By combining drone technology, genomic data and advanced models, one of the world’s most important forage crops, lucerne, can be bred more quickly and efficiently.
In his PhD research, Harkingto Harkingto is investigating how drones can be used in lucerne breeding. After flying over trial fields and collecting data, the results show that drones can provide reliable measurements of both plant height and yield in small trial plots. This is an important finding, as such measurements are usually carried out manually. This is both time-consuming and costly, which has limited the amount of data plant breeders have been able to collect.
The research is based on field trials conducted at the Centre for Climate-Controlled Plant Research (SKP), the open-field department at Vollebekk in Ås, where two different types of genetic material were tested. One trial included 135 half-sib families from a broad European reference population, whilst another consisted of 44 historical Norwegian breeding populations.
Both trials were measured both manually and using a drone. This made it possible to compare traditional methods with new, automated solutions.
The results show that specific vegetation indices provided stable and strong correlations with actual yield. When different types of data were combined in models, the accuracy increased further. Drone-based phenotyping provides reliable estimates of both crop height and dry matter yield. Furthermore, the use of advanced machine learning models results in very good yield predictions.
Genes reveal the varieties of the future
In addition to drone technology, Harkingto has investigated how genetic information can be used in plant breeding. DNA analyses can predict which plants will perform best, even before they have been fully tested in the field. The results showed moderate to high accuracy for several key traits, such as flowering time and yield.
Harkinto’s PhD project shows that drones can provide reliable measurements of plant height and yield in lucerne field trials, and can replace time-consuming manual recordings
The fact that genomic prediction makes it possible to select promising plant material earlier in the process means that the development of new varieties can proceed more quickly. This is important in the face of climate change and the need for sustainable food production.
This PhD provides both new knowledge and practical tools that can be put to use in the field, and demonstrates how modern technology can make plant breeding both faster and more precise by combining drones, genomic data and statistical models.
More robust and productive fodder crops can contribute to more climate-friendly agriculture, for example by reducing the need for fertiliser and increasing the stability of yields.
Harkingto Harkingto is 43 years old and comes from Medan in Indonesia. On 3 July, he will defend his doctoral thesis entitled: ‘Alfalfa adaptation to a northern climate: Characterisation of genetic material and development of high-throughput airborne phenotyping and genomic prediction’.
Harkingto’s principal supervisor has been Professor Åshild Ergon (NMBU). His co-supervisors have been Sahameh Shafiee (NMBU), Nelson Nazzicari (CREA) and Helga Amdahl (Graminor).