Intensification of domestic ungulate grazing delays secondary forest succession: evidence from exclosure plots

Questions What is the relative importance of direct herbivory compared to microsite modification in ungulate impacts on secondary forest succession? Do domestic ungulate impacts differ between small-seeded pioneers and large-seeded late successional species? Location Birchbeech secondary forest (Betula celtiberica and Fagus sylvatica), Bizkaia, Northern Spain. Methods We conducted a detailed spatial analysis of 216 permanent 1-m2 subplots that were distributed among six plots, of which three were fenced (each plot was 3600-m2) and three were unfenced (each plot was 1296-m2). Within each subplot, the emergence and survival rates of all tree, shrub and vine species were monitored. In total, 21 censuses of the subplots were conducted across 4 yr (19982001). Ungulate abundance was measured by using the pellet counting method. Structural equation models were applied to model the entire recruitment process. Results Emergence and survival rates differed between small- and large-seeded species, and the effect of domestic ungulates (primarily sheep) on these rates depended on seed traits. Sheep did not affect the emergence of large-seeded species (e.g. beech and ivy), but did result in a decline in their survival. In contrast, the emergence of small-seeded species (e.g. birch) was promoted through an increase in bare soil cover because of trampling by the ungulates, but no significant effect on survival was observed. Interestingly, the sheep prompted differences in the role of the understorey vegetation: the understorey composition was able to promote the survival of large-seeded species growing in unfenced conditions by reducing the risk of herbivore predation, but had a negative effect on the survival of small-seeded species because of low light availability under the shaded shrub canopies. Conclusions Our results clearly highlight that sheep can change the seedling bank structure in a secondary temperate forest, thereby affecting the dynamics and structure of forest remnants. This impact should be included as a critical driver in current predictive models of forest dynamics in temperate regions, since herbivore pressure is increasing in forests across Europe due to an increase in wild ungulate populations and livestock.