Prophylactic Avoidance of Hazardous Prey by the Ant Host Myrmica rubra

Abstract

Ants are the hosts of many microorganisms, including pathogens that are incidentally brought inside the nest by foragers. This is particularly true for scavenging species, which collect hazardous food such as dead insects. Foragers limit sanitary risks by not retrieving highly infectious prey releasing entomopathogenic fungal spores. This study investigates whether similar prophylactic strategies are also developed for food associated with weak or delayed risks of fungal contamination. We compared, in Myrmica rubra ant colonies, the retrieval dynamics of dead flies that were (1) conidia-free, (2) covered with a low amount of Metarhizium brunneum entomopathogenic conidia or (3) recently fungus-killed but not yet sporulating. Foragers mostly avoided fungus-killed prey and delayed the retrieval of conidia-covered flies. A second sanitary filter occurred inside the nest through a careful inspection of the retrieved prey. Ultimately, ants mostly consumed conidia-free and conidia-covered flies, but they relocated and discarded all fungus-killed prey outside of the nest. Our study confirms that, as a host of generalist entomopathogenic fungi, Myrmica rubra ants have developed a prophylactic avoidance and a differential management of prey depending on their infectious potential. We discuss the functional value as well as the possible cues underlying pathogen avoidance and prey discrimination in ants.

Keywords: Metarhizium brunneum; Myrmica rubra; avoidance-foraging; grooming; social immunity.

Figure 1

Proportion of flies not yet retrieved inside the nest as a function of time elapsed since prey introduction on the foraging area. The two vertical dashed lines indicate the half (i.e., 1 h 30) and the end of the experiment (i.e., 3 h). Curves are drawn for the different experimental conditions. Blue solid curve: control conidia-free flies (Ctrl, n = 90); green dotted curve: conidia-covered flies (LC, n = 90); red dashed curve: flies killed by the entomopathogenic fungus (Fkill, n = 90); The coloured shade around curves represents the 95% confidence intervals for each condition.

Figure 2

(a) Length of time that dead flies remained inside the nest depending on their level of infection. Blue: control flies (n = 85), green: flies covered with conidia (n = 82), red: fungus-killed flies (n = 34). (b) Global exposure of ant colonies to prey items. This exposure level is the cumulated lengths of time that all flies remained inside the nest. Blue: control flies (n = 9), green: flies covered with conidia (n = 9), red: fungus-killed flies (n = 9). The horizontal bar within the boxes represents the median; the upper and lower boundaries of the boxes represent, respectively, the 75th and 25th percentiles, while the whiskers extend to the smallest and largest values within 1.5 box lengths. Groups sharing a common letter were not significantly different.

Figure 3

Number of ants contacting each fly inside the nest as a function of the time elapsed since its retrieval. Blue: control flies (n = 51), green: conidia-covered flies (n = 51), red: flies killed by the entomopathogenic fungus (n = 34). The points represent the mean; the bars represent confidence intervals (95%) for each condition. Time steps sharing a common letter were not significantly different.

Figure 4

Weighted number of extensive groomings (EG) inside the nest before the introduction of food (basal), after the retrieval of the first prey inside the nest and twenty minutes later. The weighted number was calculated by summing the number of EG within 5 min divided by the number of ants inside the nest. Blue: control flies (n = 27), green: conidia-covered flies (n = 27), red: fungus-killed flies (n = 21). The horizontal bar within the boxes represents the median; the upper and lower boundaries of the boxes represent respectively the 75th and 25th percentiles, while the whiskers extend to the smallest and largest values within 1.5 box lengths. For each condition, p-values are given for pairwise comparisons between observation sessions (NS = non-significant, p > 0.05). Within each observation session, conditions sharing a common letter were not significantly different.