Five quinoa genotypes Amarilla Marangani, Amarilla Sacaca, Blanca de junin, Kancolla and Salcedo INIA, were received from the Oman FAO office in 2016 for evaluation of their productivity in Oman. These genotypes reported to be originated in the high altitudes (2000-4000 m) of the Andean valley are characterized by a long growth period, high Saponin content and colored seed [29].

These genotypes were evaluated on three experimental sites located in different agro-ecological regions of Oman viz. Agriculture Research Station, Rumais in South Batinah (23°40'57.00”N, 58° 0'37.08”E), Agriculture Research Station, Alkamil in Sharqia (22°14'18.31”N, 59°10'50.17”E) and Agriculture Research Station, Wadi Hibi, Sohar, in North Batinah (24°28'15.40”N, 56°35'4.47”E) (Fig. 1).

The experiments were conducted consecutively for three years (2016/2017, 2017/2018 and 2018/2019) at layouts on sites composed of sandy loam soils during winter seasons from November to March. The electrical conductivity (EC) and pH of soil and water were recorded for each experimental site (Table 1) before the start of the experiments. The soil at Alkamil research station had the highest pH (8.1) as compared to the soil of Rumais (7.62) and Sohar (7.90) research sites. However, the EC of Sohar soil was the highest (2.5 ds m^-1). The irrigation water EC at Sohar ranged between 1.5 to 2.0 ds m^-1, whereas it was between 0.93 and 1.00 ds m^-1 at Alkamil and between 0.32 and 1.21 ds m^-1 at Rumais. Water pH ranged from 6.71 to 7.20 at Rumais and from 7.5 to 8.0 at Alkamil and Sohar (Table 2). The maximum, minimum and mean temperature conditions of the experiments at three locations viz. Rumais, Alkamil and Sohar during three cropping seasons from October to November are presented in Figs. (2a-2c), respectively.

The experiments were laid in Randomized Complete Block Design (RCBD) with four replications at each location. Seeds of each genotype were planted in six rows of a plot (3 × 4 m) within 2.5 cm soil depth and a planting distance of 35 cm was maintained along with 50 cm distance between the rows. The seed rate was 5 kg ha^-1 (53,000 seed per hectare given that the weight of seed is about 0.3 g). The fertilizers were applied at the rate of 170, 100 and 10 kg N, P and K ha^-1 using urea, triple superphosphate and potassium sulphate as sources, respectively. Half of nitrogen and whole of potassium and phosphorus were applied as basal doses at planting, while the remaining nitrogen was applied after 30 days of emergence. The plants were irrigated for 15 minutes every day in the evening hours until germination (two weeks after planting). The irrigation time was increased by 5 minutes every 15 days until physiological maturity through a well-maintained drip irrigation system. All the crop husbandry practices were followed according to national guidelines to raise a successful crop [30].

The observations on leaf SPAD chlorophyll index were measured one month before harvest (during the flowering stage) in two locations only, whereas days to maturity were recorded whenever more than 95% maturity was attained by each genotype. Leaf chlorophyll was recorded in the field using Chlorophyll Meter SPAD-502 on three sample sites of a leaf prior to the top leaf, which gives a value called SPAD value that corresponds to the amount of chlorophyll present in the leaf sample according to Ling et al. [31]. The remaining characters viz. plant height (cm), number of branches, inflorescence length (cm), and grain yield (ton/ha) were recorded at harvest (average of six plants) at all locations. The plants were harvested at grain maturity (after 30-35 days of flowering) by inspecting grains of five randomly sampled panicles in each plot, and panicles were threshed manually during March. In respect of proximate analysis, grain samples from each genotype were analyzed for moisture, protein, fat and fiber contents according to ISO 712:2014, ISO 1871:2015, ISO 11085:2015 and ISO 5498:2014, respectively [32] by outsourcing with Oman Folour Mill.

Fig. (1). Map of Oman showing three experimental locations (Rumais, Alkamil and Sohar Research Stations).

Fig. (2a). Maximum, mean and minimum temperatures in three cropping seasons during winter from October to March in 2016-2017, 2017-2018 and 2018-2019 at Rumais.

Fig. (2b). Maximum, mean and minimum temperatures in three cropping seasons during winter from October to March in 2016-2017, 2017-2018 and 2018-2019 at Alkamil.

Fig. (2c). Maximum, mean and minimum temperatures in three cropping seasons during winter from October to March in 2016-2017, 2017-2018 and 2018-2019 at Wadi Hibi, Sohar.

Data on yield and other characters were subjected to multi-factor analysis of variance (ANOVA) according to Gomez and Gomez [33], where genotype, location, and year were the factors used in the analysis using GenStat Statistical Package v12 [34].

Uniform plant height plays a significant role in determining the quinoa suitability for mechanical harvesting [37]. Among the locations, Sohar Research Station (SRS) had the highest mean plant height (123.27 cm) in comparison with that at Rumais, which had the shortest quinoa plants (89.6 cm), whereas among the genotypes, Amarllia Sacaca genotype was the tallest (120.11 cm) followed by Salcedo INIA (110.19 cm), and Blanca de junin (108.19 cm) (Table 2). Winter season of 2017/2018 witnessed the highest mean plant height (148.10 cm) at Sohar Research Station and lowest at Rumais (80.00 cm) (Table 3). The shorter plant height at Rumais can be attributed to lower prevailing mean temperatures ranging from 20^oC-29^oC as compared to 22^oC-31^oC at Alkamil and 19^oC-32^oC at Sohar during the season (Fig. 2a). This is attributed to slow enzymatic activity that causes slow and stunted plant growth [7, 37]. (Adams et al., 2001; Maliro et al., 2017). During the winter season of 2017/2018, Amarllia Sacaca was the highest in plant height (134.92 cm), followed by Salcedo INIA (117.33 cm) during the winter season of 2016/2017, whereas the lowest was obtained by Amarllia Maranagani (95.50 cm) during the winter season of 2017/2018 which did not significantly differ from Kancolla (95.58 cm) during the winter season of 2016/2017. (Table 3). The overall response of crop irrespective of varieties to the temperature climate in respect of plant height indicated quinoa attained the mean height of 107.49 cm with a range from 89.60 cm (Rumais) to 123.27 cm (Sohar). These observations are in line with the results of other researches across the contemporary world [39-42].

The results revealed highly significant differences among locations, year and their interaction (P<0.001), whereas there were no significant (p=0.438) differences found between genotypes in respect to a number of branches, in which all genotypes produced an average of 18 branches per plant except Amarllia Sacaca which produced highest of 20 branches (Table 4). Winter season of 2017/2018 was the highest in an average number of branches (22 branches per plant). However, the highest number of branches were produced by the plants at Rumais (26 branches per plant), which was significantly different from that produced in Sohar (16 branches per plant) and Alkamil (13 branches per plant). Irrespective of genotypes, years and locations, quinoa showed its ability to produce on an average of 18 number of branches. These observations are consistent with the results of earlier workers [41, 43-45].

In 2014, Riccardi and coworkers demonstrated the use of non-destructive evaluation of chlorophyll contents in the leaves of quinoa leaves using a chlorophyll meter that provides SPAD values which are directly proposal to chlorophyll contents [46]. The analysis of variance showed significant differences in genotypes (p<.05), and highly significant (P<0.001) in the year and the interaction of year×location in respect of chlorophyll content, whereas there was the location (p=0.085) and its’ interaction with genotype (p=0.076) was not significant (Table 4). Amarllia Sacaca and Blanca de junin genotypes, which had the highest Chlorophyll content (47 SPAD), were not significantly different, whereas Kancolla and Salcedo INIA contained the lowest Chlorophyll (43 SPAD). Quinoa chlorophyll at Sohar for 2016/2017 winter season was the highest (52 SPAD) followed by Rumais during 2017/2018 (48 SPAD), whereas the lowest was recorded at Sohar and Rumais during 2018/2019 (39 and 40 SPAD, respectively. The grand mean of SPAD over varieties, location and year, was found to be 44.47. The SPAD values reported in the present studies are in line with previous reports [8, 47, 48].

The statistical analysis did not show any significant (P>0.05) effect of genotypes and its’ interaction with year and location in respect of inflorescence length. The location was significant (p=0.001) in respect to inflorescence length (cm). The inflorescence length obtained in Rumais (41.87 cm) was the highest as compared to that in Sohar (32.38 cm), followed by that in Alkamil (21.63 cm), as displayed in Table 5. The interaction between location and year was highly significant (p<0.001). Inflorescence length during the winter season of 2018/2019 at Rumais was the highest (63.55 cm), followed by that obtained during 2017/2018 at Sohar (47.45 cm), whereas the lowest (14.85 cm) was obtained during the winter season of 2018/2019 at Sohar (Table 5). The grand mean of inflorescence length over gentypes, locations and years was found to be 31.96 cm. These results are similar to the reports of earlier researches that indicated inflorescence length in the range of 13.86 cm to 35.32 cm [7, 45], made by Maliro et al. 2017 (29.33 cm to 35.32 cm) and Biswas and Tanni, 2017 (13.86 cm to 29.62 cm) in their studies in quinoa.

The results on days to maturity of five quinoa genotypes are presented in Fig. (3). Significant differences (P<0.05) were found among genotypes in respect to days to maturity. Amerllia Maranagani genotypes showed the shortest period to maturity (84 days), which is significantly different from other genotypes (Fig. 3). The highest days to maturity were noted in Salcedo INIA (149 days), followed by Kancolla (124 days), Amerllia Sacaca (116 days) and Blanca de junin (103 days). The grand mean of days to maturity over the years for a location (Rumais) was found to be 115.2. These are in line with the findings of Spehar and Santos. 2005 (80-126 days) and Tan and Temel, 2018 (119- 141 days) among the genotypes studied. Belmonte et al. (2018) observed days to flowering ranging from 53.8 to 57.7 among the varieties investigated. It is reported that the physiological maturity of quinoa can be achieved within 70-90 days after flowering [49, 50].

Fig. (3). Means of days to maturity for five quinoa genotypes.

Genotype × Location × Year interaction was highly significant (p<0.001) for grain yield (t ha^-1). The genotypes Amarllia Maranagani, Salcedo INIA and Amarllia Sacaca produced higher grain yields of 4.04, 3.99 and 3.95 t ha^-1, respectively (Table 6). The grain yields obtained from these genotypes were within the range of mean grain yield (0.46 kg m^-2 equivalent to 4.6 t ha^-1) of five accessions evaluated at the International Center for Biosaline Agriculture (ICBA) during 2012 [38]. (Rao and Shahid, 2012). Locations had a highly significant effect (p<0.001) in which the highest grain yield was obtained at Rumais (5.40 t ha^-1) followed by Sohar (4.08 t ha^-1). Alkamil Research Station location was the lowest, with an average grain yield of 2.01 t ha^-1 (Table 6).

Winter season of 2017/2018 was the highest in grain yield (7.34 t ha^-1) at Rumais, which is significant from a season of 2018/2019 (5.95 t ha^-1) at the same location, whereas the lowest grain yield (1.62 t ha^-1) was found during winter 2018/2019 at Alkamil (Table 6). Among 25 genotypes tested in Brazilian Savannah in 2005, grain yield was recorded between 1.00 and 2.5 t ha-1 [51]. Relatively, the grain yields were stable during the three seasons at Sohar (p>0.05), while they were found significantly different (p<0.05) in the other two locations, with higher preferential of Rumais and Sohar in comparison to Alkamil Station. In general, the grand mean yield of quinoa irrespective of genotypes, locations and years was found to be to the extent of 3.83 t ha^-1. This mean yield was similar to yield levels obtained in Egypt (3.87 ha-1) and Lebanon (4.5 t/ha) and other countries of the world [25, 52].

The proximate analysis of quinoa genotypes are illustrated in Figs. (4-E). The moisture contents of the genotypes varied significantly between 8.75% (Salcedo INIA) to 9.48% (Amarilla Sacaca) with mean moisture of 9.17% (Fig. (4A), which are similar to earlier findings [53]. This clearly indicates that these genotypes have about 88% and higher dry matter content indicating their higher storage ability in marketing [4, 54]. In respect of protein content, only two genotypes Amarilla Marangani (17.49%) and Salcedo INIA (17.07%), had significantly higher protein content than the remaining three genotypes (15.53% to 16.3%) (Fig. 4B). The mean protein content of the genotypes was found to be 16.5%. These figures are in line with the protein % reported by earlier workers in the range from 14 to 20% [53-56]. In respect of fat content, the present studies indicated an average fat content of 4.0% in quinoa genotypes with a range from 3.57 to 4.55% (Fig. 4C). These are within the range of fat content (2% to 10%) reported earlier by Valencia-Chamorro in 2003 [49]. However, recent research reports indicated a higher range of fat contents from 5.3% to 7.8% [56-59]. In respect of fiber contents, the genotypes varied from 2.83 to 3.66%, with an average of 3.4% (Fig. 4D). These figures are in line with the fiber contents 4.2% reported in FAO-INFOODS Data base [59]. In the present study, the highest ash contents were found in a range of 2.87% (Amarilla Sacaca) to 3.74% (Amarilla Marangani) (Fig. 4E) with mean ash content of 3.4%. These results in the range of ash contents from 2.3% reported earlier [57, 58]. The ash content determines the amount and type of minerals in food; and has its own significance as the amount of minerals can determine the physiochemical properties of foods like quinoa [53, 60].

Fig. (4). Proximate analysis of five quinoa genotypes.