AI tool Spotibot accelerates rose breeding for disease resistance
en-GBde-DEes-ESfr-FR

AI tool Spotibot accelerates rose breeding for disease resistance

04/09/2025 TranSpread

By providing objective, high-resolution data, Spotibot represents the first easy-to-use tool for rapid phenotyping in roses, offering breeders a transformative way to select resistant cultivars more accurately and efficiently.

Roses belong to the Rosaceae family, alongside apples, almonds, and strawberries, and hold enormous economic and cultural significance. Global cut rose production exceeded $3 billion in 2022, with the Netherlands leading exports and the U.S. as the largest importer. Roses, cherished worldwide for their beauty and cultural value, face a persistent threat from Botrytis cinerea, the grey mould fungus responsible for devastating post-harvest losses. In roses, the pathogen causes necrotic brown lesions that rapidly spread through spores, especially during harvest and post-harvest storage. No complete resistance has been identified, and existing defense is controlled by multiple genes with small effects. Because of this complexity, precise and objective phenotyping is essential to advance breeding for Botrytis resistance.

A study (DOI: 10.1016/j.plaphe.2025.100029) published in Plant Phenomics on 19 March 2025 by Dan Jeric Arcega Rustia’s team, Wageningen University & Research, opens new opportunities for precision rose breeding by delivering rapid, objective, and scalable phenotyping.

The study first optimized deep learning models for petal and Botrytis lesion segmentation to enable rapid, objective disease phenotyping in roses. Several YOLOv8 segmentation models were tested, and results showed that the petal segmentation model achieved high accuracy regardless of model size, with YOLOv8n-seg selected for deployment due to its fast inference time and stable performance. For lesion detection, YOLOv8s-seg achieved the highest F1-score, but YOLOv8n-seg was ultimately chosen for mobile use because of its much lower inference time, which is critical when analyzing multiple petals per image. Although the models performed well across colors, orange petals posed difficulties because lesions were less visually distinct, while red, pink, and white petals were more accurately segmented. Speed benchmarking demonstrated that Spotibot could process 4,155 images in 2.7 hours on a GPU-equipped PC, compared with over 7–11 hours on standard smartphones, which still represents a major improvement over manual scoring that typically takes 23 hours. Qualitative evaluation confirmed that the algorithm consistently detected lesions, with only minor boundary errors or occasional misidentification. To validate accuracy, subjective and objective results were compared using Spearman correlation, revealing strong positive relationships across lesion area, diameter, and lesion-to-petal ratios, with coefficients as high as 0.88. These correlations held across petal colors, although orange petals again showed lower agreement. Finally, a two-way ANOVA of 212 genotypes demonstrated that objective data captured stronger differences among cultivars than subjective scores, allowing clearer discrimination between resistant and susceptible genotypes. Moreover, when challenged with different Botrytis isolates, objective measurements enabled ranking of genotypes and differentiation between aggressive and mild isolates. Overall, the optimized models and Spotibot app proved capable of delivering fast, accurate, and reproducible lesion quantification, enhancing the precision of resistance screening in rose breeding.

Breeding programs, which often require 7–20 years to release new cultivars, can now accelerate selection for Botrytis resistance. The app’s accessibility on smartphones eliminates the need for costly equipment and makes it feasible for breeders worldwide to incorporate AI-driven screening into their pipelines. Beyond roses, Spotibot’s approach may be adapted for other crops where quantitative resistance traits are controlled by subtle genetic effects. In addition to breeding, the tool could improve post-harvest quality management by enabling early detection of infection and minimizing losses during storage and transport.

###

References

DOI

10.1016/j.plaphe.2025.100029

Original URL

https://doi.org/10.1016/j.plaphe.2025.100029

Funding information

This work is part of a public-private-partnership project called “Variation in S gene and S gene dosage to increase Botrytis resistance in rose and strawberry” (Project number: LWV21.027) funded by the Dutch Topsector for Knowledge and Innovation and co-funded by the corresponding industry partners Dümmen Orange, De Ruiter Innovations, Interplant, Meilland International and United Selections.

About Plant Phenomics

Plant Phenomics is dedicated to publishing novel research that will advance all aspects of plant phenotyping from the cell to the plant population levels using innovative combinations of sensor systems and data analytics. Plant Phenomics aims also to connect phenomics to other science domains, such as genomics, genetics, physiology, molecular biology, bioinformatics, statistics, mathematics, and computer sciences. Plant Phenomics should thus contribute to advance plant sciences and agriculture/forestry/horticulture by addressing key scientific challenges in the area of plant phenomics.

Title of original paper: Spotibot: Rapid scoring of Botrytis lesions on rose petals using deep learning and mobile computing
Authors: Dan Jeric Arcega Rustia a 1, Maikel Zerdoner b c 1, Manon Mensink d, Richard GF. Visser b, Paul Arens b, Suzan Gabriëls b
Journal: Plant Phenomics
Original Source URL: https://doi.org/10.1016/j.plaphe.2025.100029
DOI: 10.1016/j.plaphe.2025.100029
Latest article publication date: 19 March 2025
Subject of research: Not applicable
COI statement: The authors declare that they have no competing interests.
Attached files
  • Figure 2. Rose Botrytis lesion detection and segmentation algorithm pipeline.Each output box includes several output values as described on the lower right of the figure, where each red-colored text represents an output value. The orange line on each detected object indicates Botrytis lesion instance segmentation results and each white line indicates petal instance segmentation results.
04/09/2025 TranSpread
Regions: North America, United States
Keywords: Applied science, Engineering, Artificial Intelligence, Science, Agriculture & fishing

Disclaimer: AlphaGalileo is not responsible for the accuracy of content posted to AlphaGalileo by contributing institutions or for the use of any information through the AlphaGalileo system.

Testimonials

For well over a decade, in my capacity as a researcher, broadcaster, and producer, I have relied heavily on Alphagalileo.
All of my work trips have been planned around stories that I've found on this site.
The under embargo section allows us to plan ahead and the news releases enable us to find key experts.
Going through the tailored daily updates is the best way to start the day. It's such a critical service for me and many of my colleagues.
Koula Bouloukos, Senior manager, Editorial & Production Underknown
We have used AlphaGalileo since its foundation but frankly we need it more than ever now to ensure our research news is heard across Europe, Asia and North America. As one of the UK’s leading research universities we want to continue to work with other outstanding researchers in Europe. AlphaGalileo helps us to continue to bring our research story to them and the rest of the world.
Peter Dunn, Director of Press and Media Relations at the University of Warwick
AlphaGalileo has helped us more than double our reach at SciDev.Net. The service has enabled our journalists around the world to reach the mainstream media with articles about the impact of science on people in low- and middle-income countries, leading to big increases in the number of SciDev.Net articles that have been republished.
Ben Deighton, SciDevNet

We Work Closely With...


  • e
  • The Research Council of Norway
  • SciDevNet
  • Swiss National Science Foundation
  • iesResearch
Copyright 2025 by AlphaGalileo Terms Of Use Privacy Statement