Five Scientific Studies that Changed the Way I Think About Gardens, Part 3: Wild Bees > Honey Bees

This article is the third in a five part series that I am writing for the Hardy Plant Society of Oregon (HPSO) Quarterly Magazine. I am grateful to the team at HPSO for their editorial skills and feedback. Part 1 (overview, and gardens as ‘islands’ in an urban ‘ocean’), and Part 2 (putting a price on nature) of this series can be found in earlier blog posts.


I distinctly remember the day that I decided I wanted to study wild bees. I was sitting in a darkened auditorium at the American Museum of Natural History in New York City, listening to Claire Kremen deliver the plenary address in a symposium focused on invertebrate conservation.

In her address, Dr. Kremen shared the results of her research on watermelon farms in California’s Central Valley. Like all cucurbits, watermelon requires insect pollination to set fruit. Watermelon, in particular, has a high pollination requirement: it takes at least eight bee visits to deposit the 500-1,000 viable pollen grains needed to set harvestable fruit in seeded watermelons. Seedless watermelons require between 16 and 24 visits by a bee in order to set fruit. Most growers meet the high pollination requirement of melons by renting and placing honey bee hives in fields. Dr. Kremen’s research suggested that a different approach might be possible.

Honey bees are important agricultural pollinators, in large part because they can be moved en mass and placed into farm fields at the exact moment that pollination services are needed.Photo by Isabella Messer.

Dr. Kremen and her colleagues studied three types of watermelon fields that varied in their pest management practices (organic or conventional) and their proximity (near or far) to native habitat in the foothills of the California coast range. These fields were: organic near, organic far, and conventional far. Watermelon fields that used conventional pest management practices and that were located near the foothills were not included in this study because this type of farm did not exist in the region. The team determined the number of pollen grains that different bee species deposited on watermelon, by presenting different bees with a single watermelon blossom that had yet to receive any insect visits. After a blossom was visited by a single bee, the flower was bagged and tagged accordingly, and the number of pollen grains deposited on the stigma by that single bee was counted in the lab. They repeated this process for 13 different bee species.

Next, the team sat in watermelon fields and observed what types of bees visited watermelon blossoms, in different types of farm fields. Watermelon flowers are only receptive to pollination visits for a single day. They recorded the sex and species of each bee visitor to the blossoms. Based upon these species specific counts, combined with the pollen deposition data (above), they were able to assess how much each particular bee species contributed to the production of harvest-ready watermelon.

Dr. Kremen found that pollination surpassed 1,000 pollen grains needed to set harvestable fruit per flower in the organic near fields but not in the organic far or the conventional far field (Fig 1, part A). Furthermore, she found that this outcome could be tied to the greater diversity and abundance of bees in organic near fields, compared to the other two types of fields (Fig 1, part B).

Figure 1. The number of pollen grains deposited on watermelon stigmas (A), and the diversity (orange circles) and abundance (yellow stars) of bees (B) on farms that were classified as organic near (ON), organic far (OF), and conventional far (CF). The number of pollen grains required to set marketable fruit (1000) is noted via a red threshold line in A. (modified from Kremen et al. 2002).

Dr. Kremen also found that, even though no single species of wild bee was as effective as managed honey bees, the collective group of wild bees surpassed the effectiveness of honey bees in organic near fields (Figure 2). Interestingly, honey bees were most effective as crop pollinators in the conventional far fields  and least effective as crop pollinators in the organic near fields. This may be because few other flowers were in bloom in the conventional far fields, so that honey bees concentrated their attention on the crop at hand. In the organic near fields, a greater diversity of flowering plants likely competed for the pollination services of honey bees.

Figure 2. The cumulative contribution of native bees, compared to the contribution of honey bees to the pollination requirement of watermelons on organic near (orange circles), organic far (navy squares), and conventional far (yellow stars) farms (modified from Kremen et al. 2002).

Wild bees were able to fully satisfy the pollination requirements of a crop with an extremely high pollination requirement because broad spectrum insecticides were not used, and the foothills provided year-round and protected habitat for the bees. This story blew my mind!

Prior to that conference, I had never given wild bees much thought. They’re mostly solitary nesters, with small bodies, that only forage for a few days to a few weeks. They tend to be inefficient foragers, particularly when compared to the juggernaut of a honey bee hive. Whereas wild bees are akin to a single vendor on Etsy, honey bees seemed the unbeatable Amazon!

Dr. Kremen’s work showed the potential value that wild bees have to agriculture. And her work was published just prior to the global onset of colony collapse disorder in honey bees in 2006. It set off a worldwide discussion about what to do about honey bee losses. Should scientists put time and effort into saving a single, non-native species (the honey bee), or should we work to conserve or build habitat around farm fields while also reducing insecticide use?

I was incredibly hopeful that the simultaneous threat to honey bees and promise of wild bees might promote heavier investments in agroecology, including the conservation of bee-friendly habitat around farms. During this time period, I was also in the early stages of documenting wild bee biodiversity in community and residential gardens, and I was surprised that abundance and diversity of garden bees was much higher than I had anticipated.

Back in 2004, I started to see gardens, and the abundance and diversity of wild bees that they host, as a potential solution to the problem of colony collapse disorder. Although I continue to be fascinated by the potential role of home and community gardens as a safe haven for bees from agricultural stresses, the urgency of this question has faded. Colony collapse disorder does not currently plague honey bees, due in large part to federal investments in studying, understanding, and mediating the factors that contribute to failing hives. With honey bees doing much better, attention has somewhat faded on the potential role of wild bees as crop pollinators. Still, work in this area continues and may rise to renewed importance, should colony collapse disorder again present a major challenge to United States agriculture.

Wild bees, including this leaf-cutter bee in the genus Stelis are also potentially important crop pollinators. However, many farm practices, such as regular insecticide sprays and mono-cultural cropping systems, make farms inhospitable to wild bees. Photo by Isabella Messer.

Kremen, Williams, Thorp. 2002. Crop pollination from native bees at risk from agricultural intensification. Proc. Natl. Acad. Sci. 99: 16812-26816. DOI: 10.1073/pnas.262413599.

2019 Native Plant Field Season Update

I’m thrilled to announce that this summer I completed the third field season of my study. This is slightly bittersweet – while I’m excited that we are done with hot fieldwork, I will miss chasing bees around the farm and the view of Mt. Hood. I’m incredibly thankful for this third season of data, as it will help account for some of the temporal variation inherent in ecological studies. In fact, pollinator communities in particular tend to be highly variable both within and across field seasons. Having three seasons of data will hopefully allow us to identify more reliable patterns of pollinator visitation between my study plants.

Lots of lab work remains, as I’m tackling the insect samples that we collected with the bee vacuum. With the help of a dissecting scope, I’m attempting to identify the each specimen to at least the taxonomic level of family to get a sense of the broader insect communities associated with each flower species in my study. It will be several months before I can share this species-richness data, but in the meantime I have bee abundance data to share with you!

Aaron and Lucas in the native plant study site, in 2017. You can see the 1m by 1 m plot in the foreground by Aaron, a second one near Lucas, and a few more in the distance.

As a refresher, we performed timed pollinator observations at each plot. This consisted of observing each blooming plot for five minutes and counting all the insects that landed on open flowers. Bees were sorted to “morpho-type” (honey bee, bumblebee, green bee, and other native bee). Though this doesn’t give us species-level information on the floral visitors, it allows us to understand which plants attracted the most pollinators overall, and allows us to detect any patterns of visitation between honey bees, bumblebees, and solitary native bees. Below is a summary of some of the highlights.

2019 overall bee abundance by plant species:

  • Origanum vulgare, Lavendula intermedia, and Eschscolzia californica were top five bee plants in 2019, just as they were in 2018.
  • In 2019, Phacelia heterophylla and Solidago canadensis jump into the top five, while Nepeta cataria and Gilia capitata fall out of the top five. It should be noted that Nepeta was the sixth most attractive plant, with about the same visitation level as Solidago.
  • Again, similar to 2018, it appears that honey bee visitation was driving the high visitation rates of the popular exotic garden species (marked with a red asterisk), while native wildflowers were being visited more frequently by native bees.
  • I’ve included the 2017 and 2018 overall abundance graphs as well, for comparison. You can see that the overall abundance was higher in 2019 for the two most popular plants, at about ~25 bees per observation period!

2017 overall bee abundance by plant species:

2018 overall bee abundance by plant species:

Since honey bee visitation drove the high abundance of many of the top pollinator plants, I took honey bee visits out of the data set and made a new graph, to compare which plants were most attractive to native bees.

2019 native bee abundance by plant species:

As you can see above, honey bees are excluded from the analysis, the top five most popular plant species completely reshuffles.

I’ve included that 2017 and 2018 native bee abundance data below for comparison.

2017 native bee abundance by plant species:

2018 native bee abundance by plant species:

Please stay tuned for more updates on the bee species richness we collected in 2019, as well as data on the other insects (pests and natural enemies) that we collected!

Unpopular Opinion: Saving Honey Bees Does Very Little to Save the Bees

Although I have been studying garden bees for the past three years, I was never focused on honey bees. From a biodiversity point of view, they are not very interesting to me. They are non-native and abundant. In fact, honey bees were the most abundant bee species that we collected in Portland-area gardens (332 individuals collected), even though we took great care not to collect more than one individual per visit, when hand-collecting.

Some of the 300+ individual honey bees that we collected from Portland area gardens, even though we took great care to not hand collect more than a single individual honey bee per garden, per site visit.

Honey bees, which hail from Europe, are only one of 20,000 bee species, worldwide. In North America, there are 4,000 species of bee. In Oregon, we have between 400-500 species of bee. From Portland area gardens, we have documented 86 species of bee (with our 2019 bees still awaiting identification).

Unlike some native bees, honey bees are not at risk of extinction. Compare this to bumblebees. We found 17 species of bumblebee in Portland gardens, two of which (12%) are at risk of endangerment or extinction, due to declining populations: Bombus fervidus and Bombus caliginosus. Across North America, more than 25% of bumblebee species are thought to be at risk of extinction.

By focusing ‘save the bee’ campaigns on honey bees, we may be neglecting the bee species that really need our help. In fact, researchers have started to call out organizations and advertising campaigns that promote feel good stories about honey bee conservation as a form of ‘bee washing’. You can visit to learn more about companies that promote their product or organization as being bee-friendly, in a less than genuine way.

Researchers have documented at least seven different ways that honey bees may harm native bee species (summarized in Cane and Tepedino, 2016):

  1. Honey bees monopolize and deplete nectar and pollen from local plant communities, which can reduce native bee reproduction.
  2. By depleting local plant resources, native bee females have to devote more time and energy to fly and find new resources, which also reduces native bee reproduction.
  3. Unlike honey bees, most bees are solitary, which means that they do not live in colonies and they do not have a queen. Solitary females who have access to fewer floral resources produce fewer daughters and more sons. Since female bees are needed to maintain a population, this skewed sex ratio can slow population growth and recovery in native bees.
  4. When females collect less nectar and pollen, they have less food to feed their young. These bees grow up to be smaller, and are more likely to die over winter, compared to well-fed bees.
  5. The longer a solitary bee mom is away from her nest, the higher risk that parasites and predators will attack her unguarded young.
  6. Honey bees can physically block native, solitary bees from preferred pollen hosts.
  7. Honey bees have many diseases. Some honey bee viruses have been found in native bee communities. Researchers think flowers that are visited by both native bees and honey bees are analogous to an elementary school water fountain: a place where repeat visitors can pick up a pathogen.

Please note that I am not suggesting that you extinguish honey bees from your garden. What I am asking, instead, is that you take the time to learn about and to notice some of the other 80+ species of bee that you might find in your garden. My group is creating a ‘Bees of Portland Gardens’ guide that we hope can help you in this journey. In the meantime, there are some great guides that are currently available. One is Wilson and Carrill’s ‘The Bees in Your Backyard: a guide to North America’s bees’. This book is available at Powell’s City of Books, as well as on Amazon. The second is August Jackson’s ‘The Bees of the Willamette Valley: a comprehensive guide to genera’. This free guide can be found online.

The first step to saving something you love is to be able to recognize it and to call it by name.


  • Cane and Tepedino. 2016. Gauging the effect of honey bee pollen collection on native bee communities. Conservation Letters 10: 205-210.
  • Jackson. 2019. The Bees of the Willamette Valley: A Comprehensive Guide to Genera. Self-Published, Online:
  • Wilson and Carrill. 2016. Bees in Your Backyard: A Guide to North America’s Bees. Princeton University Press.