Proline Powers Pre-Flight Warm-Up In The African Fruit Beetle Pachnoda Sinuata (Cetoniinae)

Many insects require thoracic temperatures (T_th) to fly that are well above ambient temperature (T_a) and, therefore, need to raise their body temperature prior to flight either passively by basking in the sun or by endothermic warm-up (see Heinrich, 1993). Large beetles are able to elevate their T_th endothermically. For example, in dung beetles of different genera, T_th can reach 38–42 °C, and this is maintained even during terrestrial activity (Bartholomew and Casey, 1977; Heinrich and Bartholomew, 1979). Male rain beetles of the genus Plecoma fly with a T_th of approximately 38 °C, even in snowstorms, to find the flightless females (Morgan, 1987). In flower scarabs, such as Cotinus texana (Chappell, 1984) and Pachnoda sinuata (Heinrich and McClain, 1986), T_th is elevated endothermically to more than 35 °C before flight.

In many insects, endothermic heat production is achieved by muscle contraction (shivering) with or without wing beats (see Heinrich, 1993). Other authors suggest that heat can also be generated by non-shivering thermogenesis. This involves substrate cycling of, for example, fructose 6-phosphate and fructose 1,6-diphosphate by the enzymes phosphofructokinase and fructose diphosphatase (Newsholme et al. 1972).

In Manduca sexta, it was demonstrated that carbohydrates are used to power pre-flight warm-up (Joos, 1987). Otherwise, little is known about the fuels involved in generating heat during this process. We therefore wanted to investigate this problem for the African fruit beetle Pachnoda sinuata, which uses proline and carbohydrates to power flight (Zebe and Gäde, 1993; Lopata and Gäde, 1994) and elevates its T_th endothermically prior to flight (Heinrich and McClain, 1986). Moreover, we also investigated how T_th influences wingbeat frequency as a parameter of flight performance.

Male fruit beetles of the species Pachnoda sinuata flaviventris (Gory and Percheron) were caught in the vicinity of Cape Town, South Africa, and were maintained as outlined previously (Zebe and Gäde, 1993). Animals in all experiments had a body mass of 1±0.18 g.

Beetles were kept in a transparent plastic container prior to experimentation. Only beetles which were absolutely quiescent in the container, i.e. did not walk around or open the elytra, were used. Experiments were performed at 24–25 °C in a laboratory close to a window which was exposed to bright sunlight or at 31 °C in a constant-temperature room. The beetle was attached via the prothorax to a bent insect pin using dental wax. The pin (mass 180 mg) could move inside a piece of metal tube (diameter 2 mm) in such a way that it was possible to detect whether the animal produced lift as during natural flight (Fig. 1). This apparatus was held in a wooden clamp mounted on a stand. When the top of the insect pin moved up and down, the insect was considered to be producing lift. In addition, this method allowed the insect to turn on its vertical and horizontal axes (Fig. 1). A frontal laminar air current of 1.6 m s⁻¹ was created by placing a fan (27.5 W) at one end of a rectangular metal tunnel (12 cm×12 cm×50 cm). Air was forced through a 1 cm thick grid made up of 64 squares (13 mm×13 mm) which was placed inside the tunnel 10 cm from the opening opposite to the fan. Laminar flow was confirmed by passing smoke through the system.

After the beetle had been attached to the pin, a 1 μl haemolymph sample was taken from the intersegmental membrane between the head and prothorax for determination of metabolite concentrations (see below). After attachment to the stand, the beetle was provided with a styrofoam ball for tarsal contact. The beetle remained quiescent until it started to warm up. As judged by T_th, which in eight beetles (measured using a thermocouple; see below) rose from 25.9±0.2 °C to 34.5±2.0 °C (mean ± S.D.) at take-off, the warm-up period took 83±23 s. Most of the beetles had a pre-warm-up T_th slightly above T_a, probably as a result of slow movement prior to experimentation. In most of the beetles, the beginning of warm-up was indicated by pumping movements of the abdomen, after which the head also started to move back and forth in rhythm with the abdomen. Thereafter, the beetle started to walk on the styrofoam ball for a few seconds. When it tried to open the elytra to take off, the beetle plus attached pin was taken off the clamp, a 1 μl haemolymph sample was taken and the beetle was replaced. This procedure took no longer than 15–20 s. The beetle was then allowed to fly while producing lift for 10 s. Animals that did not start flying voluntarily or flew without producing lift were excluded from the experiment. A third 1 μl haemolymph sample was taken immediately after the flight period. For metabolite determination in the flight muscles, resting beetles, beetles that had completed warm-up and those that had warmed up and flown for 10 s were dissected. The head, prothorax, legs and wings were cut off, and the gut was subsequently removed. The flight muscle tissue was wrapped in aluminium foil and immediately frozen in liquid nitrogen. The whole procedure took no more than 1 min. Preparation of samples and metabolite determinations were performed as described below.

Haemolymph (1 μl) was either blown immediately into 100 μl of concentrated H₂SO₄ for the determination of total lipid concentration (Zöllner and Kirsch, 1962) or carbohydrate levels (Spik and Montreuil, 1964), or pipetted into 60 μl of 80 % acetonitrile for amino acid analysis (see below).

Perchloric acid extracts from frozen tissues were prepared according to Zebe and Gäde (1993).

Glycogen was extracted as described in Zebe and Gäde (1993) and analysed using the modified anthrone method (Spik and Montreuil, 1964) with glucose as a standard.

Derivatisation of extracts with dansyl chloride and determination of proline and alanine levels by high-performance liquid chromatography (HPLC) were carried out as described previously (Zebe and Gäde, 1993).

The wingbeat frequency (f_w) during flight was determined using an infrared light barrier which was placed at 90 ° to the plane of the wing beat. The set-up for flight was the same as described above except that the insect pin was in a fixed position, so that the beetle could not generate lift. The ensuing impulses were recorded using an electronic counter and read every 20 s. In the same experiment, T_th was determined by introducing a copper–constantan thermocouple (accuracy 0.1 °C, mounted in a syringe needle) ventrally through a pre-drilled hole into the metathorax and the temperature was read at 20 s intervals. Measurements of f_w and T_th continued for 6 min. During measurements of T_th in the field, made at a farm 30 km northeast of Cape Town, South Africa, the thermocouple was inserted into the flight muscles immediately after capture. T_a was measured in the shade using the same thermocouple.

An open flow-through respirometry system was used. Changes in the oxygen concentration in the system were determined using an S-3A analyser (Applied Electrochemistry Inc., Sunnyvale, CA, USA) fitted with an N-22M oxygen sensor and an R-1 flow control. Changes in CO₂ concentration were measured using a 225 MK3 infrared gas analyser (ADC, UK). The flow rate of the whole system was 250 ml min⁻¹ for resting animals, 500 ml min⁻¹ for warm-up and 700 ml min⁻¹ for lift-generating tethered flight. The chambers had the following volumes: rest, 10 ml; warm-up, 50 ml; lift-generating tethered flight, 144 ml. Thus, the chamber air was renewed in 2.4, 6.0 or 12.3 s, respectively. These small chamber volumes and the high flow rates ensured a quick response of the instruments so that corrections for gas mixing could be omitted. Flow rate was measured using a bubble flowmeter before and immediately after each experiment. The oxygen analyser was connected downstream from the pump, followed by the CO₂ analyser. Air was drawn through the system. Before air reached the chamber, it flowed through a container filled with Carbosorb (BDH), to remove atmospheric CO₂, and was dried by means of a plastic syringe filled with silica gel. A narrow plastic pipe, 10 cm long and containing silica gel, was attached behind the chamber so that moisture generated by the beetles could be removed. The oxygen analyser was calibrated with atmospheric air, and the zero value was checked regularly with an oxygen-free gas mixture (CO₂+N₂). The carbon dioxide analyser was set to zero by filling the reference cell with CO₂-free air or 350, 750 or 1450 p.p.m. CO₂ (in N_2, oxygen-free) and calibrated with 350, 750, 1450 or 2100 p.p.m. CO₂ (in N_2, oxygen-free), respectively. All gas mixtures were from Afrox (R.S.A). Oxygen and carbon dioxide content were measured to the nearest 0.001 %. The delay between the two analysers was 5–10 s.

Measurements were taken only after the system had equilibrated following insertion of an animal into the chamber. Data from animals that became active immediately upon being placed in the chamber were discarded. The incoming air current in the chamber was directed towards the beetle’s head. For measurement of the resting value, eight animals were used at a time, while single animals were investigated in all other experiments. Prior to flight, they were kept in the chamber in the dark. Body masses were determined immediately after each measurement. These data, together with the measured flow rate, were used to calculate rates of oxygen consumption per gram per hour at standard temperature and pressure (STP).

For separate measurements of rates of oxygen consumption, the carbon dioxide analyser was omitted and the flow rate was set to 250 ml min⁻¹.

All respiration experiments were performed in a controlled-temperature room at 28 °C.

All data are presented as means ± S.D. Significance levels of differences were determined using Student’s t-tests or, where appropriate, paired t-tests.

Wingbeat frequency f_w was measured at T_a values of 25 and 31 °C. The highest wingbeat frequencies were recorded at 25 °C immediately after the onset of tethered flight; after 20 s, f_w decreased significantly with decreasing T_th (Fig. 2A). At T_a=31 °C, f_w and T_th remained constant during the entire 6 min of flight and no significant relationship was found between them (Fig. 2B). Because smaller beetles have a relatively larger heat loss than larger individuals (Bartholomew and Heinrich, 1978), data from two beetles of the same body mass are shown in Fig. 3. At a T_a of 25 °C, the beetle warmed up to 37 °C and started flying voluntarily. During the first minute of flight, T_th did not increase further, but instead began to decrease (Fig. 3B). Although a T_th of 31 °C is sufficiently high for take-off, at a T_a of 31 °C the beetle also warmed up to approximately 36 °C at take-off (Fig. 3B). Therefore, the flight temperature for P. sinuata at which heat loss and heat production are approximately equal should be between the values of T_a investigated.

The T_th of P. sinuata was measured in the field and in the laboratory during different types of activity. Beetles resting in the shade (underneath leaves) had a T_th of 24.6±0.9 °C (N=5), close to T_a (25 °C), while T_th of beetles resting in bright sunlight was several degrees higher at 31.1±1.6 °C (N=6). The latter beetles were able to take off for flight without any visible warm-up or preparation such as those in the cockchafer Melolontha melolontha (Schneider, 1980). T_th was 33.8±1.9 °C (N=6) in beetles after endothermic warm-up (tethered) in the laboratory in conditions similar to those in the field (T_a=25 °C; exposed to sunlight). Beetles that were walking slowly (in the field, not exposed to bright sunlight, T_a=25 °C) had a lower T_th of 27.4±1.5 °C (N=5). Some beetles which were resting on leaves or flowers (in the field, not exposed to direct sunlight, T_a=25 °C) were forced to fly without allowing them to warm up by dropping them from a height of 1 m. For those that opened their wings and tried to fly, T_th was 27.9±1.5 °C (N=7). We also measured T_th of beetles which had just arrived at a plant after a flight of unknown duration. We either caught the individuals in the air before landing or picked them from the plant within a few seconds after landing. For these beetles, T_th was 32.1±1.1 °C (N=7). Our results indicate that a T_th below approximately 28 °C is not sufficient for flight in P. sinuata.

As can be seen from Fig. 4, the proline concentration in the flight muscles decreased significantly during warm-up from 53.5±12.1 to 34.7±9.2 μmol g⁻¹ fresh mass and remained at this level (27.1±4.9 μmol g⁻¹) during 10 s of lift-generating tethered flight. The alanine concentration in the flight muscles, however, increased significantly from 3.1±1.8 to 15.4±5.6 μmol g⁻¹ during warm-up and remained at this level (18.0±2.6 μmol g⁻¹) during 10 s of flight.

The glycogen concentration in the flight muscles was not affected by warm-up, remaining at 58.8±9.3 μmol glucose equivalents g⁻¹ compared with 60.8±9.1 μmol g⁻¹ at rest. The glycogen level decreased significantly, however, during the following 10 s of flight to 45.3±8.1 μmol glucose equivalents g⁻¹ (Fig. 4).

Haemolymph samples were taken at rest, after warm-up and after 10 s of subsequent tethered flight from ten beetles for determination of haemolymph proline, alanine, carbohydrate and lipid concentrations. Although the variability of the resting values was high, there was a significant reduction in proline concentration during warm-up; there was a tendency for proline levels to decrease during the subsequent 10 s of tethered flight, although this was not significant. All individuals showed a significant increase in alanine levels during warm-up and a further significant increase during 10 s of tethered flight (Fig. 5).

Concentrations of total carbohydrates in the haemolymph did not change significantly during warm-up and the subsequent 10 s of tethered flight. Total carbohydrate concentration for 24 beetles was 13.0±5.4, 13.5±5.6 and 14.3±5.3 mg ml⁻¹, respectively. Haemolymph lipid levels did not change either, staying at approximately 10 mg ml⁻¹.

Metabolite stores in the fat body are not mobilised during warm-up (Heinrich and McClain, 1986); therefore, we did not measure the consumption of metabolites in this tissue during warm-up.

The rate of oxygen consumption of resting P. sinuata was 1.3±0.4 ml g⁻¹ h⁻¹ (N=8), while during warm-up a maximum value of 46.8±7.0 ml g⁻¹ h⁻¹ (N=8) was measured. Oxygen consumption rates were highest after 2 min of lift-generating tethered flight at 104.4±7.9 ml g⁻¹ h⁻¹ (N=8).

The resting RQ of P. sinuata was 0.89±0.05 (N=10), and similar values were measured during lift-generating tethered flight lasting 2 min (0.90±0.01; N=6). In contrast, a significantly (P<0.01) lower value of 0.82±0.05 (N=10) was determined during pre-flight warm-up.

In the present study, we have found that the African fruit beetle Pachnoda sinuata starts flight activity voluntarily when a T_th of at least 34 °C is reached. It was also found that the beetle stops flying when T_th drops below approximately 28 °C. Although a T_th of 31 °C is sufficient for flight, P. sinuata elevates its T_th a few degrees above this value prior to flight. This is probably a compensatory mechanism to allow for heat loss to occur when T_a is low (Figs 2, 3). Like the closely related protea beetle Trichostetha fascicularis (Nicolson and Louw, 1980), P. sinuata is unable to retain the heat generated during flight at low T_a. From our data, we can conclude that the T_a at which heat generation and heat loss during flight are equal is approximately 30 °C. An elevated T_th is necessary to maintain flight performance. As in the cockchafer Melolontha melolontha (Schneider, 1980), wingbeat frequency, and thus flight performance, is closely related to T_th: with a decrease in T_th, wingbeat frequency declines (Figs 2, 3). This is probably a result of the properties of the enzymes involved in muscle contraction. Barnett et al. (1975) found temperature optima for P. sinuata flight muscle phosphorylase and actomyosin ATPase of approximately 40 °C, below which activities declined steeply.

In contrast to certain dung beetles, which maintain a high body temperature during terrestrial activity (Bartholomew and Casey, 1977; Bartholomew and Heinrich, 1978; Heinrich and Bartholomew, 1979), P. sinuata elevates its T_th only for flight. Even then, basking is preferred to endothermic warm-up (Heinrich and McClain, 1986). This discrepancy between the potential to elevate T_th and the maintained low body temperature during non-flight activity was described as ‘laziness’ (hypothermia) by the latter authors.

It is likely that the beetles avoid metabolic warm-up in the field when the high energy cost for this process is considered. In our experiments, the maximum rate of oxygen consumption, and thus metabolic rate during warm-up, reaches 45 % of that during flight.

How do the beetles meet this high energy demand? This study provides unequivocal evidence that proline supplies most (if not all) of the energy required for the endothermic elevation of T_th during warm-up. While no changes were found in haemolymph lipid and carbohydrate levels, proline concentration decreased sharply and the concentration of alanine, the end product of the partial oxidation of proline, increased concomitantly. In addition, glycogen stores in the flight muscles seemed to be unaffected by warm-up. These data are supported by measurements of RQ. During flight as well as at rest, RQ was approximately 0.9, confirming previous results showing that P. sinuata uses a combination of proline and carbohydrates to power its flight (Zebe and Gäde, 1993; Lopata and Gäde, 1994). During warm-up, however, the RQ was significantly lower at 0.82, which is close to the theoretical value of 0.8 calculated for the partial oxidation of proline (see review by Bursell, 1981).

Little is known about metabolic fuels for warm-up in other insects. Joos (1987) found that carbohydrates are the energy source during warm-up in the endothermic sphinx moth Manduca sexta, which otherwise uses predominantly lipids during flight (Ziegler and Schulz, 1986). The high metabolic rates observed during endothermic warm-up in the present study support the suggestion of Heinrich and McClain (1986) that endothermic heat production in P. sinuata is achieved by flight muscle activity (shivering). The participation of glycolytic substrate cycles, and thus non-shivering thermogenesis, as proposed for Bombus spp. (Newsholme et al. 1972) and Acherontia atropos (Surholt and Newsholme, 1983) seems unlikely. The activities of key enzymes of these cycles in flight muscle tissue of specimens of the genus Pachnoda were found to be very low (Mitchell et al. 1977; Surholt and Newsholme, 1981).

The importance of proline as a flight substrate in certain insects has been demonstrated previously. Its quantitative role, however, varies quite substantially. In the blowfly Phormia regina, for example, proline serves in small amounts as a ‘sparker’ of the Krebs cycle at take-off (Sacktor and Wormser-Shavit, 1966; Sacktor and Childress, 1967), while it is a major fuel in combination with carbohydrates in several beetle species (see, for example, Weeda et al. 1979; Zebe and Gäde, 1993; Auerswald and Gäde, 1995). In the tsetse fly Glossina morsitans (see Bursell, 1981) and, more recently, in certain dung beetles (Gäde, 1997a,b), proline has been shown to be the sole energy source during flight. In the present study, we show that proline also plays a role during endothermic pre-flight warm-up and is probably the only substrate for this process in P. sinuata.

The authors are grateful for expert technical assistance by Ian Davidson during the oxygen and carbon dioxide measurements and to Dr Sue Nicolson for checking the English. This study was supported financially by a grant (to G.G.) and a PhD bursary (to L.A.) from the Foundation for Research Development (Pretoria) and funds from the University of Cape Town (to P.S. and G.G.).