Background
Genetic trials are central to breeding programmes: we use them for estimating genetic parameters, predicting breeding values, training genomic models, etc. Forestry trials often show strong within-trial trends, which means that we must remove substantial environmental noise to observe a clear genetic signal.
Historically, we could only remove some noise through experimental design and later improve this approach by using spatial analysis. We now routinely have additional data available from Unpiloted Aerial Vehicle (UAV) imagery and LiDAR, which we are not fully exploiting.
The project
The NZ School of Forestry at the University of Canterbury (UC) is offering a 3-year PhD project to investigate how tree and terrain related data can be used to improve within-trial analysis, by adding spatial analysis, topographic indices and competition indices derived from UAV LiDAR data. The project is funded by the NZ Radiata Pine Breeding Company (RPBC). The student will work closely with RPBC, which specialises in breeding elite radiata pine germplasm for New Zealand and Australian forest owners. The student will work with a research team led by Professor Luis Apiolaza and Dr Vega Xu from UC, and Dr Mark Paget and Dr Sai Arojju from RPBC.
The research project will include:
• Investigate the derivation of multiple competition indices based on UAV LiDAR.
• Evaluate the utility of multiple topographic indices and tree metrics derived from UAV LiDAR in the context of RPBC genetic trials.
• Compare the base genetic analysis (experimental design + genomic pedigree) to base + spatial analysis, base + spatial + topographic + competition.
Supervisors
Primary Supervisor: Vega Xu
Second Supervisor: Luis A Apiolaza
Key qualifications and skills
The ideal applicant will have a GPA of 7.0 (A-) or higher, and hold a four-year bachelor’s degree with first class honours or a Master’s degree in forestry, remote sensing or data science.
Applicants must demonstrate:
• Strong data analysis skills, with proficiency in R or Python
• Familiarity with remote sensing and geospatial data processing (e.g. LiDAR and UAV imagery)
• Ability to learn and apply quantitative genetics and advanced statistical methods in the context of tree breeding.
• A valid driver’s licence and willing to undertake field work.
Does the project come with funding
Yes: Scholarship covering full university fees and a stipend of NZ$34,000 p.a. for three years.
Final date for receiving applications
Ongoing
How to apply
By email to primary supervisor with specified documents
Keywords
remote sensing; quantitative genetics; lidar; spatial analysis; machine learning
