What’s the Buzz? Assessing Efficacy, Synergisms, and Sustainability of Pollinators in Southern Highbush Blueberry (Vaccinium  corymbosum  L.) 

Final report for GS21-244

Project Type: Graduate Student
Funds awarded in 2021: $16,493.00
Projected End Date: 08/31/2023
Grant Recipient: University of Florida
Region: Southern
State: Florida
Graduate Student:
Major Professor:
Dr. Rachel Mallinger
University of Florida
Expand All

Project Information


Southern highbush blueberry (Vaccinium corymbosum L.) is highly dependent upon insect pollination with fruit set and berry weight both decreasing dramatically in the absence of pollinators. Due to this, the abundance and efficacy of pollinators is closely tied to successful blueberry production. In Florida, previous research has shown that wild pollinator visitation rates are low. This leaves growers reliant on managed honey bees (Apis mellifera L.), and still often receiving unsatisfactory pollination due in part to honey bees limited efficacy as blueberry pollinators. Bumble bees (Bombus impatiens), which are capable of buzz pollinating and can be stocked as managed pollinators, have been shown to be more effective pollinators; we found that increasing bumble bee stocking density improves fruit set and yield despite low overall visitation rates as compared to honey bees, indicating a disproportionately positive impact on blueberry pollination. However, the use of these two managed pollinators together to achieve optimal pollination has not been well studied despite the growing proportion of farmers who invest in both pollinators during bloom. The objective of this study is to test the hypothesis that managed bumble bee deployment facilitates and improves the efficacy of non-buzz pollinators by increasing rates of pollen release. Additionally, southern highbush blueberry breeding selections will be assessed to determine their attractiveness to buzz pollinators including their pollen quantity and release rates. The proposed research will provide a better understanding of whether utilizing both managed bumble bees and honey bees together is an economically sustainable approach to improving blueberry pollination.

Final Report: 1) The project addressed the long standing concerns of farmer collaborators associated with southern high bush blueberry in the Southeastern US, particularly in Florida. Farmers have continually pursued strategies to improve pollination outcomes but have often encountered difficulties and insufficient results. This project set out to examine the effects of two managed pollinator species (honey bees (A. mellifera)and common eastern bumble bees (B. impatiens)) in tandem. Both species have been examined independently in the system but their cumulative and/or synergistic effects on blueberry pollination had not been studied. Previous findings (Mallinger et al. 2021) showed that the presence of managed bumble bees significantly and drastically improved pollination outcomes. This finding underscored the value of managed bumble bees but it was unclear how their impact was associated with honey bee presence. In this study all farmers that utilized bumble bees used them in conjunction with honey bees. This meant that it was unclear how bumble bee mediated benefits occurred. Were these benefits a result of bumble bee pollination solely or was their presence also capable of improving honey bee efficacy through increased pollen flow in the system and direct interactions between bees on flowers that lead to more movement? 2) The research approach of the project was to utilize a cage set up with two treatments of pollinators present with 60 mature blueberry bushes. The treatments were a honey bee only treatment that replicated the non-buzz pollinator system that has been historically utilized by blueberry farmers and a honey bee plus bumble bee treatment that replicated the new approach to managed pollinators being adopted by many farmers. This design allowed us to analyze the pollinator visit and yield metrics associated with each treatment and examine what impact the presence of bumble bees had on honey bee visitation, efficacy, and behavior as pollinators. Our educational approach and farmer learning outcomes were primarily achieved via targeted research and extension talks to local shareholders, an extension document highlighting the benefits of bumble bee utilization, and workshops associated with producer organizations. These cumulative efforts and collaborations with extension professionals at UF allowed for the management practices and findings of the project to be seen by dozens of farmer collaborators. Our goal associated with farmer learning outcomes were to educate farmers on the benefits of implementing bumble bees and how their behaviors improve pollination and have been shown to interact with honey bee efficacy. We also set out to improve farmer management associated with managed bumble bees. This primarily focused on when, where, and how to place bumble bee colonies in the field and how what to expect of them as flower visitors. 3)This research found that the presence of bumble bees with honey bees does alter foraging behaviors as a result of direct interactions. When both species are present, we saw significantly more interactions that caused for greater movement across plants within the experimental cages. In an open field setting this kind of interaction has been shown to improve cross pollination and yield as a result. We also found that bumble bees are more frequent visitors and provide better value when farmers plan to bolster pollination services with honey bees already occurring at recommended stocking densities. Data associated with increased honey bee pollen contact in the presence of buzz pollinating bumble bees is being finalized. However, after one season of data we found no evidence that honey bees collect or carry more pollen in the presence of bumble bees. We did find that in cage settings honey bees are capable of collecting blueberry pollen, which is infrequent in open field settings. This finding could mean that through improved breeding and management practices that honey bee efficacy could be improved by allowing for greater pollen foraging frequency. These findings provide clear evidence that bumble bees are an effective pollinator that have the potential to synergistically improve honey bee efficacy when both species are present in blueberry fields. 4) The farmer adoption outcomes that we observed were incremental increases in the number of farmers implementing managed bumble bees, higher stocking densities among those that already utilized them, and better management practices to optimize them. Examples of better management practices were farmers placing bumble bees in advantageous places in the farm such as inside of high tunnels. This places the bumble bees closer to the targeted crop and reduces the chances of honey bee robbing behavior. Farmers were also instructed to remove and properly dispose of colonies at the end of bloom. This was undertaken by farms observed by the research team and can lead to reduced risk of pathogen spill over in native bee populations. 


Project Objectives:

Objective 1: Assess synergy between buzz pollinators and honey bees. The first objective will focus on the pollen deposition of honey bees both in the presence and absence of buzz pollinators. This will provide insight into the mechanisms by which honey bees are able to encounter and deposit blueberry pollen. We know that honey bees are depositing small amounts of pollen when visiting blueberry flowers, even during nectar foraging (Rogers et al. 2013, Benjamin and Winfree 2014). However, it is not clear how they are coming into contact with pollen and what role previous visitors, including buzz pollinators, are playing. Are honey bees able to release pollen on their own despite their inability to buzz pollinate or are they encountering pollen as a result of previous buzz pollination? If the latter mechanism is correct, then the presence of buzz pollinators would likely have a synergistic impact on pollination by increasing pollen release and facilitating honey bee pollination efficacy.

Objective 2: Evaluate pollen quantity and release rates across blueberry genotypes. The second objective will assess how breeding selections vary in pollen quantity and release when buzz pollinated. This will provide a framework for assessing what breeding selections could receive optimal pollination services from both buzz and non-buzz pollinators. We predict that cultivars will vary in their total available pollen quantity as well as the release rate of pollen that occurs in response to buzz pollination. These two factors in combination would influence the amount of pollen that visitors would encounter, thereby influencing both pollinator recruitment and pollen transfer. In collaboration with the Blueberry Breeding Program at UF, we will be able to highlight certain selections to inform future breeding efforts. This outcome would provide producers with cultivars that exhibit greater pollination success.

Objective 3: Examine the influence of floral traits on pollinator visitation and behavior. Finally, the third objective will examine the influence of these pollen traits including pollen quantity and release rates on pollinator recruitment and pollination efficacy. If these traits are in fact impacting pollinator recruitment and behavior, they could play a key role in pollen transfer and pollination success.. These understudied traits could play an important role in the attraction and effectiveness of pollinators and represent an area of interest for plant breeding to improve pollination. Examining variation across selections in pollen quantity and release will enhance our understanding of why growers report differences in fruit set and yield across cultivars as well as improve our understanding of the efficacy of pollinators.


Click linked name(s) to expand/collapse or show everyone's info


Materials and methods:

Obj. 1: Synergisms Between Buzz and Non-Buzz Pollinators

            To assess the impact of buzz pollinators on honey bee pollination efficacy, I will examine pollen deposition by honey bees with bumble bees present and absent. To accomplish this, four paired large-scale blueberry high tunnel production sites will be used. The high tunnel setups will have identical stocking densities of honey bees with colonies placed outside of the tunnel with the end open to allow visits. One of the high tunnel sites per pair will receive additional managed bumble bee colonies at a density of approximately 1 quad (4 individual colonies) per acre placed inside of the tunnel (Koppert Biological Systems). At both sites, pollinators will be allowed to forage on blueberry flowers. After approximately 2 hours of active foraging within the high tunnels a minimum of 15 honey bees actively visiting blueberry flowers will be collected in each treatment per sampling date. Sampling will be replicated twice-weekly for a minimum of 6 times across the bloom period. Honey bees will be collected and individually placed in Eppendorf tubes on ice as they forage in the high tunnels. Upon returning to the lab, bees will be placed in a freezer until they are ready to be processed. Processing will begin with honey bees being dried for 96 hours at 40°C and then stored at room temperature. Honey bees will then be individually washed using a solution of 12 mL of saline with 100 µL of a 0.1% aqueous solution of malachite green stain (Vaissiére & Froissart 1996). The solution with the bee will be ultrasonically probed for 60 seconds to remove all pollen from the body of the honey bee. This washing method will be repeated to ensure that all pollen is successfully removed (Vaissiére & Froissart 1996). Blueberry pollen grains will then be identified, counted, and compared across the two treatments (with and without bumble bees) to test the prediction that bumble bees will increase the amount of blueberry pollen that honey bees encounter and transfer while foraging. Pollen present on honey bees will be analyzed using a t-test to compare the effect of bumble bee presence. Additionally, 10 bagged branches within each paired high tunnel will be observed for honey bee per visit pollen deposition within a standard genotype. Following a single visit by a honey bee to a flower, the stigma will be collected, and pollen grain deposition will be assessed following the above protocols. We will determine if the presence of buzz pollinators improves honey bee per-visit pollen deposition.

Obj. 2: Variation in Floral Traits Across Plant Genotypes

            Pollen availability and release will be assessed in 40 breeding selections cultivated by the UF Blueberry Breeding Lab in Waldo, Florida. To assess this, 10 flowers from each breeding selection will be buzzed by a handheld buzz pollination tool (VegiBee Garden Pollinators) on the highest frequency setting. Buzzes will closely match observed buzz duration of bumble bees by buzzing flowers for 3 seconds. Newly opened flowers will be selected on previously bagged branches to ensure that the flower had not been previously visited by a pollinator.  While artificially buzz pollinating, pollen grains will fall into microcentrifuge tubes. This will tell us the amount of pollen released per buzz. The buzzed flower will also be collected to assess how much pollen remains following a buzz and the total pollen quantity per flower across breeding selections. Following collection, pollen and flowers will be stored in a freezer until processing, at which point they will be dried at 40°C for 24 hours (Vaissiére & Froissart 1996). Once dried, flowers will have their anthers removed and dissected to ensure that all remaining pollen is released. After dissection, the anthers will be placed back in individual tubes with the remaining flower material. Pollen will be washed from flower material following the same protocol as described above for honey bees (Vaissiére & Froissart 1996). Buzzed pollen grains will be stained using malachite green stain. Pollen release and quantity will be assessed by counting pollen grains under a microscope. This objective has been completed in 2021, but lab processing and data analysis has not been completed as of the time of submission. Pollen quantity and release via sonication will be replicated in 2022 and 2023. Analysis will be conducted using ANOVA with pollen release rate and total pollen quantity as dependent variables and breeding selection identity as an independent variable.

Obj. 3: Floral Trait Effects on Pollinator Recruitment and Behavior

            Pollinator behavior observations will take place at the UF blueberry breeding plot in Waldo, Florida. The breeding plot is a mixed planting that includes plants of all 40 focal breeding selections. Managed buzz pollinators will be stocked at current recommendations (4 bumble bee colonies per acre) and allowed to forage throughout the plot. Pollinator visitation and behavior data will be taken on all 40 selections. Observations will take place a minimum of 3 times weekly for the duration of the flowering period of each breeding selection with a total of two 1-minute long observations per selection per sampling date. The number of flower visits for honey bees, bumble bees, and other flower visitors will be recorded along with the number of flowers present on the bush. Behavioral responses such as buzz pollinating and nectar foraging will be recorded. These behavioral responses and visitation rates will be compared across breeding selections and compared with the results of pollen quantity and release rate per selection. This protocol has been completed for the 2021 blueberry season but has not been analyzed as of the time of submission. Pollinator sampling will be replicated in 2022 and 2023. Data will be analyzed using linear mixed effects models with pollinator visitation variables treated as dependent variables and selection as an independent variable, Julian date and weather data will be included as random effects. This analysis will provide evidence of the role that floral traits such as pollen quantity and release play in pollinator behavior.


Research results and discussion:

The project successfully completed both years of planned field work for all three objectives. All data and analyses for Objective 1 have been completed with the exception of the second year of incidental and corbiculate pollen present on foraging honey bee bodies. The first year of this data has been assessed and found no significant difference in the quantity of pollen present on honey bees based on whether they were foraging with bumble bees or not. We found that cages that included honey bees and bumble bees at a cost equivalent to honey bee only cages had significantly more floral visitation and saw more direct interactions occurring between pollinators. Both of these factors in an open field setting would mean that flowers would receive a greater number of visits and would be expected to receive more cross pollen. These have been shown to be among the most important factors associated with fruit set and yield in the southern highbush blueberry system. Within our cages we saw no treatment level effect on fruit set or yield. We expect that this is a result of the extremely high stocking densities that were used in the experiment to better understand the synergisms that were part of our hypothesis. This likely means that any benefits associated with either treatment over the other were overcome by the high pollinator densities and visitation. In open field settings, the outcomes associated with the behavioral findings of the study may lead to yield effects. Objective 1 is currently having all outstanding results finalized and is being prepared for a stand alone publication.

Objective 2 was performed as planned and involved collaboration with another PhD student at the University of Florida. Plant traits including floral morphology, yield, nectar volume, and pollen release were all recorded and are included in a publication that is expected to be submitted in the first quarter of 2024. These findings of this objective indicate that floral traits such as nectar sugar quantity and flower morphology do influence pollinator attraction and are also differentially preferred by different pollinator species (i.e. honey bees and bumble bees). We also found that these traits significantly vary across genotypes and are heritable. This means that future breeding efforts can prioritize pollinator attractive traits in the breeding process to lead to better pollination outcomes in future cultivars.   

Participation Summary
1 Farmers participating in research

Educational & Outreach Activities

1 On-farm demonstrations
6 Webinars / talks / presentations
1 Workshop field days
3 Other educational activities: We participated in a tabling event at the Flora and Fauna festival in 2022 and 2023 and the Nature Day in May at Little Orange Creek Nature Park in 2022. We educated the general public on the importance of pollinators in agriculture, the kinds of bees they may see and the differences in their behavior, and how to help conserve pollinators.

Participation Summary:

25 Farmers participated
40 Ag professionals participated
Education/outreach description:

The above activities have been primarily focused on educating farmers and beekeepers about the pollination benefits associated with different species of pollinators and how to effectively implement managed pollinators into blueberry production. This has been done through extension presentations targeted at beekeepers at the UF Bee College and farmers at Florida Blueberry Grower Association events. One of these FBGA events, a blueberry field day, included members of the research team providing information to blueberry farmers on the biology, pollination benefits, and management practices associated with managed honey bees and bumble bees. In addition to our work with agricultural professionals, similar activities were completed at outreach events targeted at the general public where ~300+ people were reached. Our extension/outreach efforts have primarily focused on teaching about the benefits of pollinators and the differences between them as well as how to managed pollinators should be implemented in agricultural settings. As results are being finalized, continued extension and outreach efforts are planned to discuss the findings of the project. In specific, the synergistic benefits of implementing managed bumble bees alongside honey bees and how breeding traits influence pollinator behavior will be a focus.

Project Outcomes

1 New working collaboration
Project outcomes:

The primary outcome of this project is associated with the positive effects of managed bumble bee adoption. Through our findings, extension efforts, and workshops we have connected with dozens of farmers and contributed to pollination education effects. The largest benefit associated with these efforts are economic and social benefits to farmers. Implementing bumble bees has been shown to improve pollination outcomes, provides more pollinator visits to blueberry on a per dollar basis in our study, and can act as a reliable pollinator when conditions are not ideal for honey bees. These factors combine to provide farmers with a targeted pollination approach that has reduced the pressure of increasing honey bee stocking densities far beyond recommendations. This common practice of high stocking densities of honey bees leads to greater than necessary expenditures and could contribute to competition with native bees.

Of the 64 farmers that participated in the UF Blueberry survey nearly half (31) have reported using managed bumble bees. Through our efforts at FBGA meetings we have reached nearly all of these growers and many that do not currently use them. Stocking densities of bumble bees across farms has remained consistent over the course of the project but many of the 31 growers that do utilize bumble bees were present at our pollinator workshop that included best practices for managed bees.

Knowledge Gained:

The knowledge gained from this project is that improving pollinator diversity in the Florida highbush blueberry system is attainable with managed bumble bees. The successful implementation of managed bumble bees can be achieved by many farmers and can provide benefits due to the complementarity and potential synergisms examined in the project. This really underlines the importance of sustainable economic investment in agricultural settings, especially in regard to pollination. 

Information Products

    Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect the view of the U.S. Department of Agriculture or SARE.