Parmentier, T., Dekoninck, W. & Wenseleers, T.

Survival of persecuted myrmecophiles in laboratory nests of different ant species can explain patterns of host use in the field (Hymenoptera: Formicidae)

Myrmecophiles or ant associates are able to penetrate and survive inside the heavily defended nests of various ant species. With the exception of some highly specialized species, many of these myrmecophiles elicit a highly aggressive response and are frequently wounded or even killed by their hosts. Many myrmecophiles also appear to strongly prefer particular host species. The factors that allow the myrmecophiles to survive in these hostile environments and cause myrmecophiles to prefer particular host species are largely unknown. The aim of the present study was to examine the impact of the presence or absence of either the preferred host (wood ants of the Formica rufa Linnaeus, 1761 group) or one of several nonpreferred ant species on the long-term survival of three obligate, unspecialized beetle myrmecophiles, Thiasophila angulata (Erichson, 1837), Lyprocorrhe anceps (Erichson, 1837), and Amidobia talpa (Heer, 1841), and one facultative myrmecophile, the woodlouse Porcellio scaber Latreille, 1804. In addition, we tested whether host specificity was driven by the size of the ant host workers, because host specificity has previously been demonstrated to be inversely related to aggression towards macroparasites. Our results show that despite regular aggressive host interactions, survival of the obligate myrmecophilous beetles over a period of 20 days was not different from a control set-up without ants. By contrast, the facultative ant associate P. scaber hardly provoked any aggressive host response, but its survival was lower in presence of F. rufa workers compared with a control set-up without ants. Furthermore, the data on survival in presence of nine different ant host species show that the three obligate myrmecophile beetles survived better in presence of largerbodied ant species, and that their survival was highest in presence of their preferred host F. rufa, which also has relatively large workers. The only exception to this trend was the low survival observed in presence of the large-bodied ant Camponotus vagus (Scopoli, 1763). Finally, species that were less successful in killing the beetles in our tests are also shown to support more myrmecophilous rove beetles in nature. Overall, our results shed new light on the interaction between ants and various associated macroparasites and on the factors that drive observed host preferences.