Bridging the Gap between Field Experiments and Machine Learning: The EC H2020 B-GOOD Project as a Case Study towards Automated Predictive Health Monitoring of Honey Bee Colonies.

Honey bee colonies have great societal and economic importance. The main challenge that beekeepers face is keeping bee colonies healthy under ever-changing environmental conditions. In the past two decades, beekeepers that manage colonies of Western honey bees (Apis mellifera) have become increasingly concerned by the presence of parasites and pathogens affecting the bees, the reduction in pollen and nectar availability, and the colonies' exposure to pesticides, among others. Hence, beekeepers need to know the health condition of their colonies and how to keep them alive and thriving, which creates a need for a new holistic data collection method to harmonize the flow of information from various sources that can be linked at the colony level for different health determinants, such as bee colony, environmental, socioeconomic, and genetic statuses. For this purpose, we have developed and implemented the B-GOOD (Giving Beekeeping Guidance by computational-assisted Decision Making) project as a case study to categorize the colony's health condition and find a Health Status Index (HSI). Using a 3-tier setup guided by work plans and standardized protocols, we have collected data from inside the colonies (amount of brood, disease load, honey harvest, etc.) and from their environment (floral resource availability). Most of the project's data was automatically collected by the BEEP Base Sensor System. This continuous stream of data served as the basis to determine and validate an algorithm to calculate the HSI using machine learning. In this article, we share our insights on this holistic methodology and also highlight the importance of using a standardized data language to increase the compatibility between different current and future studies. We argue that the combined management of big data will be an essential building block in the development of targeted guidance for beekeepers and for the future of sustainable beekeeping.

Diverse pollen nutrition can improve the development of solitary bees but does not mitigate negative pesticide impacts.

Floral resource loss and pesticide exposure are major threats to bees in intensively managed agroecosystems, but interactions among these drivers remain poorly understood. Altered composition and lowered diversity of pollen nutrition may reinforce negative pesticide impacts on bees. Here we investigated the development and survival of the solitary bee Osmia bicornis provisioned with three different pollen types, as well as a mixture of these types representing a higher pollen diversity. We exposed bees of each nutritional treatment to five pesticides at different concentrations in the laboratory. Two field-realistic concentrations of three nicotinic acetylcholine receptor (nAChR) modulating insecticides (thiacloprid, sulfoxaflor and flupyradifurone), as well as of two fungicides (azoxystrobin and tebuconazole) were examined. We further measured the expression of two detoxification genes (CYP9BU1, CYP9BU2) under exposure to thiacloprid across different nutrition treatments as a potential mechanistic pathway driving pesticide-nutrition interactions. We found that more diverse pollen nutrition reduced development time, enhanced pollen efficacy (cocoon weight divided by consumed pollen weight) and pollen consumption, and increased weight of O. bicornis after larval development (cocoon weight). Contrary to fungicides, high field-realistic concentrations of all three insecticides negatively affected O. bicornis by extending development times. Moreover, sulfoxaflor and flupyradifurone also reduced pollen efficacy and cocoon weight, and sulfoxaflor reduced pollen consumption and increased mortality. The expression of detoxification genes differed across pollen nutrition types, but was not enhanced after exposure to thiacloprid. Our findings highlight that lowered diversity of pollen nutrition and high field-realistic exposure to nAChR modulating insecticides negatively affected the development of O. bicornis, but we found no mitigation of negative pesticide impacts through increased pollen diversity. These results have important implications for risk assessment for bee pollinators, indicating that negative effects of nAChR modulating insecticides to developing solitary bees are currently underestimated.

Exposure to pollen-bound pesticide mixtures induces longer-lived but less efficient honey bees.

Due to the widespread use of pesticides and their persistence in the environment, non-target organisms are chronically exposed to mixtures of toxic residues. Fungicides, herbicides and insecticides are all found at low doses in the diet of pollinators such as honey bees, but due to the lack of data on the toxicological effects of these mixtures, determining their risk is difficult to assess. We therefore developed a study combining the identification of common pollen-bound pesticide mixtures associated with poor colony development and tested their effects on bee behavior and physiology. We exposed bees to the identified pesticide mixtures during the first days of their adult life, a crucial period for physiological development. Using optic bee counters we recorded the behavior of bees throughout their lives and identified two pesticide mixtures that delay the onset of foraging and slow-down foraging activity. Furthermore, one of these mixtures hampers pollen foraging. As bee longevity is strongly influenced by the time spent foraging, bees exposed to these pesticide mixtures outlived control bees. Physiological analysis revealed that perturbations of the energetic metabolism preceded the altered behavior. In conclusion, we found that early-life exposure to low doses of pesticide mixtures can have long-term effects that translate into longer-lived but slower and less efficient bees. These surprising findings contrast with the commonly reported increase in bee mortality upon pesticide exposure, and demonstrate that exposure that may seem harmless (e.g., very low doses, pesticides not intended to kill insects) can have undesirable effects on non-target organisms.

Stress decreases pollen foraging performance in honeybees.

Foraging in honeybees is energetically demanding. Here, we examined whether stressors, which generally increase metabolic demands, can impair foraging performance. A controlled non-pathogenic stressor (immune challenge) resulted in a change in the foraging preferences of bees. It reduced pollen foraging and increased the duration of trips in pollen foragers. Stress also reduced the amount of octopamine in the brain of pollen foragers (a biogenic amine involved in the regulation of foraging and flight behaviour in insects). According to the literature, flight metabolic rate is higher during pollen foraging than during nectar foraging, and nectar gives a higher energetic return relative to the foraging effort when compared with pollen. We thus propose that stress might be particularly detrimental to the performance of pollen foragers, and stressed bees prefer the energy-rich resource of nectar. In conclusion, stress, even at low levels, could have consequences for bee foraging behaviour and thereby the nutritional balance of the colony.

A 'Landscape physiology' approach for assessing bee health highlights the benefits of floral landscape enrichment and semi-natural habitats.

Understanding how anthropogenic landscape alteration affects populations of ecologically- and economically-important insect pollinators has never been more pressing. In this context, the assessment of landscape quality typically relies on spatial distribution studies, but, whether habitat-restoration techniques actually improve the health of targeted pollinator populations remains obscure. This gap could be filled by a comprehensive understanding of how gradients of landscape quality influence pollinator physiology. We therefore used this approach for honey bees (Apis mellifera) to test whether landscape patterns can shape bee health. We focused on the pre-wintering period since abnormally high winter colony losses have often been observed. By exposing colonies to different landscapes, enriched in melliferous catch crops and surrounded by semi-natural habitats, we found that bee physiology (i.e. fat body mass and level of vitellogenin) was significantly improved by the presence of flowering catch crops. Catch crop presence was associated with a significant increase in pollen diet diversity. The influence of semi-natural habitats on bee health was even stronger. Vitellogenin level was in turn significantly linked to higher overwintering survival. Therefore, our experimental study, combining landscape ecology and bee physiology, offers an exciting proof-of-concept for directly identifying stressful or suitable landscapes and promoting efficient pollinator conservation.

Variations in the Availability of Pollen Resources Affect Honey Bee Health.

Intensive agricultural systems often expose honey bees (Apis mellifera L.) to large temporal variations in the availability (quantity, quality and diversity) of nutritional resources. Such nutritional irregularity is expected to affect honey bee health. We therefore tested under laboratory conditions the effect of such variation in pollen availability on honey bee health (survival and nursing physiology-hypopharyngeal gland development and vitellogenin expression). We fed honey bees with different diets composed of pollen pellets collected by honey bees in an agricultural landscape of western France. Slight drops (5-10%) in the availability of oilseed rape (Brassica napus L.) pollen resulted in significant reductions of all tested variables. Despite some variations in taxonomic diversity and nutritional quality, the pollen mixes harvested over the season had a similar positive influence on honey bee health, except for the one collected in late July that induced poor survival and nursing physiology. This period coincided with the mass-flowering of maize (Zea mays L.), an anemophilous crop which produces poor-quality pollen. Therefore, changes in bee health were not connected to variations in pollen diversity but rather to variations in pollen depletion and quality, such as can be encountered in an intensive agricultural system of western France. Finally, even though pollen can be available ad libitum during the mass-flowering of some crops (e.g. maize), it can fail to provide bees with diet adequate for their development.

Influence of pollen nutrition on honey bee health: do pollen quality and diversity matter?

Honey bee colonies are highly dependent upon the availability of floral resources from which they get the nutrients (notably pollen) necessary to their development and survival. However, foraging areas are currently affected by the intensification of agriculture and landscape alteration. Bees are therefore confronted to disparities in time and space of floral resource abundance, type and diversity, which might provide inadequate nutrition and endanger colonies. The beneficial influence of pollen availability on bee health is well-established but whether quality and diversity of pollen diets can modify bee health remains largely unknown. We therefore tested the influence of pollen diet quality (different monofloral pollens) and diversity (polyfloral pollen diet) on the physiology of young nurse bees, which have a distinct nutritional physiology (e.g. hypopharyngeal gland development and vitellogenin level), and on the tolerance to the microsporidian parasite Nosemaceranae by measuring bee survival and the activity of different enzymes potentially involved in bee health and defense response (glutathione-S-transferase (detoxification), phenoloxidase (immunity) and alkaline phosphatase (metabolism)). We found that both nurse bee physiology and the tolerance to the parasite were affected by pollen quality. Pollen diet diversity had no effect on the nurse bee physiology and the survival of healthy bees. However, when parasitized, bees fed with the polyfloral blend lived longer than bees fed with monofloral pollens, excepted for the protein-richest monofloral pollen. Furthermore, the survival was positively correlated to alkaline phosphatase activity in healthy bees and to phenoloxydase activities in infected bees. Our results support the idea that both the quality and diversity (in a specific context) of pollen can shape bee physiology and might help to better understand the influence of agriculture and land-use intensification on bee nutrition and health.

Nutrigenomics in honey bees: digital gene expression analysis of pollen's nutritive effects on healthy and varroa-parasitized bees.

BACKGROUND: Malnutrition is a major factor affecting animal health, resistance to disease and survival. In honey bees (Apis mellifera), pollen, which is the main dietary source of proteins, amino acids and lipids, is essential to adult bee physiological development while reducing their susceptibility to parasites and pathogens. However, the molecular mechanisms underlying pollen's nutritive impact on honey bee health remained to be determined. For that purpose, we investigated the influence of pollen nutrients on the transcriptome of worker bees parasitized by the mite Varroa destructor, known for suppressing immunity and decreasing lifespan. The 4 experimental groups (control bees without a pollen diet, control bees fed with pollen, varroa-parasitized bees without a pollen diet and varroa-parasitized bees fed with pollen) were analyzed by performing a digital gene expression (DGE) analysis on bee abdomens. RESULTS: Around 36, 000 unique tags were generated per DGE-tag library, which matched about 8, 000 genes (60% of the genes in the honey bee genome). Comparing the transcriptome of bees fed with pollen and sugar and bees restricted to a sugar diet, we found that pollen activates nutrient-sensing and metabolic pathways. In addition, those nutrients had a positive influence on genes affecting longevity and the production of some antimicrobial peptides. However, varroa parasitism caused the development of viral populations and a decrease in metabolism, specifically by inhibiting protein metabolism essential to bee health. This harmful effect was not reversed by pollen intake. CONCLUSIONS: The DGE-tag profiling methods used in this study proved to be a powerful means for analyzing transcriptome variation related to nutrient intake in honey bees. Ultimately, with such an approach, applying genomics tools to nutrition research, nutrigenomics promises to offer a better understanding of how nutrition influences body homeostasis and may help reduce the susceptibility of bees to (less virulent) pathogens.

Diet effects on honeybee immunocompetence.

The maintenance of the immune system can be costly, and a lack of dietary protein can increase the susceptibility of organisms to disease. However, few studies have investigated the relationship between protein nutrition and immunity in insects. Here, we tested in honeybees (Apis mellifera) whether dietary protein quantity (monofloral pollen) and diet diversity (polyfloral pollen) can shape baseline immunocompetence (IC) by measuring parameters of individual immunity (haemocyte concentration, fat body content and phenoloxidase activity) and glucose oxidase (GOX) activity, which enables bees to sterilize colony and brood food, as a parameter of social immunity. Protein feeding modified both individual and social IC but increases in dietary protein quantity did not enhance IC. However, diet diversity increased IC levels. In particular, polyfloral diets induced higher GOX activity compared with monofloral diets, including protein-richer diets. These results suggest a link between protein nutrition and immunity in honeybees and underscore the critical role of resource availability on pollinator health.