This project brings together specialists in entomology, macadamia pollination, bee behaviour, neuromorphic systems and machine learning from Western Sydney University to use neuromorphic machine learning to search satellite imagery for undiscovered wild populations of M. tetraphylla, the species that has lost the most habitat through clearing. Selected sites will then be monitored for pest and pollinator activity. The first step will be to perform a comprehensive analysis of satellite imagery to identify wild M. tetraphylla trees within and beyond their expected range. The morphology of newly discovered trees, nuts and seedlings will be studied to analyse their genetic diversity and assess the habitats they are located in to compare them with known populations and identify new priority conservation locations.
The second step is to deploy neuromorphic AI powered cameras around M. tetraphylla trees in different ecosystems and landscape contexts (for example, at varying distance from commercial Macadamia orchards) to identify and quantify the visitation rates of pests and pollinators, and compare the flowering and fruiting phenology of distinct subpopulations.
How trees are pollinated has a big impact on genetic diversity, as many macadamias will not produce self-pollinated seed – understanding how near the next mature tree must be for successful cross-pollination is important for conservation planning.
This Ian and Janet McConachie Macadamia Conservation Research Grant provides funding for three years, supporting three field trips and adapting technology to learn more about the interactions of plants, pests and pollinators.
Sign up to the Wild Macadamia Champions newsletter to stay updated on research findings.
Tissue Culture in the Tool-Kit for endangered Macadamia jansenii