Analysis of the factors that influence the quality of rapeseed and sunflower seeds and development of predictive models☆
This study examined the factors influencing the quality of rapeseed (Brassica napus L.) and sunflower (Helianthus annuus L.) seeds, to find out the factors having the greatest impact on protein and oil concentration. Historical data from variety and N fertilization trials were used for these analyse...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
EDP Sciences
2025-01-01
|
| Series: | Oilseeds and fats, crops and lipids |
| Subjects: | |
| Online Access: | https://www.ocl-journal.org/articles/ocl/full_html/2025/01/ocl240046/ocl240046.html |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | This study examined the factors influencing the quality of rapeseed (Brassica napus L.) and sunflower (Helianthus annuus L.) seeds, to find out the factors having the greatest impact on protein and oil concentration. Historical data from variety and N fertilization trials were used for these analyses, and to subsequently develop predictive models of seed quality.
The results showed that several environmental, agronomic and genetic factors significantly affect seed quality of oilseeds. These factors include region, soil texture, weather conditions (global radiation, precipitation, temperature), variety characteristics (maturity and flowering earliness, flowering duration, plant height, oil and protein potential concentrations), and crop practices (sowing date and density, N-fertilization).
The best-performing model for predicting protein and oil concentration in rapeseed and sunflower seeds was the Random Forest model. The model achieved good predictive accuracy, with over 84% of well-predicted values falling within acceptable ranges for rapeseed seed quality (oil and protein concentrations), and the same for sunflower protein concentration. However, some progress has to be done for sunflower oil concentration, as less than 59% of the situations were satisfactorily predicted. |
|---|---|
| ISSN: | 2272-6977 2257-6614 |