Download
Summary
This thesis attempts to quantitatively determine some current impacts of feral horses on sub-alpine and montane areas in Victoria and New South Wales. Habitat preferences were analysed in relation to activities of the horses. Abundance estimates were made for two small catchments in southern Kosciusko National Park (NSW) and related to habitat use. Impacts on soils and the associated vegetation were assessed by comparisons between trampled sites and the adjacent untrampled areas. Assessments of disturbance along streambanks were related to vegetation type.
Feral horses are not randomly distributed over different habitats within Kosciusko National Park. Horses feed between 51 per cent and 75 per cent of the time, they are concentrated in grassland and heath communities, and avoid forests. Seasonal usage of habitat varies, with horses avoiding open grassland areas throughout the day during summer.
Groups consisted of small harems with one stallion, one or two mares and a foal or yearling. Population densities (45.6 animals km2 for summer) calculated from dung counts may be overestimated if defecation rates vary in response to forage availability. Population estimates of 216 horses for two catchments were based on a ten day period during summer.
There is an extensive network of tracks throughout the sites, with more soil lost from the wider tracks. Trampled sites are considerably more compacted than untrampled areas. Because the soils had minimal clay contents, they were susceptible to compaction. Dry soils were more prone to compaction than wet soils, which were more vulnerable to structural damage. In dry soils there was a compaction threshold of 20 – 50 passes (unshod horse) for significant compaction. Therefore, an average group (four horses) would only need to pass twice daily for less than one week along a new track to cause significant compaction.
Trampling also contributed to vegetation differences. Fewer species and fewer plants were found on trampled sites. Those plants showed characteristic morphologies and life forms (e.g.. prostrate, hemicryptophtyes), and favourable responses to increased lighting found on tracks. Exotic species colonised tracks, but not at the expense of the native species. Similarly, the reduced occurrences of exotics in the untrampled areas suggested the inability of exotics to compete with native species in unstressed situations.
Streambank disturbance was greatest in open areas with easily penetrated vegetation. Horses tended to cross streams through the softer Sphagnum, rather than the less penetrable heath complex. However, when crossing streams they avoided the Sphagnum in favour of the more solid ground under grassy and herbaceous vegetation. if this was available. Sphagnum was also trampled in search of food, although the horses did not appear to graze the moss. Little indication of stream incision was found. although some lateral channel migration was evident. This could be the result of increased runoff and drainage in response to a drier, more grassy and herbaceous vegetation following grazing and bank breakdown. Incision may eventuate if grazing maintains and increases the grassy vegetation.
Little differences exist between population densities of feral horses and habitat usage in the sub-alpine and montane sites. Similarly. no differences were found between the measured effects of horses in those areas. Although existing impacts have been quantified, long term monitoring is necessary to determine rates of impact and their relationships to feral horse population dynamics.