Journal Article

Perspectives on morphology, physiology, genetic polymorphism and machine learning in cucumber grafting under zinc toxicity

Toprak, Seher, Aydın, Alim, Başak, Hakan, Mavi, Kazım, Coşkun, Ömer Faruk, Yetişir, Halit

2025

Overview

Background Heavy metal contamination in agricultural soils disrupts plant growth and metabolism. Although zinc (Zn) is a necessary element, concentrations above 50 ppm can be toxic to plants. Grafting has emerged as a potential strategy to mitigate heavy metal stress by enhancing tolerance, reducing translocation to edible plant parts, and maintaining crop productivity. Results In this study, cucumber ( Cucumis sativus L.) was grafted onto Cremna F1, Maximus F1, and TZ148 F1 rootstocks and evaluated under four Zn conditions (Zn-deficient 0 ppm, 10 ppm, 50 ppm, and control with 0.05 ppm Zn) in a hydroponic system. Grafted plants exhibited significant improvements, including a 21.5% increase in shoot length and 27.3% higher fresh root weight compared to non-grafted controls. Chlorophyll content remained stable in grafted plants (1.42 ± 0.03 mg g⁻¹ FW) but declined in non-grafted plants (1.18 ± 0.07 mg g⁻¹ FW) under 50 ppm Zn stress. Fruit set (%) decreased from 0.84 ± 0.04 in control plants to 0.61 ± 0.02 under 50 ppm Zn. ISSR band shifts (average polymorphism 91%) were treated as stress-related genotoxicity proxies rather than fixed genotypic differences. PCA indicated moderate separation among treatments rather than fully distinct clusters in grafted plants. Machine learning (ML) analyses complemented these findings: Random Forest regression achieved a five-fold cross-validated performance of R² = 0.64, with fruit curvature, root dry weight, and plant height identified as the strongest predictors of yield. The Zn Tolerance Index (ZTI) ranked TZ148-Cagla as the most tolerant combination, retaining 25% of the control yield under 50 ppm Zn. Conclusion These results demonstrate that grafting enhances cucumber tolerance to Zn stress by improving morphological, physiological, and biochemical performance while maintaining yield stability. ML analyses strengthened these conclusions by quantitatively ranking trait importance and providing a predictive framework for rootstock selection. Together, these findings highlight grafting as an effective strategy for reducing heavy metal accumulation in edible plant parts and sustaining crop productivity, thereby supporting sustainable practices in Zn-affected systems, including contaminated irrigation scenarios.

• Publisher: BioMed Central,BioMed Central Ltd,BMC
• Subject: Agriculture / algorithms / Biomedical and Life Sciences / Botanical research / Botany / Chlorophyll / Chlorophyll - metabolism / Cucumbers / Cucumis sativus / Cucumis sativus - anatomy & histology / Cucumis sativus - drug effects / Cucumis sativus - genetics / Cucumis sativus - growth & development / Cucumis sativus - physiology / Ethylenediaminetetraacetic acid / food plants / fruit set / fruits / Genetic aspects / genetic polymorphism / Genetic polymorphisms / Genetic research / genotoxicity / Grafting / Heavy metal tolerance / heavy metals / hydroponics / irrigation / Life Sciences / Machine Learning / metabolism / Morphology / Physiological aspects / plant growth / plant height / Plant Roots - drug effects / Plant Roots - genetics / Plant Roots - growth & development / Plant Roots - physiology / Plant Sciences / Polymorphism, Genetic / Rankings / Rootstock-scion interaction / rootstocks / Soil Pollutants - toxicity / Stress, Physiological / Structure / Tree Biology / Zinc / Zinc - toxicity / Zinc compounds / Zinc in the body / Zinc stress