Juvenile agile frogs spatially avoid ranavirus-infected conspecifics

Abstract

Exposure to contagious pathogens can result in behavioural changes, which can alter the spread of infectious diseases. Healthy individuals can express generalized social distancing or avoid the sources of infection, while infected individuals can show passive or active self-isolation. Amphibians are globally threatened by contagious diseases, yet their behavioural responses to infections are scarcely known. We studied behavioural changes in agile frog (Rana dalmatina) juveniles upon exposure to a Ranavirus (Rv) using classic choice tests. We found that both non-infected and Rv-infected focal individuals spatially avoided infected conspecifics, while there were no signs of generalized social distancing, nor self-isolation. Avoidance of infected conspecifics may effectively hinder disease transmission, protecting non-infected individuals as well as preventing secondary infections in already infected individuals. On the other hand, the absence of self-isolation by infected individuals may facilitate it. Since infection status did not affect the time spent near conspecifics, it is unlikely that the pathogen manipulated host behaviour. More research is urgently needed to understand under what circumstances behavioural responses can help amphibians cope with infections, and how that affects disease dynamics in natural populations.

Keywords: Behavioural flexibility; Behavioural immunity; Disease transmission; Frog Virus 3; Host-pathogen interactions; Ranavirus.

Fig. 1

A schematic timeline of the experiment, the time shown as the development stage according to Gosner (A), and the compartments of the choice chambers as viewed from above (B).

Fig. 2

A summary of potential pathogen-induced behaviours of non-infected and infected hosts, along with corresponding predictions that were tested in this study. The red-coloured choice area represents the one adjacent to the infected stimulus while the blue-coloured choice area represents the one adjacent to the non-infected stimulus. Red × symbols denote reduced time relative to the areas marked with green checkmarks.

Fig. 3

Time spent in the choice areas and the neutral area by focal individuals during the no-stimulus phase (no conspecific present) assessing baseline behaviour (A) and in the stimulus phase (an exposed-and-infected and a non-infected conspecific present) assessing social and sickness behaviours (B) by all focal individuals (exposed-and-infected and non-infected animals pooled; N = 24). In each violin plot, the white dot and the grey box represent the median and the interquartile range, respectively, and the violins are Kernel density plots. The black diamond with whiskers denotes the mean ± standard error estimated from the final statistical models. The grey rectangle in the no-stimulus phase indicates that the stimulus phase was two times longer than the no-stimulus phase. Size differences between the choice areas and the neutral area were corrected for unit size (i.e. the size of a choice area).

Fig. 4

Space use by focal individuals during the stimulus phase of the trials. Each line connects the data points of one individual. Note that this figure complements Fig. 3 by presenting not only “exposed.and-infected” and “non-infected” focals (i.e. detectably Rv-infected and Rv-naïve focal individuals, respectively) but also “exposed but Rv-negative” focals, i.e. those focal individuals that had been exposed to Rv but were not detectably Rv-infected at the end of behavioural assays.