Our first hypothesis regarding the presence of bats in most Grenadian homes was correct, as 60 % of respondents reported bats in their homes (Table 1). Regarding roof type and arbovirus infection however, the results showed no significant correlation between the presence of bats with arbovirus infection or with roof type (P= 0.34, P=0.28 respectively).
Homes in all parishes were surveyed and reported bats; but a shortcoming was the use of the St. George bus stop as our location. While the bus stop was used to obtain surveys from a wide range of Grenadians from multiple parishes, our results showed that this did not allow us to obtain surveys evenly from all areas of the island. The sample size of 210 was met, however they were not consistently distributed among the parishes of Grenada. Certain areas of the island such as Grenville were particularly underrepresented. Grenville is the second largest town in the country, located on the east side of the island and therefore people may use a different bus stop for that route. Non-specific information of where the participants lived: community (data not shown), was also collected in the data from participants. The data was summarized and analyzed by parish, however most of the data has more specific information on location/community. This was to allow the study to compare areas of rural vs. suburban, looking at a possible correlation to having more prevalent cases of bats roosting in homes that were closer to natural reserves or wildlife areas. A similar study made this correlation this in northeastern Brazil, where they looked at areas associated within close distance to the wildlife reserves and the association with human-bat interactions (9). Grenada’s key reserves, parks, sanctuaries are detailed in Fig. 2, however, most of Grenada’s homes are surrounded by wildlife and green areas, therefore, 60 % of participates reporting bats inhabiting their houses was not unexpected.
One significant finding was that the majority of residents who reported having bats in their homes, had attempted one or more methods to remove the bats (P=0.0053). When questions were asked if the household had attempted to get rid of the bats, the answers were sometimes answered with simple yes or no, but some had more detail. Even more so, in some cases, the actual method was hard to interpret such as “bat proofed the house”, thus the data was simplified on a yes or no basis for analysis. This suggests bats are viewed as a problem to Grenadian families, although this is most likely due to the destruction of their property rather than the awareness of potential zoonotic diseases, this information was not collected in this study. Nevertheless, our results indicate that citizens of Grenada would likely be receptive to strategies that would help reduce the presence of bats in people’s homes, such as construction plans to help prevent bats from accessing roofs, and the creation of wooden bat houses adjacent to populated areas.
Another shortcoming worth reporting was that while responding to the survey, some individuals who participated were embarrassed to admit or discuss the presence of bats in their homes perhaps due to their cultural association with uncleanliness, which could have impacted other responders and thus, the real number of affected homes may be higher.
Results show that both fruit-eating bats as well as non-fruit-eating bats were present in homes, as suggested by the presence of seeds in their guano; however, it’s difficult to differentiate between species and this information is speculative because it cannot be confirmed (9). Information on the type of roofing (galvanized, wood etc.) and the presence of window screens was gathered for interpretation on whether bats would roost in a particular type of home building style. Looking at screens on windows showed the type of protection the occupants had from the mosquitos that could potentially transmit the arboviruses among bats and humans. Again, seeing if correlation of no window screen would be comparable with disease presence. No difference between roofs and presence of screens were observed which opposed our hypothesis.
The interview process also inquired if any of the household members had reported dengue, Zika or Chikungunya virus infections. If a household resident had encountered one of the arboviruses, further questions were asked seeing if confirmation of the virus had been done by a doctor/lab or if the stated virus type was a non-confirmed assumption. Arbovirus disease self-reporting was predominantly Zika and Chikungunya which may reflect the outbreak behavior of the virus as well as people’s awareness due to recent introduction as compared to the endemic dengue cases. The rate of infection of bats with Zika, dengue and Chikungunya is under investigation in Grenada, but a recent study in Mexico reported 9% prevalence of Zika virus infection and 32% of WNV by RT-PCR, with no detection of dengue in the region that time in that small sample (n=22) (8) .