Search

To provide necessary information for the management of biodiversity on Sable Island, Nova Scotia, we studied the effects of feral horses on vegetation using exciosures and ancillary observations. Nine plant communities inside and outside of six exclosures were compared using various vegetation parameters and Mann-Whitney tests to evaluate the significance of dif- ferences. The most important findings were as follows: (1) effects of horses were greatest in the Marram (Ammophila bre- viligulata) grassland and much less in the communities that were not dominated by Marram Grass; (2) effects on Marram grassland varied substantially among sites; (3) the cover of standing litter of herbaceous plants was on average of 9.3 times greater inside exclosures in grassland habitats; (4) the cover of living foliage was usually higher inside exclosures, but not all differences were significant; (5) species richness and species diversity were not substantially affected; (6) the average cover of Marram Grass, the most abundant plant and a key sand binder on the island, was greater inside exclosures in six of seven study sites, significantly so in three of them; and (7) there were inconsistent differences in cover of other species at different sites. Wetland habitats cover a relatively small portion of Sable Island, but they support much of the plant biodiver- sity. There is evidence of strong but variable effects of horses on wetland vegetation. "Horse lawns" are littoral habitats dominated by Agrostis stolonifera (Carpet Bentgrass) and other low-growing plants. The lawn habitats represent less than 1% of the island's vegetation, and their presence is believed to be due to grazing and trampling by horses. [ABSTRACT FROM AUTHOR]

Indicators are being sought for monitoring the ecological integrity of forests and other kinds of ecosystems. Biological measures are commonly used as indicators because of their inherent ecological importance and ability to provide insight into environmental change. Such measures are commonly based on data from sets of permanent plots in which the abundances of plant species are monitored. However, the data may be difficult to interpret, especially if corresponding information on natural and anthropogenic stressors is lacking. In this review, we examine general principles of indicator use and discuss the types of plot-based compositional measures obtained from vegetation that may be most relevant for monitoring ecological integrity. Our focus is on the ground vegetation of forested ecosystems, but the principles discussed are relevant to other vegetation types. Individual plant species, guilds, aliens, diversity indices, Ellenberg indicator values, the floristic quality assessment index, multivariate and multimetric indicators are examined, as well as concepts of threshold changes and the need for reference states. The usefulness of any given approach tends to be highly context specific. In particular, the value of using individual species as indicators is highly dependant on factors such as the character of the floristic community of interest and the types and intensities of anthropogenic stressors. Alien species are considered to be especially valuable indicators of changes in ecological integrity due to their established relationships with anthropogenic stressors, known historical state, relevance to all floristic communities, and ability to cause undesirable changes to biodiversity and ecological processes. [ABSTRACT FROM AUTHOR]