Selecting Better Potatoes with Drone-Gathered Data

Jim Crocker
7th March, 2024

Selecting Better Potatoes with Drone-Gathered Data

Potato (Solanum tuberosum)

Photographer: Paula Greer

Key Findings

  • Study at Heinrich-Heine-University found drones can predict potato traits like yield and maturity
  • Combining drone data with genetic data improves prediction for 20 out of 22 potato traits
  • This method could speed up breeding new potato varieties, aiding global food security
In the quest to feed a growing global population, scientists are constantly seeking innovative ways to boost crop production. One such crop at the forefront of this battle is the potato, a staple food with a wide distribution and significant potential for enhancing food security, particularly in developing nations[2]. A recent study from Heinrich-Heine-University has made strides in the field of predictive breeding—a technique that could revolutionize how we select and breed potato varieties with superior traits[1]. Predictive breeding involves using models to predict the future performance of plant varieties, which can save time and resources in agricultural production. There are two main types: genomic selection (GS), which uses genetic markers, and phenomic selection (PS), which uses observable characteristics. The Heinrich-Heine-University study sought to explore the capabilities of phenomic prediction in potato breeding, compare it to genomic prediction, and assess the predictive power of combining both types of data. The study's findings are significant as they build upon previous research showing the potential of phenomic data, gathered by technologies like drones, to predict plant performance. For instance, earlier studies demonstrated that temporal phenomic prediction (TPP) using drone data could outperform genomic predictions in maize[3]. Similarly, in wheat breeding, models that combined near-infrared spectroscopy data from different environments were as accurate as genomic selection[4]. The Heinrich-Heine-University study evaluated a diverse range of potato traits, including yield, maturity, foliage development, and emergence. The researchers found that phenomic predictions varied greatly but showed high predictive abilities for certain traits. For example, yield and maturity had high predictive abilities of 0.45 and 0.88, respectively. This suggests that phenomic data can be particularly useful for predicting complex traits in potatoes, echoing findings in triticale where phenomic prediction was more effective for complex traits than for those controlled by a few genes[5]. Interestingly, when the study combined genomic and phenomic data, the predictive abilities increased for 20 out of 22 traits. This indicates that genomic and phenomic information is complementary, and using both can enhance the accuracy of predictions. This mixed approach aligns with the concept that combining different types of data can lead to better results, as seen in wheat breeding, where a model combining genotyping and near-infrared spectra yielded the highest prediction ability[4]. The study's results are particularly relevant for early stages of potato breeding, where the number of tubers per entry is limited, and the costs of genotyping are prohibitive. Phenomic selection could allow breeders to apply predictive breeding principles during these initial stages, potentially accelerating the development of new potato varieties and enhancing the efficiency of breeding programs. In conclusion, the Heinrich-Heine-University study provides valuable insights into the utility of phenomic prediction in potato breeding. By demonstrating that phenomic data, especially when combined with genomic data, can predict a variety of important traits, this research paves the way for more resource-efficient breeding strategies. As the world grapples with the challenge of increasing agricultural productivity, such advancements in predictive breeding could prove crucial in ensuring food security for future generations.

AgricultureBiotechPlant Science

References

Main Study

1) Using drone-retrieved multispectral data for phenomic selection in potato breeding.

Published 6th March, 2024

https://doi.org/10.1007/s00122-024-04567-3

Related Studies

2) The Potato of the Future: Opportunities and Challenges in Sustainable Agri-food Systems.

https://doi.org/10.1007/s11540-021-09501-4

3) Temporal phenomic predictions from unoccupied aerial systems can outperform genomic predictions.

https://doi.org/10.1093/g3journal/jkac294

4) Phenomic selection in wheat breeding: identification and optimisation of factors influencing prediction accuracy and comparison to genomic selection.

https://doi.org/10.1007/s00122-021-04005-8

5) The performance of phenomic selection depends on the genetic architecture of the target trait.

https://doi.org/10.1007/s00122-021-03997-7


Related Articles

An unhandled error has occurred. Reload đź—™