies at the various flight forces (1.44±11.7·Hz, 1975; Lehmann et al., 2005; Weis-Fogh, 1973). Altogether, the mean difference ± s.d., N=5 forces, Fig.·3, Table·1). Previous results suggest that respiratory gas exchange based on the studies on the mechanisms of unsteady aerodynamics in usage of multiple thoracic and abdominal spiracles appears to flapping insect wings have shown that flight force production be beneficial for maintaining an elevated efficacy of linearly decreases with decreasing wing velocity, thus aerodynamic force production in the fruit fly. The ultimate following conventional aerodynamic laws (Ellington, 1984a; explanation for this finding might be that changes in the Lehmann and Dickinson, 1998). Apparently, wing velocity (the arrangement of spiracle conductance alter the contraction product between stroke amplitude and frequency) in dynamics or muscle stiffness of the IFM and thus the unmanipulated animals decreases to a greater extent than in flies movements of the mechanical thoracic oscillator (Josephson, in which flight force reduction is forced by a reduction in 1999; Josephson et al., 2001; Josephson and Stokes, 1999; diffusive exchange area. As a consequence, mean lift Vigoreaux, 2001; Vigoreaux et al., 2000). These parameters coefficient in unmanipulated flies varies less dramatically with were not covered by our measurement technique. changing flight forces than in the spiracle-manipulated animals (Fig.·3E). From these findings, we hypothesize that alterations An alternative explanation in the spatial distribution of gas exchange areas ultimately alter Alternatively, we should consider whether the above results the fine structure of wing motion such as the angle of attack, are simply due to the different mechanisms used by the flies the velocity profile during wing translation or the wing’s for modifying flight force production. While unmanipulated rotational speed and timing during the stroke reversals. For flies ‘voluntarily’ altered flight forces in response to the example, it has been shown that in a robotic model of vertical motion of the visual lift stimulus, the spiracle-blocked Drosophila, changes in wing rotation may alter both flight costs flies probably reduced flight force due to the restriction in and lift production (Dickinson et al., 1999; Sane and Dickinson, mechanical power output of the flight musculature. 2001). An 8% advanced rotational timing, during which the Consequently, our findings can be interpreted such that wing rotates prior to the stroke reversal, may reinforce changes in wing kinematics mediated by active control of flight aerodynamic force production by more than 70% of total force, steering muscles produce more favourable stroke kinematics compared to a delayed rotation that occurs at the beginning of than a passive change via mechanical power limits (Götz, each half stroke (Dickinson et al., 1999). Somewhat smaller 1983; Heide and Götz, 1996). Numerous kinematic and increases in lift production have been reported for increases in electrophysiogical studies have shown that 17 flight control the wing’s angular velocity during rotation (Sane and muscles (steering muscles) tune several aspects of wing motion Dickinson, 2002). Changes in rotational speed and timing, during manoeuvring flight, such as stroke amplitude, stroke moreover, may alter the benefit of the wake capture mechanism frequency, the timing of wing rotation, angle of attack or the and clap-and-fling lift enhancement that also contribute to wing trajectory in Drosophila (Götz, 1983; Heide and Götz, Drosophila’s high lift coefficient (Lehmann et al., 2005). 1996; Lehmann and Götz, 1996) (for a review, see Dickinson Besides the changes in relative lift coefficient, the findings and Tu, 1997). In this case, the data plotted in Fig.·6 would in Figs·3F, 4B, 5B suggest that changes in the arrangement of highlight the energetic and aerodynamic consequences of wing spiracle conductance may also decrease the ratio between lift kinematic alterations due to different flight control strategies and drag coefficients and thus increase the relative profile rather than reflect the significance of respiratory constraints. It power requirements for wing flapping. However, since lift and appears to be difficult to distinguish unambiguously between drag forces are vectors of total flight force, an attenuation in lift both interpretations in our experiments because changes in the would likely cause a relative decrease in drag rather than a arrangement of spiracle conductance also involve changes in relative increase. The same 3-dimensional robotic model wing total diffusive area. However, the comparison between gas flux of Drosophila mentioned above has demonstrated a tremendous through the meso- and metathoracic spiracles should be noted increase in wing drag towards higher angles of attack as an exception (Fig.·7). As already mentioned, the mean (Dickinson et al., 1999; Usherwood and Ellington, 2002). While difference in the performance measures of 26% approximately at angles of attack <45° lift and drag coefficients are positively matches the difference in spiracle opening area between meso- correlated, lift and drag coefficient are inversely correlated at and metathoracic spiracles, which suggests a negligible 2.4% angles >45°. Increases in angle of attack above this threshold difference in oxygen supply rate between the meso- and or

THE JOURNAL OF EXPERIMENTAL BIOLOGY 1676 N. Heymann and F.-O. Lehmann ——— metathoracic spiracle. However, due to the large ipsilateral P*mech mean flight-specific and flight muscle mass-specific tracheal trunks and air sacs, this ipsilateral effect was expected mechanical power —— to be small a priori (Miller, 1950). P*MR mean flight-specific and flight muscle mass-specific metabolic power Conclusions —— P*pro mean flight-specific and flight muscle mass-specific The in-depth evaluation of the significance of tracheal gas profile power —— exchange in Drosophila potentially provides several new insights P*RF mean flight-specific and flight muscle mass-specific onto how the spatial distribution and the size of spiracle exchange Rankine–Froude power areas determine the function of the flight motor in a flying insect. PT tracheal partial pressure of a gas The present results provide direct evidence for the general SEM scanning electron microscopy assumption in respiratory research that the tracheal development sp tracheal and abdominal spiracles of a simple diffusion-based system matches the respiratory need STP standard temperature and pressure at maximum metabolic activity of the animal. Under those wb body weight of the animal conditions, respiratory water loss would be minimal, which in ⌽ wing stroke amplitude turn prevents water stress on animals living in xeric environments ␤ capacitance coefficient (Lighton, 1994; Lighton, 1996). Moreover, our findings ␩A aerodynamic efficiency apparently show that changes in the arrangement of spiracle ␩M muscle efficiency conductance primarily effect aerodynamic phenomena in ␩T total efficiency of flight motor addition to flight muscle mechanical power output, but not ␶ time constant predominantly muscle efficiency. The exact reason why relative muscle mechanical power output depends more strongly on the We like to thank the two anonymous referees for their spatial distribution of spiracle areas than muscle efficiency helpful comments and Ursula Seifert for carefully reading the remains unknown and will require further research on the indirect manuscript. This work was generously funded by the flight musculature in the behaving animal. Since it has been BioFuture grant 0311885 of the German Federal Ministry for assumed that insects have no anaerobic capacity, the magnitude Education and Research to F.O.L. of oxygen and CO2 buffer capacity might play a crucial role in breathing behaviour and spiracle control in the fruit fly (Ziegler, References 1985). On the one hand, CO2 buffer capacity may explain why Bailey, L. (1954). The respiratory currents in the tracheal system of the adult spiracles have to open immediately after flight initiation and honey bee. J. Exp. Biol. 31, 589-593. match their opening area to metabolic need (Lehmann, 2001). On Barton, B., Ayer, G., Heymann, N., Maughan, D. W., Lehmann, F.-O. and Vigoreaux, J. O. (2005). Flight muscle properties and aerodynamic the other hand, the small CO2 buffer capacity might also partly performance of Drosophila expressing a flightin gene. J. Exp. Biol. 208, explain why inbred lines of Drosophila reared on commercial 549-560. food do not exhibit a clear DGC pattern during resting Birch, J. M. and Dickinson, M. H. (2001). Spanwise flow and the attachment of the leading-edge vortex on insect wings. Nature 412, 729-733. metabolism (Williams and Bradley, 1998). Eventually, in Burkett, B. and Schneiderman, H. A. (1974). Role of oxygen and carbon conjunction with the small safety margin for tracheal respiration dioxide in the control of spiracular function in Cecropia pupae. Biol. Bull. at maximum locomotor capacity, we conclude that Drosophila Mar. Biol. Lab. Woods Hole 147, 274-293. Casey, T. M. (1989). Oxygen consumption during flight. In Insect Flight (ed. may apparently maximize the efficiency of its locomotor system G. J. Goldsworthy and C. H. Wheeler), pp. 257-272. Boca Raton: CRC for flight by well-balancing respiratory gas flow between the four Press. large spiracles in the fly’s thorax. Casey, T. M. and Ellington, C. P. (1989). Energetics of insect flight. In Energy Transformations in Cells and Organisms (ed. W. Wieser and E. Gnaiger), pp. 200-210. Stuttgart: Thieme. de Sanctis, D., Dewilde, S., Vornhein, C., Pesce, A., Moenz, L., Ascenzi, List of symbols and abbreviations P., Hankeln, T., Burmester, T., Ponassi, M., Nardini, M. et al. (2005). Bis-histidyl heme hexacoordination, a key structural property in Drosophila A total spiracle exchange area —T melanogaster hemoglobin. J. Biol. Chem. 280, 27222-27229. CD mean drag coefficient for wing flapping Demerec, M. (1965). Biology of Drosophila. New York, London: Hafner — Publishing Company. CL mean lift coefficient for wing flapping Dickinson, M. H. and Lighton, J. R. B. (1995). Muscle efficiency and elastic D diffusion coefficient storage in the flight motor of Drosophila. Science 268, 87-89. DGC discontinuous gas exchange cycle Dickinson, M. H. and Tu, M. S. (1997). The function of Dipteran flight muscle. Comp. Biochem. Physiol. 116A, 223-238. FT total flight force Dickinson, M. H., Lehmann, F.-O. and Sane, S. (1999). Wing rotation and G spiracle conductance the aerodynamic basis of insect flight. Science 284, 1954-1960. IFM indirect flight muscle Duncan, F. D. and Byrne, M. J. (2002). Respiratory airflow in a wingless dung beetle. J. Exp. Biol. 205, 2489-2497. LT typical tracheal length Ellington, C. P. (1975). Non-steady-state aerodynamics of the flight of M rate of gas flux Encarsia formosa. In Swimming and Flying in Nature, vol. 2 (ed. T. Y. Wu, n wing stroke frequency C. J. Brokaw and C. Brennan), pp. 783-796. New York: Plenum Press. Ellington, C. P. (1984a). The aerodynamics of insect flight. IV. Aerodynamic PA partial pressure of the gas in the ambient air —— mechanisms. Philos. Trans. R. Soc. Lond. B Biol. Sci. 305, 79-113. P*ind mean flight-specific and flight muscle mass-specific Ellington, C. P. (1984b). The aerodynamics of insect flight. VI. Lift and power induced power requirements. Philos. Trans. R. Soc. Lond. B Biol. Sci. 305, 145-181.

THE JOURNAL OF EXPERIMENTAL BIOLOGY Spiracle function in Drosophila 1677

Ellington, C. P., Machin, K. E. and Casey, T. M. (1990). Oxygen Cataglyphis bicolor: regulation of carbon dioxide release from the thoracic consumption of bumblebees in forward flight. Nature 347, 472-473. and abdominal spiracles. J. Insect Physiol. 39, 687-699. Götz, K. G. (1983). Bewegungssehen and flugsteuerung bei der fliege Lighton, J. R. B., Garrigan, D. A., Duncan, F. D. and Johnson, R. A. Drosophila. In BIONA-report 2 (ed. W. Nachtigall), pp. 21-34. Stuttgart: (1993b). Spiracular control of respiratory water loss in female alates of the Fischer. harvester ant Pogonomyrmex rugosus. J. Exp. Biol. 179, 233-244. Gulinson, S. L. and Harrison, J. F. (1996). Control of resting ventilation rate Lipp, A., Wolf, H. and Lehmann, F.-O. (2005). Walking on inclines: in grasshoppers. J. Exp. Biol. 199, 379-389. energetics of locomotion in the ant Camponotus. J. Exp. Biol. 208, 707-719. Hadley, N. F. and Quinlan, M. C. (1993). Discontinuous carbon dioxide Machin, J., Kestler, P. and Lampert, G. J. (1991). Simultaneous release in the Eastern lubber grasshopper Romalea guttata and its effect of measurements of spiracular and cuticular water loss in Periplaneta respiratory transpiration. J. Exp. Biol. 177, 169-180. americana: implications for whole-animal mass loss studies. J. Exp. Biol. Hankeln, T., Jaenicke, V., Kiger, L., Dewilde, S., Ungerechts, G., Schmidt, 161, 439-453. M., Urban, J., Marden, M. C., Moens, L. and Burmester, T. (2005). Manning, G. and Krasnow, M. A. (1993). Development of the Drosophila Characterization of Drosophila hemoglobin. Evidence for hemoglobin- tracheal system. In The Development of Drosophila Melanogaster (ed. M. mediated respiration in insects. J. Biol. Chem. 277, 29012-29017. Bate and A. M. Arias), pp. 609-686. Cold Spring Harbor: Cold Spring Harrison, J. F. and Roberts, S. P. (2000). Flight respiration and energetics. Harbor Laboratory Press. Annu. Rev. Physiol. 62, 179-205. Miller, A. (1950). The internal anatomy and histology of the imago of Harrison, J. F., Hadley, N. F. and Quinlan, M. C. (1995). Acid–base status Drosophila melanogaster. In Biology of Drosophila (ed. M. Demerec), pp. and spiracular control during discontinuous ventilation in grasshoppers. J. 420-534. New York: John Wiley. Exp. Biol. 198, 1755-1763. Miller, P. L. (1966). The supply of oxygen to the active flight muscles of some Heide, G. and Götz, K. G. (1996). Optomotor control of course and altitude large beetles. J. Exp. Biol. 45, 285-304. in Drosophila is achieved by at least three pairs of flight steering muscles. Miller, P. L. (1981). Ventilation in active and in inactive insects. In J. Exp. Biol. 199, 1711-1726. Locomotion and Energetics in Arthropods (ed. C. F. Herreid II), pp. 367- Josephson, R. K. (1999). Dissecting muscle power output. J. Exp. Biol. 202, 390. New York: Plenum. 3369-3375. Miller, P. L. (1982). Respiration. In The American Cockroach (ed. H. J. Bell Josephson, R. K. and Stokes, D. R. (1999). The force–velocity properties of and K. G. Adiyodi), pp. 87-116. London: Chapman & Hall. a crustacean muscle during lengthening. J. Exp. Biol. 202, 593-607. Nikam, T. B. and Khole, V. V. (1989). Insect Spiracular Systems. New York, Josephson, R. K., Malamud, J. G. and Stokes, D. R. (2001). The efficiency Chichester, Brisbane, Toronto: Ellis Horwood. of an asynchronous flight muscle from a beetle. J. Exp. Biol. 204, 4125- Quinlan, M. C. and Hadley, N. F. (1993). Gas exchange, ventilatory patterns, 4139. and water loss in two lubber grasshoppers: quantifying cuticular and Kestler, P. (1985). Respiration and respiratory water loss. In Environmental respiratory transpiration. Physiol. Zool. 66, 628-642. Physiology and Biochemistry of Insects (ed. K. H. Hoffmann), pp. 137-186. Ramamurti, R. and Sandberg, W. C. (2001). Computational study of 3-D Berlin: Springer. flapping foil flows. In 39th Aerospace Sciences Meeting and Exhibit, pp. 605. Krogh, A. (1913). On the composition of air in the tracheal system of some Reno, NV: AIAA. insects. Skand. Arch. Physiol. 29, 29-36. Sane, S. and Dickinson, M. H. (2001). The control of flight force by a flapping Lehmann, F.-O. (1999). Ambient temperature affects free-flight performance wing: lift and drag production. J. Exp. Biol. 204, 2607-2626. in the fruit fly Drosophila melanogaster. J. Comp. Physiol. B 169, 165-171. Sane, S. and Dickinson, M. H. (2002). The aerodynamic effects of wing Lehmann, F.-O. (2001). Matching spiracle opening to metabolic need during rotation and a revised quasi-steady model of flapping flight. J. Exp. Biol. flight in Drosophila. Science 294, 1926-1929. 205, 1087-1096. Lehmann, F.-O. (2002). The constraints of body size on aerodynamics and Schneiderman, H. A. (1960). Discontinuous respiration in insects: role of the energetics in flying fruit flies: an integrative view. Zoology 105, 287-295. spiracles. Biol. Bull. 119, 494-528. Lehmann, F.-O. (2004). Aerial locomotion in flies and robots: kinematic Slama, K. (1994). Regulation of respiratory acidemia by the autonomic control and aerodynamics of oscillating wings. Arthropod Struct. Dev. 33, nervous system (coelopulse) in insects and ticks. Physiol. Zool. 67, 163-174. 331-345. Slama, K. (1999). Active regulation of insect respiration. Ann. Entomol. Soc. Lehmann, F.-O. and Dickinson, M. H. (1997). The changes in power Am. 92, 916-929. requirements and muscle efficiency during elevated force production in the Snyder, G. K., Sheafor, B., Scholnick, D. and Farrelly, C. (1995). Gas fruit fly, Drosophila melanogaster. J. Exp. Biol. 200, 1133-1143. exchange in the insect tracheal system. J. Theor. Biol. 172, 199-207. Lehmann, F.-O. and Dickinson, M. H. (1998). The control of wing Usherwood, J. R. and Ellington, C. P. (2002). The aerodynamics of kinematics and flight forces in fruit flies (Drosophila spp.). J. Exp. Biol. revolving wings II. Propeller force coefficients from mayfly to quail. J. Exp. 201, 385-401. Biol. 205, 1565-1576. Lehmann, F.-O. and Dickinson, M. H. (2001). The production of elevated Vigoreaux, J. O. (2001). Genetics of the Drosophila flight muscle myofibril: flight force compromises flight stability in the fruit fly Drosophila. J. Exp. a window into the biology of complex systems. BioEssays 23, 1047-1063. Biol. 204, 627-635. Vigoreaux, J. O., Moore, J. R. and Maughan, D. W. (2000). Role of the Lehmann, F.-O. and Götz, K. G. (1996). Activation phase ensures kinematic elastic protein projection in stretch activation and work output of Drosophila efficacy in flight-steering muscles of Drosophila melanogaster. J. Comp. flight muscles. In Elastic Filaments of the Cell (ed. Granzier and Pollack), Physiol. 179, 311-322. pp. 237-250. London, Amsterdam: Kluwer Academic/Plenum Publishers. Lehmann, F.-O. and Heymann, N. (2005). Unconventional mechanisms Wasserthal, L. T. (2001). Flight-motor-driven respiratory air flow in the control cyclic respiratory gas release in flying Drosophila. J. Exp. Biol. 208, hawkmoth Manduca sexta. J. Exp. Biol. 204, 2209-2220. 3645-3654. Weis-Fogh, T. (1964a). Diffusion in insect wing muscle, the most active tissue Lehmann, F.-O., Dickinson, M. H. and Staunton, J. (2000). The scaling of known. J. Exp. Biol. 41, 229-256. carbon dioxide release and respiratory water loss in flying fruit flies Weis-Fogh, T. (1964b). Functional design of the tracheal system of flying (Drosophila spp.). J. Exp. Biol. 203, 1613-1624. insects as compared with the avian lung. J. Exp. Biol. 41, 207-227. Lehmann, F.-O., Sane, S. P. and Dickinson, M. H. (2005). The aerodynamic Weis-Fogh, T. (1973). Quick estimates of flight fitness in hovering animals, effects of wing–wing interaction in flapping insect wings. J. Exp. Biol. 208, including novel mechanisms for lift production. J. Exp. Biol. 59, 169-230. 2075-3092. Wigglesworth, V. B. (1972). The Principles of Insect Physiology. London: Lighton, J. R. B. (1988). Discontinuous CO2 emission in a small insect, the Methuen (Chapman & Hall). formicine ant Camponotus vicinus. J. Exp. Biol. 134, 363-376. Williams, A. E. and Bradley, T. J. (1998). The effect of respiratory pattern Lighton, J. R. B. (1994). Discontinuous ventilation in terrestrial insects. on water loss in desiccation-resistant Drosophila melanogaster. J. Exp. Biol. Physiol. Zool. 67, 142-162. 201, 2953-2959. Lighton, J. R. B. (1996). Discontinuous gas exchange in insects. Annu. Rev. Ziegler, R. (1985). Metabolic energy expenditure and its hormonal regulation. Entomol. 41, 309-324. In Environmental Physiology and Biochemistry of Insects (ed. K. H. Lighton, J. R. B., Fukushi, T. and Wehner, R. (1993a). Ventilation in Hoffmann), pp. 95-118. Berlin, Heidelberg, New York, Tokyo: Springer.

THE JOURNAL OF EXPERIMENTAL BIOLOGY