Image-based morphometric analysis combined with seed sorting reveals quality differences in Urochloa seedsABSTRACT:
This study evaluated the physiological potential of Marandu grass (Urochloa brizantha cv. Marandu) seeds, focusing on their size and color. Two commercial seed lots were cleaned and sorted by color (light- and dark-colored) and size (small, medium, and large). Eight replications of 100 seeds from each group were weighed to determine the thousand-seed weight. Seeds were then photographed on a blue EVA foam sheet, and digital images were analysed with the EBImage package in R to measure seed area, length, and perimeter. Tests for moisture content, germination, first count, and germination speed index were performed using a completely randomized design in a 2 × 2 × 3 factorial arrangement. Statistical analysis revealed that seed lot, seed coat color, and seed size influenced seed physiological performance. Seed size influenced germination performance, with small seeds generally showing lower germination and first count values, whereas medium-sized seeds exhibited higher germination speed. Dark-colored seeds generally showed superior physiological quality. These findings demonstrate that seed size and seed coat color influence the physiological performance of Urochloa seeds, highlighting the potential of image-based morphometric analysis as a rapid, objective, and non-destructive approach for seed characterization.
Index terms:
digital phenotyping; germination; seed classification; Urochloa brizantha cv. Marandu
RESUMO:
Este estudo avaliou o potencial fisiológico de sementes de capim-marandu (Urochloa brizantha cv. Marandu), com foco no tamanho e na coloração. Dois lotes comerciais de sementes foram beneficiados e classificados quanto à coloração (claras e escuras) e ao tamanho (pequenas, médias e grandes). Oito repetições de 100 sementes de cada grupo foram pesadas para determinação da massa de mil sementes. Em seguida, as sementes foram fotografadas sobre uma folha de EVA azul, e as imagens digitais foram analisadas com o pacote EBImage no R para mensuração da área, comprimento e perímetro das sementes. Foram realizados testes de teor de água, germinação, primeira contagem e índice de velocidade de germinação, utilizando delineamento inteiramente casualizado em esquema fatorial 2 × 2 × 3. A análise estatística revelou que o lote, a coloração do tegumento e o tamanho das sementes influenciaram seu desempenho fisiológico. O tamanho das sementes influenciou o desempenho germinativo, com sementes pequenas apresentando, em geral, menores valores de germinação e primeira contagem, enquanto sementes de tamanho médio apresentaram maior velocidade de germinação. Sementes de coloração escura apresentaram, em geral, qualidade fisiológica superior. Esses resultados demonstram que o tamanho e a coloração do tegumento influenciam o desempenho fisiológico das sementes de Urochloa, destacando o potencial da análise morfométrica baseada em imagens como abordagem rápida, objetiva e não destrutiva para a caracterização de sementes.
Termos para indexação:
fenotipagem digital; germinação; classificação de sementes; Urochloa brizantha cv. Marandu
INTRODUCTION
The genus Urochloa plays a fundamental role in agriculture, particularly in establishing pastures, as it includes resilient and adaptable species. Among them, Marandu grass (Urochloa brizantha Hochst. ex A. Rich cv. Marandu) is considered the most representative in Brazil’s pasture seed production fields (Landau et al., 2020), primarily due to their adaptability, high forage yield, persistence, strong regrowth capacity, tolerance to adverse conditions, and high production of viable seeds (Abrantes et al., 2025).
Although these species produce a significant quantity of seeds, they do not always exhibit the desired quality and vigor. Several factors can make it difficult to obtain commercial seed lots with high physiological potential, including irregular maturation and natural seed shattering in production fields (Silva et al., 2022). These problems can result in seeds with significant variations in size and color.
Seed size is considered an important physical characteristic because it may reflect reserve accumulation during seed development, which can contribute to seedling establishment under favorable conditions. However, the relationship between seed size and physiological quality is species-dependent and may also be influenced by environmental conditions during seed development (Adji et al., 2022; São José et al., 2024). Seed size is an important physical attribute in seed production because it directly influences seed processing, commercial classification, sowing uniformity, and, potentially, seedling establishment (Melo et al., 2016; Mazon et al., 2025; Nomada et al., 2026). Furthermore, advances in image-based analysis have increased the importance of objective morphometric traits as non-destructive indicators of seed quality. Although the relationship between seed size and physiological performance has been investigated in several cultivated species, information for tropical forage grasses, particularly Urochloa spp., remains limited.
Digital image analysis has expanded the possibilities for rapid and non-destructive seed quality assessment. Morphometric descriptors extracted from digital images, such as area, length, perimeter, and shape-related attributes, provide objective information that can support seed classification, processing, and quality control while reducing the subjectivity associated with visual evaluations. This approach has also been successfully applied to Urochloa seeds, demonstrating the usefulness of image-derived morphometric descriptors for genotype discrimination and seed characterization (Freitas et al., 2021). Image-based digital morphometric analysis represents promising tool to complement conventional physiological tests and to support the development of automated seed classification systems (Freitas et al., 2021; Morebise et al., 2026).
Although differences in seed size can be compensated for over time under favorable environmental conditions, this initial variation in vigor can be critical for Urochloa species. These grasses are widely used to establish pastures in regions subject to adverse environmental conditions. In these challenging environments, vigorous seeds are essential to withstand such stresses and ensure successful crop establishment (Silva et al., 2019).
Seed coat color has been associated with physiological maturity, tegument characteristics, and deterioration processes in several species (Silva et al., 2025; Mendes et al., 2025; Sujatha et al., 2025). However, its relationship with seed physiological quality is not universal and may vary according to species and seed lot characteristics (Sujatha et al., 2025). Seed coat color has been recognized as an important visual attribute associated with seed physiological quality in some species (Silva et al., 2025; Sujatha et al., 2025). Furthermore, advances in digital image analysis have shown that color and morphometric descriptors extracted from images can be used as non-destructive predictors of germination success, highlighting the potential of image-based phenotyping for seed quality assessment (Morebise et al., 2026; Novikova et al., 2026).
Because seed size and seed coat color are readily detectable by digital imaging systems, understanding their relationship with seed physiological performance is important not only for seed quality evaluation but also for the development of automated seed sorting technologies. Therefore, this study evaluated the influence of seed size and seed coat color on the morphometric characteristics, germination, and early seedling development of Marandu grass seeds using image-based morphometric analysis.
MATERIAL AND METHODS
The study was conducted in the municipality of Ilhéus, Bahia, Brazil. Two seed lots of Marandu grass (Urochloa brizantha) acquired from specialized companies, were used. The package labels stated a germination potential of 80% and a minimum purity of 60%.
Initially, a blower was used to separate pure seeds from inert material (dead seeds, soil particles, and plant debris). Subsequently, the seeds were visually classified into two seed coat color categories (light- and dark-colored) by the same evaluator under uniform laboratory lighting conditions. Within each color class, the seeds were manually sorted into three size categories (small, medium, and large) using tweezers (Figure 1). The color classification was based exclusively on the predominant external appearance of the seed coat and was adopted only as a practical criterion for grouping seeds, without instrumental color measurements or validation as an indicator of physiological maturity. Because no instrumental color measurements were performed, the adopted color classes represent visual categories rather than objective physiological classes. This resulted in six treatments: T1, small light-colored seeds; T2, medium light-colored seeds; T3, large light-colored seeds; T4, small dark-colored seeds; T5, medium dark-colored seeds; and T6, large dark-colored seeds.
Workflow for cleaning and manual classification of Urochloa brizantha cv. Marandu seeds according to seed coat color (light and dark)
Then, for each treatment within each lot, eight 100-seed replications were weighed on an analytical balance to determine the thousand-seed weight (TSW) (Brasil, 2025). After weighing, the seeds were arranged on a blue EVA foam sheet and photographed with a smartphone mounted on a tripod at a fixed height. All images were acquired under standardized conditions, maintaining the same lighting environment, camera position, and background throughout the experiment to minimize variations in image acquisition. A reference object with known dimensions was included in each image to calibrate the spatial scale, allowing the conversion of pixel-based measurements into millimeters. The images were processed using the EBImage package (Pau et al., 2010) implemented in R software (R Core Team, 2026). Image processing included channel selection, Otsu thresholding for seed segmentation, application of the fillHull function to fill internal gaps in the segmented objects, removal of image noise based on object area, and automatic extraction of seed morphometric traits, including area, length, and perimeter. Four hundred individual seeds from each treatment were analyzed for morphometric characterization (Figure 2).
Workflow of digital image acquisition and morphometric analysis for automatic extraction of seed area, length, and perimeter measurements.
After image acquisition, seed physiological quality was assessed through germination and seedling development measurements. The experiment was conducted using a completely randomized design in a 2 × 2 × 3 factorial scheme, corresponding to two seed lots, two seed coat colors, and three seed size classes. The seed physiological quality analyses are described below.
Seed moisture content was determined using the oven method (Brasil, 2025). Two replications of 1.5 mL of seeds were placed in previously identified aluminum capsules and dried in an oven at 105 °C for 24 hours. The results were expressed as a percentage on a wet basis. To carry out the germination test, four replications of 50 seeds per treatment were sown in plastic boxes containing sheets of moistened filter paper (2.5 mL.g-1 of dry paper) and placed in a germination chamber with an alternating temperature of 20-35 °C and a 16-hour photoperiod. The germinated seeds were counted after 21 days, considering only normal seedlings with essential structures (Brasil, 2025).
Concurrently with the germination test, the first germination count (FGC) and the germination speed index (GSI) were determined. For the FGC, normal seedlings were counted on the seventh day after sowing, with the data expressed as a percentage (Brasil, 2025). The GSI was determined from daily counts of germinated seeds (Maguire, 1962). Mean germination time (MGT) was calculated from the daily germination counts (Labouriau, 1983), expressing the average number of days required for seed germination.
At the end of the germination test, normal seedlings were used to determine seedling development. Shoot length (SPL) and root length (RL) were measured with a graduated ruler, and total seedling length (TL) was obtained by summing shoot and root lengths. Ten normal seedlings per replication were evaluated. For seedling dry mass (DM) determination, the evaluated seedlings were dried in a forced-air oven at 65 °C until constant weight and subsequently weighed on an analytical balance. Results were expressed as milligrams per seedling (mg.seedling⁻¹).
The data were tested for normal distribution of errors using the Shapiro-Wilk test and for homogeneity of variances using Bartlett’s test. Subsequently, the data were subjected to an analysis of variance (ANOVA). When the analysis of variance indicated significant effects (p ≤ 0.05), differences between factors with two levels were assessed directly by the F-test, whereas means of factors with three levels were compared using Tukey’s test (p ≤ 0.05). For significant interactions, simple effects were analyzed following the same criterion. Statistical analyses were performed using the ExpDes.pt package in R software version 4.6.1 (R Core Team, 2026).
RESULTS
When comparing the seed water content between light- and dark-colored seeds (Table 1), in L1, no effect of color on water content was observed in the various size classes, remaining within the range of 11.1 to 12.0%. In L2, however, the large dark-colored seeds presented a higher water content, reaching 12.7%.
Regarding the influence of seed color on seed dimensions, light-colored seeds in L1 exhibited greater area, length, and perimeter values across all size classes (Table 1). In L2, light-colored seeds in the large size class showed a significantly greater area than dark-colored seeds. They also exhibited greater perimeter values in the small and large size classes, whereas no significant effect of color was observed for seed length. In L1, there was no significant interaction between the factors of color and seed size in relation to the germination performance of the seeds (Table 2). Seed size did not result in a significant difference in germination, first germination count, and average germination time of the seeds. Despite this, the smaller seeds showed a faster germination rate, evidenced by a higher germination speed index (7.96), followed by the medium-sized seeds (6.56), which exhibited an intermediate speed (Table 2).
Seed coat color significantly affected the physiological quality of the seeds in L1 (Table 2). Dark-colored seeds showed higher germination (46%), corresponding to a difference of 19.87 percentage points compared with light-colored seeds. In addition, these seeds presented a higher GSI of 8.6, along with a shorter MGT, of only 2.9 days.
In contrast to L1, in L2 there was a significant interaction between the factors of color and seed size in relation to germination performance (Table 3). Within the light-colored seed group, seed size did not significantly affect germination, first germination count, or GSI. However, the large light-colored seeds required a longer period for the germination process, with an MGT of 6.38 days.
Within the dark-colored seed group (Table 3), germination differed significantly among seed size classes. Small seeds exhibited lower germination, with a 37% reduction compared to medium sized seeds and 29% compared to large seeds. In addition, small dark-colored seeds had lower GSI (7.12) and lower FGC (37%).
Similarly to L1, seed coat color influenced the physiological quality of seeds in L2 (Table 3). Dark-colored seeds showed superior germination performance. Significant differences in germination between seed colors were observed for the medium- and large-sized seeds, whereas no significant difference was detected for the small-sized seeds. In addition, dark-colored seeds exhibited higher FGC, higher GSI, and lower MGT across all seed size classes.
In the analysis of initial seedling development as a function of seed size and color (Table 4), in L1, there was a significant interaction of the analyzed factors only for root length (RL) and overall length (TL).
Neither the size nor the color of the seeds influenced the shoot development of the seedlings in L1 (Table 4). Light-colored, medium-sized seeds produced seedlings with lower root length. In seedlings originating from dark-colored seeds, the difference in seed size did not affect root length. However, when comparing seed coat colors within each seed size class, seedlings originating from large, dark-colored seeds showed greater root length.
The different seed size classes did not influence the overall length of seedlings, whether originating from light- or dark-colored seeds (Table 4). In the analysis of the influence of color within each seed size, seedlings generated by large and dark-colored seeds showed greater total seedling length.
Seed coat color did not affect seedling dry matter (Table 4). Seed size significantly affected seedling dry matter accumulation, with seedlings originating from small seeds accumulating approximately 17% less dry matter than those from the other size classes.
There was no significant interaction between the factors analyzed in the initial development of seedlings from L2 (Table 5). Root length was not affected by seed size. However, seedlings originating from small seeds showed lower SPL, TL, and DM than those originating from medium and large seeds.
Among the seedling development parameters evaluated (Table 5), seed coat color affected only SPL and TL, as seedlings originating from dark-colored seeds exhibited greater shoot and total seedling length.
DISCUSSION
For the proper conservation of Urochloa seeds, the water content of the seeds should be between 11% and 12% (wet basis) (Abrantes et al., 2025). From this, it can be stated that the water content presented by the seeds of the evaluated lots fell within the ideal range. Being within the ideal range is important because high moisture levels can accelerate seed deterioration and increase metabolic activity, due to the intensification of seed respiration (Barbosa et al., 2021).
Although the dark-colored, large seeds of L2 presented a higher percentage of water content, this difference is not very discrepant. Therefore, the color difference did not result in a difference in water content in the seeds of the evaluated lots. Furthermore, the water content values of the seeds were very close; this proximity of values is important so that the test results are not affected by the difference in metabolic activity due to the respiration rate (Melo et al., 2016).
Beyond the biological interpretation of seed size and color, the morphometric analysis demonstrated the usefulness of digital image analysis for objectively characterizing physical variability within commercial seed lots. The image-based measurements allowed subtle differences in seed dimensions to be quantified without destructive procedures, reinforcing the potential of computer vision tools to support seed classification and quality control in forage seed production (Freitas et al., 2021).
Based on the morphometric results, no consistent association between seed coat color and seed dimensions was observed across the two seed lots. In L1, light-colored seeds exhibited greater seed measurements, whereas in L2 the effect of color was restricted to specific variables and size classes. Seed coat color and seed dimensions may vary among seed lots due to differences associated with seed development, tegument characteristics, environmental conditions during seed production, or deterioration processes (Silva et al., 2025). Therefore, seed coat color alone cannot be considered a consistent indicator of seed size.
The classification of seeds into S, M, and L affected the physiological performance of the seeds in the lots analyzed. In both lots, the difference in size affected seed vigor, specifically the germination rate. In L1, the small seeds showed a better germination speed index, and in the second lot, the large, light-colored seeds needed more time to germinate.
Typically, small seeds germinate faster because they are more permeable and absorb water more rapidly than larger seeds (São José et al., 2024). In contrast, larger seeds generally provide greater reserves for seedling establishment and are frequently associated with superior germination and early seedling growth, although these responses depend on the species and environmental conditions during seed development (Adji et al., 2022; Mazon et al., 2025; Nomada et al., 2026).
This observation is consistent with the results of this study and with findings reported for soybean seeds, in which smaller seeds tend to exhibit a higher germination speed when evaluating the relationship between seed size and primary root growth rate (Mazon et al., 2025).
The differences observed in germination speed may be associated with the physical characteristics of the seeds, particularly their size, which can influence the initial water uptake and reserve mobilization during germination. Although seed size did not consistently affect the final germination percentage, it influenced the germination rate, indicating that seed dimensions may alter the dynamics of early germination events without necessarily affecting seed viability (Mazon et al., 2025; Nomada et al., 2026).
In addition to affecting germination speed, seed size may also influence germination performance by altering the amount of reserves available to support the early stages of seedling establishment (Adji et al., 2022; Mazon et al., 2025). This is consistent with the present findings, in which small, dark-colored seeds exhibited lower germination, suggesting that reduced seed size may negatively affect physiological performance under specific seed-lot conditions. The influence of seed size on germination and early seedling development has been consistently reported across different species, although the magnitude of this effect depends on species-specific characteristics and seed-lot conditions (Steiner et al., 2019; Adji et al., 2022; Mazon et al., 2025; Nomada et al., 2026).
The difference in coloration between seeds from the same lot influenced the physiological quality of the evaluated lots. In the present study, dark-colored seeds generally exhibited superior physiological performance. However, because seed coat color was established by visual classification only, the observed differences cannot be unequivocally attributed to differences in physiological maturity. Seed coat color may reflect variations associated with seed development, tegument characteristics, or deterioration processes, and these factors were not independently evaluated in the present study.
However, contrasting with the results obtained in the present study, previous studies have reported different associations between seed coat color and physiological quality. Light-colored seeds have been associated with superior physiological quality in some species (Sujatha et al., 2025), whereas dark-colored seeds have been associated with greater germination and vigor in others (Mendes et al., 2025). These contrasting findings reinforce that the relationship between seed coat color and physiological quality is not universal but rather depends on species-specific characteristics and seed lot conditions.
When analyzing, in general, the reflection of the difference in seed color and size on the initial development of seedlings, it is not possible to state which specific part of the seedling will or will not influence which factor. However, from the analysis of the results of the two lots, the present results indicate that size and color can influence the initial development of Marandu grass seedlings.
In the two lots analyzed, seedlings generated from small seeds had less dry matter mass, which may be a consequence of stored reserves, which are generally greater in larger seeds. In bean seedlings, the smaller seeds also showed the lowest values of DM accumulation when considering the dry mass of the seedling produced (Siega et al., 2023).
The total of stored reserves is one of the factors that influence the physiological quality of seeds, as the seedling uses it until it becomes capable of performing photosynthesis (Huo et al., 2025). This observation is consistent with previous studies showing that seed size influences early seedling performance and physiological quality, although the magnitude of this effect depends on the genotype (Mazon et al., 2025).
In addition to DM accumulation, seed size also influenced seedling length. In L1, medium-sized light-colored seeds produced seedlings with shorter root length, whereas in L2, seedlings from small seeds showed reduced development compared to the others. The influence of seed size on seedling growth has been widely reported, with larger seeds generally associated with greater seedling length and biomass accumulation, although this response may vary depending on the species and seed lot conditions (Morebise et al., 2026; Nomada et al., 2026).
Color influenced seedling dry mass only in seedlings from L1, where seedlings from light-colored seeds had greater accumulation. Despite this, dark-colored seeds resulted in larger seedlings. Dark-colored seeds also exhibited superior germination performance and produced larger seedlings. However, the present results refer to early seedling development under controlled laboratory conditions and should not be interpreted as evidence of superior field performance. (Siega et al., 2023).
Urochloa spp. is characterized by uneven flowering, which contributes to heterogeneous seed maturation and consequently to variation in the physical appearance of seeds. As a result, commercial seed lots frequently contain seeds with different morphometric characteristics and physiological performance. Therefore, physical attributes such as seed size and seed coat color may represent useful criteria for improving seed sorting and obtaining more uniform seed lots. In addition, the objective characterization of these attributes through digital image analysis provides valuable information for developing automated seed classification systems. The morphometric descriptors obtained by digital image analysis demonstrate the potential of computer vision as a rapid, non-destructive, and reproducible approach to support seed quality assessment and processing. Nevertheless, because seed coat color was classified visually, its biological interpretation should be made cautiously, and future studies combining digital color measurements with physiological and biochemical analyses may further improve automated classification models.
CONCLUSIONS
Seed size and seed coat color affect the physiological quality of Marandu grass seeds. Seed size influences germination performance and early seedling development, with smaller seeds generally exhibiting faster germination, whereas larger seeds are associated with greater initial seedling development.
Seed coat color also influences seed physiological quality, with dark-colored seeds generally exhibiting superior germination performance and early seedling development. Furthermore, image-based morphometric analysis proved to be a rapid, objective, and non-destructive approach for characterizing physical differences among seed lots, supporting its potential application in seed quality assessment.
ACKNOWLEDGMENTS
We would like to thank CAPES (Council for the Improvement of Higher Education Personnel - funding code 001), FAPESB (Bahia State Research Support Foundation) and CNPq (National Council for Scientific and Technological Development) for their support of this study.
REFERENCES
-
ABRANTES, F.L.; MACHADO-NETO, N.B.; CUSTODIO, C.C. Seed storage and germination of three grass species: effect of spikelet weight and dormancy. Brazilian Archives of Biology and Technology, v.68, e25241009, 2025. https://doi.org/10.1590/1678-4324-2025241009
» https://doi.org/https://doi.org/10.1590/1678-4324-2025241009 -
ADJI, B.I.; AKAFFOU, D.S.; REFFYE, P.; SABATIER S. Maternal environment and seed size are important for successful germination and seedling establishment of Pterocarpus erinaceus (Fabaceae). Journal of Forestry Research, v.33, p.977-990, 2022. https://doi.org/10.1007/s11676-021-01412-x
» https://doi.org/https://doi.org/10.1007/s11676-021-01412-x -
BARBOSA, R.M.; JESUS, M.A.; PEREIRA, R.A.; GOMES-JUNIOR, G.A. Electrical conductivity and ethanol release to assess red rice seed vigor. Caatinga, v.34, n.4, p.791-798, 2021. http://dx.doi.org/10.1590/1983-21252021v34n406rc
» https://doi.org/http://dx.doi.org/10.1590/1983-21252021v34n406rc -
BRASIL. Regras para Análise de Sementes (RAS) Ministério da Agricultura e Pecuária - MAPA, Brasília, 2025. https://wikisda.agricultura.gov.br/pt-br/Laborat%C3%B3rios/Metodologia/Sementes/RAS_2025/Amostragem
» https://wikisda.agricultura.gov.br/pt-br/Laborat%C3%B3rios/Metodologia/Sementes/RAS_2025/Amostragem -
FREITAS, M.N.; DIAS, M.A.N.; GOMES-JUNIOR, F.G.; ABUD, H.F.; ARAÚJO, L.B.; MORAES, T.F. Discrimination of Urochloa seed genotypes through image analysis: morphological features. Agronomy Journal, v.113, n.6, p.4930-4944, 2021. https://doi.org/10.1002/agj2.20839
» https://doi.org/https://doi.org/10.1002/agj2.20839 -
HUO, J.; ZHENG, D.; FENG, N.; ZHANG, R.; XUE, Y.; KHAN, A.; ZHOU, H.; SHEN, X.; ZHAO, L.; LI, Y.; LIU, X.; LIU, J.; LI, J.; MEI, W.; DU, X.; FENG, Y. Transcriptome and metabolome analysis of senescent rice (Oryza sativa L.) seeds: insights into the mechanism of germination vigor and seedling morphogenesis. BMC Plant Biology, v.25, n.591, p.1-16, 2025. https://doi.org/10.1186/s12870-025-06458-1
» https://doi.org/https://doi.org/10.1186/s12870-025-06458-1 - LABOURIAU, L.G. 1983. A germinação das sementes Washington, DC: Organization of American States. 174 p.
- LANDAU, E.C.; RESENDE, R.M.S.; MATOS-NETO, F.C. Evolução da área ocupada por pastagens In: LANDAU, E.C.; SILVA, G. A.; MOURA, L.; HIRSCH, A.; GUIMARAES, D.P. (Eds.) Dinâmica da produção agropecuária e da paisagem natural no Brasil nas últimas décadas: produtos de origem vegetal. Brasília: Embrapa, 2020.
-
MAGUIRE, J.D. Speed of germination aid in selection and evaluation for seedling emergence and vigor. Crop Science, v.2, n.2, p.176-177, 1962. https://doi.org/10.2135%2Fcropsci1962.0011183X000200020033x
» https://doi.org/https://doi.org/10.2135%2Fcropsci1962.0011183X000200020033x -
MAZON, A.S.; KONZEN, L.H.; RODRIGUES, D.B.; MENEGHELLO, G.E.; TUNES, L.V.M. Seed size and its effects on the physiological quality of produced seeds. Revista Ciência Agronômica, v.56, e202392564, 2025. https://doi.org/10.5935/1806-6690.20250013
» https://doi.org/https://doi.org/10.5935/1806-6690.20250013 -
MELO, L.F.; MARTINS, C.C.; SILVA, G.Z.; BONETI, J.E.B.; VIEIRA, R.D. Beneficiamento na qualidade física e fisiológica de sementes de capim-mombaça. Revista Ciência Agronômica , v.47, n.4, p.667-674, 2016. https://doi.org/10.5935/1806-6690.20160080
» https://doi.org/https://doi.org/10.5935/1806-6690.20160080 -
MENDES, K.G.; SILVA, L.J.; PICOLI, E.A.T.; AGUIAR, V.P.; ALMEIDA, L.S.; LIRA, J.M.S. Evaluation of Cedrela fissilis Vell. seeds with color heteromorphism using near-infrared spectroscopy and their relationship with physiological quality. Journal of Seed Science, v.47, e202547216, 2025. https://doi.org/10.1590/2317-1545v47297619
» https://doi.org/https://doi.org/10.1590/2317-1545v47297619 -
MOREBISE, A.S.; NANA, H.R.T.; AZEVEDO, S.R.; CRUZ, C.M.; BARBOSA, R.M. Digital morphometric phenotyping of Genipa americana seeds: post-germinative root-development classification and seed-lot morphometric differentiation. Forests, v.17, n.7, p.786, 2026. https://doi.org/10.3390/f17070786
» https://doi.org/https://doi.org/10.3390/f17070786 -
NOMADA, R.T.; PEREIRA, L.K.; BRICEÑO-PINZÓN, I.D.; PEREIRA, W.V.S.; CARVALHO, E.R.; PIRES, R.M.O. Size as a fast and non-destructive marker for predicting wheat seed physiological quality. Journal of Seed Science , v.48, n.1, e202648003, 2026. http://dx.doi.org/10.1590/2317-1545v48300679
» https://doi.org/http://dx.doi.org/10.1590/2317-1545v48300679 -
NOVIKOVA, T.P.; TYLEK, P.; PETRISHCHEV, E.P.; NOVIKOV, A.I. Higher plant seed container germination success predicted by smart farming optical RGB approach. Scientific Reports, v.16, p.13021, 2026. https://doi.org/10.1038/s41598-026-42258-9
» https://doi.org/https://doi.org/10.1038/s41598-026-42258-9 -
PAU, G.; FUCHS, F.; SKLYAR, O.; BOUTROS, M.; HUBER, W. EBImage-an R package for image processing with applications to cellular phenotypes. Bioinformatics, v.26, n.7, p.979-981, 2010. https://doi.org/10.1093/bioinformatics/btq046
» https://doi.org/https://doi.org/10.1093/bioinformatics/btq046 -
R CORE TEAM. R: A language and environment for statistical computing R Foundation for Statistical Computing, Vienna, Austria, 2026. https://www.R-project.org/
» https://www.R-project.org/ -
SÃO JOSÉ, J.F.B.; HERNANDES, M.A.S.; LISBOA, B.B.; VOLPIANO, C.G.; SCHLINDWEIN, G.; TRINDADE, J.K.; LATTUADA, D.S.; BENEDUZI, A.; VARGAS, L.K. Seed size and Azospirillum brasilense Ab-V5 and Ab-V6 inoculation influences germination and early seedling vigor of Acacia mearnsii Ciência Florestal, v. 34, n. 4, e85546, 2024. https://doi.org/10.5902/1980509885546
» https://doi.org/https://doi.org/10.5902/1980509885546 -
SIEGA, Y.P.; PADILHA, M.S.; COELHO, C.M.M.; EHRHARDT-BROCARDO, N.C.M. O tamanho de sementes de feijão afeta a utilização das reservas armazenadas durante a germinação. Revista de Ciências Agroveterinárias, v.22, n.3, p.529-537, 2023. https://doi.org/10.5965/223811712232023529
» https://doi.org/https://doi.org/10.5965/223811712232023529 -
SILVA, A.Q.; OLIVEIRA, L.A.; SILVA, C.P.; MENDES, C.T.E.; FERREIRA, A.M.O.; SÁFADI, T.; CARVALHO, M.L.M. Seed quality of Brachiaria brizantha by X-ray image analysis using a Bayesian approach. Acta Scientiarum. Agronomy, v.44, e55960, 2022. https://doi.org/10.4025/actasciagron.v44i1.55960
» https://doi.org/https://doi.org/10.4025/actasciagron.v44i1.55960 -
SILVA, G.Z.; MARTINS, C.C.; JEROMINI, T.S.; PEREIRA, F.E.C.B.; BRUNO, R.L.A. Production regions, physical and physiological quality of Brachiaria brizantha cv. BRS ‘Piatã’ seeds. Revista Ciência Agronômica , v.50, n.4, p.552-561, 2019. https://doi.org/10.5935/1806-6690.20190065
» https://doi.org/https://doi.org/10.5935/1806-6690.20190065 -
SILVA, T.R.B.; MONTEIRO, M.E.S.; ABATI, J.; SILVA, M.F.; NOLLA, A.; ALVEZ, C.Z.; SECCO, D.; MARINS, A.C. Physiological quality of chia seeds as a function of coat color and fungicide treatment. Revista Ciência Agronômica , v. 56, n.1, e202392518, 2025. https://doi.org/10.5935/1806-6690.20250021
» https://doi.org/https://doi.org/10.5935/1806-6690.20250021 -
STEINER, F.; ZUFFO, A.M.; BUSCH, A.; SOUSA, T.O.; ZOZ, T. Does seed size affect the germination rate and seedling growth of peanut under salinity and water stress? Pesquisa Agropecuária Tropical, v.49, p. e54353, 2019. https://doi.org/10.1590/1983-40632019v4954353
» https://doi.org/https://doi.org/10.1590/1983-40632019v4954353 -
SUJATHA, K.; SINDHU, S.; KEERTHANA, D.; ARUN, K.P.; ELAMPARITHI, R. Seed coat colour variation and its impact on seed quality parameters of horsegram var. Paiyur 2. Agricultural Science Digest, v.45, n.3, p.457-459, 2025. https://doi.org/10.18805/ag.D-5451
» https://doi.org/https://doi.org/10.18805/ag.D-5451
