The effects of hemiparasitic plant removal on community structure and seedling establishment in semi-natural grasslands

Question: Hemiparasitic plants can profoundly affect the structure of the community in which they occur, mainly due to parasitic suppression of hosts. As a consequence, non-host species have the opportunity to colonize resulting gaps. In contrast to most grassland species, hemiparasites are generally short-lived and can reach high densities; as a consequence, vegetation gaps are left after their death. These gaps form microsites more suitable for seed germination and therefore might increase recruitment of other species. Which species suffer or profit from parasitism and is there a positive effect on seedling establishment? Location: Semi-natural grasslands in northern Belgium. Methods: We selected two hemiparasitic plant species from contrasting vegetation types: Rhinanthus angustifolius growing in mesotrophic grassland and Pedicularis sylvatica growing in oligotrophic heath-grassland. A weeding experiment was set up at six sites in which the hemiparasite was repeatedly removed in half of the plots during three growing seasons. The abundance change of individual species was compared between weeded and control plots. After the second growing season, seeds of up to ten species were added. The number of seedlings in the third year was then compared between weeded and control plots. Results: Rhinanthus removal significantly affected the abundance of species relative to control plots, both positively and negatively, and decreased the species evenness. Pedicularis removal only increased the abundance of some species. Only Juncaceae (but no other graminoid families) increased after Rhinanthus and Pedicularis weeding, and there was considerable variation within functional groups. Moreover, based on our observations, we propose as a new hypothesis that species with persistent clonal spread are more vulnerable to parasitism. Once attached, resources are potentially drawn from a whole network of interconnected ramets. Finally, only half of the sown species successfully established seedlings; hemiparasite removal had a significantly negative effect on seedling number for two of these species. Conclusions: Effects of hemiparasites on species differ considerably, also within functional groups. Persistent clonal spread emerges as an important plant trait determining vulnerability to hemiparasites. Finally, our results suggest that hemiparasitic plants might have a limited positive effect on seedling establishment in these semi-natural grasslands where chances for successful establishment were shown to be low