Synchronized shift of oral, faecal and urinary microbiotas in bats and natural infection dynamics during seasonal reproduction

Dietrich, Muriel; Kearney, Teresa; Seamark, Ernest C. J.; Paweska, Janusz T.; Markotter, Wanda

For M. natalensis, among the bats sampled at two (n = 85) or three (n = 30) body habitats, 45% were infected by two infectious agents and 17% were infected by the three infectious agents. Leptospirapositive urine samples in M. natalensis clustered in two genetic groups (figure 2a), one associated with L. borgpetersenii samples previously identified in M. natalensis in South Africa [26] and the other closely related to L. interrogans and L. kirschneri. High Leptospira shedding prevalence was observed across the reproductive season, varying between 71% (±18%) and 87% (±10%). Despite some variation over time and a peak in November that coincided with parturition (figure 1a), we found no significant effect of the sampling month for Leptospira shedding prevalence in urine. This may be due to a lack in statistical power, because of the limited number of urine samples available in September (n = 4). Only a significant effect of age class on Leptospira shedding prevalence was found (GLM 1 : χ 2 1 = 6.851, p = 0.009; electronic supplementary material, table S1), but this was because none of the two urine samples collected from juveniles were positive for Leptospira, compared to an overall shedding prevalence of 83% (±9%) in adults. Our results are consistent with previous studies reporting bats as a reservoir of pathogenic Leptospira worldwide [36] , and similar to the dynamic found in the insectivorous bat Mormopterus francoismoutoui in Reunion Island, where a peak of shedding was observed in late-pregnant females [12] . AdVs detected in faeces of M. natalensis were almost all identical and closely related to mastadenoviruses found in the bat Myotis ricketti in China [30] . One sample (UP5980) clustered with a sequence from the fruit bat Pteropus giganteus in Bangladesh (figure 2b). We found a strong variation of AdV shedding prevalence over the reproductive season (GLM 2 : χ 2 2 = 65.197, p < 0.001), as shedding only occurred in November with 72% (±14%) of the bats excreting AdV (figure 1a). Our limited number of faeces samples collected in September (n = 3) may hide a higher prevalence in the M. natalensis colony, but results from January (n = 28 samples) showed that lactating and post-lactating females, as well as young juveniles, were not shedding AdV. This contrasts with the stable prevalence observed by Drexler et al. [13] in Myotis myotis in Germany, before and after parturition. The absence of AdV shedding in young juveniles may be explained by the transfer of maternal antibodies, as demonstrated experimentally for henipaviruses [37, 38] and largely suggested by several field studies [12, 13, 39, 40] . However, the absence of AdV shedding in lactating and post-lactating females suggests that even though persistence of DNA viruses on the level of individual bats might occur [13] , shedding is highly dynamic over the reproduction period in female individuals.