Wolbachiastrengthens the match between pre-mating and early post-mating isolation in spider mites

We are grateful to Inês Santos for the maintenance of the spider mite populations and the plants, to Leonor Rodrigues and Élio Sucena for useful advises with experimental designs, and to Vitor Sousa and Alexandre Blanckaert for useful comments that improved the manuscript. Funding This work was funded by an FCT -ANR project (FCT -ANR//BIA-EVF/0013/2012) to SM and Isabelle Olivieri, and by an ERC Consolidator Grant (COMPCON, GA 725419) to SM. MC was funded through an FCT PhD fellowship (SFRH/BD/136454/2018), and FZ through an FCT Post-Doc fellowship (SFRH/BPD/125020/2016) when experiments were performed. This is contribution ISEM-2024-XXX of the Institute of Evolutionary Science of Montpellier (ISEM). Conflict of interest The authors declare that they have no conflict of interest with the content of this article. Data availability All datasets and R scripts are available at Zenodo (https://doi.org/10.5281/zenodo.11160702). .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 12, 2024. ; https://doi.org/10.1101/2024.05.09.593295doi: bioRxiv preprint 2

Endosymbiotic reproductive manipulators are widely studied as sources of post-zygotic isolation in arthropods, but their effect on pre-zygotic isolation between genetically differentiated populations has garnered less attention. We tested this using two partially isolated populations of the red and green colour forms of Tetranychus urticae, either uninfected or infected with a Wolbachia strain inducing or not cytoplasmic incompatibility. We first investigated male and female preferences, and found that, in absence of infection, females were not choosy but all males prefer red red-form females. Wolbachia effects were more subtle, with only the CI-inducing strain slightly strengthening colour- form based preferences. We then performed a double -mating experiment to test how incompatible matings affect subsequent mating behaviour and offspring production, as compared to compatible mating. Females mated with an incompatible male (infected and/or heterotypic) were more attractive and/or receptive to subsequent (compatible) matings, although analyses of offspring production revealed no clear benefit for this remating behaviour ( i.e., apparently unaltered first male sperm precedence). Finally, by computing the relative contributions of each reproductive barrier to total isolation, we showed that pre-mating isolation matches both host-associated and Wolbachia-induced post-mating isolation, suggesting that Wolbachia could assist speciation processes in this system.

Haplodiploidy, mate choice, reproductive interference, sperm precedence, cytoplasmic incompatibility, reinforcement. .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 12, 2024. ; https://doi.org/10.1101/2024.05.09.593295doi: bioRxiv preprint 3

Understanding the evolution of reproductive barriers between taxa ha s long been a major focus of evolutionary biology (Coyne and Orr 2004) . While speciation research has traditionally viewed species divergence as a process inevitably leading to full reproductive isolation (biological species concept; Mayr 1942), recent evidence has shown that partial isolation occurring along the speciation continuum (Stankowski and Ravinet 2021) can be reversible (Taylor et al. 2006; Bhat et al. 2014; Kearns et al. 2018), or may even be selected in some circumstances (Servedio and Hermisson 2020). Studying population pairs for which reproductive barriers are incomplete is of great value to understand these processes, as it can provide insight into which type of reproductive barrier is more likely to evolve first , then drive the evolution of others (Baack et al. 2015; Lackey and Boughman 2017). On the one hand late post-zygotic barriers leading to costly hybridization can evolve first (e.g., in allopatry), then promote the evolution of pre- and/or early post -zygotic barriers at secondary contact (i.e., reinforcement following the definition of Coughlan and Matute; 2020; but see Bank et al. 2012) . On the other hand , by limiting gene flow, pre -zygotic barriers should lead to faster accumulation of genetic differences between populations in sympatry , thereby promoting the evolution of greater post-zygotic barriers (e.g., Lackey and Boughman 2017) . In addition , reproductive isolation may be driven not only by the genetics of the organisms themselves, but also by their endosymbionts (Brucker and Bordenstein 2012) . This is especially true for endosymbionts that directly manipulate the reproduction of their hosts (Duron et al. 2008; Engelstädter and Hurst 2009; Brucker and Bordenstein 2012). Wolbachia is a widespread endosymbiotic bacterium (Weinert et al. 2015) that manipulates its host reproduction in different ways to increase its own transmission and invasiveness (Werren et al. 2008; Engelstädter and Hurst 2009) . The most common of such manipulations is cytoplasmic incompatibility (CI), a conditional sterility phenotype which results in embryonic mortality of offspring from crosses between infected males and uninfected females (or females infected with an incompatible strain; Shropshire et al. 2020). Although the contribution of Wolbachia to post-zygotic isolation has been extensively studied in different systems , its contribution to pre -zygotic isolation (both pre- and post-mating) between hosts has received comparatively less attention (see Shropshire and Bordenstein 2016; Bi and Wang 2020; Kaur et al. 2021) , especially when acting alongside host genetic incompatibilities. Theory predicts that Wolbachia could drive reinforcement between undifferentiated host populations (i.e., females may evolve avoidance of incompatible males to escape CI ; Champion de Crespigny et al. 2005; Telschow et al. 2005), but empirical studies have produced contrasting results, most of them showing no (or weak) evidence for CI -driven assortative mating (reviewed by Shropshire and Bordenstein 2016; Bi and Wang 2020) . Such discrepancy could be explained by uneven abilities of hosts to detect Wolbachia infection in their mates (e.g., Wolbachia may alter the chemical profiles of some host species only; Richard 2017; Fortin et al. 2018; Schneider et al. 2019), or because avoidance of CI might be more likely when the infection is associated with pre -existing host traits that can be used for mate recognition (Engelstädter and Telschow 2009). If such is the case, CI avoidance should be more commonly found between already differentiated populations. In line with this, the rare studies focusing on genetically differentiated hosts showed that pre-mating isolation was strengthened (possibly even caused) by Wolbachia infection (e.g., Jaenike et al. 2006; Koukou et al. 2006; Miller et al. 2010); but see Shoemaker et al. 1999). Finally, Wolbachia infection may also .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 12, 2024. ; https://doi.org/10.1101/2024.05.09.593295doi: bioRxiv preprint 4 drive other forms of pre -zygotic isolation, including those occurring after mating . For instance, Wolbachia infection can have deleterious effects on the production of sperm (Snook et al. 2000), the transfer of fertile sperm (Lewis et al. 2011), the fertilization success (Bruzzese et al. 2021), and the receptivity to and/or the effectiveness of re -mating; (De Crespigny and Wedell 2006; Champion De Crespigny et al. 2008; Liu et al. 2014; He et al. 2018) . However, to our knowledge, no study has specifically disentangled the relative role of Wolbachia from that of host genetic factors on different types of post-mating pre-zygotic barriers. Tetranychus spider mites are an excellent system to address the interplay between host - associated and symbiont -induced incompatibilities (Cruz et al. 2021) . Indeed, Wolbachia is ubiquitous in this genus (Breeuwer and Jacobs 1996; Gotoh et al. 2003; Xie et al. 2006; Zhang et al. 2013, 2016; Zélé et al. 2018a), and its effects have been widely studied in the two-spotted spider mite T. urticae. In this host species, the bacterium induces highly variable degrees of different types of CI (mortality or development as male of fertilized eggs in incompatible crosses; e.g., Breeuwer 1997; Perrot-Minnot et al. 2002; Vala et al. 2002; Gotoh et al. 2007; Suh et al. 2015; Zélé et al. 2020; Wybouw et al. 2023) , and has variable effects on pre-mating isolation (either no effect: Zhao et al. 2013b; Rodrigues et al. 2022; or avoidance of infected males by uninfected females: Vala et al. 2004). However, in spider mites , as in many other arthropod species, its contribution to post-mating pre- zygotic isolation has seldom been studied, which is at odds with the critical role that this symbiont may play in the speciation processes currently ongoing in this group. Given the wide and overlapping distribution of m any spider mite species (Migeon and Dorkeld 2023), as well as the high variability in genetic distances both between and within species (e.g., Matsuda et al. 2018; Villacis-Perez et al. 2021), spider mites commonly suffer various degrees of reproductive isolation . In particular, there is ample evidence of variation in all possible post- zygotic reproductive barriers (zygote and juvenile hybrid mortality, hybrid sterility, hybrid breakdown), both between (Keh 1952; Helle and Van de Bund 1962; Hill and O’Donnell 1991) and within spider mite species (e.g., Van de Bund and Helle 1960; de Boer 1982a,b; Sugasawa et al. 2002; Knegt et al. 2017; Cruz et al. 2021) . Several studies also revealed variable post-mating pre-zygotic isolation in this group (e.g., fertilization failure resulting from gametic or mechanical incompatibilities), as evidenced by a reduction in the production of female offspring in this system, because spider mites are arrhenotokous haplodiploids (haploid males develop from unfertilized eggs and diploid females from fertilized eggs; Helle and Bolland 1967) . Hence, whereas no female offspring are produced in crosses between well-formed species (e.g., Helle and Van de Bund 1962; Hill and O’Donnell 1991; Chain-Ing and Sheuan-Ping 1995; Clemente et al. 2016, 2018), male-biased sex ratios are often reported in crosses between genetically differentiated ‘forms’ of the same species or even between genotypes of the same form (e.g., Gotoh et al. 1993; Navajas et al. 2000; Sugasawa et al. 2002; Auger et al. 2013; Cruz et al. 2021; Villacis-Perez et al. 2021). In addition, because spider mites exhibit first male sperm precedence (only the first male that mates with a female sires all the offspring; Helle 1967; Rodrigues et al. 2020), females usually cannot restore their fitness through re- mating. Therefore, post-mating incompatibilities are particularly costly and should select for earlier pre-zygotic barriers through reinforcement. Yet, highly variable degrees of pre-mating isolation can be found both between (Sato et al. 2014, 2016; Clemente et al. 2016; Sato and Alba 2020) and within species (e.g., Murtaugh and Wrensch 1978; Gotoh et al. 1993). .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 12, 2024. ; https://doi.org/10.1101/2024.05.09.593295doi: bioRxiv preprint 5 To improve our understanding of the role that Wolbachia can play in ongoing speciation processes of its hosts, we aimed at disentangling the relative contributions of Wolbachia and host genetic factors to the strength of both pre - and post-mating pre-zygotic barriers between two colour forms, green and red, of the two spotted spider mite T. urticae (sometimes also referred to as T. urticae and T. cinnabarinus; Xie et al. 2006; Auger et al. 2013; Lu et al. 2017, 2018) . A previous study focusing on the joint effects of Wolbachia-induced and host-associated post-mating incompatibilities between populations of these two forms revealed full reproductive isolation due to late post-zygotic barriers (hybrid sterility and hybrid breakdown) that were independent of Wolbachia infection (Cruz et al. 2021). However, this study also revealed partial and asymmetrical earlier post-mating barriers (pre- and/or post-zygotic), resulting from a combination of host -associated and Wolbachia-induced incompatibilities (Cruz et al. 2021). Host genetic incompatibilities led to an increased proportion of haploid sons to the detriment of diploid daughters (‘male development’ or MD-type incompatibility, likely due to fertilization failure ), whereas incompatibilities attributed to Wolbachia infection in males were expressed as an increased embryonic mortality of daughters (‘female mortality’ or FM - type CI). Furthermore, both types of incompatibility ha d additive effects and act ed in the same direction of crosses (Cruz et al. 2021) , which hint ed at a possible role of Wolbachia-induced incompatibilities in host population divergence and subsequent evolution of intrinsic reproductive barriers, as found in Nasonia wasps (Bordenstein et al. 2001). Here, we first performed male and female choice tests to determine their preference for infected or uninfected mates from their own or a different colour -form population (i.e., test for pre- mating isolation). Second, we used a no-choice test to determine the effect of female mating history (virgin or previously-mated with a compatible vs incompatible male) on their mating behaviour, and to investigate whether eggs are more likely fertilized by compatible than incompatible sperm (i.e., test for ‘homotypic’ sperm precedence). Finally, we used data gathered throughout all experiments stemming from this study and the previous one (Cruz et al. 2021) to estimate the relative contribution of each measured host-associated or Wolbachia-induced individual barrier to total reproductive isolation in this system.

Spider mite populations Two populations of spider mites, each belonging to a different colour form of T. urticae (‘red’ or ‘green’), and either infected or uninfected with Wolbachia, were used in this study. These populations were previously used to assess post-mating isolation caused by both host-associated incompatibilities (HI) and Wolbachia-induced reproductive barriers (Cruz et al. 2021; see also Zélé et al. 2018 for field collection, and Zélé et al. 2020 for the effects of Wolbachia following laboratory maintenance). Briefly, the Wolbachia-infected population ‘Ri’ and its uninfected counterpart ‘Ru’ (‘Ri1’ and ‘Ru1’ in Cruz et al. 2021) belong to the red form of T. urticae, whereas the Wolbachia-infected population ‘Gi’ and the uninfected population that derived from it, ‘Gu’, belong to the green form of T. urticae. The Ru and Gu populations used in the present study were obtained from the antibiotic treatments performed for Experiments 2 and 1 in Cruz et al. (2021) , respectively. All populations were subsequently reared under the same standard laboratory conditions (24±2ºC, 16/8h L/D) at high .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 12, 2024. ; https://doi.org/10.1101/2024.05.09.593295doi: bioRxiv preprint 6 numbers ( >1000 females per population) in mite -proof cages containing bean plants ( Phaseolus vulgaris, cv. Contender seedlings obtained from Germisem, Oliveira do Hospital, Portugal). All behavioural observations were conducted during daytime at constant room temperature (25 ± 2°C). Mate preference and behaviour of males and females in choice tests To determine whether spider mites discriminate between mates to avoid Wolbachia-induced and/or host-associated incompatibilities, individual males and females were provided two mates from different populations and/or infection statuses. All combinations of choice tests performed are described in Table 1. To obtain a large number of individuals of similar age, age cohorts were created for each population twelve to fourteen days prior to the onset of each mate choice observation ( i.e., each cohort was used for two to three sequential days of observation). To this aim, 50 mated females or 50 virgin females (to obtain cohorts of females or males, respectively) from each population laid eggs during 3 days on detached bean leaves placed on water -soaked cotton in p etri dishes under standard laboratory conditions (24±2ºC, 16/8h L/D). Ten to twelve days later, female and male deutonymphs undergoing their last moulting stage ( i.e., teleiochrysalids) were randomly collected from each of the female and male cohorts, respectively, and placed separately on bean leaf fragments (ca. 9cm2) to obtain virgin adult females and males of similar age two days later. As conversely to females, males cannot easily be identified based on their body colouration, they were painted before each observation with either blue or white water-based paint (randomized across treatments) using a fine brush. Previous experiments showed no effect of this paint on spider mite mate choice and behaviour (Rodrigues et al. 2017, 2022) . Subsequently, a pair of virgin mates was installed on a 0.4cm2 leaf disc (called ‘arena’ hereafter) and each observation started when the focal individual (a virgin female or a virgin male) was introduced to the arena. The colour of the mate that first copulated with each focal individual was registered, and later ass igned to the corresponding treatment (thus ensuring that the observer was blind to the treatment to which mites belonged). Simultaneously, the time until the beginning of copulation (‘latency to copulation’) and its duration (‘copulation duration’) were recorded using an online chronometer ( http://online-stopwatch.chronme.com/). Each observation lasted until the end of a first copulation or for 30 minutes if no mating occurred. Male and female choice tests were performed separately, each with one replicate of each treatment observed simultaneously per session and four sessions of observations carried out per day. In total, 60 replicates per treatment were obtained over the course of 15 days for each of the two tests. (Re)mating behaviour and offspring production in the no-choice test Mating behaviour in the first mating event Spider mites may possess pre-zygotic mechanisms other than mate discrimination to avoid and/or reduce the cost of incompatibilities. For instance, an incompatible mate could be rejected after a copulation has started (e.g., in Littorina snails; Rolán-Alvarez et al. 1999). Moreover, as copulations lasting for less than 30 seconds can be insufficient to fully fertilize a spider mite female (Potter and Wrensch 1978; Satoh et al. 2001), shorter copulations might explain the excessive production of male offspring (i.e., arising from unfertilized eggs) to the detriment of female offspring ( i.e., arising from fertilized eggs) previously observed in crosses between green females and red males (Cruz et al. .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 12, 2024. ; https://doi.org/10.1101/2024.05.09.593295doi: bioRxiv preprint 7 2021). To test whether such post -copulatory mechanisms of avoidance of incompatibilities occur in spider mites, we performed a no -choice test, where the mating behaviour of virgin females placed with a single male was observed. Given the workload involved in this experiment, we only performed the crosses allowing to test for the single and combined effects of host -associated and Wolbachia- induced incompatibility, along with their respective controls (cf. Table 2). Also, because only few individuals could be test ed per day in this experiment , producing age cohorts was not necessary. Instead, male and female teleiochrysalids were directly sampled from the base populations two days prior to observation, and isolated on bean leaf fragments to ensure their virginity. For each treatment, one male and one female were installed together on a 0.5cm 2 bean leaf disc and observed for 60 minutes. During that time, multiple mating could occur. Thus, in addition to the mating propensity (i.e., the probability that mating occurred at least once), copulation frequency ( i.e., the number of copulations during the observation period) was also registered. The latency to the first copulation and the duration of each copulation occurring during the observation period were recorded as in the mate choice experiment. The cumulative copulation duration of each couple was subsequently computed. At the end of the observation period, females for which at least one copulation occurred were individually placed on a 2cm2 bean leaf disc and kept for the next step ( cf. below), while non-mated females and all males were discarded. Mating behaviour in the second mating event In species with first-male sperm precedence such as T. urticae (i.e., the first male that had mated with a female sires all of her offspring; Rodrigues et al., 2020), females usually have low receptivity to a second mate (Clemente et al. 2016) . However, if the first copulation is interrupted or (at least partially) ineffective, females may show increased receptivity to second matings that could effectively contribute to fertilization (Helle 1967; Clemente et al. 2016; Costa et al. 2023). To test this, females for which at least one copulation occurred during the first mating event were placed with a second compatible male 24 hours later (cf. Table 2) and their mating behaviour was recorded. Behavioural observations were carried out for 60 minutes as in the first mating event. The mating propensity, mating frequency, latency to first copulation and cumulative copulation duration with the second male were simultaneously registered. At the end of the observation period, males were discarded and females were kept individually on 2 cm2 bean leaf discs placed on water-soaked cotton in petri dishes in a climatic chamber (25 ± 2°C, 60% RH, 16/8 h L/D). Given the workload and the multiplicity of tasks involved in this experiment , only 9 couples were observed simultaneously per session of observation, corresponding to one or two replicates per treatment. Four sessions of observation were performed per day (hence 6 replicates of each treatment per day), each day corresponding to an experimental ‘block’. In total, 19 blocks, each separated by 3 days, were performed to obtain ca. 100 replicates per treatment. Offspring production and strength of post-mating incompatibilities To test whether the second copulation with a compatible male could restore female offspring production, the offspring produced over 3 days of oviposition by females mated with either a single or two different males was compared , and female mortality during that period also registered . The number of unhatched eggs was counted 6 days later (day 9), and the numbers of dead juveniles, adult males and females were counted 3 and 6 days later (days 12 and 15). Then, to determine the proportion of offspring affected by host-associated MD-type incompatibility (i.e., “Male Development”), and/or .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 12, 2024. ; https://doi.org/10.1101/2024.05.09.593295doi: bioRxiv preprint 8 Wolbachia-induced FM-type incompatibility (i.e., “Female Mortality”), we computed two indexes as fully described in Cruz et al. (2021): the 𝑀𝐷!"## index, which calculates the overproduction of sons in the brood, and the 𝐹𝑀!"## index, which calculates the embryonic mortality of fertilized offspring (i.e., only females in haplodiploids), both relative to the control crosses (thereby accounting for

variation). Finally, as in Cruz et al. (2021) , we also computed the proportion of F1 females over the total number of eggs (FP) to determine the combined effect of FM- and MD-type incompatibilities on the total proportion of daughters in each cross . Raw data are shown in the Supplementary Figure S1. Strength and contribution of each reproductive barrier to total isolation Strength of reproductive isolation for each reproductive barrier (RIn) To estimate the strength of pre- or post-mating reproductive barriers identified for each type of cross within and between the green - and red-form populations, we used a combination of the pre -mating data obtained in the present study and the post-mating data obtained in the previous study (Cruz et al. 2021), respectively. A ll reproductive barriers found to play a role in reducing gene flow among the spider mite populations were considered: mate preference (RI1), fertilization failure ( RI2), hybrid inviability (RI3), hybrid sterility ( RI4), and hybrid breakdown ( RI5). As we found no evidence for homotypic sperm precedence in the no -choice test ( cf. Results), this barrier was not considered. Similarly, female choice data were not used when computing RI1, as females overall showed no mate preferences in the choice test (cf. Results). To determine the strength of pre -mating isolation (RI1), we applied a sexual isolation index, which varies between zero and one, to the male choice data. This index, adapted from Bateman (1949) and Merrell (1950) by Malogolowkin-Cohen et al. (1965), is given by: 𝑅𝐼$(&) = (𝑛&& − 𝑛(& ) (𝑛&& + 𝑛(& ) where nxx is the number of copulations observed between females and males of a population x, and nyx is the number of copulations observed between females of a population y and males of the population x. As RI1 represents the degree to which a population x is isolated from a population y due to mating preferences, it was set to 0 in the case of preference for heterotypic mates (i.e., no negative impact on gene flow). To determine the strength of post -mating barriers, we used the data published in Cruz et al. (2021), as late reproductive barriers ( i.e., hybrid sterility and hybrid breakdown) were not measured here. Moreover, earlier post -mating barriers (fertilization failure and hybrid inviability) have been estimated more precisely in the previous study than in the current one (i.e., larger sample sizes). For fertilization failure (RI2) and hybrid inviability ( RI4), we used the median values of the MD corr and FMcorr indexes, which correspond to the percent increase in non -fertilized eggs and in embryonic mortality o f fertilized eggs, respectively ( cf. ‘Offspring production and strength of post -mating incompatibilities’ above). For hybrid sterility ( RI5) and hybrid breakdown ( RI6), we computed the percent decrease in ovipositing F1 females and increase in embryonic mortality of F1 females’ eggs relative to compatible crosses, respectively. Contribution of each reproductive barrier (Cn) to total isolation (T) We employed a method previously adapted from Coyne and Orr (1989, 1997) by Ramsey and colleagues (2003), in which total (cumulative) reproductive isolation between two populations or .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 12, 2024. ; https://doi.org/10.1101/2024.05.09.593295doi: bioRxiv preprint 9 species is computed as a multiplicative function of the strength of each reproductive barrier (𝑅𝐼); see above), so that the contribution of each barrier to reducing gene flow at a stage 𝑛 in life history is calculated as: 𝐶) = 𝑅𝐼) .1 − 0 𝐶* )+$ *,$ 1 Thus, a given reproductive barrier eliminates gene flow that has not been prevented by earlier barriers, and for 𝑚 reproductive barriers, total reproductive isolation is given by: 𝑇 = 0 𝐶* - *,$ Statistical analyses Analyses were carried out using the R statistical software (v3.6.1). The general procedure for building all statistical models is detailed in the Supplementary Tables S1 and S2. Time-to-event data (latency to copulation and copulation duration) were analysed using Cox proportional hazards mixed models (coxme procedure; coxme package), a non -parametric method that does not assume any particular error distribution (Crawley 2007) . All other data were analysed using generalized linear mixed models (glmmTMB procedure; glmmTMB package ; Brooks et al. 2017) . P roportion data were computed either as binary response variables (e.g., mated or not, chosen mate) or using a concatenated response variable binding the number of successes and failures with the function cbind to account for the number of eggs in the analyses of female proportion (i.e., number of daughters vs. number of eggs – number of daughters). F or the corrected variables 𝑀𝐷!"## and 𝐹𝑀!"##, which are continuous variables bounded between 0 and 1, we weighted each individual datum by the number of observations (i.e., a “weights” argument was added to the model s). These data were subsequently analysed with a binomial or (zero -inflated) betabinomial error distribution when errors were overdispersed. Count data were analysed with a Poisson error distribution, and continuous data with a Gaussian, except for daily oviposition which was Box-Cox transform ed (λ=0.549) to improve normality (Crawley 2007) and analysed with a zero-inflated Gaussian. For the analyses of each response variable of the choice tests, the ‘type’ of the focal individual (i.e., combination of population and infection status), and either the combination of provided mates (for the analyses of mating propensity and mate preference) or the chosen mate (for the analyses of latency to copulation and copulation duration), were fit as fixed explanatory variables, whereas the day and session (nested within day) of observation, the colour with which chosen males were painted (in the female choice test), and the combination of provided mates (for the analyses of latency to copulation and copulation duration) were fit as random explanatory variables. In addition, f or the analyses of mate preference, the intercept of the models was forced to zero to obtain the estimate of the fixed factor as the difference to a 0.5 probability (particularity of models with categorical factors and binomial distribution; Crawley 2007). For the analyses of the no -choice test, mating propensity data obtained in the two mating events were analysed altogether, and the mating event was fit as fixed explanatory variable, along with the type of cross (i.e., female x first male) and the interaction between these two factors, whereas the day and the session (nested within day) of observation were fit as random explanatory variables. To avoid pseudoreplication i n the analyses of copulation frequency, latency to copulation and .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 12, 2024. ; https://doi.org/10.1101/2024.05.09.593295doi: bioRxiv preprint 10 copulation duration, data obtained in each mating event were analysed separately and only the type of cross was fit as fixed explanatory variable. Subsequently, to determine whether these response variables differed when females mated with the first or with the second male, the differences in copulation frequency, latency, and duration between 1st and 2nd mating events were computed for each individual female that mated at least once in both events . For that, the type of cross was fit as fixed explanatory variable, the day and session (nested within day) were fit as random explanatory variables, and the intercept was forced to zero to obtain the estimate of the fixed factor relative to no difference between mating events . Finally, for the analyses of offspring data, the type of cross, the mating status of the females (i.e., mated with only the first or both males), and the interaction between these two factors, were fit as fixed explanatory variables, whereas the day and session (nested within day) during which they mated were fit as random explanatory variables. For all analyses, maximal models containing the complete set of explanatory variables were subsequently simplified by sequentially eliminating non -significant terms to establish a minimal model (Crawley 2007). The significance of the explanatory variables was established using chi-square tests with the Anova function (car package; Fox et al. 2019). The significant values given in the text are for the minimal model, whereas non -significant values correspond to those obtained before deletion of the variable from the model (Crawley 2007). When explanatory variables with more than two levels were found significant in the analyses of behavioural data, a posteriori contrasts between factor levels were carried out by aggregating factor levels together and testing the fit of the simplified model using a likelihood ratio test (function anova; Crawley 2007). When interactions between two explanatory variables were significant, the two variables were concatenated (e.g., crosses and mating events) and a posteriori contrasts between the factor levels of the concatenated variable were carried out as described above . In all cases, Holm-Bonferroni corrections ( i.e., classical chi -squared Wald test for testing the global hypothesis H0; Holm 1979) were subsequently applied to account for multiple testing. Note that contrast analyses were not carried out for offspring production in the no - choice test because the results did not differ qualitatively from th ose of the previous study (Cruz et al. 2021). Finally, to determine the difference to random mating in the analyses of mate choice , as well as changes in mating behaviour between 1 st and 2 nd mating event s in the no -choice test , coefficients (estimated as the difference with the zero -intercept) obtained from the maximal models for each combination of focal individual and pair of provided mates (mate choice test), or for each cross (no-choice test), were analysed with Z -tests using the function test (emmeans package; Lenth et al. 2018).

Male and female mating behaviour in choice tests Overall, the propensity of females to mate with one of two provided males depended on their own population (‘focal’: χ23=10.06, p=0.018; Model 1.1 in Table S1; Figure 1a), with green females being, on average, ca. 20% less likely to mate than red females (Tables S3 and S4). However, their mating propensity was unaffected by the type of males they were offered (‘mates’: χ23=2.70, p=0.44; Model 1.1), and none of them showed any clear mating preference ( ‘focal’: χ24=2.91, p=0.57, and ‘mates’: χ24=3.47, p=0.48; Model 1.2; Figure 1b and Table S5 ). Conversely, the mating propensity and the mate choice of males were independent of their population ( ‘focal’: χ23=4.13, p=0.25 and χ24=5.01, p=0.29 in Model 1.5 and 1.6, respectively ), but strongly affected by the type of females provided .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 12, 2024. ; https://doi.org/10.1101/2024.05.09.593295doi: bioRxiv preprint 11 (‘mates’: χ23=15.72, p=0.001 and χ24=50.24, p<0.0001 in Model 1.5 and 1.6, respectively; Figures 1c and 1d). Indeed, males that were given the choice between two green females were less likely to mate than those that were given the chance to mate with a red female ( ca. 40% vs 68% mated males on average; Figure 1c, Tables S3 and S4), and males of either colour form showed a preference for red females (ca. 60% to 80% preference; Tables S3 and S4). In contrast to the mite colour form, Wolbachia infection had no effect on mite mating propensity and only a small effect on their mate preference (cf. Figure 1, Table S3 and contrasts in Table S4). Neither uninfected females nor infected males showed any preference between infected or uninfected mates of the same colour (cf. category I in Figures 1b and 1d), but Wolbachia infection in red males (CI -inducing Wolbachia strain; Cruz et al. 2021) strengthened their preferences for red females. Indeed, although the mate preference of Ru and Ri males did not differ significantly, Ru males showed no significant difference from random mating (cf. category II in Figure 1d) whereas Ri males significantly preferred red females over green females (cf. categories III and IV; see also Table S5). In addition, whereas red females overall did not differ in their mate preference, and showed no preference between males of either colour form when these were from the same infection status as themselves (cf. categories II and III in Figure 1b ), Ru females preferentially mated with Gi males over Ri males , suggesting avoidance of the CI induced by Wolbachia infection in red males (cf. category IV; see also Table S5). Conversely, the non-CI-inducing Wolbachia strain infecting green males (Cruz et al. 2021) had no effect on the strength of mate preference of both males and females. Finally, latencies to copulation did not differ significantly among focal females or chosen males in the female choice test (χ23=6.76, p=0.08 and χ23=1.35, p=0.72, respectively; Model 1.3), nor among focal males or chosen females in the male choice test (χ23=1.33, p=0.72 and χ23=1.03, p=0.79, respectively; Model 1.7, Figure 2a,b), but copulation duration differed between males of different colours (Figure 2c) and between females of different infection status (Figure 2d) . Regardless of Wolbachia infection (although Ru males showed intermediate copulation durations in the female choice test; Figure 2c; Table S6 and S4), green males copulated on average 37 and 40 seconds longer than red males in the female and male choice test, respectively (‘chosen’: χ23=7.92, p=0.048, and ‘focal’: χ23=27.09, p<0.0001; Model 1.4 and 1.8, respectively ). Conversely, the copulation duration of females was not affected by their colour form (although Gi females showed intermediate copulation durations in the female choice test; Figure 2d and Table S6; cf. contrasts in Table S4), but that of infected females was, on average, ca. 29 and 34 seconds shorter than that of uninfected females in the female and male choice test, respectively (‘focal’: χ23=10.64, p=0.014, and ‘chosen’: χ23=24.70, p<0.0001 in Model 1.4 and 1.8, respectively). (Re)mating behaviour in the no-choice test On average, 58% of the virgin females placed on a leaf disc with a single male mated within 1 hour, whereas less than 20% of those mated females re-mated when placed with another male 24 hours later. In line with this, a reduced copulation frequency ( 1.6±0.1 vs 2.1±0.1 copulations per couple) and copulation duration (118±13 vs 252±8 seconds) and an increased latency to copulation (1582±108 vs 986±46 seconds) , were observed, on average , between the first and second mating events for couples that mated at least once (Figures 3,4 and Table S7). However, this reduction in the willingness to mate varied across types of crosses for all behavioural traits tested but copulation frequency (no statistically significant differences found among crosses for either o r among the two mating events; cf. models 2.2 to 2.4 in Table S2). .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 12, 2024. ; https://doi.org/10.1101/2024.05.09.593295doi: bioRxiv preprint 12 For the mating propensit ies observed during the first mating event, we found the same tendencies as in the choice tests: Gu females were less likely to mate than Ru females (except when paired with Ru males here), and Wolbachia infection in red males seemed to promote mate discrimination (Ri males were less willing to mate with Gu females than Ru males , whereas both types of males mated as much with Ru females ; Figure 3; Table s S7 and S8). Then, although no statistically significant differences were found among crosses in the second event, not all crosses led to the same reduction in mating propensity between the two mating events ( cross x mating event interaction: χ25=16.56, p=0.005; model 2.1; Figure 3; Table S 8): Gu females showed a lower reduction in their tendency to re-mate than Ru females, especially when they were previously mated with an incompatible Ri male (hence when both types of incompatibilities were at play; Figure 3; Tables S7 and S8). In this experiment, conversely to the previous experiment in which virgin individuals could choose their mate and were given only half an hour to mate, we found significant differences among latencies to copulation of couples that mated at least once during the first mating event (χ25=13.19, p=0.02; model 2.5) . Gu females took, on average, 5 more minutes than Ru females to engage in copulation with their first partner, regardless of the form or infection status of the latter (although Ru x Ru crosses had intermediate latencies to copulation ; cf. circles in Figure 4a; Tables S7 and S9). Also, in this first mating event, as in the previous experiment (choice tests), the cumulative time spent copulating was longer for green males than for red males regardless of their infection status and the female they mated with (ca. 39 seconds difference; χ25=21.19, p<0.001; model 2.4; Figures 2c,d and 4c; Tables S7 and S9). Then, when females that mated during the first mating event were placed with a second male 24 hours later, their latency to copulation increased by almost 10 minutes, and their copulation duration was more than 2 minutes shorter, than when they were virgin ( cf. diamonds in Figure 4a,c; Table S7). Despite no significant differences being found among types of crosses for both latency to copulation and cumulative copulation duration in the second mating event (χ25=5.16, p=0.40; model 2.6; Figure 4a, and χ25=2.78, p=0.73; model 2.9; Figure 4c, respectively), behavioural changes between the two mating events at the female level (for those who mated in both events) varied depending on the type of cross (χ26=12.47, p=0.05; model 2.7 ; Figure 4b , and χ 26=43.73, p<0.0001; model 2.10 ; Figure 4d , for latency to copulation and cumulative copulation duration, respectively). Thus, i n line with the mating propensity observations , differences in latency to copulation and copulation duration between mating events tended to disappear for females that had first mated with an incompatible male (except for the copulation duration of Gu females mated with Ri males; Table S10). Effect of remating on offspring production in the no-choice experiment The pattern of offspring production for females that mated only with one male (Figure 5a) was consistent with that described in our previous study (Cruz et al. 2021) . Briefly, (i) we found an overproduction of males (MD -type incompatibility) in crosses between green females (Gu) and red males (Ru or Ri ) as compared to the other crosses (χ25=76.30, p<0.0001; model 2.12; Figure 5b) . Moreover, because copulations were observed in the present study, it further unambiguously revealed a high variability for this barrier: among the 66 Gu females that mated only with a Ru or Ri mate and oviposited (i.e., 85 Gu females mated with a Ru or Ri male subsequently refused to mate wi th a second male; Table S7, and 19 of these females did not lay a single egg), 20 produced only sons (i.e., full incompatibility), whereas 18 did not produce a more male-biased sex ratio than the controls (i.e., no incompatibility); (ii) we found an increased female embryonic mortality (FM -type CI quantified .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 12, 2024. ; https://doi.org/10.1101/2024.05.09.593295doi: bioRxiv preprint 13 as a decreased hatching rate of fertilized eggs) in crosses between uninfected females (Gu or Ru) and males infected with a CI -inducing Wolbachia strain (Ri males), as compared to the other crosses (χ25=76.78, p<0.0001; model 2.13; Figure 5c); and (iii) we found a reduction in the proportion of daughters (FP) in crosses affected by either (or both) type(s) of incompatibility (i.e., Ru x Ri, Gu x Ru and Gu x Ri, female x male crosses ) as compared to compatible crosses ( χ25=87.65, p<0.0001; model 2.14; Figure 5d). However, no difference in offspring production was found between females that mated with only one or two different males (daily fecundity: χ 21=3.19, p=0.07; model 2.11; MDcorr: χ21=0.63, p=0.43; model 2.12; FM corr: χ21=0.14, p=0.71; model 2.13; FP: χ 21=0.17, p=0.68; model 2.1 4), regardless of whether the first male was compatible or not ( i.e., no significant interactions between the type of cross and whether females mated with one or two males; daily fecundity: χ25=4.57, p=0.47; model 2.11; MD corr: χ25=0.89, p=0.97; model 2.12; FM corr: χ25=9.33, p=0.10; model 2.13; FP: χ25=7.17, p=0.21; model 2.14; Figure 5; see also Figure S1 and Table S7). Contribution of intrinsic and Wolbachia-induced reproductive barriers to reducing gene flow Although hybrid breakdown is the strongest reproductive barrier in both directions of crosses between the two studied spider mite populations (100% F2 embryonic mortality; Cruz et al. 2021), it ultimately contributes very little to total isolation due to the occurrence of earlier barriers (Figure 6 and Table S11). Red females and green males are mainly isolated due to hybrid sterility (98 to 100% isolation regardless of Wolbachia infection), as no other reproductive barrier exist s in this cross direction. However, despite having the same strength in both directions of heterotypic crosses, hybrid sterility acts along with other reproductive barriers in crosses between green females and red mal es, which strongly reduced its contribution to total isolation (ca. 12% and 29% in crosses with infected and uninfected males, respectively). In this cross direction , our estimations revealed that assortative mating and fertilization failure are in fact the main sources of reproductive isolation, contributing to 27-61% and 23-71% of total isolation, respectively. Moreover, although hybrid inviability caused by the CI-inducing Wolbachia strain infecting red males only has a weak contribution to total isolation in heterotypic crosses (as compared to homotypic crosses: ca. 5.5 to 6.4% in crosses between green females and Ri males vs 32% in crosses between Ru females and Ri males; Table S11), infection of males with this Wolbachia strain clearly potentiates pre-mating isolation (Figure 6) . Indeed, t he strength of assortative mating increases from ca. 27% in crosses between Gu females and Ru males (non-significantly different from random mating; cf. Figure 1 and above) to ca. 61% in crosses between Gu females and Ri males (Figure 6) and to ca. 49% in crosses between Gi females and Ri males (Table S11).

In this study, we sought to shed light on the potential role played by Wolbachia as an agent of pre - zygotic isolation between genetically differentiated colour forms of the spider mite Tetranychus urticae. Our results revealed that Wolbachia infection had no effect on the mating preference of both males and females in homotypic crosses, but the CI-inducing strain infecting the red form exacerbated pre-existing colour-based mate preferences. Whereas both types of males showed a preference for red females , this preference seemingly disappeared when red males were cured from Wolbachia infection. In line with this, females showed no mate preferences in the absence of Wolbachia infection, but uninfected red females showed a preference for green infected males (which do not .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 12, 2024. ; https://doi.org/10.1101/2024.05.09.593295doi: bioRxiv preprint 14 carry a CI-inducing Wolbachia strain) over red infected ones (which do carry a CI -inducing Wolbachia strain). Then, we also found that (i) females that had engaged in matings where both types of incompatibility occur red (Wolbachia-induced and host -associated) were more likely to re -mate with a compatible male, and (ii) females exposed to either type of incompatibility did not significantly increase their latency to re-mate, nor reduce their copulation duration when re-mating, as compared to their first matings. Yet, the significance of such behaviour is puzzling, as we found no evidence of sperm contribution by the second (compatible) males following copulations with incompatible mates (i.e., adult offspring production and sex -ratios were not more similar to the controls). This further indicated that ‘homotypic sperm precedence’ was not a reproductive barrier at play in our experiment. Finally, our estimations of the relative contribution of each reproductive barrier to reproductive isolation between the studied populations clearly illustrate the strong asymmetries that occur in this system: red females are isolated from green males due to late intrinsic post-zygotic barriers (hybrid sterility and hybrid breakdown), whereas green females are isolated from red males by a combination of early and late reproductive barriers (pre-mating, post-mating pre-zygotic, and post-zygotic), either directly caused (hybrid inviability due to CI) or strengthened (assortative mating) by Wolbachia infection in males. A system driven by male rather than female mate preferences In most tested scenarios, T. urticae females did not choose between mates with different colour forms. This corroborates earlier results, where females from different morphs or infection status did not discriminate between different types of males (Murtaugh and Wrensch 1978, Zhao et al. 2013b; Rodrigues et al. 2022, but see Vala et al. 2004; Rohrscheib et al. 2015) , and indicates no differences in male competitive ability as well (Murtaugh and Wrensch 1978 ; Wagner 1998) . In fact, several other studies revealed an absence of mate choice in spider mite females (e.g., Magalhães et al. 2009; Zhou et al. 2020). This is surprising, as females invest more energy than males in their reproduction (Kokko et al. 2006 ), and spider mites have first-male sperm precedence (Helle 1967; Satoh et al. 2001; Rodrigues et al. 2020) , hence the choice of the first male has enormous consequences for females (Wittenberger and Tilson 1980; Howlett 1988; Griffith et al. 2011) . Possibly, this weak female choice is a consequence of male guarding of females just before their emergence as virgin adults (Potter et al. 1976), leading to little opportunity for females to choose their mate (Everson and Addicott 1982; Oku 2014). In contrast, we found strong mate preferences in males, which is also in line with earlier studies on spider mites (e.g., Everson and Addicott 1982; Rodrigues et al. 2017), and in other arthropods in which males invest time and energy in pre - and/or post-copulatory guarding (reviewed in Bonduriansky 2001). Asymmetric reinforcement could explain the match between pre- and early post-mating barriers In this system, one might expect assortative mating ( i.e., homotypic preference in both cross directions) to be selected for due to severe costs of hybridization in both cross directions (Cruz et al. 2021). Instead, our results revealed an asymmetry in pre-mating isolation (only red males prefer homotypic females) . Possibly, post-mating pre -zygotic barriers ( e.g., fertilization failure due to cytonuclear incompatibility) first evolved incidentally between green -form females and red -form males in allopatric populations. The resulting asymmetrical maladaptive hybridization may have subsequently ( i.e., at secondary contact) led to asymmetrical levels of reinforcement in areas of sympatry (Noor 1999; Servedio and Noor 2003; Coyne and Orr 2004), thereby driving the evolution .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 12, 2024. ; https://doi.org/10.1101/2024.05.09.593295doi: bioRxiv preprint 15 of homotypic mate preferences by red males only (e.g., as observed between Drosophila recens and D. subquinaria populations due to unidirectional CI induced by Wolbachia; Jaenike et al. 2006). This may explain the match between pre-mating and early-acting post-mating barriers in this system (sex ratio distortion likely due to fertilization failure in crosses between red males and green females; Cruz et al. 2021), as found in other systems (reviewed in Ortiz-Barrientos et al. 2009; see also Giesbers et al. 2013; Yukilevich et al. 2018). Alternatively, asymmetric barriers acting in the same cross direction could as well be due to genetic linkage between barriers ( e.g., Merrill et al. 2011) , a possibility not yet investigated in spider mites . Subsequently, the two forms might have further diverged due to limited gene flow, leading to the establishment of strong late post-zygotic barriers in both directions (Servedio and Sætre 2003) . In line with this , previous work has shown that barriers acting early in reproduction tend to evolve faster than those acting later (Coyne and Orr 1989; Servedio 2001; Turissini et al. 2018). Aside from pre-mating isolation, reinforcement could also drive the evolution of other types of pre-zygotic barriers, including those occurring after mating, such as conspecific sperm precedence (Coughlan and Matute 2020). Although preferential use of sperm from conspecific (or ‘homotypic’) males would reduce the negative effects of mating with heterospecifics (e.g., Price 1997; Fricke and Arnqvist 2004; Noriyuki et al. 2012) , we did not find any evidence for such reproductive barrier . Nevertheless, we observed that the latencies to copulation and copulation duration s of green-form females previously mated with red-form males tended to remain the same as when they were virgins. This contrasts with the behaviour of females mated with fully-compatible males, which become less receptive to subsequent males (increased latency and reduced copulation duration), in line with the first-male sperm precedence pattern (Helle 1967; Rodrigues et al. 2020). Other studies also found that spider mites first mated with (fully or partially) incompatible males behave as virgins in subsequent matings (Clemente et al. 2016; Costa et al. 2023) . Moreover , when the incompatibilities are not caused by fertilization failure (as in the case of Wolbachia-induced CI), a reduced latency to copulation and increased copulation duration with a second male could increase the likelihood that the sperm of the latter outcompete the sperm of the first male (Potter and Wrensch 1978; Satoh et al. 2001). However, contrarily to these earlier studies, the results obtained here do not indicate any use of the sperm from second males, which could rescue limited offspring production resulting from incompatible matings. Yet, this pattern may be jeopardized under other conditions than those tested in the current study. Here, we allowed for several copulations with the first male, and mated females were exposed to a second male only 24 hours later . This was done to detect potential issues with sperm transfer or storage when an excess of male offspring is found ( i.e., in crosses between green females and red males) . If this was the case, significant effects of double mating on offspring production could not be unambiguously attributed to changes in the sperm precedence patterns (García‐González 2004) . However, the timing used might have been excessive to enable sperm competition/cryptic sperm choice (Potter and Wrensch 1978; Satoh et al. 2001) , and future studies are necessary to uncover potential benefits of the behaviours observed here. Wolbachia-induced CI might strengthen asymmetrical reinforcement Our results concerning t he effect of Wolbachia infection on mate preferences between mite colour forms corroborate those of previous studies showing that Wolbachia infection strengthens assortative mating between genetically differentiated hosts (e.g., Jaenike et al. 2006; Koukou et al. 2006; Miller et al. 2010). In particular, the fact that Wolbachia infection alone (i.e., in homotypic crosses) has no significant effect on mate choice (as in Rodrigues et al. 2022), supports the hypothesis that avoidance .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 12, 2024. ; https://doi.org/10.1101/2024.05.09.593295doi: bioRxiv preprint 16 of CI might evolve more readily when the infection is associated with pre-existing host traits that can be used for mate recognition (Engelstädter and Telschow 2009) . Moreover, the fact that mating preferences contribute more to total reproductive isolation when infected red males , which carry a CI-inducing Wolbachia strain, were involved ( cf. fig. 6) , suggests that CI could be a mechanism driving asymmetrical reinforcement between spider-mite colour forms. Consistent with a previous study on incompatibilities between different geographic strains of green-form T. urticae, in which the only females receptive to a second mate were those previously mated with a genetically incompatible male carr ying a CI-inducing Wolbachia strain (Navajas et al. 2000) , we also found that only uninfected green females previously mated with a red infected male (hence carrying a CI -inducing Wolbachia strain) were as likely to mate with a second male as when they were virgins. In line with this, only when uninfected females (both red and green) had mated with an infected red male (with the CI-inducing strain) did their latency to copulation and copulation duration remain as when they were virgin . Hence, together these findings revealed that Wolbachia can affect other mating behaviours beyond mating preferences ( as in other systems; reviewed in Bi and Wang 2020) , and suggest that Wolbachia-induced CI could assist reinforcement processes in this system. Not just a missing barrier: Heterotypic mate preference may be an adaptive strategy Although reinforcement is a seductive hypothesis to explain why red-form males prefer red females, it does not explain why green-form males also prefer these females. The occurrence of such seemingly maladaptive behaviour suggests that other, or additional, evolutionary forces are at play. One possibility could be that heterotypic mating preference is a by-product of inbreeding avoidance in the green -form population. Spider mites effectively avoid related individuals (Tien et al. 2011; Bitume et al. 2013; Yoshioka and Yano 2014) , but it is not clear whether this extends to more distantly-related individuals. For instance, males of both T. evansi and T. urticae preferentially mate with T. urticae females (Sato et al. 2016 ; but see (Clemente et al. 2016) , but this occurs even when T. evansi females are non-kin (Sato et al. 2016) . In line with this, e vidence that inbreeding avoidance can drive the evolution of disassortative mating is also lacking in other systems (van den Berg et al. 1984; Juola and Dearborn 2012; Huchard et al. 2013; Galaverni et al. 2016). Another possibility could be that preference of both types of males for red-form females is due to these females being more attractive. For instance, a new trait (e.g., a pheromone profile) may have evolved in red females in response to intense female competition (i.e., their sex ratio is more female-biased than that of green mites when they oviposit in groups; unpublished data), and this trait may then be fortuitously preferred by green males if it stimulates the same coding system as the ancestral trait (Endler and Basolo 1998) . Alternatively, both types of male may have conserved an ancestral preference for a trait that has been lost or diverged in green-form females (Endler and Basolo 1998). This could occur if the rate of evolution of male preference is slower than that of the female trait. The observed male preferences may also be caused by differences in female reluctance and male vigour (e.g., van den Berg et al. 1984) in response to stronger sexual conflicts in the green -form population. This hypothesis is supported by the fact that green females are less likely to mate than red females even in the absence of choice , whereas green males spend longer periods of time copulating than red males do (suggesting longer post-copulatory guarding; (Satoh et al. 2001). In line with this, theory predicts that sexual conflicts can drive the evolution of mate preferences, increasing reproductive isolation and, consequently, the rates of speciation (Parker and Partridge 1998). Finally, building upon the recent idea that partial reproductive isolation may be an adaptive optimum (Servedio and Hermisson 2020) , we considered the possibility that heterotypic mating .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 12, 2024. ; https://doi.org/10.1101/2024.05.09.593295doi: bioRxiv preprint 17 preference might be selected for under reproductive interference (Gröning and Hochkirch 2008), as the two colour forms have overlapping distribution and host plant range (Migeon and Dorkeld 2023), and often co -occur on the same individual host plant (Lu et al. 2017, 2018; Zélé et al. 2018b) . Although most conditions that have been theoretically considered to promote the evolution of ‘disassortative mating’, such as a heterozygote advantage (e.g., Maisonneuve et al. 2021), are not met in our system (hybrids are sterile or suffer breakdown ; Cruz et al. 2021) , heterotypic mating preference may still confer higher benefits than costs to the green-form population in the presence of red-form competitors. Indeed, this behaviour should be highly costly for red females due to first male sperm precedence, but green males may only pay relatively small cost s as they can mate multiply (Krainacker and Carey 1989). Thus, similarly to how CI induced by Wolbachia increases the relative fitness of infected females, the ‘spiteful’ behaviour of green males might be selected for as it confers an indirect fitness advantage to their green sisters (Hamilton 1970; Gardner and West 2004; Engelstädter and Charlat 2006) . Disassortative mating may thus act synergistically with sex-ratio distortion (i.e., the overproduction of sons) in crosses between green females and red males (cf. Cruz et al. 2021) to promote the exclusion of the red form population (cf. Grether et al. 2017; Cruz et al. 2023). Conversely, homotypic mating preference by red males should decrease the strength of reproductive interference for the red population, as it reduces the prevalence of crosses between green females and red males (hence the overproduction of green males stemming from these crosses) and should prevent (non-choosy) red females from having a higher chance to mate with a green male . Following this hypothesis , the CI -inducing Wolbachia strain naturally infecting the red -form population seems to favour its own host population by increasing the likelihood that red males mate with compatible (red) females, whereas it has no control over heterotypic mating preference by green males. Testing whether such an ‘escalating arms race’ could indeed occur in response to reproductive interference (involving or not Wolbachia-induced CI) is of high relevance for future speciation studies.

In this study, we identified a mechanism through which Wolbachia could assist host speciation processes. Our results show that Wolbachia infection in T. urticae males indirectly contributes to pre- mating isolation between genetically differentiated T. urticae colour forms by strengthening pre- existing preferences. These preferences match early post -mating barriers in the system , as crosses that are affected both by host-associated and Wolbachia-induced incompatibilit ies are generally avoided. Our results also further highlight the importance of pre-mating isolation in this system , as they revealed that, in our experimental conditions, females of either form are unable to compensate for incompatible crosses by re -mating. Overall, our comprehensive study of pre - and post-zygotic reproductive barriers allowed identify ing asymmetries in patterns of isolation between the two populations, hinting at a possible history of reinforcement followed by an interruption of gene flow . These findings also open new research avenues, such as to study the impact of complex patterns of isolation on population dynamics, and of the resulting selection pressures on the evolutionary trajectories of populations. .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 12, 2024. ; https://doi.org/10.1101/2024.05.09.593295doi: bioRxiv preprint 18

Auger, P., A. Migeon, E. A. Ueckermann, L. Tiedt, and M. Navajas. 2013. Evidence for synonymy between Tetranychus urticae and Tetranychus cinnabarinus (Acari, Prostigmata, Tetranychidae): Review and new data. Acarologia 53:383–415. Baack, E., M. C. Melo, L. H. Rieseberg, and D. Ortiz -Barrientos. 2015. The origins of reproductive isolation in plants. New Phytol 207:968–984. Bank, C., J. Hermisson, and M. Kirkpatrick. 2012. Can reinforcement complete speciation? Evolution 66:229–239. Bateman, A. J. 1949. Analysis of data on sexual isolation. Evolution 3:174–177. Bhat, S., P. A. Amundsen, R. Knudsen, K. Ø. Gjelland, S. E. Fevolden, L. Bernatchez, and K. Præbel. 2014. Speciation reversal in European whitefish ( Coregonus lavaretus (L.)) caused by competitor invasion. PLoS One 9:e91208. Bi, J., and Y. F. Wang. 2020. The effect of the endosymbiont Wolbachia on the behavior of insect hosts. Insect Sci 27:846–858. Bitume, E. V., D. Bonte, O. Ronce, F. Bach, E. Flaven, I. Olivieri, and C. M. Nieberding. 2013. Density and genetic relatedness increase dispersal distance in a subsocial organism. Ecol Lett 16:430–437. Bonduriansky, R. 2001. The evolution of male mate choice in insects: a synthesis of ideas and evidence. Biol Rev 76:305–339. Bordenstein, S., F. P. O’Hara, and J. H. Werren. 2001. Wolbachia-induced incompatibility precedes other hybrid incompatibilities in Nasonia. Nature 409:707–710. Breeuwer, J. A. J. 1997. Wolbachia and cytoplasmic incompatibility in the spider mites Tetranychus urticae and T. turkestani. Heredity 79:41–47. Breeuwer, J. A. J., and G. Jacobs. 1996. Wolbachia: Intracellular manipulators of mite reproduction. Exp Appl Acarol 20:421–434. Brooks, M. E., K. Kristensen, K. J. van Benthem, A. Magnusson, C. W. Berg, A. Nielsen, H. J. Skaug, M. Mächler, and B. M. Bolker. 2017. glmmTMB balances speed and flexibility among packages for zero-inflated generalized linear mixed modeling. R J 9:378–400. Brucker, R. M., and S. R. Bordenstein. 2012. Speciation by symbiosis. Trends Ecol Evol 27:443 – 451. Bruzzese, D. J., H. Schuler, T. M. Wolfe, M. M. Glover, J. Mastroni, M. M. Doellman, C. Tait, W. L. Yee, J. Rull, M. Aluja, G. R. Hood, R. Goughnour, C. Stauffer, P. Nosil, and J. L. Feder. 2022. Testing the potential contribution of Wolbachia to speciation when cytoplasmic incompatibility becomes associated with host‐related reproductive isolation. Mol Ecol 31:2935–2950. Chain-Ing, T. S., and S. Sheuan -Ping. 1995. Copulation competition of male Tetranychus urticae Koch and Tetranychus kanzawai Kishida (Acarina: Tetranychidae) with conspecific and heterospecific females and their isolation mechanism. Chin J Entomol 15:239–255. Champion de Crespigny, F. E., R. K. Butlin, and N. Wedell. 2005. Can cytoplasmic incompatibility inducing Wolbachia promote the evolution of mate preferences? J Evol Biol 18:967–977. Champion De Crespigny, F. E., L. D. Hurst, and N. Wedell. 2008. Do Wolbachia-associated incompatibilities promote polyandry? Evolution 62:107–122. .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 12, 2024. ; https://doi.org/10.1101/2024.05.09.593295doi: bioRxiv preprint 19 Clemente, S. H., L. R. Rodrigues, R. Ponce, S. A. M. Varela, and S. Magalhães. 2016. Incomplete species recognition entails few costs in spider mites, despite first-male precedence. Behav Ecol Sociobiol 70:1161–1170. Clemente, S. H., I. Santos, R. Ponce, L. R. Rodrigues, S. A. M. Varela, and S. Magalhães. 2018. Despite reproductive interference, the net outcome of reproductive interactions among spider mite species is not necessarily costly. Behav Ecol 29:321–327. Costa, S. G., S. Magalhães, and L. R. Rodrigues. 2023. Multiple mating rescues offspring sex ratio but not productivity in a haplodiploid exposed to developmental heat stress. Funct Ecol 37:1291–1303. Coughlan, J. M., and D. R. Matute. 2020. The importance of intrinsic postzygotic barriers throughout the speciation process. Phil Trans R Soc B 375:20190533. Coyne, J. A., and H. A. Orr. 1989. Patterns of speciation in Drosophila. Evolution 43:362–381. Coyne, J. A., and H. A. Orr. 1997. “Patterns of speciation in Drosophila” revisited. Evolution 51:295- 303. Coyne, J. A., and H. A. Orr. 2004. Speciation. Sinauer, Sunderland, MA. Crawley, M. J. 2007. The R Book. John Wiley & Sons, Ltd, Chichester. Cruz, M. A., O. Godoy, I. Fragata, V. C. Sousa, S. Magalhães, and F. Zélé. 2023. Sex in the kitchen: non-additive effects of competition for food and reproductive interference on coexistence outcomes between closely related species. doi: 10.1101/2023.11.09.566372. Cruz, M. A., S. Magalhães, É. Sucena, and F. Zélé. 2021. Wolbachia and host intrinsic reproductive barriers contribute additively to postmating isolation in spider mites. Evolution 75:2085–2101. de Boer, R. 1982a. Laboratory hybridization between semi-incompatible races of the arrhenotokous spider mite Tetranychus urticae Koch (Acari: Tetranychidae). Evolution 36:553–560. de Boer, R. 1982b. Partial hybrid sterility between strains of the arrhenotokous spider mite, Tetranychus urticae complex (Acari, Tetranychidae). Genetica 58:23–33. De Crespigny, F. E. C., and N. Wedell. 2006. Wolbachia infection reduces sperm competitive ability in an insect. Proc R Soc B 273:1455–1458. Duron, O., D. Bouchon, S. S. S. Boutin, L. Bellamy, L. Zhou, J. Engelstädter, and G. D. Hurst. 2008. The diversity of reproductive parasites among arthropods: Wolbachia do not walk alone. BMC Biol 6:27. Endler, J., and A. Basolo. 1998. Sensory ecology, receiver biases and sexual selection. Trends Ecol Evol 13:415–420. Engelstädter, J., and S. Charlat. 2006. Outbreeding selects for spiteful cytoplasmic elements. Proc R Soc B 273:923–929. Engelstädter, J., and G. D. D. Hurst. 2009. The ecology and evolution of microbes that manipulate host reproduction. Annu Rev Ecol Evol Syst 40:127–149. Engelstädter, J., and A. Telschow. 2009. Cytoplasmic incompatibility and host population structure. Heredity 103:196–207. Everson, P. R., and J. F. Addicott. 1982. Mate selection strategies by male mites in the absence of intersexual selection by females: a test of six hypotheses. Can J Zool 60:2729–2736. Fortin, M., C. Debenest, C. Souty-Grosset, and F. J. Richard. 2018. Males prefer virgin females, even if parasitized, in the terrestrial isopod Armadillidium vulgare. Ecol Evol 8:3341–3353. Fox, J., and S. Weisberg. 2019. An R Companion to Applied Regression. Third. Sage, Thousand Oaks. .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 12, 2024. ; https://doi.org/10.1101/2024.05.09.593295doi: bioRxiv preprint 20 Fricke, C., and G. Arnqvist. 2004. Conspecific sperm precedence in flour beetles. Anim Behav 67:729–732. Galaverni, M., R. Caniglia, P. Milanesi, S. Lapalombella, E. Fabbri, and E. Randi. 2016. Choosy wolves? Heterozygote advantage but no evidence of MHC-based disassortative mating. J Hered 107:134–142. García‐González, F. 2004. Infertile matings and sperm competition: the effect of “ nonsperm representation” on intraspecific variation in sperm precedence patterns. Am Nat 164:457–472. Gardner, A., and S. A. West. 2004. Spite and the scale of competition. J Evol Biol 17:1195–1203. Giesbers, M. C. W. G., S. Gerritsma, J. Buellesbach, W. Diao, B. A. Pannebakker, L. van de Zande, T. Schmitt, and L. W. Beukeboom. 2013. Prezygotic isolation in the parasitoid wasp genus Nasonia. Pp. 165 –192 in P. Michalak, ed. Speciation: Natural Processes, Genetics and Biodiversity. Nova Biomedical, New York. Gotoh, T., J. Bruin, M. W. Sabelis, and S. B. J. Menken. 1993. Host race formation in Tetranychus urticae: genetic differentiation, host plant preference, and mate choice in a tomato and a cucumber strain. Entomol Exp Appl 68:171–178. Gotoh, T., H. Noda, and X. Y. Hong. 2003. Wolbachia distribution and cytoplasmic incompatibility based on a survey of 42 spider mite species (Acari: Tetranychidae) in Japan. Heredity 91:208– 216. Gotoh, T., J. Sugasawa, H. Noda, and Y. Kitashima. 2007. Wolbachia-induced cytoplasmic incompatibility in Japanese populations of Tetranychus urticae (Acari: Tetranychidae). Exp Appl Acarol 42:1–16. Grether, G. F., K. S. Peiman, J. A. Tobias, and B. W. Robinson. 2017. Causes and consequences of behavioral interference between species. Trends Ecol Evol 32:760–772. Griffith, S. C., S. R. Pryke, and W. A. Buttemer. 2011. Constrained mate choice in social monogamy and the stress of having an unattractive partner. Proc R Soc B 278:2798–2805. Gröning, J., and A. Hochkirch. 2008. Reproductive interference between animal species. Q Rev Biol 83:257–282. Hamilton, W. D. 1970. Selfish and spiteful behaviour in an evolutionary model. Nature 228:1218– 1220. He, Z., H. B. Zhang, S. T. Li, W. J. Yu, J. Biwot, X. Q. Yu, Y. Peng, and Y. F. Wang. 2018. Effects of Wolbachia infection on the postmating response in Drosophila melanogaster. Behav Ecol Sociobiol 72:146. Helle, W. 1967. Fertilization in the two-spotted spider mite (Tetranychus urticae: Acari). Entomol Exp Appl 10:103–110. Helle, W., and H. R. Bolland. 1967. Karyotypes and sex -determination in spider mites (Tetranychidae). Genetica 38:43–53. Helle, W., and C. F. Van de Bund. 1962. Crossbreeding experiments with some species of the Tetranychus urticae group. Entomol Exp Appl 5:159–165. Hill, R. L., and D. J. O’Donnell. 1991. Reproductive isolation between Tetranychus lintearius and two related mites, T. urticae and T. turkestani (Acarina: Tetranychidae). Exp Appl Acarol 11:241–251. Holm, S. 1979. A simple sequentially rejective multiple test procedure. Scand J Statist 6:65–70. Howlett, R. 1988. Sexual selection by female choice in monogamous birds. Nature 332:583–584. Huchard, E., A. Baniel, S. Schliehe‐Diecks, and P. M. Kappeler. 2013. MHC‐disassortative mate choice and inbreeding avoidance in a solitary primate. Mol Ecol 22:4071–4086. .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 12, 2024. ; https://doi.org/10.1101/2024.05.09.593295doi: bioRxiv preprint 21 Jaenike, J., K. A. Dyer, C. Cornish, and M. S. Minhas. 2006. Asymmetrical reinforcement and Wolbachia infection in Drosophila. PLoS Biol 4:1852–1862. Juola, F. A., and D. C. Dearborn. 2012. Sequence -based evidence for major histocompatibility complex-disassortative mating in a colonial seabird. Proc R Soc B 279:153–162. Kaur, R., J. D. Shropshire, K. L. Cross, B. Leigh, A. J. Mansueto, V. Stewart, S. R. Bordenstein, and S. R. Bordenstein. 2021. Living in the endosymbiotic world of Wolbachia: a centennial review. Cell Host Microbe 29:879–893. Kearns, A. M., M. Restani, I. Szabo, A. Schrøder -Nielsen, J. A. Kim, H. M. Richardson, J. M. Marzluff, R. C. Fleischer, A. Johnsen, and K. E. Omland. 2018. Genomic evidence of speciation reversal in ravens. Nat Commun 9:906. Keh, B. 1952. Mating experiments with the two-spotted spider mite complex. J Econ Entomol 45:308–312. Knegt, B., T. Potter, N. A. Pearson, Y. Sato, H. Staudacher, B. C. J. Schimmel, E. T. Kiers, and M. Egas. 2017. Detection of genetic incompatibilities in non -model systems using simple genetic markers: hybrid breakdown in the haplodiploid spider mite Tetranychus evansi . Heredity 118:311–321. Kokko, H., M. D. Jennions, and R. Brooks. 2006. Unifying and testing models of sexual selection. Annu Rev Ecol Evol Syst 37:43-66. Koukou, K., H. Pavlikaki, G. Kilias, J. H. Werren, K. Bourtzis, and S. N. Alahiotis. 2006. Influence of antibiotic treatment and Wolbachia curing on sexual isolationg among Drosophila melanogaster cage populations. Evolution 60:87–96. Krainacker, D. A., and J. R. Carey. 1989. Reproductive limits and heterogeneity of male twospotted spider mites. Entomol Exp Appl 50:209–214. Lackey, A. C. R., and J. W. Boughman. 2017. Evolution of reproductive isolation in stickleback fish. Evolution 71:357–372. Lenth, R., H. Singmann, J. Love, P. Buerkner, and M. Herve. 2018. Emmeans: estimated marginal means, aka least-squares means. Lewis, Z., F. E. C. De Crespigny, S. M. Sait, T. Tregenza, and N. Wedell. 2011. Wolbachia infection lowers fertile sperm transfer in a moth. Biol Lett 7:187–189. Liu, C., J. L. Wang, Y. Zheng, E. J. Xiong, J. J. Li, L. L. Yuan, X. Q. Yu, and Y. F. Wang. 2014. Wolbachia-induced paternal defect in Drosophila is likely by interaction with the juvenile hormone pathway. Insect Biochem Mol Biol 49:49–58. Lu, W., Y. Hu, P. Wei, Q. Xu, C. Bowman, M. Li, and L. He. 2018. Acaricide-mediated competition between the sibling species Tetranychus cinnabarinus and Tetranychus urticae . J Econ Entomol 111:1346–1353. Lu, W., M. Wang, Z. Xu, G. Shen, P. Wei, M. Li, W. Reid, and L. He. 2017. Adaptation of acaricide stress facilitates Tetranychus urticae expanding against Tetranychus cinnabarinus in China. Ecol Evol 7:1233–1249. Magalhães, S., E. Blanchet, M. Egas, and I. Olivieri. 2009. Are adaptation costs necessary to build up a local adaptation pattern? BMC Evol Biol 9:182. Maisonneuve, L., T. Beneteau, M. Joron, C. Smadi, and V. Llaurens. 2021. When do opposites attract? A model uncovering the evolution of disassortative mating. Am Nat 198:625–641. Malogolowkin-Cohen, Ch., A. S. Simmons, and H. Levene. 1965. A study of sexual isolation between certain strains of Drosophila paulistorum. Evolution 19:95–103. .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 12, 2024. ; https://doi.org/10.1101/2024.05.09.593295doi: bioRxiv preprint 22 Matsuda, T., T. Kozaki, K. Ishii, and T. Gotoh. 2018. Phylogeny of the spider mite sub -family Tetranychinae (Acari: Tetranychidae) inferred from RNA-Seq data. PLoS One 13:e0203136. Mayr, E. 1942. Systematics and the Origin of Species. Columbia University Press, New York. Merrell, D. J. 1950. Measurement of Sexual Isolation and Selective Mating. Evolution 4:326–331. Merrill, R. M., B. Van Schooten, J. A. Scott, and C. D. Jiggins. 2011. Pervasive genetic associations between traits causing reproductive isolation in Heliconius butterflies. Proc R Soc B 278:511– 518. Migeon, A., and F. Dorkeld. 2023. Spider Mites Web: a comprehensive database for the Tetranychidae. Miller, W. J., L. Ehrman, and D. Schneider. 2010. Infectious speciation revisited: impact of symbiont- depletion on female fitness and mating behavior of Drosophila paulistorum . PLoS Pathog 6:e1001214. Murtaugh, M. P., and D. L. Wrensch. 1978. Interspecific competition and hybridization between twospotted and carmine spider mites. Ann Entomol Soc Am 71:862–864. Navajas, M., A. Tsagkarakou, J. Lagnel, and M. J. Perrot -Minnot. 2000. Genetic differentiation in Tetranychus urticae (Acari: Tetranychidae): polymorphism, host races or sibling species? Exp Appl Acarol 24:365–376. Noor, M. A. F. 1999. Reinforcement and other consequences of sympatry. Heredity 83:503–508. Noriyuki, S., N. Osawa, and T. Nishida. 2012. Asymmetric reproductive interference between specialist and generalist predatory ladybirds. J Anim Ecol 81:1077–1085. Oku, K. 2014. Sexual selection and mating behavior in spider mites of the genus Tetranychus (Acari: Tetranychidae). Appl Entomol Zool 49:1–9. Ortiz-Barrientos, D., A. Grealy, and P. Nosil. 2009. The genetics and ecology of reinforcement: implications for the evolution of prezygotic isolation in sympatry and beyond. Ann NY Acad Sci 1168:156–182 Parker, G. A., and L. Partridge. 1998. Sexual conflict and speciation. Philos Trans R Soc Lond B 353:261–274. Perrot-Minnot, M. J., B. Cheval, A. Migeon, and M. Navajas. 2002. Contrasting effects of Wolbachia on cytoplasmic incompatibility and fecundity in the haplodiploid mite Tetranychus urticae. J Evol Biol 15:808–817. Potter, D. A., and D. L. Wrensch. 1978. Interrupted matings and the effectiveness of second inseminations in the twospotted spider mite. Ann Entomol Soc Am 71:882–885. Potter, D. A., D. L. Wrensch, and D. E. Johnston. 1976. Aggression and mating success in male spider mites. Science 193:160–161. Price, C. S. C. 1997. Conspecific sperm precedence in Drosophila. Nature 388:663–666. Ramsey, J., H. D. Bradshaw, and D. W. Schemske. 2003. Components of reproductive isolation between the monkeyflowers Mimulus lewisii and M. cardinalis (Phrymaceae). Evolution 57:1520–1534. Richard, F. J. 2017. Symbiotic bacteria influence the odor and mating preference of their hosts. Front Ecol Evol 5:143. Rodrigues, L. R., A. R. T. Figueiredo, T. Van Leeuwen, I. Olivieri, and S. Magalhães. 2020. Costs and benefits of multiple mating in a species with first -male sperm precedence. J Anim Ecol 89:1045–1054. .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 12, 2024. ; https://doi.org/10.1101/2024.05.09.593295doi: bioRxiv preprint 23 Rodrigues, L. R., A. R. T. Figueiredo, S. A. M. Varela, I. Olivieri, and S. Magalhães. 2017. Male spider mites use chemical cues, but not the female mating interval, to choose between mates. Exp Appl Acarol 71:1–13. Rodrigues, L. R., F. Zélé, I. Santos, and S. Magalhães. 2022. No evidence for the evolution of mating behavior in spider mites due to Wolbachia-induced cytoplasmic incompatibility. Evolution 76:623–635. Rolán-Alvarez, E., J. Erlandsson, K. Johannesson, and R. Cruz. 1999. Mechanisms of incomplete prezygotic reproductive isolation in an intertidal snail: testing behavioural models in wild populations. J Evol Biol 12:879–890. Sato, Y., and J. M. Alba. 2020. Reproductive interference and sensitivity to female pheromones in males and females of two herbivorous mite species. Exp Appl Acarol 81:59–74. Sato, Y., J. M. Alba, and M. W. Sabelis. 2014. Testing for reproductive interference in the population dynamics of two congeneric species of herbivorous mites. Heredity 113:495–502. Sato, Y., H. Staudacher, and M. W. Sabelis. 2016. Why do males choose heterospecific females in the red spider mite? Exp Appl Acarol 68:21–31. Satoh, Y., S. Yano, and A. Takafuji. 2001. Mating strategy of spider mite, Tetranychus urticae (Acari: Tetranychidae) males: postcopulatory guarding to assure paternity. Appl Entomol Zool 36:41– 45. Schneider, D. I., L. Ehrman, T. Engl, M. Kaltenpoth, A. Hua -Van, A. Le Rouzic, and W. J. Miller. 2019. Symbiont -driven male mating success in the neotropical Drosophila paulistorum superspecies. Behav Genet 49:83–98. Servedio, M. R. 2001. Beyond reinforcement: the evolution of premating isolation by direct selection on preferences and postmating, prezygotic incompatibilities. Evolution 55:1909–1920. Servedio, M. R., and J. Hermisson. 2020. The evolution of partial reproductive isolation as an adaptive optimum. Evolution 74:4–14. Servedio, M. R., and M. Noor. 2003. The role of reinforcement in speciation: theory and data. Annu Rev Ecol Evol Syst 34:339–364. Servedio, M. R., and G.-P. Sætre. 2003. Speciation as a positive feedback loop between postzygotic and prezygotic barriers to gene flow. Proc R Soc Lond B 270:1473–1479. Shoemaker, D. D., V. Katju, and J. Jaenike. 1999. Wolbachia and the evolution of reproductive isolation between Drosophila recens and Drosophila subquinaria. Evolution 53:1157–1164. Shropshire, J. D., and S. R. Bordenstein. 2016. Speciation by symbiosis: the microbiome and behavior. mBio 7:e01785-15. Shropshire, J. D., B. Leigh, and S. R. Bordenstein. 2020. Symbiont -mediated cytoplasmic incompatibility: what have we learned in 50 years? eLife 9:e61989. Snook, R. R., S. Y. Cleland, M. F. Wolfner, and T. L. Karr. 2000. Offsetting effects of Wolbachia infection and heat shock on sperm production in Drosophila simulans: analyses of fecundity, fertility and accessory gland proteins. Genetics 155:167-178. Stankowski, S., and M. Ravinet. 2021. Defining the speciation continuum. Evolution 75:1256–1273. Sugasawa, J., Y. Kitashima, and T. Gotoh. 2002. Hybrid affinities between the green and the red forms of the two -spotted spider mite Tetranychus urticae (Acari: Tetranychidae) under laboratory and seminatural conditions. Appl Entomol Zool 37:127–139. Suh, E., C. Sim, J. J. Park, and K. Cho. 2015. Inter -population variation for Wolbachia induced reproductive incompatibility in the haplodiploid mite Tetranychus urticae. Exp Appl Acarol 65:55–71. .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 12, 2024. ; https://doi.org/10.1101/2024.05.09.593295doi: bioRxiv preprint 24 Taylor, E. B., J. W. Boughman, M. Groenenboom, M. Sniatynski, D. Schluter, and J. L. Gow. 2006. Speciation in reverse: morphological and genetic evidence of the collapse of a three -spined stickleback (Gasterosteus aculeatus) species pair. Mol Ecol 15:343–355. Telschow, A., P. Hammerstein, and J. H. Werren. 2005. The effect of Wolbachia versus genetic incompatibilities on reinforcement and speciation. Evolution 59:1607–1619. Tien, N. S. H., G. Massourakis, M. W. Sabelis, and M. Egas. 2011. Mate choice promotes inbreeding avoidance in the two-spotted spider mite. Exp Appl Acarol 54:119–124. Turissini, D. A., J. A. McGirr, S. S. Patel, J. R. David, and D. R. Matute. 2018. The rate of evolution of postmating-prezygotic reproductive isolation in Drosophila. Mol Biol Evol 35:312–334. Vala, F., M. Egas, J. A. J. Breeuwer, and M. W. Sabelis. 2004. Wolbachia affects oviposition and mating behaviour of its spider mite host. J Evol Biol 17:692–700. Vala, F., A. Weeks, D. Claessen, J. A. J. Breeuwer, and M. W. Sabelis. 2002. Within - and between- population variation for Wolbachia-induced reproductive incompatibility in a haplodiploid mite. Evolution 56:1331–1339. Van de Bund, C. F., and W. Helle. 1960. Investigations on the Tetranychus urticae complex in north west Europe (Acari: Tetranychidae). Entomol Exp Appl 3:142–156. van den Berg, M. J., G. Thomas, H. Hendriks, and W. van Delden. 1984. A reexamination of the negative assortative mating phenomenon and its underlying mechanism in Drosophila melanogaster. Behav Genet 14:45–61. Villacis-Perez, E., S. Snoeck, A. H. Kurlovs, R. M. Clark, J. A. J. Breeuwer, and T. Van Leeuwen. 2021. Adaptive divergence and post -zygotic barriers to gene flow between sympatric populations of a herbivorous mite. Commun Biol 4:853. Wagner, W. E. 1998. Measuring female mating preferences. Anim Behav 55:1029–1042. Weinert, L. A., E. V. Araujo -Jnr, M. Z. Ahmed, and J. J. Welch. 2015. The incidence of bacterial endosymbionts in terrestrial arthropods. Proc R Soc B 282:20150249. Werren, J. H., L. Baldo, and M. E. Clark. 2008. Wolbachia: master manipulators of invertebrate biology. Nat Rev Microbiol 6:741–751. Wittenberger, J. F., and R. L. Tilson. 1980. The evolution of monogamy: hypotheses and evidence. Annu Rev Ecol Syst 11:197–232. Wybouw, N., E. Van Reempts, J. Zarka, F. Zélé, and D. Bonte. 2023. Egg provisioning explains the penetrance of symbiont-mediated sex allocation distortion in haplodiploids. Heredity 131:221– 229. Xie, L., H. Miao, and X.-Y. Hong. 2006. The two-spotted spider mite Tetranychus urticae Koch and the carmine spider mite Tetranychus cinnabarinus (Boisduval) in China mixed in their Wolbachia phylogenetic tree. Zootaxa 1165:33–46. Yoshioka, T., and S. Yano. 2014. Do Tetranychus urticae males avoid mating with familiar females? J Exp Biol 217:2297–2300. Yukilevich, R., L. S. Maroja, K. Nguyen, S. Hussain, and P. Kumaran. 2018. Rapid sexual and genomic isolation in sympatric Drosophila without reproductive character displacement. Ecol Evol 8:2852–2867. Zélé, F., I. Santos, M. Matos, M. Weill, F. Vavre, and S. Magalhães. 2020. Endosymbiont diversity in natural populations of Tetranychus mites is rapidly lost under laboratory conditions. Heredity 124:603–617. .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 12, 2024. ; https://doi.org/10.1101/2024.05.09.593295doi: bioRxiv preprint 25 Zélé, F., I. Santos, I. Olivieri, M. Weill, O. Duron, and S. Magalhães. 2018a. Endosymbiont diversity and prevalence in herbivorous spider mite populations in South -Western Europe. FEMS Microbiol Ecol 94:fiy015. Zélé, F., J. L. Santos, D. P. Godinho, and S. Magalhães. 2018b. Wolbachia both aids and hampers the performance of spider mites on different host plants. FEMS Microbiol Ecol 94: fiy187. Zélé, F., M. Weill, and S. Magalhães. 2018c. Identification of spider -mite species and their endosymbionts using multiplex PCR. Exp Appl Acarol 74:123–138. Zhang, Y.-K., Y.-T. Chen, K. Yang, G.-X. Qiao, and X. -Y. Hong. 2016. Screening of spider mites (Acari: Tetranychidae) for reproductive endosymbionts reveals links between co-infection and evolutionary history. Sci Rep 6:27900. Zhang, Y. -K., K. -J. Zhang, J. -T. Sun, X. -M. Yang, C. Ge, and X. -Y. Hong. 2013. Diversity of Wolbachia in natural populations of spider mites (genus Tetranychus): evidence for complex infection history and disequilibrium distribution. Microb Ecol 65:731–739. Zhao, D. -X., X. -F. Zhang, D. -S. Chen, Y. -K. Zhang, and X. -Y. Hong. 2013. Wolbachia-host interactions: host mating patterns affect Wolbachia density dynamics. PLoS One 8:e66373. Zhou, P., X. Z. He, C. Chen, and Q. Wang. 2020. No evidence for inbreeding depression and inbreeding avoidance in a haplodiploid mite Tetranychus ludeni Zacher. Syst Appl Acarol 25:1723–1728. .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 12, 2024. ; https://doi.org/10.1101/2024.05.09.593295doi: bioRxiv preprint 26 Table 1. Choice tests to assess the mating behaviour and preference of male s or females that were given the choice between two mates of different colour forms and/or Wolbachia infection status. CI: cytoplasmic incompatibility; HI: host-associated incompatibility. Category Choice Female tests Male tests I – Avoidance of Wolbachia- induced CI Mates with different infection statuses Ru♀ x (Ri♂ vs Ru♂) Gu♀ x (Gi♂ vs Gu♂) Ri♂ x (Ri♀ vs Ru♀) Gi♂ x (Gi♀ vs Gu♀) II – Avoidance of HI in absence of Wolbachia Uninfected mates from different populations Ru♀ x (Ru♂ vs Gu♂) Gu♀ x (Ru♂ vs Gu♂) Ru♂ x (Ru♀ vs Gu♀) Gu♂ x (Ru♀ vs Gu♀) III – Avoidance of HI in presence of Wolbachia Infected mates from different populations Ri♀ x (Ri♂ vs Gi♂) Gi♀ x (Ri♂ vs Gi♂) Ri♂ x (Ri♀ vs Gi♀) Gi♂ x (Ri♀ vs Gi♀) IV – Avoidance of HI, Wolbachia-induced CI, or both Mates both with different infection statuses and from different populations Ru♀ x (Ri♂ vs Gi♂) Gu♀ x (Ri♂ vs Gi♂) Ri♂ x (Ru♀ vs Gu♀) Gi♂ x (Ru♀ vs Gu♀) Table 2. No-choice tests to assess the behaviour and offspring production of virgin females (♀) placed with a compatible or incompatible male (1 st ♂ of a similar or different colour form and/or Wolbachia infection status, respectively), then (for those that had mated with the first male) with a second compatible male (2 nd ♂ from their own population). CI: cytoplasmic incompatibility; HI: host-associated incompatibility. Category First crosses ♀ 1st ♂ 2nd ♂ A – Controls Intra-population crosses between uninfected ♀ and ♂ Ru Ru Ru Gu Gu Gu B – HI but no effect on F1 production1 Inter-population crosses between uninfected red ♀ and green ♂ Ru Gu Ru C – HI with reduced F1 production Inter-population crosses between uninfected green ♀ and red ♂ Gu Ru Gu D – Wolbachia-induced CI2 Intra-population crosses between red uninfected ♀ and infected ♂ Ru Ri Ru E – HI and Wolbachia-induced CI2 Inter-population crosses between green uninfected ♀ and red infected ♂ Gu Ri Gu 1HI in this cross direction leads to F1 female sterility and hybrid breakdown (i.e., late post-zygotic isolation). 2Only Ri males were used as only the Wolbachia strain infecting the red population induces CI (Cruz et al. 2021). .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 12, 2024. ; https://doi.org/10.1101/2024.05.09.593295doi: bioRxiv preprint 27 Figure 1. Mating propensity and mate choice of spider mites of different colour forms and/or Wolbachia infection status. For each type of choice test, bars represent mean (± s.e.) proportion of mated (a) females and (c) males, and of mates chosen by (b) females and (d) males ( dotted: Wolbachia- infected mates; plain: uninfected mates; orange: red mates; blue: green mates). Identical or absent superscripts indicate non-significant differences at the 5% level among treatments, asterisks indicate a difference to random mating (white dotted line). .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 12, 2024. ; https://doi.org/10.1101/2024.05.09.593295doi: bioRxiv preprint 28 Figure 2. Latency to copulation (a, b) and copulation duration (c, d) of virgin mites during the choice and no-choice tests. Dots represent mean time (± s.e.) in seconds observed for males (a, c) and females (b, d) in the male and female choice tests (white and grey dots, respectively), and in the first mating event of the no-choice test (black dots). Overall, no significant differences were found among males or females for latency to copulation in the choice tests (statistical results are not given for the no-choice test as latencies to copulation exceeding 30 minutes were excluded from the means displayed in this figure to allow comparisons across experiments). For copulation duration, identical superscripts indicate non-significant differences at the 5% level among crosses within each test (Italic: male choice test, lowercase: female choice test, uppercase: 1 st mating event of the no -choice test). Note that infected females were not used in the no -choice test (hence, black dots are not displayed for Ri and Gi females in panel d). .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 12, 2024. ; https://doi.org/10.1101/2024.05.09.593295doi: bioRxiv preprint 29 Figure 3. Mating propensity observed in two successive mating events in the no-choice test. For each cross category, circles and diamonds represent mean (± s.e.) proportion of females that mated during the first and the second mating event, respectively. The population of the female is displayed at the bottom level of the x -axis and the population of the first male at the top level (the population of the second male is always the same as that of the female ). Identical superscripts indicate non - significant differences at the 5% level among crosses across mating events. .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 12, 2024. ; https://doi.org/10.1101/2024.05.09.593295doi: bioRxiv preprint 30 Figure 4. Latency to copulation (a, b) and copulation duration (c, d) observed in two successive mating events in the no-choice test. In (a) and (c), circles and diamonds represent mean time (± s.e.) in seconds for each cross category during the first and the second mating event, respectively. Identical or absent superscripts indicate non-significant differences at the 5% level among crosses within each mating event. In (b) and (d), squares represent the mean time difference (± s.e.) observed between the two mating events for each female that mated with both males (i.e. [time spent for the second mating] – [time spent for the first mating]). Superscripts indicate significant differences from zero at the 10% level (‡: p<0.10) and at the 5% level ( *: p≤0.05). In all panels, t he population of the female is displayed at the bottom of the x-axis and the population of the first male at the top (the population of the second male is always the same as that of the female). .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 12, 2024. ; https://doi.org/10.1101/2024.05.09.593295doi: bioRxiv preprint 31 Figure 5. Effect of re-mating on offspring production in crosses affected by Wolbachia-induced CI and/or HI. (a) Outcome of egg development from each cross category, with bars representing the mean (± s.e.) relative proportions of unhatched eggs ( i.e. embryonic mortality), adult daughters and sons. (b) Boxplot of the proportion of males produced in all crosses relative to control crosses (MDcorr). (c) Boxplot of the proportion of estimated unhatched female eggs relative to control crosses (FMcorr). (d) Boxplot of the proportion of F1 adult females in the brood (FP). The population of the female (♀) is displayed at the bottom level of the x-axis, that of the first male (1st ♂) at the middle level, and that of the second male (2nd ♂) at the top level (“-” indicates females that did not mate with the second male). .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 12, 2024. ; https://doi.org/10.1101/2024.05.09.593295doi: bioRxiv preprint 32 Figure 6. Contribution of the different reproductive barriers to reducing gene flow within and between populations. Percent contributions to total reproductive isolation were computed based on the estimated strength of reproductive isolation (RI) caused by a given reproductive barrier. They are shown for the six most representative types of cross in this system (see Table S11 for all crosses). Mate preference corresponds to male preferences only, as females overall showed no mate preferences. ‘ns’ indicates no significant difference to zero at the 5% level. .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 12, 2024. ; https://doi.org/10.1101/2024.05.09.593295doi: bioRxiv preprint