By analyzing the spatial variability of leaf traits with imaging and predictive models, researchers devised a reliable risk index that forecasts cold damage in strawberries with over 92% accuracy, paving the way for climate-resilient fruit production.
Strawberry (Fragaria × ananassa Duch.) is one of the world’s most consumed fruits, valued for its antioxidants and nutritional benefits. Yet its optimal growth requires mild temperatures of 15–25 °C, while exposure to low temperatures—even above freezing—can severely impair development. In over 90% of strawberry-growing regions, cold spells routinely occur between autumn and spring, making frost and chilling injury a persistent agricultural threat. Conventional weather-based warnings lack the resolution to capture real-time crop physiology, while current plant imaging approaches often overlook the spatial heterogeneity of stress responses. To address these gaps, researchers explored whether phenotypic spatial variability in chlorophyll fluorescence and hyperspectral imaging could serve as a foundation for a purely phenotype-based monitoring and forecasting system.
A study (DOI: 10.1016/j.plaphe.2025.100041) published in Plant Phenomics on 6 May 2025 by Zaiqiang Yang’s team, Nanjing University of Information Science & Technology, provides a powerful and reliable approach for monitoring cold stress and predicting damage risk in strawberries, offering valuable tools for precision agriculture under climate challenges.
In this study, ‘Toyonoka’ strawberry plants were exposed to four cooling gradients and three durations of cold stress to examine physiological responses and develop a predictive framework for cold damage. Researchers monitored the net photosynthetic rate (Pmax), relative electrolyte conductivity (REC), and total chlorophyll content (Chla+b) alongside imaging-derived phenotypic traits to capture changes in photosynthetic efficiency, tissue integrity, and pigment concentration. Results revealed complex dynamics: under mild gradients (19/9 °C and 16/6 °C), Pmax declined sharply after six days but partially recovered after nine, whereas under severe stress (10/0 °C) it plummeted by nearly half compared with controls. REC steadily increased with decreasing temperature and longer stress, exceeding 100% above controls under the harshest conditions, reflecting pronounced cell membrane damage. Chlorophyll content consistently decreased under moderate stress, reaching about 79% of controls after nine days, while under the lowest temperatures it initially rose but then dropped to its lowest levels, indicating early compensatory adjustments followed by pigment degradation. To integrate these parameters, principal component analysis generated a Photosynthetic Physiological Potential Index (PPPI), which effectively summarized plant vitality, declining progressively with lower temperatures and prolonged stress, even approaching zero under the harshest treatment. Building on this, machine learning models—XGBoost, AdaBoost, and RandomForest—were trained to link imaging features with PPPI and to calculate a Cold Damage Risk Index (CDRI). Six key spatial variability features were identified, with the fluorescence quenching parameter qP in the leaf vein direction emerging as the most critical predictor. Among models, XGBoost achieved the best performance (R²=0.98, RMSE=0.34), classifying cold damage risk with 92.13% accuracy and a Kappa of 0.904. By combining PPPI with Relative Negative Accumulated Temperature (RNAT), the researchers defined a robust CDRI with five risk levels, providing both real-time assessment and early-warning capability. Collectively, these results highlight how phenotypic imaging and AI can capture complex stress responses and offer reliable tools for safeguarding strawberry production against cold damage.
This research provides a blueprint for a new generation of agro-meteorological early-warning systems that do not rely solely on weather data but instead interpret the real-time physiological state of crops. For strawberry growers, such tools could inform timely interventions—such as protective coverings or irrigation scheduling—before irreversible damage occurs.
###
References
DOI
10.1016/j.plaphe.2025.100041
Original URL
https://doi.org/10.1016/j.plaphe.2025.100041
Funding information
This work was supported by the National Natural Science Foundation of China [grant number 42275200]; the Postgraduate Research & Practice Innovation Program of Jiangsu Province [grant number KYCX24_1446]; and the National Natural Science Foundation of China [grant number 32360443].
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.