Phenotyping is the measurement of tree characteristics, such as height, or needle colour, which represent the combined effects of the tree’s genetics and growing environment. Emerging technologies such as laser scanning (LiDAR), dense image matching, and multispectral sensors, operated from UAV or ground-based platforms are providing opportunities to achieve high volume, high quality, cost effective, phenotypic data for tree breeders. Recently developed techniques such as deep learning, and machine learning offer opportunities to identify and segment objects such as trees from two- and three-dimensional data sets and identify relationships hidden in large data sets. There are significant challenges in processing high density data sets to extract tree phenotypes. There are further challenges in utilising such unprecedented levels of information in statistical analyses of trials. Current methods are designed to process phenotypic observations for a handful of traits. High throughput phenotyping could produce large numbers (tens to thousands) of phenotypic observations per tree. It can also provide detailed spatial information about the trees growing environment. Such as tree locations, terrain characteristics, and competition levels. Such large amounts and types of phenotypic data potentially demand novel analytical and modeling methods.
The Radiata Pine Breeding Company (RPBC) is sponsoring a PhD scholarship, offered by the University of Canterbury, for the development of novel tree phenotyping methods to accelerate rates of genetic improvement. RBPC breeds elite genetic material, and provides knowledge, support and tools to continuously improve profitability for Australasian radiata pine (Pinus radiata) forest owners. RPBC and research partner Scion are committed to ongoing development and application of state-of-the-art methods to support these goals. The scholarship will cover university tuition fees, a stipend of $25,000 p.a. and some research-related costs. It is expected the PhD will be completed in a 3 year period. A preferred starting date is March 2020.
Potential areas for research focus could include one or more of the following, or related, topics:
The ideal applicant will have a GPA of 8.0 (A-) or higher, a four-year bachelor degree with first class honours, or a Masters degree in a relevant field; this may include forestry, ecology, geomatics, computer science, mathematics, statistics or other related field. Preference maybe given to applicants that have demonstrated research skills in remote sensing. Candidates who have published research in peer-reviewed journals are strongly encouraged to apply. Candidates should have a valid drivers licence and be willing and capable to undertake field work, dependant on the nature of the agreed topic.