The decision by individuals to disperse – leaving their natal territory in search of breeding opportunities (Greenwood, 1980) – and the distance that they travel, can be affected by inbreeding avoidance, by competition for mates, and by the distribution of resources in the environment (Gaines and McClenaghan, 1980; Bowler and Benton, 2009; Clutton-Brock, 2016). During dispersal, individuals must navigate through an unfamiliar environment, and for most terrestrial mammals this presents considerable mortality risks (Lambin et al., 2001; Maag et al., 2022). Such risks have been suggested to be particularly high in subterranean species whose morphological and ecological adaptations to underground living may make them ill-suited to dispersing long distances (Buffenstein, 2000; Stein, 2000). Because of their morphological adaptations and reliance on underground burrows, the natal dispersal distances of subterranean mammals have been suggested to be shorter than those of surface-dwelling species (Nevo, 1979).

Dispersal distances are known to vary widely within and between species, and the distribution of dispersal distances can provide information on a species’ ecology (Paradis et al., 2002; Nathan et al., 2012; Whitmee and Orme, 2012). For example, by comparing sex differences in dispersal across species it becomes apparent that the relative distance that males and females travel when dispersing is not necessarily directly associated with their overall tendency to disperse (Pope, 2000; Hanski and Selonen, 2009; Clutton-Brock, 2016). Despite such insights, collecting data on dispersal dynamics remains challenging. Following dispersing individuals is time and labor intensive and often requires tracking individuals beyond the bounds of a study area (Tomkiewicz et al., 2010; Maag et al., 2022), or relies on the deployment of biologgers which makes tracking possible without the need for direct resighting or recapture (Wilmers et al., 2015; Cozzi et al., 2020). These challenges have rarely been met for subterranean mammals and quantitative data on dispersal are relatively uncommon in these taxa (Busch et al., 2000). Recent molecular studies on subterranean rodents have shown that not all species have low vagility (Bray et al., 2012; Welborn and Light, 2014; Visser et al., 2018), but whether the dispersal distances of subterranean rodents are generally lower than those of surface-dwelling rodents remains unclear.

Damaraland mole-rats (Fukomys damarensis) are cooperatively breeding subterranean rodents (Family Bathyergidae) with high reproductive skew: a single reproductive pair monopolize reproduction, and their philopatric offspring delay dispersal until ecological constraints on burrowing (e.g., hard soils) are relaxed (Jarvis and Bennett, 1993; Bennett and Faulkes, 2000). Breeders can have long life spans, and non-breeders have a low chance of inheriting the burrow system (<1%; Schmidt et al., 2013; Torrents-Ticó et al., 2018; Thorley et al., 2021). Thus offspring of either sex must disperse from the natal group to acquire reproductive opportunities, even though strong ecological constraints make successful settlement rare. Indeed, very few individuals within a Damaraland mole-rat populations breed at some point in their lives (<8%; Jarvis and Bennett, 1993; Torrents-Ticó et al., 2018; Mynhardt et al., 2021; Thorley et al., 2021). Previous field studies of social mole-rats have shown that when individuals disperse, they frequently do so alone and rarely in coalitions (Jarvis and Bennett, 1993; Torrents-Ticó et al., 2018), and evidence suggests that dispersal occurs both above and below ground (Hazell et al., 2000; Bray et al., 2012; Patzenhauerová et al., 2013; Finn, 2021). By blocking themselves off into a section of the natal burrow, individuals can form a new territory via “territory budding” (Jarvis and Bennett, 1993; Lövy et al., 2013; Mynhardt et al., 2021). Genetic and re-trapping data suggest that new groups often form from unrelated pairs (Burland et al., 2002). Taken together, this demographic information identifies features that are likely to have been important selective forces in the formation of social groups in mole-rats. For example, the fact that males sometimes immigrate into established groups, whereas females rarely do (Patzenhauerová et al., 2013; Torrents-Ticó et al., 2018; Thorley et al., 2021), might explain the relatively large male-biased sexual size dimorphism in Fukomys mole-rats (Caspar et al., 2021). However, mole-rat populations are known to occupy habitats with varying climate and geology and the structure of groups often varies (Spinks et al., 2000; Finn et al., 2018). In this context, dispersal could play a crucial role in structuring population differences and it is important to understand sex-specific dispersal patterns across mole-rat populations.

In this study we investigate the dispersal distances of Damaraland mole-rats in the southern Kalahari Desert. Our study serves two aims. First, we use longitudinal data from a capture-mark-recapture study to test whether the distances traveled vary between the sexes or as a result of body condition or group size. Second, we conducted a phylogenetically controlled comparative analysis of published dispersal distances in rodents to test whether the below ground lifestyle constrains dispersal distances.

Our study investigated the dispersal dynamics of the Damaraland mole-rat. We did not find a significant difference in the dispersal distances of males and females. Both sexes were capable of traveling well over 1,000 m, and in two cases over 4,500 m. This similarity may be due to the social structure and mating system of social mole-rats, where mole-rat families are composed of an unrelated monogamous pair (Bennett and Faulkes, 2000). This structure forces both sexes to disperse in order to reproduce with unrelated individuals, rather than inheriting breeding positions in their natal burrow. Females can travel long distances searching for new territories with food resources (i.e., plant tubers), and males can travel long distances searching for mates.

In most monogamous and cooperatively breeding mammals, like the Damaraland mole-rat, females often settle in close proximity to their mother’s territory (Dobson, 2013). Therefore, we expected that females would also follow this pattern as observed in a related social mole-rat (Lövy et al., 2013). Whereas short dispersal distances do occur, we found that females mostly traveled more than 500 m from the natal burrow, and as shown previously, establish a new burrow system while waiting for dispersing males to join them (Thorley et al., 2021). The widely scattered food resources may force females to travel far to find suitable territory. Alternatively, females may disperse farther from the natal group to avoid competing with neighboring groups since over time burrow systems can become quite large. When the population density is low, the chance of an immigrant male locating a single female is low and therefore single females may have to wait several years before finding a mate (Thorley et al., 2021). The exceptional longevity of mole-rats for their size (Buffenstein et al., 2020; Thorley, 2020), combined with low predation rates once settled in a burrow, may allow females to endure these considerable periods without reproductive opportunities while waiting for male dispersers to find them.

Dispersal distance can be affected by body mass, territory size, availability of resources or vacant territories, and breeding opportunities (Bowman et al., 2002; Clobert et al., 2009; Whitmee and Orme, 2012). In our study, we found that animals from larger groups traveled further, and that natal group size had a stronger effect on dispersal distance than the body condition of dispersing individuals (Supplementary Material). The cause of this pattern is unclear, but one possibility is that group and burrow size may be influenced by the distribution of food resources in the landscape. In the Kalahari, the food resources for mole-rats exhibit a scattered but patchy distribution which may affect the dispersion and size of family groups in the landscape (Lovegrove, 1988). In particular, one could hypothesize that large groups occupy larger burrow systems and that this may force females to move further to find a vacant habitat with sufficient food resources. Thus, since natal group size and burrow size are closely related (Lovegrove, 1988; Thomas et al., 2016) the effect of group size may indicate that burrow size affects dispersal distances in mole-rats. Variation in group size might also reflect the localized abundance of plant tubers, and this in turn might affect the distances that individuals are required to move to find vacant habitat or receptive mates. Lastly, increases in group size may also lead to increased competition for food or space between individuals within a group (Cooney, 2002; Zöttl et al., 2016), which may cause heightened stress levels and trigger dispersal to avoid competition (Quirici et al., 2011). Heightened stress levels from increasing group size (and limited burrow capacity) may cause animals to disperse farther to escape a perceived stressor. Adult Damaraland mole-rats exhibit higher glucocorticoids than juveniles (Medger et al., 2018), and cortisol levels increase during the rainy season likely due to dispersal opportunities (Young et al., 2010), but the effects of group size on stress levels has not been investigated in mole-rats.

Despite the challenges of following dispersal in a subterranean rodent, the low rate of successful dispersal conforms to the general view that dispersal in mole-rats and other subterranean rodents carries a high mortality risk. Mole-rats are specialized diggers with low visual acuity (Kott et al., 2010; Gomes Rodrigues et al., 2016; Montoya-Sanhueza, 2020), and like other similarly adapted species, it has long been assumed that an underground lifestyle limits vagility (Nevo, 1979). Adaptations for subterranean life may make mole-rats particularly susceptible to predation while dispersing. Indeed, diet studies indicate that mole-rats may present a seasonally available prey item to jackals (Canis mesomelas) and serval (Leptailurus serval; Humphries et al., 2015; Ramesh and Downs, 2015). Despite the apparent constraints on dispersal, we present evidence showing that dispersal distances in subterranean rodents fall within the expected range of rodents based on their body size. Within the Family Bathyergidae previous studies have recorded a maximum dispersal distance of 2,149 m in Cape dune mole-rats (Bathyergus suillus; Bray et al., 2012), and 2,400 m in naked mole-rats (Heterocephalus glaber, Braude, 2000). Our study found that 5 dispersers, including individuals of both sexes, traveled over this distance, with the greatest distance traveled being 5,232 m from the natal burrow. Our study shows that the subterranean niche does not necessarily limit long distance dispersal, and some individuals can settle successfully far away from their natal territory. The long distances observed in this study suggest that traveling above ground may be the primary method of natal dispersal for Damaraland mole-rats and perhaps many other subterranean rodents, as previously suggested in geomyids (Welborn and Light, 2014).

Like in other field studies of natural populations the estimations of dispersal frequency and distance are likely to be conservative underestimates. The estimation of dispersal distances are based on the recapture of known dispersers and it is likely that the likelihood of finding these individuals decreases with increasing dispersal distance (Tomkiewicz et al., 2010; Maag et al., 2022). The number of unknown dispersers is also an approximation because we were unable to assess relatedness among individuals in each group. Some unknown dispersers found in established groups during the study may be recruits born to the reproductive pair which evaded capture and were not immigrants. However, our data suggests that these are rare occurrences because longitudinal resampling suggests that we rarely miss individuals repeatedly. Because of these reasons we believe that both of these sources of bias either lead to conservative underestimates or to very minor inaccuracies in frequencies of immigration.

By focusing on successfully established dispersers and the distances they moved, our study can only get us part way toward a complete understanding of sex-specific dispersal distances in Damaraland mole-rats. Dispersal is a three-stage process covering emigration, transience, and settlement, and different factors (e.g., mortality) may affect the sexes at each of these stages (Clobert et al., 2009; Maag et al., 2022). The examination of settlement patterns provides information on the probability that members of either sex acquire breeding opportunities, and the circumstances under which they do so, but additional sources of information must also be used to understand individual decisions during emigration and transience. As mole-rats cannot be observed continuously, emigration patterns have often been inferred from the duration of philopatry (Torrents-Ticó et al., 2018; Thorley et al., 2021), under the assumption that a sizeable proportion of disappearances are due to dispersal rather than in situ mortality. Such data suggests that in the social mole-rats, males emigrate from groups earlier than females (Torrents-Ticó et al., 2018). However, there is also no clear evidence of lifelong philopatry of non-breeding females (Torrents-Ticó et al., 2018; Thorley et al., 2021). While males may disperse earlier, the difference in the absolute number of male and female emigrants may be very small, at least in our population where the adult sex ratio is equal to that of dispersers (cf. Hazell et al., 2000). Direct information on the transience phase of dispersal in mole-rats is largely non-existent, save for one individually tracked animal (Lövy et al., 2013). It remains to be seen whether the two sexes face differing mortality risk during dispersal, such as from competitive aggression (Busch et al., 2000), or through increased search times (Šumbera et al., 2008). Molecular techniques have advanced our understanding of dispersal patterns in fossorial rodents by using microsatellite markers to determine the presence of sex biases in dispersal, gene flow between populations, and allow for an estimation of dispersal distances without the need for longitudinal sampling (Cutrera et al., 2005; Fernández-Stolz et al., 2007; Bray et al., 2012; Patzenhauerová et al., 2013; Welborn and Light, 2014; Visser et al., 2018; Mynhardt et al., 2021). Comparing variation in dispersal patterns within a single species across populations living in varying population densities, or environmental conditions (e.g., rainfall or soil type) would greatly assist in determining the extent to which demographic and ecological factors affect dispersal opportunities, frequency of intergroup movement, and gene flow in subterranean rodents.

The datasets presented in this study can be found on the online repository FigShare: https://doi.org/10.6084/m9.figshare.19196732.v1.

The study was approved by the University of Pretoria Ethics Committee (EC032-13, EC006-15, EC050-16, and EC059-18) and subsequently approved by Northern Cape Nature Conservation.

KF, JT, and MZ conceived and designed the study. KF, JT, HB, and MZ collected the data. KF and JT analyzed the data. KF wrote the first draft of the manuscript. All authors made significant contributions.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.