Early potassium fertilization enhances physiological and health quality of soybean seeds

A adubação antecipada com potássio melhora a qualidade fisiológica e sanitária das sementes de soja

ABSTRACT

Potassium fertilization plays a crucial role in optimizing soybean yield by enhancing seed vigor and reducing the incidence of pathogens. This study aimed to evaluate the physiological and health qualities of soybean seeds from cultivars grown under varying potassium chloride application rates and timing. We used a randomized block design in a split-split-plot arrangement, with three replications. Two soybean cultivars (Brasmax Olimpo IPRO and CZ 58B23 I2X) were planted in the main plots. In the subplots, three potassium application timings were tested: 100% early (30 days before sowing), 50% at sowing, and 50% at 30 days after emergence (DAE), and 100% at 30 DAE. The sub-subplots included five potassium rates (0, 50, 100, 150, and 200 kg K₂O ha^-1). Seed quality parameters assessed included thousand-seed mass, water content, seedling emergence, electrical conductivity, first germination count, germination percentage, shoot and root dry matter, accelerated aging, and pathogen incidence. The results showed that fertilization of Brasmax Olimpo IPRO with 200 kg K₂O ha^-1, either 100% early or 100% at 30 DAE, resulted in superior seed physiological and health quality.

Index terms:
Glycine max; seed vigor; pathogen incidence

RESUMO

A adubação potássica desempenha papel fundamental na otimização da produtividade da soja, ao promover maior vigor das sementes e reduzir a incidência de patógenos. Este estudo teve como objetivo avaliar a qualidade fisiológica e sanitária de sementes de cultivares soja cultivadas sob diferentes doses e épocas de aplicação de cloreto de potássio. O delineamento experimental utilizado foi em blocos casualizados, em esquema de parcelas subsubdivididas, com três repetições. Duas cultivares de soja (Brasmax Olimpo IPRO e CZ 58B23 I2X) foram alocadas nas parcelas principais. Nas subparcelas, foram avaliadas três épocas de aplicação de potássio: 100% antecipado (30 dias antes da semeadura), 50% na semeadura e 50% aos 30 dias após a emergência (DAE), e 100% aos 30 DAE. As subsubparcelas foram constituídas por cinco doses de potássio (0, 50, 100, 150 e 200 kg de K₂O ha^-¹). Os parâmetros de qualidade de sementes avaliados incluíram massa de mil sementes, teor de água, emergência de plântulas, condutividade elétrica, primeira contagem de germinação, porcentagem de germinação, massa seca da parte aérea e das raízes, envelhecimento acelerado e incidência de patógenos. Os resultados indicaram que a adubação da cultivar Brasmax Olimpo IPRO com 200 kg de K₂O ha^-¹, aplicada de forma 100% antecipada ou 100% aos 30 DAE, proporcionou qualidade fisiológica e sanitária superior das sementes.

Termos para indexação:
Glycine max; vigor de sementes; incidência de patógenos

Introduction

Soybean (Glycine max (L.) Merrill) is one of the world’s most significant crops, valued for its high nutritional and economic importance. Rich in protein and vegetable oil, soybean protein content ranges from 30% to 53%, making it an essential component in both human and animal diets (Silva et al., 2022). Brazil is the leading producer, with an estimated output of 166.04 million Mg in the 2024/2025 crop, yielding an average productivity of 3,499 kg ha^-1 (Companhia Nacional de Abastecimento - CONAB, 2025). To maintain high productivity and quality, effective nutritional management is essential, especially in terms of potassium (K) supply. K plays a crucial role in regulating photosynthesis, assimilate translocation, water-use efficiency, and stress tolerance (Cao et al., 2025; Zi-Qiang et al., 2025), in addition to the importance of using high-quality seeds with robust physiological and health attributes (Sousa et al., 2024).

Seed quality is a key determinant for crop establishment, directly influencing germination, seedling vigor, and subsequent growth. Genetic, physical, and health factors all contributed to seed performance, including germination and seedling vigor (Ruppin et al., 2019). In addition, proper fertilization can enhance both seed vigor and health. K fertilization, in particular, can enhance the physiological quality of plants (Mamun et al., 2022). K plays a direct role in seed development, primarily by supporting sugar formation and boosting plant vigor, which leads to higher yields and improved physiological and health qualities of seeds (Cao et al., 2025).

The application of K at balanced rates is essential for optimizing agronomic performance and seed health (Lara et al., 2018). Both excessive and insufficient K levels can disrupt key physiological processes such as photosynthesis and assimilate transport (Liu et al., 2024; Zi-Qiang et al., 2025), thereby diminishing seed quality. While K deficiency impairs cell membrane integrity and reduces seed vigor, an adequate K supply enhances cellular integrity and physiological performance. However, in fertile soils, high K rates may result in limited or inconsistent responses (Lara et al., 2018). Moreover, potassium fertilization has been shown to affect the sanitary quality of soybean seeds. K deficiency is associated with an increased incidence of seed-borne pathogens, whereas adequate K supply enhances plant health, reduces disease susceptibility, and promotes the production of healthier seeds (Zambiazzi et al., 2017).

The timing of K application is another key factor influencing agronomic outcomes. Early or well-timed applications enhance K utilization, reducing nutrient losses through leaching or soil fixation (Lago et al., 2023). Applying K in sync with critical growth stages ensures optimal nutrient availability during periods of high physiological demand (Dong et al., 2023). Anticipating K fertilization 30 days before sowing improves nutrient availability, supports seedling vigor, and minimizes stand variability caused by mechanical operations.

This study aims to evaluate the physiological and health qualities of soybean seeds from different cultivars grown under varying potassium chloride application rates and timings.

Material and Methods

Experimental site

The experiment was conducted at the ‘Accert Pesquisa e Consultoria Agronômica’ experimental station, located on the Pequizeiro Farm in Balsas, Maranhão, Brazil (latitude: 07°31’57’’ S, longitude: 46°02’08’’ W, altitude: approximately 283 m). The climatic data recorded during the experiment, along with historical climate data from 1991 to 2020, are presented in Figure 1.

Figure 1:
Monthly averages of rainfall and temperature during the 2024/2025 crop year, compared with historical averages (1991-2020) for Balsas, MA. Source: INMET (2025).

The soil at the experimental site was classified as yellow oxide with a sandy texture, according to the Brazilian soil classification system (Santos et al., 2018). Soil samples were collected from two layers: 0-20 cm and 20-40 cm, and their chemical and physical properties are summarized in Table 1.

Table
1: Main chemical properties of the soils used in the experiment.

Depth	pH	OM	TOC	H+Al	Al3+	Ca2+	Mg2+	K +	CEC	BS
Cm	H2O	---- dag/kg ---		--------------------- cmolc dm-3 -----------------------						%
0-20	5.92	1.57	0.91	1.58	0.01	3.39	1.04	0.15	6.16	74.34
20-40	5.03	1.07	0.62	1.76	0.37	1.81	0.49	0.29	4.35	59.58
	PMehlich -1	B	Cu	Fe	Mn	Zn	S	Clay	Silt	Sand
	-----------------------------mg dm-3-----------------------------							------------%------------
0-20	45.71	0.16	0.25	6.92	38.16	0.01	10.29	24.24	9.26	66.49
20-40	30.24	0.18	0.24	20.52	11.51	0.37	19.32

• Abbreviations used: pH - Soil pH (Potential of Hydrogen), OM - Organic Matter, TOC - Total Organic Carbon, H+Al - Potential Acidity (Hydrogen + Aluminum), Al³+ - Exchangeable Aluminum, Ca²+ - Exchangeable Calcium, Mg²+ - Exchangeable Magnesium, K+ - Exchangeable Potassium, CEC - Cation Exchange Capacity, BS - Base Saturation, P (Mehlich-1) - Phosphorus, B - Boron, Cu - Copper, Fe - Iron, Mn - Manganese, Zn - Zinc, S - Sulfur.

Experimental design

A randomized block design in a split-split-plot arrangement with three replications was employed. In the main plots, two soybean cultivars were sown: Cultivar 1 (Brasmax Olimpo IPRO) and Cultivar 2 (CZ 58B23 I2X). In the subplots, three potassium (K) application timings were tested: (1) 100% K applied early, 30 days before sowing; (2) 50% K applied at sowing and 50% 30 days after emergence (DAE); and (3) 100% K applied 30 DAE. Additionally, five K rates were applied in the sub-subplots: 0, 50, 100, 150, and 200 kg K₂O ha^-1. Each experimental unit consisted of four rows, spaced 0.50 m apart and 3.0 m long, covering a total area of 6.0 m². The two central rows were considered the useful area, with 1 m discarded from each end, resulting in an effective area of 1.5 m².

Soybeans were sown on December 7, 2024, using a seeder-fertilizer with a rod-type furrower mechanism, at a depth of approximately 3 cm. The rows were spaced 0.50 m apart, with 15 seeds per meter, aiming for a final population of approximately 300,000 plants per hectare. Weeds, pests, and diseases were controlled throughout the crop cycle using chemical pesticides, as needed.

At full maturity (R₈ growth stage), soybean plants were manually harvested and threshed. Seed water content was determined immediately after harvest, following the methodology described by Brasil (2025). Subsequently, seed samples were standardized to a moisture content of 13% by slow drying under shaded conditions. After drying, the seed lots were uniformly mixed, thousand-seed mass was determined (Brasil, 2025), and the standardized seed samples were used for physiological and sanitary quality analyses. Several seed quality parameters were evaluated:

Germination: Seeds were placed on three sheets of germination paper moistened with distilled water at a volume equivalent to 2.5 times the dry paper mass. The paper was rolled, and the rolls were incubated in a BOD-type germinator at 25 °C. Germination was evaluated on the 5th and 8th days after sowing, according to the criteria established in Brasil (2025).

Shoot and Root Dry Matter: Ten randomly selected seedlings per treatment were used. Seedlings were separated into shoots and roots and dried in an oven at 60 °C for 72 h. Dry matter was expressed in mg per seedling.

Accelerated Aging: Seeds were distributed on stainless steel mesh suspended inside plastic boxes containing 40 mL of water (Marcos Filho, 2020) and incubated in a BOD chamber at 41 °C for 96 h. After the aging period, seeds were subjected to the germination test, with normal seedlings counted on the 5^th day after sowing, as recommended by Brasil (2025).

Seedling Emergence: Seeds were sown in plastic trays containing Carolina Soil^® substrate, using four replicates of 50 seeds per treatment. The trays were maintained in a greenhouse with temperatures ranging from 25 to 30 °C. Seedling emergence (E) was evaluated 12 days after sowing.

Electrical Conductivity: Fifty seeds per replicate were weighed and placed in plastic cups containing 75 mL of deionized water, then incubated at 25 °C for 24 h (Vieira & Marcos-Filho, 2020). The electrical conductivity of the soaking solution was measured using a conductivity meter (MS Tecnopon^® - mCA150), and the results were expressed in μS cm ^-1 g ^-1.

Seed Health: Seed health was assessed using the Blotter test. Five replicates of 40 seeds were placed in an incubation room at 20 °C under a 12-hour photoperiod for seven days (Brasil, 2025). After incubation, seeds were examined under a light microscope for fungal identification, following the procedure described by Barnett and Hunter (1998).

Statistical analyses

Analysis of variance (ANOVA) was performed to assess the effects of genotype, environment, and their interaction on physiological, morphological, and health traits. The harmonic mean of the relative performance of predicted genetic values (MHPRVG) was calculated, as proposed by Resende (2004), to identify superior genotypes across environments. This method integrates performance, stability, and adaptability across environments based on predicted genetic values obtained from mixed models. The MHPRVG was calculated according to Equation (1):

${\text{MHPRVG}}_i=\frac{n}{{{\displaystyle \sum }}_{j=1}^n\frac{1}{V{g}_{ij}}}$ (1)

where n is the number of distinct conditions under which cultivar i was evaluated, and Vg_ij is the genotypic value of cultivar i under condition j, expressed as a proportion of the mean for that environmental condition.

Principal Component Analysis (PCA) was performed based on the correlation matrix of the normalized data. The GGE biplot model used is Yij -yj = y1 εi1ρj1 + y2 εi2ρj2 + εij, where Yij represents the physiological and health quality of genotype i in environment j; yj is the general mean of all genotypes in environment j; y1 εi1ρj1 corresponds to the first principal component (PC1); y2 εi2ρj2 represents the second principal component (PC2); and y1 and y2 are the eigenvalues associated with the first and second principal components (PCA1 and PCA2). In this model, ε1 and ε2 are the PC1 and PC2 values for genotype i, while ρj1 and ρj2 are the corresponding values ​​for environment j. Finally, εij represents the error associated with the i-th genotype in the j-th environment (Yan & Rajcan, 2002).

All analyses were performed using the R statistical software (R Core Team, 2024; v. 4.4.1) with the following packages: dplyr (Wickham et al., 2023) for data manipulation and aggregation; MASS (Venables & Ripley, 2002) for Box‒Cox transformations in linear models; lme4 (Bates et al., 2015) for mixed model; agricolae (Mendiburu, 2023); and functions from the utils and base packages. For GGE Biplot analysis, graphical outputs, and other analyses, additional functions from GGEBiplot, statistics, and graphics packages were used.

Results and Discussion

The average quality of the soybean seeds evaluated in this experiment was approximately 80%, meeting the minimum germination standard required for soybean seed lots in Brazil, which is 80%, as established by Brasil (2013). The ANOVA results (Table 2) indicate that the dose and timing of K application, combined with cultivar selection, significantly influence most physiological and health characteristics of soybean seeds. Additionally, most variables exhibited significant interaction effects, both double and, in some cases (e.g., water content (WC), electrical conductivity (EC), emergence (E), root dry matter (RDM), and thousand-seed mass (1000-S)), triple interactions involving time × cultivar × dose. This complexity suggests that the effect of one factor depends directly on the levels of the other factors, making isolated interpretations using mean tests exhaustive, fragmented, and difficult to synthesize. Given the high interdependence of the factors, a suitable approach for a more comprehensive understanding of the observed phenomena is the application of multivariate statistical techniques, such as Principal Component Analysis (PCA), which simplifies data complexity and identifies similarity patterns between treatments in a more practical way.

Table 2:
Mean squares estimates for water content (WC), first germination count (FGC), germination (G), electrical conductivity (EC), seedling emergence (E), accelerated aging (AA), shoot dry matter (SDM), root dry matter (RDM), thousand-seed mass (1000-S), and incidence of Fusarium sp. (FUS), Cercospora sp. (CER), and Phomopsis sp. (PHO) in two soybean cultivars (Brasmax Olimpo IPRO and CZ 58B23 I2X), evaluated under three potassium (K) application times: 100EA (100% early, 30 days before sowing), 50SO-50EM (50% at sowing and 50% at 30 days after emergence (DAE)), and 100TO (100% topdressing at 30 DAE), and five application rates (0, 50, 100, 150, and 200 kg K₂O ha^-1).

	WC		FGC		G		EC
Time (T)	7.372	**	64.889	ns	19.389	ns	0.125	ns
Cultivar (C)	14.553	**	88.889	ns	174.222	*	0.125	ns
Rate (R)	0.815	**	134.222	*	95.259	ns	317.421	**
T × C	2.602	**	107.556	ns	45.389	ns	1082.292	**
T × R	1.422	**	166.667	**	123.759	*	457.421	**
C × R	2.352	**	52.444	ns	49.037	*	305.940	**
T × C × R	3.866	**	20.889	ns	34.204	ns	287.884	**
	E		FUS		AA		CER
Time (T)	50.375	**	2006.347	**	5.056	ns	1476.389	**
Cultivar (C)	968.000	**	4035.014	**	40.500	ns	953.389	**
Rate (R)	72.333	**	612.421	*	134.574	ns	140.500	ns
T × C	285.875	**	72.347	ns	174.500	*	231.056	ns
T × R	112.708	**	540.866	*	274.019	**	549.278	**
C × R	34.333	**	54.718	ns	23.463	ns	173.241	ns
T × C × R	77.208	**	44.940	ns	110.130	ns	143.352	ns
	SDM		RDM		PHO		1000-S
Time (T)	1276.355	**	1.436	ns	17.167	ns	75.608	ns
Cultivar (C)	2120.850	**	9.901	**	3842.722	**	3261.627	**
Rate (R)	642.341	**	6.815	**	136.500	**	36.480	ns
T × C	323.133	ns	17.151	**	110.056	*	49.448	ns
T × R	394.439	ns	9.522	**	55.389	ns	53.357	ns
C × R	1126.722	**	2.922	ns	101.981	*	4.163	ns
T × C × R	365.114	ns	6.522	**	31.537	ns	34.643	*

• ^NS = not significant; ^* and ^** indicate statistical significance at the 5% (p ≤ 0.05) and 1% (p ≤ 0.01) levels, respectively, according to the F-test.

PCA revealed variability among treatments regarding the physiological and health variables of the soybean seeds (Figure 2). The first two principal components together explained 50.65% of the total variance in the data, with 29.95% attributed to component 1 (PC1) and 20.7% to component 2 (PC2). Analysis of the PCA plot showed that treatments associated with the cultivar Brasmax Olimpo IPRO were predominantly located in the right quadrant (highlighted in green). We observed positive correlations with most of the evaluated variables, including WC, 1000-S, FGC, G, E, SDM, RDM, AA, PHO, and CER (Figure 2). This distribution indicates that this cultivar showed superior performance under the evaluated experimental conditions. These results align with those reported by Coelho et al. (2023), who found that Brasmax Olimpo IPRO outperformed other cultivars in terms of germination and seed health.

Figure 2:
Principal Component Analysis (PCA) of variables: WC (water content), 1000-S (thousand-seed mass), FGC (first germination count), G (germination), E (emergence), SDM (shoot dry matter), RDM (root dry matter), EC (electrical conductivity), AA (accelerated aging), PHO (Phomopsis sp.), FUS (Fusarium sp.), and CER (Cercospora sp.), in two soybean cultivars (Brasmax Olimpo IPRO and CZ 58B23 I2X), evaluated under three potassium (K) application times: 100EA (100% early, 30 days before sowing), 50SO-50EM (50% at sowing and 50% at 30 days after emergence (DAE)), and 100TO (100% topdressing at 30 DAE), and five application rates (0, 50, 100, 150, and 200 kg K₂O ha^-1).

In contrast, treatments with cultivar CZ 58B23 I2X were predominantly positioned in the left quadrant (highlighted in purple), indicating lower performance, particularly in terms of vigor and seedling emergence. This separation highlights the differential physiological responses of the two cultivars to varying K fertilization strategies. These results confirm the findings of Ul-Allah et al. (2020) and Wasaya et al. (2021), who observed that different genetic materials respond distinctly to fertilization strategies. Specifically, genotypes with lower K-use efficiency tend to exhibit reduced vigor and emergence.

The harmonic mean of the relative performance of predicted genetic values (MHPRVG), as proposed by Resende (2004), indicated significant differences in the physiological, morphological, and health traits of seeds from CZ 58B23 I2X and Brasmax Olimpo IPRO cultivars (Table 3). The MHPRVG approach is valuable for selecting superior genotypes evaluated across various environmental conditions (Resende, 2004), as it allows for the integrated interpretation of performance, stability, and adaptability.

Table 3:
Harmonic mean of the relative performance of the predicted genetic values (MHPRVG) of the soybean cultivars Brasmax Olimpo IPRO and CZ 58B23 I2X for the variables WC (water content), 1000-S (thousand seed mass), FGC (first germination count), G (germination), E (emergence), SDM (shoot dry matter), RDM (root dry matter), EC (electrical conductivity), AA (accelerated aging), PHO (Phomopsis sp.), FUS (Fusarium sp.), CER (Cercospora sp.).

---------------------------------------------------------- HMRPPGV -------------------------------------------------------
	-- FGC--	-- G--	-- EC --	-- WC --	-- E --	-- ECA --
CZ 58B23 I2X	73.67	78.08	54.63	8.40	76.50	36.00
Brasmax Olimpo IPRO	76.33	79.75	54.75	8.94	79.13	36.83
	-- 1000-S --		-- CER --	-- FUS --	-- PHO --	-- SDM --
CZ 58B23 I2X	161.47		17.67	53.46	12.08	4.86
Brasmax Olimpo IPRO	163.03		24.08	41.67	14.75	4.38

• Interpretation is based on the absolute value of the MHPRVG.

These results were corroborated by the multivariate analysis using PCA (Figure 2), which clearly distinguished the two cultivars. The cultivar Brasmax Olimpo IPRO showed superior performance across most of the variables, which was associated with greater seed physiological performance and a higher incidence of Phomopsis sp. and Cercospora sp. Conversely, cultivar CZ 58B23 ICX showed a higher incidence of Fusarium sp. Araújo Neto et al. (2024) reported that Brasmax Olimpo IPRO exhibited high germination rates even when seeds were harvested late, demonstrating good physiological quality. Furthermore, Chang et al. (2020) noted that Fusarium sp. is one of the primary fungal pathogens that negatively impact seed quality by reducing seed weight, decreasing germination percentage, and causing seed rot. This underscores the importance of selecting cultivars resistant to fungal pathogens to ensure seed health.

The results suggest that Brasmax Olimpo IPRO exhibits greater vigor and better physiological quality, likely due to its adapted genetics and higher tolerance to stress during seed formation. These findings are consistent with those of Cavalcante et al. (2022), who also observed cultivar-specific responses to fertilization strategies, emphasizing the importance of selecting adapted genotypes for high physiological potential. Our results confirm that the cultivar Brasmax Olimpo IPRO exhibited superior seed vigor and physiological quality, reflecting its genetic adaptation to the experimental environment. These findings reinforce the importance of selecting cultivars that respond effectively to fertilization strategies.

PCA also allowed us to examine the responses of the two soybean cultivars (Brasmax Olimpo IPRO and CZ 58B23 I2X) to different K application dates and rates (Figures 3A and 3B). For Brasmax Olimpo IPRO, treatments with 200 kg ha^-¹ of K₂O, either 100% early (100EA) or 100% at 30 DAE (100TO), resulted in improved physiological and health variables such as germination (G), emergence (E), first germination count (FGC), root dry matter (RDM), water content (WC), and incidence of Phomopsis sp. (PHO) (Figure 3A). These treatments were associated with superior seed physiological and health quality. Mamun et al. (2022) and Ortel et al. (2024) noted that adequate K availability mitigates stress and maintains seed physiological performance, whereas K deficiency leads to reduced plant vigor, lower biomass accumulation, and increased vulnerability to adverse conditions. K is also linked to the synthesis of defense compounds and stabilization of cell membranes, which may hinder pathogen entry (Brito et al., 2024).

Figure 3:
Analysis of the principal components of the variables: WC (water content), 1000-S (thousand-seed mass), FGC (first germination count), G (germination), E (seedling emergence), SDM (shoot dry matter), RDM (root dry matter), EC (electrical conductivity), AA (accelerated aging), PHO (Phomopsis sp.), FUS (Fusarium sp.), and CER (Cercospora sp.). Analyses were performed separately for two soybean cultivars, Brasmax Olimpo IPRO (A) and CZ 58B23 I2X (B), evaluated under three potassium (K) application times: 100EA (100% early, 30 days before sowing), 50SO-50EM (50% at sowing and 50% at 30 days after emergence (DAE), and 100TO (100% topdressing at 30 DAE), and five application rates (0, 50, 100, 150, and 200 kg K₂O ha^-1).

In contrast, for CZ 58B23 I2X, the K rates of 100, 150, and 200 kg ha^-¹, applied either 50% at sowing and 50% at 30 DAE, or 100% at 30 DAE, were associated with better performance in FGC, G, E, and AA variables, indicating a positive response to split K fertilization. Mamun et al. (2022) also highlighted that the splitting of K fertilizer ensures nutrient availability during periods of stress, leading to greater stability in production and seed quality.

However, for cultivar CZ 58B23 I2X, treatments with no K application (0-K) or low rates with early application alone were associated with increased incidence of Fusarium sp. (FUS) (Figure 3B). Cai, Tao and Guo (2020) reported that K plays a crucial role in soybean resistance to Fusarium sp., where a K deficiency is associated with higher susceptibility to this pathogen, while adequate K doses help reduce the disease index and enhance seed health. Our results suggest that combining high K₂O rates with fractional or late application strategies may improve seed physiological quality and health, with varying responses between cultivars.

The GGE biplot revealed that Brasmax Olimpo IPRO demonstrated the best overall performance across treatments, as indicated by the apex of the blue arrow (Figure 4A). This finding is consistent with that of Sousa et al. (2025), who reported that the cultivar Brasmax Olimpo IPRO was positively correlated with health variables, even under delayed harvesting conditions. Goi et al. (2024) noted that cultivars with higher pathogen tolerance tend to exhibit superior development. Conversely, cultivar CZ 58B23 I2X was positioned at the opposite end, reflecting lower overall performance.

Figure 4:
GGE biplot showing the ranking of the best cultivar compared with other treatments and variables studied (A) and the ranking of the best potassium (K) fertilization strategies (B).

The treatments with 200 kg ha^-¹ of K₂O applied either 100% early (100 EA) or 100% at 30 DAE (100 TO) showed the greatest performance, positioned near the center and upper sectors of the biplot (Figure 4B). Slaton et al. (2020) reported that soybean K absorption efficiency remains high, even with late fertilizer applications, up to 83 days after sowing. This supports the good performance of the 100TO treatment and suggests that the cultivar Brasmax Olimpo IPRO has a high capacity for nutrient use in later stages. However, the findings regarding K dose contrast with those reported by Barbosa et al. (2025), who suggested that the ideal dose is 78 kg K₂O ha^-1, regardless of application time. In contrast, treatments involving CZ 58B23 I2X cultivar, using lower K rates or no fertilization (0K0), were grouped in opposite regions, indicating lower efficacy (Figures 4A and 4B). These results highlight that, in addition to cultivar selection, proper management of K fertilization is crucial for optimizing the physiological and health qualities of soybean seeds.

In terms of K soil levels, Sousa and Lobato (2004) consider K concentrations above 0.13 cmolc dm^-³ as adequate for crop development. In our study, the K levels in the experimental area were 0.15 and 0.29 cmolc/dm³ in the 0-20 cm and 20-40 cm layers, respectively, indicating sufficient initial K availability. However, the application of the highest K₂O dose (200 kg ha^-1) significantly enhanced seed physiological and health qualities, even under conditions of adequate initial availability. These results emphasize that K is essential not only for plant growth but also for seed filling and formation, ensuring higher vigor, germination, and pathogen resistance. Thus, the judicious management of K fertilization is crucial for maximizing crop performance and ensuring the production of high-quality seeds.

The observed results can be attributed to the essential roles of K in soybean physiology and nutrition. Potassium is directly involved in osmotic regulation, stomatal opening, photoassimilate transport, and the activation of more than 60 enzymes critical to plant metabolism. These functions promote balanced plant development and the formation of high-quality seeds (Oliveira et al., 2024). In addition, K influences K-use efficiency in modern genotypes by enhancing their capacity for nutrient uptake, translocation, and utilization, especially under low-K availability conditions typical of Brazilian Cerrado soils (Gomides et al., 2023). Thus, the differences observed in physiological and health traits among cultivars can be attributed to variations in their ability to acquire and utilize K, reflecting nutritional mechanisms that directly impact seed formation and viability.

The Pearson correlation matrix (Figure 5) provides valuable insights into the relationships among the physiological, physical, and health variables evaluated. A highly significant positive correlation was observed between germination (G) and the first germination count (FGC) (r = 0.92; p < 0.001), suggesting that seeds with higher initial germination speeds tend to achieve higher final germination percentages. Djedje et al. (2024) reported that seeds with better performance in FGC exhibited higher germination rates. Similarly, Araújo Neto et al. (2024) found that FGC decreases as seed quality declines, directly affecting final germination percentages.

Figure 5:
Pearson’s correlations between the evaluated variables. Abbreviations used: 1000-S: mass of one thousand seeds, WC: water content, E: emergence, EC: electrical conductivity, FGC: first germination count, G: germination, SDM: shoot dry matter, RDM: root dry matter, AA: accelerated aging, CER: Cercospora sp., PHO: Phomopsis sp., FUS: Fusarium sp.

Positive correlations were also observed between water content (WC) and thousand-seed mass (1000-S) (r = 0.36; p < 0.01), and between 1000-S and the presence of Phomopsis sp. (PHO) (r = 0.38; p < 0.001) (Figure 5). Additionally, a positive correlation was found between WC and root dry matter (RDM) (r = 0.46; p < 0.001). Chiluwal, Kawashima and Salmeron (2022) demonstrated that seeds with higher water content accumulate more dry mass, which is reflected in the increased mass of one thousand seeds. A positive correlation between water content and the presence of Phompsis sp. was also reported by Wang et al. (2020), who showed that higher water content accelerates fungal growth, resulting in seed deterioration.

Conversely, a negative correlation was observed between the incidence of Cercospora sp. (CER) and Fusarium sp. (FUS) (r = -0.61; p < 0.001) (Figure 5). This negative correlation may be attributed to competitive interactions between these fungal genera during seed colonization. Saracoglu et al. (2024) suggested that Fusarium has a high capacity for adaptation and seed colonization, often inhibiting the growth of competing microorganisms through the production of secondary metabolites and mycotoxins. Viljoen Truter and Kritzinger (2025) further reported that Fusarium predominates in environments with high fungal diversity and tends to exhibit low simultaneous occurrence with other fungi.

Conclusions

The combination of the Brasmax Olimpo IPRO cultivar with 200 kg ha^-¹ of K₂O, applied either 100% early or 100% at 30 days after emergence, resulted in superior seed physiological and health quality. Proper K management, considering both application rate and timing, is essential for optimizing seed vigor, germination, and pathogen resistance. These findings demonstrate that K fertilization improves soybean seed quality and supports management practices that enhance seed health and overall crop performance.

Acknowledgments

The authors also thank the Universidade Estadual do Maranhão (UEMA) for granting a scientific initiation scholarship to the first author. The authors are grateful to Accert Pesquisa e Consultoria Agronômica for providing financial support through agricultural inputs and for making the experimental area available.

Data Availability Statement

Data available upon request to authors.

References

• Araújo Neto, J. P. et al. (2024). Qualidade de sementes de soja submetidas ao retardamento de colheita. Contribuciones a las Ciencias Sociales, 17(13):e13813.
• Barbosa, M. C. et al. (2025). The effects of potassium dose, timing, and source in soybean crops in Brazilian savannah oxisol. Sustainability, 17(3):934.
• Barnett, H. L., & Hunter, B. B. (1998). Illustrated genera of imperfect fungi St Paul, MN: APS Press. 218p.
• Bates, D. et al. (2015). Fitting linear mixed-effects models using lme4. Journal of Statistical Software, 67(1):1-48.
• Brasil. Ministério da Agricultura, Pecuária e Abastecimento. (2013). Padrões para produção e comercialização de sementes de soja (Glycine max (L.) Merrill). Instrução normativa n° 45, de 17 de setembro de 2013. Diário Oficial da União, Brasília, DF, 20 set. 2013. Seção 1, p. 25.
• Brasil. Ministério da Agricultura e Reforma Agrária. (2025). Regras para análise de sementes Brasília: Ministério da Agricultura e Pecuária.
• Brito, S. et al. (2024). Efeito da nutrição de plantas na resistência a pragas e doenças. Revista Multidisciplinar do Nordeste Mineiro, 9(1):1-16.
• Cai, H., Tao, N., & Guo, C. (2020). Systematic investigation of the effects of macroo-elements and iron on soybean plant response to Fusarium oxysporum infection. The Plant Pathology Journal, 36(5):398.
• Cao, W. et al. (2025). Progress in the study of plant nitrogen and potassium nutrition and their interaction mechanisms. Horticulturae, 11(8):930.
• Cavalcante, J. P. J. et al. (2022). Potencial fisiológico de sementes de cultivares de soja no oeste da Bahia. Research, Society and Development, 11(12):499111234793.
• Chang, X. et al. (2020). Diversity of the seedborne fungi and pathogenicity of Fusarium species associated with intercropped soybean. Pathogens, 9(7):531.
• Chiluwal, A., Kawashima, T., & Salmeron, M. (2022). Soybean seed weight responds to increases in assimilate supply during late seed-fill phase. Journal of Crop Improvement, 36(2):222-238.
• Coelho, Y. P. et al. (2023). Physiological quality of soybean seeds as a function of organic-mineral fertilizers. Contribuciones a las Ciencias Sociales, 16(10):23632-23650.
• Companhia Nacional de Abastecimento - CONAB. (2025). Acompanhamento da safra brasileira 2024/2025 (Vol. 12, No. 12, 132 p.). Brasília: Conab.
• Djedje, I. et al. (2024). Efeito da salinidade na germinação de soja (Glycine max). Revista de Ciências Agrárias, 47(3):443-453.
• Dong, Z. et al. (2023). Effect of different NPK fertilization timing sequences management on soil-petiole system nutrient uptake and fertilizer utilization efficiency of drip irrigation cotton. Scientific Reports, 13:14287.
• Goi, R. J. et al. (2024). Incidence of diseases and selection of soybean cultivars (Glycine max L. Merril) in a water deficit environment. Revista de Ciências Agroveterinárias, 23(2):211-220.
• Gomides, J. F. F. B. et al. (2023). Identification of modern high-yield soybean genotypes for potassium-use efficiency in sandy soil of the Brazilian Cerrado.Agronomy,13(10):2639.
• Instituto Nacional de Meteorologia- INMET. (2025). Banco de Dados Meteorológicos Available in: <https://bdmep.inmet.gov.br/>.
  » https://bdmep.inmet.gov.br/
• Lago, B. C. et al. (2023). Potassium application timing to improve corn K-fertilizer use in the oat-corn sequence: A tracer study for high yielding corn. Journal of Plant Nutrition, 46(4):618-629.
• Lara, L. M. et al. (2018). Quality of soybean seeds cultivated on different potassium fertilization management. Journal of Agricultural Science, 10(11):211-218.
• Liu, J. et al. (2024). Effect of nitrogen, phosphorus and potassium fertilization management on soil properties and leaf traits and yield of Sapindus mukorossi Frontiers in Plant Science, 15:1300683.
• Marcos-Filho, J. (2020). Teste de envelhecimento acelerado In F. C Krzyzanowski. et al. (Eds.) Vigor de sementes: Conceitos e testes Londrina: ABRATES, (pp.185-246).
• Mamun, M. A. A. et al. (2022). Application of potassium after waterlogging improves quality and productivity of soybean seeds. Life, 12(11):1816.
• Mendiburu, F. de. (2023). agricolae: Statistical procedures for agricultural research (R package version 1.3-7). Available in: <https://CRAN.R-project.org/package=agricolae>.
  » https://CRAN.R-project.org/package=agricolae
• Oliveira, F. A. de. et al. (2024). Potássio nos sistemas de produção de soja Londrina: Embrapa Soja. 255p.
• Ortel, C. C. et al. (2024). Interaction of drought stress and potassium deficiency on soybean vigor and leaf temperature. Agrosystems, Geosciences & Environment, 7(4):e20576.
• R Core Team. (2024). R: A language and environment for statistical computing Vienna: R Foundation for Statistical Computing. Available in: <https://R-project.org/>.
  » https://R-project.org/
• Resende, M. D. V. (2004). Métodos estatísticos ótimos na análise de experimentos de campo 1. ed. Colombo, PR: Embrapa Florestas. 57p.
• Ruppin, N. W. et al. (2019). Caracterização morfológica de sementes de diferentes cultivares de soja armazenadas sob condições não controladas. Agrarian Sciences Journal, 11:1-8.
• Saracoglu, K. et al. (2024). Fusarium species and assessments of mycotoxin (deoxynivalenol), in wheat seeds from different regions of Türkiye. Phytopathologia Mediterranea, 63(2):233-253.
• Santos, H. G. et al. (2018). Sistema Brasileiro de Classificação de Solos (5ed.). Brasília: Embrapa. 356p.
• Silva, F. et al. (2022). Soja: do plantio à colheita São Paulo: Oficina de Textos. 304p.
• Slaton, N. A. et al. (2020). Irrigated soybean response to granular fertilizer potassium application timing. Agronomy Journal, 112(5):4344-4357.
• Sousa, C. H. C. et al. (2025). Physiological and sanitary quality of soybean seeds in response to harvest delay. Revista Ciência Agronômica, 56:e202493440.
• Sousa, D. M. G., & Lobato, E. (2004). Cerrado: Correção do solo e adubação (2ed.). Brasília: Embrapa Informação Tecnológica. 416p.
• Sousa, C. H. C. et al. (2024). Physiological and sanitary quality of soybean seeds in response to harvest delay. Revista Ciência Agronômica, 56:e202493440.
• Ul-Allah, S. et al. (2020). Potassium application improves grain yield and alleviates drought susceptibility in diverse maize hybrids. Plants, 9(1):75.
• Venables, W. N., & Ripley, B. D. (2002). Modern Applied Statistics with S-PLUS (4th ed.). New York: Springer.
• Vieira, R. D., & Marcos-Filho, J. (2020). Teste de condutividade elétrica. In F. C. Krzyzanowski. et al. (Eds.). Vigor de sementes: Conceitos e testes Londrina: ABRATES, (pp.333-388).
• Viljoen, J. J. F., Truter, M., & Kritzinger, Q. (2025). Mycoflora and mycotoxins associated with stored cowpea (Vigna unguiculata L. Walp) seeds from smallholder farmers in South Africa. Phytoparasitica, 53:44.
• Wang, R. et al. (2020). Effects of deterioration and mildewing on the quality of wheat seeds with different moisture contents during storage. RSC Advances, 10(25):14581-14594.
• Wasaya, A. et al. (2021). Foliar potassium sulfate application improved photosynthetic characteristics, water relations and seedling growth of drought-stressed maize. Atmosphere, 12(6):663.
• Wickham, H. et al. (2023). dplyr: A grammar of data manipulation (R package version 1.1.4). Available in: <https://CRAN.R-project.org/package=dplyr>.
  » https://CRAN.R-project.org/package=dplyr
• Yan, W., & Rajcan, I. (2002). Biplot evaluation of test sites and trait relations of soybean in Ontario. Crop Science: 42(1):11-20.
• Zambiazzi, E. V. et al. (2017). Desempenho agronómico e qualidade sanitária de sementes de soja em resposta à adubação potássica. Revista de Ciências Agrárias, 40(3):543-553.
• Zi-Qiang, H. et al. (2025). Effect of potassium fertilizer application on photosynthetic effect of spring maize. Chinese Journal of Agrometeorology, 46(3):384.

• Editor de seção:
  Renato Paiva

Publication Dates

• Publication in this collection
  20 Apr 2026
• Date of issue
  2026

History

• Received
  25 Oct 2025
• Accepted
  02 Feb 2026