1. Introduction
The mite
Varroa destructor (Acari: Varroidae) [
1], is the main health problem for the beekeeping industry worldwide [
2]. This is because the mite has had a relatively recent association with
Apis mellifera, the western honey bee, and has rapidly spread to almost all regions where western honey bee colonies are managed [
3,
4]. The mite feeds on the hemolymph and fat tissue of the brood and adult bees, inhibiting their immune system, making them more susceptible to bacteria, viruses, fungi, and pesticides [
5,
6,
7,
8,
9]. Additionally,
V. destructor transmits several viruses to its hosts [
10,
11,
12]. It also shortens the lifespan of parasitized bees [
13], reducing the populations and honey yields of their colonies [
14,
15,
16]. For all these reasons,
V. destructor is one of the main factors associated with the high loss of colonies worldwide [
17,
18,
19].
Besides the damage caused by
V. destructor to honey bee colonies, the application of acaricides (synthetic and organic) used for its control represents another problem because, in general, all have shown adverse effects to the bees [
20,
21,
22,
23,
24].
To control mite populations in honey bee colonies and at the same time reduce the use of acaricides, it is necessary to identify the appropriate times to apply treatments. This knowledge would help reduce the negative effects of acaricides on bee health, the selection pressure for mites to develop resistance to the active ingredients of the chemicals, the risk of contamination of hive products, and the production cost for beekeepers due to unnecessary or excessive treatment applications [
25,
26,
27,
28,
29].
Several reports on the population dynamics of
V. destructor in honey bee colonies from different countries have been published [
30,
31,
32,
33,
34], but none from Mexico. These reports are valuable for specific countries, but it is critical to take into consideration that the population dynamics of the mite varies regionally due to the seasonality of the brood period in the colonies and its effects on mite reproduction [
27,
28]. Additionally, very few studies have jointly evaluated the population levels of
V. destructor in its different life phases (phoretic and reproductive) along with recording the daily mite drop and the population and food stores of honey bee colonies.
Studies on the population dynamics of V. destructor in honey bee colonies provide information on its evolution and changes in a given region and allow for the establishment of timely control strategies. This work reports for the first time in Mexico the population dynamics of V. destructor, considering its phoretic and reproductive phases, as well as natural mite drop, in relation to the population dynamics and food stores of honey bee colonies established in a temperate semi-arid climate.
3. Results
The colonies did not differ at the beginning of the study for infestation levels in adult bees (F1,9 = 2.51, P = 0.35), brood (F1,9 = 2.42, P = 0.36), and daily fall of V. destructor (F1,9 = 3.0, P = 0.33). There were also no differences between colonies for initial bee population (F1,9 = 1.72, P = 0.41), amount of brood (F1,9 = 0.81, P = 0.53), and areas of pollen (F1,9 = 0.48, P = 0.61) and honey (F1,9 = 1.69, P = 0.42).
During the course of the study, significant differences were found between sampling periods for V. destructor infestation rates in adult bees (F4,46 = 4.08, P = 0.008) and brood (F4,46 = 11.82, P < 0.0001), as well as for the number of fallen mites (F4,46 = 7.81, P = 0.0001), amount of brood (F4,46 = 39.27, P < 0.0001), adult bee population (F4,46 = 13.33, P < 0.0001), mite population in brood (F4,46 = 10.33, P < 0.0001), mite population in adult bees (F4,46 = 5.36, P = 0.001), total mite population (F4,46 = 13.51, P < 0.0001), areas of pollen (F4,46 = 15.19, P < 0.0001), and areas of honey (F4,46 = 25.59, P < 0.0001).
In September, the colonies had the lowest infestation rates in brood and adult bees and the lowest number of estimated total mites, while in December, the colonies had the highest mite infestation rate in the brood. In May, the colonies had the highest number of mites fallen onto the adhesive sheets, and the highest estimated number of total mites in the colonies (
Figure 1).
Figure 2 shows that there was a greater population of adult bees and brood during the months of September and May, while the estimated number of
V. destructor mites on adult bees showed significant differences only between September and May. Additionally, the estimated number of mites in the brood was significantly higher in May compared to the period between the previous June and December. Moreover, a reduction in the population of
V. destructor in adult bees was observed in response to an increase in the amount of brood in the colonies, and an increase in the population of
V. destructor in adult bees was observed when the amount of brood in the colonies decreased. It was also observed that in December, there was the lowest estimated number of
V. destructor mites in the brood compared to the number of mites on adult bees.
A decrease (although not significant) in the estimated total number of
V. destructor mites (including mites in both the reproductive and phoretic phases) relative to the populations of brood and adult bees, which increased significantly, was observed from July to September. The estimated total population of
V. destructor in the colonies was lowest in September, with an average of 782±168 mites, a number that progressively and significantly increased to reach 2,715±350 mites in May. Following a similar pattern, the daily average of fallen mites showed no significant differences from July to February, with an average of 27.6±1.6 parasites, while the mean of that variable was significantly higher in May (75.3±5.6;
Figure 3).
From September and until February, there was a parallel and significant increase in the total population of
V. destructor and the honey and pollen stores in the brood chambers of the experimental colonies (
Figure 4).
Significant correlations were found between the rates of
V. destructor infestation (in adult bees and brood) and daily mite drop, total
V. destructor population, and amount of brood. The daily mite drop was related to the infestation rate in adult bees. Additionally, the amount of brood was also related to the adult bee population, while the honey areas were significantly correlated with the pollen areas of the colonies (
Table 1).
4. Discussion
The differences found in the estimated populations of
V. destructor in brood, adult bees, and total mites, as well as for the number of parasites fallen between the analyzed time periods, are likely due to changes in environmental conditions and availability of food resources for the colonies over the 10 months monitored. These factors are known to influence variations in the queen's egg-laying rate [
27,
28,
40], and consequently, the availability of brood for the reproduction of
V. destructor.
There is a close relationship between the population dynamics of the parasite and its host, which is apparently confirmed by the reduction in mite infestation rate in brood and adult bees between July and September. This appears to be due to the increase in the brood and worker bee populations observed during that period, causing a dilution effect on the mite infestation rates. The increase in the amount of brood encouraged that many mites in the phoretic phase left the adult bees to enter cells containing larvae to reproduce. Conversely, the increase in the infestation rate of
V. destructor in adult bees in December is likely a consequence of the mite reproduction between July and September and a reduction in the bee and brood populations (see
Figure 1 and
Figure 2). From September to January, the amount of brood decreased, thereby reducing the possibilities for mite reproduction and population growth. From February to May (spring blooming season), the amount of brood increased, leading to a significant multiplication of
V. destructor, probably due to greater availability of food resources for the colonies during that time of year, resulting in greater production and availability of larvae for the mite to reproduce. Although there was no significant correlation between
V. destructor population and the stores of honey and pollen,
Figure 4 shows an increase in the mite population when food stores increased. In September and October, as well as between February and May, there are many plants blooming in the study area, which favored colony development and the storage of honey and pollen in the combs. It has been reported that the amount of pollen in the colonies is related to the fertility of the mite [
41].
The results of this study show that when the amount of brood increases (between July and September), the mite infestation rate in adult bees decreases because many parasites migrate from the adult bees to the brood. Inversely, when the amount of brood decreases (September to December), the number of mites and the infestation rate in adult bees increases because there are not enough larvae for the parasites to reproduce [
34].
The estimated total mite population tended to decrease when the bee population in the colonies increased (see
Figure 3). This is possibly because before July there was little brood available for the parasite to reproduce, and this is simply a delayed effect of the population dynamics resulting in a decrease in mite population due to low reproduction during the period of little brood before July. However, from September onward, there was a progressive increase in the mite population.
The results presented in
Figure 1 and
Figure 3 show that monitoring
V. destructor populations by recording fallen mites is more reliable than determining mite infestation rates in bees, as mite fall has a dynamic pattern similar to that of the estimated total mite population. Additionally, the techniques for determining infestation levels in brood and adult bees are destructive, negatively impacting the colony population, particularly when infestation levels are determined frequently [
42]. This conclusion is reinforced by the correlation found between daily mite drop and the estimated total mite population in the colonies. Overall, the results confirm that mite fall is a relatively effective technique for determining
V. destructor infestation levels in colonies.
Contrary to what was reported by Branco et al. [
42], our results suggest that the rate of
V. destructor in adult bees and brood is an imprecise and uninformative measure for estimating the total mite population in the colonies. This is due to the migration of
V. destructor during the phoretic and reproductive phases of the mite and the fact that the population dynamics of the parasite are related to the cycle and amount of brood and adult bees in the colonies, which vary across different seasons of the year [
27,
28]. The increase of brood and adult bee population in the colonies reduces the proportion of mites relative to the bee population, while a decrease in the colony population leads to a higher mite infestation rate [
43]. Additionally, when the amount of brood decreases and the number of adult bees increases, the percentage of mite infestation in adult bees rises, as occurred during December, the time of year with the least brood area in the colonies, which caused most mites to remain in the phoretic phase, concentrating the
V. destructor population on adult bees (see
Figure 2).
Clearly, the period from November to December is optimal for applying an acaricide, as most of these chemical products do not affect mites inside capped brood cells but do affect them in the phoretic phase. Thus, during these months (at the end of the fall honey harvest), there is a window of opportunity to apply an acaricide treatment in regions with a temperate semi-arid climate of the Mexican highlands. Specifically, this is the time of year when products such as oxalic acid are most effective [
44]. Additionally, it is important to consider that acaricide treatment during this period will limit
V. destructor reproduction before the expected increase in bee population in the early part of the year (starting in February). It is also important to consider that the treatment threshold suggested for Mexico by SAGARPA [
45] is 5% infestation in adult bees and/or 10 mites fallen in 24 h, which aligns with the results of a study conducted in the central highlands region of Mexico, where it was found that infestation levels of 4.5% in adult bees do not significantly affect population parameters, food stores, and colony weight [
46]. Proper and timely application of a treatment helps reduce the negative effects of acaricides on bees and brood, as well as the selection pressure for resistant mites, the risks of contamination of hive products, the production cost for beekeepers due to unnecessary treatment applications, and prevents relaxing the selection pressure to develop mite-resistant honey bee colonies [
25,
26,
27,
28,
29].
In a study conducted with Africanized honey bee colonies in the Mexican tropics, it was found that the population of
V. destructor decreased by more than 1,000 mites over twelve months. This was attributed to the low production of fertile mites in each reproductive cycle of
V. destructor (0.7) during a time of year with high mortality of worker larvae [
47]. Conversely, in the present study conducted in a temperate semi-arid region of Mexico with colonies of bees with greater European ancestry, an increase of 1,136 mites per colony on average over ten months was observed, representing a 26% increase in the total mite population. Considering the record of mites fallen onto the adhesive boards of the hives, there was a 39% increase in final infestation levels compared to the initial levels (33±1.5 mites in July vs. 75.3±5.6 in May). The increase in mite population observed in this study is similar to that reported in honey bee colonies in Costa Rica [
32]. The population size of drone brood in the colonies and the low expression of resistance mechanisms to the mite may have been factors that partly explain the population increase of
V. destructor in this study, although many more factors not studied here could have contributed to the mite's population growth and its temporal variations.
The population dynamics of
V. destructor are variable and depend on multiple factors and their interactions [
44,
48]. The main characteristics that determine the population growth rate of
V. destructor and its pathogenicity in honey bee colonies are its reproductive ability and longevity [
44,
49], as well as the climatic conditions and nectar flow that influence the availability of brood, drones, and the generation of swarms [
27,
28,
40], management practices [
48], and the overcrowding of colonies, which can favor bee drift and robbing [
50,
51,
52]. Other factors include the genotype of the colonies [
53,
54,
55] and the expression of bee defense mechanisms that limit the reproduction and survival of the mite. The mechanisms of resistance that most restrain the population growth of
V. destructor in honey bee populations of Latin America include hygienic behavior, grooming behavior, low brood attractiveness, suppression of mite reproduction and other mite non-reproduction related mechanisms that result in low fertility and fecundity of the mite [
56,
57].
Studies have reported variable results regarding the correlations between colony population parameters and
V. destructor infestation levels, as well as correlations between infestation levels measured by different methods [
26,
30,
32,
41,
42,
47,
53]. Therefore, there is no consensus on the effect of these parameters on the population dynamics of the mite. Moreover, complex multifactorial interactions exist that make it difficult to accurately predict mite populations in colonies [
45]. The results of this study provide new insights into the population dynamics of
V. destructor in a temperate semi-arid climate region and suggest that more studies are needed to identify the limiting factors of mite reproduction, as well as the seasonality of mite population dynamics in different regions and climates. This knowledge will help establish sustainable control strategies for this harmful parasite of honey bees.