elocation-id: elocation-id: e4195
Soybeans are a strategic source of plant protein for animal nutrition; however, the presence of antinutritional factors such as trypsin inhibitors, phytates and saponins limits their utilization when the grain is used unprocessed. The research was conducted in 2025 and evaluated the effects of four postharvest conditioning methods (soaking, boiling, germination and roasting) on antinutritional factors, crude protein and lipid concentrations in six soybean varieties grown in Puebla (HSO, RGNA, BM2, HSVA, Varita and Tamesí). Crude protein, total lipids, phytates, trypsin inhibitors, and saponins were determined and analyzed using a completely randomized design with a two-factor (variety x method) factorial arrangement and Tukey’s multiple comparisons (α= 0.05). The results showed differences between varieties and conditioning methods. The Hoja Seca Original variety stood out for having the highest crude protein content, intermediate and stable total lipid values, and an efficient reduction of the main antinutritional factors after treatments. In contrast, the BM2 variety had the lowest protein content and the greatest sensitivity to heat treatment. Raw grain had the highest concentrations of antinutritional factors. Roasting and boiling achieved the greatest reductions in trypsin inhibitors, while germination and soaking decreased phytates and saponins with less impact on protein content. The varieties differ in chemical composition, but this variation is less than the effect produced by the conditioning method, with roasting being the one that best reduced antinutrients.
family livestock farming, phytates, trypsin inhibitors.
Soybeans are the most important plant protein source for the livestock industry because of their high protein and energy content, essential amino acid profile, and highly digestible oil. In Mexico, their use is strategic in animal nutrition, particularly for poultry, cattle, and pigs, whose diets require crude protein concentrations between 11 and 23% (INIFAP, 2019).
Nonetheless, the country is dependent on imports, since domestic production meets between 4 and 8% of domestic consumption, being subject to international price fluctuations, which limits the profitability of small livestock producers (SADER, 2023).
A second obstacle to the use of grain is the presence of antinutritional factors, mainly trypsin inhibitors, saponins and phytates, which reduce protein digestibility and mineral bioavailability (De la Luna, 2006). Therefore, it is necessary to apply inactivation processes before use in animal nutrition.
The variation in the response of soybeans to conditioning methods is known to depend on both the nature of the antinutritional factor and the characteristics of the genotype (Singer et al., 2023). Heat treatments favor the reduction of thermolabile compounds, particularly trypsin inhibitors, through protein denaturation; however, intense treatments can affect amino acid availability due to side reactions such as the Maillard reaction (Kong et al., 2022).
On the other hand, processes such as soaking and germination can reduce phytates through solubilization and enzymatic activation, representing alternatives with lower technological demand for production systems with limited resources (Sharma, 2021; Kong et al., 2022; Wang et al., 2022).
Despite the agricultural relevance of soybeans, information is incomplete regarding variation in antinutrient content among varieties adapted to rainfed conditions in temperate zones and dry tropics subject to easy-to-apply and low-cost inactivation methods, and how they can affect the composition of grain used in family livestock production systems.
Therefore, the objective of the research was to evaluate the variation in the contents of crude protein, lipids and antinutritional factors in soybean varieties subjected to inactivation conditioning.
The commercial varieties Tamesí and BM2 were used, along with the experimental varieties Hoja Seca Original (HSO), Hoja Seca Vainas Abundantes (HSVA), Varita and RGNA, which are grown under rainfed conditions in Atlixco, Puebla, Mexico. The climate is subhumid temperate, with an average annual temperature of 16.1-21.8 °C, average rainfall of 906 mm and sandy loam soil.
The crop was established with conventional management practices in a completely randomized experiment with four replications, from which the grain was obtained, keeping each replication separate.
The harvested grain, in its raw form, was dehydrated in a forced-air oven at 40 °C for 72 h until it reached a constant weight, then pulverized in an electric mill. Soaking lasted 5 h at 20 °C, using three times the mass of water relative to the grain weight; later, the grain was sun-dried, spread on metal trays.
Germination consisted of soaking for 48 h in purified water, after which the grain was placed on plastic trays in the shade, maintaining moisture for five days when the radicle emerged. The grain was then removed and placed on metal trays for sun-drying.
Boiling lasted 30 min; the grain was drained through a strainer and placed on trays for sun-drying. Roasting was carried out in an electric oven at 150 °C for 20 min, with the grain spread evenly on metal trays.
All the grains were dehydrated in a forced-air oven at 40 °C for 72 h to be pulverized in an electric mill at 30 000 rpm.
Crude protein (CP) was determined by Kjeldahl (method 2001.11; AOAC, 2012a); total lipids by Soxhlet (method 920.39; AOAC, 2023); saponins by sequential extraction with ether, 80% ethanol and n-butanol (Hostettmann et al., 1995); trypsin inhibitors by spectrophotometry using BAPNA as a substrate [Welham and Domoney, 2000; AACC 22-40, 2009; AOCS (2017) Ba 12/12a; ISO 14902 (ISO, 2001)], and phytates by the method of Wheeler and Ferrel (1971), quantifying iron by colorimetry (AOAC 2012b, method 986.11).
The factors variety, method of inactivation, and their interaction generated significant differences in most of the three sources, except for total lipids (Table 1). The variety factor had a greater effect on CP and total lipids (60 and 90% of the total variation explained, respectively); by contrast, the method-of-inactivation factor was decisive for trypsin inhibitors, phytates and saponins, explaining 99, 97 and 99%, respectively. On the other hand, the interaction contributed little, explaining at most 6% in the CP variable.
The varieties differed in CP content p< 0.0001 (Table 2). HSO had the highest concentration (32.6%). RGNA ranked second (30.3%); Tamesí, Varita and HSVA were similar, with intermediate values (28% on average), and BM2 had the lowest concentration (25.9%).
The difference between HSO and BM2 was 6.7%, which was a significant variation. These differences were consistent with reports by several authors (Rotundo and Westgate, 2009; Panthee et al., 2016), who attribute the variation to the genetic basis and the relative protein-to-oil ratio in soybeans.
Usually, in soybean seeds, protein concentration can present an inverse relationship with lipid concentration, because both components share metabolic pathways during reserve accumulation in the grain.
Likewise, breeding programs can prioritize increases in protein or oil, depending on the productive objective of the varieties developed, which can also lead to differences among genotypes.
The CP concentration varied with the inactivation method (p< 0.0001). Raw grain had the highest value (30.4%). There was a 0.9% decrease with the soaking and boiling methods; germination caused a 2% reduction, and roasting had the lowest value (26.3%), with a 4.1% reduction compared with raw grains.
According to Liu (2019), roasting can slightly reduce CP content, an aspect related to Maillard reactions, which influence CP quantification. The other losses occur due to solubilization (Sharma et al., 2013) and the use of reserves. Although roasting presented the greatest reduction in antinutritional factors, its application implies availability of thermal equipment and energy consumption, factors that may limit its adoption in family livestock production units.
In contrast, methods such as soaking and germination require less infrastructure and may be low-cost alternatives, although they have less capacity to reduce trypsin inhibitors. Therefore, selecting the conditioning method should consider not only inactivation efficiency, but also the availability of resources, the scale of production and the grain’s final destination.
Regarding the variety x treatment interaction, the RGNA and BM2 varieties showed the greatest combined effect (p≤ 0.05). In both cases, the roasting treatment produced the most pronounced decreases in protein compared with the other methods, an aspect that Liu (2019) related to genetic differences in the protein fraction and the protein-oil ratio.
The varieties were different (p< 0.0001) in total lipid concentration (Table 3), but with a reduced range (20-22%). The RGNA variety had the highest concentration (21.4%), whereas HSO and Varita had the lowest values (20%).
According to Hou and Chang (2004); Rotundo and Westgate (2009); Liu (2019), these small differences are due to the lipid content in soybeans presenting moderate genetic variability, but it can be influenced by factors of accumulation efficiency during maturation and the composition of the integument or as Wilson (2004) points out, it can be due to differences in fatty acid biosynthesis pathways.
Inactivation methods did not affect lipid concentration (p= 0.7955). Hu et al. (2002) point out that, in soybeans, lipids are not very susceptible to losses due to moderate heating (≤ 150 °C) or short soaking because of their location in the seed.
The varieties differed in trypsin inhibitor concentration p< 0.0001 (Table 4). Varita and Tamesí differed from HSO and HSVA by having 3.8% more on average (402 TIU g-1). Kakade et al. (1973); Liener (1994) describe that variation in inhibitor concentration is mainly attributable to genetic factors of the seed.
Inactivation methods significantly reduced (p< 0.0001) the concentration of trypsin inhibitors. Raw grain had the highest value (38 321 TIU g-1), whereas conditioning treatments reduced the inhibitory activity by 71-121%. Boiling had an effective reduction of 5 700 TIU g-1 (89%), and roasting was more efficient, with a reduction of 2 509-3 431 TIU g-1 (91-94% less than raw grain).
The variety x conditioning interaction was not uniform (p< 0.0001) among varieties, with RGNA showing the greatest variation in the soaking, boiling, and germination methods. Panthee et al. (2016); Liu (2019) mention that, among varieties, there are certain genotypes that vary in the relative proportions of Kunitz and Bowman-Birk inhibitors and in their thermal stability, which could be related to this variety.
The varieties showed differences in phytate content p< 0.0001 (Table 5). RGNA had a higher concentration than Varita, HSO and BM2; the latter had the lowest value (1.17 ppm). This variability is attributable to genetic differences, which aligns with observations by Lott et al. (2000); Raboy (2001), who observed that phytate levels in soybean seeds vary among genotypes due to differences in the accumulation of phytic phosphorus during grain filling and to the proportion of cotyledon relative to the total seed.
Inactivation methods, on the other hand, had a differentiated effect (p< 0.0001) on phytate concentration. Raw grain presented the highest value (3.3 ppm); in contrast, soaking, boiling and germination showed similar reductions (1-1.1 ppm). This is related to the partial solubilization of phytates in water during soaking, the activation of endogenous phytases at the beginning of germinative metabolism, and the diffusion of phytic acid into the water.
Roasting recorded the lowest concentration (0.9 ppm), indicating that dry heat contributes both to the partial denaturation of phosphorus complexes and to the thermal degradation of phytic acid. These results coincide with reports by Sandberg et al. (1999); Luo et al. (2009), who point out that thermal and biological processes effectively reduce phytates through enzymatic hydrolysis, solubilization or heat decomposition.
In the interaction, the most pronounced changes were observed in HSVA in germination; Tamesí showed greater stability and Varita had abrupt reductions.
The varieties were different (p< 0.05) in saponin content (Table 6), although the range was narrow (0.4-0.42%). This aligns with Shimoyamada et al. (1990), who indicated that the total saponin content in soybeans remains relatively stable among yellow-type genotypes, varying mainly with culture conditions and physiological maturity, rather than with genetic factors.
For their part, the inactivation methods reduced the concentration (p< 0.0001) compared with the raw grain. Soaking reduced the value to 0.09%, but roasting had the greatest reducing effect, reaching 0.23 percentage units. The variety x conditioning interaction was highly significant (p= 0.0011). Although all varieties followed the same general trend, the intensity of the decline was not uniform.
HSO had the largest drop, and raw grain had the greatest variation. Water conditioning methods had intermediate values. This aligns with the findings of Gu et al. (2002), where water causes partial degradation of DDMP saponins by oxidation associated with the initial metabolism of seed germination.
The varieties differ in the concentrations of CP, lipids and inhibitors; however, the ranges of variation are small.
Grain conditioning reduces the concentrations of trypsin inhibitors, phytates and saponins, without affecting lipid concentration, with roasting being more effective in reducing antinutrients but significantly decreasing crude protein.
Nonetheless, the implementation of roasting must be assessed considering protein loss, energy requirements and technological availability. In family production systems, methods such as soaking or germination can represent practical alternatives; nevertheless, they require optimizing their capacity to reduce antinutrients.
The variety x conditioning method interaction is small; however, it confirms that there are specific biochemical and structural characteristics of each genotype whose response depends on the method of conditioning applied.
Therefore, the choice of variety and processing type can be decisive for the use of grain in animal nutrition. Roasting was the most effective at reducing antinutritional factors.
AOAC. 2012a. International. Official Method 2001.11: protein (crude) in animal feed, forage (plant tissue), grain and oilseeds: block digestion method using copper catalyst and steam distillation into boric acid. In: official methods of analysis of AOAC international. AOAC International. Gaithersburg, MD, USA.
Singer, W. M.; Lee, Y. C.; Shea, Z.; Vieira, C. C.; Lee, D.; Li, X.; Cunicelli M.; Kadam, S. S.; Khan, M. A. W.; Shannon, G.; Mian, M. A. R.; Nguyen, H, T. and Zhang, B. 2023. Soybean genetics, genomics, and breeding for improving nutritional value and reducing antinutritional traits in food and feed. The Plant Genome. 16(4):e20415. https://doi.org/10.1002/tpg2.20415.
Wilson, R. F. 2004. Seed composition. In: Soybeans: Improvement, production and uses. Boerma, H. R. and Specht, J. E. (Eds.). 3th. edition. American Society of Agronomy. Crop Science Society of America and Soil Science Society of America. Madison, WI, USA. 621-677 pp. https://doi.org/10.2134/agronmonogr16.3ed.c13.