Herbage Yield and Nutritive Value of Selected Grasses in Subhumid Agroecological Environments in Ethiopia

,


Introduction
Tere is substantial evidence that the livestock sector signifcantly contributes to Ethiopia's national economy and livelihood system.Te sector accounts for approximately 40 percent of agricultural GDP, up to 20 percent of national GDP [1].Tis sector serves multiple functions, including food and nutritional security, as a source of milk and meat, income, manure, and draught power for smallholder farmers.However, it contributed far less than expected, mainly because of feed shortages [2,3].Te availability of feed in Ethiopia's crop and livestock systems depends mainly on natural pastures and crop residues [4].Whereas these feed resources are threatened by overexploitation and lack of high-quality feed [3,5].Furthermore, pressures arising from climate change and variability exacerbate these longstanding feed problems [6].Tese drawbacks highlight the importance of introducing improved forage in addition to the existing feeding systems [7].
Brachiaria (syn.Urochloa) are among the potential grasses that are important for sustainable forage production and are resilient to the detrimental efects of climate change [8,9].It is the source of many tropical grass species that originate in Africa and are widely grown in tropical Latin America and South Asia [10].Consequently, there has been a surge of interest in Brachiaria grass in East Africa [11][12][13].
Tis is due to its adaptability to a wide range of soils, climates, and growing conditions in both the tropics and subtropics [14,15].Moreover, Brachiaria grasses have shown promising results in improving livestock productivity as they have better nutritional quality and biomass production [16].Furthermore, Brachiaria grasses have several environmental benefts, such as the ability to sequester carbon, increase nitrogen use efciency through biological nitrifcation inhibition (BNI), efectively cover crops to control soil erosion, and crop pests through push-pull pest management [17][18][19].
Desho grass (Pennisetum glaucifolium) is an indigenous forage grass species in Ethiopia.In addition to its use as animal feed, Desho grass conserves soil water and serves as a means of income generation for smallholder farmers.Desho thrives well in diferent soil types, has the potential to produce large amounts of biomass per unit area, is suitable for different forage production strategies, is acceptable for diferent livestock species, and increases livestock productivity [20].Morphologically, it is closer to the genus Brachiaria, which shares the acidic and wetter areas of Ethiopia [21].Rhodes grass (Chloris gayana) is a tropical grass widely used in grazing and cut and carry systems in warm areas.Chloris gayana is best suited to areas with 600-1200 mm of rainfall.It is moderately drought tolerant, grows best at high temperatures, and is relatively frost tolerant [22].
To capitalize on these potential forages and close the feed defcit gap, the comparative evaluation of Brachiaria with other important grass species should be the main research approach to demonstrate the suitability of forage for greater adoption in Ethiopia [7].As a result, understanding the infuence of environmental factors and management practices on forage productivity is critical for optimizing feed production and utilization to increase animal productivity [23,24].Information regarding these factors is critical for optimizing feed production and utilization to increase animal productivity.In addition, it is also important to document information on the nutritional value of feeds for their inclusion in livestock feeding programs [25].
Tis study aimed to compare the growth, herbage accumulation, and nutritive value of four Brachiaria cultivars, B. brizantha (DZF-13379), B. humidicola (DZF-9222), B. decumbens (DZF-10871) and B. mutica (DZF-483), and two local grass cultivars, Desho grass (Pennisetum glaucifolium var.Kindu Kosha) and Rhodes grass (Chloris gayana var.Massaba) in two subhumid agroecological in Ethiopia.Te information generated will help identify promising grass cultivars for further utilization under the subhumid agroecological conditions.(2020)(2021)(2022).Te rainfall pattern is bimodal with the main rainy season (Kiremt) occurring between June and September and the short rains (Belg) from February through May [26].Te soil type in the Debre Zeit center is predominantly Eutric vertisol, Vitric Andosols, and Haplic Andosols [27], while the experimental plots were laid out on Eutric vertisol with a clay texture and a neutral to slightly alkaline pH.Te dominant soil type in the Bako center is Alfsols, which are clay in texture and acidic [28].[7], and a popular grass species, Rhodes grass, were used as controls.

Plant Materials
Experiments were conducted in each location using a completely randomized block design with three replicates.Tree blocks, each containing 6 plots of a well-prepared 12 m 2 (3 by 4 m) were used for the experiment.Te vegetative materials were transplanted with the spaces between rows and within plants of 50 cm and 25 cm, respectively.Te transplanted materials were of the same age, and 2-3 plants were placed in each hole at a depth of 10 cm.Nitrogen and phosphorus fertilizers were applied at a rate of 18 N and 46P kg/ha (DAP) at sowing and soon after harvesting in a band along the planting furrow.Hand weeding management practices were performed when appropriate.

Data Collection and Measurement.
Te trial plots were regularly monitored, and data on growth performance including plant height and herbage accumulation were measured during the rainy seasons (June-Sept).Plant height (cm), which is the distance from the soil surface to the uppermost point of the stem as the mean of fve randomly selected plants (one per crown) from the middle row of each plot, was measured immediately during each harvesting.Te frst harvest was carried out three months after the establishment period; in the following year, after clipping at the early onset of the rainy season, it was harvested at an interval of 60 days until the end of the rainy season, leaving a stubble height of 10 cm above the ground by hand using a sickle.Te fresh herbage yield was taken in the feld immediately after the entire plot was mowed for each harvest, and a sample of 300 g was taken.Te sample was immediately cut into small pieces and placed in a draft oven at 65 °C for 72 h to determine the dry matter yield.Simultaneously, a composite sample of 400 g for each treatment was collected and dried in 2 International Journal of Agronomy the shade for laboratory analysis.Herbage DM yield was calculated by multiplying fresh forage biomass by the respective DM concentration of the samples.Herbage DM yield (HDY) was calculated as follows: HDY (DMha −1 ) � HDY•ha −1 .Crude protein (CP) yield (CPY•ha −1 ) was calculated as HDY × CP concentration in the forages (determined after laboratory analysis).
2.4.Laboratory Analysis.Chemical analyses of the feed samples were performed at the Debre Zeit and Holetta Agricultural Research Centers' Feed and Animal Nutrition Laboratory.Te grass samples were dried at 105 °C overnight and ground to pass through a 1 mm sieve.Te total DM was then determined.Te nitrogen (N) content was determined using the Kjeldahl method, and CP was calculated as N × 6.25.Te total ash content was determined by igniting the samples in a mufe furnace overnight at 550 °C.Neutral detergent fber (NDF), acid detergent fber (ADF), and acid detergent lignin (ADL) were determined according to [29].
In vitro organic matter digestibility (IVOMD) was determined using the modifed Tilley and Terry method [30].

Statistical Analysis.
All data were analyzed using a linear mixed model approach through the Lmer function in R, v. 4.1.2(R Foundation for Statistical Computing, https://www.r-project.org).For mean comparisons, cultivars, location, year, and their interactions were treated as fxed factors.In all cases, the block, the interaction of location, and the year were treated as random factors.Mean comparisons of the efect were performed using the "lsmeans" package in R [31].
We assessed these lsmeans over locations and years, given their general interest for subhumid environments of Ethiopia.However, we assessed the value of the individual cultivars in the single sites and years, for plant height, DM yield, and CP yield traits that displayed signifcant cultivar × location and cultivar × year interactions.Nutritive value parameters were analyzed using one-way ANOVA for the signifcant diference among the grass cultivars.Tukey's honestly signifcant diference post hoc test was used to separate signifcant diferences between cultivars.Te statistical model used for the analysis of DM yield and other related agronomic traits was as follows: where Yijkx is the response variable, μ is the overall mean, Cj is the efect of the j th grass cultivar, Lk is the efect of the k th location, Yx is the efect of x th year, Cj * Lk is the interaction between cultivars and locations, Cj * Yx is the interaction between cultivar and year, Cj * Lk * Yx is the interaction among cultivar, location, and year, and eijxk is the random error.Compared to the experimental year, a relatively low dry matter yield was obtained during the year of establishment (2020) for all cultivars.Desho grass, followed by B. mutica and Rhodes grass, showed signifcantly (P < 0.05) the highest DM yield in 2020.Brachiaria (humidicola and brizantha) showed the lowest DM yield during the establishment year.In 2021, all cultivars showed their own maximum DM yield, except for B. brizantha, which had shown the highest in 2022.B. mutica showed the highest DM yield in 2021 but did not difer (P > 0.05) from Brachiaria (decumbens and brizantha) or Desho.B. humidicola, which exhibited the lowest DM yield, did not difer (P > 0.05) from Rhodes.In 2022, B. brizantha, followed by B. mutica, showed signifcantly (P < 0.05) the highest DM yield.Low DM yield was observed for Rhodes, while it did not difer (P > 0.05) from Brachiaria (humidicola and decumbens) Desho grass.

Efects of
Like for DM yield, B. mutica attained the highest plant height but did not difer (P > 0.05) from B. brizantha and Rhodes grass (Table 3).In contrast, B. humidicola achieved a lower plant height.B. mutica showed signifcantly (P < 0.05) the highest plant height in both locations.B. brizantha had the second highest plant height but did not difer (P > 0.05) from Rhodes grass in Bishoftu, while vice Te ash content was signifcantly (P < 0.01) infuenced by cultivar (Table 5).Te highest ash content was observed in B. humidicola, followed by Brachiaria (brizantha and decumbens).Low ash content was observed for Desho grass and Rhodes grass but did not difer for B. mutica.Te lignin concentrations of the tested cultivars did not difer significantly (P > 0.05).In vitro dry matter digestibility was signifcantly (P < 0.001) infuenced by the cultivar (Table 5).B. humidicola was superior to all investigated grass cultivars in terms of in vitro dry matter digestible content but did not difer (P > 0.05) from B. brizantha.B. mutica showed the third in vitro dry matter digestibility content but did not difer (P > 0.05) from Desho and B. decumbens.Rhodes grass had a low in vitro dry matter digestibility content.

Discussion
Te test locations represent the diversity of rainfall, temperature patterns, and soil types that characterize the subhumid environments of central and western Ethiopia.Te amount of rainfall that covers more than 70% during the  4 International Journal of Agronomy main rainy season (Kiremt) and the higher temperatures that occurred in the test years relative to the long-term data are consistent with the predicted efects of climate change in the region [32] and add interest to our results from the perspective of future climate scenarios on forage grasses.Because growing circumstances varied between test sites and years, it was hypothesized that cultivar × location and cultivar × year interactions would show up in forage yield attributes.Indeed, the patterns of microclimate (rainfall and temperature) and soil difered slightly between the experimental sites and years, which afected the yield of the grass cultivars.Tis implies a change in the ranking order of cultivars over location and year.Te signifcant interaction of the yield traits in the current study was in agreement with the report by Wassie et al. [33] which showed a signifcant efect of the interactions (cultivar × altitude and cultivar × harvest date) on Brachiaria brizantha in northwest Ethiopia.
Our fndings supported earlier reports that Brachiaria mutica (var.DZF-483) is a high forage yielding in subhumid conditions [34,35].In addition to its DM yield, B. mutica had the highest CP yield among the tested cultivars.Te dry matter yield of forage Brachiaria mutica (DZF-483) obtained in our results (13.2 t•ha −1 ) was comparable to the result (13.3 t•ha −1 ) reported during its variety registration [34] and higher than the result of 11.8 t•ha −1 that was reported in the northwest highlands of Ethiopia by Bantihun et al. [35].
Brachiaria brizantha (DZF-13379) ranked second and showed an increasing trend from year to year in forage dry matter yield in our study, which indicated the potential and importance of this cultivar.Te Debre Zeit Agricultural Research Center Forage program team has chosen this cultivar as a candidate variety for registration, and our results validate and complement their earlier work (unpublished data).Te average result of 5.4 t•ha −1 reported for Brachiaria brizantha ecotypes cultivated at low and midaltitudes in northwest Ethiopia was less than the dry matter yield of B. brizantha, which was 11.34 t•ha −1 in the current study [33].
In contrast, the dry matter yield of Desho grass (P.glaucifolium cv.Kindu Kosha) varied with location (altitude) from high in Bishoftu to moderately low in Bako, from 12.94 to 9.67 t•ha −1 , respectively.Te current results for the dry matter yield of Desho grass were also lower than those reported in other studies conducted at higher altitudes in Ethiopia [36][37][38].Tis is in agreement with Mengistu et al. (2019), who indicated that highland environments were more ideal for the growth and development of Desho grass species.
Te dry matter yield obtained for Brachiaria decumbens (10.39 t•ha −1 ) in our study was lower than that reported for U. decumbens (11.40 t•ha −1 ) under supplementary irrigation conditions in Holetta, Ethiopia, by Faji et al. [36], which might be due to irrigation supplementation.Rhodes grass (Chloris gayana cv.Massaba) forage dry matter yield in our study, 9.71 t•ha −1 , was much lower than that reported by Faji et al. [36], which was 14.93 t•ha −1 and reported irrigation supplementation in Holetta, Ethiopia.On the other hand, although the dry matter yield of Brachiaria humidicola was the lowest, this cultivar has higher CP and IVDMD.It  International Journal of Agronomy ranked third in CP yield, indicating better nutritional importance.Other researchers also confrmed that, in addition to its good forage, this cultivar is known for its restoration ability when established on degraded land [39].
In this study, all cultivars had CP concentrations >7%, which met the minimum crude protein requirements (7%) for the synthesis of microbial proteins in the rumen that can support at least the maintenance requirements of ruminants [40].Te CP concentration for Brachiaria mutica in our study was slightly higher than the result of 11.6, reported by Bantihun et al. [35].Te B. decumbens Desho and Rhodes cultivars were also higher than the result reported by Faji et al. [36] in Holetta under irrigation supplementation.On the contrary, the CP concentration of Rhodes grass is much lower than the result reported by Jayasinghe et al. [41].Furthermore, as Faji et al. [36] cited Lonsdale (1984), the feeds were classifed based on their CP concentration as low, medium, and high protein sources with percentages of <12%, 12-20%, and >20% CP, respectively.Brachiaria (humidicola and mutica) was classifed as a medium-protein feed source, whereas the other tested cultivars were classifed as low-protein feed sources.
Fiber content (defned by NDF and ADF), an estimate of the amount of plant cell wall rather than cell content, is inversely related to digestibility and forage intake.Terefore, a high content of NDF indicates indigestible portion of the feed, which is an indicator of the quality of the feed.Our results for the NDF content for all grass cultivars fell above 65%.According to Van Soest [42], NDF content of tropical grasses within the limit (65% and above) is classifed as lowquality roughage feed.However, our results disagreed with the result reported by Faji et al. [36] that perennial grass species varied in NDF content.Similarly, the ADF content is an indicator of forage intake.Forage grasses with a higher ADF content have a lower intake [42].In the present study, B. mutica (var.DZF-483) can be considered to have a higher intake than other evaluated perennial grass cultivars.Te varied ADF content of the cultivars observed in our study could be due to the genotypic and phenotypic characteristics of the grass cultivars.
Te ash content of any feed is a positive indicator of the inorganic content (minerals) and an indirect estimator of the organic matter content.Most forages had ash contents ranging from 5 to 12%, which is in agreement with the results of Wassie et al. [33].According to Van Soest [42], a lignin content greater than 6% negatively afects the digestibility of grass forage and the evaluated grass cultivars in the current study had higher values.In the current study, in vitro dry matter digestibility declined slightly with an increase in NDF, ADF, and ADL.Tis may be due to the fact that higher levels of NDF, ADF, and ADL induce lignin to be deposited in the cell walls and increase the percentage of stems, which are less digestible than the leaf section that has a high fber and lignin content [42].Our results were consistent with Faji et al. [36], who reported that IVDMD levels of perennial grasses were inversely proportional to NDF, ADF, and ADL, which are more slowly degraded in the rumen, afecting microbial synthesis and animal performance.Similarly, the variation in the nutritional composition in our results is consistent with the results of Faji et al. [36], who reported that the nutritional composition of forage crops can vary with genotypic characteristics and environmental conditions.

Conclusions
Tis study comprehensively evaluated the growth, herbage accumulation, and nutritional value of various forage grass cultivars in two subhumid agroecological zones in Ethiopia.Te fndings underscore the signifcance of considering environmental conditions and cultivar characteristics in selecting forage grasses.Brachiaria mutica (var.DZF-483) appeared as the top-performing cultivar, excelling in dry matter yield and crude protein yield across the test locations and years.Following closely, Brachiaria brizantha (DZF-13379) showed promising potential with the second highest dry matter yield.Our fndings have important implications for selecting target grass cultivars for adaptation and utilization.Terefore, we recommend Brachiaria mutica (var.DZF-483) and Brachiaria brizantha (DZF-13379) as preferentially recommended in the study areas as alternative forage grasses due to their higher biomass production and acceptable nutritional value.Finally, further studies should be conducted on the performance of animals fed these grass cultivars as a basal diet, aiming to fne-tune the concentrate supplementation recommended for livestock producers.
However, despite the variability in environmental conditions across the two growing locations, CP yield was not afected by the cultivar × location interaction.
Cultivars, Location, Year, and Teir Interactions on Yield and Yield Components.Variations among cultivars over locations and years were signifcant (P < 0.001) for plant height, dry matter (DM) yield, and crude protein (CP) yield (Table1).In addition, cultivar × year and cultivar × location interactions were signifcant (P < 0.05) for the DM yield and plant height.3.2.Growth and YieldPerformance.Combined analyses showed that B. mutica was a signifcantly (P < 0.05) topranking cultivar in terms of dry matter yield (Table2).B. brizantha and Desho grass ranked the second and third highest DM yields, respectively, but they did not difer (P > 0.05) from all cultivars except for B. humidicola.Low DM yield was observed for B. humidicola but did not difer (P > 0.05) from B. decumbens and Rhodes.In Bishoftu, B. mutica showed the highest DM yield but did not difer (P > 0.05) from all other cultivars except for B. humidicola.Similarly, although a low DM yield was observed for B. humidicola, it did not difer (P > 0.05) from Rhodes and B. decumbens.B. mutica, followed by B. brizantha, had the highest DM yield in Bako.However, B. brizantha did not difer (P > 0.05) from Desho and B. decumbens.Low DM yield was observed for Rhodes grass but did not difer (P > 0.05) from Brachiaria (humidicola and decumbens) and Desho.

Table 1 :
Statistical probabilities of the F criterion of six cultivars tested for three years (2020-2022) in subhumid environments in central and western subhumid environments.
versa in Bako.Troughout the years, the highest plant height was observed for B. mutica, although it did not difer signifcantly (P > 0.05) from that of B. brizantha in 2022.Low plant height was observed in B. humidicola throughout the experimental period.3.3.Crude Protein Yield.Te crude protein yields of perennial grass cultivars over three years in the two locations are shown in Table4.Te results of the combined analysis showed that the crude protein yield was signifcantly (P < 0.001) infuenced by the cultivars.B. mutica had the highest crude protein yield but did not difer (P > 0.05) from B. brizantha.Similarly, the highest crude protein yield was observed for B. mutica, followed by B. brizantha, at 3.4.Nutritional Content.Te combined analysis of the nutritive composition of the forage grass cultivars grown for three years in two locations is presented in Table5.Te crude protein content (CP%) was signifcantly (P < 0.01) infuenced by cultivar.A higher CP content was observed in Brachiaria (humidicola, mutica, and brizantha).A low CP content was observed for Desho grass.Te NDF content did not difer (P > 0.05) in any of the grass cultivars tested.Te ADF content was signifcantly (P < 0.01) infuenced by cultivar.Te highest ADF value was observed for Rhodes grass, but it did not difer signifcantly (P < 0.05) from all cultivars tested except for B. mutica.

Table 2 :
DM yield (t ha −1 ) (LS means) of grass cultivars tested for three years in two locations.

Table 3 :
Plant height (cm) (LS means) of grass cultivars was tested for three years at two locations.
a,b,c,d,e means with diferent letters are signifcantly diferent.

Table 4 :
Crude protein yield (t•ha −1 ) (LS means) of the cultivars tested for three years at two locations.

Table 5 :
Nutrient composition of grass cultivars tested in central and western subhumid areas.Means with diferent letters are signifcantly diferent; CP: crude protein; NDF: neutral detergent fber; ADF: acid detergent fber; ADL: acid detergent lignin; IVDMD: in vitro digestibility of dry matter.