From bees in space to living sensors: how we read colony stress

The first space hive: NASA's 1984 bee experiment

In 1984, NASA sent a small honey bee colony into space. The Bee Enclosure Module (B.E.M.), the first "space hive" of its kind, made it possible to test whether social insects could function in microgravity ("weightlessness"). The aim of the study was to understand how honey bees survive and behave in microgravity, and whether they can build honeycomb in such an environment.

During mission STS-41-C, the space shuttle Challenger carried the B.E.M., a small transparent hive with three frames and one wall made of acrylic glass, so that the astronauts could watch the colony. The colony consisted of about 3,500 workers and one queen. A second, identical module with an equivalent colony stayed on Earth as the control group. It was also exposed to simulated shuttle launch conditions in a centrifuge, so that scientists could compare bees that had spent seven days in orbit with bees that stayed in normal gravity on Earth [1].

What happened to honey bees in microgravity

On Earth, bees receive one constant piece of information throughout their lives: which way gravity points. They use it to orient themselves and to build combs in their characteristic hexagonal arrangement. So what happened when that information disappeared?

The "space bees" became disoriented when they tried to fly and collided with the walls and with each other. Even so, after about four days some of them worked out how to move around and stopped hitting obstacles. The colony also kept functioning. The workers built comb comparable to that of the colony on Earth, with one difference in the inclination of the cells. This points clearly to the role of gravity as a reference for comb construction. The "space queen" also laid about 35 eggs in orbit. Unfortunately, none of them survived after the return to Earth, and the reason was never established.

Perhaps the most interesting observation was made on Earth. About 120 dead bees were found in the "space" group, while the Earth group had nearly three times as many, about 350. The researchers pointed to differences in the age structure of the two colonies, so this cannot simply be read as an effect of the flight. It is still a good example of how hard it is to interpret data from living systems, and why we need good, continuous measurement instead of occasional snapshots.

Despite its age, the experiment remains interesting for two reasons. It is still one of the few direct studies of a honey bee colony in space. It also raises a question that has no complete answer yet: can a colony not only survive but also reproduce in conditions vastly different from the ones evolution prepared it for?

The bee colony as a superorganism

A single bee is unremarkable. A colony, on the other hand, is a system of tens of thousands of individuals that respond to change as a whole. A colony has no central "brain". Decisions such as choosing a nest site [2], changing the pace of foraging or reacting to a threat [3,4] come from interactions and communication between workers. For that reason a bee colony is often described as a superorganism, an intelligent biological system that monitors its own state and adjusts its behaviour to external conditions.

Because of this, the colony constantly "talks" on several channels at once. Its sound changes with activity, temperature, swarming preparations or the loss of the queen. Beekeepers often recognise the distinct sound of a colony that has lost its queen, or of a queen communicating with the workers. Through comb vibrations during the well-described waggle dance, scouts pass on the direction and distance of a food source. With chemical signals, such as queen and alarm pheromones or the scent of brood and food, bees coordinate the life of the colony.

Bees as living sensors

For a researcher, all these signals are a great opportunity. When the variables inside the hive are measured continuously and non-invasively, the colony becomes a source of data, both about itself and about the environment it lives in. It tells us far more than occasional inspections can. In this sense, bees work as living sensors of their environment.

At Apisense, we started from the assumption that understanding a colony's response to stress does not require sending it into space. It is enough to listen carefully to the biological signals the colony sends and to interpret them correctly. How does the temperature change through the season? Do bees behave differently during heatwaves? What happens to VOC levels when a hornet appears in the apiary? We collect such data from around 3,000 hives worldwide, and dedicated models are trained to recognise various problems through pattern recognition. The system works, but its accuracy depends strongly on how much clearly described data we have. Simply put, the more data, the better the system.

Hypergravity: queen bees in a centrifuge

Another space-related experiment was carried out in 2019 on the centrifuge at the Military Institute of Aviation Medicine in Warsaw. In a study one of us co-authored [5], queen bees and their attendants were exposed to the acceleration profile of a launching rocket. After the experiment, all queens from the test and control groups were moved to an experimental apiary for further observation.

The results are cautious but intriguing. After the acceleration, egg laying by the test queens diverged in two directions: some laid more eggs and others fewer, while the control queens kept a similar pace. In the test colonies, drone brood appeared earlier and in larger amounts, which may point to problems with the stored sperm. Overwintering of the colonies was not disturbed.

Acceleration is a clearly defined stress stimulus, and the response to it can be followed in the colony's signals. Measuring in-hive signals during such an experiment could show which variables are most affected when honey bees are exposed to extreme conditions. Knowing that, we could interpret data from other colonies better. As in 1984, the question is not only how many bees survived, but how the colony as a system responded and whether it regained its balance.

Biological Intelligence: reading colony stress before symptoms appear

Biological Intelligence is an approach in which the signals produced by a living system are treated as data about its state, instead of relying only on sensors that measure the surroundings. The sensors are only the ear. The point is to learn to read what the colony itself communicates: changes in sound, vibration and scent that come before visible symptoms of a problem.

Why is the response to stress so valuable? A healthy colony has its own "response pattern", a characteristic way of reacting to a disturbance and returning to balance. When a colony is weakened, sick or short of stores, this pattern can change: the response may be stronger, longer or more chaotic. By comparing responses to controlled stress, such as acceleration in a centrifuge, with what happens in apiaries under cold, heat, pesticides or disease, algorithms can be trained to recognise early signs of danger.

Why colony health matters for agriculture

For AgTech this has concrete significance. Bees pollinate close to three quarters of the most important food crops, at least in part [6], so the health of bee colonies is also the health of the food supply chain. Monitoring based on biological signals can shorten the time between the appearance of a problem and the beekeeper's reaction, and reduce the number of unnecessary inspections, which themselves stress the colony.

From orbit back to the hive: our next steps

Four decades have passed since the Challenger flight, but the question remains the same: how does a living system cope when the conditions it adapted to disappear? The difference is that today we can answer it not only by counting dead bees, but by continuously following the signals through which a colony communicates its state, adapts and makes decisions.

This is the aim of our research at Apisense: to understand how living systems communicate, how they react and what changes under stress. Space is a frontier laboratory, but the real field of application remains the hive. Our next steps follow from that:

• Grow and describe our dataset. Every well-documented case of heat, cold, disease or hornet pressure from the 3,000 hives we monitor helps our models tell a normal stress response from a dangerous one.
• Measure the colony during controlled stress, not only after it. Recording in-hive signals during experiments such as the centrifuge would show which variables react first and how quickly the colony returns to balance.
• Compare controlled stress with real apiaries. Matching the responses from experiments with what we see in the field will give us reference patterns of healthy and weakened colonies, the basis for earlier warnings for beekeepers.

And who knows, maybe one day our centrifuge experiments will help in a real extraterrestrial apiary, pollinating crops in a Martian greenhouse.

References

1. Vandenberg, J.D.; Massie, D.R.; Shimanuki, H.; Peterson, J.R.; Poskevich, D.M. Survival, Behavior and Comb Construction by Honey Bees, Apis Mellifera, in Zero Gravity Aboard NASA Shuttle Mission STS-13. Apidologie 1985, 16, 369-384.
2. Perdriau, B.S.; Myerscough, M.R. Making Good Choices with Variable Information: A Stochastic Model for Nest-Site Selection by Honeybees. Biol. Lett. 2007, 3, 140-143, doi:10.1098/rsbl.2006.0599.
3. Russell, S.; Barron, A.B.; Harris, D. Dynamic Modelling of Honey Bee (Apis Mellifera) Colony Growth and Failure. Ecol. Modell. 2013, 265, 158-169, doi:10.1016/j.ecolmodel.2013.06.005.
4. Khoury, D.S.; Barron, A.B.; Myerscough, M.R. Modelling Food and Population Dynamics in Honey Bee Colonies. PLoS One 2013, 8, 1-7.
5. Stasiowska, D.; Kolasa, M. Hypergravity Impact on Fertility of Apis Mellifera Carnica Queens - Case Study. Microgravity Sci. Technol. 2024, 36, doi:10.1007/s12217-024-10111-1.
6. FAO. Why Bees Matter. The Importance of Bees and Other Pollinators for Food and Agriculture. 2018.

‍