top of page

A Pollen Foraging Behavior

A new era in breeding honey bees began in the 1950’s when artificial insemination became a practical option. Managed crop pollination was in its infancy, and for the next fifty years selection was made on honey production, temperament and disease resistance (Oxley and Oldroyd, 2010). Meanwhile, demand for pollination services grew exponentially and soon became a critical component of US agriculture. Commercial beekeeping was no longer all about the honey, leading to new ideas about which traits had the most economic value. Today, it is not all together uncommon for a healthy colony to expand normally and perform well in the springtime yet fail to store enough honey to last the winter. Is this phenomenon the result of strong demand for a more efficient pollinator?

pollen-hoarding heritability and behavior

Fifty years ago, researchers from the USDA bee lab in Logan, Utah discovered that certain behaviors of honey bee foragers were hereditary. Later research identified trait loci for resource preference and pollen load size, followed by field studies that resulted in the successful development of high and low pollen-hoarding strains. Unlike the unrestricted flow of nectar into the hive, pollen stores do not typically exceed more than a few pounds. Artificial selection can produce high pollen-hoarding strains that store six times as much pollen in just three generations, indicating high heritability of the trait. The increase in stores, however, comes at the expense of honey production. A larger proportion of foragers from high pollen-hoarding strains collect only pollen, and the size of individual pollen loads increases. The opposite is true concerning nectar, as the proportion of nectar foragers declines, as does nectar load size. These changes may not be noticeable in springtime when division of labor mechanisms take full advantage of a nectar flow. But high pollen-hoarding bees are particularly sensitive to stimuli that call for gathering more pollen, and when nectar availability declines, nectar is as good as gone. Come early July along the Wasatch Front, honey production abruptly ends in many areas. (Nye 1965, Hunt 1995, Fewell 1992, Page 1995/1995, Rueppell 2004)

value of pollen vs nectar foraging

All honey bee foragers are not created equal. Some unfertilized flowers make better use of a nectar forager, while others depend on bees that collect pollen. Artificial selection can be used to alter the foraging behavior of bees in ways that improve pollination of certain crops. For example, high pollen-hoarding strains have been shown to significantly improve cranberry pollination. Bees in the bog foraging protein transfer four times as much pollen to the next flower than a nectar forager. Likewise, the rate of pear pollination is over 50% when blooms are visited for pollen but falls to under 20% when bees gather only nectar. In the almond groves, at least ten visits by a nectar forager are required for pollination to occur, compared to as little as three by a pollen picker upper. In general, flowering crops that offer both pollen and nectar rewards are more efficiently pollinated by pollen foragers. (Cane 2001, Monzon 2004, Bosch 1994)

Honey produced in the execution of a pollination contract remains the property of the beekeeper. It serves as a wild card in negotiating an agreement, because its value depends on the crop being pollinated. Many crops produce a highly marketable surplus for the beekeeper, while the grower benefits from a reciprocal reduction in rental rates. Pollination fees are about 50% less for crops that yield a honey harvest. (Sumner 2006, Rucker 2012)

who cares?

The market for commercial beekeepers was worth 830 million dollars in 2016. Honey production accounted for 343 million, crop pollination 338 million and sales of bees, wax and other hive products 149 million (National Agricultural Statistics Service). On a national scale, beekeepers earned as much revenue from pollination contracts as they did from honey production. The two sectors engage in a mutually exclusive relationship, because pollination happens when bees produce honey and honey is the byproduct of pollination. The customer base, however, shares no such association. The honey market is wide and varied, largely unorganized and exists in fragments. In contrast, 85% of the pollination market comes from almond growers, represented by a single organization called the Almond Board of California. Their contribution to beekeeping revenue in 2016 was 272 million bones! That means one out of every three dollars came from almond growers. It’s reasonable to assume their influence on queen production is considerably more than marginal.

nectar of the almond gods

Because almonds are produced at the mercy of insect pollinators, the flowers have evolved to be highly rewarding. Almond trees produce more nectar than any other member of the prunus genus. It is secreted throughout the day and continues to flow even after the flower has been fertilized and petals have dropped to the orchard floor. The relatively high sugar concentration makes the blooms especially attractive to honey bees. For us human types, however, not so much. The bitter and unpleasant flavor makes almond honey barely marketable. Hive density during the bloom spreads the nectar too thin for producing a surplus, anyway. Worse than worthless, honey production represents an opportunity cost for almond growers and can deplete the trees of critical carbohydrate reserves required for the nut’s development. These were, perhaps, minor issues when almond pollination was underpriced. (Thake 1999, Connell 2000, Traynor 2018)

course changes direction

Even before the coining of Colony Collapse Disorder, market forces in 2004 had driven the per-hive fee for almond pollination to an all-time high of $55. A year later it jumped to $73, putting growers on notice that a market correction could soon impact the bottom line. Another year passed, and in 2006, the bottom dropped out in the wake of large-scale colony losses that threatened to devastate the nation’s supply of managed pollinators. Practically overnight, fees cut through the stratosphere to nearly $140. A multifaceted response by the almond industry included further collaboration with commercial beekeepers in the formation of non-profit group, Project Apis m (PAm). Their mission: “To fund and direct research to enhance the health and vitality of honey bee colonies while improving crop pollination.” And it sounds perfectly lovely. The notion that honey bees can somehow have their health enhanced while simultaneously being exploited is not grounded in the reality thus far. Research funded and directed by PAm has included the genetic evaluation of honey bee strains for identification of superior lines of pollinators and the use of artificial brood pheromone to increase pollen foraging ( Whatever it was beforehand, the arrival of CCD solidified demand for a better pollinator, and it seems the almond industry was quick to get a hand on the wheel to help steer the future course of honey bee breeding.

other options

The interest in artificial brood pheromones represents an additional channel for improving crop pollination. Foraging behavior is subject to environmental inputs at both the individual and colony level that can have an even larger influence on trait expression than genotype. This makes the honey bee highly vulnerable to behavioral manipulation. Foraging is influenced by food storage levels, availability of nectar and pollen, amount of brood and other environmental factors that impact division of labor. Not only can a populous colony bred for pollen collection produce surplus honey during a nectar flow, unselected strains can be made to collect more pollen. Techniques include supplementary feeding, pollen trapping, artificial brood pheromone and requeening. Each scheme has its own drawbacks and may not ultimately improve pollination of the target crop. Even when it does, it’s probably not enough to maximize almond pollination. There’s no crop thinning in almonds, and anything less than 100% fruit set amounts to lost revenue. Growers literally want every grain of pollen stripped from the trees daily, and that is likely to require multiple avenues of influence. (Rueppell 2004, Duff 1986, Goodwin 1997, Free 1960, McGregor 1976)

what say the breeders?

It wouldn’t take much to convince friends in the queen and package industry to reproduce the pollen-hoarding trait. Queen producers maintain tens of thousands of hives that also participate in almond pollination. The crop blooms in February, which tasks beekeepers with the considerable challenge of a midwinter buildup. Both growers and beekeepers have much to gain. Pollen equals brood production, and brood production creates stimulus for further brood rearing. The entire pace of colony development picks up, provided pollen continues to flow into the hive (Sagili 2009). In this way, the goals of breeders and almond growers are in alignment. Keep in mind, however, that revenue from crop pollination and honey production are about equal on a national scale. Queen producers are faced with the challenge of promoting their product to equal but opposing markets.

Strachan Apiaries, world’s largest producer of Carniolan honey bees, solves this problem by saying very little. And that’s exactly what one can learn about their queens after visiting the company website. For the most part, however, queen producers seem to openly admit selection for the pollen-hoarding trait by advertising bees that build up rapidly in spring, a very common descriptor. Almond grower Koehnen & Sons describes the packages and queens they sell in this way:

They have many fine qualities: the queens are very prolific, known for their rapid Spring buildup, and they are extremely gentle. They are proven to be very hygienic.

That’s a bit of a riddle, because they really haven’t said what traits are being selected. They’ve merely made a statement about their own observations and perceived reputation. Concerning honey production, readers are left to draw their own conclusions, and that is practically guaranteed to result in misinterpretation that favors the vendor. Wildflower Meadows goes a step further by suggesting their bees are bred for both buildup and honey production:

From beginner to expert, beekeepers rely on Wildflower Meadows for gentle, mite-resistant stock that builds up rapidly for early season pollination and honey production.

colony development vs honey production

Recall that a negative correlation exists between pollen foraging (colony development) and nectar forging (honey production). However, it is also true that colony development translates to population increase, and larger populations have higher honey production. Honey bee genetics is a trip down the rabbit hole, if ever there was one, and an important disclaimer here is my entry level understanding of it all. That said, variability in honey production caused by a difference in colony population is an example of an environmental effect, and this result can be expected without breeding specifically for honey production. Selection made on observed behavior (phenotype) does not necessarily translate to genetic improvement of that trait (genotype), because environmental effects are not hereditary. In other words, an increase in honey production cannot serve as a measure of progress in breeding bees for increased honey production. There is too much environmental variation to draw such a conclusion. Milne Jr (1985) found that laboratory testing, rather than field observations, was necessary for identifying a genetically superior honey producer. Bar-Cohen (1978) found that “number of brood cells could not serve as criterion for selection for improvement of honey production.” Bielefeld (1991) concluded that, “Spring development does not show much genetic relationship to honey production.” Few queen producers explicitly claim selection for honey production, relying instead on reader interpretation of carefully chosen words. If breeding for bees that make more honey is not the objective, is breeding for a better pollinator the reason?

what say the numbers?

The new Cost of Pollination report (USDA National Agricultural Statistics Service) has only been published the last two years, so the data set is small. The average fee per hive in 2017 rose 2% to $171, barely covering inflation. The average cost to pollinate one acre of almonds fell 5%, indicating a decline in hive density. The nation’s hive inventory didn’t decline, so presumably efficiency improved. The reason for that could be Independence, a second-generation self-fertile almond tree that has performed well since its 2008 introduction. It could also be the result of hive management strategies, variations in colony strength or selective breeding. It is notable, however, that honey production declined 9% over the same period, and has been gradually declining in the twenty years since the first strain of high pollen-hoarding bees was produced. Overall, trends moved exactly as would be expected if bees were being bred for pollen specialization. Unfortunately, days before release of the 2018 data, the USDA decided to discontinue the report.

the verdict

In addition to common references to spring buildup, oft repeated claims made by queen producers include selection for gentleness and mite resistance. Inbreeding results in gentle bees without the need for selection. As for Varroa, have a look around. Does anybody really believe that any serious effort is being made to develop mite resistance? Commercial operators are treating four to six times per year, or more, yet it takes less mites to bring down a colony now than it did ten years ago. If they are selecting for mite resistance, they should stop. It isn’t working.

Nearly 75% of the nation’s commercial beehives are needed to pollinate California’s almond groves, which cover more than a million acres. The goal is to strip all pollen every day from every flower on every tree in an area larger than Grand Canyon National Park. One thing is clear: They aren’t selecting for honey production. Growers have a hard-enough time keeping bees off the sauce. There’s no way they can allow their orchards to be saturated with nectar hungry bees at bloom time. It’s the same conclusion that can be drawn from the statements made by queen producers: selection is for pollen bees, not honey bees. And as Tammy Olivarez graciously reminds us, “whatever we’re sending out is going to affect the whole beekeeping industry” (Marin Independent Journal 2013). Olivarez Honey Bees is partially owned and was built under the leadership of almond grower Dan Cummins, who spent nearly a decade as CFO before his 2014 departure.

The overall strategy seems to be selection for the pollen-hoarding trait to facilitate buildup for almond pollination. Beekeepers then have populous colonies heading into spring for selling packages, honey production, making increase or providing further pollination services. The same performance, however, should not be expected from daughter queens. Inbreeding causes a substantial loss of vigor in the next generation (Harbo 1980). Production queens are an end-product. The beekeeper must re-queen annually with another queen from the breeder to maintain the desired traits. The young queen then joins strategically placed hives in areas of warmer winter weather ahead of the next almond bloom. This also eliminates the need to select for improved overwintering. Those with colonies under a foot of snow and clinging to life in February can buy new bees when they come out of almonds.

missed behavior

We are not a population that easily lets pass an opportunity to shoot ourselves in the foot. Rather than an independent, selectable trait, it was discovered that pollen hoarding is more accurately defined as a behavioral syndrome, which refers to a framework of multiple, overlapping behaviors and non-behaviors that intersect across common gene groups. The implication is that independent evolution of foraging behavior is prevented by fundamental processes that are the basis of this Pollen Hoarding Syndrome (PHS). Selection for pollen-hoarding has far reaching physiological and developmental effects, many of which have no obvious connection to hoarding pollen. The PHS causes bees to make an early transition to forager, which reduces life span. Premature foraging also happens to be the driving force behind the rapid depopulation of a collapsing colony. Isn't it reasonable to assume there is a risk associated with substantially increasing the proportion of bees in a colony that have a reduced life span? Bees from high pollen-hoarding strains also require more time to pupate. The longer pupation of drones is the reason mites prefer to invade drone cells. It allows time for more mite offspring to mature. A logical conclusion is that selection for pollen hoarding increases mite loads.

Other effects include higher sensitivity to certain stimuli, smaller body size and larger ovaries. Brain chemistry also changes. Vitellogenin and juvenile hormone are processed differently in selected strains, which is certain to impact social structure and division of labor in more ways than intended. But why let unknown consequences prevent unbridled tinkering with the honey bee genome? What are the implications if hygienic behavior is found to be a component of the PHS, as it has in Africanized honey bees? Is there a negative correlation between mite resistance and pollen hoarding? Will PAm help find the answer? Optimization of an individual trait in honey bees does not appear to be an option. However, marker-assisted selection (MAS) and other technological advancements in breeding techniques promise to yield more reliable results and reduce unintended consequences. Let us hope so. The current body of research strongly implies that whatever causes colonies to collapse is getting plenty of help from the careless execution of artificial selection. This is not surprising, as selective breeding has long been identified as a suspected contributor to the plight of honey bees. (Page 2012, Rueppell 2014)

bitter end

The entire landscape of commercial beekeeping was summed up nicely by Dr Jeffrey Pettis, former research leader at the USDA Bee Research Laboratory in Beltsville, Md at a congressional hearing in 2014: “We have looked at honey bees as something kind of mystical... They are livestock. And they have one mission and that is pollination...” How discouraging that such a perspective is held in a place of power over the honey bee. He was right the first time. She is mystical, revered through the millennia, she has transcended time, religion and culture so that all the world’s many civilizations might know what it means to live peacefully and productively together, selflessly, for the greater good of something larger than ourselves. Now she sits among cows and pigs as livestock. The future of almond pollination is self-fertility, enhanced by a handful of mason bees. At that point, the honey bee will be fully discarded and permanently broken. Dinosaurs, meteorites and colliding continents will have failed to do what took man a mere century or two. If there is any hope at all, it lies with those who keep bees close to nature, unmotivated by financial gain. For now, her future is in the hands of profit, in the name of business and enterprise, her health most unfortunately held hostage by a bunch of nuts in California.


The genetic architecture of honey bee breeding. P Oxley and B Oldroyd, 2010

Preliminary report on selection and breeding of honeybees for alfalfa pollen collection. WP Nye, 1965

Major quantitative trait loci affecting honey bee foraging behavior. GJ Hunt, 1995.

Colony state and regulation of pollen foraging in the honey bee. JH Fewell ML Winston, 1992

The effects of colony-level selection on the social organization of honey bee colonies. RE Page, 1995

Genetic determinants of honey bee foraging behavior. RE Page, 1995

Pleiotropy, epistasis and new QTL: the genetic architecture of honey bee foraging behavior. O Rueppell, 2004

Do honey bee colonies selected for pollen‐hoarding field better pollinators of cranberry? JH Cane, 2001

Foraging behavior and pollinating effectiveness of Osmia cornuta and Apis mellifera on “Comice” pear. VH Monzon, 2004

Foraging behavior and pollinating efficiency of Osmia cornuta and Apis mellifera on almond. J Bosch, 1994

Honey bee pollination markets and the internalization of reciprocal benefits. RR Rucker W Thurman M Burgett, 2012

Bee-conomics and the leap in pollination fees. DA Sumner H Boriss, 2006

USDA National Agricultural Statistics Service – Honey Report, 2017

Almond blossoms and their avian nectar feeders. MA Thake, 1999

Pollination of almonds: practices and problems. JH Connell, 2000

Don’t Over-Pollinate Your Almonds. J Traynor, 2018

Pollen trapping honey bee colonies in Minnesota. SR Duff B Furgala, 1986

Feeding sugar syrup to honey bee colonies to improve pollination: a review. RM Goodwin, 1997

USDA Agricultural Handbook 496; Insect Pollination of Cultivated Crop Plants. SE McGregor, 1976

The need for using laboratory tests in breeding honeybees for improved honey production. CP Milne Jr, 1985

Progeny testing and selecting Italian queens for brood area and honey production. Bar-Cohen G Alpern R Bar-Anan, 1978

Genetic Correlations Among Several Colony Characters in the Honey Bee. K Bienefeld F Pirchner, 1991

USDA National Agricultural Statistics Service – Cost of Pollination Report, 2017

Orland honeybee company a hive of activity, Marlin Independent Journal. November 17, 2013

Breeding and Genetics of Honey Bees. J Harbo E Rinderer, 1980

Complex pleiotropy characterizes the pollen hoarding syndrome in honey bees. RE Page MK Fondrk O Rueppell, 2012

The architecture of the pollen hoarding syndrome in honey bees. O Rueppell, 2014

bottom of page