Instantaneous Oxygen Consumption and Muscle Stroke Work in Malacosoma Americanum During Pre-Flight Warm-Up

In many insects, heat is produced by the flight muscles to elevate thoracic temperature (T_th) prior to take-off (Dotterweigh, 1928; Krogh & Zeuthen, 1941). Although all heat production occurs essentially in the flight muscles, their quantitative performance during warm-up has been difficult to evaluate. Estimates of the metabolism per contraction of flight muscle of synchronous fliers during pre-flight warm-up have been made based on measurements of heat storage and heat loss of the thorax (Heinrich & Bartholomew, 1971; May, 1979; Casey, Hegel & Buser, 1981). However, those studies underestimate total heat exchange because they do not account for heat loss and heat storage from the head and the abdomen (Hegel & Casey, 1982). Moreover, measured instantaneous rates of oxygen consumption during warm-up significantly exceeded values obtained from calculations of heat exchange parameters (May, 1979; Bartholomew, Vleck & Vleck, 1981). However, these authors did not report wing stroke frequency (n) and, therefore, important details of muscle performance cannot be characterized from their data. Although the role of T_th in determining stroke frequency is well known (Kammer & Heinrich, 1978; Kammer, 1981), metabolism during warm-up is dependent not only on frequency of muscle contraction, but also on the energy expended per contraction (stroke work input, E/n). The latter cannot be derived with sufficient precision without simultaneous measurement of n, T_th and ⁠.

The flight muscles serve different purposes in warm-up and flight. During warmup, the muscles raise T_th to the minimum temperature necessary for continuous flight. This process has been postulated to occur as rapidly as possible to reduce the time when the moth is grounded and vulnerable to predators (Bartholomew & Heinrich, 1973; Heinrich & Casey, 1973). Thus, it should be advantageous for the muscle to produce as much heat as possible (i.e. to operate at or near maximal rates). During flight, the muscles supply the amount of lift appropriate to flight conditions. Estimates of metabolism of moths during warm-up based on heat exchange calculations suggest that energy expenditure is lower during warm-up than during flight at the same T_th (Heinrich, 1974; Casey et al. 1981; Heinrich & Mommsen, 1985). This result is puzzling if the moths are warming up as rapidly as possible. In an attempt to resolve this paradox we undertook to re-examine the energetics of pre-flight warm-up in the tent caterpillar moth (Malacosoma americanum). The present study reports instantaneous rates of ⁠, T_th and muscle contraction frequency during pre-flight warm-up. These data are used to determine energy expenditure throughout warmup, and to quantify the effects of muscle temperature on the energetics of muscle contraction. We conducted our experiments at several T_a values to examine the effects of T_a on the energetics of warm-up.

Male moths were collected in the Hutcheson Memorial Forest (Somerset County, New Jersey) in June 1982. Procedures for collection, storage and handling were similar to those described in an earlier paper (Casey et al. 1981).

Due to the rapidly changing rates of oxygen uptake during pre-flight warm-up (see Bartholomew et al. 1981) experiments were conducted in a high flow (240–300 ml min⁻¹), small volume respirometer. The chamber, a 35ml plastic, transparent, Coulter-counter vial was connected to an Applied Electrochemistry Inc. Oxygen Analyzer (S3A-sensor N22M) via a T-tube. A small opening was made in the lid of the vial for thermocouples, impedance leads and incurrent air flow. An identical chamber was attached to the other end of the T-tube. During experiments, the empty chamber was clamped off with mosquito forceps. The fractional oxygen concentration of incurrent air prior to and immediately following each experiment was determined by clamping off the animal respirometer and allowing room air to flow through the empty chamber. Prior to gas analysis the air was dried by passing it through a column of dessicant. The effective volume of the chamber and rates of instantaneous oxygen consumption were determined by the method described by Bartholomew et al. (1981). A respiratory quotient of 0·85 was assumed (Joos, 1986). All gas volumes were converted to STPD. Heat production (HP) was calculated from oxygen consumption data assuming that 1 ml O₂ is equivalent to 20·1 J.

A small area of scales was removed from the dorsal thorax using microforceps. We implanted 44-gauge copper-constantan thermocouples into the dorsal thorax after punching a small hole in the cuticle with a microsurgical needle. During implantation the moth was placed in a plastic box filled with crushed ice. This treatment kept thoracic temperature well below 10°C and made the use of anaesthesia unnecessary. Implanted wires were held in place by dried haemolymph and required no further attention. The moths were usually allowed a quiescent period in the ice after implantation (about 30min). They were then transferred to the appropriate thermal regime. Since they were not anaesthetized, they often began typical pre-flight warmup behaviour immediately after T_th had passively warmed to 10–15°C.

Thoracic and ambient temperature were monitored using alternate channels of a two-channel Bailey Instruments Laboratory Thermometer whose output was attached to another servo channel of the polygraph. This arrangement allowed us to obtain continuous, simultaneous data for thoracic temperature and oxygen uptake.

Wing stroke frequency (n) was measured by implanting 44-gauge constantan wire into two holes made in the dorsal thorax (see above) on either side of the dorsal midline. The wires were attached to an impedance converter whose output was recorded on a polygraph.

During warm-up, T_th of Malacosoma americanum increased linearly with time at rates which were strongly and linearly related to T_a (Fig. 1). Correlation coefficients for linear regressions of thoracic temperature versus time exceeded 0·99 at all T_a values. Mean rates of warm-up of 2·6°C min⁻¹ at T_a of 13°C and 10°C min⁻¹ at T_a of 25 °C were not significantly different from values predicted from regression analysis of our previous data (Casey et al. 1981). Therefore, although the previous data were collected in still air, it is apparent that airflow within the respirometer did not markedly affect rates of convective cooling and subsequent rates of warm-up.

The wing stroke frequency during warm-up was tightly coupled to thoracic temperature, varying from 15 s⁻¹ at T_th = 15°C to about 55 s⁻¹ at T_th = 35°C. The temperature of the muscle rather than the ambient temperature determined the frequency of muscle contraction. Slopes and intercepts for linear regressions relating n to T_th at T_a values of 13, 20 and 25°C were indistinguishable from the relationship describing all data regardless of T_a. Furthermore, frequency data were similar to those we presented previously (Casey et al. 1981).

As in the study of Bartholomew et al. (1981), changes in T_th during warm-up reflect changes in oxygen consumption. A slight curvilinearity occurred in the rate of oxygen uptake which was not reflected in the change in T_th. Fig. 2 shows a typical polygraph trace of oxygen consumption at T_a = 25°C. Despite the design of our system, this curvilinear change of apparent fractional oxygen concentration of excurrent air with time appears to be an artifact due to time lags and rapid rates of change of oxygen uptake (Bartholomew et al. 1981). The difference between apparent and actual instantaneous was always greatest at high T_a and was much less pronounced or completely absent at low T_a. After appropriate correction factor had been applied, all data, regardless of T_a, increased linearly with time.

Rates of heat production during warm-up are directly related to thoracic temperature, varying from about 10mW at T_th of 15°C to 75mW at T_th of 40°C (Fig. 3). HP was dependent only on T_th. Thus, a moth warming at T_a of 13 or 25°C had an instantaneous rate of heat production of approximately 45 mW at T_th of 30°C. These values are similar to predictions based on heat exchange analyses in their linear relationship to T_th and their independence of T_a, but are substantially greater in magnitude.

Since heat production is a function of thoracic temperature and independent of T_a (Fig. 3), while the rate of thoracic temperature increase is directly related to T_a, the rate of change of the metabolic rate during warm-up should show a similar dependence on ambient temperature. Metabolic rate increases continuously during warm-up and the rate of increase is linearly related to T_a (Fig. 4). There is good correspondence between the rates of increase of metabolic rate and of T_th during warm-up. Between T_a values of 15 and 25°C, each parameter increases about three times (Figs 1,4).

Due to its very small mass (X̄ = 90 mg), heat loss, and therefore T_a, should have a strong effect on the total cost of both warm-up and cooling in M. americanum. Fig. 5 illustrates the rates of energy expenditure during pre-flight warm-up and post-flight cooling at ambient temperatures of 13 and 25 °C. At high T_a, a moth has a higher initial rate of heat production due to a higher muscle temperature. It also increases its rate of heat production more rapidly because more of the heat is being used to warm the thorax. Both of these factors reduce the duration of warm-up at the higher T_a. Indeed, over a 12°C range of T_a, the duration of pre-flight warm-up varies almost eight times (Fig. 5). Since the total cost of warm-up at each T_a equals the area under each of the trapezoids in Fig. 5, T_a is obviously a major determinant of the total cost of warm-up in M. americanum, in sharp contrast to the situation in flight where energy metabolism is independent of T_a (Casey, 1981 a).

The total energy expended during pre-flight warm-up and post-flight cooling is shown in Fig. 6. Over the range of T_a from 13 to 25 °C, the cost of warm-up decreases about five times. Such large differences are not apparent in the energetics of postflight cooling. There was a slight, significant difference between the total energy expenditure during cooling at different T_a values (1·25 J vs 0·5 J, Fig. 6). Although the energy expended during cooling is greater at the lower T_a the relative increase in the cost of warm-up is much greater. As a consequence, cooling represents only about 5 % of the energetic episode at T_a of 13 °C compared with about 10% at T_a of 25 °C (Fig. 6).

Metabolism during warm-up can increase as a result of an increase in n, an increase in the energy expended per wing stroke (the stroke work input, E/n) or both. E/N is considerably higher than previously calculated (Fig. 7) and at low thoracic temperatures it is temperature-dependent. At T_th greater than 25 °C, however, stroke work input is essentially independent at about 1·1 mJ. Thus, most of the increase of energy metabolism as warm-up proceeds is mediated by increased frequency of muscle contraction which occurs as a result of increased T_th. Measured values for stroke work during warm-up are comparable to values obtained for M. americanum during free hovering flight.

The rate of heat production is related to T_th, but not to T_a (Fig. 3). Since heat loss is proportional to T_th– T_a, rate of thoracic temperature increase is strongly dependent on T_a. Similarly, rate of change of heat production (mWmin⁻¹) is also strongly related to T_a (Fig. 4). While metabolic data from the present study agree qualitatively with previous understandings of the energetics of warm-up (Heinrich & Bartholomew, 1971; Heinrich, 1975; May, 1979; Casey et al. 1981; Hegel & Casey, 1982), the magnitude of heat production is much greater than previous estimates for M. americanum (Casey et al. 1981) because they were based solely on heat storage and heat loss in the thorax, and did not include the head, abdomen or respiratory system as additional avenues of heat exchange. Measured rates of heat exchange from the head and abdomen of the sphingid Manduca sexta accounted for about 24–27 % at T_a values from 16 to 30°C (Hegel & Casey, 1982). Data reported by Bartholomew et al. (1981) also indicate a significant difference between measured ⁠, values and heat exchange values.

Our data suggest that the energetic cost of cooling is a very small fraction of the total cost of warm-up at all T_a values (Fig. 6). These results are in marked contrast to those of Bartholomew et al. (1981), who report that the cost of cooling amounts to 69–75 % of the cost of warm-up. The sphingids and saturniids spend substantially more time in cooling than in warm-up (Bartholomew et al. 1981, their figs 2, 3, 4). The small size and high thoracic conductance of M. americanum compared with that of sphingids and saturniids is probably responsible for much of this discrepancy. Due to these factors, the cost of elevating T_th is very high and high levels of heat loss reduce the effectiveness of heat storage, thereby increasing the total duration (and therefore the total cost) of warm-up, particularly at low T_a. However, the same factors facilitate rapid post-flight cooling. Consequently, M. americanum achieves resting thermal states much more rapidly than larger moths, which reduces its total cost of cooling.

Our data indicate that heat production by Malacosoma is considerably greater than was previously reported but they do not necessarily indicate that this represents a maximal effort. A strong selective pressure for maximal output during warm-up based on predator avoidance has routinely been presented (see Introduction) and our data are consistent with that interpretation. However, perhaps a more compelling argument applies for M. americanum because these moths routinely fly during early morning hours when T_a is 15°C or less (Casey, 1981a). At this temperature, cost of warm-up is very high due to the low rate of thoracic temperature increase and consequent long time required to reach flight temperature (Fig. 5). If the moth could increase metabolic rate above measured levels (polygon C in Fig. 8A) it could substantially reduce the time necessary for warm-up. Since there is a steep inverse relationship between total cost of warm-up and the rate of warm-up (Fig. 8B), small changes in the latter would result in substantial reductions in the cost of warmup at ecologically relevant air temperatures. Thus, under most conditions warming as rapidly as possible is cheaper than warming at a slower rate.

It is clear that T_a is an important determinant of the energetics of warm-up. In the sphingid, M. sexta, a drop in T_a from 30 to 16°C triples the cost of warm-up (Hegel & Casey, 1982). Our data indicate that thermal effects are even greater for smaller moths (Figs 6, 8). Consequently, scaling estimates of the energetics of warm-up (Bartholomew & Casey, 1978; Bartholomew et al. 1981) are difficult to evaluate when based on a single T_a value. Conclusions based on such data should be made with care due to the large numbers of variables involved, strong interactions between T_th and n for moths of different morphology (Kammer & Heinrich, 1978; Casey et al. 1981) and between various conductances (head, thorax, abdomen) with mass, T_a and insulation (Hegel & Casey, 1982). Comparative data on during warm-up similar to those obtained by Bartholomew et al. (1981) for moths of different sizes at low T_a would be extremely useful, as would studies of muscle frequency vs T_th for moths differing in size and wing morphology.

It is generally assumed that mechanical power requirements and metabolic rate should be closely related during flight (Weis-Fogh & Alexander, 1977; Casey, 1981; Ellington, 1985). During warm-up, however, it should be adaptive to convert as much of the expended energy into heat as possible (i.e. be very inefficient) to minimize the duration and the cost of warm-up. For insects, as well as for ectothermic vertebrates, muscle performance is very temperature-sensitive (Josephson, 1981; Bennett, 1985) and much of the thermal sensitivity is associated with temporal characteristics of muscles while tension development appears to be relatively insensitive. These characteristics are consistent with our data for energy expenditure of M. americanum during warm-up. Most change in energy expenditure results from change in contraction frequency while change in E/n is relatively insensitive to temperature (Fig. 7).

The frequencies of muscle contraction during warm-up and flight are comparable in M. americanum at the same T_th. Since the energy expended per muscle contraction is also similar whether the animal is warming up or flying, our data suggest a comparable neural input to the muscles during each activity. Obviously, the change in muscle function in the transition from warm-up to flight (providing lift and aerodynamic power) need not require different energy expenditure. Change in mechanical power output can be accomplished by a change in the phase of contraction between the elevators and depressors (Kammer, 1968). During warm-up the muscles are contracting almost simultaneously so that virtually all the energy is degraded to heat as the muscles work against each other. By contracting alternately each muscle set can do useful work on the surrounding air. Thus, change in muscle function between warm-up and flight at the same thoracic temperature represents a change in muscle efficiency which is consistent with the functions they serve in each activity. The moths must retain the capacity for additional increases in power during flight to accommodate sudden aerobatic manoeuvres and climbing flight. Double firing of the motor neurones innervating the flight muscle (Wilson & Weis-Fogh, 1962; Kammer & Heinrich, 1978) would allow for increased power output by increasing frequency and stroke amplitude (Kammer & Rheuben, 1981).

Results from the present study are very similar to those obtained for bumble-bees, which also show equivalent muscle metabolism per action potential at a given T_th, regardless of whether they are in warm-up or flight (Kammer & Heinrich, 1974). While synchronous and asynchronous fliers may differ in muscle morphology and physiology (Ellington, 1985), motor patterns (Kammer & Rheuben, 1981) and in the magnitude of energy expended per muscle contraction (Casey, May & Morgan, 1985), they show comparable responses to temperature and similar patterns of muscle energetics in the transition from warm-up to flight. For M. americanum as well as for the bees, the limitation to take-off is a thermal dependence of muscle contraction frequency. Once T_th is elevated to a sufficient level for the muscles to operate at appropriate wing stroke frequencies flight commences.

We are pleased to thank Drs W. A. Buttemer, C. P. Ellington, B. A. Joos, M. L. May and R. D. Stevenson for stimulating and illuminating discussions. Supported by NSF grant PCM8219311 and by the New Jersey State Experiment Station (Project 08511).