The ecological and physiological drivers and the development of improved landscape detection of the invasive perennial, Eragrostis curvula, across the southern tablelands of NSW, Australia

Abstract

This thesis addresses managing and quantifying the associated impacts of the invasive species Eragrostis curvula, a non-native grass that poses ecological and economic threats in Australia. It aims to understand its ecological impacts to improve management strategies for sustainable ecosystems. The study utilised convolutional neural networks to assess the feasibility of using drone imagery for early detection of E. curvula infestations, a relatively new endeavour in invasive species management in Australian grassland systems. Key findings from this research highlight E. curvula's significant association with species richness and diversity, revealing a reduction in biodiversity in areas dominated by the grass. Eragrostis curvula appears to exhibit some degree of drought tolerance in the form of biomass retention, which in part likely explains its dominance over other species. Moreover, the presence of herbicide-resistant E. curvula populations underscored the challenges of relying on chemical controls and the necessity for integrated management strategies. The investigation into E. curvula's ecological dynamics revealed that variables, such as temperature, proximity to roads and watercourses, significantly influence its distribution. This suggests that E. curvula's spread can be partly attributed to human activities and climatic suitability, underlining the importance of considering these factors in management plans. The significant findings advocate a shift towards integrated management approaches incorporating technological advancements for early detection and monitoring alongside traditional control methods. The insights generated from this research hold implications for natural resource management agencies and restoration ecologists, providing a robust foundation for informed decision-making in ongoing invasive species management.