Activation of the Fibrillar Muscles in the Bumblebee During Warm-Up, Stabilization of Thoracic Temperature and Flight

The fibrillar muscles of bumblebees may be used to drive the wings during flight and during fanning for nest-temperature regulation (Hasselrot, 1960), to produce buzzing sounds, to vibrate the substrate during pollen collecting (Michener, 1962), and to generate heat prior to flight or during the incubation of brood (Heinrich, 1972 b). During fanning, buzzing and the production of other vibrations, the generation of heat is incidental to the performance of mechanical work, but during warm-up and brood incubation the primary task of the fibrillar muscles is heat production.

During warm-up the indirect muscles are usually excited by synchronous impulses (Mulloney, 1970; Kammer & Heinrich, 1972), but the motor patterns during the stabilization of thoracic temperature in stationary bumblebees, and during flight at different ambient temperatures, have not been examined previously. Little is known about the relationship between muscle activation and (1) the rate of heat production during warm-up at different thoracic and ambient temperatures, and (2) the temperature excess maintained during stabilization of thoracic temperature at various ambient temperatures.

Bumblebees fly at low as well as at high ambient temperatures (Heinrich, 1972 a) and require a high thoracic temperature in order to be able to continue flying (Krogh & Zeuthen, 1941). However, during flight the primary function of the fibrillar muscles is to drive the wings, and the production of heat is obligatory and continuous. Since the fibrillar muscles can be used specifically for heat production during warm-up, it is of interest to know whether their activation is varied also during flight in order to control heat production.

We here report on the activation of the fibrillar muscles at different muscle temperatures during warm-up, during the stabilization of thoracic temperature at a variety of ambient temperatures, and during fixed flight at different ambient temperatures.

Concurrent observations were made on the thoracic temperatures and the electrical activity of the fibrillar flight muscles of queen bumblebees, Bombus vosnesenskii Radowskowski. Muscle potentials were recorded extracellularly from bees glued to a support, as in previous experiments (Kammer & Heinrich, 1972). The data wet recorded on magnetic tape and film for subsequent processing. The number of muscle potentials (spikes) per second was determined by manually counting the filmed recordings. When the firing occurred in bursts it was sometimes difficult to make accurate counts of individual spikes because the extracellular potentials became small and were summed. They could thus be confused with the ‘noise’ of small potentials representing distant muscle units (cf. Fig. 7). Each small potential change considered to come from the unit in question was counted as one spike.

Thoracic temperatures (T_Th) were measured by means of a small bead thermistor placed about 2 mm below the thoracic notum. These data were also recorded on magnetic tape with the muscle potentials. Ambient temperatures (T_A) were measured with a mercury thermometer placed 1–3 cm from the bee. The ambient temperature was changed by placing the bee and its support inside either a Styrofoam box containing a layer of ice or a glass box within a heated water bath.

Thoracic temperatures were also measured on bumblebees free to move about within a wooden box. In these bees thoracic and abdominal temperatures were measured with 46-gauge copper-constantan thermocouples with cotton insulated leads. Temperatures were recorded at 3 sec intervals with a Honeywell recording potentiometer.

Rates of heat production were calculated from the observed temperature excesses (T_Th – T_A) and the rates of change of T_Th. In making these calculations we assumed that: (1) only the thorax was heated; (2) cooling was passive and, according to Newton’s Law of Cooling, a function of the temperature excess ; and (3) the thermal capacity of the thorax was 0·8 cal °C⁻¹.g⁻¹ (Krogh & Zeuthen, 1941). Data from the rate of cooling in still air of a dead B. vosnesenskii queen which weighed 0·4 g were used to calculate the rate of heat loss as different values of (T_Th – T_A). An average value of 0·23 g was used as the thoracic weight of the experimental insects. Using a mean thoracic weight is a valid approximation because the thoracic weights of queen B. vosnesenskii were similar (0·22 – 0·25 g, N = 5). The assumption that only the thorax is heated appears to be valid during preflight warm-up, but during the stabilization of T_Th in stationary bees the abdomen may also be heated (Heinrich, 1972b) and calculations of heat production based on T_Th therefore underestimate the actual rates of heat production.

The rate of heat production by the bumblebees during warm-up depended on their thoracic temperature. At low T_Th (< 10 ° C) the bees were unable to warm themselves. However, when a bee commenced activity at higher T_Th, activation of the fibrillar muscles (Fig. 1) resulted in a progressive rise in thoracic temperature (Fig. 2). As the warm-up proceeded and increased, spike frequency increased (Figs. 2, 3).

The estimated rate of heat production during warm-up was correlated with spike frequency (Fig. 4). Within the limits of error in the estimated rate of heat production, there was no effect of temperature on the heat produced per muscle activation (Fig. 5). Therefore, heat production was primarily a function of spike frequency. This frequency was always low at low T_Th and higher but more variable at higher T_Th.

Bumblebee queens when relatively stationary during ‘brooding’ may maintain a high and relatively constant thoracic temperature independent of T_A for more than a day (Heinrich, 1972 a). In bees walking about and in bees wired for the recording of muscle potentials there were short-term fluctuations of T_Th around a high value (Figs. 6 – 8). For purposes of data analysis in the present paper stabilization was defined as the maintenance of T_Th within 1 ° C for at least 20 sec. Under the experimental conditions of the present study thoracic temperatures were usually stabilized at 33 – 36 ° C over ambient temperatures ranging from 5 – 30 ° C. However, the bees did not warm up at the lowest T_A unless was already partially elevated when they were placed in the lower T_A. Bumblebee queens may also stabilize the temperature of the abdomen, but at very different temperatures (Fig. 6; Heinrich, 1972 a), indicating that different amounts of heat can be transferred from the thorax to the abdomen.

Muscle potentials recorded during the stabilization of T_Th showed bouts of intense activity, periods of relative inactivity, and wide ranges of intermediate frequencies (Fig. 7). The mean number of spikes/sec during the stabilization of T_Th was usually lower and more variable than that during warm-up at approximately the same thoracic temperature (Figs. 2, 8).

During warm-up in most of the bees, heat production was a function of spike frequency (Fig. 4), and there is no reason to think that the same generalization does not apply to heat production during the stabilization of T_TH. AS expected, there was a clear correlation between spike frequency and the temperature excess (T_Th – T_A) maintained during the stabilization of T_Th in some bees (Fig. 9). However, in other bees the temperature excess, and the estimated heat production during the stabilization of T_Tb (assuming only the thorax is warmed) was often less than half the predicted values at a given spike frequency. It follows that thoracic temperature is differently affected by a given spike frequency because of heat loss from the thorax. Since the abdomen may contain as much heat as the thorax at any one time (Heinrich, 1972b), we suggest that during the stabilization of T_Th heat may be transferred from the thorax to the abdomen, producing elevated abdominal temperatures (Fig. 6). However, in stationary bees heat transfer is obviously not a prerequisite for the stabilization of T_Th.

The bumblebees usually initiated flight spontaneously after a vigorous warm-up. At the beginning of flight spike frequency was high, and then decreased (Figs. 1, 10; fig. 1 in Kammer & Heinrich, 1972). Spike frequency during fixed flight was independent of the ambient temperature. During fixed flight at high T_A thoracic temperature increased but at low T_A thoracic temperature usually decreased (Fig. n). However, in one case we observed a T_Th= 35·8 ° C at a T_A of 7 ° C during fixed flight lasting approximately 1 min. Spike frequency also stayed high during this flight, namely between 20·3 and 21·7/sec. In the same animal while not flying, and presumably while losing less heat by convection, stabilization of T_Th at 34·7 ° C at 5 ° C was accomplished with only 10·2 spikes/sec. Thus a high T_Th can be maintained at a low T_A during fixed flight. In most of these experiments, however, T_Th was not stabilized during fixed flight, and varied with T_A and with spike frequency.

The thoracic temperature of bumblebees, as of any insect, depends on the balance between the rate of heat input and the rate of heat loss. The main determinant of heat input (in shade) is the activity of the flight muscles, as indicated by the frequency of muscle potentials. The rate of heat loss depends primarily on the difference between thoracic and ambient temperatures, but heat loss can be varied by the transfer of heat from thorax to abdomen. The relationships among these variables during warm-up, stabilization of T_Th, and fixed flight are considered in the following discussion.

Like many other insects, bumblebees warm up prior to flight and do not initiate flight until the thoracic temperature is near 30 ° C (Krogh & Zeuthen, 1941). Heat production of sufficient magnitude to elevate body temperature above ambient temperatures occurs only when the flight muscles are activated. In honeybees the metabolic rate during warm-up as well as during flight is directly related to spike frequency (Bastian & Esch, 1970). Similarly, in bumblebees, spike frequency and rate of heat production are correlated during warm-up. Both increase with thoracic temperature (Figs. 2 – 4) so that T_Th rises slowly at low T_A (Fig. 2) and rapidly at high T_A (Fig. 8). At T_Th< 10 ° C the rate of heat production is lower or equal to the rate of heat loss. Consequently the bees are unable to attain the high T_Th required for the rapid production of heat which is necessary to maintain a large temperature excess. Since bumblebees may forage with a T_Th near 36 ° C, even at dawn at a T_A of 2 ° C (Heinrich, 1972 a), it is probable that they keep warm throughout the night in the nest and that, at such low T_A, they maintain a high T_Th during the day in order to remain active. Temperature regulation in the nest at low T_A (Hasselrot, 1960) may thus be functional not only in accelerating the rate of brood development (Himmer, 1932), but also in making it possible for the bees to initiate foraging early in the morning before T_A has risen to levels where warm-up from low T_Th (and low T_A) is possible.

In most bees the rate of heat loss is probably not varied during pre-flight warm-up. The linear relation between spike frequency (Fig. 4) and heat production (calculated from T_Th and passive rate of cooling) and the maintenance of a relatively low abdominal temperature (Heinrich, 1972b) suggest that heat is sequestered in the thorax rather than being dissipated to the abdomen. Such a retention of heat in the thorax occurs in the sphinx moth, Manduca sexta, during flight at low T_A (Heinrich, 1970) and during warm-up (Heinrich & Bartholomew, 1972).

It is intuitively obvious that the minimum amount of heat which is required to stabilize a given T_th is inversely related to T_A. However, in several bees we did not observe a correlation between spike frequency (and presumably heat production) and the difference between T_Th and T_A (Fig. 9). Abdominal temperature was often variable while T_Th was relatively stable (Fig. 6). These results suggest that, at the higher spike frequencies, these bees were producing more heat than the minimum which was necessary to stabilize T_Th and the additional heat was being transferred into the abdomen. Heating of the abdomen is functionally significant during brood incubation; heat is transferred from the thorax into the brood clump via the abdomen (Heinrich, 1972b). Transport of heat from the thorax to the abdomen has also been observed in flying moths. In Manduca sexta during free flight at ambient temperatures higher than 20 ° C heat is carried by the blood into the abdomen, thereby preventing overheating of the flight muscles and stabilizing T_Th (Heinrich, 1970).

During fixed flight in bumblebees spike frequency was not related to T_A. The heat production, which is a function of spike frequency, was sufficient to cause an increase of T_Th during fixed flight at high T_A. At low T_A, however, the heat production was insufficient to maintain the temperature (⪖ 30 ° C) which is necessary for level free flight. These results indicate that T_Th is not regulated at specific lower set-points during fixed flight. However, the bees must have a T_Th ⪖ 30 ° C during free flight in the field, and bumblebees in the field are capable of continuous free flight even at T_A< 10 ° C. If a bee in free flight at these T_A allowed T_Th to decrease to the same extent as in fixed flight, it would soon become incapable of generating sufficient lift for further flight. Nevertheless, the bees could remain in free flight without regulating their T_Th, per se, if they control their altitude and flight velocity on the basis of information received from exteroceptors such as the compound eyes and/or mechanoreceptors. Increasing the flight effort would automatically increase T_Th especially in an insect that is sufficiently insulated with pile (Church, 1960) to retard the rate of convective cooling.

We are grateful to the Department of Zoology, University of California, Davis, for making available most of the equipment used during the course of this work. Supported by NSF Grant GB-31542.

It has recently been proposed (on the basis of enzyme activities in muscle homogenates) that bumblebees elevate their T_Th by non-shivering thermogenesis (Newsholme et. al., 1972, Biochem. J. 128, 89 – 97). However, the elevation of T_Th in bumblebees with inactive muscles has so far not been demonstrated by direct measurement. On the basis of our correlation of action potentials and heat production, and on the basis of Ikeda & Boettiger’s (1965) correlation of action potentials and mechanical activity in bumblebee fibrillar muscle (J. Ins. Physiol. 11, 779 – 789), we conclude that the fibrillar muscles produce heat of sufficient quantity to elevate T_Th only when they are excited by impulses from the central nervous system, when they are mechanically active.