The parasitoids known to attack 191 phytophagous species of gall midges (Cecidomyiidae) were used to examine factors influencing parasitoid assemblage size. The number of parasitoid species a midge species supports was tested against nine variables describing geographical, biological and ecological attributes of hosts. The apparency of midge larvae was found to have the greatest influence on parasitoid assemblage size; highly visible species support more parasitoids than less visible ones. Pupation site and midge voltinism also significantly affect associated parasitoids, at least for highly apparent hosts. Biogeographic region, host-plant architecture and the plant parts infested were found to be of secondary importance. The surface texture of infested plant parts, the number of midge larvae occupying galls and the diversity of plant tissues infested have minimal apparent effects on parasitoid richness. Parasitoid assemblage size and total parasitism rates were also found to be positively correlated for 73 galling and nongalling midge species, and gallers typically suffer higher levels of parasitism than non-gallers. Using these data to test the "enemy hypothesis", which proposes that the galling habit has evolved to escape attack from parasitoids, we conclude that parasitoid pressure cannot account for the presence of galls in the Cecidomyiidae.
Read full abstract