A Field Trial of Probiotics: Part 2
Contents
Analysis of the Effect of Probiotics Upon the Gut Microbiomes. 5
A Field Trial of Probiotics
Part 2
First Published in ABJ May 2024
Randy Oliver
ScientificBeekeeping.com
Back in 2020, probiotics for honey bees were all the rage, and beekeepers were spending a ton of money on them. I received a lot of questions as to whether feeding of the products currently on the market was worth the cost. Intrigued, I ran two field trials to give them a fair test. I’ve already written about the practical results of my first trial, but Covid shut down the USDA lab that was about to run the genetic analyses necessary to determine the effects of the probiotics upon the gut microbiomes of the treated bees. I can now share the results –– better late than never!
Introduction
The honey bee gut provides a nutrient-rich environment that is dynamically exploited by various bacteria and fungi. However, all races of Apis mellifera worldwide share a consistent “core” gut microbial community consisting of eight groups of endosymbiotic bacteria [[1]] (plus a large number of other “cryptic” opportunistic bacteria that occur sporadically or at low numbers). It is not yet completely clear as to which of the core bacteria are merely commensals (not actually benefitting the bee), or beneficial mutualists (helping to fight pathogens, assisting in digestion, or producing critical nutrients or other substances), but at least one can be an opportunistic parasitic pathogen [[2]].
There have been a few studies indicating that the feeding of specific strains of the “native” core bacteria as probiotics can be of benefit to the bee [[3]]. However, most commercially-available products consist of bacteria and fungi “foreign” to the honey bee –– species commonly found in mammal guts (easily obtained from one’s own feces) or the soil, and (perhaps most importantly) amenable to mass culture (Figure 1).
Fig. 1 A photograph of the ingredients of one of the tested “direct fed microbials” (DFMs). None of the listed species are native to honey bee guts, the hive, or the floral environment (with the possible exception of Enterococcus faecium).
The question then is whether such bacteria can actually function as probiotics –– generally defined as “live microorganisms that, when administered in adequate amounts, confer a health benefit to the host.” Some beekeepers, upon hearing extravagant claims and gushing testimonials, hope that feeding of these foreign bacteria will confer health or vitality benefits to their colonies.
Or they may do it for another reason. Wikipedia defines probiotics as “live microorganisms promoted with claims that they provide health benefits when consumed, generally by improving or restoring the gut microbiota.” Based upon this supposition, some beekeepers feed probiotics following their applications of antibiotics used to control AFB or EFB, in the hope that the probiotic will restore the gut microbiome.
There is also the possibility that the introduced organisms might function as prebiotics –– “ingredients that beneficially affect the host by selectively stimulating the growth or activity of the beneficial bacteria in the gut.”
Practical application: A recent review summarizes it nicely –– “There is substantial evidence from in-vitro laboratory studies that suggest beneficial microbes could be an effective method for improving disease resistance in honey bees. However, colony level evidence is lacking and there is urgent need for further validation via controlled field trials” [[4]] (emphasis mine).
It just so happens that controlled field trials are something that I do …
My Two Field Trials
At the trade shows of beekeeping conventions, I make a point of asking the sellers of products promoted to improve bee health or performance, for any hard data that they have to support their claims (I usually get an eyes-down shake of the head). At the suggestion of one of the manufacturers of the two probiotics in Figure 2 that I run a trial myself, I undertook two connected large-scale experiments to see whether the feeding of either probiotic would confer measurable benefit to colonies during our stressful dry summer and pollen-deficient autumn in the California foothills (since I assumed those conditions would give the products the best chance to prove themselves).
In these trials my aim was to impartially and open-mindedly provide these probiotics every opportunity to support the claims that I’d heard for them –– either by their salespersons or expectations by beekeepers themselves. So I blinded both myself, my helpers, and my collaborators at the USDA lab as to which treatments were which, by having an uninvolved visiting researcher place the two probiotics and a powdered sugar control into three bottles labeled A, B and C. We did not unblind ourselves until our analyses of the data were completed.
Trial #1
For this trial I used 78 hives, divided between two separate apiaries (for replication of the experiment), and then in each apiary randomly subdivided them into three test groups, to two of which we applied one of the commercial probiotics (Figure 2), with the third serving as a Control group, to which we applied a sham treatment of powdered sugar alone.
Fig. 2 The two commercial DFM probiotics tested. They both make testable claims, including “for gut health and digestive well-being,” “this is a naturally occurring bacteria that is found in the gastrointestinal tract (GIT) of healthy animals,” “healthy honeybees naturally have these bacteria in their GIT,” and “restores bees’ healthy bacteria, revitalizing their immunity.” My question was whether these beguiling claims would be supported by evidence.
So per label instructions, we applied monthly applications of the probiotics, over the course of four months spanning from the end of our honey flow through (intentionally) our stressful summer dearth period, during which we equally fed pollen subs (Healthy Bee, followed by Mann Lake Bulk Soft) and 1:1 sugar syrup to supplement the minimal natural pollen and nectar flows. By the end of the trial, the colonies had largely filled their upper brood chambers with honey, honeydew, and sugar syrup “honey.”
I’ve previously presented my preliminary results and added additional histograms at my website [[5]], but will here visualize the results in a different manner, and can now include the results of metagenomic analysis.
Note on experimental design: For this trial I chose to use colony strength and weight gain as the metrics to evaluate economic benefit, since these are the two outcomes that put money into my wallet.
Results of Field Trial #1
I can now say that Group C was the Control group given a sham treatment of powdered sugar rather than probiotic. I see no need for me to identify which probiotic was which. Allow me to first present the colony strength and weight results visually in Figures 3 & 4.
Fig. 3 I found (in my opinion) that the best way to visually present the results was to sort the changes in colony strengths from smallest to largest, in order to compare the test groups, combining the data from both experimental yards. Most clusters graded smaller in December than in July (despite the supplemental feeding and considerable weight gain, but that was also likely an artifact of it being cooler when we graded them). The only consistent difference was that Control group C appears to have generally retained or gained more strength than did the probiotic-treated colonies (as evidenced by the shorter blue columns of the colonies that lost strength, and the taller blue columns for those that grew).
Fig. 4 Similar to colony strength, there was no consistent difference in colony weight gain, with the Controls again exhibiting consistently better performance (as evidenced by the taller blue columns).
For the statistical geeks, Table 1 shows the averages numerically.
Table 1 There was no appreciable difference in either change in colony strength or pounds gained between the test groups. If anything, the probiotic-treated colonies overall appeared to slightly underperform relative to the group C controls.
Analysis of the Effect of Probiotics Upon the Gut Microbiomes
Since I wanted to give the probiotics every chance to prove themselves, I collaborated with Dr. Kirk Anderson of the ARS Tucson Lab to perform high-throughput sequencing of 16S rRNA bacterial genes to determine the microbiome structure of individual hindguts taken at the start and end points of this trial, in order to compare the gut microbiomes of the probiotic-treated vs. the untreated Control colonies (Figure 5).
Fig. 5 We collected samples of ~50 bees from every hive at each time point, immediately packed them in dry ice, and then shipped them overnight (at considerable cost) to the Anderson Lab for metagenomic analysis, in order to quantify the prevalence of every strain of bacteria in their guts.
Results of the Metagenomic Analysis
I waited to write this article until our results had been published in an open-access peer-reviewed paper [[6]], with the finding that the microbiota of colonies treated with either probiotic for four months did not differ from those of the Control colonies, and that there was only a scattered and sparse presence of microbes that might have come from the probiotics [[7]].
Conclusions from Trial #1
Under the conditions of this trial, we unfortunately found (1) no benefit from long-term feeding of either probiotic upon colony strength or weight gain, (2) nor upon the composition of the bees’ gut microbiomes, (3) nor did the microbes introduced by the probiotics establish a presence in the bees.
Practical application: Many beekeepers and pollen sub manufacturers add these off-the-shelf probiotics to their hives or products, based solely upon glowing testimonials, rather than any clear supportive hard data that they are actually of benefit (much less worth the cost).
But to be fair, there could conceivably have been two more subtle benefits, so I used the remaining healthy-appearing colonies for a follow-up trial.
Trial #2
The failure of the probiotics to improve colony performance didn’t necessarily mean that they couldn’t confer other benefits. So we reused the probiotic-treated colonies to continue with a second experiment related to how the feeding of probiotics might affect the recovery of the bees’ gut microbiomes after treatment with antibiotics, with the aim of determining:
- To what extent treatment with either oxytetracycline or tylosin disrupts the gut microbiome.
- How long it takes for the gut microbiome to “recover” after treatment with antibiotics.
- Whether the feeding of a probiotic helps to reconstitute the core gut microbiome.
- Whether the feeding of a probiotic will suppress the prevalence of pathogens in the recovering antibiotic-stressed bees (fungal load, EFB, nosema, or the troublesome viruses deformed wing virus, black queen cell virus, and chronic bee paralysis virus).
Experimental metrics: Since the objectives of this experiment all related to invisible microorganisms in the bees’ bodies, and since we already had baseline data on the bees’ gut community structure, I sent additional samples of bees taken at specific time points to the Anderson Lab for quantitative PCR and (expensive) high-throughput metagenomic sequencing.
Experimental Setup
To set up this experiment, I unblinded myself as to which was the Control group (which we had not treated with probiotics). We then removed any weak colonies remaining from Trial 1, and divided the remainder into seven test groups (Figure 6).
Fig. 6 I kept the probiotic treatments consistent for each hive for the entire course of the two experiments, and randomly assigned antibiotic treatments to each of the three probiotic test groups, in order to be able to tease out every combination of effect due to probiotic and antibiotic.
We had already taken baseline bee samples on December 2 (Time point 0), so began the three antibiotic applications at 4-day intervals on Dec. 7 — similar to how many beekeepers apply antibiotics going into winter (Figure 7).
Fig. 7 We applied the three antibiotic treatments of either oxytetracycline or tylosin at 4-day intervals, sprinkling the label doses of 200 mg a.i. (active ingredient) over the top bars and bees between the two brood chambers.
We then took bee samples at:
Time Point 1: Three days after the third antibiotic application, under the assumption that at that point the antibiotics would have had their maximum effect upon the gut bacteria. Two days later (assuming that the antibiotics would have degraded by then) we then fed each hive a patty consisting of natural pollen and sugar to provide a gut substrate for bacterial growth, and applied the same probiotic that they had been treated with during the summer.
Time Point 2: One week after feeding the probiotics.
Time Point 3: Three weeks after feeding the probiotic, to determine how well the probiotic-treated bees had reestablished their gut microbiome communities compared to the Control bees (Figure 8).
Fig. 8 Timeline of antibiotic treatments (red), feeding of probiotics (green), and the taking of bee samples at four timepoints (yellow).
The weather was unexpectedly warm during the course of the experiment (Figure 9), so the colonies were able to bring in a bit of natural pollen –– perfect conditions to determine the effect of the probiotics.
Fig. 9 I intentionally ran this experiment during the early winter, when the treated workers would be expected to have extended longevity, so that we could track the long-term impact of antibiotic treatment and subsequent recovery of the bees’ gut microbiomes. As it turned out, we enjoyed an extended autumn that year, and the colonies did not go completely broodless during the course of the trial (so at least some the bees sampled at Time Point 3 may have emerged after the application of the antibiotics). Weather chart courtesy personal weather station KCAGRASS50.
The Anderson Lab sequenced the genetics of the gut microbiota from 240 individual bees, taken from 60 hives belonging to the seven treatment groups, identifying 229 “types” of bacteria (operational taxonomic units or OTUs), as well as performing quantitative PCR to identify pathogens. Their informative huge raw data set is readily available at [[8]].
Results of Trial #2
Let’s start with our finding that treatment with antibiotics caused long-term dysbiosis of the gut microbial communities of non-probiotic-treated bees, treated with either oxytet or tylosin, relative to untreated bees (Figure 10, snipped from our published paper).
Fig. 10 Note how consistent, over time, the gut microbiome was in bees from the untreated Control hives, compared to the disruption and poor recovery of the microbiome communities in antibiotic-treated bees. Not shown above is that much (or nearly all in some cases) of the gut microbiome was reduced in all groups at the 7-day time point –– presumably due to cross-colony transfer of the antibiotics.
It’s pretty clear that treatment with antibiotics greatly disrupts the universal gut microbiome that Apis mellifera has coevolved with over millions of years. But what many beekeepers are interested in is whether one can mitigate that adverse effect by feeding a probiotic to help the bees to recover and reestablish a healthy gut microbiome. The Anderson Lab’s analysis statistically concluded that feeding probiotics didn’t make a difference, but I wanted to confirm that visually for the benefit of my readers. So I processed their huge data set to create a chart comparing the gut community structures of individual bees taken from different hives in each test group, arranged in before-and-after matching pairs of charts (Figure 11):
Fig. 11 To produce the above charts, I combined the subspecies of the eight core bacterial groups, and combined the rest of the 221 recorded “cryptic” OTUs under “Unclassified.” The lefthand columns of each pair of charts represent bees sampled before antibiotic treatment (all colonies received treatment with antibiotic); the righthand columns represent the microbiomes 21 days after the feeding of a “recovery” probiotic (none were given to the Controls).
Note that the gut microbiomes of bees sampled prior to treatment with antibiotics were fairly similar in the left-hand columns of all test groups. The feeding of either probiotic did not appear to improve the reestablishment of their core gut microbiome community structures in the right-hand columns, since the “recovered” structures were still out of balance in all groups, whether having received a probiotic or not [[9]]. Also note the relatively small proportion of unclassified bacteria (whether from the probiotics or elsewhere) present in most of the bees.
In summary, the Anderson Lab found that:
- Consistent with other findings [[10]], application of the antibiotics oxytet or tylosin strongly decreased (or sometimes nearly eliminated) the gut microbial community in treated bees [[11]], and caused a persistent dysbiotic effect on the community structure of the hindgut microbiome lasting at least three weeks.
- They could not detect any measurable benefit from the feeding of the tested probiotics upon the reestablishment of the bees’ gut microbiomes [[12]].
- They could not detect any correlation between feeding of a probiotic and the abundance or prevalence of the seven common pathogens they tested for [[13]].
Discussion
As much as I would have liked to have confirmed the claimed benefits for the two DFM (direct-fed microbial) probiotics that I tested, our data did not provide any supportive evidence for those claims ––as far as improvement in colony strength or performance, or via genetic analysis of their effect upon the endosymbiont microbiomes, or pathogens in the bees’ bodies.
Practical application: Many beekeepers spend money on probiotics as an “insurance policy,” or in the expectation that they will in some way “make their colonies healthier,” “improve their performance,” or “suppress pathogens.” This might well have been the case had the DFMs contained beneficial “native” strains of bacteria able to establish in the bees’ guts, but that didn’t occur with the products tested.
Aside from that, one finding of great interest is our confirmation of the very negative impact of our commonly-used antibiotics on the normally well-established beneficial endosymbiotic “core” microbiome in healthy bees. In general, either antibiotic treatment was hard on the gut bacteria, disrupted the established community structure, and generated an environment ripe for exploitation by “outside” bacteria, or strains of the core species exhibiting resistance to those antibiotics.
Practical application: A number of beekeepers are under the impression that “it’s good idea” to feed a probiotic after antibiotic treatments, hoping that it will help their bees’ “normal” gut microbiomes to recover. This brings us to …
A needed disclaimer: As one of the manufacturers themselves recently clarified, probiotics consisting of “foreign” microbes are not intended to help in the reestablishment of the gut microbiome, since those microbes would not be expected to “take up residence,” nor evoke any detectable change in the microbes that are normally present. I personally feel that it behooves the sellers of honey bee probiotics to make this clear on their label!
A Bright Future?
This doesn’t mean that a “recovery” probiotic couldn’t be developed. A recent paper presented “new experimental results showing that non-native bacterial strains from a commercial probiotic product fail to establish in the worker bee gut, but a mixture of native gut bacterial strains colonizes robustly and resembles a natural microbiota in eliciting expression of bee genes related to immunity and metabolism. Though some questions are unanswered, the future of probiotics for honeybees is bright. It may be possible to design specific communities of natural gut isolates that stably replenish gut communities disrupted by the many stressors bees face and that are economical and efficient for use in apiaries” [[14]].
Final thoughts
It’s conceivable that we may find or develop strains of bacteria that could be introduced to colonies to fight disease, help the bees to digest certain foods, or to produce essential nutrients or other beneficial substances. But keep in mind that honey bees, over the millennia, have already been exposed to virtually every species of bacteria on this planet, and evolutionary pressure is continually selecting for strain-specific core microbiomes “that work well together,” not just in a few labs, but in at least four trillion individual honey bees every day!
Not only that, but every bee gets naturally inoculated with those endosymbionts within hours of emerging from its pupa, so generally would need no help from a beekeeper. So the “proof of benefit” is really up to those selling a product. I ran these experiments as an example of the sort of hard data that we beekeepers would like to see, so that we can make informed decisions as to how we spend our money.
The results of this particular study are for a single trial in the California foothills (objective, impartial, blinded, and replicated). However, similar results have been found by other research groups (pers comm), including one [[15]], that noted: “Probiotics, in theory and concept, are a promising solution to enhance bee health, but the current market available products for beekeepers are making claims that far outreach the ability of their products.”
Acknowledgements
Metagenomic analysis is not cheap, so I see why there are not a lot of hard data on this subject –– it cost me more than $8000 for the collection, shipping, and processing of the samples. And that’s not counting the extensive amount of time involved in lab work and the tedious analysis of the data performed by Dr. Kirk Anderson’s group. So I thank the ARS crew, as well as all you beekeepers who have donated to support our research!
Citations and Notes
[1] Bobay, L, et al. (2020) Strain structure and dynamics revealed by targeted deep sequencing of the honey bee gut microbiome. Msphere 5(4): 10-1128.
From the above paper: “The fact that bees from different hives and states present similar strain profiles, whereas many bees from the same hive have completely different strain compositions, suggests that there are complex strain dynamics in the honey bee microbiota.”
The players involved in the “core” microbiome appear to be in continual dynamics, and new metagenomic analyses suggest that the taxonomy is open to revision, with some currently-named “species” exhibiting unique strains and perhaps continually mutating and evolving.
[2] Engel, P, et al (2015). The bacterium Frischella perrara causes scab formation in the gut of its honeybee host. MBio, 6(3): 10-1128.
[3] Patruica, S & I Hutu (2013) Economic benefits of using prebiotic and probiotic products as supplements in stimulation feeds administered to bee colonies. Turkish Journal of Veterinary & Animal Sciences 37(3): 259-263.
Borges, D, et al (2021). Effects of prebiotics and probiotics on honey bees (Apis mellifera) infected with the microsporidian parasite Nosema ceranae. Microorganisms 9(3): 481.
Daisley, B, et al (2020) Novel probiotic approach to counter Paenibacillus larvae infection in honey bees. The ISME journal 14(2): 476-491.
[4] Rodríguez, M, et al (2023) Probiotics and in-hive fermentation as a source of beneficial microbes to support the gut microbial health of honey bees. Journal of Insect Science 23(6): 19.
[5] Oliver, R (2021) A Field Trial of Probiotics American Bee Journal May 2021 https://scientificbeekeeping.com/a-field-trial-of-probiotics/
[6] Anderson, K. E., Allen, N. O., Copeland, D. C., Kortenkamp, O. L., Erickson, R., Mott, B. M., & Oliver, R. (2024) A longitudinal field study of commercial honey bees shows that non-native probiotics do not rescue antibiotic treatment, and are generally not beneficial. Scientific Reports 14(1): 1954. https://www.nature.com/articles/s41598-024-52118-z
[7] Supplementary Figure 1 https://static-content.springer.com/esm/art%3A10.1038%2Fs41598-024-52118-z/MediaObjects/41598_2024_52118_MOESM1_ESM.tif
Supplementary Table 4 https://static-content.springer.com/esm/art%3A10.1038%2Fs41598-024-52118-z/MediaObjects/41598_2024_52118_MOESM5_ESM.xlsx
[8] The raw data is in Supplementary Table 3. See below.
[9] If you’re interested in diving deeper, I’d be happy to share my spreadsheet to show you how to easily do it.
[10] Daisley, B, et al (2020) Lactobacillus spp. attenuate antibiotic-induced immune and microbiota dysregulation in honey bees. Communications Biology 3(1): 534 https://www.nature.com/articles/s42003-020-01259-8
Powell, J, et al. (2021) Field-realistic tylosin exposure impacts honey bee microbiota and pathogen susceptibility, which is ameliorated by native gut probiotics. Microbiology Spectrum 9(1): 10-1128. https://journals.asm.org/doi/pdf/10.1128/spectrum.00103-21
[11] Supplementary Table 3 https://static-content.springer.com/esm/art%3A10.1038%2Fs41598-024-52118-z/MediaObjects/41598_2024_52118_MOESM4_ESM.xlsx
[12] Supplementary Table 5 https://static-content.springer.com/esm/art%3A10.1038%2Fs41598-024-52118-z/MediaObjects/41598_2024_52118_MOESM6_ESM.xlsx
[13] Supplementary Table 2 https://static-content.springer.com/esm/art%3A10.1038%2Fs41598-024-52118-z/MediaObjects/41598_2024_52118_MOESM3_ESM.xlsx
[14] Motta, E, et al (2022) Prospects for probiotics in social bees. Philosophical Transactions of the Royal Society B 377(1853): 20210156.
[15] Damico, M, et al (2023) Testing the effectiveness of a commercially sold probiotic on restoring the gut microbiota of honey bees: A field study. Probiotics and Antimicrobial Proteins 1-10. https://www.biorxiv.org/content/10.1101/2023.09.13.557574v2.full