Pflanze-Herbivore-Parasitoid Interaktionen auf Wildrosenarten und ihren Hybriden entlang eines geographischen Gradienten

Abstract

Phenotypes of closely related plant species have been shown to affect higher trophic levels. Multitrophic interactions between host plant species, herbivores and natural enemies may change depending on phenotypic differences between the host plant species and their geographical location. During this study it was investigated how the phenotypic plasticity of dog roses affects next trophic levels across a geographic gradient. The study focuses on three closely related dog rose species which differ in their phenotypes (on the lower leaf surface: glabrous Rosa canina, hairy R. corymbifera, glandular scented R. rubiginosa).The consumers Rh. alternata and D. rosae showed higher densities on R. rubiginosa, while Phragmidium spp. showed higher density on R. canina and R. corymbifera. Little host plant specialisation of Rh. alternata was linked to little impact on host performance despite high densities of Rh. alternata. High specialisation of Phragmidium spp. was related to high host impact reducing hip density and leaf cover in the following year. D. rosae galls showed intermediate patterns. This is in line with the concept of interactive systems locked in a highly specialised plant-consumer arms race, in contrast to non-interactive systems.The genetic differentiation of Rh. alternata between populations was very low on the three host species and across central Europe. There are three possible explanations for this low genetic differentiation. First, the gene flow between populations of Rh. alternata must be high. Second, the pattern of genetic differentiation is based on recent expansion of the distributional range of Rh. alternata and the ongoing gene flow between rose species of section Caninae act as a hybrid bridge.The scented R. rubiginosa showed remarkably higher gall densities, than R. canina and R. corymbifera. The survival rate of D. rosae was highest and percent parasitism lowest on R. rubiginosa. Hence, this rose species appeared to provide an important refuge for the gall-maker population. Gall-parasitoid interactions did not change with geographical position. The trophic interactions were driven by host plant identity, not by geographical location.The invertebrate communities showed only minor differences in the abundances of invertebrate taxa (Aphidina, Collembola, Aranea, Hymenoptera and Coleoptera) between the three rose species. Maybe possible species-specific differences of insect and phytophagous species could not be shown on this taxonomic level.The generalist Spodoptera littoralis showed no principal differences in leaf consumption between hybrids and non-hybrids, but between the rose phenotypes with and without glandular trichomes. Hence, there was no evidence that herbivores contribute to stabilization of rose species identity, but the mother-based phenotype determines rose palatability. This opens a possibility for introgression of new hybridogenic genetic combinations into rose populations and is one basis for the diversity-generating potential of hybridisation.Multitrophic interactions between plants, herbivores and natural enemies affected the distribution of herbivores on the different host plant species. Investigations on different scales, like the host plant species, the consumer type and the geographical location could only reveal the relationship between host plant and herbivore. Further plant-herbivore studies should not be limited on simple interactions, but consider multitrophic interactions on different scales.