The effects of livestock grazing in a snow tussock (Chionochloa species) grassland, Mt Grand, Otago

Uncertainty surrounds the extent to which livestock grazing drives changes in tussock grassland vegetation. The overall aim of this study is to examine the effects of livestock grazing on high altitude snow tussock grassland, Mt Grand, Otago. The study is divided into three objectives. First, I examined changes in vegetation composition and structure across 40 transects in snow tussock grassland, above 1000 m on Mt Grand. These transects were established in December 1996 and were remeasured in December 2001. Averaged across all transects there was a 7% decline in vegetation biomass between measurements. Fertiliser treatment was the strongest predictor of change in biomass. Fertilised transects declined in biomass on average 20 %, while unfertilised transects changed very little. Vegetation trends in fertilised transects generally reflected the trends of key species or species groups. In fertilised transects, grass species A. odoratum (- 67%) and F. novae-zelandiae (- 44%) declined in biomass, where they were more abundant in 1996, while Hieracium species increased significantly, led by a 32% increase in H. lepidulum. However, in unfertilised transects the lower abundance and smaller decline of A. odoratum (-37%) and F. novae-zelandiae (-9%), was offset by a disproportionately larger increase in the biomass of H. lepidulum (69%). In unfertilised transects, biomass of Chionochloa species declined by 12%, while large herb species (> 2cm in height) increased by 25%. Large herb species declined in biomass in fertilised transects by 23%. Low soil fertility, spring drought, competition with dominant species, and low grazing pressure are all possible explanations for the observed vegetation change in these grasslands. Second, I quantified the effects of grazing exclusion on tussock grassland vegetation by comparing vegetation response in 12-paired grazed and exclosure plots. The plots were established in December 1997 and remeasured in December 2001. After four years, grazing exclusion had no significant effect on vegetation composition and structure, as all plots remained similar irrespective of grazing treatment. General vegetation trends in all plots were similar to transect trends, with a significant decline in the biomass of tussock species and inter-tussock species, particularly introduced species, and a widespread increase in Hieracium species. Aspect was the most significant predictor of change in biomass with shady plots (-28%) declining more than sunny plots (-7%). It is likely that four years was insufficient time to detect significant vegetation differences due to grazing exclusion, taking into consideration low grazing pressure, high elevation and low fertility conditions in this grassland. Third, I assessed 12 plant species for evidence of grazing. The same species were analysed for digestibility and mineral concentration to determine if these characteristics were associated with grazing by sheep. Heavily grazed species were mainly introduced species (Lotus pedunculatus, Trifolium repens, A. odoratum and H. praealtum), with higher mineral concentration and digestibility than native species. Several native species (Celmisia lyalli, Chionochloa species and Aciphylla aurea) were also grazed within exclosure plots, possibly by hares. Most of the 12 species were grazed in developed sites, where they were more abundant, however grazing of Chionochloa species was greatest in the undeveloped community, where the species was less abundant and also declined significantly in biomass. This could suggest that sheep tend to graze Chionochloa species where preferred introduced species are less abundant, and is of concern for Chionochloa persistence in undeveloped areas. There is evidence of significant vegetation change occurring in this snow tussock grassland, with a decline in biomass of native tussock species, inter-tussock species and an increase in the biomass of Hieracium species. However, the cause of vegetation change is not clear and appears to be unrelated to the effects of grazing, although the significance of some vegetation trends may not be apparent after a period of 4-5 years. Appropriate management will depend on whether objectives are production or conservation based and will need to be site specific, as vegetation change varies in undeveloped and developed blocks. The dramatic increase in H. lepidulum biomass is an issue, particularly under current low intensity grazing regimes that are unlikely to halt the species invasion in Chionochloa grasslands