ABSTRACT
The sun navigation of honey bees has been investigated in a region of the southern hemisphere where the sun was moving counter-clockwise during the observations.
Foragers from a strain long established in the region were fed in the evening on a dish in a particular geographical direction and transferred overnight to a new locality unknown to them. During the next day the majority of bees were at all hours searching in the direction of their previous training.
Foragers which were the offspring of queens recently imported in an inseminated state from the northern hemisphere showed after similar training systematically false orientation on the day of observation. The direction of their search shifted by about 30° per hour counter-clockwise from a direction about 90° clockwise from the training direction in the morning to the correct direction in the evening.
Bees of hybrid (local and imported) origin also showed false orientation.
The existence of innate mechanisms is postulated compensating in northern bees for the sun’s clockwise movements and in southern bees for the counterclockwise movement of the sun. The change in the direction of compensation must have occurred during the last 425 years, and a possible mode of this evolutionary process is discussed.
INTRODUCTION
Honey bees can orientate by means of the sun, taking the hour of the day into account (v. Frisch, 1952); colonies containing workers which had been trained during a sunny evening to search for food in a particular geographical direction were transferred overnight to a locality unknown to them; during the following sunny morning the trained bees searched predominantly in the ‘learned direction’, in spite of the fact that during the training in the evening the sun had been in the west, while during observation on the following morning it was in the east. This faculty of recognizing the geographical direction, which bees share with some other arthropods and with birds (see Kalmus, 1954), can be explained by assuming the working of an internal mechanism or reference system which compensates for the daily movements of the sun’s azimuth. In temperate northern regions, where v. Frisch’s observations were made, the daily course of the sun is clockwise ; it seemed interesting to study the behaviour of honey bees trained in a similar way in a region of the southern hemisphere, where the sun moves counter-clockwise.
Investigations were carried out on bees whose ancestors had lived for several centuries in the southern hemisphere and on bees which were of recent northern hemisphere origin and whose parents had been transferred to the southern hemisphere.
MATERIAL AND METHODS
The experiments were performed during the south Brazilian winter in May and June 1955, when fairly dry weather, the paucity of bee crops and the low elevation of the sun at noon offered suitable conditions. The region was that of Piracicaba (220 40’ S.) in the state of São Paulo; there vast areas planted with sugar cane as well as fallow and orchards provided numerous sites poor in landmarks, and the existence of a well-equipped bee department in an agricultural college made the organization of the work possible.
The following types of colonies were used in the investigation :
Black bees belonging to a local bee keeper; these were said to be mainly descended from European colonies imported from Portugal during the last two to four centuries, with some possible admixture of later imports also from Europe.
Locally reared light offspring from ‘Italian Queens’ bred and inseminated in California and brought to a Brazilian apiary. (No workers actually reared in the northern hemisphere and transferred to Brazil were available.)
‘Hybrid ‘bees, descendants from ‘Italian queens ‘imported during the last decade from California into Brazil, showing by their coloration various degrees of genetical admixture with local black drones.
The existence of a mechanism compensating for the daily changes of the sun’s direction has been demonstrated in two different ways as described by v. Frisch (1952): ….
By observing a gradual change in the direction of dance of a few accidentally observed bees which for unknown reasons continued to dance for an hour or more instead of for only a few minutes.
By shifting colonies containing workers trained during the evening to search for food in a particular direction, and by observing the direction in which they search on the following morning. Only this second method was used in the present work ; details can be gathered from the following descriptions of experiments.
RESULTS
The results of five experiments are described in some detail and all other results are summarized.
A. Black Brazilian honey bees
In the evening of 6 June, hive no. LA containing a flourishing colony of local black bees was put in the middle of a large flat field covered with about 1 ft. high sugar cane in rows 1 m. apart, running in an approximately north-south direction. The sugar cane in adjacent fields was higher; no larger trees were anywhere nearer than a mile. In the morning of 7 June, which was a sunny day throughout, the hive entrance (which was facing north-west) was opened, and at noon a dish of scented syrup was put on each of three wooden stands near it. Feeding bees were observed after hr. The stands were then gradually moved away from the hive in a direction of 15° E. of N. ; a distance of 140 m. was reached at 16.30 hr. Two stands were now removed, leaving only one. The initially small number of visitors now increased rapidly so that it was possible to mark 217 bees between 16.43 hr- and sunset at about 17.36 hr. Half an hour later the hive entrance was blocked and the colony was moved to a young orchard some 5 km. away. There were no rows of sugar cane and higher trees were entirely absent.
Early in the morning of 8 June, four stands with scented syrup dishes similar to those used on the previous day were placed at a distance of 140 m. from the colony in the following directions: 15° E. of N., 15° S. of E., 15° W. of S., 15° N. of W. Observers were posted to kill and record all honey-bee visitors to these dishes, and at 8.38 hr. the hive entrance, now facing north-east, was unblocked. Apart from some haze during the first 2 hr. of observation the weather was sunny and only an occasional breeze from the north interrupted the quiet air. Fig. 1 shows the visits of marked and unmarked bees which were caught on the four dishes during the day. Only one visitor escaped, and it did not return. Bees which circled in the neighbourhood of the observers but did not settle were noted but are not included in the figure.
Fig. 1 indicates that at all times most bees were caught at the dish in the direction to which they had been trained. The results of this and of three similar experiments with black local honey bees trained in different geographical directions are summarized in Table 1 ; in the four experiments 81, 76, 33 and 50% of the bees were caught before noon at the geographical direction of training.
The proportions of successfully orienting bees (Table 2) show remarkable agreement with the results obtained by v. Frisch (1952) in Bavaria. Thus the local honey bees of southern Brazil show the same degree of efficiency in sun navigation as the bees in Bavaria.
B. Locally reared workers of Italian Stock from queens bred and inseminated in California
In the evening of 11 June hive NA containing an ‘Italian’ queen bred and inseminated by an ‘Italian’ drone in California and her locally reared worker offspring, was blocked and transferred from the apiary to an experimental field of the Genetical Department some 2 km. away. On the morning of 12 June, which was a sunny day throughout, the hive entrance (facing due north) was unblocked. From 14.00 hr. onwards three stands with scented syrup dishes were moved from near the entrance in a generally western direction until at 14.36 hr. a distance of 160 m. was reached in a direction 15° N. of W. Two dishes were now removed. The small number of bees which had visited the dishes during the ‘pulling out period ‘was now rapidly increased by offering more concentrated syrup, and until 17.30 hr. 400 bees were marked while feeding. The next morning (13 June) was a rainy one and no bees went foraging. The weather, however, improved in the afternoon, and many bees went to the syrup dish 160 m. distant and 15° N. of W., where they were fed between 15 and 17 hr. More than 100 bees marked the previous day received an additional mark and thirty were newly marked. After sunset the hive was blocked) and transported to a field of young sugar cane some 17 km. away.
During the next day (14 June) slight cloud somewhat obscured the sun during short periods in the morning; later it was sunny throughout. Four observation stands were erected 160 m. away in the following directions: 15° N. of W., 15° E. of N., 15° S. of E., 15° W. of S. The hive was unblocked at 8.21 hr.
Fig. 2 shows the numbers of bees captured at the four dishes, and in Table 3 these results and those of two similar experiments are summarized. The results reveal a pattern of visits quite different from that of the black Brazilian bees. Only seventy-four (40%) out of 185 captured workers were collected at the dish in the direction of training, and most of these (49) only after 15.00 hr. The visits of the bees showed a definite trend, the maximum of visitors moving counter-clockwise from dish to dish in the course of the day.
C. Hybrids
Fig. 3 shows the outcome of an experiment with colony LB of black local bees and a colony of multicoloured hybrids (HyA) descended from light ‘Italian’ stock (imported about 8 years previously from California) which had mated with local black drones. On the evening of 28 June colony LB was put in the middle of a large field of young sugar cane, with the hive entrance facing due east. On 29 June a number of the dark workers were trained to forage at 150 m. due south, and while feeding there between 16.00 and 17.43 hr-forty of them were marked. In another field some 5 km. away hybrids from colony HyA (hive entrance facing east) were also trained out 150 m. due south, and fifty-three of these workers were marked while feeding between 16.05 hr. and sunset at about 17.52 hr. In the evening both hives were blocked and placed side by side in a third field, both entrances facing north. The field was 4 km. from the first site and km. from the second one. On the morning of 30 June, four observation dishes were placed at 150 m. distance in the four directions of the compass and the hives were unblocked at 07.20 hr. It was rather windy and the sun was frequently hidden by clouds. Fifty marked hybrids but only seven marked bees from LB were caught during this day; il appears that while the few local bees were searching in the direction of training during the morning, the hybrids were not, but behaved rather like recently imported ‘northerners’. Table 4 summarizes these results as far as the hybrids are concerned and the results of a second similar experiment.
D. Experiment with eight observers
An attempt was made to test simultaneously in one experiment local honey bees, bees from a queen recently imported from the northern hemisphere, and hybrids, using at the same time eight observers instead of four. Although this experiment met only with limited success its results are given below (Fig. 4).
On 21 June colony HyC, with workers mainly descended from Italian Stock but with some black Brazilian drones in their ancestry, was put in the genetics grounds and opened facing west. On the same day hive ND, containing an Italian queen imported after insemination from California with her offspring, and hive LE, containing black local bees, were put side by side in a field of young sugar cane some 14 km. from the site with the hybrids. Both these hives were also facing west. In the evening of 22 June bees from all three hives were gradually trained due west and between 16.00 hr. and sunset ninety-eight black Brazilians, twenty-five light ‘Italians’ and 273 hybrids were marked with different paints, while feeding at dishes 200 m. away from their hives in their respective fields. The three colonies were blocked in the evening and transported to an orchard 8 km. from the genetics ground and 6 km. from the sugar-cane field. Here they were put side by side all facing due north. Unfortunately, the hybrid colony was imperfectly blocked so that many marked bees escaped during the transport.
On 23 June eight dishes were laid out in the directions indicated in Fig. 5 and the hives were unblocked at 8.26 hr. The day was mostly sunny, but a cold south wind kept the activity of the bees at a low level. The results represented in Fig. 4 may be regarded as additional evidence for the conclusions reached above concerning the orientational behaviour of the three kinds of honey bees.
DISCUSSION
It now remains (1) to correlate the visits of the foragers with the daily movements of the sun, (2) to discuss the differences of navigational behaviour of the various strains and (3) to give some indication of the way in which such differences may arise. Some results obtained by other workers are also included in the discussion.
Our basic assumption is that the navigational abilities of bees are regulated by phenomena associated with the daily movements of the sun and in particular by the sun’s azimuth. The rate of displacement of the sun’s azimuth varies with locality and time. When the vertical equatorial sun moves through the zenith the azimuth changes instantaneously from east to west.* In temperate regions where v. Frisch’s original observations were made, as well as in California and in Pira-cicaba, during the months of our observations displacement of the sun’s azimuth is gradual but not quite uniform. Calculations made to correlate the visits of the bees with the exact position of the sun during the day were inconclusive and are not reported here; probably results from this type of experiment are inherently too inaccurate for such a purpose (v. Frisch, 1954). For the following interpretation a gross uniform hourly displacement of 150 per hour is being assumed.
The top row of Fig. 5 illustrates the sun navigation of bees bred and trained in temperate northern regions; fed at 16.00 hr. due south, i.e. ϕ = 6o° counter-clock-wise, from the sun they searched due south at 9.00 hr. the following morning. In the interval the sun’s azimuth has been displaced in a clockwise direction by the angle +αN =205°, and the hypothesis is that an internal reference system has compensated for this ; this is indicated by the introduction of the counter-clockwise angle—αN of equal magnitude.
The behaviour in similar circumstances of southern hemisphere bees is illustrated in the middle row of Fig. 5. Here the sun moves counter-clockwise and the azimuth displacement of + ±S is added to ϕ = 120°. This has been compensated by the clockwise angle—±S and the bees thus strike out in the correct direction, i.e. due south.
The difference between the counter-clockwise compensation of northern bees in the northern hemisphere and the clockwise compensation of southern bees in the southern hemisphere could be explained in two different ways: (1) by assuming that an individual forager learns the direction of the daily course of the sun and compensates for it accordingly and (2) by assuming that the sense of compensation is innate and that an evolutionary change has occurred since the importation of the ancestors of the southern bees from their original northern home.
The results obtained with recently imported bees make the latter assumption more reasonable (Fig. 5, bottom row). Here the direction of training, as well as the position of the sun during training and observation are the same as in the experiment described in the middle row. However, instead of compensating for the counter-clockwise displacement of the sun’s azimuth by the addition of the clockwise angle—<Xg, the disorientated bees add the counter-clockwise angle—±N as if they, like their immediate ancestors, were living in the northern hemisphere. Thus they strike out in the direction D which is almost opposite to south, the direction of their training.* Generalizing from the assumption illustrated in Fig. 5 one can predict the direction of search of the wrongly orientated bees at any particular hour; this must change by 720° in a counter-clockwise direction in 24 hr. The bees are then expected to start in the morning in a direction about 90° clockwise from south, the direction of their training ; and the direction of their search would change by about 30° per hour in a counter-clockwise direction until they reach south at 16.00 hr., the time of their previous training. The trends in visiting shown in Fig. 2 and Table 3 agree with this hypothesis.
It now remains to discuss the change in the innate compensating mechanism which we infer as having occurred during the period since the first honey bees were shipped from Portugal to Brazil, i.e. since 1530 (Schenk, 1938; Emelen, 1945). The exact time and rate of this change is unknown, but our observations on hybrids suggest that it may not have occurred very quickly. A few words may be said on one aspect of such an evolutionary transformation. Local bee keepers agree that recently imported Italian-Californian bees are at least as successful as foragers as the old Brazilians, and one might infer from that that sun navigation does not play an important role in the foraging of workers. It may, however, be important in other situations, e.g. when the bees swarm (Lindauer, 1951). While nothing is known concerning the sun navigation of queens and drones it is not unreasonable to assume that similar orientational faculties exist in the three castes of honey bees, and if so, sun navigation may be important for the nuptial flight. Thus it is possible that the change in navigational behaviour, which we infer occurred in Brazil, was brought about by the selection of the sexual forms. But no quantitative data exist on this point.
Honey bees, and especially queens, are being shipped in increasing numbers across the equator, and it might be interesting to repeat the above experiments and to see whether similar changes in navigational behaviour can be detected in different circumstances. If meliponids and trigonids should be found capable of sun navigation their transfer to the opposite hemisphere would also be of experimental interest.
ACKNOWLEDGEMENTS
The experiments described in this paper were performed during a stay in Brazil made possible by the award of the André Dreyfus prize for genetics for 1954, and the receipt of a travelling grant from the Rockefeller Foundation. Personnel and facilities of the Genetics Department of the Escola Superior di Agronomía Luiz Queiros in Piracicaba, Estado São Paulo, were kindly put at my disposal by Prof. F. G. Brieger. Dr Warwick E. Kerr, of the same department, and Dr Erico Amaral, of the Bee-keeping Department, as well as several technical assistants were indispensable helpers.
Dr R. H. Garstang of the Department of Astronomy, University College London, has advised me on the calculations of the azimuths.
REFERENCES
A corresponding sudden change in the orientation of Apii indica, a close relative of the honey bee, has been reported by Lindauer (v. Frisch, 1955), who transported a colony containing direction-trained foragers across the equator and observed them afterwards.
A type of false orientation similar to that of the Italian-Califomian bees described in the present paper seems to have occurred in individuals of the amphipod Tallinn laltator, which Papi (i955) transported from Italy to the Argentine. This seems to be the only experiment in the literature which has a bearing on our results.