by Dave Goulson
This was not a perfect experiment; I am sure you will have realised that many foraging bees might have happened to be in the nest at the time we shut the doors. Nonetheless it showed very clearly that the very smallest workers were all in the nest, suggesting that they rarely if ever go foraging – in fact, unless you look inside a bumblebee nest you will never see these diminutive creatures. Medium-sized workers seemed to include a mix of nest bees and foragers, while the very largest workers were almost all foragers, just as had been suspected.
This is all very well, but why are foragers bigger than nest bees? It is easy to understand why ants which are to defend the nest must be large, but less easy to see why carrying food necessitates being large. You may well be thinking, ‘Surely a large bee can carry more?’ It would be surprising if they could not, but it isn’t as simple as that. Big workers presumably require more food to rear in the first place. Let’s suppose, for the sake of argument, that it takes the same amount of food to rear one big worker or two small ones of half the weight. Further, let’s suppose that the big worker could carry twice as much food. On each foraging trip, the two small workers would bring back the same weight of food between them as the larger bee. However, they would be back sooner as they would only have to visit half as many flowers to gain a full load – all else being equal – so they would be able to make more trips per day, and should be able to bring back more food in total. It would only be worthwhile to produce large foragers if they were somehow a lot better at gathering food than small ones, or if they had some other significant advantage of which we were unaware.
James set out to discover how efficient bees of different size were at gathering food. His experimental set-up was simple. We rigged up a buff-tailed bumblebee nest in my lab with a clear plastic tube attached to the nest door. When the door was opened, bees could walk down the tube, through a clear plastic chamber constructed over the pan of an electric balance, along a further tube, and out through the window to the outside world. We could individually identify every bee since we had carefully glued a numbered plastic disc to the thorax of each. This is an enormously fiddly job. The glue takes a minute or two to set, during which time the bee tries to dislodge the disc by hooking a foreleg over its back. As often as not it succeeds, and the process has to be repeated. Not infrequently the bee pushes the sticky disc forwards on to its head, effectively covering its eyes, and then blunders around bumping into things. I once tried using quick-drying superglue instead, but after ending up with a bee with its foot stuck to its back I quickly gave up.
As each bee left the nest over the pan of the balance, its number and weight were manually recorded. Initially, bees tended to fly across the chamber rather than walking over the pan, but this problem was solved by covering the chamber with transparent red plastic film; bees can’t see red light very well so the chamber appeared dark, and bees won’t fly in the dark, so they were forced to walk. Once they left the tube protruding from the lab window, the bees invariably hovered around for a minute or two, presumably trying to memorise the location, before disappearing off into the suburban gardens of Southampton. It was very exciting watching the first few bees depart, and we sat about eagerly awaiting their return.
After about half an hour, the first bee appeared back at the window, with large balls of pollen on its legs. Unfortunately it didn’t seem able to find the end of the plastic tube, so it flew backwards and forwards along the long row of glass windows. It was soon joined by others, and quickly a cloud of confused bees built up outside the window. It became apparent that there was a flaw in our plan. The lab was on the third floor of a very sizeable and ugly concrete building, with seven floors in all. Viewed from the outside, there were seven rows of identical windows, one above the other, and each row was perhaps 50 metres long. All looked much the same, and the little plastic tube protruding from the corner of one of them was exceedingly inconspicuous. Our poor bees knew their nest was there somewhere, but couldn’t find the correct window. Before long the frustrated bees started coming in any open window, often a floor or two above or below my own, invading offices, biochemistry labs and so on. Since the bees all had numbered discs attached it didn’t take anyone long to realise who was to blame, and we soon found ourselves bombarded with phone calls and having to run around the building with nets and pots to catch our strays. We shut the nest door so that no more bees could go out, put back as many as we could recover, and sat down to discuss what to do. Eventually we decided on attaching a large orange funnel to the nest tube, providing an eye-catching target for returning bees. To make life even easier for them, Ben Darvill, then an undergraduate student, somehow acquired a traffic cone and gaffer-taped it to the outside of the building just below the correct window. How he got it up there I never found out, since it was too large for him to have put it out through the window. When I left Southampton five years later it was still there.
The funnel and traffic cone served their purpose and bees began foraging from the nest, running out through the plastic tubes and returning laden with pollen or nectar. Once this was all working, James and a team of undergraduate helpers set about recording and weighing every bee that left or came back into the nest. Someone had to be by the nest at all times during daylight hours, so they set up a rota between them. The nest lasted for about one month, after which we replaced it with a second, and eventually a third, so the team, between them, spent three months sitting at the lab window with notebook in hand. The difference between the outgoing and incoming weights of each bee allowed us to calculate how much food they had gathered on their trip, and since we also knew how long this had taken them, we were able to calculate their efficiency in terms of forage gathered per minute. Some bees were notably more dynamic than others, sprinting down the tubes, returning within a few minutes with full loads, then immediately setting off again. Others ambled along the tube, stopping occasionally, sometimes getting to the exit funnel and then turning back as if they’d changed their mind and decided they didn’t really fancy going foraging after all. The keen bees were enormously impressive. Bumblebees carry pollen on their legs and nectar in their honey stomach, a chamber which, when full, almost fills their abdomen. Some of them brought back up to 150 milligrams of food, close to their own body weight, having gathered it in twenty minutes or less, and they did so many times during the day.
Bees got better at foraging with experience. On their first trip from the nest, many came back weighing less than when they set out, presumably because they had flown about burning energy and failed to find any food. Over time their success rate increased, and on average it took about thirty or so trips before they reached peak efficiency. Presumably during this time they were learning about how to navigate through the landscape and experimenting with different flowers to see which were most rewarding. On their first foraging trips bees tended to collect nectar, usually returning with no signs of pollen on their legs, but as they got older and more experienced they turned to collecting more pollen. I suspect that this is because drinking nectar is much easier than gathering pollen, for the latter requires the bee to brush pollen from the anthers with her hairy legs, then comb the grains from her body into her pollen baskets, using a little nectar to glue them together. An experienced bee makes this look easy, but it must take quite a bit of practice.
A few largish bees never went foraging but sat for most of their time just inside the door to the colony, or just outside in the tube. We dubbed these ‘guard bees’, since honeybee colonies have individuals that sit in the nest entrance and try to prevent intruders, but in truth we don’t really know if these bees were actually doing any guarding. They never did anything when foragers came past them – often literally walking over them – perhaps because all the bees coming in were legitimate members of the nest. If a cuckoo bee had attempted to get in, or a worker from another nest hoping to lay a few eggs, they might have sprung into action.
It turned out the bigger foragers were, on average, more efficient at
gathering food, bringing back more per unit time. The smallest of the bees that left the nest rarely returned with a net profit and so contributed very little to colony growth. This presumably explains why foragers tend to be large, but in doing so it raises another question: why are bigger bees better at foraging? I discussed this with James, and our first idea was that it might relate to keeping warm. To fly, a bee needs to keep her core temperature above about 30°C, and bigger bees ought to be better able to do this since they have a smaller surface area to volume ratio, but they might be prone to overheating on hot days. Bees working in the nest don’t need to worry about keeping warm since they don’t have to fly and in any case the nest itself is kept very cosy by insulation and the combined activities of the bees inside. If a bee becomes too cold when foraging she will be grounded, unable to return home, and at great danger of being eaten by a predator, so this might explain why small bees avoid foraging. James tried testing whether the smaller of the bees that went foraging were affected more badly by cold weather, but this didn’t seem to be the case. He also went through the data set carefully to see whether smaller foragers tended to go out only on warmer days, but this didn’t seem to be true either.
James and I didn’t really get to the bottom of what made bigger bees better at foraging, but part of the answer has since been provided by a German researcher named Johannes Spaethe, working in the lab of Lars Chittka at Queen Mary University, London. He found that larger workers have much more acute vision, in part simply because their eyes are much bigger. This is likely to be enormously important when searching for particular types of flowers in the landscape, or when memorising and distinguishing between landmarks for navigation, but of no value whatsoever to a bee operating in the dark confines of the nest. He also found that having larger eyes enables bigger bees to fly in darker conditions, presumably helping them to get home if they find that it is starting to get dark when they are out collecting food. Spaethe went on to study the sensitivity of antennae to odours, and discovered that larger workers also have a better sense of smell. This would benefit them greatly when trying to locate or identify flowers by smell. Larger bees also tend to have bigger brains and so might have enhanced learning ability; this too would help in the complex job of traversing the landscape in search of food, although it has not yet been studied. Finally, larger bees might just be more likely to survive when out and about in a dangerous world. I have commonly seen bumblebees tear themselves from spiderwebs, and I imagine that this would be much harder for a small bee. Despite their large size, the life expectancy of foragers – at two to three weeks, roughly the same as a First World War pilot – is much lower than that of nest bees, which often survive for months. This difference might be much worse if the foragers weren’t larger.
Interestingly, honeybees have a different strategy in this respect. Rather than divide up jobs according to size (workers are all the same size), honeybees divide up jobs by age. The younger bees stay in the nest, looking after the brood, while the older bees do the foraging. It seems likely that older bees are given the most hazardous job because they have less residual value to the colony; to take an extreme example, a very old bee is likely to die any day anyway so why not send her on a perilous mission, whereas a young bee has many weeks of work left in her, and it would be a greater loss to the colony if she were to die. There is a possible parallel with humans, although you might find this far-fetched. In the Hadza, a hunter-gatherer tribe of Tanzania, older, post-menstrual women gather a disproportionate amount of food, spending longer out in the savannah picking berries and digging for tubers than their daughters and granddaughters. In evolutionary terms, it may be a wise strategy for post-reproductive women to spare their children such arduous and potentially dangerous tasks and take the risks themselves, since they are unlikely to live too much longer and they are not able to reproduce again themselves. In evolutionary terms, granny’s genes would rather she were eaten by a lion than that her grandchildren were eaten. I should add that I am not suggesting this is a conscious strategy, just as I do not suggest that worker bees consciously choose to help rear their sisters because they are aware of patterns of relatedness.24
Let us return to size variation in bumblebees. The question as to why foraging bumblebees tend to be large can be turned on its head by asking, why are nest bees small? It seems likely that the small bees are more nimble in the confined space of the nest, so that it benefits the colony to have small workers for within-nest tasks and big ones for food-gathering. If this were so, nests with workers of a range of sizes ought to fare better than nests with only large or only small workers. Such nests don’t naturally exist, but James and I set out to create them by moving bees around. In many social insects, it is not possible to move workers between nests; for example, if one places a worker ant in a different nest she is usually swiftly attacked and killed. Social insects can usually detect the odour of non-nest mates, and are not welcoming to visitors. However, I had noticed that bumblebees from commercial nest boxes often enter the wrong nest when several are placed near each other, and they seem to be accepted. I presume that this is because they have been reared in identical conditions, and so perhaps they all smell much the same. Thus I thought that it might be possible to juggle workers between nests to create all-big or all-small colonies.
I experimented with a giant home-made pooter. Pooters were invented as a device for picking up small insects without harming them; such creatures are all too easy to squash with forceps or fingers. Ordinarily, a pooter consists of a small glass bottle with a cork in the top. The cork has two holes in it, through which two long, flexible, transparent tubes are pushed. Importantly, the end of one of these tubes, where it protrudes below the cork inside the bottle, is covered with fine mesh or netting. To use a pooter, one tube is placed in the mouth – this must be the one with the mesh attached – and the other tube is pointed at the insect. A swift suck on the former usually results in the insect shooting up the latter in a rush of air and ending up, unscathed, in the bottle. It is best to avoid doing this with ants as they squirt out formic acid when frightened which is not kind to the lungs and results in a lot of coughing, but otherwise this is an excellent device for handling tiny beasts. It had occurred to me that a super-sized pooter based on a large jam jar and tubes large enough to suck up bumblebees might be just the thing for efficiently moving bumblebees between nests. I constructed one, and tried it out.25
I had a nest in our darkroom which was equipped with red lighting so that the bees would not fly when I removed the lid. Several workers were scurrying around nervously on top of the nest, aware of the disturbance. When handling bumblebee nests it is important not to breathe on them as this causes the bees to become very agitated, presumably because they think they are under attack from a large mammal such as a badger. I carefully pointed one tube at a bee, and sucked hard on the other. The bee clung on, confused by the sudden draught. I gave the mightiest suck I could manage, and was starting to go blue in the face when the bee suddenly rocketed up the tube and into the jar. Pleased with my success, I paused to get my breath back, then tried a second bee. Unfortunately I had forgotten the all-important netting on the tube going to my mouth. As I sucked, the bee that was already in the jar shot up the tube and into my mouth. Before I could spit it out it stung me, causing my lower lip to swiftly swell to a ridiculous size. As you might imagine, James and my other students made fun of me mercilessly for the rest of the day.
My giant pooter never really caught on, but nonetheless James did manage to juggle bees between nests to create four different experimental treatments: nests with thirty large workers, nests with thirty small workers, nests with about sixty small workers of the equivalent total weight as thirty large workers, and nests with thirty workers of assorted sizes. These he placed out on the university campus to see how they fared. We predicted that the nests with only thirty small bees would be poor at foraging and so grow very little. The nest with thirty large bees ought to be good a
t collecting food but poor at looking after their young. Comparing nests with sixty small bees versus thirty bees of twice the size ought to reveal whether two small bees are better than one big one. Finally, we predicted that the nests with a mix of large and small bees, as occur naturally, would perform best of all.
When the bees had had three weeks of looking after themselves, James gathered in the nests and counted how many young they had reared. Our predictions were wrong. The nests with only big workers reared the most offspring, performing better than nests with a natural mix or with the equivalent weight of small worker bees. Taken at face value, this suggests that bumblebees have got it wrong. They shouldn’t bother rearing any small workers, but should just concentrate on producing big ones, even if this means having fewer of them. This may be true – no organism is perfectly adapted to its environment – but I suspect that the results might be misleading. The weather during James’s experiment was awful, with near-constant rain, and none of the nests were thriving. Perhaps larger bees are better at coping with such adverse conditions. It may serve the queen’s interests to have mainly small workers in the nest, for they would be easier for her to dominate and prevent from laying their own, male, eggs. Perhaps mixing bees from different nests is hopelessly unnatural and causes strife within the nest, disrupting its normal functioning. We still don’t fully understand the roles that bees of different sizes play in bumblebee nests.
