Digital tools for beekeeping and biomonitoring

Table of contents

Share this article!

Bzz, Bzz! The bee is THE great national cause of the year 2022 in France ! Yes, it is, I swear. The bee was in the spotlight but you certainly didn’t know about it… The latest IPBES reports on biodiversity (please tell me you’ve at least heard of this acronym) are, without much surprise, ever more terrifying, showing a massive decrease in the abundance of living organisms and terrestrial and marine ecosystems. Invertebrates, largely less followed than vertebrates, are the big losers of the case – few people seem to really pay attention to them. Among them, however, are some bee species that, because of their interest in beekeeping and their social sex appeal, are given special treatment.

The deployment of digital tools in beekeeping is not new, but it is clear that apart from connected scales and temperature sensors, there is not much that is operational in the field. You might be surprised by the release of this blog file, as it seems to be a terse statement. Nevertheless, you will see that by digging a little into the subject, the digital devices are finally quite varied and offer the promise not only to accompany the monitoring of apiaries and colonies for beekeeping production but also to provide knowledge for environmental monitoring and biomonitoring on a large scale. Promise or illusion? As usual, the answer is complex and far from being binary. I will try to provide you with a variety of reading keys.

This report on beekeeping is also an opportunity to highlight all the knowledge that is beginning to be capitalized on the directory of digital tools for agriculture. At the time of writing, we have exceeded 1500 digital tools referenced! In addition to serving as a collaborative watch, this platform is now used to take a step back on the digital tools in place and to identify trends.

As usual, for readers of the blog, this article is based on video interviews with industry players (whose names you will find at the end of the article) whom I thank for the time they were able to give me. Several articles, reports and seminars have allowed me to complete the feedback from the interviews.

Enjoy reading!

Translation : Bees officially recognized as an endangered species. “We have been wearing our yellow vests for over 80 million years”. “But nobody cares”

Soutenez Agriculture et numérique – Blog Aspexit sur Tipeee

Find here the main infographics of the file:

Infographic 1. Digital tools in beekeeping. Source: Author. Adapted from Marchal et al. (2020).

Infographic 2. Parameters monitored in beekeeping with digital tools. Source: Author. Inspired by Marchal et al (2020). Caution: The digital tools of a category do not measure all the parameters of this category. Some parameters are only measured in research activities

Important preamble

My previous files have mainly focused, in one way or another, on the plant sector. Here, I’m going to try my hand at an animal sector, a sector I knew very little about before I started working on it. While writing this file, I was naturally struck by the imposter syndrome, this classic discomfort that pushes you to continuously question your legitimacy to even dare to give your opinion on a subject that you are far from mastering completely despite the amount of hours spent on it. So I position myself here with humility and I hope that my naive entry on this sector and my experience on the digital ecosystem in agriculture in general will offer you some relevant reading elements. Am I entitled to ask for a little indulgence for this first attempt despite the few spikes I couldn’t help but throw =) ? It’s hard to separate myself from my cashier style of writing. Haters gonna hate…

I will add that this work on beekeeping has allowed me to tackle subjects oriented towards bio-monitoring and the environment – themes that my previous popularization files tended to exploit relatively little in favor of climate and energy. The links between biodiversity and climate are however very close and our propensity to focus on climate seems to make us forget all or part of what is happening to biodiversity. This is all the more serious as almost all actions to save life are also beneficial for the climate – the reciprocal being much less true.

Figure 0: Effects (positive and negative) of actions to mitigate climate change on actions to mitigate biodiversity loss (top), and of actions to mitigate biodiversity loss on actions to mitigate climate change (bottom) (IPBES-IPCC, 2021). The graph seems to be a bit hard to read but basically, if we start on actions in line with the climate issue (bottom left), it leads to positive (blue) and negative (red) for biodiversity (top left). But if we start to focus on actions to save the living (bottom right), we have much less negative actions on the climate (red).

Some orders of magnitude of beekeeping

State of the beekeeping market

Have you ever heard of apidology? I haven’t, and yet it is the name given to the branch of study that concerns bees! We will all sleep less stupid tonight … From chestnut, alfalfa, acacia, or even sunflower, honey in all its forms – liquid or solid – will not finish enchanting our taste buds (Figure 1). Even if a true honey will never be really composed at 100% of the same flower – some beekeepers will rather sell their honey under a more vague name “spring honey”, “all flowers” – it is also the mixture of the gathered plants which will give its own identity to the honey. Of course, we can always do pollen research in honeys to study their purity (and beekeepers will have to do this to certify a species-specific honey), and some visual characteristics of honey (whiteness of wax for a lavender honey, very liquid aspect of acacia honey…) will allow to lift some veils, it remains nevertheless rather difficult to follow completely the trajectory of our foragers and to verify that they will be interested only in a particular type of flowers (difficult to make sure that a honey is really completely organic too). Has a little bramble and clover in a chestnut honey ever really hurt anyone?

In France, the France Agrimer census of 2020 shows a little more than 70.000 declared beekeepers (I insist here on the fact that they are the ones who are declared) for a little more than 1.7 million hives. The beekeeping sector is totally different from the other breeding sectors for one simple reason: more than 90% of beekeepers have less than 50 hives (France Agrimer, 2022b). These beekeepers are thus considered as amateurs in the sense that it is hardly possible to make a living from this only artisanal production with “so few” hives. Beware, however, of misunderstandings. If the majority of the beekeepers in number are indeed amateurs, they are not the ones who produce the majority of the French honey but the so-called professional or semi-professional beekeepers (around 5000 professionals for a little more than 75% of the total volume), that is to say those who, at the big ladle, have at least 150 to 200 hives. The turnover of the beekeeping sector is around 100 million euros (France Agrimer, 2022b), far behind the other animal sectors. The production of honey is mainly conditioned in pot, especially by the amateurs, that the French will buy in a priority way in direct sale. As for the professionals, they also turn to a conservation in barrels or honey tanks to avoid having thousands of pots to fill themselves.

Still in France, the beekeeping production would have visibly decreased between the 90s and the much more recent years. The year 2021, particularly unfavorable in terms of meteorological conditions, will have led to a production of nearly 20 T of consumption honey, well below the previous annual averages (globally around 30-40 T in the 90s). Whether they are amateurs, professionals, sedentary or itinerant, the main challenge for beekeepers remains the same: to keep bee colonies alive in order to produce the coveted honey. The economic pressure will obviously not be the same for a leisure beekeeper, potentially more interested in one of his favorite hobbies and in the sublime side of the functioning of a beehive, as for a professional beekeeper who will need to break the bread… The bee colonies will remain for the latter his work tool. Let’s add to this that the competition is tough, with a lot of imported honey and counterfeits more and more original: honey mixed with water, random imports, honey mixed with imported honey, honey cut with glucose syrup. The labelling of EU (European Union) and non-EU honeys remains an important subject of contention.

Figure 1: French national honey production in 2021. Source: France Agrimer (2022b).

Honey, but not only…

In France, the vast majority of the turnover of beekeepers is linked to honey production (France Agrimer, 2022b). The rest is divided between :

The breeding and sale of queens, which are then reintroduced into the apiaries
The breeding and sale of swarms of bees: even if the quantity of bees seems to remain relatively stable, it hides in reality a rather frightening turnover, we will have the occasion to speak about it again
The production of royal jelly as a food supplement for humans, which requires specialized beekeepers to raise hives without a queen (the queen feeding exclusively on this substance produced by the nurse bees)
The production of beeswax to make candles
The production of pollen
Production of propolis used mainly in pharmacopoeia
Pollination services when a beekeeper provides hives on a rental basis to increase the presence of pollinators on agricultural plots. A little reminder of your biology lessons: Pollination is the transport of pollen from the anthers (the male part of the flower that produces pollen) to the female part, the stigma. This pollen can either come from the same flower or from another flower of the same plant or from another plant (this is called cross-pollination)

That’s a lot of human-centered services, you may say… I would like to insist here also on the pollination services which, in France, are not really monetized compared to other countries of the world. One thinks, for example, of the gigantic almond fields in the United States where some beekeepers arrive with hundreds of hives to support pollination. For some critics, the bees are thus sent to the pillories, as the almond fields are quite extensively treated with phytosanitary products.

Pollination services – whether monetized or not – pose major challenges in terms of food security and governance. No less than 75% of the world’s major food crops and nearly 90% of wild flowering plants depend, at least in part, on animal pollination, with wild pollinators also playing an absolutely fundamental role (Figure 2). In particular, insect pollination is necessary for the fertilization of a majority of flowering plant species that we cultivate for their seed (rapeseed, sunflower, buckwheat), fruit (apple, pear, kiwi, melon), root or bulb (carrot, radish, onion), or foliage (cabbage, salad). And the volume of production of pollinator-dependent crops would only increase, making us ever more dependent on these natural pollination services. These figures are obviously to be put into perspective depending on the type of crops and the regional agricultural economies, but they allow us to put our finger on the main orders of magnitude.

Some studies tend to emphasize that the crops most dependent on insect pollination are also those with the greatest economic value. While this type of study should alert us, it is important to take a step back and examine the assumptions of these studies. For example, the studies do not necessarily take into account the impact of a decline in pollinators on seed production, which is very important for many forage, vegetable and horticultural crops, nor the effects on wild flora. Others of these studies imagine a total disappearance, not a gradual decline, of pollinator populations, and do not necessarily incorporate the policy responses that we (beekeepers and civil society) would make to deal with this disappearance.

Figure 2. The dependence of global crops on pollination. Source: IPBES (2018).

Moving beyond a purely food prism, pollinators – all of them – are also directly used in the production of medicines, biofuels (such as rapeseed and palm oil), fibers (cotton and flax), building materials (timber), musical instruments, or even arts and crafts. There are also a whole bunch of economic and social arguments that come into the equation since millions of jobs depend on these pollination services.

All this being said, I let you go and have a look at the Fresque de la Biodiversité – a collaborative gameplay to discover the stakes of biodiversity – just to give you a good dose of humility and questioning.

The life of bees

Pollinators visit flowers primarily to collect or feed on nectar (the source of sugar) and/or pollen (the source of proteins, lipids, vitamins and minerals), although some specialist pollinators may also collect other substances such as oils, fragrances and resins produced by some flowers (e.g. honeydew is a syrupy liquid made by aphids sucking the sap from trees). In temperate regions, bees hibernate during the winter, begin production in the spring and continue through the summer, stopping production and returning to a state of hibernation in the fall. In warm, tropical regions, pollen and nectar sources may be available year-round, so bees do not hibernate.

Able to search up to about 3km from the hive during the day, our branchy-haired mellipher bees prioritize one plant species on a trip, which greatly improves pollen transport efficiency.

When discovering new food locations, bees have the fascinating ability to communicate to their fellow hives the distance and direction of their booty with abdominal oscillations and vibrations, both in a round or figure-of-eight “dance” system in the hive and in flight navigation. However, this communication is not very precise and the bees use their sense of smell once they are in place to help them navigate. The bees would have an extremely developed sense of smell and sensitive to a whole bunch of chemical signals. Bees are used either in the laboratory on test tables or in the field as free-flying traceable biosensors (for example for tracking metal or chemical pollutants over large areas and/or in landmines, or for tracking dead bodies or chemical signals produced by diseased plants and animals). Some would even use bees as Covid detectors, progress is unstoppable… Others would use bees for biocontrol, using honeybees and bumblebees to transport or vectorize microbial biocontrol products.

Sensitive souls, please refrain! Foraging bees store nectar in their crop, at the end of the esophagus, and regurgitate it into the crop of the nurse bees (this is called trophallaxis) once they return to the hive, which will then regurgitate it into the cells dedicated to storing it (nectar is hardly ever consumed by the foragers) The nectar, passing from mouth to mouth, so to speak, is transformed into honey thanks to the enzymes present in the bees’ crop. The air currents in the hive help to dehydrate the honey and the last bees close the honeycomb with wax cappings once the honey is at the right hygrometry degree.

The bee has no concept of honey storage. The colony will continue to store as long as there is room in the hive. This is why beekeepers install honey supers (frames) to collect the surplus honey that the bees will not need to last the winter (provided that the beekeepers leave them enough basic food and that they do not steal all of their collection…) [Figure 3]. The pollen is transported to the hive via small bags on the hind legs of the bees – a forager bee being able to carry half its weight in pollen.

Figure 3: Diagram of a Dadant hive.

In the apiary, the cells are separated into two main categories, which are not of the same color: those which receive the nectar (and thus the future honey – there are also pollen cells) – we have already spoken about this above – and those which will receive the future bees (females) and drones (males) of the colony (we speak here of brood cells). The queen bee, fed from birth with royal jelly, is carefully selected from the colony’s larvae by the workers. After her nuptial flight, where she will be fertilized by the drones of the colony (which will not serve much else), the queen returns to the hive to lay thousands of eggs per day and prevents the workers from laying eggs on their own from the secretions she emits (the spermatozoa remain fertile for several years in the queen). As mentioned above, keep in mind that some beekeepers breed selected queens for introduction into the colonies. The larvae, fed on honey and pollen, become future worker bees or drones 21 days later. The role of the bees in the hive, as worker, feeder or forager, is defined over time.

To raise the brood, the bees regulate the temperature of the hive in an extremely fine way (I will come back to this when I talk about digital tools). The environment close to the queen is maintained at a temperature of nearly 30° when the bees in the periphery can be at more than a dozen degrees difference. The hair of the bees is comparable to goose down. They certainly have a role to play in collecting food, but they also play a major role in keeping the temperature at an acceptable level when it is cold. The honey bee has the ability to overwinter in its nest thanks to the survival of part of the adult workers (the cluster) which will maintain positive temperatures by producing endothermic heat, and thus allow the survival of the queen.

Depending on the health conditions of the colony and/or the age of the queen, the bee colony is likely to swarm, i.e. to leave the current hive to find a new place to live. This swarming is certainly one of the greatest fears for beekeepers who will then see their apiary emptied. If the queen is too old, the workers prevent her from laying eggs and push her to leave with her faithful courtiers. The bees then choose a new place to live (and may even die if they can’t agree on the location), aerate the area and secrete odors so that their fellow bees can find each other. The bees are even able to measure the size and depth of the place they are considering by forming a long chain hand in hand (or paw in paw – I don’t know if it’s really said).

Some orders of magnitude to keep in mind:

Bees can fly up to 3 km away from the hive (sometimes even a little more)
The hives contain between 30 and 40,000 bees. In wild colonies, the number can be much higher
The production of 1L of honey will require the bees to collect nectar from about 10 million flowers which, for the foragers, represents a total of 10.000 hours of flight and 60.000 km of distance.
The production of 1kg of wax will require the consumption of 7kg of honey by the bees
A hive can contain several thousand bees and the foragers go out on average 4 times a day to pollinate about fifty different flowers.
Only a small part of the totality of the flowers is pollinated

The bee is far from being the only pollinator!

Beekeeping is still concentrated on a few specific species of bees, notably the well known Apis Mellifera (the western honey bee), or the Apis Cerana (the eastern honey bee). Some bumblebees, some stingless bees and some solitary bees are found to a much lesser extent

The western honey bee – Apis Mellifera – represents in reality a very small part of the pollinator species; it is necessary to get this into our heads! The vast majority of pollinator species are wild (Figure 4). There are more than 20,000 species of bees in the world (and around 2,000 in Europe), some species of flies (syrphid flies for example), butterflies and moths, wasps, beetles, thrips, birds, bats, lizards, and so on. We could play the game of the 7 families (or species) for insects: Diptera, Hymenoptera, Lepidoptera, Coleoptera … So, what do we put in?

Don’t think that all these pollinators are species living in groups. Some species are on the contrary very solitary – especially for wild bees. All these species, wild and domesticated, interact with each other. They are also referred to as specialist pollinator species (those that visit a relatively small number of flowering plant species) and generalists (those that visit many species). The corollary is the same for plants – specialist plants are pollinated by only a fairly small number of species, whereas generalist plants (certainly much more open-minded) are pollinated by many species. Some flowers will for example have a unique pollination because of the size of the insects (some pollinators are small, others have a too big rostrum…). Our dear Apis Mellifera is therefore far from being the only one we have to worry about – all other species being impacted by similar issues (neonicotinoids, climate change…). We will talk about it later.

Figure 4. Simplified mapping of some pollinators in the World. Source: IPBES (2018).

Digital in beekeeping

Theoretically, each activity of bees (including flight, defense, cleaning, and brood rearing) requires specific energy that is transferred :

into heat (resulting in changes in temperature, humidity and gas content),
sound (which can be detected by sound and vibration sensors)
and other measurable activities.

Add to this all the things that can worry the beekeeper (swarming, water and food availability, queen absence, brood absence, colony death, starvation and the first spring cleaning flight; diseases and parasitic pressures…), and it is hardly surprising that a plethora of digital tools have been developed to try to measure a whole bunch of parameters directly or indirectly. I insist on the fact that we did not wait for digital tools to monitor bees! Transects, Floral Observations, Traps, Crowdsourcing …. This blog file is obviously oriented and we are talking here about approaches that use digital devices in one way or another.

One could imagine many ways to organize the presentation of digital tools in beekeeping, and there are many in the literature. Some works are only interested in sensors – it is true that there are quite a few different ones, but it is a pity to limit oneself to that – and propose, for example, to differentiate them into four main categories: (1) weight; (2) temperature, humidity and gases; (3) sounds and vibrations; and (4) forager traffic. Others will prefer to work at different scales, for example, categorizing digital tools according to monitoring at the level of an individual bee, a colony of bees in a hive, or even an entire apiary consisting of several hives. Others will have preferred a more functional approach, categorizing the tools according to what they allow to monitor from an agro-environmental point of view. In this section, I propose two large infographics that mix all these approaches:

A first one on the existing digital tools in beekeeping (Figure 5)
A second one on the parameters monitored in beekeeping with these tools (Figure 6)

I insist here on the fact that the second infographic is more theoretical and informational. The digital tools of a category do not measure all the parameters of this category. Some parameters are only measured in research activities. So don’t make a direct link between the two infographics…

These infographics are subject to change and update.

Figure 5. Digital tools in beekeeping. Source: Author. Adapted from Marchal et al. (2020).

Figure 6. Parameters monitored in beekeeping with digital tools. Source: Author. Inspired by Marchal et al (2020). Caution: Not all digital tools in a category measure all parameters in that category. Some parameters are only measured in research activities

Conditions in the hive – between weight and temperature

Connected scales – that’s pretty telling. We can imagine them placed under the hives to follow the evolution of the weight of the hive over time. And some of them have quite original shapes (Figure 7). We are interested in the relative weight of the hive – the instantaneous weight of a hive is of little interest. Increases in hive weight can provide information on a whole range of direct or indirect parameters: the occurrence of nectar flow during the foraging season (beginning and end of nectar flow) or the daily gain in nectar reserves, food consumption during periods of no foraging, the occurrence of swarming events through a decrease in hive weight, or the estimation of the number of foragers (I will let you go and read the paragraphs on bee counters). The idea is to try to imagine a link with the state of the colony or an environmental stress (pesticides, availability of nectar and pollen in the surroundings…)

The follow-up of the weight curve is not that obvious. The weight of an inhabited hive is already the sum of a lot of things: the weight of the box, the weight of the combs containing the food reserves and the weight of the bees living in it (I remind you that there are a lot of them). The time series of hive weights fluctuate a lot, even within a day, which is why their analysis is not completely trivial – more on that later. The weight curves can indeed be disturbed by water loss due to nectar drying, breathing during the night, early morning departure of the foragers, or rain accumulated on the hive roof.

The state of the bee colonies can also be approached by monitoring the temperature inside the hive because the bees regulate it in a rather impressive way. This ability of the colony to regulate its temperature is dependent on a number of things including the subspecies of bees and the genetic diversity within the colony. Bees raise the temperature by clustering together and contracting their muscles, or lower the temperature by dispersing and evaporating water via their wing beats. All of this incidentally also affects the humidity of the hive. The monitoring of the hygrometry in the hives would obviously not yet have allowed very concrete applications.

The complexity of temperature monitoring lies in the temperature gradients within the hive, which require careful attention to the location of the temperature sensors in order not to get the wrong range of values. The brood – where the larvae are pampered – is said to be stenothermic because its survival and development depend on maintaining the temperature in a reduced range (33 to 36°C), whereas the adults can withstand high temperature variations (they are then called eurythermic adults). Sensors in the brood cluster (rather well inside the hive) will be less affected by temperature conditions outside the hive than sensors near the hive exit. Moving away from the brood cluster, temperature and relative humidity will be more influenced by atmospheric conditions and will be less representative of the integrity of the colony’s homeostasis functions. A temperature probe close to the brood will reflect queen oviposition activity and a worker population able to maintain thermal homeostasis, whereas probes further from the brood, for example at the edges of the hive, will be more indicative of increasing population size as queen oviposition reaches these peripheral areas.

Bees move and change cluster size throughout the year and reduce or eliminate it in winter when temperature gradients are, for example, more intense between the outside and inside of the hive, or between the top and bottom of the hive. In addition to temperature sensors, other work has proposed, for example, the use of infrared to examine the thermal profiles of brood nests and heater bees (or to study the movement of bees inside the hive during the wintering period).

Some sensors have also been developed for monitoring the respiratory gas content in the hive. Bees are said to maintain low oxygen levels in the hive to lower their metabolic activity and thus conserve energy for longer periods of time. Gas sensors in the hive must be checked fairly regularly because bees have a habit of covering any foreign objects with propolis or wax, which interferes with the movement of air through the sensor.

Colony dynamics

We are also beginning to see bee counters operationalized to measure the flow of foragers in and out of the colony throughout the day. You could argue that an observer with a good old-fashioned stopwatch might be enough (this observer could look at the vitality of the bees on the flight board, the speed of entry/exit of the bees, and the number of foragers…) but the activity of the bees on a flight board is more like the Paris metro than a gentle stroll along the beach in winter. These counters are not new, but their use is still mainly confined to the academic world due to the complexity of the measurement: several tens of thousands of foragers pass through a hive in one day (and the trains of bees within the channels make the identification of each individual more complex). It may be added that some bees only make clean-up flights (and thus do not forage) but the measurements from the counters at least allow trends to be identified.

The recognition of the bees is thus done on the flight board of the hive (Figure 3) with two particular desires. Either to follow the colony as a whole – in this case classical cameras with different resolutions and possibly additional lighting are used or counters with infrared transmitters and receivers whose beams are blocked by the presence of bees. Either to follow each bee individually (a forager, a worker, the queen…) or rather a small cohort of bees. For this fine tracking, there are for example quite classical devices like RFID chips (or barcodes on plastic paper) placed on the thorax of the bees and which are read by readers in the hive. We can also find metallic tags, also installed on the thorax of our little hymenoptera. These tags are detected by an electromagnetic field created by an inductive sensor installed in the hive. The bees marked with a chip or pellet are marked manually and one by one – so you can imagine that we can’t track an infinite quantity either.

And the applications of these counters seem to be multiple:

Monitoring of the mortality rate in the colony and detection of the time of colony collapse. When night falls, the bees do not come out anymore, and it is then possible to make the difference between the number of bees at the beginning and at the end of the day
Understanding how bees react to a variety of external stresses. The classic bee inflow/outflow profile would look like a trapezoid with an inflow/outflow that increases (more outflow than inflow), stabilizes, then decreases (more inflow than outflow). During a heat wave, large peaks of bee outflow appear in the morning. The bees ventilate the hive more to save the brood and the conditions are not good for nectarification when the flowers are dry. As soon as the weather is cooler, the bees go out in priority to find water.
Individual bee monitoring is also underway to continue to evaluate the effect of neonicotinoids on bee cognition (more on this later). Bees, equipped with RFID chips, can be subjected to sub-lethal doses of neonicotinoids so that their behavior can be studied and the dangerousness of the pesticides can be judged before they are put on the market.
Evaluation of the dynamics of the colony or of the bees inside the hive: for example, one can imagine monitoring the behavior of the queen in the hive and her interaction with the surrounding bees.

In general, tagging bees offers a whole host of potential applications that we will discuss further in the course of the dossier: visualizing and analyzing bee movement patterns in free flight, deciphering bee dance and agitation, exploring the state of the hive (population size, health, dynamics), using chemical sensory abilities of constrained and free-flying bees, evaluating acceptance of a new queen or queen failure, monitoring colony health during winter. The possibilities seem endless.

Sound plays a crucial role in the life of bees. Bees transmit such signals through the vibration of their thorax and wing muscles, and the movement of their wings. Sound is transmitted through the air and the vibrations are propagated through the wax combs or by contact between two individuals. The sounds produced by bee colonies would not only be altered when bees are exposed to different stressors, and even potentially discriminating according to a type of chemical, pest or disease. Beekeepers have understood this. They listen to their hives and certain sounds are no longer unknown to them (for swarming, for example) and bees do not emit the same frequencies in the morning as in the afternoon

Measuring the sound itself is not a complicated task, but the location of the sensor and the influence of environmental sounds other than those of the bees must be analyzed. It is also important to determine the correct frequency at which sound monitoring of bee colonies can be performed. Vibrations and sounds are difficult to track because they can be analyzed both from a temporal perspective (tracking over time) but also from a frequency and amplitude perspective. Bees have specialized receptors in their legs to receive low frequency signals. But obviously, it is a very broad spectrum of vibration frequencies that bees are capable of generating – and we know very little about the range actually used to communicate. From there to predict a swarming a few days, even a few weeks before its realization (and thus to recover the swarm or to prevent the swarming) we are not completely there yet…

We will therefore find here sound sensors for the monitoring of buzzes and frequencies in the hives. Some researchers have proposed laser vibrometry devices by focusing a laser on the wall of a comb cell next to bees with particular behaviors. The detailed knowledge of vibration frequencies could even allow researchers to impose frequencies and see how the bees react and thus detect a particular state of the colony.

Figure 7. Examples of digital devices in beekeeping: digital scales, anti-theft devices, temperature sensors, bee counters.

The use of space by bees

The entry and exit of bees tells us something about a flow, but ultimately does not tell us that much about what the bees are doing outside the hive. Some researchers have proposed to follow the bees with X-band radar, using the same system of electronic chips placed on the thorax of the bees. The radar would still be more suitable for large masses of insects, but it can detect small insects at short range and large insects at long range. Others will use Lidar systems to study flight paths and gathering areas.

Managing pest pressures

Bees are subject to virulent pressures. Among them, there are mainly certain species of hornets that literally devour the bees as they leave the hive and others, such as the varroa mite, that will enter the hives and cling to the bees. Research projects are underway around microphones and sound sensors to detect the frequencies of vibrations characteristic of hornets. Some would even go as far as developing laser systems that, supported by an image analysis system, would take the life of hornets with a high intensity, short distance beam.

To return to the bee counters using video systems presented above, the applications can also be applied to the counting of parasitic pressures – in particular the varroa mite hanging on the bees (other parasitic pressures are more visible at the level of the colony as a whole: Asian hornet, small hive beetle, foulbrood, wax moth). Mirror systems, both to detect the presence of a varroa mite on the ventral side of the bee, and to illuminate in specific wavelengths, can be used. There are also sporadic smartphone applications designed to take pictures of frames taken out of the hives by beekeepers (for a post-processing varroa count) but these applications do not yet seem to be very effective. Another suggestion is to place a white plastic cloth under the hive and count the varroa mites that fall out or that the bees discard. The fine counting and discrimination of varroa mites and bees falling into the cloth does not seem completely trivial…

The state of the environment and external environmental conditions

Environmental conditions are significantly different between inside and outside the hive – we have talked about the temperature gradients inside the hive itself. Knowledge of weather factors outside the hive is perhaps more useful for recontextualizing data acquisition inside the hive than for factual information about what is happening outside. Examples of this are rain and wind, which can affect the quality of the measured data – for example, rain artificially increasing the measured weight of the hive, or wind affecting temperature gradients or humidity levels in the hive. In addition, high atmospheric pressure makes bees sensitive, and having this information in advance would make it possible to avoid disturbing them too much at that time.

Bees also operate in an environment where resources can be limited and this lack of food and water may actually be one of the main reasons for the decline in bee populations (more on this below). Tools are in place to map meliphytic areas of interest to beekeepers and/or to use open land-use databases (such as the graphical parcel register or information from natural history museum sites, for example) to find out what bees will or will not be lucky enough to find around the hives and to indirectly gain information about adjoining agricultural practices. These tools are also used indirectly to reflect on the transhumance strategy for beekeepers. Cameras, initially used to count bee flows, could also be used to quantify the amount of pollen brought back to the hive by the foragers.

Some propose to use this information to calculate eco-scores on sites and territories by cross-referencing information from open databases and geo-located data near the hives.

Sensors designed to detect the presence or concentration of particular compounds can contribute to disease or contamination monitoring.

The cognition of bees

You may have thought that insects were machine animals. On the contrary, they have a nervous system and a brain like ours and are able to perform a whole range of mental operations. Bees can indeed copy other bees, analyze strategies and improve them, count, and feel positive emotions. For example, researchers have shown that a sucrose solution given to bees allowed them to respond positively to ambiguous cues, but also that the bees returned to foraging more quickly after a simulated predator attack (Perry et al., 2016). Some even question their potential state of consciousness and push the boundaries of animal intelligence.

The Econect project wishes to propose cognitive games to bees (the project is not only focused on bees but also on birds, fishes or gastropods…) taken in the environment with various devices (light, sweet water…) to refine our knowledge on the memorization and the cognitive state of bees It is now relatively clear that bees decline because they are subjected to environmental stresses. Behavioral and cognitive metrics are interesting in that they can reflect sensitive responses where, when measuring mortality or lethality, little is actually observed. Even with biomarkers on biological tissue, we would actually have less fine-grained responses than on behavioral or cognitive responses of bees

Researchers are developing more or less automated mazes to evaluate the capacity of bees to learn in relation to a given instruction (capacity to construct learning but also to deconstruct learning by changing the rules of the game for example). Odor or visual stimuli (based on different colored LEDs) are used to reward the bees and metrics are measured to assess how quickly the bees learn or associate a reward. The bees are marked with a barcode, recognized by camera at the entrance of the maze and individually recorded during their passage, and can self-train for hours and days to provide learning and memory curves. These tests can then be used to compare given stress situations by subjecting the bees to various levels of pesticides, heavy metals or nutrition, for example, and then measure how the bees react. This is eco-toxicology beyond survival since, once again, we do not necessarily see the bees die. There are also devices in the form of connected flowers, but it seems that the maze format is a bit more promising.

These digital devices are still in the research phase and it will be essential to test them outdoors because the effects of stress on bees are multi-factorial (heavy metals, lack of food availability, chemical pollutants, micro-particles in the air, etc.) and combined – we talk about cocktail effects. It will also be necessary to move towards population approaches (on the whole colony) and not only on isolated bees.

Risk management and security

It is sad to say that hive theft (often between beekeepers) seems to be a fairly common practice. With the high mortality of bee colonies and the problem of swarm availability, reselling (or reusing) swarms is quite easy. There is also a risk of losing traceability (of the honey, of the bees…) because the bees are not each ear-marked. Anti-theft systems have thus been proposed for the sector, either based on a relatively simple GPS tracking (with alert or not according to the distance to the farm, to the apiary, or to a reference site), or based on sensors of inclination or orientation of hives. It is then considered that if the hive changes its position, something must have happened to it…

Let us also point out the depredations made by the surrounding fauna: wild boars, bears, rodents, birds, and even by other bee colonies in need of honey.

Apiary management

In the straight line of the FMIS software in the more classical plant and animal productions, we also find some inclinations for beekeeping. Digital apiary management tools are used for example to organize and plan the beekeeper’s work, to follow his accounting, or to trace transhumance (since some transhumance routes can bring more or less difficulties on pollination), sanitary care, position of the apiaries or harvest.

Let’s try to take a step back

Is digital technology a panacea in beekeeping?

Mainly logistical interests

Precision beekeeping, as it is called by the main stakeholders (I am quite openly critical of the term precision), holds the promise of a better understanding of the apiary so that the beekeeper can synchronize himself as well as possible with it, to the nearest day or couple of days, so as not to miss a honey flow, to avoid having colonies which decline if there is a lack of water or food (by supplementing their food), or to prevent swarming or blocking of egg-laying (as soon as a bee is born and leaves its cell, the bees will replace the cell with honey and then the colony will collapse). The beekeeper will then decide to visit his colonies on the basis of a first diagnosis of the situation from a distance. In view of the context of super variability and uncertainty in which beekeepers evolve, it is understandable that any tool that can accompany them and reduce their mental load could find interest. A 2019 survey by ITSAP reports on the main data entry practices, expectations and reluctance of beekeepers for digital. Ultimately, little is known about how many of these digital devices are present in the field. The organization of the sector with more amateurs necessarily leaves more room for particular uses (some will work on electro-magnetism, for example) and for a somewhat cultural and empirical aspect of production. This little mysticism is perhaps also quite seductive in the end.

A beekeeper regularly goes to check the activity and the health of his bees. The controls are still quite visual (by looking at the flight board, the frames) and manual (by weighing the hive, for example). These practices are the result of the beekeeper’s training and experience. Currently, digital tools seem to intervene mainly to support the beekeepers on logistic aspects. The remote follow-up of the hives is especially a means to rethink the displacements of the beekeeper, especially when one learns that the transhumance of the hives is the principal post of expenditure, for approximately 40% of the costs of honey production. From there to study the mobility carbon footprint of a beekeeping operation, there is only one step! Because he lives in a climate of uncertainty and does not completely decide on his production, the beekeeper must carefully consider his transhumance strategy to move their hives according to the different flower theaters available and make sure that the flowering calendar coincides correctly with the health of the bee colonies. The apiaries are not necessarily close to where the beekeeper lives, and for professional beekeepers the amount of hives to be moved during transhumance is such that it would be a pity not to try to optimize the movements. Transporting several hundred hives requires very rigorous logistics with the use of large trailers (and potentially on night shift).

Europe and France seem to be very focused on hive productivity and the use of connected scales. In the United States, for example, the trend would be more towards the use of internal sensors in the hive to monitor the temperature and condition of the colony. The scales would have indeed made more inroads than anything else in France. There is a plethora of products available in this area (Figure 5). For brood monitoring, the proposals are more limited and the complexity will lie in the analysis and exploitation of potential temperature curves.

However, I would add here a small limitation of connected scales for transhumance management. First problem: stacking hives on top of each other generates a certain weight on the bottom hives of the stack. You can imagine that a scale can support one hive or even a little more, but not necessarily a whole stack of hives. The second problem is that scales also take up weight and space. If we imagine a scale under every hive, it would require additional logistics for the transhumance. And we can’t leave all the scales on the spot either, so let’s hear it…

From the point of view of working time, one can indeed question the interest of moving to one’s apiary if the supers are not filled with honey or if the bees are a little too excited (I take this opportunity to remind you that the smokers used by beekeepers to calm the bees before opening a hive are not used to put the bees to sleep but to cut off communication between them). When the hive is honeying well, one can see the point of coming back to install or replace empty supers, but otherwise, is there really any point in going back and forth for nothing? Regular opening of the hives is not really recommended as it is time consuming and energy consuming for the bees to rebalance the temperature inside the hive once it has been opened. With the sensors in place, the hives can be monitored without being disturbed, even during periods when invasive hive inspections are contraindicated, such as in winter or during periods of colony stress.

However, beekeepers are not willing to share just any information. We’re starting to hear a lot about consent to share data in the digital ecosystem in agriculture, but in this case, the subject is much more down to earth. Beekeepers are not necessarily very open to talking about geo-location for two main reasons. The first one, and we have already talked about it, has to do with hive theft. Pointing precisely to the location of these hives can raise some fears. There is also competition between beekeepers to find the most interesting locations in terms of abundance or diversity of flora for their bees. It is a bit like the mushroom patches, some protect them body and soul…

Working on voice input doesn’t seem to be too much of an exaggeration either, to make it easier for beekeepers, who are often dressed in a lot of protective clothing, to get information back to the apiary and manage it.

In beekeeping, perhaps more than anywhere else, the sensors must be robust and last long enough to not have to change them too regularly. And these constraints are not necessarily easy to overcome. Temperature sensors can indeed be disturbed by wax, propolis or honey in the hive. Humidity, gas and sound sensors must be cleaned or even replaced when contaminated. With the transhumance routes of beekeepers, it is no fun to have to recharge a sensor’s battery all the time either. Some areas, particularly isolated (in the mountains or elsewhere) can only use satellite communication.

Let’s remain clear on the fact that in any case, a relatively small part of a beekeeper’s activity can be automated, which leaves plenty of room for both digital technologies and beekeepers.

Biomonitoring or bio-monitoring?

Despite the 100 years of professional research we have on the honeybee, we should actually say that we have been working for 100 years on the honeybee apis mellifera. For an insect like the bee that has been around for 80 million years, we still have things to learn. Our knowledge has obviously evolved, both on how it reproduces, lives, uses its brain, feeds itself, and all the links maintained with pollination, but there are still many things that we do not yet understand well (for example, around swarming).

Networking apiaries on the territories – by relying for example on unions and or referent beekeepers – would allow to obtain data chains on different regions to centralize information, to detect weak signals, and to generate knowledge, both for the beekeeping sector, but also to follow the state of biodiversity. For example, there are many hives in vineyards, on the premises of polyvalent farmers, or in cities – hives that, if equipped with sensors, would go from “showcase” hives to “control” hives.

Leaving the logistical prism we discussed in the previous section, could digital devices be seen as anything other than gadgets? If one wanted to be a little prickly, one could argue that the boxes do not prevent hives from dying (although this would still be a good argument for saying that digital technologies are only tools), or that a tipped-over hive will remain tipped over until a beekeeper intervenes. But these tools are mostly used to provide a lot of data that is not yet interpreted much. Knowing that the hive is empty or full? Well, okay… The weight and temperature curves show daily and seasonal variabilities of great amplitude but, from there to know what to do with them, it’s a whole other story. The tools collect data and help to generate knowledge but are not yet widely used to generate a decision and apply it in the field.

Despite all the potentially measurable parameters for monitoring bee colonies, we must admit that the operational reality is much less dreamy. The tools focus mainly on measuring the weight and temperature of the hives, which is sometimes a bit far from all the applications that we have detailed before (Figure 6). The remote connection in real time allows us to see if the living organism is in good condition, but does not allow us to monitor the biology (and therefore potentially the surrounding stresses) or to make concrete decisions. To push the explanation to the end, it is a bit like installing a thermometer and saying that we are taking action against climate change. Measurement and control are two different things. The image may be a little (too) strong, but it has the merit of being meaningful.

Apart from the purely honey production activity, it is finally not so sure that the data collected for a use in biomonitoring will be valued by the beekeepers themselves. We rather think of the territories or the private companies which will be able to benefit from all that. The data collected on a much larger scale than the apiaries can be of interest to develop a more detailed knowledge of the honey productions by type of territory and by honey flow period, or to participate in the creation of regional and national observatories of the honey production. It would be necessary to stop talking about biodiversity at the scale of a site, but always to see how this site is included in its territory, to try to have a more macro vision of the treated subject.

The bee is certainly a tiny drop in the ocean of pollinator species, but its study, also made possible by the fact that beekeeping exists, helps to inform on the general state of insect populations. By being at once a flagship species (attracting public support), an umbrella species (whose conservation needs incidentally protect other species), an indicator species (sensitive to change/degradation) and a keystone species (whose ecological impact is disproportionate to its abundance), pollinating insects, including the honeybee, offer an unparalleled showcase of life.

Because it is an important link in the ecosystem, the bee – as a sentinel of the environment or marker of the living – will represent various impacts on the fauna, the flora, the humans that we can try to approach by translating what the bee lives in its daily life, in the difficulties of its life and adaptation to its environment. By taking a step back on the research envelope covered by the bees (a few km in diameter around a hive), it is actually several hundred million or even billions of plants that are gathered by the bees of a single hive, thus giving access to a rather impressive quantity of information on the state of the surrounding environment.

In addition to the bees, there are actually many different matrices to dig through: pollen, honey, nectar, wax – all of which provide information about the context around the apiary. Pollen is particularly interesting because it is possible to collect it in traps before the bee has even entered the hive (a kind of raw information available). Pollen binds pollutants and also contains information about the plants foraged by the bees, even in trace amounts.

In contrast to the other animal sectors, it is clear that the bee evolves in a completely different environment. Bees work in their hive and are able to explore the environment several kilometers away – sometimes to the point of complete exhaustion if the food resources are too far away (so it is not necessarily a good sign if the bees go very far from their hive). When we talk about beekeeping, all the semantics associated with more classical beekeeping are also used: reproduction, genetic selection, transhumance… But in fact, the notion of freedom surrounding the bee and its hive comes strongly into play.

If there were no domesticated bee species, would we be or have been able to start quantifying the decline in bee numbers? It is true that the practice of testing cognition (see the section on digital tools) seems a bit more complicated with wild bees. Of course, this does not mean that we should look away from other pollinating insects. Some large European projects are starting to include other species such as bombus and some solitary bees. Maybe we will even manage to install sensors on solitary bees as well. Connected hives can indeed be a way to have an indirect cursor on wild pollinators. Could we go so far as to say that this would be a new marker of biodiversity?

The bee is a flagship species that we must continue to study, that we know better and better and that allows us to understand a lot of things about biology and cognition. It is necessary to go further and to see what is happening in other species to, in a way, move from a vision centered on the bee to an open vision on biodiversity. And I am not just talking about considering other species of pollinators, but rather, for the bee, to no longer simply study it as an individual, but to see it as a super-organism in its environment, so as to be able to imagine a more ecological and holistic approach to apiculture.

The honey bee doesn’t seem to be really in danger. We breed them, and there may be more and more of them. But biodiversity as a whole is declining. We can no longer simply extrapolate the results obtained on our domestic bees to wild bees.

Measured data is still rather complicated to use

But why is the raw data left as is and rarely exploited by the main tools on the market? Simply – if I may say so – because it is complicated… Bias, bias, bias, it’s everywhere! Most of the time, the data acquisition time varies from a few minutes to a few days, whatever the type of data collected. Since each colony is unique, the analysis of a few hives for a small part of the bees’ life cycle is not enough to understand all the subtleties of a complete apiary. Biases are even more present with models trained on data from a single hive. Research initiatives that aim to understand the lives of bees must include at least an entire year’s worth of data, as bees have dramatically different lifestyles between summer and winter, and their behavior at the beginning and end of the peak season also differs

A selection of situations to consider:

On weight measurements
- The weight curve of a wooden hive is likely to be affected by changes in the moisture content of the wood, by stagnant water on the hive after a rain, or by quick interventions of the beekeeper on the hive (installation, removal of the supers…)
- The weight curve informs rather on the state of the colony (adults, brood and food reserves) than on the individual only. Even if we could theoretically have more precise information on foragers specifically (compared to other bees), the risk of confusion with the gains and losses of water and pollen remain troublesome.

On temperature measurements
- The surface area for brood rearing varies greatly depending on the colony, and also on the period within the same colony
- The location of the temperature probe is even more complex in winter than in season because the cluster of bees can move in the hive. On less than 15 cm, one can observe a gradient of temperatures going from 5°C in periphery of the cluster to 25°C in the heart of this one

On gas measurement
- Gas measuring devices require a controlled air flow, which can influence the microclimate of the honeycomb.

On humidity measurements
- Humidity sensors are more expensive and must be kept clean and protected from bees because water vapor cannot overcome wax or propolis to reach the sensor

On vibration and sound measurements
- Vibration data from a hive is very noisy.

Bee counters
- Bees may tend to cluster around the sensor, causing distorted readings of bees entering and leaving the hive
- Bees may back up
- Mechanical and electronic bee counters at least alter air movement, removal of dead bees from the hive, and probably other bee activity.
- For counters with cameras, measurements are affected by the placement and orientation of the camera around the hive, but also by shadows of bees in the background that can lead models to count each bee twice, or to count a bee even if it is not in the camera angle
- Hives are sometimes modified to allow for better shooting conditions (lighting or forcing bees to follow a specific path) which can make it more complicated to compare data between different hives.

Economic models still complicated to find

As for any digital tool, the question of the economic model associated with the main beekeeping sensors remains. Some people will say that scales and temperature sensors are too expensive, justifying their statement by the fact that at several hundred euros per tool, it is not economically feasible to install one per hive and even less one per bee – even taking into account the life span of the equipment. In the plant sector, where we reason on several hectares, or in the animal sector where we work with much larger animals (a cow, a sheep), the cost per hectare or per animal is obviously not comparable. In beekeeping, even if it is true that as a good accountant, it is difficult to find contradictory arguments, is it finally so serious to follow only a part of one’s apiary? It is then all the expertise of the beekeeper that must be mobilized to position these tools in areas and hives representative of the complete apiary. More advanced and expensive sensors can be placed in priority sites and hives to ensure the reliability of reference data, while low-cost sensors can be deployed more massively – accepting their lower reliability or calibrating them against reference sensors – to identify trends in the apiary. The questions then become: do I equip them or not? And, if so, how many do I equip and where do I best position them? Even if it seems obvious, we can imagine that the uses are very different between all beekeepers: amateurs, semi-professionals, professionals.

If the sensors themselves are too expensive, new economic models are to be imagined: sponsorship of boxes, sponsorship of hives by individuals or companies, pollination services paid by farmers, services of monitoring the eco-score of the territory by the bees. The question will then arise as to how to activate these segments of biodiversity which are clearly not very profitable for the moment…

The digital sector in beekeeping suffers from all kinds of buzz and announcements. You will tell me that we can extend this observation to the digital ecosystem as a whole…. Nevertheless, the fight to save bees is an argument that rings in many ears and often tends to hit the bull’s eye. The result is that new digital devices are multiplying and, for many of them, aborting, simply because, if the initial message was well marketed, the construction and industrialization of the digital tools meet a more painful reality. Sometimes you just have to go to a few company websites to realize that the pages are either no longer maintained or that it is simply impossible to buy anything material from the company.

Some of the digital solutions offered are more do-it-yourself than anything else. Many beekeepers are passionate hobbyists and it is understandable that they want to prototype by themselves, copy ideas here and there and deploy their own tool. Don’t see any slander here – I’m not questioning whether these Do It Yourself tools work or not – I’m just saying that the DIY approach is difficult to reconcile with the industrialization and operationalization of the tool. There is a great risk that the promises are not fulfilled and that beekeepers who have tested one of these technologies are left on the side of the road a few weeks or months later. Do-it-yourself sensors will necessarily be low-cost, but the effects of announcements on very low-cost sensors may cause all the structures that have tried to industrialize their sensors to explode in flight. A connected beehive at 100€ should be a real eye-opener.

It is also difficult to know who does what. Figure 5 shows indeed a whole patchwork of digital devices, but when we try to go into a little more detail in terms of parameters actually monitored or interpretation of raw data, we are often not at the end of our troubles. No, not everyone is doing mortality analysis, nor is everyone doing swarm prediction or automatic honeybee detection. For many, data is “just” being at least presented or displayed on a mobile app or web tool.

Do digital tools bother bees?

Many of the sensors used in beekeeping are intrusive – they come to sit directly inside the hive. Can we say that they completely disrupt the functioning of the colony? The bees will certainly adapt a little to the intrusion. Bees also produce propolis, which they generally use to plug up unwanted holes and insulate them from foreign objects such as digital devices. The debate about the invasiveness of digital tools is not completely over. Besides that, low frequency waves can disturb bees and the arrival of new digital devices necessarily raises questions about the frequency bands used by these tools.

A mass extinction of invertebrate biodiversity

For many, we have a strong belief that the insect population is declining. We felt this recently during brief periods of outings during Covid-19 when we felt we were seeing much more biodiversity than before (or perhaps we were just paying more attention). The windshield phenomenon is often used as an authoritative argument for the fact that we used to (not sure when) remember having windows (especially the windshield) studded with insects after any car trip. The main problem is that we have no old quantified data of population status and thus we cannot gauge the abundance and diversity of populations when our windshields (or those of our parents and great-grandparents) were strewn with insects. We are thus only comparing the new data we acquire to a potentially distorted (or largely underestimated) reference dating from the first measurements made. We are then dealing with the concept of a shifting baseline in the sense that there is a significant risk that younger generations will take the current, already reduced, abundance of pollinators as normal.

The vast majority of population studies point to significant declines in pollinator species in terms of abundance but also in terms of narrower geographic distribution of species. Although you will see that these studies have their limitations and that the understanding of the decline is partial (I invite you to read the following section), these observed declines are very worrisome in several ways.

If we focus on beekeeping, in France, the bee population does not seem to be in decline. It would in fact hide a very important turnover. But this observation is not identical in all parts of the world. In North America, for example, it seems more difficult to have enough swarm production to meet the demands of pollination services. For honey bees, some speak of 20% average annual mortality in apiaries, with peaks sometimes above 30%. There are thus a lot of colonies to renew.

More particularly on bees, even if we can obviously think that the regular decrease observed in the number of species is due to changes in data collection strategies (difficulty in acquiring certain data, reduction in sampling coverage…), it is still likely that the number of bees has decreased. ), it is still likely that all of the work actually reflects a global decline in bee diversity as many species become rarer and less likely to be found, while fewer species become dominant and perhaps even increase in abundance (or decrease in abundance, perhaps at a slower rate).

In general, endangered insects would not only be restricted to specialists with narrow ecological requirements (which would also depend on specific plants) or to ecological niches, but also to quite common generalist species. One might indeed expect that the species with the most specialized pollination requirements would be the most endangered, but there is no concrete evidence for this.

Typically, plant-pollinator interactions are asymmetrical and usually nested, with a core of generalist species playing key roles, specialized pollinators often depending on generalist plants, and specialized plants often depending on generalist pollinators. Because generalist species are often less vulnerable to change than specialist species, they may be able to partially sustain the network structure under altered conditions. These asymmetric networks are particularly redundant, making them relatively robust to the loss of species and cascading interactions. Plant and pollinator species also have a relatively high rate of rewiring (acquisition of new interactions). This suggests that these species will be flexible to some extent in changing partners in response to the extinction of previous partners With global change underway that affects not only species presence, but also species interactions and interaction pathways, there may be concern that plant-pollinator networks are reaching a tipping point and collapsing despite the prevalent mesh and network structure.

Even if some declining insects can be replaced by others, it is difficult to imagine how a net decline in overall insect biomass could be offset.

A synthetic review of the causes of decline

The causes of decline are so diverse and varied that it is hard to know where to start. I admit that it is rather difficult here to get out of a list like the Prevert but I will try to be as synthetic as possible and to bring some keys of reading:

Climate deregulation potentially impacts all kinds of different levels of organization: the individual level, population genetics, changes at the species level (plant phenology, water use) but also changes at the community level. Climate change is currently threatening pollinators and this factor is likely to become increasingly important as major temporal and spatial decouplings and desynchronizations occur in plant-pollinator interactions (disrupted phenologies, changing species distributions). Species ranges can change both in size and in locality – and some species simply cannot shift their range, making them highly sensitive to rapid climate change.

Bee species may have different thermal tolerances. Future years of drought and heavy rainfall are likely to continue to further weaken bees (those without thermal tolerance, for example), exacerbate other causes of decline (less water availability due to droughts, less nectarification), and increase their mortality. In Corsica, for example, beekeeping has suffered from a chronic lack of water due to drought for several months in a row. The lack of baseline data in many parts of the world hinders our ability to track the effects of climate change as they unfold. On a positive note, climate disruption could still have a positive impact on species with higher temperature tolerances.

The Gaucho, Regent, Cruiser and Poncho scandals have revealed the powerful impact of systemic insecticides (which are found in all plant organs) on insect populations and the presence of phytopharmaceutical lobbies to protect the registration of their products. Systemic insecticides can be sprayed as such or coated around the seeds to be sown (this is called seed treatment). These insecticides have a strong persistence and are found in a lot of different matrices (water, soil, pollen, wildflower nectar…) which implies that they do not act “only” directly on insect populations but on the whole surrounding ecosystem (insect food web, natural biological control mechanisms…).

These systemic products, and in particular the neonicotinoids (which we recently saw reversed in France on beets), are particularly insidious because they often do not directly cause the death of insects (pollinators or not) but act more perniciously at sub-lethal or chronic levels by affecting their behavior and their life span. Their presence, even in small quantities, tends to disorientate the bees, to affect their capacity of memorization and learning (the bees thus risk not finding their hive), to play on their physiological state (malformation of the wings, reduction of the growth), to impact their daily activity (drastic fall of the honey production), or to induce metabolic effects (phenomena of hypoglycemia for example). The bee colonies often mask what they are undergoing and, when a colony dies, it had actually been trying to adapt for a long time. These products are also difficult to dose and are often present in minute quantities in bee corpses, which also decompose very quickly. The evidence for their pernicious impact on all insect species is not yet overwhelming; the effects on honey bee colonies, for example, are still contradictory. Long-term impacts, too, remain controversial. Could herbicides also play all these roles? Research is ongoing.

Agricultural practices are intimately linked to the state of insect populations. We have of course talked about systemic products and particularly neonicotinoids but agriculture has a much broader role than that. Monocultures also reduce the availability of nesting resources and micro-habitats (hedges, dead wood, wetlands, shrubs…). Early mowing of meadows, oversimplification of crop rotation and the lack of leguminous surfaces limit the quantity and diversity of food (polliniferous and nectariferous) available to bees. Let us also remember that the fight against seed treatments (I refer you to the section on neonicotinoids) should not make us forget other questions such as the possible toxicity of certain cultivated species. GMOs also pose the risk of a decrease in food resources because they directly or indirectly contribute to a reduction in neighboring weed populations.

Lack of food availability is a dire consequence of many of these effects, cumulative or otherwise. And let’s not just focus on quantity, but also open our eyes to the quality of the resources provided to bees because the composition of nectar and bees plays into bee attraction and loyalty. The diversity and abundance of wild bees is strongly related to the availability of floral resources in the foraging area (interannual spatial continuity of plants and floral resources). Efforts should focus on periods of deficit so that the bees can work on the balance of their diet throughout the season to ensure that additional food and resources are available. The installation of hives or the acceptance of new hives on a site (e.g., a nature park) must be considered in relation to the amount of resources available.

We have talked about agricultural practices, but it would be too simple to limit the scope to farmers. Beekeepers also have their role to play, but be careful not to fight the wrong battle. In an increasingly anthropized environment, beekeeping is not an easy job. Some beekeepers can be considered as honey hunters because they try to optimize their number of hives and because they regularly move the colonies (a kind of excessive transhumance), which inevitably generates stress for the bees. feeding the hives with glucose syrup, sugar loaf and antibiotics, regularly opening the hives (which creates important thermal shocks for the bees who spend an inordinate amount of time to regain a temperature equilibrium), using ready-made wax plates (instead of letting the bees produce their own wax and thus give a clear identification to their hive) adding empty supers too regularly to boost honey production (the bees will then not succeed in heating the heart of the hive properly) or recovering a stock of honey that should have been left to the bees to spend the winter, are all practices that will also contribute to the decline of the bees. Some beekeepers would even tear off the wings of the queen to avoid swarming… The general tendency is of course to be benevolent, but certain practices must change.

Known as the white wolf in beekeeping, the Varroa Destructor (it sounds like an old Hollywood movie) is one of the mites that has shown its face as a result of the worldwide displacement of classic western honey bees outside their original range. Nevertheless, it is not the only sanitary issue (the small beetle of the meeting, the American lope…) but all of them have not yet arrived in France… It is necessary to note that beekeepers also use insecticides with organic or synthetic molecules (and this in their hive) to protect their bees from these invasions. So there are not only the bad farmers but the subject is perhaps a little too taboo. The Varroa mite would not tolerate too high temperatures that the warm body of the bee cluster could reach relatively easily. The IPBES reports show that the whole sector would benefit from working on hygiene and the control of pests and pathogens in domestic pollinating insects. Pathogens are not new and have continuously coexisted and evolved with bee species, but it is the new introduction of organisms, more or less exotic, that contributes more or less to the decline of insect populations.

As in any ecological system, biological invasions of bees and exotic plants can threaten the stability of native plant-pollinator interaction networks. Invasive plants can locally decrease the abundance of native plant species and associated pollinator communities. Some invasive bee species may compete for nesting and floral resources with other bee species and facilitate the spread of pathogens. It is also conceivable that interbreeding between native and exotic pollinators may affect the genetic diversity of native populations. However, we should not forget that exotic plants can also provide additional sources of pollen and nectar and thus act as a buffer against potential nectar (and pollen) shortages under environmental change. Is the competition that steep? The subject is also still quite controversial

It is important to remember that bees in bee farms are domesticated. These bees have thus been selected for a long time and are necessarily weakened on certain particular traits (which do not serve the direct interest of the beekeepers, i.e. the production of honey). Very few beekeepers in France are also breeders and only a few beekeepers have organized themselves collectively to conduct a breeding program. Beekeepers still remain for many dependent on a selection which comes from the countries of southern Europe and South America and which, not surprisingly, raises legitimate questions on the sanitary and environmental level (I refer you to the part on pathogens and various invasions).

And the replacement of an old queen by a younger one (this is called requeening) is also a factor. Some queens, too, are selected and reared under special conditions to be reintroduced into the farms. We can then question our capacity to choose better queens than the bees and workers themselves (and thus seek a financial manna by this activity) and to reintroduce them correctly into the farms so as not to lose them after a few weeks. The queens will potentially be selected for low swarming, which implies selection against colony reproduction from an evolutionary point of view. Selection programs also tend to select for gentleness as a preferred parameter so as not to have overly excited bees. However, this risks encouraging the entry of foreign worker bees and their mites. Introduced queens are not the mothers of the bees in the hive, and it will take a long time for the bees to die and be replaced by the worthy daughters of the new queen. The queen that remains after swarming will also have to be impregnated, potentially by any drones in the neighborhood (from managed colonies or wild drones). Depending on the genetics of the drone, the bees to be born are likely to be aggressive, unproductive and resource hungry.

I will stop here for the list but we could go further (industrial pollution, metal pollution, noise pollution, light pollution…). In short, if we look, we find.

Understanding the decline remains a terribly complicated task

Measuring is one thing – and we have seen that the actors of the beekeeping sector do not lack ideas in terms of developing digital tools. Understanding and interpreting is another. Understanding the decline of pollinators is devilishly complicated. The causes of this decline – as we have just discussed – are already extremely varied, resulting in a multi-factorial analysis and a combination of intertwined parameters. Some factors are collinear, i.e. they impact the decline in the same or opposite direction without us being able to see very clearly to which factor we can attribute the trends. Other factors are confounding in the sense that they indirectly impact the decline. One example is the case of GMOs, whose effects vary by crop type, but are relatively difficult to separate from crop rotation and plot organization because GMO crops are often produced in monocultures. Monocultures also tend to be a more preferred location for the use of plant protection products. Similarly, most herbicide-tolerant crops are generally accompanied by a reduction in weed populations, reducing the food resources available to pollinators. The effects may also be interactive and additive or not (referred to as cocktail effects)-and despite commendable research efforts, most studies still work on factors in isolation. Evidence of these interaction or cocktail effects is therefore relatively scarce. Cause or consequence, correlation or causation, we always come back to the problems that statisticians know well…

It is important to recognize, however, that the importance of a particular driver varies with geographic and ecological context, as well as with the pollinator species in question. It is therefore very difficult to universally rank drivers on a quantitative basis. Some of these factors may also have only a visible long-term influence. One example is the effects of climate change, which would only become apparent too late due to the reaction and adaptation time of ecosystems.

A second major issue is that so far we have mainly measured local declines – both in terms of spatial distribution but also in terms of the species monitored. The global status of bee (and other pollinator) decline has thus not really been measured (IPBES, 2018). The vast majority of population studies are concentrated in a few well-targeted countries, particularly in Europe and North America. Thus, it is primarily in these localities (and at different spatial scales) that historical data are most likely to be available. One might even add the tendency to sample well-populated and more accessible areas rather than more remote and potentially more biodiversity-rich areas. Studies (both research and participatory science) are mainly oriented towards easily identifiable taxa. And the downward trend in the richness of species records is accompanied by a trend towards increasing dominance of records by a few species. There is a concern that we will miss some key species with important functional roles that would reduce the resilience of the ecosystem as a whole. Some declining species may also have become sufficiently rare in some parts of their range that they are difficult to detect.

Data are therefore lacking, both on the monitoring of species per se (honeybees, wild bees and other pollinators) and in relation to the factors that may be affecting bee decline. We are sorely lacking in access to openly available and highly resolved data (such as that related to pesticide use). Formal establishment of local extinction would require more intensive sampling than is currently being done. It must be assumed that it is relatively difficult to observe bees in their normal environment without significantly disrupting colony functioning. The honey bee is an eusocial species that lives in colonies of several thousand individuals with tasks well distributed between castes.

In addition to the lack of data, data are also disjointed and monitoring programs rarely coordinated in that data acquisition methods and practices vary from jurisdiction to jurisdiction, making it more difficult to reconstruct geographic ranges and species abundances, and to attempt to synthesize global patterns in pollination ecology. Changes in collecting patterns (due to shifts in museum priorities, restrictions on the movement of biological material from biodiversity hotspots, abandonment of natural history research and taxonomy, or implementation of systematic monitoring programs) could lead to a shift from rare to common species, creating a false signal of decline in apparent species richness. As difficult as it is to monitor a population at a point in time “t”, dynamic monitoring (spatial and temporal) of this same population is all the more singular because it requires repeated measurements in time and space.

Studies that focus on understanding the decline – not simply quantifying and measuring the decline – often show results obtained under controlled laboratory conditions. This work makes any form of extrapolation quite touchy because, as discussed above, the environmental conditions surrounding an apiary are extremely varied and the factors influencing bee decline are multifactorial and potentially confounding, additive, or interactive. The actual exposure of bees in the field must be taken into account in order to derive general explanations (but knowledge of this actual exposure is also lacking). To return to the example of pesticides, bees – in their natural environment – are often exposed to mixtures of pesticides which, when applied individually, have no toxicity, but which can induce negative effects when tested in a cocktail. Bee feeding protocols must therefore be as close as possible to field exposure conditions.

Spatially densified monitoring with harmonized sampling methods is certainly a major challenge for a better understanding of the decline of pollinators, and instrumented approaches may have an important role to play in this. One should not expect everything from automated digital devices either, in the sense that although data collection can be quite intense at the individual, colony and apiary levels, these devices do not necessarily inform on what is actually happening outside the hive.

Caution and rigor are therefore required…

Between taboos and debates

A sector dominated by amateur beekeepers

Several interviewees have presented beekeeping as one of the last hunter-gatherer professions, with the ability of beekeepers to read nature and to try to find the best sites for installing hives. The craze for social bees is helping to increase the number of amateur beekeepers. The beekeeping industry is characterized by a large number of these hobbyists and one can then see things from two different perspectives. On the one hand, one could be tempted to say that amateurs necessarily work less well than professionals, because they are less trained (simply because it is not their main activity), that they do not necessarily declare all their hives (and therefore their sanitary interventions are more approximate and less protective than expected), or that they might tend to sell some honey under the table. There would obviously be a lot of missing numbers in beekeeping. Another way of looking at it would be to think that amateur beekeepers are passionate and that they will spend a lot of time on their hobby-passion, to take care of their colonies (less numerous than the pros), and to better understand and follow their health status. The reality is undoubtedly between the two. Some hobbyists are particularly technical in the best sense of the word, and other professionals are at the other end of the tunnel, with a rather low technical level.

Maintaining a line of conduct and supervising the market remains very complicated anyway in a sector with so many non-professionals. But passionate people have the advantage to develop a very strong attention. In countries where beekeeping is managed in a much more intensive way, are the practices really much cleaner? One could argue that those who manage to make a living from beekeeping are people who know their colonies well – otherwise they would have gone out of business a long time ago (or are on the way to doing so).

Being a hobby beekeeper or a professional is certainly much more difficult now than it was a few decades ago. Pressures are growing, hazards and variabilities are more and more uncertain. The cost of raw materials is exploding in beekeeping, such as feed for example, which is essential in a context of global warming, especially when there is not much left to eat (the weight of the hives is decreasing). Bee overwintering is made more complicated by more erratic cold rhythms (in intensity and repetition) that are difficult for the bees to synchronize with. Parasites, at least those that were supposed to die by the cold, continue to develop. There is a high risk of having a very difficult awakening due to mortality in spring 2023. The genetics of the bees may not be mastered and the colonies may cross with others from neighboring apiaries. The production routes imply follow-up routes (follow-up book, traceability of transhumance, certification and control). The production of organic honey for the beekeepers is not obvious either: a good half minium of the surface in flowers foraged must be organic, in fallow land or in natural zone. The varroa mite has to be contained with organic solutions (while there are always phytosanitary products scattered everywhere anyway).

Bees pollinate a large part of the crops but beekeepers do not have the competence to pollinate; they bring the bees, period. Perhaps one way to enhance their profession would be to know as much about them as the farmers do to provide them with a real value-added service. Beekeepers could advise to bring bumblebees because they know it will be cool, or bring hives near the plots because they know the plots will bloom in the coming days.

Do honeybees really put pressure on wild bees?

Does the presence of honeybees really significantly impact the abundance and diversity of wild species? The question is a bit touchy, and some will argue that this is not really the problem – and that our focus should be on the general lack of resources available to bees. Still, if we return to the question at hand, it would seem that the relationship between honeybees and wild bees is quite intimate in terms of pollination. Honey bees are not in fact always the most efficient pollinators of all crops, and there would be a great deal of variation in the pollination efficiency of bees both between species and within honey bees. In fact, wild and honey bees, through their multiple behavioral interactions, would together improve the level of pollination in the field. And this ability is particularly important for cross-pollinated plants (one flower pollinated by pollen from another flower) such as sunflowers. Domesticated species, sometimes more specialized on male or female flowers would indeed be less likely to facilitate cross-pollination while wild bees, by foraging a bit on both, would increase the frequency of honey bee transfers from male to female plants. In some academic studies where honey bee populations were high, some plots did not have all of their flowers pollinated, suggesting that production may be limited by a lack of pollination.

The abundance and diversity of wild bee communities is often associated with increased crop pollination, and sometimes it is species diversity rather than abundance that is important. For example, species can be complementary in pollinating different parts of a flower with different sizes, longer or shorter tongues, and different buzzing behaviors. Some crops simply cannot be pollinated by honeybees. For example, strawberries and tomatoes require sonication pollination (the insect must vibrate very strongly to drop the pollen and it is mainly bumblebees that do this.

Even so, the diversity of (wild) pollinators alone cannot claim to take care of the pollination of entomophilous crops as a whole. Although wild bees are generally more efficient, the diversity of wild bees visiting entomophilous crops is only an extremely limited subset of regional species pools, and conversely, the abundance of common wild bee species is more important than their diversity in ensuring efficient crop pollination. Thus, it actually seems important to separate the two concepts of abundance and diversity when addressing the causes of pollinator decline and their different interactions with each other. Often, a few abundant pollinators provide the bulk of pollination services to plants, and reductions in the abundance of these common species may be more disruptive to pollination than losses in pollinator species richness. Thus, it may be only a few species that are responsible for the majority of network stability. Whether the loss of these core species can lead to network collapse remains an open empirical question.

City bees or country bees?

City or Country? It would seem that our environmental sentinels are quite happy in the city. If some interpret this by the escape of too intense rural conditions (neonicotinoids, lack of resources), others explain it more simply by the fact that in the city, there are finally complex mosaics of different land uses and ecological habitats (private gardens, public spaces, family and community gardens, flowered parks, lawns…). Residents and cities, sensitive to the cause of bees, plant mellipheric species (and sometimes also use synthetic products that are much worse than some used in agriculture). And cities are full of areas built with plant species (cemeteries, schoolyards, university campuses, industrial areas, green roofs …).

Again, if we try to take a step back, the issues raised remain somewhat the same. With too many hives in the city, the competition for resources will also be exacerbated (their number and location should therefore be chosen carefully). Remember the figures given in the introduction: a bee explores up to 3km around a hive. The choice of species to be planted is also important so that all the plants do not bloom at the same time (we should therefore avoid planting mono-specific plants…). Meadows and areas flowered with ornamental species (often industrially produced in different countries of the world) are sometimes rather to be considered as bee-killing plants because these plants – very bright and attractive – are deficient in pollen and thus exhaust the bees which spend their time going back and forth.

The presence of honeybees should not be at the expense of a broader biodiversity, where reduced taxonomic diversity, a loss of specialists and an overrepresentation of super-generalist taxa would be favored. The cities should especially have the mission of informing on the state of pollinators with their beehive in presence.

Towards a change in agricultural and other practices and awareness

It becomes fundamental to establish a dialogue between beekeepers and farmers so that each profession becomes aware of its practices. In some countries, in Quebec or in Canada for example, there are applications to connect beekeepers and farmers. Why don’t we have that here? The subject is certainly still too taboo. Apart from a few madmen, almost all farmers do not take malicious pleasure in using plant protection products…

If there is a collision with a sprayer or a new product on the market, it seems normal that the beekeepers around are kept informed (and/or consulted). If plant cover is going to be shredded (it’s great that it is), the risk is that some foragers will be shredded when the machine passes by. By collecting information on the practice of farmers in the vicinity of the hives, it also becomes easier to synchronize the transhumance of the hives with the flowering of certain mellipheric species. And we can of course imagine a whole range of conservation practices implemented by farmers to promote biodiversity actions: strengthening existing diversified agricultural systems, investing in ecological infrastructures, ecological intensification… It is not only farmers who have to make efforts and it is necessary that beekeepers also provide data on their current practices because the plot management software is not as developed as those that can be found in other plant and animal sectors.

Numerous policies in place to monitor biodiversity

After the rather unconditional focus on carbon (and all the carbon assessment methods – I refer you to another blog entry), it is now biodiversity that is starting to draw all the crowds (and without being a great prophet, we should be on the water in a short time). In France, Article 29 of the 2019 Energy and Climate Law requires investors to report on biodiversity. Regulations are starting to arrive, notably at the European level with the Green Taxonomy, the SFDR (Sustainable Finance Disclosure Regulation) and the CSRD (Corporate Sustainability Reporting Directive) to frame reporting, transparency and the integration of biodiversity impacts and dependencies into corporate management. The actors who have developed climate reporting are also often those who are now investing in biodiversity.

Approaches to monitoring impacts on biodiversity are beginning to appear at all levels, already at the public level (Globio, ABD Index, LPI), and at the private level (notably the GBS method of CDC Biodiversity) in addition to the more general methods of life cycle assessment. These approaches propose aggregated or non-aggregated indicators, dynamic or non-dynamic assessments of the state of biodiversity, and generally focus on either species abundance (MSA for Mean Species Abundance) or potentially disappeared ecosystem functions (PDF for Potentially Disappeared Function). These metrics can also be translated to a unit of area by multiplying them by an impacted area.

Following a vote in the National Assembly in 2021, the bee has become the great national cause of the year 2022. Let us rejoice even if, as I have repeated several times in this dossier, the call to save the bees should not mask the enormous complexity behind a global decline of insects. If we only look at the release of new policies and legislative proposals, we might think that biodiversity is a given. In particular, France reports:

an ABC plan (communal biodiversity atlas) to help carry out local inventories in the territories,
a biodiversity plan launched by the government in 2018,
a national action plan (PNA) “France Land of pollinators” for the period 2016-2020,
a national pollinator plan for the period 2021-2026, for which each region will have to implement its own version.
A national Biodiversity 2030 strategy

As part of its recovery plan, France has also launched its PACTE action plan for biosecurity and animal welfare. Whether there is any redundancy or overlap is another matter.

At the European level, the same battle is going on:

The United Nations Convention on Biological Diversity (CBD) launched its international pollinator initiative with an updated plan for the period 2018-2030 – an initiative that has led to a whole bunch of national action plans and strategies including France’s National Biodiversity Strategy 2030 and international.
The Commission also launched in 2020 the European Union Pollinator Information Hive to list, on a common web platform, conservation actions for wild pollinators.
The European regulation “Animal Health Law” (AHL) should also involve the actors of the beekeeping sector in a more global approach “One health”.
We can also highlight the Zero Pollution action plan (water, air, and soil) which, indirectly, will also serve to protect insects.
Quite original, the European Commission has also launched the Pollinator Park initiative to artificially immerse (in virtual reality) the population in a world without bees.
Other major projects include: https://b-good-project.eu/ led by Wageningen University and https://worldbeeproject.org/

European programs to support the beekeeping sector should also be implemented in the national strategic plans (NSP) of the CAP (Common Agricultural Policy) with a whole bunch of existing or future measures. One thinks for example of the conditionality (maintenance of a minimum surface for elements favorable to biodiversity: melliphere fallows, permanent meadows, wetlands, buffer strips along waterways or sensitive meadows), of eco-regimes, of the HVE (High Environmental Value) labeling, or more specifically of agri-environmental and climatic measures (MAEC) which have their small declination in beekeeping (MAEC API). These specific measures encourage the use of mixtures favorable to the development of pollinating insects, as well as commitments that improve the pollinating potential of bees. Nevertheless, one can still be concerned about the critical opinions of the Environmental Authority (Ae) on the ambition of the French NSP CAP…

As the IPBES reminds us, however, the implementation of many measures for the benefit of pollinators is not that obvious. Multi-level administrative units that do not necessarily communicate well with each other, or the discrepancy between the diversity of practices at the local level and the homogenization of government policies at a more global level clearly hinder the implementation of practices. Policy objectives between sectors are still quite contradictory… We can cite an interesting work of researchers who have identified all the sub-national policies adopted by US state legislatures between 2000 and 2017 on insect pollinators (Hall & Steiner, 2020). Add to this that the effectiveness of these initiatives is largely untested and some have the potential to negatively affect pollinators and their associated native plants. One example is bee hotels, which may actually be used primarily by invasive bees and may even increase the abundance of invasive wasps to the detriment of native bees. Another initiative often put forward is the planting of flowers known to attract pollinators but which are often introduced in areas (natural strip, urban gardens…) without necessarily looking upstream at the proportion of native species present in these places, and therefore potentially threatened by this introduction.

The lack of data on the status of insect pollinators makes it impossible to establish solid red lists and to properly fulfill the IUCN criteria for assessing the level of threat to pollinator species. Without such lists, it is even more difficult to integrate pollinator conservation issues into public planning and conservation policies.

Our anthropocentrism around the bee

Biodiversity is still very much seen as a burden or an expense for the average person. There still seems to be this notion of obligation to take an interest in biodiversity (for example in environmental assessment files). We still consider biodiversity too much not for what it is but for what it brings to man. The famous ecosystem services – widely criticized by many authors – are used as a defence line when we question the actions of a company or a community.
Beekeeping is fashionable and the trend is to install beehives everywhere on rooftops, natural parks or even protected sites. A whole bunch of actors are having a field day and the marketing arguments of CSR and greenwashing professionals are taking advantage of the generalized craze and the enthusiasm of the public to save bees. Announcements and technological promises are also numerous and quite widely supported at their start. But are we really trying to save the bees?

The advertising campaigns pushing for the rescue of bees and the efforts in the field of awareness and popularization are obviously to be commended. They have served the cause and made the bee attractive, reduced the fear of these potential stinging insects, and increased the interest and willingness to protect bees. Several unions have a strong will to raise awareness around them with educational kits (digital or not) and by installing apiaries schools everywhere. Honestly, who doesn’t want to help Maya the Bee? But are these awareness actions numerous enough? The media do not seem to be particularly supportive. Some researchers show, for example, that the topic of pollinators gets far less media attention than climate disruption – which is already not particularly highlighted. Articles on pollinator populations would still remain largely marginalized in scientific reports and the back sections of mainstream media.

In reality, beekeepers and hive installers are breeding a domesticated species. And the call to save the bees could be interpreted as a call to stabilize our current industrialized agriculture as pollinating insects have an important role in the success of our agricultural productions.

I introduced at the beginning of this blog the absolutely fascinating amount of bee species, so much larger than just the honey bees that are in the spotlight. The majority of insects are invisible to the general public and this invisibility is carried over to science – insects are much less studied than vertebrates. We are ignorant about the real diversity of bees and the widespread call to save (honey) bees masks the complexity of the global decline of insects (I refer you to the corresponding section of the blog file) and hides biodiversity issues that are much bigger than us. It is a bit like comparing the work of ornithologists to the study of chickens. Beekeeping and apidology are a bit more than just honey bees…

Some work in biotechnology strives to reduce the demand for pollinators, that is, to limit the dependence of crop varieties on insect pollination. One example is parthenocarpy, a phenomenon that gives plants the ability to produce fruit without fertilization – and therefore without the need for external pollination; and this can be improved through breeding or genetic research.

We must also be concerned about any transhumanist action. When we know that millions of years of evolution have allowed us to improve pollination by plants and insects, should we really try to substitute ourselves to nature and turn to transhumanism? I had already started to talk about this subject in a previous dossier on robotics – some robots (there are not many of them either) being precisely put forward for their capacity to replace insect pollination. Some people are also thinking about the development of robotic bees – with touches of Black Mirror dystopia, but when you look a little closer, the constraints of cost, feasibility, efficiency or ethics appear very complicated to overcome.
Even if we can imagine that there are indeed means and methods to compensate for a lack of pollination in the short term, we cannot decently think that we will be able to compensate for a chronic loss of pollination services in the long term. Can we imagine for a moment the cost of hand pollination when we see the thousands of flight hours of our dear sentinels of the environment? China is paying the price for some crops, but how can we believe for a moment that this solution is not doomed to failure?

The study of ecosystem services and conservation proposals are anthropocentric quests. But is it really that much of a problem? Perhaps it is these human, social, and cultural dimensions that we need to activate first to get people to (re)act and make a real change to ensure insect diversity and abundance in general. The first to benefit from ecosystem services are the wildlife themselves. We must not forget this.

In conclusion

If the first image of digital in beekeeping remains that of a sector playing with scales and temperature sensors, a more in-depth analysis shows the diversity of digital tools available. The operational tools are still mainly focused on beekeeping as such and indirectly on honey production with mainly logistical interests (avoiding unnecessary round trips, lower fuel costs, remote monitoring…). There is not really any fear of a complete digitalization of the apiaries as it remains difficult to automate a large part of the beekeepers’ work. And anyway, as the sector is mainly represented in number by amateurs and enthusiasts, it is difficult to imagine a complete automation.

The promise is made to have tools for a wider biomonitoring of the environment even if, as we have seen, the fine monitoring of insects and the decline of pollinators is particularly complex to apprehend. Between multifactorial, additive, collinear or confounding issues, digital tools will certainly be useful to navigate and see more clearly but they are not a panacea. Digital tools must help to move away from the prism of the bee as an individual towards broader landscape and territorial approaches, and in particular to monitor sanitary practices in a coordinated manner, or to discriminate between territories and the factors influencing the state of health of colonies. But are we ready to put money into this?

The bee is a showcase without common measure on the living. It is at once a flagship species (attracting public support), an umbrella species (whose conservation needs incidentally protect other species), an indicator species (sensitive to change/degradation) and a keystone species (whose ecological impact is disproportionate to its abundance). The public is ready to save bees, and we need to play on the human, social, and cultural dimensions to go further and get people to (re)act and undertake real change to ensure insect diversity and abundance in general. The generalized propensity to save bees should not make us forget the diversity of pollinator and insect species.

The lack of data and our lack of knowledge of all the processes involved should not prevent us from acting. There is no time to simply continue “counting books while the library burns”. Pollinator monitoring and research activities (with digital tools or not) must be part of planning and management actions at a whole host of spatial scales.

Again, most of the current tropism is focused on climate deregulation. Even if the issue is commendable, we cannot and must not forget the living. The crisis of life and biodiversity should worry us much more than the “simple” climate crisis. All actions to save life are also beneficial for the climate – the reciprocal being much less true.

Bibliography complementary to the interviews

Allier, F., et al. (2022). MIELLEES – Le Système Informatique (SI) MIELLEES, un outil de mutualisation des données des balances connectées de ruches pour décrire les miellées. Innovations Agronomiques, 85

Althaus, S.L., et al. (2020). No buzz for bees: Media coverage of pollinator decline. PNAS, 118, 2

Baldock, K.C.R (2020). Opportunities and threats for pollinator conservation in global towns and cities. Current Opinion in Insect Science, 38, 63-71

Bota, G., et al. (2022). Passive acoustic monitoring for estimating human-wildlife conflicts: The case of bee-eaters and apiculture. Ecological Indicators, 142

Bromenshenk, J.J et ak. (2015). Bees as Biosensors: Chemosensory Ability, Honey Bee Monitoring Systems, and Emergent Sensor Technologies Derived from the Pollinator Syndrome. Biosensors, 5, 678-711

Cameron et al., (2009). Patterns of widespread decline in North American bumble bees. PNAS, 108, 2

Decourtye, A., et al. (2019). La ruche connectée : labeille sous surveillance numérique. CAIRN.INFO. Sciences, eaux & Territoires, 29

Decourtye, A., et al. (2019). Toward the protection of bees and pollination under global change: present and future perspectives in a challenging applied science. Current Opinion in Insect Science, 35, 123-131

Dicks, L.V., et al. (2021). A global-scale expert assessment of drivers and risks associated with pollinator decline. Nature Ecology & Evolution.

European Commision (2020). FUTURE BRIEF: Pollinators: importance for nature and human well-being, drivers of decline and the need for monitoring. Science for Environment Policy, 23

France Agrimer (2022a). Observatoire de la production de miel et gelée royale 2022 (données 2021) https://www.franceagrimer.fr/Actualite/Filieres/Apiculture/2022/Retrouvez-la-synthese-de-l-etude-observatoire-de-la-production-de-miel-et-de-gelee-royale-donnees-2021

France Agrimer (2022b). Apiculture – Fiche Filière : https://www.franceagrimer.fr/fam/content/download/68209/document/FICHE%20FILIERE%20APICULTURE%202022.pdf?version=4

Greenleaf, S.A. & Kremen, C. (2006). Wild bees enhance honey bees’ pollination of hybrid sunflower. PNAS, 103, 37

Hadjur, H., et al. (2022). Toward an intelligent and efficient beehive: A survey of precision beekeeping systems and services. Computers and Electronics in Agriculture, 192

Hall, D.M. & Martin, D.J. (2020). Human dimensions of insect pollinator conservation. Current Opinion in Insect Science, 38, 107-114

Hall, D.M. & Steiner, R. (2020). Policy content analysis: Qualitative method for analyzing sub-national insect pollinator legislation. MethodsX, 7

IPBES (2016). Rapport d’évaluation sur les pollinisateurs, la pollinisation et la production alimentaire. Résumé à l’intention des décideurs.

IPBES-IPCC (2021). Co-sponsored workshop biodiversity and climate change. Scientific outcome.https://ipbes.net/sites/default/files/2021-06/2021_IPCC-IPBES_scientific_outcome_20210612.pdf

Knight, T.M., et al. (2018). Reflections on, and visions for, the changing field of Pollination Ecology. Ecology Letters, 21, 1282-1295

Meikle, W.G., and Holst, N. (2015). Application of continuous monitoring of honeybee colonies. Apidologie,

Ministère de la transition écologique & Ministère de l’agriculture et de l’alimentation (2021). Plan national en faveur des insectes pollinisateurs et de la pollinisation 2021-2026, 96p

Lettman, M., & Chauzat, M-P. (2018). Les outils connectés en apiculture : évaluation de leurs applications auprès des apiculteurs français, Anses

Marchal, P., et al. (2020). Automated monitoring of bee behaviour using connected hives: Towards a computational apidology. Apilodogie

Mollier, P. et al. (2009). Le déclin des abeilles, un casse-tête pour la recherche. INRA Magazine, n°9

Panziera, D., et al. (2022). The Diversity Decline in Wild and Managed Honey Bee Populations Urges for an Integrated Conservation Approach. Frontiers in Ecology & Evolution, 10

Perry, C., et al.J. (2016). Unexpected rewards induce dopamine-dependent positive emotion–like state changes in bumblebees. Science, 353, 6307

Potts, et al. (2010). Global pollinator declines: trends, impacts and drivers. Trends in Ecology and Evolution, 25, 6

Ruckelshaus, M.H., et al. (2020). The IPBES Global Assessment: Pathways to Action. Trends in Ecology & Evolution, 35, 5

Sanchez-Bayo, F., and Wyckhuys, K.A.G (2019). Worldwide decline of the entomofauna: A review of its drivers. Biological Conservation, 232, 8-27

Van Klink, R., et al. (2020). Meta-analysis reveals declines in terrestrial but increases in freshwater insect abundances. Science, 368, 417-420.

Zacepins, A. et al. (2015). Challenges in the development of Precision Beekeeping. Biosystems Engineering, 130, 60-71

Zattara, E.E, & Aizen, M.A. (2021). Worldwide occurrence records suggest a global decline in bee species richness. One Earth, 4, 114-123

Personnes Interviewées

NAME	STRUCTURE
Dominique CASSOU RIBEHART	Des Abeilles et Nous
Alexandre DANGLEANT	Itsap
Arnaud ELGER	CNRS
Bertrand LAURENTIN	Apiculteur
Yves LECONTE	INRAE Avignon
Mathieu LIHOREAU	CNRS – Centre de recherche sur la cognition animale
Christian LUBAT	Beeguard
Lorenzo PONS	Mellisphera – Broodminder
Maurice RIVIERE	Abeille et Sagesse & Apiculteur
Charles VALLET	Beeodiversity

Other structures were contacted but the interviewees did not wish to be named