Grain, masa, and tortilla quality of native maize races from Central Mexico: valuing agrobiodiversity for conservation and food use

1 Introduction

Maize (Zea mays L.) is recognized as a strategic element for food sovereignty and security in Mexico, Central America, Africa, and Asia, due to its relevance in food, economic, social, cultural, and political spheres (González-Cortés et al., 2016). In Mexico, 59 native maize races are available, although it is estimated that there are up to 64 races in total (Kato et al., 2009; Caballero-García et al., 2019). A race is defined as a group of populations that share similar characteristics and are adapted to specific agroecological conditions (Wellhausen et al., 1951). A variety, on the other hand, is a distinct population within a maize race, defined by specific characteristics such as grain color and texture, as well as the duration of its growing cycle (Sierra-Macías et al., 2018). Accurately establishing the total number of maize varieties in Mexico is challenging, since varietal diversity is constantly changing because of ongoing processes of selection, conservation, exchange, and adaptation by producers. Native maize races are used in the preparation of a wide variety of traditional dishes, in addition to other uses the plant may have (González-Cortés et al., 2016; Njeri et al., 2017; Singh et al., 2020). The tortilla is considered fundamental to food security and has been one of the main foods in the Mexican diet for the past 3, 500 years; the annual per capita consumption in urban areas reaches 57 kg, while in rural areas it is 80 kg (Escobedo, 2023; Ramítrez-Rivera et al., 2024).

There is considerable maize diversity of maize in Mexico, with native maize races adapted to diverse climatic, geographic, and agricultural management conditions, giving rise to a variety of forms, colors, and sizes of both the ears and the plants (Kato et al., 2009). It has thus been proposed that the broad genetic diversity present in native maize races is the result of millennia of selection by farmers, as well as the interaction between different genotypes and the diverse environmental conditions found in the rainfed agroecosystems of these regions (Muñoz, 2005). This gives native maize a high potential for adaptation to events associated with climate change (Muñoz, 2005). This potential has been demonstrated by native corn achieving higher grain yields than improved varieties, due to its potential for adaptation and stability, developed indirectly year after year by producers (Santiago-López et al., 2017; López-Morales et al., 2017).

In 2023, the area of maize harvested in the state of Puebla totaled 420, 898 hectares, with an average yield of 2.36 tons per hectare (SIAP, 2024). Ninety percent of that total was native maize, used primarily for household consumption as food for both humans and animals, as well as being used as seed for the subsequent agricultural cycle (Muñoz, 2005; Sierra-Macías et al., 2016). In this state, there are approximately 994 thousand small-scale maize producers distributed across 28.4 thousand communities focused on subsistence production (SIAP, 2024). These native maize producers play a crucial role in the conservation of traditional local varieties through the selection and planting of seeds with desirable phenotypic characteristics. However, this system faces significant challenges, including the abandonment of cultivation and the loss of genetic diversity (Flores-Cruz et al., 2014; Rivera et al., 2021).

Several studies have evaluated the quality of maize grain, masa, and tortillas in different regions of Mexico to identify traits that enhance their suitability for industrial processing, functional food production, and the development of value-added products (Martínez-Bustos et al., 2001; García-Cruz et al., 2023). However, studies focused on evaluating grain and tortilla quality in native maize from the state of Puebla are scarce (Vázquez et al., 2003, 2011), despite the fact that the state constitutes one of the main centers of maize race diversity in the country (Pérez-García, 2023). At least 14 maize races have been reported in Puebla, reflecting the importance of this state for the conservation of maize genetic resources in Mexico (Pérez-García, 2023).

The objective of this study was therefore to evaluate the physical and colorimetric characteristics of the grain, its hydration properties during nixtamalization, and the texture of tortillas prepared from 16 native maize varieties from the state of Puebla, Mexico, in order to determine their technological potential and suitability for tortilla production.

2 Materials and methods

2.1 Evaluated germplasm

Samples were collected from 12 municipalities in the state of Puebla, Mexico, during November and December 2022. A total of 16 native maize races were identified. Each race was represented by 10 morphologically representative ears, which were used for visual characterization based on ear shape and size, number of kernel rows, and kernel type and color (SADER-SNICS, 2024). Dr. Juan Manuel Hernández-Casillas, head of the Germplasm Bank at the National Institute of Forestry, Agricultural and Livestock Research (INIFAP), identified the races via visual examination supported by specialized literature. Table 1 presents the race names, identification codes, collection municipalities and localities, collection altitudes, and kernel colors. After collection, the maize ears were shelled, and the kernels were dried at room temperature until they reached a moisture content of 13% (Castorina et al., 2023). The samples were subsequently evaluated at the Maize Quality Laboratory of the Valle de México Experimental Station (CEVAMEX-INIFAP), located in Coatlinchán, Texcoco, Mexico State.

Table 1. Location and identification of the 16 maize races collected in the State of Puebla, Mexico (2022).

2.2 Grain and tortilla characteristics

The flotation index (FI) was used as an indirect indicator of grain hardness, in accordance with the NMX-FF-034/1 standard (SE, 2020a), which establishes the appropriate nixtamal cooking times based on the value of this index. Specifically, maize classified as very hard (FI of 0–12) was cooked for 45 minutes; hard (FI of 13–37), for 40 minutes; intermediate (FI of 38–62), for 35 minutes; soft (FI of 63–87), for 30 minutes; and very soft (FI of 88–100), for 25 minutes. These criteria are adopted by both the Masa and Tortilla Industry (Industria de Masa y Tortillas; IMT) and the Nixtamalized Flour Industry (Industria de Harina Nixtamalizada; IHN). In addition, hectoliter weight (HW) was determined and, as an indirect measure of grain size, the weight of 100 kernels (100KW) was measured, in accordance with the NMX-FF-034 standard (SE, 2020b).

To determine tortilla quality, the conventional nixtamalization method was used, which comprises the grain, dough, and tortilla stages, as described by Vázquez-Carrillo et al. (2016). The moisture percentages of the nixtamal (NM) and the dough (DM) were evaluated and recorded using the AACC method 44-10 (2000). Likewise, tortilla moisture was evaluated 2 hours after preparation (TM), and after 24 (TM24) and 48 (TM48) hours, respectively, using method 44-10 (AACC, 2000). Furthermore, the solids loss (SL) from the nejayote (cooking water) was evaluated according to the methods described by Salinas and Vázquez (2006). Once cooled, the fresh tortillas were placed in Ziploc brand resealable plastic bags and stored at 4 °C until the evaluation times (24 and 48 hours), following the procedure described by López-Morales et al. (2023b).

2.3 Kernel color and tortilla texture

Grain color (Luminosity, Hue Angle, and Chroma) was determined using a Hunter Lab MiniScan XE Plus Model 45/0-L colorimeter with a D/65 illuminant and a 10° viewing angle, using the CIELAB color space as a reference according to the method described by Salinas and Vázquez (2006). The force and distance required to tear the tortillas after 2, 24, and 48 hours of storage were quantified to characterize the texture of the tortilla. For this, we used a Brookfield Model CT3 texture analyzer (Middleboro, MA, U.S.A.) with a 10 mm diameter spherical probe. Breaking force results were expressed in gram-force (g_f), while breaking distance was expressed in millimeters (mm) (Vázquez-Carrillo and Santiago-Ramos, 2019; López-Morales et al., 2021a). All of these variables were measured in duplicate.

2.4 Statistical analysis

Populations of 16 native maize varieties were collected in the state of Puebla, Mexico (Figure 1). In the laboratory, each population was divided into two samples, which were processed and analyzed independently under the same experimental conditions to determine the physical characteristics of the grain, nixtamalization variables, and tortilla texture. Each processed sample constituted an experimental unit for statistical analysis. For each variable, an analysis of variance (ANOVA) was performed under a completely randomized design, considering the maize variety as a fixed factor. When significant differences were detected (P ≤ 0.05), the means were compared using Tukey’s honestly significant difference (HSD) test. The assumptions of normality of residuals and homogeneity of variances were previously assessed using the Shapiro-Wilk and Levene tests, respectively. Statistical analyses were performed using SAS version 9.0 (SAS Institute, 2002).

3 Results and discussion

Highly significant differences (P ≤ 0.01) were observed among the sixteen maize races for all variables analyzed, according to the analysis of variance (Tables 2, 3). The races showed wide variability in endosperm hardness, ranging from very hard to very soft, as well as considerable differences in hectoliter weight and grain size, which ranged from small to very large. This result is consistent with recent studies that have documented high genetic and phenotypic variability in the physical properties of maize grain, attributing it mainly to high genetic diversity and its interaction with the environment (Salinas-Moreno et al., 2024a; Oliva-García et al., 2026). The diversity of kernel colors (white, cream, yellow, blue, orange, and red) described in Table 1 reflects the heterogeneity of the races collected in the state of Puebla (Figure 1) (Salinas et al., 2013). Likewise, differences in tortilla breaking strength and breaking distance were also evident, as was observed in maize populations from the state of Sonora, Mexico that present different kernel colors (Vázquez-Carrillo et al., 2014).

Table 2. Physical and nixtamalization characteristics of native maize grains collected in the State of Puebla, Mexico (2022).

**: significance at P ≤ 0.01 in the ANOVA for the factor maize race, respectively; HW: hectoliter weight, HGW: hundred-grain weight, NM: nixtamal moisture, DM: dough moisture, TM, TM24, TM48: moisture of freshly made tortillas, at 24 hours, and at 48 hours, respectively, SL: solids loss. MSE: Mean square error, Root MSE: Root mean square error, CV: Coefficient of variation. Means with different letters within columns are statistically different (Tukey α 0.05). HSD: honestly significant difference.

Table 3. Grain color characteristics and tortilla texture in native maize races collected in the state of Puebla, Mexico (2022).

**: significance at P ≤ 0.01 in the ANOVA for the factor maize race, Lu: Luminosity, Hu: Hue, Ch: Chroma (Grain color); F, F24, F48: breaking force of freshly prepared tortillas and after 24 and 48 hours, respectively. D, D24, D48: breaking distance of freshly prepared tortillas and after 24 and 48 hours, respectively. MSE, Mean square error; Root MSE, Root mean square error; CV, Coefficient of variation. Means with different letters in the same column are statistically different (Tukey, α = 0.05). HSD, honestly significant difference.

Maize is a crop characterized by extraordinary genetic, morphological, and colorimetric diversity, which is reflected in its wide range of culinary uses in Mexico (Preciado-Ortiz and Vázquez-Carrillo, 2022). Consequently, producers select different maize varieties for specific purposes based on their physical and technological characteristics. For example, the Arrocillo Yellow (AA) and Arrocillo Red (AR) varieties have not traditionally been used for tortillas, but rather for popcorn, as they belong to a group of native Mexican maize varieties with kernel-popping capacity when exposed to heat (Pérez et al., 2024). However, although their potential for tortilla production has not been fully characterized, and further evaluation of their technological properties related to masa and tortilla quality could broaden their potential applications and diversify markets for maize-based products.

3.1 Physical characteristics of the kernels

Nixtamalization, or alkaline cooking, is a widely used traditional process that is primarily used to produce masa, an ingredient for many dishes in Mexican cuisine. In this procedure, the hardness of the corn kernel is a determining factor in establishing the cooking time and the final quality of the tortillas, with softer kernels generally resulting in higher tortilla quality (Preciado-Ortiz and Vázquez-Carrillo, 2022). In the present study, the Pepitilla (PE) race had very hard kernels, with a flotation index (FI) ranging from 0 to 12, as well as a high average hectoliter weight of 79.2 kg hL⁻¹ and a small hundred-grain weight (HGW = 26.1 g). The Arrocillo Yellow (AA), Arrocillo White (AB), Arrocillo Red (AR), Arrocillo Cream (AC), and Arrocillo (AZ) had hard kernels, with FI between 13 and 37 (Table 2). This group of Arrocillo races had similar hectoliter weight and hundred-grain weight to Arrocillo races in the state of Hidalgo (Vázquez et al., 2010). The differences in hectoliter weight and hundred-grain weight among the evaluated races may be attributed to variation in kernel size, shape, and density, as well as to genotypic adaptation across altitudinal gradients and agroecological conditions in the highlands of northeastern Puebla, Mexico. These findings are consistent with recent studies showing that grain physical characteristics are influenced by genotype–environment interactions, which directly affect kernel weight and hectoliter weight (Gambin et al., 2023; Qu et al., 2024; Arora et al., 2024).

The kernels of the Conico (CO) and Conico Yellow (CA) races had intermediate hardness, with FI values ranging from 38 to 62. They also presented a hectoliter weight of 76 kg hL⁻¹ and a hundred-grain weight ranging from 20 to 32 g, indicating medium-sized kernels. On the other hand, the Pepitilla (PP), Elotes Conicos (EL), and Tuxpeño (TU) races had soft grains (FI between 63 and 87); intermediate to low hectoliter weight values, ranging from 64 to 73 kg hL⁻¹; and medium to large kernel size, with hundred-grain weight values ranging from 24 to 34 g. The Elotes Conicos (EC), Chalqueño (CH), Elotes Conicos (ET), Zapalote Chico (ZC), and Elotes Conicos (EN) races had very soft endosperm, with an FI between 88 and 100; intermediate to high average hectoliter weight (between 59 and 75 kg hL⁻¹); and kernel size values ranging from intermediate to high, with hundred-grain weight between 28 and 58 g (Table 2). These results are consistent with recent studies conducted on native maize varieties, which have shown that grains with floury or soft endosperm have lower hectoliter weights, a higher flotation index, and faster hydration during nixtamalization, characteristics that result in higher dough and tortilla quality (Acosta-Estrada et al., 2023; Salinas-Moreno et al., 2024b; Vázquez-Carrillo et al., 2024). The observed differences in hardness, hectoliter weight, and hundred-grain weight (physical characteristics of the grain) could be attributed to cross-pollination between maize plants, since these characteristics are shared among genotypes (Mazo and Rodríguez, 2021).

3.2 Tortilla quality

The moisture content of the nixtamal (NM) ranged from 42.8% (AA race) to 50.6% (PP race). These values exceed the 36–42% range established by the NMX-034 standard (SE, 2020b) (Table 3). As for the moisture content of the masa (DM), a range of 54.1% (CH breed) to 60.4% (CA race) was recorded, higher than the 50–58% range reported by Antuna et al. (2008). These differences were inversely proportional to the percentage of moisture loss from the grain and reflect the variations in the water absorption capacity of each race during nixtamalization, a phenomenon that is closely related to endosperm hardness, pericarp structure, and the physical characteristics of the grain (Serna-Saldivar, 2021; Mariscal-Moreno et al., 2022; Barajas-Olivares et al., 2025). The moisture content of freshly made tortillas (TM), as well as the moisture after 24 h (TM24) and 48 h (TM48) of storage differed significantly among the evaluated races (P ≤ 0.05). Optimal moisture values are 42–46% for TM, 43–50% for TM24, and 44–49% for TM48 (Vázquez-Carrillo et al., 2020). The values ranged from 37.2% in the AA race to 48.3% in the CH. In particular, for HT24, the races AR (45.8%) and PP (46.0%) had the highest values, which were statistically different from the rest, while the races with the lowest values were CO (38.2%) and CH (38.1%). Finally, for HT48, the values were between 46.6% in the PP race and 39.7% in the CO race, which were also significantly different from each other.

The grains that absorbed the most water in the nixtamal (NM) and in the dough (DM) corresponded mainly to varieties with soft or very soft endosperm, while the lowest moisture contents were recorded in materials with harder endosperm. This behavior coincides with what has been described for the nixtamalization process, where water absorption is determined by the pericarp structure and the endosperm texture, which regulate the diffusion of water into the grain and condition the final moisture content of the nixtamal and the dough (Antuna et al., 2008; Santiago-Ramos et al., 2018; in Puebla, López-Morales et al., 2023a). Finally, regarding the moisture content of freshly made tortillas (TM), and tortillas stored for 24 (TM24) and 48 hours (TM48), comparable results were obtained to those reported by López-Morales et al. (2023b) for some races studied in the State of Puebla.

In the 16 native maize races evaluated, solids loss (SL) during the nixtamalization process ranged from 1.9% to 5.2%. This range was slightly higher than that reported by Vázquez et al. (2010) for native maize varieties from the Mexican highlands (2.4–4.4%). However, the values obtained fall within the range described for native maize processed using traditional nixtamalization, where SL depends primarily on endosperm texture, the degree of pericarp detachment, and the intensity of grain washing (Campechano et al., 2012; Valenzuela et al., 2024). The PE race showed the greatest SL, with an average of 5.2%, attributable to its shorter cooking time, set at 25 minutes, and its high-hectoliter weight, averaging 79.2 kg per hectoliter. In contrast, the AR race, which was subjected to a cooking time of 45 minutes, exhibited the lowest SL, with an average value of 1.9%. The Conico race exhibited an SL consistent with that observed in maize with intermediate-hardness endosperm, suggesting that grain texture directly influences pericarp detachment and the amount of dry matter lost during nixtamalization. This has been widely documented in recent studies on the relationship between grain physical properties and nixtamalization efficiency (López-Morales et al., 2023a; Santiago-Ramos et al., 2018).

3.3 Grain color and tortilla texture

The grain color of the Pepitilla and Tuxpeño races were comparable to those found by Vázquez et al. (2003) for those same races. With respect to luminosity (Lu), hue (°Hue), and Chroma, the highest values corresponded to the CH race, with a luminosity of 71.8% and a hue of 85.8. On the other hand, for the chroma parameter, the CA and CO race showed the highest averages, with values of 37.1 and 38.0, which were statistically equal. The Conico race had a high Chroma value (23.8), comparable to that recorded by López-Morales et al. (2021b) for the same race. The similarity in the geographic origin of the samples, collected at the same locality, could explain the similarity between the values. The minimum values were observed in the AR race, with 21.6% luminosity, and for °Hue, the races that had the highest values were EC (34.0°Hue), AR (37.5°Hue), and EN (32.5°Hue), which were statistically similar in this variable. Regarding chroma, the lowest values were 4.6 and 4.3 for the EL and ET races, respectively, which were also statistically equal for this variable. The colorimetric parameters obtained are comparable to the ranges reported by Aragón et al. (2012) and López-Morales et al. (2021b) only from an instrumental point of view, since all the studies used the CIELAB system to quantify grain color. However, these comparisons should be interpreted with caution because those studies predominantly evaluated white-grained maize, while the present work included varieties with a wide diversity of colors, which increases the variability of Luminosity, Hue, and Chroma values.

Regarding the breaking force of freshly prepared tortillas (F), as well as the evaluations after 24 (F24) and 48 hours (F48), the results showed that the maize races with the highest and lowest values for this property were the AC race, with a value of 251.0 g_f, and the PP race, with 90.0 gf, respectively. For F24, the ZC race had the highest value, 472.0 g_f, whereas the lowest value corresponded to the AZ race, with 134.0 g_f. For F48, the AC (310.0 g_f), EC (325.0 g_f), and ZC (340.0 g_f) races had the highest values, which were statistically similar, while the PP race recorded the lowest value, 113 g_f.

On the other hand, the breaking distance of freshly made tortillas (D), as well as the evaluations at 24 (D24) and 48 hours (D48), showed statistically significant differences among all maize races. The results indicated that race AB had the highest D value, at 8.6 mm, while four races exhibited statistically equal and lower values: AA at 5.7 mm, ET at 5.9 mm, EC at 5.8 mm, and EL at 5.7 mm. These last three races were classified as Blue Conical maize, suggesting that tortillas made with this corn are very soft and fresh. For D24, the AC race obtained the highest value, 11.4 mm, whereas the AA race had the lowest value, 6.8 mm. Finally, for D48, the highest values corresponded to the CA race, with 11.7 mm, while the ET race showed the lowest value, 7.5 mm, indicating that tortillas made with this race maintain their softness both when freshly made and after 48 hours.

Fracture resistance and the distance required to break the tortilla increase progressively as storage time advances: freshly prepared tortillas exhibit lower breaking force values than those stored for 48 hours, while greater distances are required to fracture the tortilla. Similar results were reported by Salinas-Moreno and Aguilar-Modesto (2010). The texture of a tortilla, expressed by its breaking strength and distance, is closely related to its moisture content, as higher water content promotes the plasticization of the starch-protein matrix, increasing flexibility and decreasing the force required for breakage. In contrast, moisture loss during storage accelerates starch retrogradation and tortilla hardening. In the present study, tortillas evaluated at 2, 24, and 48 h showed higher moisture contents than those reported in other studies, which could be attributed to differences in the germplasm used and the processing conditions during nixtamalization (Leung et al., 2024; Barajas-Olivares et al., 2025). However, breaking strengths were similar to those described by Gaytán-Martínez et al. (2013), particularly for races with similar physical characteristics. The increase in breaking strength during storage can be attributed to starch retrogradation and moisture redistribution, phenomena widely recognized as the main causes of tortilla hardening over time (Acosta-Estrada et al., 2023; Salinas-Moreno et al., 2024b).

Interestingly, Vázquez-Carrillo et al. (2014) reported substantially higher breaking force and breaking distance values than those found in the present research—nearly double—at all three evaluation intervals. It is important to note that this phenomenon may be attributed to the fact that that study was conducted using improved maize varieties, which, according to previous research, tend to exhibit greater resistance in these tests, especially in freshly prepared tortillas (Jiménez-Juárez et al., 2012; 2015).

4 Conclusions

The 16 native maize races present in the State of Puebla, Mexico exhibit remarkable diversity in grain quality characteristics, including variations in endosperm hardness, ranging from very hard to very soft, as well as differences in grain dimensions and weight. Likewise, differences were observed in nixtamalization and tortilla-making quality, as evidenced by varying moisture levels during the nixtamalization process, as well as differences in tortilla texture and grain color, which included cream, white, yellow, blue, and red shades. The flotation index reflected the full range of grain textures, from very hard to very soft, with a 41% difference in hectoliter weight and a 34% difference in hundred-grain weights between the highest and lowest values. The variables related to moisture during nixtamalization, and dough preparation showed lower variation, ranging between 8 and 6 units, respectively, across all races. Moisture content in freshly prepared tortillas and those measured after 24 (TM24) and 48 (TM48) hours showed similar contrasts in each case: 9 units for TM, 8 for TM24, and 7 for TM48. The Pepitilla race recorded the highest values for the moisture-related variables, whereas the Chalqueño race exhibited lower values. On the other hand, differences among races in terms of Luminosity, hue angle (Hue), and Chroma were highly pronounced, attributable to the diversity of maize grain colors. Specifically, the races showing the greatest contrasts were Arrocillo Red and white-grain Chalqueño, with the most notable differences observed in Luminosity and Hue.

Overall, these findings demonstrate the broad technological and quality diversity of native maize races from Puebla and support their potential use in tortilla production, food diversification, conservation, and maize breeding program.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Ricardo Serna-Lagunes, Universidad Veracruzana, Mexico

Reviewed by: Fulgencio Alatorre-Cobos, Centro de Investigación Científica de Yucatán, Mexico

José Luis Del Rosario Arellano, Universidad Veracruzana, Mexico

Alejandra Ramírez-Martínez, Colegio de Postgraduados, Mexico