The quinoa (Chenopodium quinoa Willd.) Is native to Andean countries and was domesticated about 3,000 to 5,000 years ago (Mujica, Izquierdo, & Marathee, 2001). It is considered as a pseudocereal or pseudo grain, since its morphology and chemical composition is similar to cereals (Bazile, Bertero, & Nieto, 2014). The importance of this grain lies in its quality as food, the use of the complete plant and its adaptation to agroecological conditions (Mujica & Jacobsen, 2006). It is an important source of protein, amino acids, minerals and vitamins; in addition, it contains polyphenols, phytosterols and flavonoids with possible nutraceutical benefits (Abugoch-James, 2009; Bergesse et al., 2015).
In 1996, quinoa was classified by the Food and Agriculture Organization (FAO) as one of mankind’s most promising crops, not only because of its great beneficial properties and multiple uses, but also by considering it as an alternative to solve the serious problems of human nutrition (FAO, 2011).
The degree of relationship between protein, starch and other components deposited in perisperm cells (Figure 1) varies among quinoa varieties (Apaza, Cáceres, Estrada, & Pinedo, 2013). These relationships define how hard or soft the perisperm is between one seed and another. Differences in grain hardness are of great importance as they significantly influence the determination of physiological maturity at harvest time (Bazile et al., 2014), the physical properties of the seed, its milling and industrialization (Bergesse et al., 2015; Salinas-Moreno & Aguilar-Modesto, 2010). In this regard, Taverna, Leonel, and Mischan (2012) reported that there is a close relationship between the hardness and the quality of the flour in quinoa.
Grain hardness refers to the resistance of grain to a mechanical force, or to the energy required to reduce the structures of the grain into flour or semolinas (Ballón & Coca-Cadena, 1989). There are several methods to determine grain hardness, and depending on the characteristics of the grain some are more suitable than others.
Some researchers have used scales based on milling time (Ballón & Coca-Cadena, 1989); others, using a texturometer, measured the force required to break the material (Bergesse et al., 2015; Taverna et al., 2012). López, Guzmán, Santos, Prieto, and Román (2005) indicate that the texturometer measures only the hardness of the grain surface, while other procedures can measure it in a more comprehensive way.
Salinas, Martínez, and Gómez (1992) analyzed seven methods to obtain hardness in maize grains, namely endosperm texture, pearling index, flotation index, density, infrared reflectance, hectoliter weight and milling time, and they determined that the most appropriate was the flotation index. The hectoliter weight of a sample is an indirect way of determining its hardness. Salinas-Moreno and Aguilar-Modesto (2010) reported in maize that the greater the grain hardness the greater the hectoliter weight and the lower the flotation index. Peña (2003) determined that wheat grain hardness is related to the amount of insoluble protein, this being of great influence in wheat processing.
There is little research on the physical characteristics of quinoa seed; therefore, the aim of this study is to develop an experimental methodology to indirectly determine the hardness of quinoa (Chenopodium quinoa Willd.) seed. In addition, moisture and hectoliter weight were evaluated, since these characteristics are directly related to hardness.
Materials and methods
The research was conducted in the Department of Agroindustrial Engineering’s Cereals Workshop at Autonomous Chapingo University. The samples analyzed were: Blanca Canadá, Blanca de Tlachichuca (BT), Roja de Tlachichuca (RT), Negra de Tlachichuca (NT) and Ontifor, all grown in Mexico, the first in Tula, Hidalgo and the rest in Chapingo and Puebla. The variables evaluated were moisture (%), hectoliter weight (kg·hL-1) and hardness (%). The tests were performed in triplicate, except for the moisture one that was made in duplicate.
Moisture was determined using a Sartorius™ model MA37 electronic moisture analyzer, with 7 g of quinoa seed placed on each dish. Once the initial weight condition was met, the measurement was started. The approximate analysis time varied from 20 to 25 min. The result was expressed as a percentage.
A Seedburo Equipment Co. scale was used for hectoliter weight. The sample was dropped into the upper cone (285 mL) of the apparatus. Subsequently, the sample was scraped with a wooden ruler with rounded edges to level off the container in three zigzag movements. The container with the sample was weighed and the result was expressed in kg·hL-1.
In order to have seed of homogeneous size, several tests were carried out to select the appropriate sieves for the application of the methodology to be proposed.
The initial test to standardize seed size was performed with 200 g of quinoa and No. 10 (2 mm opening), 14 (1.41 mm opening) and 18 (1 mm opening) mesh sieves, and a retention tray. The sieving was conducted mechanically with a Montinox® sieving machine, performing circular movements (homogeneous) for 3 min. Subsequently, on an Adventurer™ Pro model AV2101 analytical balance, the quinoa retained on each sieve was weighed. From the mesh with the highest retention percentage, 30 g were weighed and processed in a Mr. Coffee® coffee bean grinder for 2 s. The ground sample was sieved in No. 14, 18, 20, 24 and 30 meshes, with a tray underneath, for 3 min using the same Montinox® sieving machine. Finally, the retention percentage of each sieve was obtained.
To determine grain hardness, a hedonic scale was generated (Table 1) with as many categories as possible.
Analysis of variance was conducted using a completely randomized design, Tukey’s range test (P ≤ 0.05) was performed using the Statistical Analysis System package (SAS, 1994) and the Pearson correlation coefficient was obtained with the factors hectoliter weight and sieve retention percentage.
Results and discussion
From the initial seed selection, 80 % retention was obtained in 14 mesh (from 1.41 to 2mm), this being the grain size used to perform the analyses.
The average moisture obtained was 8.88 ± 0.27 %. Table 2 shows that Ontifor had the highest moisture content (9.15 %), while NT showed the lowest (8.72 %). Considering that the maximum moisture content should be 12 % (Instituto Ecuatoriano de Normalización [INEN], 1988), the results show that the varieties used comply with the standard.
Low moisture content gives seeds longer shelf life and helps prevent insect attack (Bazile et al., 2014).
|Variety||Measurement 1||Measurement 2||Average|
The mean value of this variable was 66.17 ± 4.07 kg·hL-1. Table 3 shows that Ontifor had the highest hectoliter weight (70.63 kg·hL-1), coinciding with that reported in moisture. For its part, the BT sample had the lowest value (62.60 kg·hL-1). According to Ecuadorian technical standard INEN 1 673 (INEN, 1988), the minimum hectoliter weight should be 62 kg·hL-1 to be considered as a first quality grain. The analyzed samples meet this requirement.
|Variety||HW1 1||HW 2||HW 3||Average|
The great variability in the results could be due to the extensive genetic diversity conserved by producers.
The varieties with the highest hectoliter weight were the same ones that had the highest moisture, coinciding with the results reported by Coşkuner and Karababa (2007) and Vilche, Gely, and Santalla (2003) who found a linear relationship between moisture and hectoliter weight.
The retention percentage in each sieve showed significant statistical differences (P ≤ 0.05). No. 14 mesh presented 55.24 % retention (Table 4), thus being the most important to evaluate hardness.
|Sieve Number||Retention (%)|
|Retention tray||22.17 b|
Table 5 shows the retention percentage of each sample per mesh. The sieve with the greatest amount of sample was the 14, conserving, on average, more than 50 %; therefore, the determination of hardness was focused on this sieve.
|Mesh number||Blanca Canadá||BT||RT||Ontifor||NT|
With the selected sieve (No. 14), a hedonic scale relating retention percentage (non-fractured seed) to grain hardness was created. Table 1 shows the category assigned to each retention percentage range.
According to the results, Blanca Canadá (47.67 %) and NT (42.33 %) are slightly soft grain, while BT (54.89 %), RT (55.22 %) and Ontifor (51.11 %) are slightly hard (Table 5). The BT, RT and Ontifor samples do not present significant statistical differences (P ≤ 0.05), being those of greater hardness, that is, a higher retention percentage in 14 mesh (Table 6).
|RT||55.22 az||Slightly hard|
|BT||54.88 a||Slightly hard|
|Ontifor||51.11 ab||Slightly hard|
|Blanca Canadá||47.67 bc||Slightly soft|
|NT||42.33 c||Slightly soft|
It was expected that the variety with the greatest hectoliter weight and moisture percentage would also present the greatest hardness; however, although Ontifor was classified as “slightly hard” it was not the hardest. On the other hand, the NT variety was the softest, coinciding with the moisture result but not with that of the hectoliter weight.
To determine the hardness of quinoa grain, Ballón and Coca-Cadena (1989) created five categories (soft, semi-soft, semi-hard, hard and very hard) based on the time required to mill the grain. Their results show that most of the varieties used were soft, followed by hard and semi-soft. The variation in the results is due to the different varieties used.
The correlation between hectoliter weight and sieve retention percentage was 0.638. This analysis allowed deducing the proportion at which an increase in the hectoliter weight of the grain will increase the retention percentage in the sieve, thereby resulting in a harder grain.
The methodology used is an efficient way to indirectly quantify the hardness parameter. In the case of quinoa, the recommended sieve size is No. 14, since it was the one that presented the greatest particle retention. This procedure may also be viable in other small grains (such as amaranth, quiwicha and chia, among others), although further testing is recommended to standardize and obtain better results. The fact that the hardness of the samples was different was to be expected, since, as in other grains, each variety has its own characteristics and can therefore be used for different purposes. The correlation between hectoliter weight and the mesh retention percentage (grain hardness) is not very high (r2 = 0.6384), which may influence the moisture of the grain.