The novel finding in this study was that an individual female E. helvum
, which had specific antibodies to both EBOV-Zaire and LBV, was alive over 13 months post-sampling. This bat appeared to be healthy and showed typical migratory movements for E. helvum
in this region of West Africa 
(). Athough fruit bats seropositive to EBOV and MARV have been detected elsewhere, antibodies to filoviruses have not been previously demonstrated in E. helvum
, including E. helvum
sympatric with R. aegyptiacus
seropositive to EBOV and MARV 
, and long-term survival of seropositive bats has not previously been shown.
Fruit bats are recognised reservoirs of lyssaviruses and henipaviruses, and can have high viral seroprevalences 
. In contrast, most studies have reported that filovirus seroprevalence in bats is low (<10%) 
, although 24% H. monstrosus
were found to be positive against EBOV following an EBOV epidemic in humans, and where PCR-positive bats were detected in the region shortly after the outbreak, suggesting recent virus circulation 
. The positive predictive values of tests used in all of these studies are unknown. In the absence of a population of E. helvum
known to be uninfected, formal investigation of specificity in our study is impossible. Similarly, the sensitivities of filovirus serological assays on bat samples are unknown. Efforts to ensure that results are accurate here were based on good laboratory approach and the use of western blot analysis to confirm IFA results.
One in 262 sera is clearly a low seroprevalence and it is likely that virus prevalence is extremely low in E. helvum
in Ghana. Prevalences of other bat viruses, such as lyssaviruses and henipaviruses, for example, are very low even when seroprevalences are relatively high (e.g. around 40%) 
. Western blot analysis indicated that the antibody response in our study was specific to EBOV-Zaire, which is highly pathogenic to people, rather than to EBOV-Reston (data not shown), which has low pathogenicity in humans and which has not been detected in Africa 
. Whilst the pathogenicity of EBOV-Zaire to E. helvum
cannot be determined by our study, the presence of detectable anti-EBOV antibody demonstrates that there has been sufficient exposure to viral antigen for the bat (#49) to develop an adaptive immune response. The most parsimonious explanation is infection followed by sufficient virus replication for the animal to mount an adaptive immune response. Insufficient data exist to determine if infection of E. helvum
with EBOV can be productive, thus enabling it to serve as an effective reservoir for EBOV, or if this species develops clinical signs when infected with EBOV. Also, studies in individual bat roosts should not be extrapolated to all bats or roosts of that species, but population level infection dynamics must be considered when determining the reservoir status of a species.
Spillover of infection into a large susceptible population, such as E. helvum
, which roosts in urban areas, must be considered a possible public health risk. Because E. helvum
typically lives in large urban colonies and is a source of bushmeat in some regions 
and direct bat to human transmission of EBOV has been reported 
, further studies should determine if this species forms a reservoir for EBOV from which spillover infections into the human population may occur.