Host‐handling behaviours in parasitoids of the black scale: a case for ant‐mediated evolution

Abstract

We hypothesize that differences in host‐handling times among three closely related encyrtid parasitoids ( Metaphycus anneckei Guerrieri & Noyes, Metaphycus hageni Daane & Caltagirone and Metaphycus lounsburyi [Howard]) are influenced by the defences of their host, black scale ( Saissetia oleae [Olivier]). Two forms of host defence were examined: (i) the scale’s possession of a hard integument, and (ii) the effect of ants tending S. oleae . Duration of host assessment, host rejection, drilling/oviposition and the presence or absence of host feeding were measured for M. anneckei , M. hageni and M. lounsburyi . Results show that M. anneckei oviposits through the soft ventral integument while M. hageni and M. lounsburyi oviposit through the harder dorsal integument. M. anneckei was never observed to host feed while both M. hageni and M. lounsburyi use hosts for either oviposition or host feeding. The ventral drilling location and absence of non‐concurrent host feeding significantly reduced the overall host‐handling time of M. anneckei , relative to M . lounsburyi and M . hageni . Oviposition success of M. anneckei and M. hageni in the presence of ants was compared. An observation arena was established with the numbers of Argentine ants, Linepithema humile (Mayr) and S. oleae manipulated on potted oleanders. M. hageni and M. anneckei were placed, individually, onto the observation arena and their interaction with ants and S. oleae recorded. While 55·3% of M. anneckei successfully oviposited, M. hageni was unable to oviposit or host feed in the presence of foraging ants. Field tests were conducted using potted oleander plants, infested with S. oleae , to determine the effect of L. humile on scale density, parasitism levels and parasitoid species composition. Ant‐tended plants had significantly more scale, lower scale parasitism and lower levels of host feeding (as indicated by unknown scale mortality levels). There were also significant differences in parasitoid species composition between ant‐tended and ant‐excluded treatments, with the relative frequency of parasitoid species with faster host‐handling times more common in the ant‐tended treatment.