Presence of coronaviruses in the common pipistrelle (P. pipistrellus) and Nathusius´ pipistrelle (P. nathusii) in relation to landscape composition

Abstract

Changes in land use can modify habitat and roosting behaviour of bats, and therefore the transmission dynamics of viruses. Within bat roosts the density and contact rate among individuals increase and may facilitate the transmission of bat coronaviruses (CoVs). Landscape components supporting larger bat populations may thus lead to higher CoVs prevalence, as the number of roosts and/or roost size are likely to be higher. Hence, relationships between landscape composition and the presence of CoVs are expected to exist. To increase our understanding of the spread and shedding of coronaviruses in bat populations we studied the relationships between landscape composition and CoVs prevalence in the species Pipistrellus pipistrellus and Pipistrellus nathusii. Faecal samples were collected across The Netherlands, and were screened to detect the presence of CoV RNA. Coordinates were recorded for all faecal samples, so that landscape attributes could be quantified. Using a backward selection procedure on the basis of AIC, the landscape variables that best explained the presence of CoVs were selected in the final model. Results suggested that relationships between landscape composition and CoVs were likely associated with optimal foraging opportunities in both species, e.g. nearby water in P. nathusii or in areas with more grassland situated far away from forests for P. pipistrellus. Surprisingly, we found no positive association between built-up cover (where roosts are frequently found) and the presence of bat-CoVs for both species. We also show that samples collected from large bat roosts, such as maternity colonies, substantially increased the probability of finding CoVs in P. pipistrellus. Interestingly, while maternity colonies of P. nathusii are rarely present in The Netherlands, CoVs prevalence was similar in both species, suggesting that other mechanisms besides roost size, participate in the transmission of bat-CoVs. We encourage further studies to quantify bat roosts and colony networks over the different landscape compositions to better understand the ecological mechanisms involved in the transmission of bat-CoVs.

Fig 1

Partial coefficients and odd ratios of all predictor variables in the final model for the species P. pipistrellus (A, C) and P. nathusii (B, D). The figures above show the proportion of variance explained by each predictor variable in the final model, while accounting for the effects of other predictors (inclusive R²). The figures below display the relative odd ratios of the estimates, with positive relationships represented in blue and negative relationships represented in red. Values located near 1 on the x-axis indicate smaller effect sizes. Note that values are represented with their corresponding confidence intervals, and are standardized using a mean = 0; standard deviation = 1. For the variable sample source, “roost” is the reference. A detailed description of the inclusive R² and odd ratios’ values can be found in S4 and S5 Tables in S1 File.

Fig 2. Presence of CoVs in faecal samples collected from non-roosting bats, small roosts and large roosts.

On the y-axes, 0 = absence of CoVs (light grey) and 1 = presence of CoVs (dark grey). The proportion of positives/negatives is given per sample source and is represented on a scale from 0 to 1. In P. nathusii no samples were available from large roosts as maternity colonies are rarely found in The Netherlands. The width of the bars is relative to the sample size of each sample source category.