A 3-year study (2016–2018) was conducted on two no-till farms; one in Genesee, ID (Jensen farm) (46.611134° N Lat., −117.009925° W Long.) and the other in Almota, WA (Aeschliman farm) (46.791288° N Lat., −117.438473° W Long.). Both Genesee and Almota soils are part of the Palouse series soil and are classified as silt loams (Donaldson, 1980; Frazier and Cheng, 1989). Genesee received total annual precipitation of 565 mm in 2016, 657 mm in 2017, and 446 mm in 2018 and Almota received total annual precipitation of 455 mm in 2016, 535 mm in 2017, and 392 mm in 2018. The average maximum day temperature recorded during the growing season in Genesee were 23.47°C in 2016, 24.96°C in 2017, and 24.14°C in 2018. The total precipitation recorded during the growing seasons 2016, 2017, and 2018 was 68.06, 65.45, and 86.36 mm, respectively. In Almota the average maximum day temperature recorded during the growing season 2016, 2017, and 2018, was 25.55, 26.23, and 25.90°C, respectively. The total precipitation recorded during the growing seasons 2016, 2017, and 2018 was 55.12, 61.96, and 76.75 mm, respectively. Most of this precipitation occurred during the winter and early in the spring (AgWeatherNet, 2021; Weather Underground, 2021).

Previous crops grown in the field where experiments were conducted are as follows: in Genesee, winter wheat (WW), winter peas, and WW in 2016, 2017, and 2018, respectively. In Almota, WW in 2016, 2017, and 2018.

Statistical analysis was performed using the statistical software SAS 9.4 University Edition (SAS Institute IN., Cary, NC). The mixed-effects methodology was used to analyze data and was performed using a mixed model with PROC GLIMIX. Model assumptions were verified using marginal and conditional studentized residuals from PROC MIXED and studentized residuals from PROC GLIMMIX. A logarithmic transformation was used for yield and plant height to satisfy the homogeneity of variance assumption. Contrasts were calculated to show which of the two varieties differed by N and seeding rates. Means and least significant differences for all the data in the table were reported. The statistical significance level was set at α = 0.05 unless otherwise noted. ANOVA was performed with all factors using a mixed model with PROC GLIMIX. Pearson's correlation coefficients were calculated based on mean trait values and used to estimate phenotypic relationships between traits of interest. Fisher's LSD test was used to compare treatment means.

Nitrogen did not have a significant effect on β-glucan across all 3 years and at both locations (Table 1 and Figure 1). However, there was a significant N × variety interaction for β-glucan (Table 1). There was a significant difference between varieties for β-glucan content across all five N rates in Almota and Genesee in all 3 years of the study (Table 1) and grain β-glucan content in the varieties varied by year (Table 2). When comparing varieties' β-glucan content, Julie consistently had significantly higher β-glucan content than Havener in both locations-−1.6% higher in Almota and 1.5% higher in Genesee (Table 3).

Nitrogen had a highly significant effect (p < 0.001) on protein in Almota in 2017 and 2018 (but not 2016) and in Genesee in 2016 and 2018 (but not 2017) (Table 1). Grain protein responded positively to increasing N rates in both varieties, where the highest protein content was observed in response to 162 kg N ha⁻¹ (Table 2 and Figure 1). The average protein content across all N rates, varieties, and years was 10.9% at both Almota and Genesee (Table 2 and Figure 2). There was a difference in protein content between varieties across the N rates in Almota and Genesee (Table 3). When comparing varieties' protein content, Julie had an average of 3.5% higher protein content than Havener across all years and locations (Table 3).

In all 3 years, there was a highly significant N × variety interaction for yield in Almota and Genesee (p < 0.0001) (Table 4). N rates caused a difference in grain yield across Havener and Julie and 3 years in Almota and Genesee (Table 4). In all years at both locations, there was an increase in grain yield with increasing N rates across both Havener and Julie (Table 5). The higher N rates increased grain yield of both varieties with the highest grain yield being observed in response to 129 kg N ha⁻¹ in half of the six site-years, while the maximum yield occurred with 162 kg N ha⁻¹ at both locations in 2016 and the highest yield at the 2018 Almota site was with 95 kg N ha⁻¹ (Tables 5, 6). Across the five N treatments in all 3 years, the grain yield in Genesee was 336 kg N ha⁻¹ higher than the grain yield in Almota (Table 7).

When comparing varietal grain yield across years, location, and N treatments, the results showed a difference in grain yield between Havener and Julie (Tables 8, 9). Havener was consistently higher yielding than Julie with average of 806 kg N ha⁻¹ higher grain yield than Julie (Table 9).

In Almota, there was a significant seeding rate × variety interaction for yield in 2016 and 2018 (p < 0.0001) (Table 4). In Genesee, there was a seeding rate × variety interaction for yield in 2017 and 2018 (Table 4). Increased seeding rates 250, 310, and 375 seeds/m⁻² increased grain yield 3,440, 3,453, and 3,616 kg N ha⁻¹, respectively in all the varieties in all years and locations (Table 11).

Nitrogen did not have an impact on plant emergence in Almota or Genesee in 2016 and 2018 (Table 4). However, in 2017, N had a significant impact on plant emergence in both locations (Table 4). Across all N treatments in all years, emergence in Almota was 3% higher than in Genesee (Table 7). In Almota, there was a highly significant difference (p < 0.0001) in emergence rates between varieties in 2016 and 2017, but not in 2018 (Table 4). In Genesee, there was a highly significant difference (p < 0.0001) in varieties' emergence rates in 2016 and 2018 (Table 4). Across all N rates and two locations, Havener had on average 11% higher emergence rate than Julie (Table 9). Seeding rate only had a significant effect on emergence at Genesee in 2016 and 2017 (Table 4).

Nitrogen positively affected plant height in 2016, 2017, and 2018 in Almota (p < 0.0001); in Genesee, N positively affected plant height in 2016 and 2018 (Table 4). Increasing N rates increased plant height across both varieties in all years and locations; the maximum plant height of 88 cm was recorded at both 129 and 162 kg N ha⁻¹ (Table 6). Plant height did not differ between varieties with the exception of Almota in 2017 where Havener was significantly taller than Julie (Table 8). When comparing the plant height of varieties across all N rates, years and locations, Havener was taller than Julie (Table 9). Seeding rate did not have a significant effect on plant height (Tables 4, 11).

Analysis of variance revealed that N rates affected days to heading in 2017 at both Almota (p = 0.02) and Genesee (p ≤ 0.0001) (Table 4). However, there was not a clear trend to suggest that the N rate has an influence on days to heading. Over the course of the study, Almota was 2 days earlier than Genesee for days to heading (Table 7). In five of the six site-years, there was a significant difference in days to heading between varieties (Table 4). When comparing varieties' days to heading across all N and seeding rates in both Almota and Genesee, Havener reached heading 3 days earlier than Julie (Tables 8, 9). Seeding rates did not have any effect on days to heading (Tables 10, 11).

Nitrogen rate affected days to maturity at both Almota and Genesee in 2016, 2017, and 2018 (Table 4). Increased N rates increased days to maturity of all the varieties; across varieties in all years and locations, the highest number of days to maturity was observed in response to 162 kg N ha⁻¹ (Table 6). Across all N treatments in all 3 years, the barley varieties in Almota matured 5 days earlier than in Genesee (Table 7). When comparing varieties' days to maturity across all five N rates in both Almota and Genesee, Havener matured 3–8 days earlier than Julie (Table 9). Seeding rate did not have an effect on days to maturity (Tables 10, 11).

The effect of different N rates on test weight was significant in Almota in 2017 (p < 0.0001) and at Genesee in 2016 (p = 0.01) and 2018 (p < 0.0001) (Table 4). Across all N treatments in all years, Almota was 3 kg hL⁻¹ higher in test weight than Genesee (Table 7). Increasing N rates usually resulted in decreased test weight; across all locations and years the 0 N treatment resulted in the highest test weight, and the 162 N treatment resulted in the lowest test weight (Tables 5, 6). There was a difference between Havener and Julie in test weight across all five N rates in both Almota and Genesee; Havener was 2 kg hL⁻¹ higher in test weight than Julie (Tables 8, 9). There was no difference in test weight caused by seeding rates across all 3 years at Almota and Genesee (Tables 10, 11). For all 3 years in Almota, there was a N × variety interaction for test weight (Table 4). In Genesee there was a N × variety interaction for test weight in 2017 (p = 0.04) and 2018 (p < 0.0001), but not in 2016 (Table 4).

Nitrogen did not have a significant impact on percent plump kernels at Almota in any of the 3 years (Table 4). However, N had a significant impact on percent plump kernels 2017 at Genesee (p < 0.0001) but not in 2016 or 2018 (p > 0.05) (Table 4). Increased N rates resulted in decreased percent plump kernels only in 2017 in Genesse (Table 5). Across all N treatments in all years, Almota was 10% higher in percent plump kernels than Genesee (Table 7). There was a difference between varieties in percent plump kernels across all 3 years (Table 4). When comparing varieties' percent plump kernels across all five N rates in both Almota and Genesee, Julie was 14% higher in percent plump kernels than Havener (Table 9).

Seeding rate had a significant effect on percent plump kernels in 2018 at both Almota (p < 0.0001) and Genesee (p = 0.03). However, seeding rates did not affect percent plump kernels in either location in 2016 or 2017 (Table 4). Across all locations and years, increased seeding rates resulted in decreased percent plump, the 250 seeds/m⁻² treatment resulted in the highest percent plump kernels and the 375 seeds/m⁻² treatment resulted in the lowest percent plump kernels (Table 11).

Nitrogen rate had a significant impact on percent thin kernels in Almota in 2017 (p = 0.03) (Table 4). There was a difference between locations for percent thin kernels across all N treatments in 3 years and barley grown in Genesee had 2% more thin kernels than Almota (Table 7). There was a difference between varieties in percent plump kernels across all 3 years in Almota and Genesee (Table 4). When comparing varieties' percent thin kernels, Havener was 6% higher in percent thin kernels than Julie (Table 9).

Seeding rates had an effect on percent thin kernels in 2018 at Almota (p = 0.01), but not in 2016 or 2017. There was no significant difference in percent thin kernels associated with the seeding rate at Genesee (Table 4). Increased seeding rates resulted in increased percent thin kernels; across all locations and years, the 375 seeds/m⁻² treatment resulted in the highest percent thin kernels and the 250 seeds/m⁻² treatment resulted in the lowest percent thin kernels (Table 11).

Across all the varieties in all years, no correlation was found between grain yield and other measured traits in Almota; however, in Genesee, grain yield had a strong negative correlation with test weight (r = −0.94; p = 0.02) and a strong positive correlation with plant height (r = 0.90; p = 0.04) (Table 12). When measuring between emergence rates and other traits, the results showed that in both Almota and Genesee, there was no correlation between emergence rate and other measured traits (Table 12).

For plant height in Almota, there was a strong negative correlation between plant height and test weight (r = −0.90; p = 0.03), and a strong positive correlation between plant height and days to maturity (r = 0.99; p = 0.0001) (Table 12). In Genesee, plant height had a strong negative correlation with percent plump kernels (r = −0.94; p = 0.02) (Table 12). Days to heading did not correlate with any other measured traits (Table 12). However, in Almota, there was a strong negative correlation between days to maturity and test weight (r = −0.95; p = 0.01). In Genesee, there was a strong negative correlation between days to maturity and percent plump kernels (r = −0.90; p = 0.04) (Table 12). No correlation was found between percent thin kernels and other traits in either location (Table 12).

Cultivar, environmental factors, and nitrogen fertilization have been reported to affect yield, chemical composition, and grain quality of barley (Widdowson et al., 1982; Åman and Newman, 1986; Knudsen et al., 1987; Conry, 1994; Zhang et al., 2001; Anker-Nilssen et al., 2008; Choi et al., 2020). β-glucan and protein are major components of barley (Åman et al., 1985). Given its unique functional and nutritional properties, β-glucan is the most important fiber component in barley. It is found throughout the grain endosperm and germ, but germ β-glucan has higher dietetic value (Fincher, 1975; McNeil et al., 1975; Bacic and Stone, 1981; Salomonsson et al., 1984). β-glucan content is associated with both genotype and environmental factors (Gill et al., 1982; Stuart et al., 1988; Güler, 2003; Goudar et al., 2020). Similarly, in this study there were differences among varieties in grain β-glucan content; Julie had higher β-glucan content than Havener in all years.

Nitrogen fertilizers are usually employed to increase grain yield. However, Brunner and Freed (1994) reported that β-glucan was affected by N levels, location, and years. The grain β-glucan content reported in our study was affected by location (Figure 2). Additionally, the nitrogen treatment affected the β-glucan content of Havener but not of Julie across all sites and years (Table 13). Our findings are in agreement with those of Sørensen and Truelsen (1985), Henry (1986), and Oscarsson et al. (1998), who reported that barley β-glucan content increased with increasing nitrogen rate. Brunner and Freed (1994), Thomason et al. (2012), and Aghdam and Samadiyan (2014) also found that higher nitrogen levels generally increased grain β-glucan in barley and oats. However, Gohl et al. (1978); Aastrup (1979); Hesselman et al. (1981), and Zhang et al. (2001) argued that the β-glucan content has not been consistently influenced by nitrogen, but rather that it varies across locations, typically decreasing under moist growing conditions and high temperature.

Differences were found in β-glucan content between Almota and Genesee, where Genesee was 0.4% higher than Almota in β-glucan content across all varieties and years. This is in agreement with Åman (1986), Lehtonen and Aikasalo (1987), Åman et al. (1989), Zhang et al. (2001), and Goudar et al. (2020) who reported significant differences in barley β-glucan content across locations. They reported that drought conditions may be related to increased β-glucan concentration. Peterson (1991) observed a significantly higher mean β-glucan concentration for oat and barley cultivars grown under dryland vs. irrigated conditions at two locations in Idaho. Ehrenbergerová et al. (2008), reported higher precipitation during the flowering time and grain filling period and lower temperatures during the flowering time have negative effects on concentration of β-glucans. Conversely, drier and warmer weather in enhanced the content of β-glucans. This is also in line with Tiwari and Cummins (2009), who reported that temperature and precipitation were the most important climatic factors affecting β-glucan content. Because β-glucan accumulates late in grain development, stress from temperature and precipitation factors can cause the early end of grain development, and therefore, reduced β-glucan concentration in grains (Tiwari and Cummins, 2009). Thus, conditions favorable to endosperm development would increase the accumulation of β-glucan in the grain (Zhang et al., 2001; Goudar et al., 2020).

Generally, protein content increases with increasing nitrogen supply (Sørensen and Truelsen, 1985; Conry, 1994; Therrien et al., 1994). Dubetz (1961) and Hutcheon and Paul (1966) found that moisture and N availability were two important factors affecting the protein content of wheat. Soper and Huang (1963) and Nuttall et al. (1971) also demonstrated that percent protein and total protein per unit area could be predicted by measuring available soil N plus added N fertilizer. In our study, grain protein responded positively to increasing N rates in both barley varieties, where the highest protein content was observed in response to 162 kg N ha⁻¹. Considering the combined results across all six site-years, an increase in N rates resulted in increased protein content. Janković et al. (2011) also found that increased nitrogen N rates significantly affected the total protein content in malting barley grain, where an increase in the total nitrogen rate (50, 70, 90, and 110 kg N ha⁻¹) increased the protein content (12.6, 12.8, 12.9, and 13.5%).

Grain yield and quality traits of spring barley can vary greatly due to growing conditions (Leistrumaite and Paplauskiene, 2005). Mouchova et al. (1996) found that the efficiency of applied nitrogen fertilizers and their loss are influenced by the interaction of many factors, such as type of soil, crop, agricultural practice, including the rate, the form, the timing of fertilization and year-to-year weather patterns. The amount of nitrogen that a barley crop needs to maximize yield and quality depend on seasonal conditions, soil type, rotational history of the soil, and yield potential. Nitrogen is needed for early tiller development, without which the yield potential of the crop is limited. As expected, increased N rates increased grain yield of all the barley varieties in all years and locations; the highest grain yield observed in response to 95, 129, and 162 kg N ha⁻¹. These results are consistent with previous reports by Singh et al. (1992) and Alazmani (2014) who reported increased wheat and barley grain yield resulting from increased nitrogen fertilization in India and Iran. Ryan et al. (2009) reported that nitrogen increased leaf area, tiller formation, leaf area index and leaf area duration and that this increase led to much greater production of dry matter and barley grain yield in Morocco. However, Singh et al. (1992) reported that increasing N fertility beyond a certain limit ultimately decreased wheat grain yield in India. In this study, Havener had a higher average grain yield (806 kg N ha⁻¹) than Julie across all N rates, years and locations. It is possible to select barley cultivars and different growing locations, but environmental factors are much more difficult to control. The interaction between predetermined and environmental factors could play an important role in the yield and quality of barley.

Nitrogen fertilizer application, while often required to increase grain yield and protein composition of the barley grain may also reduce physical grain quality traits. Increased N rates resulted in decreased test weight. Across all site-years, the 0 N treatment resulted in the highest test weight, and the 162 N treatment resulted in the lowest test weight. However, the percent plump kernels and percent thin kernels remained constant despite the change in N rates. Jackson et al. (1994) and Bleidere et al. (2013) also found that a decline in test weight of naked spring barley, resulted from increasing N rates. Our results also indicated no correlation between days to heading and days to maturity. Tsenov (2009) found that the growing conditions significantly affected the time to heading and physiological maturity of wheat, especially under heat stress. Tsenov (2009) also reported that the period between heading and physiological maturity varied according to the temperatures in respective years. Higher temperature sums should, in principle, accelerate the date to heading and maturity. Moreover, in warmer years, highest degree maturation is accelerated, which increase the correlation between days to heading and days to maturity, while in cooler years combined with more available soil moisture during grain filling, maturation is retarded, i.e., the correlation between days to heading and days to maturity decreases. Bullrich et al. (2002) and Blake et al. (2009) reported a lack of correlation between days heading and days to physiological maturity, and that the period of grain filling was markedly less affected by the temperature, moisture, and light than time to heading, but rather, was dictated exclusively by genotype.

The seeding rates currently used for barley vary between agroclimatic zones. In the drier areas with a mean annual precipitation <250 mm, the recommended rate of 257 plants m⁻² (Turk, 1998) is much lower than for semi-arid areas that receive a mean annual precipitation of 250–500 mm (Munir, 2002). Higher seeding rates are recommended in areas with a higher mean annual precipitation in part because they promote greater crop competition against weeds (Kirkland, 1993). In this study, only 1 in 6 site years showed an increase in yield due to seeding rate increases; the highest grain yield was observed in response to the 375 seeds/m⁻² treatment. According to Kirkland (1993), yield increase observed alongside increased seeding rates is a function of more spikes being produced as a result of more plants being established. In Kirkland (1993), grain yields increased as seeding rates increased, with maximum yields obtained at 400 plants m⁻². This suggests that the effects of seeding rate on grain yield are derived from increased production of spikes per unit area, not through the increased production of fertile tillers per plant. In general, fertility and seeding rate were directly associated with the yield of barley. Similarly, our study revealed that a seeding rate of 375 seeds/m⁻² combined with N rates from 95 to 162 kg N ha⁻¹ is optimal for realizing the maximum grain yield.