Thermoregulation of African and European Honeybees during Foraging, Attack, and Hive Exits and Returns

The African honeybee, Apis mellifera adansonii, is well-known for aggressiveness, for rapidity of flight in and out of the hive, for willingness to forage at lower temperatures than the European bee, Apis mellifera mellifera, and for great honey production (see Chandler, 1976; Michener, 1973), as well as for less perfect nest temperature regulation than its European relative (Darchen, 1973).

In other flying insects, maximum activity rates in any one species are a direct function of thoracic temperature (Heinrich, 1974). Furthermore, at least in bumblebees, foraging speed depends on thoracic temperature (Heinrich, 1972). Similarly, during aggressive contests among African dung ball rolling beetles, those with a higher thoracic temperature move the fastest and have a competitive advantage in securing and defending their food in inter- and intra-specific contests (B. Heinrich & G. A. Bartholomew, unpubl.). Might greater activity rates and aggression in the African honeybee also be linked with a higher thoracic temperature?

I here compare the thoracic temperature of free-living individual bees of both sub-species in order to determine if some of the behavioural traits in the two bee varieties may be related to thermoregulatory differences.

All body temperatures were measured with a 40-gauge copper-constantan thermos couple threaded through a hypodermic needle (O.D. = 0·5 mm) and glued in place so that the couple was in the orifice of the needle. Temperatures were read to the nearest 0·5 °C using a battery-operated Omega Engineering Thermocouple Thermometer.

Foraging and attacking bees were captured either with an insect net or they were grasped with a pair of forceps directly from flowers or from my body. Within about 3 s after being grasped either with forceps or by gloved hand, the thermocouple probe was thrust (through the netting, if present) first into the thorax until the highest temperature was encountered, and then into the abdomen. I examined the accuracy of the measurements by continuously recording the thoracic temperature of heated dead bees in the laboratory with a Honeywell recording potentiometer, grasping these bees with known thoracic temperature, and measuring their thoracic temperature with the portable thermometer. The two readings were within 1 °C in bees heated and maintained at 20 ° above air temperature.

Immediately after body temperatures were recorded the bees were killed by crushing the thorax, placed into small paper triangles, and stored in a sealed container at 100% R.H. They were usually weighed within 1 h to the nearest milligram with a Roller-Smith torsion balance.

Returning bees were grasped immediately (within a second) after they landed on the entrance-board of the hive. Bees leaving the hive were grasped from the entrance-board immediately after they wiped their antennae, the usual indicator that they were about to fly. After a bout of measurements at any one hive where I was capturing, probing and crushing bees, there were generally attacking bees (those that clearly attempted to sting through clothing, or which pursued me), particularly at the African hives.

All of the observations on African bees were made in Kenya during January 1977. Body temperature measurements at ambient temperatures (T_A) of 12 °C or less were taken at Molo, at about 2700 m, or at Njoro at about 1800 m. The measurements at T_A> 15 °C and < 25 °C were taken in the vicinity of Nairobi (1500 m), and hose at > 25 °C were taken nearby at Athi River. Thoracic temperatures (T_Th) of European bees were also measured after timed durations of continuous flight at different T_A in a temperature-controlled Modu-lab room regulating to within 1 °C. Individual bees were flown after being captured (in an Erlenmeyer flask) as they were leaving an observation hive immediately adjacent to the room. The bees generally flew towards the fluorescent lights against the gauze covering the ceiling. If they landed they were immediately tapped and stimulated to resume flight. Measurements were taken of bees immediately after they landed, or after they were captured between gloved thumb and forefinger.

Body temperatures of free-flying European bees were recorded in Berkeley, California during January-March 1978. Those in the Modu-lab room were measured in August and September, 1978.

Thoracic volumes were determined by the water displaced by over 60 thoroughly hydrated thoraces that had been wetted in detergent solution and rolled across filter paper to take up excess external moisture.

African honeybees begin to forage early in the morning at low temperatures. I observed ‘near maximal’ traffic in and out of African hives at ambient temperatures (T_A) of 8–10 °C, when sunshine was available at 08.00 h. However, in the absence of sunshine, few bees were leaving the hive at temperatures < 9 °C.

I compared the body temperatures of bees inside the brood nest with those at the hive entrance while leaving. The thoracic temperatures (T_Th) of 15 bees perched upon the brood combs varied from 31·0 to 34·5 °C, with a mean of 32·4 °C. Abdominal temperatures of these bees ranged from 28 to 34 °C, with a mean of 3·0 °C.

The T_Th of bees leaving the hive was on the average 5 °C higher than that of the bees in the brood nest, and it was up to 14 °C above 27 °C, the minimum for level flight. Thus, as bees leave the hive from the warm interior they nevertheless increase the temperature of their thoracic musculature before initiating flight. Thoracic temperature of bees leaving was also independent of external T_A from at least 7·5 to 17 °C (Fig. 1). Bees leave the hive in rapid flight that may be a consequence of their high T_Th

Abdominal temperature (T_Ab) was variable, but on the average it was approximately 12 °C above T_A. The wide range in abdominal temperature, and its dependence on T_A, suggest that the abdomen cools passively after the bees leave the warm brood nest.

At T_A < 25 °C in shade the of bees returning to the hive was, at near 30 °C, on the average 7 °C lower than that of bees leaving (Tables 1 and 2). At least between 10 and 16 °C the T_Th of returning bees was independent of T_A (Fig. 2). The lowest T_Th of returning bees was 27 °C, the minimum to sustain level or descending decelerating flight. At T_A near 24 °C worker bees returning in sunshine were on the average 3 °C farmer than those returning during overcast. However, drones returning during overcast had an average T_Th of 41 °C, being about 7 °C warmer than workers under the same conditions. The workers apparently allow their T_Th to rise passively during flight under moderate T_A. The results do not differentiate whether the drones were warmer than the workers due to the regulation of a higher set-point, or due to passive temperature increase related to lower rates of passive cooling because of their greater mass.

At a given T_A abdominal temperatures of bees that had been in flight outside the hive were uniformly low in comparison with that of bees leaving the hive. The average difference in temperature between T_A and T_Ab was 5 °C in shade and this difference was independent of T_A< 24 °C.

The loaded returning foragers had lower T_Th than the unloaded bees leaving to initiate foraging, and their T_Th was independent of their weight (Fig. 3). However, since average nectar and pollen loads were relatively small (Table 3), about 6 mg (vs. about 23 mg in the European bee), a possible effect of wing-loading on T_Th cannot be discounted.

Attacking bees flew rapidly back and forth as well as around and at persons near the hive. The T_Th of these ‘attacking’ bees were as high or higher than those of bees leaving the hive (Fig. 4, Table 1), even though the bees often spent a minute or more in rapid and continuous flight before they were captured and their body temperatures measured. Abdominal temperatures had decreased, being markedly lower than those of bees inside the hive, and in contrast to T_Th, T_Ab were closely correlated with T_A.

The difference between T_Th and T_A was inversely correlated. For example, although the average T_Th, at T_A of 24 °C was 39 °C, while that at 8 °C was 35–5 °C, the difference between T_Th and T_A was nearly two times greater (27·5 vs. 15 °C) at 8 °C than at 24 °C (Fig. 4).

There was a tendency for the bees attacking in the immediate vicinity (1–3 m) of the hive to be warmer than those attacking at greater distances. For example, at Molo, at T_A of 9 °C, the average T_Th of attacking bees was 35·4 °C 1–5 m from the hive and 33·1 °C at 5–10 m, even though T_A was 3 °C higher when the measurements at the greater distance were made. The difference is significant (P < 0·01). However, at higher T_A (26 °C), average T_Th were higher (40–41 °C), but insignificantly different at 2 and 50 m from the hive. Apparently bees returning to the hive sometimes attack as well as those leaving; some of the attacking bees still carried pollen loads.

Independence of T_Th from T_A was observed during foraging (Fig. 5), as it was during the other three activities examined. Thoracic temperatures were regulated (31–32 °C) at lower levels than during exit from the hive and during attacking, but they were 1–2 °C higher than during return to the hive at given T_A (Tables 1 and 2). Abdominal temperatures were low and unregulated, as during other activities. At high T_A, sunshine caused increases of both T_Th and T_Ab. However, neither flight activity nor sunshine were necessary to create a high T_Th; relatively stationary bees feeding on honey in combs outside of the hive at 7–5 °C before sun-up maintained an average T_Th of 34·1 °C, which was 2 °C higher than that of bees foraging from flowers at 21 °C in sunshine at noon.

Abdominal temperatures varied more widely than T_Th. They averaged about 5 °C above T_A during both foraging and return to the hive in both bee varieties, bees leaving the hive, however, had relatively high T_Ab. At low T_A (7–10°C), the temperature excess (T_Ab-T_A) of the adansonii abdomen averaged 13 °C. (That of mellifera was 5 °C higher.) Bees attacking in the hive vicinity had intermediate temperature excess of the abdomen between leaving and returning bees, but those attacking > 50 m from the hive had low T_Ab, like foraging bees.

In contrast to African honeybees, there were few bees entering and leaving the colonies at 9 °C during a Eucalyptus honeyflow. Nevertheless, thoracic temperatures of the European honeybee were insignificantly different (P < 0·001) from the African variety during hive exits and returns, during foraging, and while attacking (Table 1). As in adansonii, there were highly significant differences (P < 0·001) in T_Th between bees leaving, returning or foraging. Thoracic temperatures of bees attacking were significantly (P < 0·01) different from bees exiting from the hive but this difference was less pronounced in comparison with the other activities (Table 2).

The T_Th of A. m. mellifera were also measured in a temperature controlled room, both as a function of duration of continuous flight, and as a function of T_A. Thoracic temperatures were independent of flight duration from 15 s to 4 min at T_A of 20 and 30 °C (Fig. 6). However, at a T_Aof 10 °C some of the bees had a near 27–28 °C, the lower limit for level flight, after half a minute of flight. None of the bees remained in continuous flight at 10 °C for longer than 2 min, even when they were repeatedly prodded to attempt to stimulate them to continue flying.

Bees that were allowed to land at 10 °C sometimes increased their up to 39 °C by warm-up. The average T_Th of 16 bees that had not been prodded to induce them to fly (the bees engaged in intermittent rather than continuous flight) was 34·5 °C after 2 min at 10 °C. The passive cooling rate (determined in dead bees) is described by the equation log (T_Th— T_A) = 1·56 –0·29 min, which predicts that a bee with a T_Th of 40 °C at T_A of 10 °C should cool to 17·6 °C within 2 min if it did not actively resists its passive rate of cooling.

Bees which were in continuous flight did not regulate their T_Th at T_A from 10 to 25 °C; as already mentioned, they did not fly for more than 2 min at 10 °C, and from 15 to 25 °C their T_Th paralleled 15 °C above T_A (Fig. 7). However, at T_A > 25 °C the bees depressed the temperature excess they generated during flight from 15 °C to an average of 5 °C at 40 °C. Bees flew uninterruptedly at 40 °C for as long as they were allowed (5 min), without stopping to cool, even though T_Th reached 46 °C.

Some of the bees flown in the temperature-controlled room attacked me, concentrating their attention on the stinger-studded black gloves used to catch them. The tendency to attack was strongly correlated with T_A (and hence T_Th). At 15 °C none of the bees attacked. But the attack frequency increased to 27, 46 and 73% at T_A of 25, 30, and 40 °C, respectively.

Thermoregulation in honeybees has been examined as a function of the hive response both to low temperatures (Southwick & Mugaas, 1971) and high temperatures (Lindauer, 1954; Lensky, 1964). Metabolic responses of groups of bees have been examined (Cahill & Lustick, 1976; Brückner, 1975). Esch (1960) measured the body temperature of individual bees during different activities in and out of the hive, and extended the work with Bastian (Bastian & Esch, 1970) to examine the physiological bases of heat production in the thorax. However, despite the great wealth of information on honeybee thermoregulation, it has not been conclusively demonstrated whether or not honeybees maintain their thoracic temperatures (T_Th) independent from ambient temperatures (T_A).

The data from the present study show that the European honeybees do not fly continuously at 10 °C or less, apparently because they cannot maintain the minimum T_Th for flight. Furthermore, during continuous flight, T_Th is not regulated at T_A from 15 to 25 °C; it increases directly with increasing T_A. Thoracic temperature during continuous flight is regulated only at T_A> 25 °C. However, the bees had thoracic temperatures independent from T_A, indicating thermoregulation, during hive exits, foraging, attack and return to the hive at T_A as low as 7 °C and up to 25 °C.

At the present time it is not known how the bees maintain their T_Th independent from T_A at T_A< 25 °C when they are not necessarily in continuous flight for long durations. However, bees warmed up after cooling when not forced to remain continuous flight; they maintained T_Th up to 39 °C at T_A of 10 °C by intermittent warm-up following intermittent flight. Intermittent flight occurs naturally during foraging, affording the opportunity for warm-up while perched on flowers at low T_A. Possibly the T_Th of bees returning to the hive at low T_A were sufficiently high for flight because the return flights had been short and the bees had not had time to cool. Alternatively, if the bees had been foraging far from the hive they could have stopped to warm up en route after cooling to T_Th near 27 °C.

Foraging at low T_A may be costly both in terms of time and energy for warm-up. But it is not known if the reluctance of most bees to forage at low T_A is due to few flowers offering nectar at these temperatures, to energetic costs that would minimize profits, or to individual differences among the bees of any one hive. In the present study both the African and the European bees were observed foraging from Eucalyptus flowers at the relatively low T_A near 10 °C.

Thoracic temperatures were stabilized during continuous flight of 20 s or less at T_A of 30 °C (Fig. 6), and after 2 min of continuous flight at T_A of 25–15 °C (Fig. 7). The bees could fly without stopping to warm up, or to cool down, at T_A from 15 to 40 °C. Under laboratory conditions the bees could maintain a high T_Th at T_A< 15 °C by stopping flight and engaging in warm-up, and they maintained continuous flight at T_A from 30 to 40 °C by physiologically preventing their flight muscles from exceeding 46 °C. During continuous flight at intermediate T_A (15–25 °C) they allowed to drift passively; T_Th followed approximately 15 °C above T_A.

The greater the flight metabolism, the greater the work load, and the heat production rate, and sometimes the T_Th (Heinrich, 1975). However, in the present study, loaded bees returning to the hive had a lower rather than a higher than ‘empty’ bees leaving. The reason for this apparent discrepancy could be that the bees leaving the colony had not yet been in flight, while those returning had engaged in continuous flight before their temperatures were measured. However, whether by behavioural or by physiological mechanisms, the different represent regulated states, rather than passive functions of different rates of heat production simply as a by-product of their respective flight metabolisms, since T_Th were independent of T_A, and hence they were maintained despite different passive cooling rates. Additionally, since the bees leaving the hive were hotter than those within it, it is clear that the bees which already have a sufficient for flight while they are within the colony engage in additional warm-up before leaving. Possibly at T_A ⩽ 10 °C, this pre-flight warm-up behaviour allows them to fly farther before they have to stop and warm up again in order to continue flight to their destination.

Bees defending the nest (attacking) typically pursue, circle and rapidly manoeuvre in flight. Obviously such flight behaviour is not possible with a low just sufficient to maintain level flight. Are the attacking bees hotter than other bees because they have a higher flight metabolism, or is attack behaviour directly correlated with a high T_Th?

At least in the African bees, aggressiveness was in part correlated with T_A (and hence T_Th). At T_A of 26–29 °C, bees pursued for at least 50 m from the hive, while at 8 °C, the bees pursued only up to about 5 m. At the higher T_A, unlike at low, many retuning foragers had high (> 40 °C), and at least some of these foragers (those carrying pollen loads) may have attacked because they were already hot. At 7 °C the bees were capable of maintaining a minimum T_Th for flight and foraging, but at increasing distance from the hive convective cooling caused T_Th to decline below the temperature normally maintained while initiating attack. It is thus possible that the drop T_Th in stopped the attack behaviour, even though it did not preclude flight. Since the attacking bees were hotter the closer they were to the hive (and the less distance they had flown) it is probable that their high T_Th is not a consequence of the attack behaviour, but on the contrary, at least in part a cause of it. Similarly, in the flight room, only those European bees attacked which had a high T_Th because they were in flight at a high T_A. However, attack behaviour is obviously not only a function of T_Th; the very aggressive African bees were, on the average, no hotter than the European bees.

Unlike some other insects (Heinrich, 1974), particularly moths and beetles Bartholomew & Heinrich, 1973, 1978) in whom T_Th during flight is a function of size, the African honeybees, which are about 66% the mass of European bees when empty though only 17% lower in thoracic mass (Table 3), did not show a correspondingly lower T_Th. Since they are smaller they must necessarily have a greater passive cooling rate, which would result in lower body temperatures in flight, if their metabolic rates were equal. These data indicate that, on the average, they have a greater energy expenditure during foraging, aggression, as well as flight to and from the hive, than do the European bees.

I am grateful to Trevor Chandler and Jim Nightingale for providing African bees and equipment, and for giving valuable information and advice. Jim Nightingale provided facilities that made the study of African bees possible in the short time available. He also gave generous hospitality that made it memorable. Christel Lehmann provided invaluable assistance in data gathering. The study was supported in part by a grant from the Guggenheim Foundation and by N.S.F. grant DEB 77-08430.