We confirm that Cuba is home to the world’s largest European honeybee population that has naturally become Varroa resistant, with an estimated 220,000 colonies being maintained for more than two decades without any form of chemical treatment19 although there was some drone trapping during the early years of the transition period This despite the presence of the mite’s K haplotype20 and the widespread occurrence of DWV19 throughout Cuba. Therefore, the Cuban honey bee population is the first major case of Varroa-resistant European bees to cover an entire country of great size (109,884 km²).2). In Europe, the proportion of Varroa resistant honeybee populations is highly variable in each country21.22, but they still consist of small, isolated populations in each country. For example, the second largest known area of European Varroa-resistant honeybees is in North Wales, UK, where 104 beekeepers have managed approximately 500 honeybee colonies over an area of 2,500 km2.2 without treatment for more than a decade23.
It has long been known that African and African honeybees south of the Sharan Varroa are resistant and that both populations cover much larger areas than Cuba, but these honeybee varieties cannot thrive in temperate regions or are rejected by beekeepers in the Northern Hemisphere. Previous studies on African/African and European honeybees4,5,6,9 they all seem to have evolved with the same resistance mechanism7 and Cuban honeybees follow this pattern, displaying high repetition behavior, high mite removal behavior, and low mite reproduction (Fig. 1, 4, Table 1).
The strongest evidence that increased replication behavior is a direct response to the presence of Varroa is the very low recurrence rates in Varroa naive colonies. This is evidenced by the summary baseline data now collected from four different Varroa-naïve (Varroa-free) honeybee populations (Australia, UK [two populations] and Hawaii [this study]) all of which yield similar results (Fig. 1). A total of 9542 worker cells from 15 colonies were studied in the four populations with a mean recurrence rate of 2.0% (+ SD 3.2). Interestingly, only two of the colonies had atypical recurrence rates of 8.5% and 10.7%, from Australia and Kauai, respectively. This may indicate an increased susceptibility in these colonies, as no clear causes were found in either colony, eg wax moth or dead pupa. The data summary in FIG. 1 indicates that even in Varroa-treated populations, workers are still able to detect mite-infested cells, but the mean consistently falls significantly lower than that found in resistant populations. That is, in uninfected working cells, recurrence rates are significantly higher in resistant populations compared to susceptible populations (Fig. 1) t4.5= 4.185, p= 0.0023 as well as for infected cells t4.5= 6.905, p= 0.00007.
The ability of Cuban honeybees to detect infected cells not only causes high levels of recaps, but also high removal rates of cells artificially infested by mites. An average removal rate of 81% is one of the highest recorded in Apis mellifera7. The 45% mean control percentage is driven by three colonies that all removed more than 75% of the controls, while the mean of the remaining seven colonies was 28%. During the mite removal studies in March 2022, the natural Varroa infestation was 23%, while in December 2021 it was only 13%. This is due to declining worker brood breeding, caused by a shortage of nectar during the annual dry season. During this time, there is an increase in hygienic behavior in the colonies24which could help explain the higher than expected removal of control cells.
The reproductive capacity of Varroa to produce viable, i.e., paired, female offspring (r ) in infected work cells in resistant colonies in South Africa4 (r= 0.9), Brazil4 (r= 0.8), Mexico18 (r= 0.73), Europe3 (r= 0.84) is comparable to the 0.87 found in Cuba (this study). in Cuba’r‘ decreases to 0.77 when considering both single and multiply infected cells. This reduction in mite reproduction, relative to susceptible colonies with values of rgreater than one is directly related to the increased ability of resistant workers to both detect and remove the infected pupa through cannibalization. This ensures that the invading mite cannot reproduce7 or reduces the fertility of mites through the recap process4. Although no significant difference was found in the reproduction of Varroa in recapped or non-recapped cells in this study, this supports the findings of two previous studies.5.9. Therefore, repetition can play a minor role in resistance. However, recapping remains the best indicator or proxy of resistance within the vast majority of honeybee populations, as it is easier, faster and requires less skill to measure recap rates than the rate of removal of mites. However, recapitulation is a highly variable property7hence, both many cells (200-300) per colony and many colonies (>10) per population should ideally be studied to help reduce variability, including in temperate countries. because cell recoating is spatially associated with affected cells11.
Despite the current focus on what happens in work cells, studies focusing on the role of recaping in drone brood are still in their infancy. Currently, data is only available from South Africa9 (Fig. 1) and now Cuba (this study). Interestingly, both studies show no significant difference in the number of repeats between infected and non-infected brood. This is because some colonies do not replicate drone brood, while some colonies do recapitulate cells, but in a non-targeted manner. While there is a significant increase in the size of the cut-back area between infected (3.1 mm) and unaffected (2.3 mm) worker cells (Fig. 3), this does not occur in drone brood, as it appears that exploring the holes completely. However, the lack of removal of infected drone brood may play an important role in mite resistance (see below).
Worker cell mite infestation currently ranges between 23 and 13% in Cuba (this study), about 25 years after it was first discovered (1996). While in Mexico and Brazil worker brood contamination rates have fallen from about 20% in 1996/1999 to 4% in 2018/197. Although Varroa was first discovered much earlier in Brazil, in 197225 and the Africanized honeybees adapted to the mite and spread northward to replace the sensitive European colonies. Therefore, we predict that worker infestations in Cuba will continue to decline over the next 20 years, especially if mite removal rates continue. Accordingly, we would expect drone brood infestation rates (currently 40%) to remain high, as mites may avoid reproduction in working cells. This may be an important, but currently overlooked part of the resistance mechanism. Since an empirical model26 indicated that negative mite population growth only occurs in (resistant) African honeybee colonies when the initial drone cells are present. This is thought to be because mites also have a tenfold preference to reproduce in drone cells (which make up only 1-5% of all honeybee brood) and they quickly become overcrowded as the mite population increases. This leads to inter-mite competition for limited food and space, increasing mite mortality27, resulting in negative reproductive success for mites invading these crowded drone cells. Thus, the growth of the mite population in drone brood cells is limited by a density dependent mechanism. In Cuba, it has been observed that strong colonies typical with drone brood do not weaken during the drought season, while colonies without drone brood are weak and often die during the drought (APP personal comm).
Although Cuban beekeepers have known about their mite-resistant honeybees for 15 to 20 years, the situation in Cuba has only recently come to light16.18. The main reason for Varroa resistance in Cuba is due to the centralized decision to allow natural resistance to evolve, as was also successfully done in South Africa3, rather than getting locked into using miticides, as has happened throughout the Northern Hemisphere. The central decision by the CIAPI and the veterinary services to ‘not treat’ was greatly aided by the fact that all Cuban beekeepers are professional, registered and embedded in a strong local beekeeping community where colony movements and queen exchanges take place within each province.
There is also a large wild population and due to Cuba’s subtropical climate, queens are replaced annually in managed colonies due to almost continuous egg-laying, similar to honeybees in Hawaii. This rapid queen change speeds natural selection over honeybee populations in more temperate climates. Finally, Cuba’s 60-year ban on honeybee imports has helped isolate the country from the invasion by African bees, which has occurred in many nearby regions (e.g. Mexico, Southern US, Puerto Rico, neighboring Dominican Republic13 and Haiti (D. Macdonald, Apiary Inspector, Min. of Agi BC, Canada, pers. Comm.). Cuba has many managed European colonies coupled with many queen rearing stations. These colonies are prolific and gentle. Thus, Cuba is an excellent example of the power of natural selection in honeybees when they can adapt naturally to Varroa with minimal human intervention.