Sperm-depleted males of the two-spotted spider mite can replenish sperm in a few hours

In many animals, males increase their reproductive success by mating with as many females as possible. The number of females a male can fertilize is often limited by male competition for access to females, sperm competition, and the cost of sperm production. Especially, recent studies have shown that sperm production is more costly than previously expected. In the two-spotted spider mite, Tetranychus urticae Koch, the number of females a male can inseminate is limited mainly by male competition for access to females. However, in the absence of rivals, males mate with so many females that they can become sperm-depleted. Mating without sperm transfer does not produce any offspring, although it takes time and energy. Therefore, a question arises as to why males continue to mate even after sperm depletion. In this study, we hypothesized that males continue to mate because sperm is replenished after a short period. To test the hypothesis, we investigated how long it takes for sperm replenishment after sperm depletion. We found that in 3 h, sperm can be replenished enough to inseminate a few females. As 3 h is sufficiently short not to lose the next mating opportunity, the results support the hypothesis. However, copulation duration in the sperm-replenished males was significantly longer than in the sperm-depleted males but shorter than in males before sperm depletion. To explain the differences, further research would be necessary. In addition, anatomical physiology study in males is also required to confirm that sperm is indeed depleted and replenished.


Introduction
In many animals, males may maximize their reproductive success by fertilizing as many females as possible (e.g., Bateman 1948).The number of females a male can mate is often limited by male competition for access to females, sperm competition and the cost of sperm production (Pitnick and Markow 1994a;Rittschof et al. 2012).Especially, recent studies have shown that sperm production is more costly than previously expected, and also that males readily experience sperm depletion by multiple mating in a short period even in synspermatogenic species, in which males produce sperm throughout their adult life (Partridge and Harvey 1992;Van Voorhies 1992;Pitnick and Markow 1994b;Pitnick et al. 1995;Olsson et al. 1997;Preston et al. 2001;Rubolini et al. 2007;Radhakrishnan et al. 2009;Dowling and Simmons 2012;Macartney et al. 2021).Therefore, how long it takes for sperm replenishment after depletion would be important in determining the optimal number of matings and the best reproductive strategy in males of synspermatogenic species.
Tetranychus urticae Koch is a spider mite infesting numerous crops, and infamous as a cosmopolitan pest (Helle and Sabelis 1985).In this species, females often mate with several males, but they use the sperm only from the first mating for egg fertilization if there are no abnormalities in the first mating (Boudreaux 1963;Helle 1967;Potter and Wrensch 1978;Satoh et al. 2001).Therefore, males compete for virgin females, and it makes competition among males very intense.As a result, males often mount females in the teleiochrysalis stage, which is the quiescent stage just before molting to adult female (premating guard) (Potter et al. 1976a, b).Males also show alternative reproductive tactics; whereas many males fight rival males for teleiochrysalis females (fighters), some males do not fight, but quietly stay on teleiochrysalis females' dorsum and are not attacked by other males during premating guard (sneakers) (Sato et al. 2013(Sato et al. , 2014(Sato et al. , 2016;;Schausberger andSato 2019, 2020;Schausberger et al. 2021).Therefore, in mites, the number of females a male can fertilize is limited mainly by male competition for access to females.However, in the absence of rivals, males mate with a large number of females and experience sperm depletion.For example, a previous study reported that a 0-day-old male mated with 5-25 females (12-13 females on average) (Kobayashi et al. 2022).In the study, the proportion of daughters in offspring decreased as the mating number increased, and males did not inseminate females after around the 11th mating at all.Mating without sperm transfer does not produce any offspring, although it takes time and energy.Therefore, the question arises why males of this mite species continue to mate even after sperm depletion.
In several species of parasitoid wasps, male behavior to continue to mate after sperm depletion has been found (Simmonds 1953;Laing and Caltagirone 1969;Gordh and DeBach 1976;Nadel and Luck 1985;Ramadan et al. 1991).In Trichogramma evanescens Westwood, it has been observed that mating without sperm transfer reduces the ability of mated females to store sperm from other males (Damiens and Boivin 2006).In Anisopteromalus calandrae (Howard), a study has found that mating reduces females' subsequent mating receptivity regardless of the amount of sperm transferred (Abe 2019).Therefore, in parasitoid wasps, sperm-depleted males may continue to mate to interfere with the reproduction of rival males, so as to increase their own relative fitness.In the spider mite, mated females readily remate with males, and when the first mating is abnormal (e.g., shorter copulation than in normal mating), their eggs can be fertilized by sperm from remating (Helle 1967;Potter and Wrensch 1978;Satoh et al. 2001;Morita et al. 2020).Therefore, mating without sperm transfer may not be expected to function such as to reduce subsequent mating receptivity in females.However, remated spider mite females produce fewer fertilized eggs (Oku 2010;Macke et al. 2012;Rodrigues et al. 2020).So, mating without sperm transfer likely has a function to interfere with reproduction of females fertilized by other males, to reduce the proportion of other males' offspring in the next generation.In addition, mated females are less attractive compared to virgin females (Oku 2010;Rodrigues et al. 2017).Premating guard is common, but males sometimes mate with virgin females that have molted without being guarded by males or rejected a guarding male as mating partner.It is unclear if females mated without sperm transfer are also less attractive; however, there is a possibility that mating without sperm transfer can reduce the efficiency of other males to find virgin females.To understand why males continue to mate after sperm depleted, it would be necessary to investigate such indirect effects.However, in the first place, if sperm can be replenished very quickly and given that the premating guarding phase takes several hours, males may continue to mate even after sperm depletion in order not to lose mating opportunity.In a previous study, a T. urticae male that mated 25 times in 3 h could inseminate the 23rd female slightly, although the male failed to inseminate the 13th to the 22nd female consecutively (Kobayashi et al. 2022).Therefore, it is highly expected that sperm can be replenished in a short period in T. urticae males.
In this study, we investigated how long it takes for sperm replenishment in T. urticae males.We allowed males to rest for 24, 12, 6 or 3 h after sperm depletion, then provided virgin females to them.We determined whether sperm was replenished by checking the insemination status of mated females and by comparing the insemination status between females mated with males before or after rest and also between females mated with males after rest or virgin males.Spider mites are haplodiploid: females develop from fertilized eggs and males develop from unfertilized eggs.Hence, we used the production of female offspring in the determination of insemination status.As it is known that copulation duration without sperm transfer is much shorter than that with sperm transfer (Kobayashi et al. 2022), we also checked and compared the copulation duration.

Mites
In this study, we used the strain "Houten-1" of T. urticae obtained from the Institute for Biodiversity and Ecosystem Dynamics, University of Amsterdam, in 2014, which was also used in a previous study reporting sperm depletion in males (Kobayashi et al. 2022).We reared the mites at University of Tsukuba on detached leaves of common bean, Phaseolus vulgaris var.Naga-Uzura, placed on wet cotton wool in styrofoam and plastic trays.The colony started from more than 50 individuals and was reared at 18-25 ℃ and L15:D9 h photoperiod.Starting roughly 1 month before experiments, the colony was kept at 25 ℃ and L15:D9 h photoperiod and colony size was kept large enough for collecting more than 500 teleiochrysalis females.
To obtain virgin males, we collected teleiochrysalis males from the mite colony and placed them individually onto bean leaf discs (1.5 cm diameter) which were placed on wet cotton wool in insect breeding dishes (5 cm diameter, 1.5 cm high; SPL Life Sciences, Gyeonggi-do, Korea) (hereafter called small leaf arena).We maintained the males separately to avoid getting injury from male fights.We kept them at 25℃ and L15:D9 h photoperiod and used the males which emerged within 24 h (0-day-old males) in the experiment.To obtain virgin females, we collected teleiochrysalis females from the mite colony and placed them onto the small leaf arena (20-30 females per leaf disc).We kept them at the same conditions as the males and used the females which emerged within 24 h (0-day-old females) in the experiment.To prepare large numbers of virgin females at the same time, we stored part of the collected teleiochrysalis females in a refrigerator (8 ℃) for a few days before keeping them at 25 ℃ and L15:D9 h photoperiod and used the females which emerged within 24 h in the experiment.We mixed these virgin females with those which were not stored in a refrigerator and used them in the experiment.

Experiment
The experiment was carried out in four steps: (1) sperm depletion, (2) confirmation of sperm depletion, (3) rest, and (4) confirmation of sperm replenishment (Fig. 1).In step (1), sperm depletion, we allowed a 0-day-old virgin male to mate with up to 32 0-day-old virgin females in a 4-h period in the small leaf arena.We determined the number of females and the mating time for sperm depletion based on the previous study reporting sperm depletion (Kobayashi et al. 2022).It is known that males easily mate with mated females repeatedly.To promote mating with virgin females, we introduced eight virgin females onto the small leaf arena where a virgin male was present, and replaced the females with eight new virgin females every 1 h (in total, 32 females for 4 h).The females used in step (1) were discarded after being replaced and were not used more than once.
In step (2), confirmation of sperm depletion, we removed all females used in step (1) from the small leaf arena, then introduced three new 0-day-old virgin females onto the small leaf arena.We observed their mating behavior using microscopes until the male mated with all three females, and measured copulation duration of each mating with a stopwatch.We transferred the mated females individually onto bean leaf discs (2.5 cm diameter) which were placed on wet cotton wool in insect breeding dishes (5 cm diameter, 1.5 cm high; SPL Life Sciences, Gyeonggi-do, Korea) (hereafter called 'large leaf arena').We allowed the females to oviposit for 4 days.We removed the females from the large leaf arena, reared the offspring until they developed into adults, and recorded their gender.Because females develop from fertilized eggs and males develop from unfertilized eggs in this spider mite, we used the sex ratio of offspring (proportion female offspring) to estimate the sperm transfer status.If the 2nd and 3rd mated females did not produce any daughters, we regarded the male mate as a sperm-depleted male and used it in the analysis; note that we did not confirm sperm depletion by dissection.Data from males that did not meet this criterion for sperm depletion were discarded (17 males on a total of 68).
In step (3), rest, we allowed the males to have a rest for 24, 12, 6 or 3 h, i.e., remain alone on the small leaf arena for this time period.Because sperm depletion was observed in a 3-h mating trial in a previous study, 3 h was used as the time unit and rest periods were set in multiples of 3, with 3 h as the minimum rest period.In step (4), confirmation of sperm replenishment, we introduced three 0-day-old virgin females onto the small leaf arena with the male, and observed their mating behavior using microscopes until the male mated with all three females, and measured copulation duration of each mating with a stopwatch.We checked the gender of their offspring in the same way as described for step (2).Some mated females died before completing 4-day oviposition.In that case, we discarded the offspring of that female and used the sex ratio of offspring produced by the other female(s).
As controls, we prepared males who had gone through steps ( 1), ( 2) and (3) in absence of any female mates (hence, no sperm depletion), that were then, as virgin males, subjected to step (4).We carried out all manipulations at 25 ℃ and L15:D9 h photoperiod in climate chambers and a laboratory.

Statistical analysis
To determine whether sperm was replenished after 3, 6, 12 and 24 h rest, we compared offspring sex ratio produced by females mated with males before vs. after rest.In the comparison, we used a generalized linear mixed model (GLMM), in which the dependent variable was offspring sex ratio in each mated female, and explanatory variables were timing of mating [before/after rest, i.e., step (2) or step (4)], rest time (3 / 6 / 12 / 24 h) and their interaction.We incorporated male individual as a random effect into the model because the same males were used in the matings before and after rest and one male mated with three females in each confirmation of sperm depletion (step 2) and confirmation of sperm replenishment (step 4).We applied beta-binomial as the error distribution because we detected an overdispersion problem when we constructed the GLMM with binomial error distribution.We also analyzed the copulation duration for males before and after rest by using a GLMM in which the dependent variable was copulation duration, using the same fixed and random effects as in the GLMM for offspring sex ratio.We applied gamma (link: log) as the error distribution because Shapiro-Wilk normality test showed the distribution was significantly different from a normal distribution (W = 0.966, P < 0.001).
To determine whether sperm-replenished males with replenished sperm can inseminate females as well as virgin males, we compared the offspring sex ratio of males after sperm replenishment [i.e., from matings in step (4)] with that of males in the control treatment [i.e., still virgin when mating in step (4)].In this analysis, we used a GLMM, in which the dependent variable was offspring sex ratio from mated females and explanatory variables were treatment (sperm replenishment / control), rest time (3 / 6 / 12 / 24 h) and their interaction.We again incorporated male individual as a random effect because each male mated with three females.We applied beta-binomial as the error distribution because again we detected an overdispersion problem when we constructed the GLMM with binomial error distribution.We analyzed copulation duration of males in the sperm replenishment treatment and males in the control treatment by using a GLMM in which the dependent variable was copulation duration, using the same fixed and random effects as in the GLMM for offspring sex ratio.We applied gamma (link: log) for the same reason with the GLMM for the comparison between copulation duration before and after rest in the treatment.
We performed these comparisons separately using different models, because the structure of the random effect was different between them.Specifically, the same males were used in the comparison before and after rest for the sperm replenishment effect, and different males were used in the comparison for insemination ability between treatment and control.In each model, we tested the effect of explanatory variables by likelihood ratio tests (LRTs) for GLMMs.If the effect of interaction was not significant, we reconstructed the models without the interaction, and used the reconstructed models to test the effect of each explanatory variable.For the statistical analyses, we used the software R v.4.0.2 (R Core Team 2022) and the packages glmmTMB (Brooks et al. 2017).
In the comparison of copulation duration, the interaction between sperm replenishment treatment and rest time did not have a significant effect (gamma-GLMM: LRT = 0.006, P = 0.94).Copulation duration was longer in males after rest than before rest (gamma-GLMM: LRT = 101.97,P < 0.001; Fig. 3), but was not significantly different among different rest times (gamma-GLMM: LRT = 0.503, P = 0.48; Fig. 3).

Fig. 2
Sex ratio in offspring from males before and after rest for 3, 6, 12 or 24 h after sperm depletion (before, after) and in males who did not experience sperm depletion (control).The upper and lower boxes in the boxplots indicate the third and first quartiles, respectively, the line in between the boxes indicates the median, the whiskers indicate 1.5× the interquartile range values, and the dots are outliers.The boxplots were created by using the mean sex ratio in offspring from each male, as each male mated with three females.The numbers of males are shown above each boxplot

Discussion
We found that 3 h rest was sufficient for enough sperm replenishment to fertilize a few virgin females.In T. urticae, males usually mount teleiochrysalis females as premating guard (Potter et al. 1976a, b), even if rival males are absent and await female emergence on the female dorsum (Sato et al. 2014).Teleiochrysalis females usually take 1-2 days to molt into the adult phase (Laing 1969).Although males would look for teleiochrysalis females which likely emerge relatively soon for their premating guard, they likely mount for about 3 h or more until female emergence.Their ability to replenish sperm in 3 h is thought to be sufficient to continue their reproductive activity even though they experience temporary sperm depletion.Therefore, we conclude that the hypothesis that males continue to mate because sperm can be replenished in a very short period is supported by this study.3 Mean (± SE) copulation duration (s) in males before and after rest for 3, 6, 12 or 24 h after sperm depletion (before, after) and in males who did not experience sperm depletion (control).The mean values were calculated from the mean values for each male, as each male mated with three females.The numbers of males are shown above each bar How long it takes for sperm replenishment has not been well studied in invertebrates.In the amphipod Gammarus pulex, a single mating reduces the number of sperm that a male reserves substantially, but sperm is fully replenished within 6 days (Lemaître et al. 2009).In the moth Parasemia plantaginis, a single mating reduces the number of sperm that a male contains by around half, but after 10 h sperm is restored to the same level as that of virgin males (Chargé et al. 2016).Our findings cannot be compared directly with the previous studies because these did not make males experience complete sperm depletion and our study did not count the sperm that a male contains.Also, our experimental setup contains a possible bias: not all males (25%) were sperm depleted (or not anymore) after the first phase of the experiment, and their data were subsequently discarded.Then a bias could occur if the measured traits have a relation with the vigor of the males, i.e., with how many matings are required before the sperm is depleted.However, the results do suggest that among invertebrates T. urticae has a high ability to replenish sperm fast.Various factors may account for differences in ability to replenish sperm in a short period among taxa, such as genital structure and physiological traits but also ecological traits, because sperm production ability and characteristics vary depending on characteristics of sperm competition and mating system (e.g., Aron et al. 2015;Michalik and Rittschof., 2011).For example, T. urticae forms colonies on host leaves, which can become very large over time.In such colonies, males are likely to take one mating opportunity after another without searching for females.Other species of spider mites are solitary and do not form apparent colonies (Saito 1995).Males may take time to find virgin females and there may be no opportunity to mate with many females in a short period.Comparing sperm replenishment ability among species of different life type may provide some insight into the evolution of the sperm replenishment ability in male spider mites.
In this study, copulation duration in the sperm-replenished males was significantly longer than in the sperm-depleted males but significantly shorter than in males before sperm depletion.We can think of two possible explanations for this phenomenon.First, males may vary their copulation duration according to the amount of sperm that they retain.Sperm was replenished to the extent that they can inseminate a few females by 3-24 h rest, but still the amount of sperm would be less than that of a virgin male.This idea may be inconsistent with the result that copulation duration did not increase with rest time.However, if there is a threshold for sperm amount that changes copulation time, and 24 h rest was not enough to replenish sperm above the threshold, then there is no discrepancy.Second, males may change their copulation duration in response to their copulation experience.Copulation for 40 s is enough for inseminating females in T. urticae (Satoh et al. 2001 -note that a different strain and form was used in this study).Prolonged copulation is considered as a form of post-mating guarding to inhibit females from immediately mating with other males and using sperm from them (Satoh et al. 2001).Therefore, males may change the copulation duration according to the importance of post-mating guard, and they may value full paternity for the first few copulations, but not so much for later copulation, and even less so if sperm is depleted.To investigate which explanation is more plausible, further research is necessary.However, we should notice that it is of course also possible that females end the mating, for example, because of no sperm transfer, even though we have been discussing as if the male exclusively determines the copulation duration.
The finding that T. urticae can replenish sperm fast appears to suggest that sperm is cheap to produce in T. urticae.The suggestion is supported by other traits.For example, costs of sperm production have been found mainly in organisms experiencing intense sperm competition.In Drosophila flies, a female mates with several males and uses the sperm of all mates for egg fertilization, resulting in intense sperm competition.To win the intense sperm competition, Drosophila males produce long spermatozoa, resulting in larger costs of sperm production.Indeed, it is found that nutrient diet during development affects the amount of sperm transferred to females in D. melanogaster (Macartney et al. 2021) and that the number of sperm bundles that the male contains is positively correlated with male body size in D. hydei (Pitnick and Markow 1994a).In spider mites, females use sperm only from the first mating if the mating is not disturbed, yet readily mate multiple times (Oku 2010;Rodrigues et al. 2020) indicating that sperm competition is present but not as strong as in fruit flies.Therefore, costs of sperm production are expected to be lower in spider mites.However, before concluding that sperm is cheap in T. urticae, it is necessary to evaluate costs of sperm production, for example by investigating the effect of nutrient status on the number of sperm that the males contain and the effect of sperm replenishment on longevity.Besides, anatomical physiology study is required to confirm that sperm was really depleted and replenished in the males, and that sperm was transferred to the females, as it is also possible that males did not transfer sperm whereas sperm was replenished and that females did not use the sperm for egg fertilization whereas sperm was transferred because of cryptic female choice.

Fig.
Fig.3Mean (± SE) copulation duration (s) in males before and after rest for 3, 6, 12 or 24 h after sperm depletion (before, after) and in males who did not experience sperm depletion (control).The mean values were calculated from the mean values for each male, as each male mated with three females.The numbers of males are shown above each bar