Seed coating improves agricultural handling and germination, but obscures visual identification and complicates quality control and regulatory compliance. Conventional de-coating is destructive, time-intensive and raises concerns about reliability and analyst safety. To compensate for these limitations, we introduce PelletRayTion, a non-destructive, automated alternative for contaminants detection in pelleted seeds using high-throughput tomography and unsupervised machine learning. Our fully unsupervised pipeline for contamination detection using (3-VAE was trained exclusively on pure tomography data. This approach was evaluated on Beta vulgaris and Cichorium endivia, artificially contaminated with three other species, and reciprocally Cichorium endivia or Beta vulgaris. Our models reached sufficient generalisation with a data size of 800 volumes (64 & times;64 & times;64) for Beta vulgaris and 700 (64 & times;64 & times;64) for Cichorium endivia, using latent space dimensionality of 512 and 256, respectively. Across 1-50% contamination, they achieved balanced accuracy of 98.45 +/- 0.95% (Beta vulgaris) and 97.47 +/- 0.95% (Cichorium endivia), with 0% false negatives and low false alerts (3.10% - 5.06%). In conclusion, these results demonstrate that PelletRayTion provides a rapid, safe, and highly accurate alternative to conventional pelleted seed testing. This approach eliminated the need for massive annotated datasets while maintaining strong performance across different species and contaminants. It holds a significant potential for routine testing, biosecurity and regulatory workflows.
Silene takeshimensis (Caryophyllaceae) is endemic to the Korean Peninsula; however, information on its seed dormancy is scarce. Understanding the seed dormancy and germination characteristics of this endemic species is essential for elucidating its reproductive ecology and establishing effective seed propagation and conservation protocols. We investigated the germination responses of S. takeshimensis to temperature regimes (4/4–30/20°C), cold stratification (CS; 0–12 weeks at 4°C), GA 3 treatment (0–1,000 mg L -1 for 24 hours), dry after-ripening (DAR; 0–12 weeks at room temperature) and seed storage at 1°C (5 or 20 months) to determine its dormancy status. The seeds have fully developed embryos and a permeable seed coat, ruling out morphological and physical dormancy. Final germination (FG) was significantly reduced at 4/4°C (< 5%), whereas mean germination time (MGT) decreased at relatively warm alternating temperatures (20/10–30/20°C). Although CS did not increase FG, it reduced MGT to within approximately one week. DAR for four and eight weeks significantly increased FG at 15/6°C, reaching 100 and 82%, respectively, and reduced MGT at 15/6°C (approximately 23 days), indicating the presence of non-deep PD in a portion of the seed population. However, germination remained lower at 15/6°C than 25/15°C, suggesting that some seeds may have been conditionally dormant under suboptimal thermal conditions. Furthermore, compared with freshly harvested seeds from the same seed collection, seeds stored dry at 1°C for eight months exhibited higher germination percentages at 15/6°C, indicating conditional dormancy in freshly harvested seeds. Freshly harvested S. takeshimensis seeds exhibited conditional dormancy under cooler incubation conditions, which was alleviated by short-term DAR or cold dry-storage.
This work was conducted to test whether grafting enhances seed longevity in watermelon. ‘Crimson Sweet’ plants grafted onto three different interspecific rootstock hybrids (‘Nun9075’, ‘Carnivor’ and ‘Strongtosa’) and an ungrafted control were grown at two sites, Adana (site 1) and Manisa (site 2). Seeds were harvested from mature fruits (i.e., 40-45 days after anthesis) and stored hermetically at 35°C with 17.0 ± 0.2% seed moisture content over 56 days. Survival curves were constructed by taking sequential samples and testing for germination. K i (initial seed lot viability), -1/σ (the rate of viability loss) and p 50 (days to reach 50% viability) were calculated through probit analysis. The highest longevity was obtained for seeds from plants grafted on ‘Nun9075’, for which K i was 3.28 and 2.85 probits and p 50 was 46 and 40 days for seeds from sites 1 and 2, respectively. Both values were significantly different from those of seeds from plants grafted onto the other rootstocks and ungrafted control ( p < 0.01). The slope (1/σ) was -0.0712 and did not significantly differ among seed lots ( p > 0.05). Results indicate that grafting in watermelon can enhance seed longevity, particularly when using ‘Nun9075’ as the rootstock in this work.
Seed treatment (ST) with phytosanitary products is essential for crop protection, yet products may cause phytotoxic responses, particularly after storage. How these responses vary among products, after different storage periods and depending on water availability during germination remains poorly characterised. Phytotoxicity was evaluated in soybean (cv. Olimpo) seeds treated with 12 seed treatment products (five fungicides, five insecticides and two nematicides). Seeds were evaluated at 0 and 3 months of storage using germination tests, seedling emergence tests under controlled and uncontrolled conditions with and without water deficit, and measurements of root and total seedling length. A phytotoxicity index was calculated against the untreated control and an exploratory multivariate approach was used to compare response profiles among products within each test condition. The imidacloprid + thiodicarb product showed the greatest phytotoxic response, reducing normal seedlings by up to 22 percentage points after three months. Storage for three months was associated with lower physiological performance for several products, although responses varied according to the assay and water regime. Fungicide products were associated with greater reductions in emergence and root development under water deficit. The fluopyram product produced a halo-effect symptom on the cotyledons without a consistent reduction in physiological performance. The multivariate approach separated products into response groups that differed among test conditions.
Sex expression in castor exhibits considerable instability, often leading to the occurrence of interspersed staminate flowers (ISFs) on pistillate lines and presenting a significant challenge for maintaining hybrid seed purity. Although rogueing of ISFs is routinely practiced, limited quantitative information exists regarding the actual competitive ability of ISF-derived pollen in relation to pollen from monoecious male parents. This study evaluated the competitive ability of ISF-derived pollen relative to pollen from monoecious male parental lines using both in vitro assays and in vivo mixed-pollen fertilisation. In vitro results showed that ISF pollen consistently exhibited higher germination (44.5-57.3%), longer pollen tubes (78.5-109.5 μm), as well as greater pollen production per anther (634-980) than monoecious pollen. In vivo grow-out tests across two years revealed that ISF pollen was capable of fertilising ovules even under highly diluted conditions. Mixed-pollen ratios of 1:50 produced high hybridity (97.75-98.25%), whereas closer ratios (1:2 and 1:4) resulted in increased selfed plants and reduced hybridity (75.25-81.25%). These results indicate that ISF pollen is highly vigorous and even small quantities of ISF pollen can compromise genetic purity in hybrid seed production, underscoring the need for stringent rogueing of ISFs in pistillate lines.
