Detection of refuge from enemies through phenological mismatching in multitrophic interactions requires season-wide estimation of host abundance

The concept of enemy-free space (EFS) refers to ways of living that reduce or eliminate the vulnerability of a species to natural enemies. It has been invoked to explain host shifts of phytophagous insects. A demonstrated cause of EFS is escape from enemies in time, through phenological mismatching of herbivore development and enemy occurrence, leading to low percentages of predation/parasitism of herbivores occurring at a certain time. The mere measurement of percentage parasitism, however, is not sufficient to demonstrate EFS in certain cases. Here we present such a case, where parasitism was studied of a phytophagous insect (Phyllotreta nemorum), using two different host plant species in the field: an atypical, relatively rarely used, plant (Barbarea vulgaris), and a more widely used one (Sinapis arvensis). At one location we found a paradoxical result: on each separate sampling day throughout the season the percentage of parasitism of P. nemorum using a patch of B. vulgaris was not significantly different from, or even significantly higher than on a nearby patch of S. arvensis. The overall season-wide proportion parasitism of the flea beetle cohort using the B. vulgaris patch, however, was lower. We conclude that, in the year and at the location we studied, the patch of B. vulgaris provided enemy-free space to the herbivore in the form of a temporal refuge, and that the importance of enemy-free space in the use of an atypical host plant should be evaluated on the basis of season-wide sampling, including estimation of host population size.