Barley (Hordeum vulgare L.) is the fourth most important cereal crop in terms of worldwide grain production (FAOSTAT, 2017). The developing countries contribute to more than half of land area planted with barley crops. Moreover, barley is a multiple-purpose crop with high economic and social importance. It is grown to produce grain for human and straw for  livestock consumption and malt for brewing  (Kaso and Guben, 2015). However, growing barley is associated with a production of large quantities of straw which is used extensively as ruminant feed specially in developing countries during dry seasons.

According to the report of Jaleta et al. (2015) cereal straws (barley teff and wheat) appears  to be preferred as livestock feed by the farmers and share about 77-88% of animal feed in mixed crop livestock  system, from cereals barley is higher than teff and wheat for livestock feed (Seyoum et al., 2020). It has been reported that crop residue biomass and nutritive value are key determinants in varietal selection by farmers in mixed crop-livestock systems (Jaleta et al., 2015; Schiere et al., 2004). Consequently, livestock researchers and crop breeders have launched research themes to upgrade straw yield and nutrient composition alongside grain yield using plant breeding approaches in pulses (Alkhtib et al., 2017; Blümmel et al., 2010; Jane Wamatu et al., 2017) and cereals (Addisu, 2018; Ertiro et al., 2013; Jensen et al., 2011).

Exploiting differences in feed traits of barley types could provide novel breeding targets for new barley varieties with potentially higher food and livestock feed value that would be particularly useful in a range of diverse environments in mixed crop-livestock systems. Thus, the current study aimed to identify superior cultivars in terms of grain yield, straw yield and food-feed potential (dual-purpose use) for mixed farming systems in Ethiopian highlands.

Experimental sites and design

The  study was  conducted  at  Kulumsa  Agricultural  Research  Center,  Bekoji and  Kofele experimental sites. The agro- ecological  description of the experimental sites is presented in Table 1. The experiment comprised of twenty food and twenty malt improved cultivars drawn from National Variety traits (NVT) of the Ethiopian Barley improvement program. The cultivars were planted using randomized complete block design with three replications during the main cropping season (June to November) under rain-fed condition, with a plot size of 2.5 m × 1.2 m. Spacing between plots and blocks was 0.5 m and 1 m, respectively. All plots were equally managed as per recommended agronomic practices for barley growing in Ethiopia, Fertilizer was applied at a rate of 50/100 kg/ha urea /DAP respectively.

Y_ijk= M + G_i + Lo_j+ B_k(L_i) + (G×Lo)_ij + E_ijk

Where Y_ijk is the response variable, M is the mean, G_i is the effect of barley cultivar i, L_j is the effect of the location j, B_k(L_i) is the effect of the block k within k location i, (G×L)_ij is the interaction between the cultivar and the location and E_ijk is the random error.

Grain yield (GY), straw yield (SY) and potential utility index (PUI) across locations

The grain yield performance of malt barley ranged from 5.2 to 6.9 t/ha. The highest grain yield was recorded for cultivars IBON174/03 (6.9 t/ha), USDF5-27 (6.5 t/ha) and IBON-HI13/14-49 (6.4 t/ha), and the lowest yield was recorded for cultivar IBON-HI14/15-126 (5.2 t/ha) (Table 2). The straw yield performance of malt barley ranged from 5.5 to 9.1 t/ha. Traveller (9.1 t/ha), HB1963 (8.4 t/ha) and IBON174/03 (8.0 t/ha) showed the highest grain yield whereas IBON-HI14/15-148 performed least (5.6 t/ha) (Table 2). The potential  utility  index  of  malt  barley ranged from MBHIBYT-22 (0.69) to IBON174/03 (0.72) (Table 2).

The grain yield performance of food barley ranged from 4.0 to 5.8 t/ha. The highest grain yield was recorded for cultivars, EH1493 (5.8 t/ha), ICARDA GP P# 44 (5.4 t/ha) and IBON-HI 13/14-P# 113 (5.2 t/ha) whereas the lowest yield was obtained for IBON-HI 13/14-P# 31 (4.0 t/ha) (Table 3). The straw yield of food barley ranged from 4.0 t/ha to 6.9 t/ha. Cultivars EH1493 (6.9 t/ha), IBON-HI14/15 P#155 (6.5 t/ha) and HB1966 (6.4 t/ha) produced most straw. The lowest straw yield was recorded for IBON-HI 13/14-P#31 (4.5 t/ha) (Table 3). The potential utility index of food barley ranged from 0.67-0.70. Based on the PUI, IBON-HI 13/14-P# 85 (0.70), ICARDA GP P# 44 (0.70) and IBON-HI 14/15-P# 165 (0.70) were the three best cultivars (Table 3).

Grain and straw nutritive quality

A wide range in grain nutrient and energy concentration was  observed  across  the  cultivars  for  food   and  malt barley cultivar: Wi: grain CP: 10.2-12.2% and 11.4.1-13.3%, grain NDF: 40-2-52.7% and 38-42.9%, grain ME: 9.9-12.3 MJ/kg and 12.1-14.5 MJ/kg, (Table 4 for malt type barley and Table 5 for food type barley).

Also for straw, large variation in nutritive value was found across cultivars for food and malt barley cultivar: straw CP: 4.1-5.7% and 4.9-6.2%, straw NDF: 73.5-76.7% and 72.9-76.1%, straw ME: 5-5.6 MJ/kg and 5.3-5.8 MJ/kg (Table 4 for malt type barley and Table 5 for food type barley). Multiple differences were found between cultivars, but with a significant cultivar-location interaction, meaning that the growing location had an impact on the performance of the cultivars and their concomitant PUI ranking. The overall grain quality in Kofele was superior to that in Bekoji. The difference between low and high yielding cultivars was presented in Table 6 (Food barley) and Table 7 (Malt barley ).

Food-fodder correlation

Table 8 presents the relationship between straw yield and nutritive value as well as grain yield and nutritive value across food and malt barley cultivars. The correlation between grain yield and straw yield was positive for both food and malt barley cultivars regardless of the location (r > 0.7). Grain yield was correlated weakly to moderately (r < 0.39) with straw nutritive value parameters for both food and malt barley in both locations.

Generally, the linear correlations between nutrient composition of grain (grain CP, grain NDF and grain ME) and nutrient composition of straw (straw CP, straw NDF and straw ME) were weak to moderate (r < 0.44) for food and malt barley in Bekoji and Kofele.

Exploiting differences in feed traits of barley types could provide novel breeding targets for new barley varieties with potentially higher food and livestock feed value that would be particularly useful in a range of diverse environments in mixed crop-livestock systems. These genotypes would promote sustainable use of resources in the farming systems by increasing biomass production for human and livestock production.

Cultivar improvement of straw traits along with grain traits requires information on the cultivar-environment interactions between grain and straw traits and the relationship between these traits across different environments. The current study showed that cultivar variation in yield and quality traits depends on location. That means that the selection of an optimal barley cultivar should be based on location. Further research is therefore warranted to identify the parameters (e.g., soil type, precipitation and slope) that could predict the location effect on barley performance.

Cultivar performance depended on growing location as shown previously in for instance chickpea (Alkhtib et al., 2018; Wosene et al., 2015) and maize (Ertiro et al., 2013). It implies that other cultivars may be selected when tested at another location.

Genotype-environment interaction is known to account for the variation in nutritive value of cereal straw (Seyoum et al., 2020)  (Birhanu et al., 2020) also identified IBON 174/03 as a high grain yielder although only grain quality and not straw quality was evaluated in that study. Our study evaluated both grain and straw quality and added HB1963 and Traveller in addition to IBON 174/03. Large variation in food and malt barley cultivars has been demonstrated in several studies (Pearce et al., 1988; Reed and Yilma, 1986; Wamatu et al., 2019). Our study walks the same trail but also adds the insight that not only grain quality but also straw quality showed an interesting range for selection. A factor that complicates easy selection of superior cultivars is the impact of growing location observed in both barley types. Tsige et al. (2020) already identified this effect but our study clarifies that some traits are more sensitive to changes in growing conditions than others. For instance, there was a huge difference between locations in ME compared to other traits. The ME at the Kofole site was higher than the ME in Bekoji which may originate from the high rainfall in Kofele compared to Bekoji. High rainfall was correlated to high ME in straw (Acone and Wootton, 1999). It was not the purpose to compare malt barley with food  barley,  but malt barley outperformed food barley in both yield and nutritive value. This difference was expected  because  of the basic difference between malt and food barley types. The higher grain and straw yield observed in malt barley type might be related to the high germination rate as a criterion  for  malt  barley  (Macleod,  2013). At  least,  our study points the necessity to consider the impact of location when selecting parental varieties (or potential food-fodder genotypes) of food and malt barley for breeding programs. This  is  especially  the case for food-feed use of barley, since features of grains and straw were clearly affected by location with different effect sizes, meaning that the overall value of food-feed barley may vary more widely than just grain yield.

The wide cultivar variation in straw yield and nutritive value combined with the poor association between yield and nutritive value for grain as well as straw allows for upgrading of straw yield and its nutritive value without decreasing grain yield and its nutritive value. This improvement could be achieved by employing appropriate breeding approaches.

Considering the environmental impact on the cultivar performance, three superior varieties were identified, that is, IBON174/03 (as grain yielder), HB1963 (as grain and straw yielder) and Traveller (as straw yielder). To evaluate how well the straw of these cultivars suits animal nutrition, a trial with regional livestock is required to evaluate the digestive and metabolic responses to the identified promising barley cultivars.

The authors have not declared any conflict of interests.

The authors appreciate the ?nancial support of the Collaborative Research Program (CRP) Livestock, Feeds and Forage Flagship, Jimma University, Ethiopia and Ghent University Special Research Fund (BOF) Program, Belgium. Gratitude to Kulumsa Agricultural Research Institute Ethiopia, for facilitating the cultivation of barley cultivars. Appreciation goes to the Animal Nutrition Laboratory of the International Livestock Research Institute (ILRI) in Addis Ababa, Ethiopia for analysis of the samples.