Hosts generally employ three strategies to reduce either exposure or susceptibility to parasites, i.e., by avoiding encounters with parasites, suppressing parasite proliferation and/or minimizing parasite damage (Råberg et al., 2009; Medzhitov et al., 2012; Hart and Hart, 2018). While parasite avoidance is largely limited to behavioral modifications, preventing parasite proliferation and reducing damage also requires physiological responses that are more costly (Viney et al., 2005; Boots et al., 2009; Medzhitov et al., 2012).

Cremer and colleagues suggested that singular breeding social insects (i.e., eusocial Hymenoptera and termites) achieve parasite protection at a group-level through the sum of defenses employed by group members that affect the exposure and susceptibility of social hosts to parasites and called this “social immunity” (Cremer et al., 2007, 2018). Similar to individual host defenses, social immunity can be distinguished into avoidance, resistance and tolerance. In addition to a behavioral and physiological component it has an organizational components that is unique to social organisms (Cremer et al., 2018). While aspects of social immunity have been the focus of many eusocial insects (Cremer et al., 2007, 2018; Stroeymeyt et al., 2014; Schmid-Hempel, 2021), its applicability to social vertebrate species has not been explored (Van Meyel et al., 2018).

There are three host traits that determine the within host-success of a parasite: the immune response of the host, host predictability and the presence of other parasites or symbionts that may compete for host resources (Schmid-Hempel, 2021).

Parasites are required to complete three steps to achieve between-host fitness gains starting with leaving the current host (Schmid-Hempel, 2021). While this step likely does not differ fundamentally between solitary and social hosts, the subsequent transmission steps (i.e., encountering and infecting) to a new hosts are characterized by substantial differences in general transmission distance (Schmid-Hempel, 2021; Figure 2).

As behavioral measures can substantially affect parasite transmission and are highly flexible at comparatively low costs, selection pressures from parasites should act strongly on behavior (Ezenwa et al., 2016; Hawley et al., 2021). The presumed higher transmission in larger groups selects for smaller group sizes (Altizer et al., 2003; Rifkin et al., 2012; Patterson and Ruckstuhl, 2013). Under such circumstances selection can instead favor preferential social interactions with certain individuals that can ultimately lead to modularity to reduces parasites transmission (Freeland, 1976, 1979; Griffin and Nunn, 2012; Nunn et al., 2015). However, evidence for this is weak possibly due to the opposing selection pressures associated with the benefits of group-living including lower predation pressure, increased foraging efficiency, transfer of protective microbes and information that may counter pressures on reductions of group size (Ezenwa et al., 2016; Townsend et al., 2020; Hawley et al., 2021). In social bathyergids the ecological constraints posed by the subterranean niche already generate between-group modularity (Figure 2C). In addition to the benefits of improved foraging efficiency in groups, ecological and behavioral factors have likely played an important role in bathyergids in tipping the balance in favor of group-living (Faulkes and Bennett, 2021). Ecological constraints to dispersal likely have also relaxed selection pressures on avoiding encounters with parasites through infected intruders. This could also account for the lack of avoidance of odors from infected males observed in highveld mole-rats and the readiness of females of several bathyergid species to engage in copulations with unfamiliar males (Lutermann et al., 2022).

Modularity can occur between as well as within groups (e.g., organizational immunity) and the selection pressure exerted by parasites is likely to contribute to modularity at a colony level (Figure 2D). The predominance of within-group transmission in social bathyergids suggests that selection for organizational immunity, including the reproductive division of labor, should be a major evolutionary trajectory in this family. Avoidance and hygienic behaviors can be observed across the full spectrum of sociality (Meunier, 2015; Van Meyel et al., 2018). This often leads to asymmetries in inter-individual contact rates which can differ widely across a range of social systems from solitary to eusocial (Sah et al., 2018). Such modularity between less connected sub-groups is common among social species and increases with group size across species (Griffin and Nunn, 2012; Sah et al., 2018; Evans et al., 2021). At the same time, parasite-mediated selection should favor greater investment in hygienic behaviors including allo-grooming, burrow hygiene or the removal/isolation of infected group members, but reduce the incidence of social behaviors that can increase parasite transmission such as agonistic behaviors (Hawley et al., 2021). This is more likely to evolve in kin groups where inclusive fitness benefits reduce competition for resources. However, the evolution of full organizational immunity, including the reproductive division of labor, is likely constrained in species with multiple breeders due to the lack of inclusive fitness benefits as a result of lower relatedness. In contrast, in closely related hosts, parasites can favor the evolution of singular breeding and organizational immunity as has been proposed for eusocial insects (Naug and Camazine, 2002; Biedermann and Rohlfs, 2017; Cremer et al., 2018). Similarly, parasites could have generated an important selection pressure for the monogamous ancestors of social mole-rats to invest into organizational immunity in response to parasitism due to the high dispersal barriers for offspring (Figure 2D). Group sizes differ substantially between social bathyergids and social insects with the latter having group sizes that may exceed those of mole-rats by three orders of magnitude. This is relevant as implementing and maintaining organizational immunity may be constrained by group size. However, it has recently been shown for singular breeding ambrosia beetles (Xyleborinus saxesenii) living in small groups, that offspring delay breeding dispersal in the presence of microparasite infection (Nuotclà et al., 2019). The additional investment in hygienic behaviors by philopatric offspring significantly reduced the detrimental effects of parasite infection and hence, lowered the costs of sociality. If similar effects can be accrued by non-breeding group members in social bathyergids this could also account for the philopatry in bathyergid species where dispersal constraints are relaxed (e.g., F. mechowii or F. anselli). At the same time, such benefits would be an additional incentive for offspring in species with high dispersal constraints (e.g., H. glaber) to forgo investment into reproduction and hence, parasites may also have contributed to the evolution of singular breeding and reproductive suppression in Bathyergidae. The division of labor for other tasks has been questioned for social bathyergids (Lacey and Sherman, 1991; Thorley et al., 2018; Siegmann et al., 2021). Recent mathematical models suggest significant fitness benefits for such division of labor (Udiani and Fefferman, 2020). However, this only applied in the presence of parasites when age-based division of labor reaped more benefits than fixed task specialization. Consequently, some forms of organizational immunity may be more flexible in bathyergids.

While both resistance and tolerance are strategies available to solitary as well as social species and although parasite burden is frequently assumed to be larger in social species (Altizer et al., 2003; Rifkin et al., 2012; Patterson and Ruckstuhl, 2013), from the arguments outlined above it is not clear that social species should exhibit a greater investment in immune defenses than solitary species (Schmid-Hempel, 2021 and references therein, Figure 2). Comparative studies in social bees and cooperatively breeding birds suggest a greater investment in immune defenses in the latter (Stow et al., 2007; Spottiswoode, 2008). However, such increased investment in immunity is energetically costly and should more likely be favored in singular breeding species as breeders can benefit from the contributions of related non-breeders (e.g., greater access to food, less investment in foraging or parental care) that in turn increase their inclusive fitness in this way. Conversely, tolerance may carry fitness costs to the host as it does not affect the parasite burden (Medzhitov et al., 2012; Budischak and Cressler, 2018). Once again, the benefits of group-living in the form of greater access to resources can, however, alleviate these costs as has been shown for other singular breeding species (Almberg et al., 2015). Similarly, theoretical work suggests that high burdens of parasites with low virulence should select for hosts to employ tolerance rather than resistance as a defense strategy (Bonds et al., 2005). Thus, organizational immunity can provide effective protection against parasites in singular breeders while additional investment in immune response may not be required or even be lower in singular breeders (Figure 2).

Due to the reproductive division of labor the effective population size of singular breeding species is smaller than in communally breeding or solitary species and this can cause a loss of genetic diversity (Hoban et al., 2020). For social insects this results in reduced negative or purifying selection across the genome (Imrit et al., 2020). Similarly, there is evidence for a link between effective population size and selection of genes involved in parasite recognition and immune activation in Bathyergidae. Kundu and Faulkes (2004) report evidence for purifying selection of exon 3 of the MHC II BLA-DQα1 gene, coding for the transmembrane protein not directly involved with parasite recognition, from solitary (Heliophobius argenteocinereus) to social bathyergids (Cryptomys hottentotus hottentotus) that was less pronounced in the two species with the largest reproductive skew (F. damarensis and H. glaber) and the smallest effective population sizes. At the same time, evidence for positive selection on exon 2, coding for the antigen recognition site, was stronger in C. h. hottentus compared to the solitary as well as the other social species (F. damarensis and H. glaber; Kundu and Faulkes, 2004). Furthermore, high mortalities observed in laboratory colonies of H. glaber as a result of artificial viral infections suggest that immune responses may be weaker in the species with the lowest effective population size (Ross-Gillespie et al., 2007; Artwohl et al., 2009). Conversely, investment in immune priming via social cues from infested conspecifics should be strongly selected for in social bathyergids to reduce within-group transmission (Schmid-Hempel, 2021).

These examples illustrate that despite the costs of parasitism, selection against group-living should be reduced for singular breeding Bathyergidae. At the same time, selection on behaviors reducing parasite transmission, particularly within colonies such as organizational immunity, should be stronger than selection on physiological immunity, particularly if they provide protection from a range of parasite taxa (Hawley et al., 2021). The kin structure of social bathyergids is a key element affecting the inclusive fitness benefits of such anti-parasite strategies and similar advantageous effects cannot be expected for plural breeders.

Social behavior of hosts can have profound effects on the population structure and evolution of virulence of parasites (i.e., parasite-induced harm to the host) (Ezenwa et al., 2016; Hawley et al., 2021). The genetic diversity and effective population size of a parasite increases in gregarious species with large groups that exhibit only weak modularity (Figure 2D). Such species also tend to sustain a larger diversity of parasite species (Côté and Poulin, 1995; Rifkin et al., 2012; Patterson and Ruckstuhl, 2013). However, in species with high modularity, such as in social bathyergids, parasite transmission is greatly impeded and consequently the genetic population structure of parasites should be much more distinct and the effective population size smaller (Hawley et al., 2021; Figure 2D). In host species that exhibit organizational immunity these effects are further exacerbated as it constraints parasite transmission within groups (Cremer et al., 2018). This should also affect patterns of parasite aggregation (Figure 2D). Parasite populations are often characterized by a skewed distribution, also called overdispersion, with a small number of host individuals sustaining the majority of parasites (Woolhouse et al., 1997; Wilson et al., 2002). While such skews may be absent in host populations with frequent contact or close proximity between host individuals, it should be pronounced in hosts where inter-individual contact rates are low (Hawley et al., 2021). These effects are illustrated for bathyergid parasites by observations that group membership is a good predictor of parasite infection (Viljoen et al., 2011a; Lutermann et al., 2013; Archer et al., 2016). Evidence for similar effects at a colony level due to organizational immunity are less clear, but differences in parasite burden between breeders and non-breeders have been reported for some social bathyergids (Viljoen et al., 2011b; Lutermann et al., 2013; Archer et al., 2016). Overdispersion of parasites among hosts results in the aggregation of parasites on certain host individuals that in turn leads to increased competition between these parasites for host resources and greater variance in reproductive success of competing parasite individuals (Poulin, 2007; Hawley et al., 2021).

These effects of host behavior on parasite population structure and aggregation also affect selection on parasite virulence (Hawley et al., 2021; Schmid-Hempel, 2021). Reductions in host connectivity or increased modularity reduce transmission rates, conditions also favored by low between-group contacts and organizational immunity, select for greater parasite virulence (Ebert, 1998; Figures 2C,D). However, in host populations with distinct modularity high virulence effectively results in a depletion of hosts. This would negatively affect parasite fitness as observed in two incidences of viral infections of naked mole-rats in captivity (Ross-Gillespie et al., 2007; Artwohl et al., 2009). These fitness implications would exert selection pressures for the evolution of low virulence on parasites (Hawley et al., 2021), a scenario also applicable to social bathyergids. The fragmentation of parasite populations due to organizational immunity could furthermore lead to reductions in parasite genome size and possibly loss of functional abilities (Figure 2D) as observed in lineages that become commensals in eusocial insects (Leggett et al., 2013; Conlon et al., 2021). In highly modular social hosts dispersal constraints would also result in closely related parasite individuals exploiting hosts. Due to the inherent inclusive fitness benefits of such a scenario selection should favor decreased transmission rates and lowered virulence to reduce kin competition among such parasites (Hawley et al., 2021 and references therein). Thus, high modularity generated by low between-group contact rates (Figure 2C) as well as behavioral barriers to within-group transmission (Figure 2D) should lower the optima for both transmission and virulence for parasites exploiting hosts such as social Bathyergidae. At the same time, this can be expected to result in lower co-infection rates, further relaxing selection for virulence in co-infecting parasites (Hawley et al., 2021). In addition, the long life expectancy observed in social bathyergids and the related increased opportunities for transmission should lower selection pressures on parasite virulence (Bonds et al., 2005).

If social interactions of hosts are biased in favor of kin, parasites transmitted between such hosts exploit hosts with a similar genetic make-up. In singular breeding species with large group sizes (e.g., naked mole-rats) this does not only resemble the conditions experienced during the serial passage of parasites through the same host (Ebert, 1998), but it is also similar to those encountered by parasites in monocultures which facilitate transmission between hosts (Baer and Schmid-Hempel, 1999; Altermatt and Ebert, 2008; Ekroth et al., 2019). Ultimately, the high degree of relatedness between hosts, as observed within groups of social bathyergids, can favor the evolution of host specialization (Figure 2C) as successful transmission and proliferation in alternative hosts is no longer required (Kassen, 2002; Bono et al., 2017). In its extreme form such specialization can not only increase parasite fitness but can be a route for the evolution of benign symbionts (Hughes et al., 2008; Biedermann and Rohlfs, 2017; Figure 2D). Similarly, host predictability should facilitate host specialization in order to increase parasite fitness (Combes, 2001). This possibility has received limited attention in the literature to date (Hawley et al., 2021).

Selection pressures experienced by parasites will also differ with the immune strategy employed by the host; since host resistance reduces parasite fitness this strategy should result in selection for increased virulence in parasites. Empirical evidence of the benefits of sociality in the form of greater access to resources suggest that social species do not necessarily exhibit more investment in immunity thus relaxing selection for virulence in parasites (Almberg et al., 2015; Ezenwa and Snider, 2016). Both strategies might be equally employed by host individuals regardless of their social system and from an evolutionary perspective the mean and variance across a host population rather than individual differences are relevant for selection (Schmid-Hempel, 2021).

Based on the arguments laid out above, parasites of social mole-rats should experience population bottlenecks that reduce their species and genetic diversity while selecting for reductions in virulence, but possibly high transmission and host specialization (Figure 2). In fact, for bathyergids there is some evidence for such evolutionary forces having acted on mesostigmatid mites of the genus Androlaelaps. Although they are not necessarily restricted to social bathyergids, the majority of the species parasitizing Bathyergidae have not been reported for any other host family (Lutermann et al., 2019). Overall, it appears that selection pressures exerted by behavioral strategies employed by social bathyergids should be stronger than those posed by immune strategies as the latter are less likely to differ from those encountered in solitary hosts or those with multiple breeders.

The forces of selection acting on both hosts and parasites are also dependent on the mode of transmission, mobility and the type of life-cycle of a parasite with the former likely evolving in response to host defenses that also affect parasite virulence (Poulin, 2007; Antonovics et al., 2017; Hawley et al., 2021; Schmid-Hempel, 2021). The mode of transmission (i.e., method used by parasite) can be either vertical or horizontal. Horizontal transmission can further be distinguished into direct transmission via physical contact, airborne (i.e., micrcoparasites), or indirect, i.e., environmentally (e.g., contaminated food, soil or water) or vector-borne (Antonovics et al., 2017). For many parasites more than one mode of transmission may be used and the relative importance of each mode in a particular host-parasite system will be important for their evolution (Antonovics et al., 2017). Generally, parasites with a vertical mode of transmission decrease in virulence because only surviving offspring that reproduces can further transmit the parasite while those with horizontal mode of transmission increase in virulence (Ebert, 2013; Antonovics et al., 2017). However, the kin structure of social bathyergids makes distinctions between these two routes challenging. On the one hand, although vertical transmission is easily possible, the high reproductive skew in mole-rat societies means that the majority of parent-offspring transmissions will not be successful for parasites with obligate vertical transmission as these offspring will never breed and are thus unable to transmit the parasite to their progeny. At the same time, within a colony, transmission will mostly happen between colony members due to their close physical and genetic proximity (see details above). Although this is technically horizontal transmission, the distinction from vertical transmission is largely formal. In contrast, between-colony transmission is clearly horizontal but occurs much less frequently. Due, to low between-group contact rates, lack of a clear distinction between vertical and horizontal mode and the limited number of group members in the case of social bathyergid hosts, the evolution of virulence of parasites using a horizontal mode should be constrained for parasites (Hawley et al., 2021). This would be further exacerbated by organizational immunity that leads to modularity and hence, further constraints to transmission within bathyergid groups. Similarly, the monogamous mating strategy of social Bathyergidae is an effective measure against directly transmitted parasites during sexual contacts (Antonovics et al., 2017).

While horizontally transmitted parasites requiring direct contact between hosts, particularly virulent ones, should constrain the evolution of sociality the dilution effects provided by group members favor selection of sociality in the presence of mobile parasites that may also act as vectors for microparasites (Hart and Hart, 2018; Hawley et al., 2021). However, the subterranean environment effectively eliminates vector-borne microparasites relying on mobile vectors (e.g., mosquitoes) for bathyergids (Figure 2B). Consequently, selection on defenses against these types of parasites should be low in social bathyergids. While no flying vectors are known for Bathyergidae, a number of relatively mobile vectors such as fleas have been reported for several bathyergids (Lutermann et al., 2019; Fagir et al., 2021). Fleas can either be directly transmitted between hosts or via shared space use such as nests (Krasnov, 2008). Intriguingly, roughly half the flea species observed for bathyergids to date are host generalists and vectors of zoonotic pathogens while the other half appear to be host specialists that only exploit social bathyergids (Table 1). If opportunities for between-group transmission are limited, even for relatively mobile parasites, such specialization should be favored by selection. This should also remove opportunities for such vectors to contract microparasites. Thus, it may not be surprising that screenings for Bartonella spp., that are most likely vectored by fleas, have been negative for social common and Damaraland mole-rats but not solitary B. suillus (van Sandwyk, 2007). However, solitary G. capensis were also negative for these microparasites. Only two tick species, less mobile vectors with environmental transmission, have been reported for solitary Bathyergidae and both species appear to be host generalists parasitizing a range of small mammals (Horak et al., 2018; Lutermann et al., 2019; Table 2). Microparasites are the least studied parasites for Bathyergidae and natural infections have only been assessed for one fungal (Emmonsia parva) and two bacterial taxa (Bacillus cereus and Mycoplasma spp.). Of these Mycoplasma spp. could be vector-borne while the other two are environmentally transmitted via contamination of the soil (Hubálek et al., 2005; Retief et al., 2017, 2021). All of these microparasites appear to be host generalists. However, while prevalences were generally greater for social compared to one solitary hosts, this did not apply to B. suillus which had prevalences exceeding those of social species (C. h. hottentotus, F. damarensis) for both bacteria.

The most speciose ectoparasite taxon infecting Bathyergidae is represented by mites with at least 18 species (Table 2). For trombiculid mites (i.e., chiggers, genera: Austracarus, Euschöngastia, Gahrliepia, and Schoutedenichia) only the first instars are generalist parasites while all other stages are soil dwelling and consequently transmission is environmental (Shatrov and Kudryashova, 2006). Thus, it is not surprising that these parasites have been reported from a range of social bathyergids but only one solitary species as the larger burrow system of the former may increase their exposure to these mites. The remaining 14 mite species reported for bathyergids rely predominately on direct transmission and are often the most prevalent and abundant parasites reported for a host species (Viljoen et al., 2011b; Archer et al., 2014; Lutermann et al., 2015, 2019; Fagir et al., 2021). Only three of these species are known host generalists (Androlaelaps marshalli, Laelaps liberiensis and Orntihonyssus bacoti) while the remaining 11 species appear to be host specialists for Bathyergidae either at the family, genus or species level (Table 2; Lutermann et al., 2019). This suggests a potentially close co-evolutionary history between these mites and bathyergids that should help to shed light on the selection pressures experienced by both actors. Blood-sucking lice (Anoplura) are the most sedentary ectoparasite taxon of bathyergids and require direct contact between hosts for transmission (Kim, 2006). This is also why most louse species have a narrow host range, often using a single host species, which also appears to apply to the two species identified in Bathyergidae with one potentially specific to the host genus (Eulinognathus hilli) and another to the species (E. lawrensis) (Zumpt, 1966; Lutermann et al., 2019; Table 1). Thus, the patterns of host specialization across the ectoparasite communities described for bathyergids to date appears to support the hypothesis for a link between parasite transmission mode, host social behavior and parasite specialization. As laid out above this specialization should also include lowered virulence in these parasites and consequently, stronger selection pressures on behavioral rather than physiological immunity for social bathyergids.

The picture arising from the helminth community described for Bathyergidae is somewhat less clear. This is partially due to the lower taxonomic resolution with most parasites only being identified to genus level (Lutermann et al., 2019; Table 3). However, cestodes have complex life cycles that require an intermediate host (often an arthropod) in their life cycle (Georgiev et al., 2006). Hence, they need to adapt to both an invertebrate and vertebrate which makes host specialization much less likely as host encounters tend to depend mostly on stochastic processes (Antonovics et al., 2017). This could not only account for the limited number of cestode species observed in Bathyergidae to date, but would also make the evolution of specific behavioral or physiological responses less likely (Antonovics et al., 2017). At the same time, basic hygienic behaviors (e.g., grooming to reduce intermediate hosts such as mites or fleas) and organizational immunity in the form of dedicated toileting areas can be effective means to reduce the transmission of cestode propagules. Nematodes have a more diverse range of transmission modes including direct, environmental or transmission via an intermediate host, but this is often not well established for a species (Anderson, 2000). However, many of the genera observed to infect Bathyergidae are host generalists with a wide geographic distribution (Anderson, 2000; Table 3). Possible exceptions are four nematode species (Ortleppsrongylus bathyergid, Mammalakis macrospiculum, M. zambiensis, and Paralibyostrongylus bathyergid) that have to date only been found in bathyergids with three of them retrieved from B. suillus (Lutermann et al., 2019; Table 3). Helminths are famous for their ability to manipulate the immune responses of their hosts and may strongly interact with symbiotic microorganism (Maizels et al., 2004). Hence, they often do not trigger strong immune responses from their hosts, at the same time, they require host survival for successful proliferation. Consequently, they may favor tolerance responses in hosts.

Rather than an individual parasites species the composition of parasite communities exploiting a host and the virulence of common parasites will determine selection on particular social behaviors as well as the extent to which such behaviors reduces the associated fitness costs (Hawley et al., 2021). This in turn will shape the evolutionary trajectories for the parasites involved. The majority of macroparasites exploiting Bathyergidae seem to fall into one of two broad categories: directly transmitted host specialists (at species, genus or family level) and indirectly (environmentally or via intermediate hosts) transmitted host generalists with the former occurring at the highest prevalences (Viljoen et al., 2011b; Lutermann et al., 2013, 2019; Archer et al., 2014, 2017; Fagir et al., 2021). Consequently, selection on the bathyergid host should be strongest on behavioral (e.g., hygienic behaviors, organizational immunity) instead of physiological control strategies while for those parasites selection will act to favor benign parasites with direct transmission.