Introduction

Varroa destructor Anderson and Trueman (Acari, Mesostigmata: Varroidae) is the major threat for beekeeping worldwide (Genersch et al. 2010; Genersch 2010; Rosenkranz et al. 2010; vanEngelsdorp et al. 2011; Dietemann et al. 2013). The life cycle of V. destructor is divided into a phoretic phase on adult honey bees and a reproductive phase inside the sealed honey bee brood cell (Rosenkranz et al. 2010). Mite reproduction is strictly limited by the duration of the capping period of the honey bee brood: Within about 12 days (worker brood) or about 15 days (drone brood), respectively, the mother mite has to lay the eggs, the offspring has to develop into adult mites and the adult male has to mate with all mature daughter mites (Dietemann et al. 2013). Approximately 70 h after the capping of the brood cell, the mother mite lays its first egg which is haploid. Due to haplodiploid sex determination the haploid egg develops into a male mite (Rehm and Ritter 1989). After the first egg is laid the mother mite lays an additional 3–5 eggs in 30-h intervals that are diploid and will develop into female mites (Ifantidis 1983, 1990; Martin 1994). During the ontogenetic development the mite offspring pass through different stages, starting with an egg-larva and followed by protonymph, protochrysalis, deutonymph and deutochrysalis to the adult stage. The whole development takes approximately 5.8 days for female mites and 6.6 days for male mites (Donzé and Guerin 1994; Ifantidis 1990; Martin 1994; Rehm and Ritter 1989). With the last molt the adult Varroa mites become sexually mature.

The mating of the mature offspring is a crucial step in mite reproduction (Ziegelmann et al. 2013a, b). Generally, the male mite mates with its sisters immediately after they have finished their adult molt. During one reproductive cycle a mother mite can—depending on the duration of the capping period—produce up to five adult female daughters which have to be mated by a single male (Ifantidis 1983, 1990). The typical mating behavior of the male mite includes several behavioral steps from walking around the adult female, mounting the dorsum and finally moving sideward to the females’ venter to transfer the spermatozoa (Ziegelmann et al. 2013a). This final behavioral step is triggered by the female sex pheromone which is exclusively elicited by attractive young females (Ziegelmann et al. 2013b) and perceived by the male by the sensory pit organ on the front leg tarsi (Häußermann et al. 2015). During multiple matings male mites transfer 30–40 spermatozoa per female mite (Alberti and Hänel 1986; Donzé et al. 1996; Ziegelmann and Rosenkranz 2014). Considering the comparatively large size of the spermatozoa (Alberti and Coons 1999), the timely production of a sufficient number of spermatozoa for all daughter mites represents a challenge to the male’s physiology.

So far, little is known about the spermiogenetic process in male mites. The male genital tract consists of one single dense testis at both anterior-lateral sides the tubular vas deferentia originate. The vas deferentia lead to the unpaired ductus ejaculatorius which has an external opening. Beyond that an unpaired accessory genital gland releases its products into the ductus ejaculatorius (Alberti and Hänel 1986). Alberti and Hänel (1986) showed that spermiogenesis of Varroa males takes place in cysts composed of somatic cells inside the male testis. The most developed spermatozoa are close-by or inside the vas deferentia. The spermatozoa have to pass the ductus ejaculatorius before they are released into the genital tract of the female mites. Unfortunately, the final step of the Varroa copulation—i.e. the transmission of sperm—has never been observed in detail. It is assumed that male mites produce a spermatophore and take it with their spermatodactyls and put it into the paired genital pores of the female mites (Rosenkranz et al. 2010). Spermatozoa get stored inside a specialized female storage organ—the spermatheca—inside the female genital tract (Alberti and Hänel 1986). The spermatozoa stored within the spermatheca have to last the whole life time of the female mite. Spermatozoa are transferred by the male mite in a roundish form that is not yet able to fertilize the egg cells. For this purpose, the spermatozoa still have to go through a maturation phase inside the female genital tract—the so-called spermatozoa capacitation (Alberti and Hänel 1986; Häußermann et al. 2016). During this process spermatozoa undergo substantial morphological changes until they are fusiform (Häußermann et al. 2016).

Importantly, male mites have to finish the mating activities before the hatching of the host bee, as they are unviable outside the sealed brood cell (Donzé et al. 1996; Ziegelmann and Rosenkranz 2014) and unmated female mites are incapable to lay fertilized female eggs (Martin et al. 1997). The reproductive success of a Varroa female therefore depends not least on the timely production of a sufficient number of spermatozoa by the male mite.

In view of the limited time for the production of the spermatozoa and their comparatively large size we here tested two hypotheses: (1) male Varroa mites produce constantly over their whole adult lifetime spermatozoa that are ready for the transmission into the female genital tract and (2) a Varroa male has at any time in the brood cell enough spermatozoa to inseminate all daughter mites. For this purpose, we focused on the time-dependent production of spermatozoa within male mites and analyzed the total number of spermatozoa that have been produced at different time points after brood cell capping.

Materials and methods

Infestation of brood cells and Varroa mite stages considered for spermatozoa counts

The mite stages were collected from Apis mellifera L. hives at the Apicultural State Institute at the University of Hohenheim. For the identification of sex and age of the Varroa mites see Dietemann et al. (2013). Briefly, adult male mites were characterized by their pear shaped body (Fig. 1c) with a size of 0.8 × 0.7 mm (Fernandez and Coineau 2006). Daughter mites differ in the intensity of sclerotization: Freshly molted female mites are characterized by their bright dorsal shield and the fine red line around their dorsum (Fig. 1b); with increasing age of the daughter mites the sclerotization becomes more and more intense (Fig. 1d) but can still be clearly distinguished from the dark reddish-brown mother mite. The size of adult female mites is 1.1 × 1.7 mm (Anderson and Trueman 2000). From the immature stages, which are not sexually active, only the male deutochrysalis was analyzed to verify whether mature spermatozoa already are present in the male’s genital tract before the final adult molt. The male deutochrysalis as the last stage before the adult molt is immobile and extends its legs forwards in a typical way (Fig. 1a), its length is 0.7 mm (Fernandez and Coineau 2006).

Fig. 1
figure 1

Varroa destructor mite stages used in this study. a immature male deutochrysalis (length 0.7 mm), b freshly molted adult female mite characterized by its bright dorsum and the fine red line around it (size 1.1 × 1.6 mm), c adult male mite (size 0.8 × 0.7 mm), d adult female mite with reddish dorsal shield (size 1.1 × 1.7 mm)

Full size image

To analyze all adult Varroa stages (i.e., male, mother mite and daughter mites) at exactly 10, 11 and 12 days after cell capping, we infested brood cells artificially with phoretic Varroa mites. We only used female mites with a dark reddish-brown coloration, as they are at least several days old and do not need further stimulation on adult bees before the start of oogenesis (Häußermann et al. 2016). Phoretic mites were sampled directly before by the powdered sugar method (Dietemann et al. 2013). Brood cells just before capping were marked on a transparency sheet and only brood cells closed within a 4-h period (‘freshly capped brood cells’) were opened with a scalpel and one Varroa mite (‘mother mite’) was introduced with fine forceps and closed again. Thereafter, the brood comb was put back to the honey bee colony until analysis of the artificially infested brood cells after 10, 11 or 12 days. At these time points we exclusively analyzed the number of spermatozoa in ‘complete’ Varroa mite families containing a certain composition of living family members (see Table 1). In total, 342 brood cells were opened to receive 80 mite families according to Table 1. The rationale behind this practice was that the spermatozoa produced by an individual male must either be stored in the male’s genital tract and/or in the genital tract of the adult female mites within the same brood cell. By adding together the numbers of non-capacitated spermatozoa from male mite, daughter mites, and mother mites we were able to determine the total number of spermatozoa that has been produced by the male mite at a distinct time point. To examine the chronological sequence of spermatozoa production, we analyzed the mites in honey bee worker brood cells at 10, 11 and 12 days after cell capping, which approximately coincide with the adult molt of the first, second, and third daughter mite.

Table 1 Minimum requirements for the composition of the analyzed Varroa destructor families on day 10, 11 and 12 after cell capping
Full size table

Dissection and spermatozoa analysis

The genital tract of female and male mites were dissected in phosphate-buffered saline (PBS, ingredients: 1.000 ml H2O, 8 g NaCl, 0.2 g KCl, 1.25 g Na2HPO4·2H2O and 0.2 g KH2PO4) under a stereo-zoom-microscope (VWR: SZT 100) with a magnification of about 30×, using DUMONT 5 tweezers. For the dissection of female mites the female ventral shields were opened and the spermatheca and rami were isolated. For the dissection of the male mites (deutochrysalis and adults) the idiosoma was carefully opened and the male genital tract including the testis, the vas deferentia and the ductus ejaculatorius were isolated. All the other body parts like intestine and legs were removed.

The male and female genital tracts were analyzed with a VWR TR 500 light microscope (magnification 100–400×). Images were taken with a CANON EOS 60 D camera. The maximum diameter of the roundish stage-0 spermatozoa inside the male genital tract was measured by analyzing the pictures with the software GIMP 2 (v.2.8.10). The genital tract of the adult male mites contains roundish ‘stage 0’ spermatozoa (Häußermann et al. 2016, Fig. 5). These stage-0 spermatozoa are transmitted during mating and can be clearly distinguished from other stages. They are exclusively found within the male’s testis, the male’s vas deferentia and in the rami of recently mated adult females (Fig. 2b–d). As stage-0 spermatozoa represent the final stage of spermatozoa production we here exclusively counted these stage-0 spermatozoa ready for transmission during mating. In daughter mites we counted all spermatozoa stages found in the rami and in the spermatheca as these daughter mites could have exclusively mated with the male (= brother) of the respective brood cell (Fig. 2e, f). In the mother mite, however, only ‘young’ spermatozoa up to stage VI (see Häußermann et al. 2016) that have not yet completed the capacitation process in the female genital tract were counted. Considering the fact that the capacitation process lasts about 5 days (Häußermann et al. 2016), these spermatozoa must have been transferred during the past 1–4 days and therefore derive from the own son as the only available male during this time period. Fully capacitated spermatozoa (= stage VII) in the spermatheca, however, must have been transferred by matings during previous reproductive cycles.

Fig. 2
figure 2

Occurrence of spermatozoa in adult male and female Varroa destructor mites. a Genital tract of male deutochrysalis, T testis, VD vas deferens, AGL Accessory gland, DE ductus ejaculatorius. No roundish stage-0 spermatozoa are visible. b Genital tract of adult male mite in the T and the VD several roundish stage-0 spermatozoa are visible (arrow). c Spermatozoa stages found in the male’s genital tract. Stage-0 spermatozoa are characterized by the roundish cell shape and the nubby cell surface (e.g. arrows). d Stage-0 spermatozoa in detail. e Spermatheca and rami of the 1st daughter mite 11 days after cell capping with several stage II spermatozoa in rami and spermatheca (arrows). f Spermatheca and rami of the 1st daughter mite 10 days after cell capping with a couple of stage II spermatozoa inside the rami (arrow). Scale bar is 100 µm

Full size image

Data analysis

All data were analyzed using the statistic software SPSS (v.22.0). All reported means are ± standard deviation (SD). Data were checked for normal distribution by the Kolmogorov–Smirnov-test and for homogeneity of variance by the Levene-test. If normal distribution and homogeneity of variances could be confirmed, we used ANOVA analysis with Bonferroni correction for the pairwise post hoc tests. If normal distribution or homogeneity of variances could be not confirmed we used ANOVA according to Kruskal–Wallis. Statistical analyses were used for comparisons of the size of the spermatozoa and for the time-dependent numbers of spermatozoa. Differences between groups with α = 0.05 were considered statistically significant.

Results

Occurrence of spermatozoa in deutochrysalis and adult Varroa males

In total we analyzed the genital tract of 12 male deutochrysalis. In none of the analyzed male genital tract the stage-0 spermatozoa that are ready for transmission during mating were present (Fig. 2a). However, the cysts in which spermiogenesis takes place were clearly visible.

In adult male mites, however, these typical roundish stage-0 spermatozoa were frequently found in the vas deferentia and in the testis (Fig. 2b). Such spermatozoa are characterized by their roundish cell shape and the nubby cell surface that reminds of a raspberry (Fig. 2c, d). The nucleus was mostly visible.

There were no significant differences in the mean diameter of these stage-0 spermatozoa singly measured in the genital tract of one to two males each at 10 (n = 16), 11 (n = 11) and 12 (n = 18) days after cell capping (Fig. 3; F2,55 = 0.637, p = 0.87, ANOVA with Bonferroni correction). Total mean (± SD) diameter (n = 58) of theses stage-0 spermatozoa was 44.9 ± 4.83 µm, with a range of 32–56 µm.

Fig. 3
figure 3

Mean (+ SD) maximum diameter of the roundish stage-0 spermatozoa in the Varroa destructor male genital tract (n, number of individual spermatozoa measured). There were no significant differences between days 10, 11 and 12 after cell capping

Full size image

Numbers of spermatozoa at different times after cell capping

In total we analyzed the genital tract of adult mites from 80 Varroa mite ‘families’ consisting of at least one mother mite, one adult daughter mite and one adult male mite at three distinct time points after cell capping (10, 11 and 12 days, respectively). In male mites we exclusively counted the roundish stage-0 spermatozoa as these spermatozoa are ready to be transferred into the genital tract of the female mites. In the daughter mites we counted all spermatozoa inside the genital tract because daughter mites must have received all spermatozoa from their brother as single male available during this time period. In mother mites we only counted non-capacitated spermatozoa (up to stage VI, Häußermann et al. 2016) as these spermatozoa must have been transferred during the past 1–4 days. Fully capacitated fusiform spermatozoa (= stage VII) must have been transferred in matings in previous reproductive cycles, and thus were not considered into account.

The mean total number of spermatozoa reveals highly significant differences between all three time points (Fig. 4, Kruskal–Wallis ANOVA, F2,77 = 51.74, p < 0.0001). Ten days after cell capping already 31 ± 16 (= mean ± SD) spermatozoa, on average, were counted. At 11 days after cell capping this number increased to 85 ± 42 spermatozoa, on average, and reached the maximum number at 12 days (shortly before the hatching of the adult bee) with an average 125 ± 38 spermatozoa (Fig. 4).

Fig. 4
figure 4

Mean total number of non-capacitated spermatozoa (stage 0–stage VI) per Varroa destructor family itemized by individual mite stages (male, daughter and mother mites) at 10, 11 and 12 days after cell capping. Different letters mark significant differences. Mean total number of non-capacitated spermatozoa per Varroa family was highly significant between day 10, 11 and 12 respectively (Kruskal–Wallis ANOVA, F2,77 = 51.74, p < 0.0001). For single mean values and respective standard deviations of each group see Online Resource S1

Full size image

The distribution of the spermatozoa among male and daughter mites changes over time with significant increase in spermatozoa counts in both, the male and the female genital tract. In the first female spermatozoa counts significantly increases over time (Kruskal–Wallis ANOVA, F2,77 = 30.798, p < 0.0001). At day 10, the time when the first daughter mite attains maturity, more spermatozoa were present in the male (18 ± 12) compared to the first daughter (13 ± 10). At day 12 the first and the second Varroa daughter stored significantly more spermatozoa 32 ± 8 and 30 ± 19 spermatozoa, respectively, in their genital tract compared to only 16 ± 10 spermatozoa in the 3rd daughter (Kruskal–Wallis ANOVA, F2,62 = 14.348, p = 0.001). At day 12 the male mite still has 53 ± 26 stage-0 spermatozoa for further matings (Online Resource S1). Only in a few cases (18.8%) also the mother mite gets mated with her son (n = 15 from a total of 80 analyzed families) with a low number of non-capacitated spermatozoa (9 ± 5) in her genital tract that has been transferred during these matings.

Discussion

Our results on the production of spermatozoa in male Varroa mites give a further example for the adaptation of the parasite to the host reproductive cycle which is predefined by the capping period of the host brood. For the female mite we could already demonstrate how oogenesis and copulation behavior are optimized in order to make sure that mated female daughters can be produced under this extreme time pressure. An immediate activation of oogenesis after the invasion of a honey bee brood cell (Cabrera Cordon et al. 2013; Frey et al. 2013; Garrido et al. 2000), a rapid preimaginal development (Steiner et al. 1994; Häußermann et al. in prep.) and mating immediately after the adult female molt (Ziegelmann et al. 2013a, b) are examples for the close synchronization of the parasite reproduction with the development of the honey bee host.

Certainly, the production of sufficient roundish stage-0 spermatozoa ready for transmission during mating is an additional major challenge for a successful Varroa reproduction. The big size of these spermatozoa with a maximum diameter of more than 50 µm in comparison to male body size of 800 × 700 µm confirms the high physiological effort involved in the spermatozoa production. Already a chain of 18 of these spermatozoa ready for transmission during mating side by side would be as long as one adult male. We could clearly show that the Varroa males continuously produce these stage-0 spermatozoa, obviously starting immediately after the adult molt. In the deutochrysalis stage the cysts in which spermiogenesis takes place are already visible, but we could not identify any stage-0 spermatozoa. Each adult male mite produces about 125 roundish stage-0 spermatozoa that are ready for transmission during mating in worker brood. Considering that only adult males can produce the respective spermatozoa, the sequence of sexes in Varroa mite offspring within the sealed brood cell (Garrido and Rosenkranz 2003) is consequential. The male egg is laid approximately 30 h before the first female egg, giving the male a timing edge for the preimaginal development (Rosenkranz et al. 2010). This temporal advance enables the production of a sufficient number of spermatozoa before the adult molt of the first female daughter.

The adult molt of the first female takes place approximately 10 days after the sealing of the brood cell and about half a day after the adult molt of the male. Immediately after the adult molt of the first female mite mating takes place. Our study indicates that at day 10—and thus shortly after the adult molt of the female mite—the freshly molted daughter mite already has received some spermatozoa while the male’s genital tract contains about 20 spermatozoa ready for transmission during mating which might be sufficient for the next matings. The relatively low number of available spermatozoa at the beginning of the mating period might be one reason for the multiple matings of the male with the different daughter mites (Donzé et al. 1996). Interestingly, freshly molted adult male mites that do not yet have produced spermatozoa ready for transmission did not reveal the typical copulation behavior of older males (Häußermann 2014). We do not know how this age-dependent mating behavior is triggered, but it ensures that the first mating will only take place when spermatozoa are available.

After the start of sexual activity, the production of spermatozoa in the adult males is an ongoing process with a consistent productivity. Already on day 11 after cell sealing, the male had doubled the stock of spermatozoa within his genital tract. At this time, the number of transferred spermatozoa in the first daughter mite had also doubled to an average of more than 25, while the second daughter has already received about 20 spermatozoa. Wendling et al. (2014) suggest that 0.99 spermatozoa are needed to fertilize one oocyte. As their data are based on a calculation it is possible that unknown parameters may have an influence on this number. However, it is likely that not many spermatozoa are needed to fertilize one oocyte. In literature, the mean number of altogether transferred spermatozoa per female mite varies between 30 and 40 spermatozoa (Alberti and Hänel 1986; Donzé and Guerin 1994; Donzé et al. 1996; Ziegelmann and Rosenkranz 2014). This should be sufficient for the insemination of all female eggs during the average number of two to three reproductive cycles per female mite (Fries and Rosenkranz 1996; Martin and Kemp 1997). On day 12 after cell sealing and thus just before hatching of the host bee, both, the first and second daughter mites have, on average, reached the required number of more than 30 spermatozoa. However, the third daughter received a lower amount of spermatozoa. It is noticeable, that in the first daughter mite the number of spermatozoa hardly increased from day 11 to 12 after cell sealing indicating that (1) the male does not transfer more spermatozoa and/or (2) the mating frequency significantly decreased in older daughter mites. The latter is in line with the age dependent production of the female sex pheromone (Ziegelmann et al. 2013b) with the consequence that daughter mites are attractive to males almost exclusively during the first 24 h after the adult molt (Ziegelmann et al. 2013a).

The fact that the third daughter receives a significant lower number of spermatozoa compared to her sister mites might limit the possible number of successful reproductive cycles. Possibly, such mites have to mate during one of the next reproductive cycles with their son to ‘fill up’ the stock of spermatozoa. This could explain the low frequency of matings between sons and mother mites in our study with a relatively low number of transferred spermatozoa. This finding is consistent with previous behavioral experiments in which a relatively low frequency and duration in mating attempts of male mites towards the mother mites was observed indicating that mother mites are less attractive compared to her daughters (Ziegelmann and Rosenkranz 2014). Additionally, mother mites move faster than the young adult daughter mites hindering male mites mounting the female’s dorsum for mating (Ziegelmann and Rosenkranz 2014). Probably, mother mites that run out of spermatozoa change their behavior supporting matings with their own son. As we did not count the number of fusiform (= capacitated) spermatozoa in the spermatheca of the mother mites, this hypothesis has to be confirmed in future studies.

The total number of about 125 roundish stage-0 spermatozoa produced by the males during a reproductive cycle in worker brood seems to be more than needed for the matings with the maximum of three daughter mites in our study. However, this overspill of about 50 spermatozoa ready for transmission during mating in the male mite could serve as a reserve for the parasitation of drone brood where the capping period is extended and more daughters are produced by the mother mite (Oldroyd 1999). Additionally, in case of brood cells that are double infested with Varroa females the male mites could not only mate with their sisters but also with the daughters from the second mother mite. Therefore, this spermatozoa surplus might also be an adaptation to such situations when sperm competition might occur (Witalinsky 1998). Further studies are needed to support this assumption. In any case the production of about 125 spermatozoa ready for transmission during mating confirms our second hypothesis that a male mite produces enough spermatozoa to inseminate all daughter mites.

A shortened duration of the male’s spermiogenesis with the production and transfer of non-capacitated spermatozoa seems to be a further factor to optimize the course of mite reproduction. So far, the details of spermiogenesis within the cysts of Varroa males are unknown and require further studies. Only Alberti and Hänel (1986) analyzed sections of the male’s genital tract and described six different stages of spermiogenesis resulting from male germ cells. For this study, we only recorded the roundish spermatozoa which are transmitted by the male into the genital tract of the female mites. These spermatozoa are still not able to fertilize the female egg cell. They are characterized by the roundish cell shape and the nubby cell surface that reminds of a raspberry and the nucleus is mostly visible. We found these spermatozoa in male mites especially in the anterior-lateral sides of the testis and in the two vas deferentia. In a previous study we detected these spermatozoa in very low amounts in the genital tract of female mites dissected immediately after mating (Häußermann et al. 2016) and defined this stage as ‘stage 0’ spermatozoa that gets transferred into the female genital tract during copulation. In the genital tract of the females these roundish spermatozoa still have to undergo the so-called spermatozoa capacitation. During this process sperm cells change their morphology substantially accompanied by an immense elongation until they appear fusiform and reach a length of more than 200 µm and finally are able to fertilize the female egg cell (Häußermann et al. 2016). Interestingly, these stage-0 spermatozoa did neither change their size nor their morphology within the female’s genital tract. However, immediately after the transfer to the female’s genital tract the capacitation process starts (Häußermann et al. 2016).

Our results also raise the question of how the spermatozoa are finally transmitted during mating. It is assumed that male mites produce a spermatophore, take it with their spermatodactyls and put it into the paired genital pores of the female mites (Alberti and Hänel 1986, Rosenkranz et al. 2010). However, the female genital pores have a size of only 7.5 µm (De Ruijter and Kaas 1983) compared to a diameter of approximately 45 µm of the spermatozoon that needs to be transferred. It is an intriguing question how the Varroa male pushes this huge spermatozoon through the small genital opening. In other gamasid mites like Hattena cometis and Veigaia sp., Di Palma et al. (2013) could show that male chelicera feature a basal opening in the movable digit that is supposed to allow the sperm shift of single sperms from the spermatophore into the lumen of the movable digit. The lumen leads through a sperm transfer tube up to the tip of the spermatodactyl. Male mites inaugurate the spermatodactyl into the female genital pores and thus transmitting the spermatozoa (Di Palma et al. 2006, 2013). We assume that spermatozoa get transmitted from male to female Varroa mites in a similar way. Perhaps male Varroa mites don’t even create spermatophores; however, this is very speculative and needs to be confirmed.

Our study clearly shows that in Varroa reproduction not only the female oogenesis and copulation behavior are optimized but also the timely production of male spermatozoa. Hence we could confirm both of our initial hypotheses, because this optimization includes (1) a rapid and continuous production of non-capacitated spermatozoa, (2) in an amount that is sufficient to inseminate all daughter mites at any time in the brood cell.