Introduction and purpose
Growing concern about the emergence of zoonotic infectious diseases has stimulated research into the dynamics and drivers of infection in wildlife reservoir hosts (Patz et al. 2004; Jones et al. 2008). Understanding factors that drive the magnitude and timing of epidemics – and which allow for the long-term persistence of pathogens – in animal host populations is of particular importance to the prediction and prevention of zoonotic spillover into humans (Plowright and Hudson 2021). Investigations into how ecological features shape infection is achieved through empirical and theoretical research at the interface of ecology and wildlife epidemiology (e.g., Peel et al. 2014; Lunn et al. 2021; Mancy et al. 2022). The availability of key ecological information on wildlife reservoir hosts is fundamental to these research efforts, yet even basic information is lacking for many wildlife species in global hotspots of infectious disease emergence (Kuzmin et al. 2011).
Seasonal birthing is a common ecological feature of wildlife species, and is thought to play a key role in the infection dynamics of populations via the addition of susceptible individuals (Peel et al. 2014; Plowright et al. 2016; Baker et al. 2020). At its simplest, the seasonal addition of naïve juveniles increases the proportion of contacts between susceptible and infected individuals, increasing the rate of transmission during and following the birthing period, thereby driving seasonal epidemic cycles (Peel et al. 2014). Ecological attributes of seasonal birthing, as well as other host and pathogen features such as population size and infectious period, shape variation in this dynamic host-pathogen relationship.
Synchronicity of birthing (the time period over which birthing occurs), and the number of birth pulses per year, are key drivers for the magnitude of epidemic cycles and contribute to the probability of pathogen extinction from a population. More synchronous birth pulses (the same number of births over a shorter time) can drive pathogen extinction by creating large amplitude oscillations in epidemics (i.e., high peaks, low troughs), which increase the probability of stochastic fade-out during troughs in the epidemic cycle (Peel et al. 2014). This is exacerbated by short infectious periods and high demographic turnover, and increases the population size required for pathogen persistence (Peel et al. 2014; Hayman 2015). Meanwhile, having more birth pulses per year can increase the likelihood of pathogen persistence by decreasing the amplitude of oscillations (i.e., lower peaks, higher troughs), meaning the number of infected individuals does not drop as low between birth pulses, and the risk of pathogen extinction via stochastic fade-out is reduced (Hayman 2015).
Modelling has shown that bi-annual birth pulses may be needed for the persistence of filoviruses (including Ebolavirus sp. andMarburgvirus sp.) in bat populations, and that persistence – even with bi-annual birthing – is sensitive to birth synchrony and host/virus attributes (Hayman 2015). Persistence of filoviruses is more likely when birth synchronicity is low, when the incubation or infectious periods are long, and when population sizes are large (Hayman 2015). These findings highlight ecological conditions under which sustained transmission of ebola- and marburgviruses may be possible and identify key traits of reservoir hosts. However, these results pertain to generic bat populations with fruit bat-like ecological traits (based on the African fruit bats, Eidolon helvum and Rousettus aegyptiacus ), and are not necessarily applicable to other possible hosts of ebolaviruses, including insectivorous bats (Hayman 2015; Leendertz et al. 2016). Species-specific investigations into plausible reservoir hosts, especially for ebolaviruses, remain limited by deficits in information on bat ecology. The synchronicity of birthing – in particular – is unknown for most bat species and warrants further attention (Kuzmin et al. 2011; Hayman 2015).
The Angolan free-tailed bat (Molossidae: Mops condylurus ) has been implicated as a reservoir host for ebolaviruses. Wild-caughtM. condylurus have tested positive for ebolavirus antibodies (De Nys et al. 2018), and demonstrate the ability to replicate and shed ebolaviruses following experimental inoculation, with little evidence of virus-induced morbidity (Swanepoel et al. 1996; Edenborough et al. 2019). More recently, this species has been associated with a new ebolavirus, Bombali virus, with detection of viral RNA in wild-caught bats from five distinct locations across their range (Sierra Leone, Guinea, two locations in Kenya, and Mozambique) (Goldstein et al. 2018; Forbes et al. 2019; Karan et al. 2019; Kareinen et al. 2020; Lebarbenchon et al. 2022). Bombali virus RNA has also been detected from a closely related co-roosting species, Mops pumilus (Goldstein et al. 2018).
Birth-pulse synchronicity and magnitude have not been estimated forM. condylurus or M. pumilus , despite the pertinence of this information to understanding infection dynamics. The temporal resolution of existing research – at best, singular capture events once per month – is insufficient to accurately estimate the spans of most reproductive states (Mutere 1973; O’Shea and Vaughan 1980; Happold and Happold 1989; Vivier and Van Der Merwe 1996; Vivier and Merwe 1997). Gestation and lactation periods have been estimated from patchy data, at roughly 85 days and 50-60 days respectively for M. condylurus , and 67-72 days and 21-28 days for M. pumilus (McWilliam 1976; Happold and Happold 1989; Vivier and Merwe 1997). Female M. condylurus show bi-annual birthing across their range, and M. pumilus up to penta-annual birthing depending on latitude (Happold and Happold 1989). Both species have a post-partum oestrus, where bats can be detected lactating and gestating simultaneously (Happold and Happold 1989).
Given current gaps in knowledge, and the importance of detailed reproductive information for investigating viral maintenance, the purpose of this study was to empirically characterise key parameters of the birth pulse in M. condylurus and M. pumilus bats. Specifically, we investigated the timing, synchronicity, and magnitude of the birth pulse in south-eastern Kenya where Bombali virus has previously been detected (Forbes et al. 2019; Kareinen et al. 2020).