Thermoregulation of foraging honeybees on flowering plants: seasonal variability and influence of radiative heat gain

1. During nectar and pollen foraging in a temperate climate, honeybees are exposed to a broad range of ambient temperatures, challenging their thermoregulatory ability. The body temperature that the bees exhibit results from endothermic heat production, exogenous heat gain from solar radiation, and heat loss. In addition to profitability of foraging, season was suggested to have a considerable influence on thermoregulation. To assess the relative importance of these factors, the thermoregulatory behaviour of foragers on 33 flowering plants in dependence on season and environmental factors was investigated. 2. The bees (Apis mellifera carnica Pollman) were always endothermic. On average, the thorax surface temperature (Tth) was regulated at a high and rather constant level over a broad range of ambient temperatures (Tth = 33.7–35.7°C, Ta = 10–27°C). However, at a certain Ta, Tth showed a strong variation, depending on the plants from which the bees were foraging. At warmer conditions (Ta = 27–32°C) the Tth increased nearly linearly with Ta to a maximal average level of 42.6 °C. The thorax temperature excess decreased strongly with increasing Ta (Tth−Ta = 21.6 − 3.6°C). 3. The bees used the heat gain from solar radiation to elevate the temperature excess of thorax, head, and abdomen. Seasonal dependance was reflected in a 2.7 °C higher mean Tth in the spring than in the summer. An anova revealed that season had the greatest effect on Tth, followed by Ta and radiation. 4. It was presumed the foragers' motivational status to be the main factor responsible for the variation of Tth between seasons and different plants.


Introduction
Honeybees need nectar and pollen to provide for their young bees and brood. Honey supplies energy for heat production to achieve a constant brood temperature and for overwintering in a temperate climate (Stabentheiner et al., 2003a. During foraging, bees are mostly highly endothermic. They may exhibit thoracic temperatures higher than 40 • C (e.g. Heinrich, 1979a;Cooper et al., 1985; Re-use of this article is permitted in accordance with the Terms and Conditions set out at http://wileyonlinelibrary.com/onlineopen# OnlineOpen_Terms Correspondence: Helmut Kovac, Department of Zoology, Karl-Franzens-University of Graz, Universitätsplatz 2, A-8010 Graz, Austria. E-mail: he.kovac@uni-graz.at 1988; Kovac & Schmaranzer, 1996;Schmaranzer, 2000;Kovac et al., 2010). Thermoregulatory investigations of honeybees during foraging on natural sources in their environment are very scarce. Heinrich (1979a) measured thoracic (core) temperatures of Apis mellifera mellifera and Apis mellifera adansonii Linnaeus during foraging on Eucalyptus sp., Bidens pilosa L., and Petrea volubilis L. Thoracic temperatures were regulated between 31 and 32 • C, differing insignificantly between the European honeybee and the African variety. Kovac and Schmaranzer (1996) measured body surface temperatures of honeybees (Apis mellifera carnica Pollman) foraging in the shade in the spring and summer at several different plants. The average thorax temperature varied in a broad range (T th = 29.3-35.7 • C, mean values per flower).
The body temperature of foraging insects is influenced by several environmental factors such as ambient air temperature, solar radiation (Willmer & Unwin, 1981), and convection (for an overview see Heinrich, 1993). The energy gain from solar radiation is of importance for the thermoregulation of foraging bees. An increase of the thorax temperature with increasing insolation was reported in Western honeybees arriving at the nest entrance after their foraging flights (Cena & Clark, 1972;Heinrich, 1979a;Cooper et al., 1985) and during nectar foraging (Heinrich, 1979a). Underwood (1991) reported the same for Indian honeybees collecting sugar syrup under sunny and overcast skies. Kovac et al. (2009a) investigated the influence of solar radiation on the thermoregulation of water-foraging wasps in detail. Vespula and Polistes increased the thorax temperature and reduced the active heat production as solar heat gain increased. In water-foraging honeybees, the relative contribution of endothermic heat production and heat gain from solar radiation on body temperature was observed by Kovac et al. (2010). Up to an ambient temperature of ∼30 • C, bees used solar heat gain for a dual purpose: to reduce energetic expenditure and to increase the thorax temperature by about 1-3 • C, in order to improve force production of flight muscles (Coelho, 1991a) and to speed up suction velocity . The aim of the present study was to investigate the contribution of radiative heat gain on the bees' thermoregulation during foraging for nectar and pollen under natural conditions. Kovac and Schmaranzer (1996), demonstrated in a comparison of honeybees foraging from 13 flowers, considerable variation of the thorax temperature. As a rule, the energy expenditure of individual foragers is balanced with the net energetic gains to the colony (Schmid-Hempel et al., 1985;Seeley et al., 1991;Seeley, 1995). The bees minimise the thermoregulatory costs during foraging by adapting their thorax temperature in response to the profitability of foraging at a food source and the colony's need for nectar and pollen (Stabentheiner & Schmaranzer, 1986, 1987Dyer & Seeley, 1987;Waddington, 1990;Stabentheiner & Hagmüller, 1991;Underwood, 1991;Stabentheiner et al., 1995;Stabentheiner, 2001;Nieh et al., 2006;Sadler & Nieh, 2011). From these investigations, we know the thoracic temperature to vary in a broad range of ∼30-44 • C. As flowers differ considerably in their profitability, i.e. as they vary in the amount of pollen and concentration and flow of nectar, the distance between single blossoms, and because the bees adapt their thorax temperature to profitability, the bees' thorax temperature at a certain flower is not predictable from measurements at other flowers. Therefore, to get a broader overview of the foragers' thermoregulation in their temperate living space, we investigated them on flowers at different locations and environmental conditions. Under Central European climate conditions, honeybee colonies undergo a typical seasonal population development, influenced by environmental and genetic parameters. The climax of the population strength and brood nest dimension is reached from the middle to the end of June (e.g. Seeley, 1985;Wille, 1985;Winston, 1987;Liebig, 1994;Imdorf et al., 1996). In spring, when the colonies have much brood and low food reserves, the bees should be more motivated to forage. In foraging honeybees, thorax temperature correlates with the insects' motivational state (e.g. Dyer & Seeley, 1987;Stabentheiner & Schmaranzer, 1987;Stabentheiner & Hagmüller, 1991;Underwood, 1991;Stabentheiner et al., 1995;Stabentheiner, 2001;Sadler & Nieh, 2011). Kovac and Schmaranzer (1996) presumed that season, beside ambient temperature, has an influence on thermoregulation. However, to test this hypothesis, data from more than 2 years and from multiple flowers were necessary, and measurements in sunshine had to be included (Kovac & Schmaranzer, 1996, had measured in shade). Our investigation covers a complete foraging season under Central European climate conditions. This allowed measurements over the entire range of ambient temperatures and solar radiation to which bees are probably exposed to during their foraging trips. Results should enable assessment of the relative importance of season and environmental factors.

Animals, field site, and measuring conditions
Measuring locations were the botanical garden in Graz and several orchards and meadows near Graz, Austria, Central Europe. We investigated honeybees (A. mellifera carnica) foraging nectar and pollen on 33 different blossoms of flowers, shrubs and trees, and collecting water from a rainwater barrel. To cover the entire foraging season and range of ambient temperatures honeybees are exposed to under Central European climate conditions, measurements were made on 26 days from March to October in 2006 (Table 1). Measurements were performed in different weather conditions, from overcast sky to bright sunshine. If no flowers were available in shade, a patch of flowers was shaded by a sunshade.

Measurements
The bees were filmed during the foraging stays at the blossoms (if possible from landing until takeoff) with an infrared camera (ThermaCam SC2000 NTS, FLIR, Stockholm, Sweden). We used infrared thermography because it allows temperature measurements without contact and behavioural impairment (e.g. Stabentheiner & Schmaranzer, 1987;Kleinhenz et al., 2003;Kovac et al., 2009a,b;Stabentheiner et al., 2010). In addition, it allows simultaneous temperature monitoring of all body parts during the entire foraging stay at one blossom. This is especially important in insects with a variable body temperature like honeybees. The behaviour of the insects was not impaired, which would not have been possible with 'grab and stab' methods with thermocouples or thermoneedles (Stone & Willmer, 1989). This outweighs the disadvantage of the method, which measures surface and not core temperatures. The surface temperature of a thorax heated to 40 • C at an ambient temperature of 21.5 • C is ∼1 • C below the subcuticular temperature (Stabentheiner & Schmaranzer, 1987;B. Heinrich, pers. comm.). The infrared camera was calibrated periodically by slotting in a self-constructed peltier-driven reference source of known temperature and emissivity (for details of calibration see Stabentheiner & Schmaranzer, 1987;. Thermographic data were stored digitally with a 14-bit resolution on a portable computer (DOLCH Flexpac-400-XG, Munich, Germany) at a rate of 3-5 frames s −1 . On 3 days, in addition to the nectar-gathering bees, water-collecting bees foraging at a rainwater barrel a few metres away from the nectar-foragers were also measured. The ambient air temperature (T a ) was measured near the foraging bees (∼1-5 cm) with thermocouples. In the near vicinity of the insects (<1 m), we also measured the relative humidity with NTC-sensors (in shade) and the solar radiation with a miniature global radiation sensor (FLA613-GS mini spezial, Ahlborn, Holzkirchen, Germany). Care was taken so that the radiation sensor was exposed to the same ambient conditions as the foraging bees. The temperature and radiation data were stored every 2 s with ALMEMO data loggers (Ahlborn, Holzkirchen, Germany).

Data evaluation and statistics
The temperature of the three bee body parts and of the blossoms' surfaces (in close vicinity to the bees' mouthparts) was calculated from the infrared thermograms ( Fig. 1) by means of the AGEMA Research software (FLIR, Stockholm, Sweden) controlled by a self-written Excel VBA-macro (Microsoft Corporation, Santa Rosa, California). The environmental data were automatically extracted from the datalogger files. Values of the body temperature during foraging were taken in regular intervals of about 3-5 s immediately after the insects' landing until their takeoff. This interval was chosen, because bees are able to increase or decrease body temperature within this time and temperature could vary considerably during foraging on one blossom (Fig. 2). The surface temperatures of the head (T hd ), thorax (T th ) and abdomen (T ab ) were calculated with an infrared emissivity of 0.97, determined for the honeybee cuticle (Stabentheiner & Schmaranzer, 1987;. Because the ThermaCam works in the long-wave infrared range (7.5-13 μm), the reflected radiation from the bees' cuticle produced only a small measurement error (0.2 • C for 1000 W m −2 ), which was compensated for. In this way we reached an accuracy of 0.7 • C for the body surface temperature of the bees at a sensitivity of <0.1 • C. The blossom surface temperature was calculated with an infrared emissivity of 0.95, representing a typical value for plants (Lamprecht et al., 2006).
The temperature gradient between the thorax and the ambient air (thorax temperature excess = T thorax − T a ) was used as a measure to assess the bees' endothermic capability. To evaluate the influence of the radiative heat gain on the body temperature, three classes of solar radiation were established: shade, <200 W m −2 , overcast sky, 200-500 W m −2 , and sunshine, >500 W m −2 . The mean of all foraging bees on one blossom type was calculated and values were divided into the three radiation classes. The values for the temperature excess of the head and abdomen were calculated in the same way.
The relationship between body temperature, temperature excess, and T a was described by linear, exponential or polynomial regression functions and tested with anova. Data analysis and statistics were performed using the Statgraphics package (Statistical Graphics Corporation, Warrenton, Virginia) and ORIGIN software (OriginLab Corporation, Northampton, Massachusetts).

Results
In 2006, we measured honeybees (A. mellifera carnica) foraging on 33 different flowering plants. Figure 1 shows thermograms of foraging bees on dandelion and apricot blossoms. From 1666 single forging stays we got 12 685 thermograms and evaluated the body surface temperatures of the head (T hd ), thorax (T th ), and abdomen (T ab ) as well as the blossom surface temperature (T blossom ) where the bees were sucking. We covered the complete foraging season (March-October) and the entire range of ambient temperatures (T a =∼ 10-33 • C) and solar radiation (50-1400 W m −2 ) to which they are likely to be exposed in their natural environment during a foraging trip in Central Europe. It must be noted that the investigated flowers often deliver both nectar and pollen (see Droege, 1989). We were unable to determine the relation of nectar and pollen load in the free-ranging individuals.

Body temperature and blossom surface temperature
The body surface temperatures during nectar and pollen collection on one blossom were not constant but fluctuated, especially during longer-lasting stays (Fig. 2). The continuous measurement with infrared thermography enabled the registration of this variability within the foraging stay. The mean body surface temperatures per plant and date varied in a wide range, T th from 23.2 to 44.2 • C, T hd from 18.6 to 43.2 • C, and T ab from 13.0 to 41.3 • C at ambient temperatures from 10.8 to 32.9 • C. A plot of all measurement data (Fig. 3) shows that at ambient temperatures of about 10-27 • C, T th was regulated rather independent of T a on average. At T a > ∼27 • C, however, it increased nearly linearly with T a (Fig. 3). The head and abdomen exhibited a stronger dependence on T a but both of them were regulated well above T a . The head was warmer and better regulated than the abdomen (Fig. 3). The abdominal temperature increased nearly linearly with T a . The relation of body temperature and ambient air temperature could be described best with an exponential function for the thorax (radiation: 0-1400 W m −2 , R 2 = 0.16185, Fig. 3, Table 2; A -E are the fit parameters): and with a simple linear regression for the head and the abdomen (radiation: 0-1400 W m −2 , head: R 2 = 0.41795, abdomen: R 2 = 0.64091, Fig. 3, Table 2): At a low T a of 10 • C, the average values of T th , T hd , and T ab derived from the regression lines were 35.6, 24.3, and 16.0 • C, respectively. In the medium range of T a , at about 20 • C, the T th decreased to 33.7 • C, the T hd increased to 29.6, and the T ab increased to 24.9 • C. At the highest T a measured (∼33 • C), T th , T hd , and T ab increased to 44.4, 37.2, and 37.5 • C, respectively. In order to allow a comparison of the results of flower-visiting bees with water-foraging honeybees (from the paper of Kovac et al., 2010), the regression lines for the three body parts of the water foraging bees are also displayed in Fig. 3 (for statistical details see Table 2).
Plotting the body temperature in dependence on three levels of solar radiation (<200, 200-500, >500 W m −2 ; Fig. 3) revealed that bees foraging in sunshine were mostly warmer than bees foraging in shade. The relation of thorax temperature and ambient air temperature could be described best with a polynomial function (radiation: <200 W m −2 : R 2 = 0.20651, 200-500 W m −2 : R 2 = 0.18030, >500 W m −2 : R 2 = 0.48709, Fig. 3, Table 2; A -D are the fit parameters): and with a simple linear regression [eqn (2)] for the head (radiation <200 W m −2 : R 2 = 0.59942, 200-500 W m −2 : R 2 = 0.72383, >500 W m −2 : R 2 = 0.72718) and the abdomen (radiation <200 W m −2 : R 2 = 0.87821, 200-500 W m −2 : R 2 = 0.82406, >500 W m −2 : R 2 = 0.76073). For further statistical and graphical details see Table 2 and Fig. 3. The temperature difference between >500 and <200 W m −2 as estimated from the regression lines of Fig. 3 was smaller at low and greater at high T a (T a = 12 The blossom surface temperature (range T bl = 9.5-42.2 • C) measured closely beside the bees' mouthparts increased linearly in dependence on T a at all three categories of radiation (Fig. 4, Table 1, statistical details in Table 2). In sunshine the blossoms' temperature was about 4 • C elevated above the ambient air temperature. Under (partly) overcast skies (200-500 W m −2 ) the T bl was also always higher than the ambient air temperature. However, the blossoms' temperature in shade was similar to the ambient air. The three regression lines differed significantly (anova, P < 0.0001, F -Ratio = 68.35, d.f. = 5), and the intercepts of values in sunshine versus the two other categories of radiation were also significantly different (P < 0.01; F -Ratio = 6. 83, 9.53, 36.32; d.f. = 1).

Temperature excess and solar radiation
The bees were always endothermic as the thorax (the centre of heat production) was clearly more elevated above the ambient air than were the other body parts. The thorax temperature excess (T th − T a ) depended strongly on T a . It decreased significantly with T a in the sunshine and in the shade (values calculated from linear regressions in Table 3; T th − T a = 20.6 − 8.2 • C at T a = 12-30 • C and radiation >500 W m −2 ; T th − T a = 21.6 − 3.6 • C at T a = 12-30 • C and radiation <200 W m −2 ; P < 0.0001). The temperature excess of the intermediate radiation range (overcast sky,  Table 2. 200-500 W m −2 ) showed a similar course. An anova confirmed the difference in thorax temperature excess between sunshine and shade (P < 0.01, F -ratio = 11.01, d.f. = 1 for intercepts, and P < 0.05, F -ratio = 4.61, d.f. = 1 for slopes). The excess temperature of the head decreased with T a as well, but the slopes were somewhat flatter than for the thorax. The decrease was still significant (P < 0.0001, Table 3). However, the temperature excess of the abdomen decreased with T a only in the shade (P < 0.05) and remained constant between 12 and 33 • C in the sunshine and overcast sky (Table 3). An anova confirmed the difference in temperature excess between the sunshine and the shade (head: P < 0.0001, F -ratio = 75.98, d.f. = 1 for intercepts, and P < 0.05, F -ratio = 4.50, d.f. = 1 for slopes; abdomen: P < 0.0001, F -ratio = 102.15, d.f. = 1 for intercepts, and P > 0.05, F -ratio = 0.77, d.f. = 1 for slopes).

Temperature and season
In Fig. 5, the mean temperatures of the three body parts during foraging in the shade are plotted against the date of observation (a) and ambient temperature (b) for each flowering plant. The T th revealed a clear dependence on the season. The average value in the spring (March-June) as calculated from the means per stay was 35.2 ± 2.3 • C, (N = 218). In the summer (July-September), it was only 31.4 ± 2.4 • C (N = 127) in spite of the higher T a in the summer. The difference Table 2. Equations of linear and non-linear regressions for the temperature of the thorax (T th ), head (T hd ), and abdomen (T ab ) of honeybees foraging on flowers or foraging for water (*, Kovac et al., 2010), and of the blossom temperature, in dependence on ambient temperature (T a ) and solar radiation (Fig. 3 Table 2. could be statistically confirmed (Mann-Whitney/Wilcoxon's test, P < 0.0001; W = 3259.0). Testing the effect of season on average T th per foraging stay and bee with anova (removing the effect of T a and radiation) showed the same result (main factor season: P 0.0001, F -ratio = 247.07, d.f. = 1; covariate T a : P 0.0001, F -ratio = 40.62, d.f. = 1; covariate radiation: P = 0.4224, F -ratio = 0.65, d.f. = 1; N = 345). F -ratios indicate that season had the greatest effect followed by T a , and radiation had no effect. Plotting the average values of T th against the ambient temperature (Fig. 5b) and calculating means for ranges of T a according to Kovac and Schmaranzer (1996) Table 1 gives an overview of body temperature and environmental parameters for each measuring day and plant divided in three classes of solar radiation (<200 W m −2 ,   Kovac & Schmaranzer, 1996). The horizontal lines are mean thorax values of two seasons (spring and summer) or ranges of ambient temperature (T a ). Values of the thorax temperature for the dandelion is marked with pink circles. Mean values between seasons (a) and between ranges of T a (b) are significantly different (P < 0.05, see Results).

Type of flower
200-500 W m −2 , >500 W m −2 ). It is of special interest that bees measured on the same day in the same environment and similar T a at different plants could exhibit remarkable differences in their thorax temperature. For example, bees foraging in the shade at apricot blossoms (Prunus armenica L.) had an average thorax temperature of 36  Table 1).

Ambient temperature and radiation
For a comprehensive description of an insect's thermoregulatory performance, it is of great advantage to investigate the entire range of ambient temperature to which it is likely to be exposed in its natural environment. Infrared thermography enabled us to measure the temperature of all three body parts of undisturbed foragers and revealed new knowledge about their thermoregulatory behaviour. An interesting result was that the bees regulated the T th at a rather constant level in a broad range of T a (10-27 • C) on average but showed a strong variation at a certain T a , depending on the plants from which they were foraging (Fig. 3). To our knowledge, there are only two similar investigations on this topic (Heinrich, 1979a;Kovac & Schmaranzer, 1996). Heinrich's study (1979a) is a pioneer work in this field. He measured A. mellifera mellifera foraging from Eucalyptus sp. and A. m. adansonii foraging from B. pilosa and Petraea volubilis (in the shade). The bees exhibited a thoracic (core) temperature of ∼30.5-33 • C at ambient temperatures of 11-22 • C. The (surface) T th of our foragers measured in the shade was clearly higher (Fig. 3), with average values ranging from ∼35 to 33 • C. Kovac and Schmaranzer (1996) reported even higher mean thorax surface temperatures of 35-38 • C in bees foraging nectar from several plants. Water foragers measured in the same environment  regulated the thorax to another 2-3 • C higher (T a = 10-27 • C; Fig. 3). At a T a above 27 • C, the T th increased somewhat more steeply in the nectar foragers than in the water foragers. At these high ambient temperatures, the bees' main problem seems not to be that their body temperature is too low. Rather, the dissipation of excessive heat becomes more important. More bees returning to the hive were shown to carry a fluid droplet at these temperatures (Cooper et al., 1985). Such droplets have a considerable cooling effect not only on the head but also on the thorax (Heinrich, 1979a,b). We suggest that cooling was more difficult for the nectar than for the water foragers because their head temperature became higher at a T a above ∼27 • C (anova, P < 0.0001; F -ratio = 296.07; d.f. = 3; Fig. 3).
At very low T a , by contrast, it seems to be more important to keep the head warm. The haemolymph circulation from the warm thorax (Heinrich, 1979b(Heinrich, , 1980aCoelho, 1991a,b) provided the head with enough heat to prevent the T hd from falling below ∼20 • C, which seems to be necessary for the proper functioning of physiological and neural processes. Regulation of the T th at a high level even at low T a allows the bees to keep the T hd at a level high enough to guarantee a high suction speed at unlimited sources . In nectar foragers a high nectar suction speed is generally not as important because the nectar is not available in an unlimited amount. Nectar foragers usually get only small portions of nectar per blossom and then have to fly or walk to the next blossom.
The temperature of the nectar foragers' abdomen was mostly below that of water foragers, probably because of the lower thorax temperature and perhaps because water foragers foraged much closer to the nest in the present study (Fig. 3). Heinrich (1980bHeinrich ( , 1993 suggested that bees use a series of aortic loops in the petiole as a counter-current heat exchanger to prevent heat leakage to the abdomen in the cold. We agree with this opinion. However, the amount of heat reaching the abdomen may differ considerably. In contrast to the present study, where the abdominal temperature was not much elevated above the ambient temperature, it was considerably increased at low ambient temperatures in other previous investigations (Kovac & Schmaranzer, 1996;Kovac et al., 2010). Digby (1955) investigated the factors affecting the temperature excess of dead or anaesthetised insects in artificial sunlight under laboratory conditions and found the temperature excess to vary directly with the radiation strength. This applies to living insects only in the ectothermic state. Foraging honeybees, however, are always endothermic at medium to low T a (Figs 3; Heinrich, 1979a;Waddington, 1990;Kovac & Schmaranzer, 1996;Kovac et al., 2010). On average the thorax temperature excess was higher in our measurements on A. mellifera carnica than in an investigation on A. m. mellifera and A. m. adansonii by Heinrich (1979a). An important result of our investigation was that the bees used the heat gain from the sun to enhance their body temperature. This enables a quicker exploitation of the flowers because a high body temperature not only increases suction speed  but also increases the bee's agility Stabentheiner et al., 2003b) and flight muscle power output (Coelho, 1991a). However, at a high T a of ∼30 • C our bees probably were only weakly endothermic. The thorax temperature excess in sunshine of ∼8 • C above ambient air was only ∼1.5 • C higher than the abdominal excess. The finding that in shade the thorax temperature excess was only ∼3.5 • C confirms that they were only weekly endothermic. At these high ambient temperatures the bees foraging in the sunshine are able to reach the optimal upper level of T th for force production and takeoff of 38-39 • C (Coelho, 1991a) without much endothermic effort.
We often observed that bees preferred flowers in sunshine to flowers in shade. Our measurements of the blossom surface temperature (Fig. 4) showed that the solar radiation elevated their temperature by about 4 • C above the ambient air. Dyer et al. (2006) found that floral temperature can serve as an additional reward for pollinator insects when nutritional rewards are also available. However, we cannot exclude from our results that bees preferred the warmer flowers in sunshine owing to greater amounts of nectar secretion, because the production and concentration of nectar depends on ambient temperature and relative humidity (e.g. Beutler, 1953;Shuel, 1970;Núñez, 1977;Corbet et al., 1979aCorbet et al., ,b, 1993Szabo, 1984;Corbet, 2003).

Seasonal variability and type of plant
A great part of collected nectar and pollen is used to provide for the brood and young bees of the colonies. Brood rearing and colony development proceed in a special periodicity. In Central Europe the majority of the brood is reared in the spring until the beginning of the summer (e.g. Seeley, 1985;Wille, 1985;Winston, 1987;Liebig, 1994;Imdorf et al., 1996). During this time, colonies need huge amounts of nectar and pollen. The presence of a brood stimulates the foraging behaviour of the bees (Pankiw et al., 2004). We had presumed that bees foraging in the spring are better motivated and should therefore have a higher T th (Dyer & Seeley, 1987;Stabentheiner & Schmaranzer, 1987;. In Fig. 5a the mean T th of each investigated plant is plotted