Improving Pollination for the Northeast: On-Farm Testing, Demonstration and Management of the Alfalfa Leafcutting Bee

Final Report for LNE94-046

Project Type: Research and Education
Funds awarded in 1994: $120,000.00
Projected End Date: 12/31/1997
Matching Non-Federal Funds: $48,908.00
Region: Northeast
State: Maine
Project Leader:
Frank A. Drummond
University of Maine, Dept of Biological Sciences
Expand All

Project Information


The goal of the present research was to develop a more sustainable approach to crop pollination by investigating an alternative to the sole use of the honey bee for pollination of highbush blueberry and cranberry. Results from two field seasons (1995-1996) involving 15 participating farmers, combined with our greenhouse bioasssay results (1995-1997) indicate that the alfalfa leafcutting bee (ALB) is a good alternative pollinator of highbush blueberry in fields that are relatively weed-free. Statistical analyses indicated that the ALB and honey bee performed similarly for both highbush and cranberry on the following parameters that measure pollinator effectiveness: fruit set, yield, berry weights, and seeds per berry. Observations of foraging behavior in the field and pollen analyses indicated that the ALB was more faithful to cranberry, whereas the honey bee was more faithful to highbush blueberry. Greenhouse flight cage studies showed that flower handling time and pollen deposition were better for the ALB than the honey bee. Reproductive output was not high on either crop, nor on alfalfa, which suggests that at least some bees would have to be purchased on an annual basis, much like the annual renting of honey bees. Economic analyses indicate that the ALB can provide a net economic benefit to growers for blueberry, but not for cranberry at the present rental price of honey bees. In 1997, the ALB was successfully field released on lowbush blueberry by one of the participating growers demonstrating some farmer adoption of this new alternative pollinator to the honey bee for small fruits.

Project Objectives:

1. Test the alfalfa leafcutting bee (ALB) as a pollinator of highbush blueberry and cranberry.

2. Develop pollination management practices for the ALB designed for highbush blueberry and cranberry.

3. Demonstrate handling, care, and management practices for the ALB to farmers.

4. Evaluate the economics and cost effectiveness of using the ALB as a pollinator of blueberry and cranberry.


Click linked name(s) to expand/collapse or show everyone's info
  • C. S. Stubbs


Materials and methods:
Project Results

Site Information

Acreage and varieties are provided in the Cooperators' Table that accompanies this report. All blueberry fields are in the southwestern portion of Maine. Soils at blueberry study sites ranged from excessively drained and deep to poorly drained and deep soils. Cranberry sites are in Washington County, Maine and soils are not yet mapped for that county. For the blueberry sites, the average annual precipitation is 108.5 cm (42.7 in), average January temp. -5.8oC (21.5 oF), and July temp. 19.8oC (67.7oF). For cranberry, the average annual precipitation is 102.2 cm (40.2 in), Jan temp. -5.8oC (21.5 oF), and July temp. 16.1oC (61.0oF).

Findings and Accomplishments

OBJECTIVE 1. Test the alfalfa leafcutting bee (ALB) as a pollinator of highbush blueberry and cranberry.

The alfalfa leafcutting bee (ALB) was field tested for two years (1995--1996) on 15 participating growers' highbush blueberry fields and cranberry bogs in Maine. Major parameters examined in order to assess its efficiency as a pollinator of these crops were: percentage fruit set, percentage yield, average berry weight, average seeds per berry, and its fidelity to the blooming crop. In 1996, we expanded our research to include cranberry sites that were stocked with honey bees in order to compare the ALB directly to them. Also, in 1996 one cranberry bog was added to the study that was stocked with one quad (one quad = 4 bumble bee colonies) of the commercially available bumble bee, Bombus impatiens. To further ascertain the ALB's potential for these crops, we also investigated flowering handling time and pollen deposition per single visit in controlled greenhouse flight cage experiments (1996--1997).

Percentage Fruit Set (Number of green fruits/number of flowers on marked branches or uprights):

Our fruit set findings demonstrated that the ALB is as effective as honey bees for this parameter and that the use of commercial pollinators improves fruit set. For highbush blueberry, fruit set was high for both the ALB and honey bees. For ALB, the average percentage fruit set was 82.0 ± 11.2% (n= 9 sites; range 38.4--98.2%). For honey bees, it was 82.4 ± 13.3 % (n = 6 sites; range 37.1--100%). Fruit set was 64.6 ± 16.9 % (range 19.1--90%) for the two sites that had no commercial bees released.

Just as in blueberry, cranberry fruit set findings demonstrated that the ALB is as effective as honey bees for this parameter. For cranberry, average fruit set was 52.8 ± 14.5% (n = 9; range 16.1--78.7%) at ALB sites, and 52.6 ± 20.5% (n= 5 sites; range = 38.8--79.8%). Fruit set was 61.2 ± 13.1% at the commercial bumble bee site.

Yield: Several aspects of yield were investigated: 1) percentage yield (number of ripe fruits / number of flowers on marked branches or uprights), 2) average berry weight, and 3) average melanized seeds per berry.

Our blueberry and cranberry percentage yield findings demonstrated that the ALB is as effective as honey bees for this parameter and that the use of commercial pollinators improves yield. For blueberry average percentage yield, although lower for the ALB (60.2 ± 15.9%; n= 9 sites, range 38.7--79.3%) than for the honey bee (69.7± 18.1%; n =6 sites, range 50.8--88.4%), was not significantly so (p = .4795, Mann Whitney). Average percentage yield was lowest at the two sites that had no commercial bees released (56.9%; range 47.3--66.6). For cranberry, average percentage yield was 29.6 ± 9.5 % (n = 9; range 14.6--43.1%) at ALB sites, and 33.6 ± 13.4 % (n= 5 sites; range = 24.3--52.8 %). Percentage yield was 26.0 ± 19.3 % at the commercial bumble bee site. Variety and time of harvest may be a factor needing further investigation as the bog with bumble bees had Stevens, which matures later and has more fruits "hidden in the understory." It was harvested on the same date (9/18/87) as one of honey bee sites with Ben Lear, which matures earlier and is less cryptic (both bogs are owned by the same grower, who thought the percentage yield we obtained was not representative for the bog with bumble bees).

Highbush blueberry weights for the two year period were not significantly different (p = .5270, Kruskal-Wallis) in fields with ALB, honey bees, and no commercial bees, which suggests that the ALB pollinates as well as honey bees and native pollinators. For ALB, the average berry weight was 1.9 ± 0.49 grams (range 0.51--3.6). For honey bees, it was 1.7 ± 0.30 grams (range .66--3.1). For sites with no commercial bees released, it was 1.7 ± 0.32 grams (range 0.41--4.0). It should be noted that some fields in Maine have sufficient field borders (generally the smaller fields) with alternate forage plants and nest sites to support an adequate number of wild bees. Cranberry average single berry weights were also not significantly different (p = .6065, Kruskal-Wallis) in bogs stocked with ALB, honey bees, and commercially available bumble bees, which provides further evidence that the ALB pollinates cranberry as well as the honey bee, and perhaps the bumble bee. (It should be noted, however, that there was only one commercial bumble bee site, so further testing is warranted). Average berry weight was 1.33 ± 0.32 grams (range 0.6--2.4) at ALB sites, 1.25 ± 0.33 grams (range 0.68--1.85) at honey bee sites, and 1.38 ± 0.15 grams (range 1.15--1.72) at the bumble bee site and those berries may have been harvested prematurely.

There were no significant differences in the average number of melanized seeds per berry (a measure of the number of pollen grains deposited on the stigma that actually fertilized an ovule) for either bee on either crop. On highbush blueberry: 50 ± 23 seeds (range 18--129) for the ALB sites and 63 ± 27 seeds (range 6--111) for the honey bee sites. On cranberry: 14 ± 3 seeds for the ALB and 19 ± 5 seeds (range 0--36) for the honey bee.

Growers, also, provided yield information for their total harvest, which was used in the economic analysis. (See Objective 4 below.)

Fidelity to the Blooming Crop: The foraging behavior of the ALB in terms of fidelity to the blooming crop versus fidelity to other plants in bloom at the same time was examined. A total of 1468 one min. observations on blueberry and 1063 on cranberry indicate that the ALB was more faithful to cranberry compared to highbush blueberry, whereas just the opposite occurred for honey bees. Native pollinators (bumble bees, andrenids and leafcutting bees) were more faithful to the blooming crop than either ALB or honey bees. Pollen analyses were consistent with our behavioral observations that ALB were more faithful to cranberry than to blueberry. For ALB released on blueberry, the average pollen load contained <1% Vaccinium pollen. For cranberry, the average was 3.6% Vaccinium (range 0--2.0%) for ALB and 0% for honey bees. Further pollen analyses may be warranted as we changed pollen load processing to a bulk loading of 10 bees per site in 1996; also bees were collected relatively late in bloom at some sites that year. Management of weeds during bloom is essential to keeping the ALB faithful to the crop, just as it is for the honey bee.

Foraging Efficiency: Greenhouse flight cage studies were initiated in 1996 and continued in the spring of 1997 to compare the relative foraging efficiency of ALB to other commercial alternative pollinators. Overall, the flight cage studies demonstrated that the ALB is a more efficient pollinator of cranberry than the honey bee based on flower handling time and pollen deposition. ALB flower handling time on blueberry and cranberry were 7.0 sec/flower and 2.8 sec/flower, respectively. Honey bee flower handling time on blueberry and cranberry were 2.6 sec/flower and 4.2 sec/flower, respectively. For the commercially available bumble bee, Bombus impatiens, flower handling time on blueberry and cranberry were 2.9 sec/flower and 6.7 sec/flower, respectively. ALB and bumble bees deposited significantly more pollen grains per single visit than honey bees for cranberry. No significant differences in pollen deposition existed among the three species for blueberry.

OBJECTIVE 2. Develop pollination management practices for the ALB designed for highbush blueberry, and cranberry.
Major management questions addressed were 1) stocking density, 2) nesting success (production of viable offspring) on the blooming crop, 3) nesting success on a second crop (migratory strategy), and 4) nesting success on alfalfa.

Stocking density is calculated by observing a decrease in fruit set along transects leading from the bee domicile. For ALB, generally, fruit set decreases with increasing distance from the bee domicile. ALB stocking varied from 5,128 to 46,666 cells per acre for blueberry and 15,000 to 33,333 cells per acre for cranberry. No decrease in fruit set was observed in our study transects with increasing distance from the domiciles for either crop. Possibilities that may account for no decrease occurring include: 1) there was poor emergence of the bees due to heavy infestation by parasites, 2) some bees may have dispersed and established nesting sites elsewhere, and 3) for cranberry, the maximum distances from the shelters were, in general, not beyond normal ALB flight range.

Nesting success is determined by comparing the number of bees produced to the number of bees released (the reproductive ratio). In all cases the number of cells (bees) produced was less than the number of cells (bees) set out. Overall, slightly more bees were produced on blueberry (av. reproductive ratio = 0.36) than on cranberry (av. reproductive ratio = 0.34). The migratory strategy of moving the bees to new forage at the end of bloom did not significantly affect the number of bees produced.

To determine if bee numbers can be increased on alfalfa in the Northeast, a total of 17,500 ALB cells were set out at a private farm in northern Maine and 60,000 at University of Maine Rogers' Farm in Central Maine for the two years of this study The average reproductive ratio on alfalfa was 0.42. It appears that farmers will have to buy some bees from suppliers on a yearly basis much like they rent honey bee hives.

OBJECTIVE 3. Demonstrate handling, care, and management practices for the ALB to farmers.
Prior to our first (1995) field season, we made a presentation to the Downeast R & D Cranberry Committee, Cherryfield Maine, introducing the SARE project and providing information on the basic biology, handling, and care of ALB. We were also interviewed by the Houlton Press Herald. (Copies of the R & D minutes and the press article accompany this report). In 1996, demonstrations to farmers included: 1) on July 7 at the "Maine Open Farm Day" at the University of Maine's Roger's Farm, Stillwater, ME; 2) at the Carl & Lorraine Estes' Highbush Blueberry Farm, Gorham, ME on August 13; and 3) on September 20-21 at the Common Ground Fair in , which is the largest organic farmers' fair in Maine. Copies of the materials disseminated at these demonstrations are enclosed with the report. Information on the research, which included the addresses of two participating farmers who were available to discuss this SARE project, was published in the June 27, 1996 University of Maine Cooperative Extension Small Fruit Newletter. (A copy is enclosed.) In 1997, we presented at the annual Spring Wild Blueberry Meeting in Maine a presentation on the Management of the ALB. At the Annual Blueberry Farm Field Day, July 16, 1997 in Jonesboro we and Travis Drake (one of the participating growers on the project), gave a demonstration on aspects of the ALB care and management. Approximately 110 people attended this Field Day. During both field seasons, we also had informal discussions relating to handling and care with the individual farmers participating in the project.

OBJECTIVE 4. Evaluate the economics and cost effectiveness of using the ALB as a pollinator of blueberry and cranberry. See section D "Economic Analysis" below.

Participation Summary


Educational approach:
Dissemination of Findings

Flight cage bioassay results were presented 24 June 1996 at the Seventh International Pollination Symposium, University of Lethbridge, Lethbridge, Alberta. A copy of the subsequent article is provided with this report. Field studies were presented at the 57th Annual Meeting of the Acadian Entomological Society, 11 August, 1997, Kentville, Nova Scotia. A copy of the abstract is included with this report. We will be participating in the Advances in Small Fruit Workshop, where we will present findings from this SARE project, next March at the Eastern Branch Entomological Society of America annual meeting in Cherry Hill, New Jersey. Also, one or more articles in refereed scientific journals are in the preparation stage. A Fact Sheet on the care, handling, and management of the ALB on wild lowbush blueberry has been published under the auspices of the University of Maine Cooperative Extension. Through this SARE project we have demonstrated that the ALB can be a good alternative to the honey bee for pollination of highbush blueberry and cranberry and thus the information contained therein is highly relevant to highbush blueberry and cranberry production. See also Objective #3 above for more on farmer outreach.

Producer Involvement

Over 300 farmers, representing a diversity of cropping systems, have attended our on-farm demonstrations. In 1996, approximately 50 farmers attended the ALB demonstration at the Estes' Highbush Blueberry Farm, Gorham, Maine and 75--100 visited the ALB demonstration at the Common Ground Fair, Windsor, Maine. 187 farmers attend the 1997 Annual Spring Blueberry growers meetings (March 17, 19, 20,22: So. Paris, Union, Ellsworth, and Machias, Maine). On 16 July 1997, approximately 100 farmers attended the demonstration that was part of the Annual Wild Blueberry Field Day in Jonesboro, Maine.

No milestones

Project Outcomes

Impacts of Results/Outcomes

This research provides an alternative to the honey bee for blueberry and cranberry pollination, which make pollination of these economically important crops more sustainable. Also, our flight cage results suggest that the ALB may be more effective as a pollinator than the honey bee on highbush blueberry. Last year, Wymans, one of two major lowbush blueberry producers (Wymans is also establishing a 15 acre cranberry bog) in Maine, successfully conducted a 5 acre field release of ALB and are planning another even larger release this year. We anticipate within the next several years more growers adopting the ALB for lowbush blueberry, highbush blueberry, and cranberry pollination. One of the participating highbush growers (see Farmer Evaluations) expressed interest in speaking with ALB suppliers this fall (1997).

New Hypotheses

Flight cage results as well as those from the field site with commercial bumble bees, suggests that the ALB may be as efficient a pollinator as bumble bees, with both species being more efficient than honey bees. Of course, the other commercially available pollinator's performance's (B. impatiens)on these crops has not been rigorously tested. We have worked extensively with it on lowbush blueberry and have found after it builds up incredibly strong colonies on this crop. Moving it to cranberry after blueberry might make this bumble bee extremely cost-effective. Another area that we think needs further investigation is native bee conservation. In some fields and bogs, particularly the smaller ones, observations of the greater fidelity of native pollinators, such as wild bumble bees, suggest that further research directed at native bee conservation will contribute to the sustainability of crop pollination.

Economic Analysis

The value of the benefit from placing the bees in the field is measured by the increase in yield provided by increased pollination times the market value of the additional fruit. Care must be taken to measure the yield benefit attributable to the bees only, eliminating any other source of yield increase. One way to do this is to measure the average change in yield in the experimental sites and compare it to the average change in yield on comparable control sites where no bees were placed. If the average yield changes are based upon a sufficient number of sites, then the difference between them should give a fair representation of the yield change as a result of ALB activity.

Highbush blueberry yield histories sufficient for measurement were obtained for three experimental sites; Estes, Rivard and Wedge. These yield histories are based on grower information and not on independent yield measurements. Based on the average reported yields before and after ALB were released on the sites and weights based on the relative acreage of each site, the benefit of the leafcutter bees averages 937.91 lb/ac. relative to the average site yields from 1991 through 1994. The price per pound obtained by the blueberry growers in the study ranged from $0.90- to $1.57/lb, with an acreage-weighted average price of $0.9974/lb. The calculated value of the benefit is, therefore, 937.91 lb/ac x $0.9974/lb = $935.44/acre. Given that this benefit is based on such a small number of sites and grower reports instead of independent yield measurements, caution should be exercised in generalizing it to other sites and years.

Insufficient yield histories exist for Maine commercial cranberry bogs to be able to compare sites across time. It is possible, however, to compare sites with ALB to sites with honey bees in 1996. The acreage-weighted yield for the four sites where alfalfa ALB were placed is 2351.6 lb/ac. The average yield for the honey bee sites is 4333.6 lb/ac. Based on these yields, the honey bees contributed an additional 1982 lb/ac when compared to the ALB. Assuming the 1996 cranberry price per pound ($0.80), the value of the honey bees relative to the leafcutters was $1585.60/acre in cranberries in 1996. It should be noted that one factor contributing to the differential in yields obtained by the growers was the fact that stem densities in 1996 may have differed considerably between the cranberry sites with ALB and the sites with the honey bee. Growers with more mature bogs were reluctant to use the "new bee" instead of the honey bee. (They were fearful that the untried ALB might not work pollinate the crop.) Based on our field data, there were no significant differences in fruit set and yield parameters for the two bees.

Another limitation to the economic analyses is that we can say nothing about the benefit of either bee species relative to the situation where no managed bees were present. This could not be calculated with the information available from the study sites. Based on the results in highbush blueberries, it seems reasonable to assume that both species provide a net benefit to growers. If honey bee supplies continue to shrink and rental prices per hive continue to rise, then ALB bees may become a profitable alternative on cranberry in the future. The costs of hive rental for honey bees in Maine in 1996 and 1997 were $40-$55/acre, up more than 30% in three years.

Farmer Adoption

Changes in practice

Two growers with large wild lowbush blueberry holdings (Jasper Wyman and Sons and Cherryfield Foods, Cherryfield, Maine) conducted test releases in 1977 in Maine. Wymans will be expanding their use of the ALB in 1998. It is our understanding that the release by Cherryfield Foods was not successful because of improper handling.

Operational Recommendations

We recommend that highbush blueberry growers consider adopting the use of the ALB over the honey bee at current prices for using either bee. Research (including economic analyses) would need to be conducted to determine whether the ALB is economically more viable than the commercially available bumble bee B. impatiens for blueberry and cranberry.

Two limitations we found are:
1. It does not fly at temperatures as low as the bumble bee, (but then neither does the honey bee).
2. If it is put into the field before there is adequate bloom (5%), it will abandon the nesting shelter.

It should be noted that some training and guidance from people experienced in the use and management of ALB is essential to success.

Farmer Evaluation

The following are quotes from participating farmers in response to our request for feed-back on this SARE project. (Grower phone numbers and addresses are provided on page 13 of this report.)

"Good project." Really unique bee; like the fact that can store the bee in the refrigerator (easy to handle, not aggressive.) Want to talk with suppliers." -- D. Wedge, highbush blueberry farmer.

"Was bothered that 1995 and 1996 were such poor years weather-wise. Wish that the SARE project had been done this year too as weather conditions were great. Had impression that there were more berries closest to the ALB hives. Would love to continue working with you, on pollination or any other research project." -- C. Estes, highbush blueberry farmer.

"ALB are Union bees -only work when sun is shining. Don't do as well as bumble bees or Italians. Was a good study." -- N. Holmes, cranberry grower.

"Quite timid for a bee; not as aggressive as the honey bee. Did not see the yield that the blossoms promised, but tipworm might have affected fruit set and yield. If you are doing anything at all with cranberry next year, would be pleased to have you use my bog again." -- T. Look, cranberry grower.

"It's a promising bee as an alternative to the honey bee for berry pollination. Their importance has increased since you started researching them. You did a good job and showed they're a good working bee." -- T. Drake, Crop Specialist, Jasper Wymans and Sons.

Areas needing additional study

As mentioned in the economic analysis section, the price of renting honey bees for pollination increased 30% above the 1995 price. Our flight cage findings indicate the honey bee may be the least efficient pollinator for blueberry. Therefore, research investigating pollination efficiency (flower handling time and pollen deposition) under field conditions for ALB and commercial bumble bees compared to the honey bee should perhaps be undertaken. Also our flight cage data, which demonstrates how efficient the ALB can be as a pollinator of highbush and cranberry suggest that the effect of weeds on pollinator performance in blueberry fields and cranberry bogs needs investigation.

Two ALB management issues that need further investigation for cranberry are the following: (1) how to best provide leaf material for the ALB for nesting purposes (the cranberry leaf is not suitable nest material); and (2) which shelter design and placement will maximize ALB efficiency on cranberry. Further farmer education is also needed because this bee is very different from honey bees. It is very small and fast moving, which necessitates some training in how to spot the bee on the crop. Also, a training course for beekeepers and farmers in the management of ALB, such as that developed for lowbush blueberry by Agriculture Canada, may be worth pursuing.

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.