Grooming is an important varroa-resistance trait, and new research might make it easier to select in breeding programs
According to the Centers for Disease Control, every year over 400,000 people are diagnosed with Lyme disease — a serious illness, caused by bacteria transmitted by ticks. Tick season is in full swing, and there are few things more disconcerting than idly touching the back of your neck and finding a swollen bump that shouldn’t be there. If you’re like me, your first instinct would be to remove it as swiftly as possible.
Some honey bees have that instinct too. When parasitized by varroa — the bee equivalent of a giant tick — workers will flail their legs, trying to scrape the mite off their bodies. Workers will also help each other with those hard-to-reach places, biting and clawing at the mites until they are injured, killed, or fall off the bee. Collectively, these behaviors are known as grooming, which is an important defense against varroa (Figure 1).1
In a breeding program originally spearheaded by Dr. Greg Hunt, professor emeritus at Purdue University in Indiana, scientists have been selectively breeding for varroa resistance, including grooming behavior, for almost a quarter century.2 In parallel, researchers in Dr. Ernesto Guzman-Novoa’s laboratory at the University of Guelph, Ontario, have been investigating the genetic link between grooming behavior and mite population growth using bees from their own selective breeding program.3,4
Now, Dr. Brock Harpur — an assistant professor at Purdue — and Dr. Nuria Morfin — a research associate at Guelph — are collaborating to capitalize on these existing breeding programs to try to understand what makes some bees good groomers. “What we’re trying to do is use the previous work as a leaping-off point to generate a holistic, mechanistic understanding of how grooming and other immune-related traits come about,” says Morfin. Morfin, Harpur and Guzman-Novoa expect that their work will also help make selective breeding programs for grooming and varroa resistance more feasible.
The Purdue breeding program began in 1997, when colonies with low mite population growth rates were selected as queen and drone sources, then mated using a combination of artificial insemination and semi-closed mating. Krispn Given, an apiculture specialist at Purdue, joined the program in 2004, quickly becoming instrumental to its success. Starting in 2007, he and the Purdue team started selecting colonies primarily for grooming behavior — a breeding program still ongoing today.
This breeding program, managed for years by Hunt and Given, has become an unexpected legacy. According to Harpur, it all started when one day, while screening for grooming behaviour, they noticed that about 5% of the mites on the bottom boards of some of the hives had significant damage to their legs and body, so they began selecting for that trait. “Now, our bees here at Purdue have upwards of 30% of mites with damage,” Harpur says.
This lineage of bees is lovingly known as the Purdue “mite-biters” or “ankle-biters” for their tendency to amputate the mites’ legs during vigorous combat (though technically, mites have no ankles — only tibiae and tarsi). “My goal is to improve overall colony health by selecting out behavioral resistance mechanisms,” says Given. “We can achieve this by flooding our mating yards with our mite-biter drones. We select the top ‘percent chew’ colonies and cross them best to best — high mite biter crossed with high mite biter — to achieve our goal.”
When distributed to commercial beekeepers for stock evaluation, twice as many mite-biter colonies survived the winter compared to unselected stock, which translated to almost four times greater honey production.4 And in research published last year by Morfin and her colleagues, varroa population growth rates in the mite-biter colonies were just one third of that in unselected Italian bee colonies.5
These are promising outcomes, but similar to other varroa-resistance traits, the scoring method on which selection is based is prohibitively tedious for widespread adoption. “With any serious bee breeding program, you have to constantly select the phenotypic traits you are looking for or they may be lost in a relatively short time,” Given explains.
To assess grooming behavior, the researchers catch fallen mites on a sticky board placed under the brood chamber, then inspect each mite under a microscope to score the damage inflicted on its body. The assumption is that the damaged mites perished due to grooming-inflicted wounds by the workers, and the proportion of damaged to undamaged mites can be translated into a grooming intensity score. Alternatively, Morfin’s research has shown that dusting workers with flour and observing subsequent grooming intensity is also an effective way to score the behavior,6 but it is arguably just as tedious.
“If you look at the success of animal breeding over the past ten to twenty years, it has been through incorporating information from the entire genome,” says Harpur. “We need more datasets with linked genomic and phenotypic information.” Through many iterations of what’s called “genomic selection,” in which breeder sources are selected based on numerous genetic markers found throughout the genome — each contributing a small amount to a desirable trait, or phenotype — beneficial characteristics can become fixed in the population. Once established, this strategy could eventually eliminate the need for tedious work under the microscope. But the road to developing this technology is long.
For grooming, that road began in 2012, when Dr. Miguel Arechavaleta-Velasco led a study, in collaboration with the Purdue lab, looking for genetic sequences associated with grooming behavior within the Indiana mite-biters, and found twenty-seven candidate genes.7 One of those genes, neurexin, is also linked to grooming behavior in mice and, curiously, social disorders in humans, leading it to become the top candidate indicator of grooming propensity.
Fast-forward five years, when researchers in Guzman-Novoa’s laboratory investigated genes linked to grooming behavior in a different way. This time, they scored individual bees’ grooming ability using a behavioural challenge, rather than scoring mite-biting. The researchers introduced a mite on a bee, and, based on several other metrics (time to initiate grooming, number of grooming attempts, number of legs used, etc.), scored the intensity at which the bee tried to groom herself.
After scoring the bees, the researchers measured the expression levels (the degree to which a gene is actually “turned on” in the cells of an organism) of six different genes, including neurexin, in intense groomers, light groomers, and non-groomers. Of all the genes, neurexin showed the starkest difference between ….