Pollination by Bumble Bees for Enhanced Clover Seed Production

Final Report for GW08-014

Project Type: Graduate Student
Funds awarded in 2008: $19,977.00
Projected End Date: 12/31/2010
Grant Recipient: Oregon State University
Region: Western
State: Oregon
Graduate Student:
Major Professor:
Dr. Sujaya Rao
Oregon State University
Expand All

Project Information


Red clover is an important forage legume and rotation crop that is raised for seed in western Oregon. Red clover growers rent hives for pollination, but honey bees are not efficient in this crop due to their preference for flowers in which nectar is more accessible. In contrast, bumble bees are better pollinators of red clover, but currently there is little information on their performance in the Pacific Northwest. This study was conducted to develop strategies for enhancing native bumble bee pollinator populations in red clover seed crops to provide economic benefits to producers through increased seed production.


Red clover (Trifolium pratense) is an important forage legume grown for seed in the Willamette Valley of western Oregon. Pollination is a critical factor in the production of red clover. Given the self-incompatibility of flowers, plants rely on cross-pollination for adequate seed set (Smith, Taylor and Bowley, 1985). Bees serve as the primary pollinating agents in red clover. Traditionally, growers have rented honey bees for pollination by incorporating one to two hives per acre. However, given the increasing difficulties in hive management attributed to parasites and diseases, honey bees have declined significantly in abundance, resulting in low availability and consequently increased costs to growers (Winfree et al., 2007). Furthermore, honey bees were found to be inefficient pollinators of red clover when compared to alternative bee species such as bumble bees (Rao and Stephen, 2006; Rao et al., 2007). Bumble bees exhibit a stronger preference for red clover over alternative flowering species, visiting two to three times more flowers per day than honey bees (Holm, 1966). In a recent cage study conducted by Rao et al. (2007) in a red clover field, bumble bees produced seed set comparable to that of honey bees.

Growers of the Willamette Valley have expressed interest in evaluating bumble bees as alternative pollinators in red clover. Since bumble bees are commercially unavailable in Oregon, growers must rely on feral populations of bumble bees for pollination. Unfortunately, the abundance and foraging behavior of the majority of feral bumble bee species are unknown in the Pacific Northwest (Stephen, 1957). These data are vital to evaluating bumble bees as viable alternative pollinators for red clover seed production.

Project Objectives:

1. Evaluate strategies for drawing native bumble bees to red clover fields.

2. Compare the pollen loads on honey bees and bumble bees visiting red clover fields.

3. Compare bumble bee populations in red clover fields and adjacent native habitats.


Materials and methods:

Objective 1. Evaluate strategies for drawing native bumble bees to red clover fields.

The following two strategies were evaluated in separate experiments based on grower input:

1. Clothes-line of attractive blue vane traps.

To determine if native bees can be drawn to target clover fields, the impacts of clothes-lines of attractive blue cross vanes were evaluated in three red clover seed fields. Each clothes-line consisted of a strong rope attached to poles that were staked firmly to the ground. On each clothes-line, 20 vanes were attached to the rope using twist ties. The following treatments were evaluated in each field:

a) a single clothes-line;
b) a cluster of four clothes-lines; and
c) no clothes-line.

To assess the impact of the cluster of clothes-lines, seed heads were collected randomly from locations close to and away from the clothes-lines. A random set of five seed heads was further selected and the numbers of seeds were recorded for each seed head.

2. Maintenance of a hedgerow border of uncut clover.

Typically, red clover seed producers cut their crop for hay in spring and harvest the second crop for seed. Our observations indicated that during early bloom in the second crop, bumble bee abundance is low. Hence, we determined if maintenance of a flowering hedgerow would draw native bees to the crop during early bloom. Due to grower concerns about impacts of hedgerow plants of other species on production practices of clover, maintenance of a border of uncut red clover was evaluated.

The study was conducted in red clover seed production fields in Polk County in western Oregon. Two growers, both with fields of approximately 30-40 acres, cooperated in this study.  One grower cut his field leaving a 5 m strip of clover at the edge of the field (Field A). A second grower left his entire field uncut (Field B).  Two other control fields (Field C and D) were both cut in early June and both bloomed by July.

Blue vane traps were used to estimate bumble bee abundance. Two traps were set out in each of the four fields and left in place for 48 hours. Trapped bees were counted and recorded.  Sampling was conducted every one to two weeks in each field until late August.

Objective 2. Compare the pollen loads on honey bees and bumble bees visiting red clover fields.

1. Pollen foraging behavior.

The foraging behavior of Bombus vosnesenskii was observed throughout the bloom period of red clover by combining both field and nest site observations.  Colonies reared from wild-caught queens were established within wooden boxes (25.4 x 21.6 x 19.1 cm) and placed within the margin of a red clover field (Figure 1).  All individual female bees were marked with distinctive colored tags (The Bee Works, Ontario, Canada) to determine the foraging activity of workers entering and leaving the nest box (Figure 2).

2. Flight range.

In examining the flight range of B. vosnesenskii, visual surveys were taken during the month of July within the same red clover field as colonies were established.  

3. Pollen load analysis.

Pollen analysis (Beil et al., 2008) is a reliable method for elucidating resource use by bees visiting flowers. Bumble bees and honey bees foraging on red clover bloom were collected in vials in the field, and subsequently their pollen loads were removed and compared. The weight of the pollen from foragers was recorded.  Pollen samples were prepared using acetolysis and analyzed under light microscopy (Erdtman, 1960).  To determine floral composition, pollen types were determined with reference to the literature (Moore and Webb, 1978; Sawyer, 1981) and by comparisons with floral material collected from fields.

Objective 3. Compare bumble bee populations in red clover fields and adjacent native habitats.

A molecular study was conducted to determine the foraging distance and nesting density of bumble bees in red clover landscapes in western Oregon. The study was focused on Bombus vosnesenskii, the dominant bumble bee species in the area.  In 2008, bumble bees were collected by walking through four fields over several days and collecting every B. vosnesenskii worker observed foraging on clover bloom. Individuals were collected using hand-held clear plastic vials with a snap cap. Samples were chilled in the field in a portable cooling box for transport to the laboratory and later frozen at -40°C.  Unfortunately, due to unforeseen circumstances, the samples were contaminated with mold. Hence, a second set of samples were collected during red clover bloom in 2009.

In the laboratory, a single leg, usually the mesothoracic leg, was cut and placed in an eppendorf tube. Bumble bee legs from the same locality and date of collection were placed in the same tube and shipped to the USDA-ARS, Bee Biology and Systematics Laboratory at Logan, Utah, for genetic analysis. DNA was extracted from each sample for comparison of eight polymorphic microsatellite loci.

Research results and discussion:

Objective 1. Evaluate strategies for drawing native bumble bees to red clover fields.

1. Clothes-line of attractive blue vane traps.

We recorded an average of 89% (SE 0.02), 86% (SE 0.02) and 87% (SE 0.04) seed set in the single clothes-line, cluster of clothes-lines, and no clothes-line treatments, respectively. While there was no difference in seed set across the treatments, there was greatest variation in the seed set in the no clothes-line control.

The study suggests that the impact of the blue cross vanes may extend further than we anticipated, which would account for the absence of any differences in seed set from locations close to and some distance away from the clothes-lines. Alternatively, it is possible that since bumble bee populations in red clover fields are high during late July – early August, small increases in parts of the field may be difficult to detect.

2. Maintenance of a hedgerow border of uncut clover.

In all four fields, bumble bee abundance was low in June and early July. Subsequently, in Fields A, C and D, populations rose rapidly and peaked in mid-August before dropping during the last sampling period. There was little difference between the three fields. In contrast, in Field B, no dramatic increase in bumblebee was observed and the populations dropped by mid-August.

The study documented that maintenance of a border of uncut red clover in Field A did not result in drawing bumble bees to the field. It is possible that the strip of uncut red clover was too small to act as a major forage source. On the other hand, Field B had an abundance of red clover in bloom in June and should have attracted bumble bees to the area. Field location may have had an impact as Field B was surrounded by agricultural fields, and it is possible that searching bees were drawn to native landscapes rather than the open farmland area around Field B.

Objective 2. Compare the pollen loads on honey bees and bumble bees visiting red clover fields.

1. Pollen foraging behavior.

A total of 1244 foraging trips consisting of 363 pollen trips were recorded over 160 hours of observation throughout the bloom period. All colonies experienced a shift in pollen collection from early- to mid-July as foragers brought back loads consisting of other plant species in addition to red clover.

2. Flight range.

A total of 17 marked bee foragers were observed along transects. These results suggest that B. vosnesenskii may not be a doorstep forager given its low abundance. However, given the limited number of observed foragers recorded, future work is needed to confirm this assumption to determine the flight range of this species.

3. Pollen load analysis.

Of the individual bumble bees and honey bees collected while foraging on red clover, 66% and 48%, respectively, had pollen. The average size load was 0.012 g per bumble bee and 0.007 g per honey bee. There was variation in time of day when bumble bees and honey bees collected pollen.

Pollen collected from pollen traps placed in honey bee hives in red clover fields during early, peak and late bloom indicated year-to-year variation towards the end of the bloom. In both years, during early and peak bloom, loads consisted primarily of red clover pollen. However, in year 1, there was no red clover in loads from late bloom, while in year 2 we detected 88% red clover pollen.

Overall, at two sites, besides red clover pollen, bumble bees and honey bees collected pollen from weeds including Queen Ann’s lace, false dandelion and thistle.

Objective 3.

Compare bumble bee populations in red clover fields and adjacent native habitats.

Workers of B. vosnesenskii from six clover fields collected during summer were genotyped using eight microsatellite loci and assigned to full sibling families (=colonies). Adequate numbers of workers (> 75) for the microsatellite analysis could not be collected from surrounding habitats. After estimation of numbers of unseen species, we inferred the presence of 162 colonies from 296 genotyped, indicating that 55% of the bees, respectively, originated from different colonies. Workers from the same colonies were observed foraging in multiple fields, indicating that many colonies were utilizing common resources within the landscape.

Bumble bee populations in clover fields were high towards the end of July. At this time, red clover is the primary food resource in the landscape, and hence the isolated fields of mass flowering appear to draw bumble bees to the fields.

Participation Summary

Educational & Outreach Activities

Participation Summary

Education/outreach description:

A display box of native bee pollinators collected from red clover fields along with a summary of current red clover research were prepared and used in a Bee Identification demonstration during the Annual Oregon Clover Grower’s Meeting and Hyslop Farm Field Day demonstration. Tours of experimental clover plots also were made each year during Hyslop Farm Field Day.


Skyrm, K., Rao, S., and Stephen, W.P. 2009. Native bumble bee abundance and foraging behavior in red clover seed production fields of the Willamette Valley. In Young, W.C., Ed., Seed Production Research, Oregon State University Publication 128:82-86.
Skyrm, K.M., Rao, S. and Stephen, W.P. 2010. The occurrence of partial bivoltinism in Bombus vosnesenskii (Hymenoptera: Apidae). (In Prep)
Skyrm, K.M., Rao, S. and Stephen, W.P. 2010. The pollen foraging behavior of the common bumble bee pollinator, B. vosnesenskii, in a mass flowering resource. (In Prep)


Red clover pollination by bees. Hyslop Farm Field Day, Corvallis, OR, May 2010.
A buzzworthy cause: Evaluating factors affecting bumble bee pollinators in Oregon agroecosystems. Pacific Entomological Society of America 94th Annual Meeting in Boise, ID. April 2010.
Comparison of honey bees and bumble bees in red clover seed production- Research update. Oregon Clover Commission meeting, McMinnville, OR, August 2009.
Red clover pollination by bumble bees. Hyslop Farm Field Day, Corvallis, OR, May 2009
The pollen foraging behavior of a native pollinator, Bombus vosnesenskii (Hymenoptera: Apidae), in an agroecosystem. Pacific Entomological Society of America 93rd Annual Meeting in San Diego, CA, April 2009.
Bumble bees and honey bees in red clover seed production: Research update. Annual Oregon Clover Grower’s Meeting, McMinnville, OR, February 2009.
The role of communication in the pollen foraging behavior of feral bumble bees. Oregon State University 1st Annual Student Research in Entomology Symposium, Corvallis, OR, February 2009.
The foraging behavior of a native pollinator, Bombus vosnesenskii (Radoszkowski) (Hymenoptera: Apidae), in flowering red clover. Entomological Society of America 56th Annual Meeting in Reno, NV, November 2008.
Pollination issues in red clover seed production. Hyslop Farm Field Day, Corvallis, OR, May 2008.
Red Clover Pollination: Research update. Annual Oregon Clover Grower’s Meeting, McMinnville, OR, February 2008.

Project Outcomes

Project outcomes:

This study provides new information related to bumble bee associations with red clover seed production in western Oregon. It highlights the diversity and abundance of a complex of bumble bee species present during bloom. Currently, red clover yields in western Oregon are higher than those achieved in majority of other states in the US. However, sustainability and build up of existing populations are critical for the future benefit of red clover seed producers. Additionally, given the high abundance of bumble bees in the area in late summer, it is possible that red clover seed producers may achieve good yield without renting honey bee hives.

We speculate that, in the absence of honey bees, bumble bee populations may be drawn to red clover during early bloom. Hence, this research on red clover may provide valuable economic benefits to producers in reducing costs of production associated with renting honey bees.

Economic Analysis

Seed set across four commercial fields was high (> 80%), documenting that existing pollinators, including rented honey bees (1-2 hives /0.4 ha) and indigenous bumble bees and solitary bees, provide close to maximum pollination in red clover in Oregon. For sustaining high red clover yields in Oregon, conservation efforts that minimize loss of nesting sites and mortality due to pesticides, as well as efforts that provide foraging sources prior to red clover bloom, are needed to protect native bees especially bumble bees. It is possible that bumble bee abundance is adequate for red clover pollination and growers can get by without investment in honey bees.

Farmer Adoption

Red clover growers are beginning to adopt the strategy of depending on native bumble bee pollinators for red clover seed production, thereby saving investment in rentals of honey bee hives.


Areas needing additional study

For achieving even higher yields, improved production practices and new cultivars with more heads/plant will be required. Given the changes in honey bee and bumble bee abundance through bloom, additional information is needed on impacts on yield on time of pollination.

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.