1.1 Economic importance of bumblebees
1.2 Life cycle
1.1 ECONOMIC IMPORTANCE OF BUMBLEBEES Bees are responsible for pollinating plants that provide much of our food; in North America it is believed that 30% of food for human consumption originates from plants pollinated by bees (Heinrich, 1979). Honeybees are generally thought of as the most common pollinator, and they are the most widely studied, but bumblebees are the chief pollinators of red clover, alfalfa, and in some areas cotton, raspberries, apple and plum blossom; and in Norwegian orchards honeybee pollination is usually limited by low temperatures, so bumblebees are the chief pollinators there (Faegri & van der Pijl, 1979). In New Zealand bumblebees were imported (they have no native bumblebees) to pollinate the forage crop clover (Trifolium pratense) as the corolla is too long for honeybees, so they cannot act as pollinators (Sladen, 1912; Faegri & van der Pijl, 1979). Bumblebees are such good pollinators of these crops for three reasons:
Bumblebees are increasingly used in glasshouse cultivation, where a honeybee hive would be too large, e.g., cabbage pollination in Holland (Free & Butler, 1959), kiwi fruits and tomatoes. So bumblebees are of great economic importance, and with the increase of glasshouse cultivation, and the spread of the mite, Varroa jacobsoni, causing a decline in honeybee populations, their importance can only increase. A decrease in the honeybee population would probably lead to an increase in the populations of short-tongued bumblebees, as happened in Colorado recently (Plowright & Laverty, 1984). Consequently it is most important that there is adequate information on the pollination ecology of bumblebees, hence this study.
1.2 LIFE CYCLE
Bumblebee colonies have a yearly cycle. Queens that have mated in late summer hibernate, usually in the soil, and emerge in spring. The queens found the new colony themselves, they search for a suitable nest site, deserted small rodent nests are popular sites, then they build a wax honey pot and fill it with regurgitated nectar (honey). This store will enable them to survive a day or two of bad weather without foraging. The queens also build up a store of pollen, some of this they eat, and the rest they form into a ball mixing it with nectar. The pollen stimulates the ovaries to produce eggs, which the queen lays in batches of 4-16 (Free & Butler, 1959), on the ball of pollen. The ball of pollen with the eggs is placed within reach of the honey pot, this enables the queen to brood the eggs and drink honey at the same time. After this stage the various species differ slightly in the way the larvae are fed. The larvae pupate and emerge as adults, the queen usually lays another batch of eggs while the first batch is still in the larval stage (Sladen, 1912).
When new workers first emerge their hair is silvery in colour, within an hour or two it changes to the colours seen in foraging bees (Sladen, 1912). The workers can start to forage after only two or three days (Heinrich, 1979), this means that the queen can spend more time egg laying. The average worker's life lasts about four weeks, and in that short time she might develop foraging preferences, this thesis aims to record the preferences of the species and of the individuals. Not all adults leave the nest to forage, some of the smallest workers may stay in the nest and perform "household" duties; these small workers may also have weak or deformed wings, but may live longer than the foragers and have less worn coats (Free & Butler, 1959) and wings, as they rarely fly.
The size a colony reaches depends on the species concerned and the food supply, some can have as few as 30 bees, and Bombus terrestris can reach as many as 400 (Free & Butler, 1959). Males are usually produced once the stores reach a sufficient quantity, or if the queen dies or loses her influence. When the adult males emerge they spend a few days in the nest, but do no work, then they leave the nest for good and forage for themselves. They can often be seen sheltering under the heads of flowers when it rains or when it gets dark. Not all workers return to the nest every night, some spend the night outside, sheltering under flowers as the males do. New queens emerge about a week or so after the males. They mate, drink lots of nectar to build up their fat body, which will enable them to survive the winter hibernation, then find a suitable place to hibernate.
Not all nests go on to produce males and queens, many fail in the early stages, some are damaged, and some never build up enough reserves to produce reproductives. Some nests produce only queens, others only males, and some nests produce both males and queens. Although it is possible for workers to lay unfertilised eggs that will be males, workers cannot produce queens or other workers.
At about the same time as the queen starts laying unfertilised eggs that will produce males, the ovaries of some workers, usually those performing household duties (Alford, 1975), may develop. When this happens aggression between worker and worker, and worker and queen increases. Some workers try to lay eggs of their own, and may even attempt to eat eggs laid by the queen. In many cases the more persistent workers will succeed in laying some eggs. The production of males usually signals the beginning of the end of the co-operation and organisation of the nest. The males drink the stores of honey, but do not forage to replace it. The queen is usually old, almost bald and may have lost some of her influence over the persistent or larger workers, and gradually the stores dwindle and the workers die.
There is some evidence that Bombus pratorum and B. hortorum may, under certain circumstances, be able to go through two colony cycles in a year (Prys-Jones & Corbet, 1987), this would mean that instead of hibernating the new queens would immediately start a colony and the queens that emerge from this colony would then hibernate through the winter. B. pratorum and B. hortorum usually have quite small colonies, and B. pratorum colonies can reach their maximum size earlier in the year than any other species.
There are some very important differences between the bumblebee life cycle and the honeybee life cycle. There is no mouth-to-mouth exchange of food between adult bumblebees, nor do adults groom each other or the queen (Free & Butler, 1959). As yet no "queen substance" has been found (Wilson, 1971); in honeybee hives workers licking the queen and each other pass the queen substance throughout the hive, and this pheromonal control enables the queen to maintain dominance. Bumblebee queens appear to maintain dominance purely by aggressive behaviour. They are usually bigger than the workers and the queen opens her mandibles and head-butts the most dominant worker from time to time (Free & Butler, 1959). This is usually sufficient until unfertilised eggs are laid, or a worker's ovaries develop.
1.3 FORAGING AND CONSTANCY
Bumblebee larvae eat most of the pollen brought back to the nest, adult bees eat very little (Free & Butler, 1959). Until the end of the final instar the larvae have a blind gut, faeces are voided all at once during the spinning of the cocoon and final instar (Alford, 1975). The highest growth rate occurs during the last instar (Alford, 1975) so most pollen will be consumed during this time. The exine (outer wall) of the pollen grain is made of a tough carotenoid polymer which is highly resistant to decay (Toothill, 1984) and so can be identified in the larval faeces.
A.D. Brian (1951) analysed the larval faeces of three nests, one each of Bombus pascuorum, B. lucorum and B. hortorum. The nests were located quite close to each other, so the bees would have probably been foraging from the same area. She found considerable differences in the kinds of pollen eaten by the larvae of the different species. In a later study (1957) she found that bumblebees prefer to forage from flowers with corollas a little shorter than their tongue length. She thought that this might increase foraging speed, and also it seemed that they did not like pushing their heads into the flowers.
Inouye (1980) found that, in general, shorter-tongued bees foraged faster on short-corolla flowers than long-tongued bees, although he found it difficult to measure the foraging times of longtongued bees on shorter corolla flowers because of "their apparent reluctance to feed on short corolla flowers". With captive bees feeding from artificial flowers in the laboratory it was found that the probing time increased gradually with depth of flower, providing the flower was shallower than the bee's tongue, but beyond that depth probing time increased much more rapidly, as the bees stretched their tongues but failed to reach the nectar (Harder, 1983).
J.B. Free (1970) measured constancy by analysis of pollen loads of bees returning to the nest, he found that 63% of B. pascuorum and 34% of B. lucorum loads contained a mixture of pollen from different flower species, however many of these "mixed" loads contained 98% or more pollen from one plant species.
The analysis of pollen indicates some of the plants the bee visits, but it does not necessarily show all the plants visited by bees, as not all bees collect pollen on foraging trips. Some bees specialise in pollen collection, some in nectar, some in both, but all will change according to the needs of the colony (Free & Butler, 1959). Pollen load analysis does not show the foraging patterns of nectar gathering bees.
Heinrich (1976) in the U.S.A. studied bumblebees foraging in an old field. He confirmed the species preferences of earlier studies, but also found that each individual preferred a small subset of the overall species group. He named the main foraging flower the "major" and the secondary flower(s) the "minor(s)".
The foraging environment of the bumblebees is constantly changing. Individual flowers and groups of flowers of the same and different species come into bloom and die at different times with different life-spans. Nectar changes throughout the day in volume and concentration (PrysJones & Corbet, 1987) and throughout the life of the flower (Real & Rathke, 1988; Willmer et al., 1994); soil conditions and daily temperature and humidity changes also affect the quantity of nectar available. Worker and queen bees aim to gather more food than they themselves will consume, in order to build up a store of surplus food. Males do not aim to build up a store, they forage only for themselves, so their foraging patterns might be expected to be different. In this constantly changing environment individual bees may get a different perception of which flower species is most rewarding. The best strategy for an individual bee is to forage from whatever species of flower it perceives to be the most rewarding, and change to another species when the original species is less rewarding and the new species is abundant enough to provide sufficient nectar.
An optimal forager would visit all flowers with rewards above a certain level, in its foraging area, to minimise energy costs and maximise energy gains (Real, 1983), but individual bumblebees regularly ignore apparently rewarding flowers. Sladen (1912) observed a Bombus lapidarius queen persistently fly against a closed window in an effort to reach a bunch of bluebells on the other side, while there were plenty of other suitable species outside that she could have foraged from. There may be constraints that limit the number of different types or species of flower that a bumblebee can forage from. Darwin (1891) thought that there might be a limit to the number of types of flower an insect could learn or remember to handle correctly. It has been shown that flower complexity can cause learning problems in bees (Heinrich, 1976; Laverty, 1980) and that bumblebees are often constant to a guild of species with the same morphology (Manning, 1956), or colour (Darwin, 1891). Real (1983), in Costa Rica, found a slightly greater bumblebee constancy in a guild of species differing only slightly in morphology, than in a guild differing only in colour. As the number of different types of flower increases the constancy to one or a few types increases (Waser, 1986). Laverty (1993) found that constancy was higher for the more morphologically complex species than for the simpler species however, even with the simpler species, bumblebees switched to a new species that had a similar corolla length and handling method, more frequently than would be expected if the switch was random. Tests done in captivity have shown that bumblebees can recognise rewarding flower "types" by their colour, odour, or shape of flower, and that they will accept flowers that are slightly different from the typical type (Dukas & Waser, 1994). However in tests with artificial flowers where one colour was rewarding and the other colour was non-rewarding, bumblebees showed a preference towards blue flowers whether they were rewarding or not (Dukas, 1995). They may also be able to recognise the general shape or outline of the whole plant (Darwin, 1891; Free & Butler, 1959; Faegri & van der Pijl, 1979), as they have been seen attempting to forage on flowers that are not yet open, while others of the same species are open.
No communication system for recruiting workers to rewarding flowers has been discovered in bumblebees (Brian, 1957; Free & Butler, 1959; Heinrich, 1979; Laverty, 1980), there appears to be no bumblebee equivalent of the honeybee waggle dance; however one bumblebee foraging in a group of flowers appears to attract others (Brian, 1957). So flowers chosen for foraging may initially be chosen at random by individual bees and then be retained, providing they are suitably rewarding and the nectar is within easy reach. The choice of the major flower might influence the choice of the minor flower(s), as the minor flower may be of the same colour and/or morphology as the major. Also the nest odour may influence their choice. If there is one species of flower that is visited by many of the bees in the nest the odour will spread throughout the nest (Faegri & van der Pijl, 1979), and so a bumblebee on its initial foraging flight may be influenced to choose a flower having the same smell, and this in turn might further influence its nest mates. This may explain why two nests of the same species, in the same area, might show great differences in the flowers that were foraged on (Free, 1970).
To help to understand the ways that bumblebees are recruited to flowers of plants, so as to maximise the benefits to these plants, we need to have a fuller understanding of the factors influencing foraging. Despite these previous studies, however, we still do not understand the relationship between the foraging patterns of individual bumblebees in relation to those of their nest mates and so to the general foraging of the "nest". This is the topic of this study.
The principle aim of this project was to discover if species and individual bumblebees have similar foraging preferences by recording the flowers that they were observed using during the course of a bee walk. To acheive this aim it was necessary to devise a way of reliably marking bees without anaesthetising them. If preferences were found it was intended to investigate whether these could be related to tongue and head measurements. To this end the functional proboscis length, and the head length and width for the species encountered in the bee walk were measured. Subsequently to record flowers visited sequentially by marked and unmarked bees during a single foraging trip, and to compare the sequence of visits of a single foraging trip with individual and species preferences over a five day period.