Natural regeneration of woody vegetation in pastoral hill country: A case study of Oashore Station, Banks Peninsula : A thesis submitted in partial fulfilment of the requirements for the Degree of Master at Lincoln University

There is a recognised need to establish more permanent indigenous forest throughout the world to mitigate climate change and enhance declining biodiversity. In New Zealand, the Climate Change Commission He Pou a Rangi has recommended the establishment of 300,000 hectares of new indigenous forest by 2035, with a focus on areas of hill country with marginal productivity for agriculture. A significant barrier to the achievement of this goal is the cost and logistical difficulties of restoration planting and direct seeding in hill country landscapes. Reliance on natural regeneration has the potential to be a more practical and economically viable option, particularly where large areas are involved. However, the rate of natural regeneration and key factors enabling it within pastoral hill country landscapes are not well understood. Based on aerial imagery, object-based image classification was used to detect changes in woody vegetation over a 16-year period on Oashore Station, a 540ha hill country property on Banks Peninsula, Canterbury. The focus was on identifying areas of the property that had changed from grassland to woody shrubland as the first stage in a successional process of natural regeneration from pasture to indigenous forest. Random forest machine learning and logistic regression were then used to evaluate the influence of different variables on the observed natural regeneration, including the impact of environmental factors and grazing regimes. The results for this study indicate that only 3.3% of Oashore Station experienced observed natural regeneration of woody vegetation in pasture between 2003 and 2019. The key factor influencing natural regeneration was the proximity of existing woody vegetation, with areas within 2m of existing woody vegetation being significantly more likely to experience natural regeneration. Cattle grazing was also found to have a significant negative impact on regeneration of woody vegetation, with regeneration most likely to occur in areas that had been free of cattle grazing for at least 10 years. Grazing by sheep and other environmental factors were not found to have a significant impact on natural regeneration. These findings have implications for hill country farmers considering natural regeneration on pastoral land with similar conditions and can help to identify locations where regeneration is most likely to occur. Areas with a higher proportion of existing woody vegetation should be prioritised for natural regeneration, with cattle grazing generally excluded from such areas. However, even in favourable conditions, the establishment of indigenous forest through natural regeneration is likely to be a slow, incremental process. This should be recognised in financial incentive schemes such as the Emissions Trading Scheme and has implications for the achievement of broader goals seeking to increase the amount of permanent indigenous forest on marginal hill country. If significantly faster rates of natural regeneration are desired, additional strategic interventions and support may be necessary