Interacting effects of grazing and habitat conditions on seedling recruitment and establishment

QUESTIONS: What is the effect of grazing on seedling recruitment and seedling establishment? To what extent is grazing dependent on local habitat conditions? LOCATION: Český Kras (Czech Karst), central Bohemia, Czech Republic. METHODS: In 2005, grazing was reintroduced on formerly grazed dry grassland. Twelve blocks of grazed and control plots were established. Seedlings of dicotyledonous plants were observed three times a year for 4 yrs, for a total of ten censuses. We tested the effects of grazing on seedling recruitment (entry of new individuals to the seedling stage) as well as seedling establishment (survival of seedlings for 1 yr after initial emergence). In addition, we explored the effect of soil depth – the most important environmental variable at our study site – on seedling response to grazing. The effects of grazing, soil depth and season on seedling number in plots were tested with generalized mixed effect models with a Poisson distribution. The effect of grazing on species composition was tested with canonical correspondence analysis. RESULTS: Grazing had a significant positive effect on seedling recruitment. Grazing affected the number of seedlings and species composition and facilitated the presence of dry grassland species. Seedlings of xerophilous species were primarily found in grazed plots, while control plots contained species preferring higher, closed sward. Grazing explained only 1.2% of the total variation in composition of recruited seedlings. We presume that the main factor determining species composition could be dispersal limitation. Grazing increased the number of recruited seedlings by 10.55%. The final number of established seedlings was unaffected by grazing. Furthermore, the interaction of grazing with soil depth had a significant positive effect on seedling recruitment, thus showing that the effect of grazing strongly depends on habitat conditions at the specific microsite within our location. CONCLUSIONS: Our results indicate that predicting higher generative reproduction on pastures based on a larger number of seedlings may be misleading if the species are limited by establishment – but not recruitment – as at our study site. Furthermore, our findings show that varying habitat conditions on small spatial scales can modify the impact of management on seedling recruitment and establishment.