Rewilding with large herbivores: Direct effects and edge effects of grazing refuges on plant and invertebrate communities

Natural grazing refuges can contribute significantly to plant and animal diversity in grazed ecosystems, particularly when herbivore densities are high. When natural grazing refuges are absent, artificial refuges could be created, for example by means of fencing. It remains, however, unclear how grazing refuges affect the diversity of various taxa in their surroundings. Edge effects can be expected to be positive for biodiversity because at these edges minimal disturbance is combined with beneficial light conditions for plant diversity. This can be expected to decrease with increasing distance and to differ between matrix vegetation types. Here, we investigated the impact of herbivore exclusion through fencing on communities of plants and various invertebrate taxa in a rewilding area on very productive soil, the Oostvaardersplassen, The Netherlands. The area is grazed year round by Heck cattle, konik horses and red deer, at a combined density of approximately 2.4 animal per ha. Ten exclosures (13 x 12 m) were erected in 2010 and plant communities were monitored for three years. In the third year, pitfall trapping, earthworm counts and plant surveys were performed in the centre, at the edge, at 10 m and at 20 m distance from each exclosure. Plant species richness declined strongly at the centre of the exclosures, but remained high outside the exclosures and at their edges. Earthworm, isopod, myriapod and ground beetle diversity increased in the exclosures, but showed small or no differences in species composition, while weevils showed a decrease. Spider, true bug, and leafhopper diversity did not differ, but showed large changes in species composition. For leaf-, dung and click beetles, neither diversity nor species composition differed systematically between treatments. For all taxa, diversity at the exclosure edges was as high as in the most diverse treatments, thus, combining all taxa to calculate multidiversity showed species richness to peak at the edge of the exclosures, but this positive edge-effect extended less than 10 metres. We conclude that when natural grazing refuges are not present, the creation of grazing refuges may thus be an effective management tool to increase diversity. Because exclosure edges support the highest multidiversity, maximising the edge length will have most beneficial effects for grassland flora and fauna, which can be achieved by creating many small, rather than few large refuges, or by creating serrated rather than straight edges. Such positive effects can be expected to be beneficial to higher trophic levels such as birds and other vertebrates