Background

Livestock production is one of the most important agricultural land use systems in the world, with grasslands covering 25% of land surface and contributing to the livelihoods of more than 800 million people [52]. However, in Ethiopia livestock production has been mainly constrained by inadequate supply and poor quality of available feed resources [18, 21]). This could be further emphasized by the fact that feed accounts for 60–70% of the costs associated with livestock production.

Nowadays, the most important livestock feed resources in Ethiopia are natural pasture, crop residues and grass hay [1]. Natural pasture hay and crop residues which provide the bulk livestock feed in Ethiopia are seasonally produced during particular periods of the year (October–January) following the main rainy season. Excess forage production is experienced during the rainy season, but more often, acute shortages occur in the dry season as observed in some parts of the country [2]. According to CSA [14] report, the major feed resources in the country are green fodder (54.59%), crop residue (31.6%), hay (6.81%) and agricultural by product (1.53%) which have poor quality. Common forage crops adapted in the farming systems in Ethiopia like Napier grass (Pennisetum sp.) and silver leaf Desmodium sp. have been affected by the global effects of climate change and Napier grass is also threatened by the emergence of stunt and smut diseases [37], which has also limited its expansion to drier areas. Consequently, finding alternative feeds for livestock is an important step to sustain livestock production in the country. This constraint can be solved through finding a climate-friendly way through the introduction of improved forage varieties of the Brachiaria genus [50]. Moreover, climate smart agriculture is one of the means that can help to reduce greenhouse gas emissions and increase livestock productivity of the country through improved livestock feed and feeding practices [19]. Brachiaria grasses tend to be drought resistant and resilient in infertile soils, and produce well with relatively low levels of fertilizer inputs. They are also resistant to many diseases affecting baseline varieties in Eastern Africa, particularly Napier stunt and smut disease [22, 31]. Thus, as a new strategy termed using ‘climate smart Brachiaria grass’, which can withstand heavy grazing is best option.

In addition to adapted to drought, diseases and low-fertility soils, Brachiaria grass sequesters carbon through its large roots system enhance nitrogen use efficiency and subsequently minimize eutrophication and greenhouse gas emissions [6, 36]. Brachiaria grass plays important roles in soil erosion control and ecological restoration. Brachiaria grass species have been important component of sown pastures in humid low lands and savannas of tropical America with current estimated acreage of 99 million hectare in Brazil alone [25]. Moreover, the grass has a potential of production of high dry matter yield [49]. Additional advantages of using the Brachiaria genus in the integrated systems are that the species produce abundant roots which contribute to the collection of water, soil aggregation and aeration [27].

The genus Brachiaria consists of roughly 100 species which grow in the tropics and subtropics. Most of these species are native to Africa, where they constitute important components of the natural savanna landscape [22]. We see the distribution of B. brizantha is high in Africa including Ethiopia [15] and need more research to exploit maximum in the region. Recent trials indicate that adoption of B. brizantha cultivars has the potential to increase baseline milk production of 3‒5 L/cow/d on participating farms in Kenya by 15‒40% [50]. A farm trial in Rwanda reported a 30% increase in milk production and a 20% increase in meat production [13]. Despite the diversity of Brachiaria spp. in Eastern and Central Africa, comparatively little information is available on their agro-morphological characteristics, yield and chemical composition.

The productivity of the different grass species could be distinctly different and is also influenced by area of origin, including temperature, light intensity, total rainfall, soil type, fertilization level, and by stage of maturity [24, 26]. In East Africa, Bogdan [11] reported that B. brizantha is very variable and several varieties show striking differences in habit, morphology and seed setting capacity. It is possible that the different varieties will perform differently in different ecological zones, but information of this kind is lacking in Ethiopia. B. brizantha grass growing in Ethiopia is the wild type and ecologically different (an ecotype) from that in other countries, which is not developed/enriched like that of other countries such as Brazil. Therefore, the study was aimed to evaluate the performance of B. brizantha ecotypes and to select the best herbage yielding and quality among selected three ecotypes so as to use the better performing ecotype for wider distribution among livestock producer communities in the country.

Materials and methods

Description of the study areas

The study was conducted in south Gondar (Amhara Regional State) at three agro-ecologies using a rain fed system. The low altitude area was represented by a location called Futan at Tach Gayint district located at 11°22′N and 28°19′E at an altitude of 1230 m above sea level. Mean minimum and maximum annual temperature of the district ranges from 13 to 27 °C and the mean minimum and maximum annual rainfall ranges from 900 to 1000 mm per annum. The mid-altitude location was represented by a place called Woreta at Fogera district which is situated at 11°58′N and 37°41′E at an altitude range of 1774 masl and is predominantly classified as mid-altitude agro-ecology. The mean annual rainfall is 1216.3 mm and ranges from 1103 to 1336 mm. According to the district Office of Agriculture, the dominant soil type is black clay (ferric vertisols) (personal communication). The highland area was represented by Farta district, Tsegure Eyesus Kebele (Kebele is the smallest administration unit in Ethiopia) at a site called Melo located near Debre Tabor Town, at 11°11′N and 38°E and at an altitude of 2650 m above sea level. The soils of Melo site are characterized by clay and sand mixture with chemical composition of 2.26% organic matter, 0.11% total nitrogen and pH of 5.47. The mean annual rainfall is about 1570 mm and the mean maximum and minimum annual temperatures were reported to be 21.5 C and 9.6 C, respectively.

Land preparation, planting and experimental design

A total area of 341 m2 was prepared for each location. The experimental land was plowed in May and harrowed in June 2017. The land was divided into three blocks each of which comprised three plots (3*3 m each). Planting materials Brachiaria brizantha ecotypes/wild (Eth. 13726, Eth. 13809 and Eth. 1377) were collect from International Livestock Research Institute (ILRI) Forage Gene Bank Addis Ababa, Ethiopia. The ecotypes were planted using vegetative root splits in rows. The experiment was laid out in a factorial arrangement of three altitudes (low, mid and high) and three harvesting stages (60 days, 90 days and 120 days) in a randomized complete block design with three replications. The spacing between rows and plants was 50 cm and 30 cm, respectively. Land preparation, planting, weeding, harvesting and related management practices were applied according to standard practice for the grass [42]. Artificial fertilizers (di-ammonium phosphate and urea) were applied at rate of 100 kg/ha and urea at 25 kg/ha during planting and after establishment based on the recommendations for the grass.

Data collection

Morphological parameters such as plant height and leaf length were measured from 10 plants that were randomly selected from middle rows of each plot at 60 days, 90 days and 120 days after planting. The number of tillers and leaves was computed as mean counts. To determine biomass yield, the forage harvesting was done by hand using a sickle leaving a stubble height of 8 cm according to recommended practice. The fresh herbage yield was measured immediately after each harvest using a portable balance with a sensitivity of 0.01 g. Representative samples were taken from each plot at each site and were dried in a draft oven at 65 °C for 72 h before chemical analysis.

Chemical analyses

The chemical composition of all samples of feeds collected during agronomic was conducted at Debre Berhan Agricultural Research Center Animal Nutrition Laboratory. The grass samples were dried at 65 °C for 72 h and ground to pass through a 1 mm sieve. Ash/organic matter (OM), dry mater (DM), crude protein and total ash were determined according to AOAC (1990). The neutral detergent fibers (NDF), acid detergent fiber (ADF), acid detergent lignin (ADL) were determined according to Van Soest and Robertson [55]. The CP was calculated as percentage of nitrogen in the sample multiplied by a factor of 6.25.

Data analysis

The collected data were analyzed using general linear model of statistical analysis system (SAS) procedures of 2002 version 9.0. Least significant difference (LSD) test was employed for variables whose F values declared a significant difference (P < 0.05). The statistical model for data analysis was:

$$Yijk \, = \, \mu + \, Ai \, + \, Hj \, + Ck + \, Ai*Hj + + A*C + H*C + \, \varepsilon ijk$$

where Yijk is the response (plant morphological parameters, chemical composition, and yield of B. brizanthagrass) at each altitude and harvesting days

µ = overall mean,

Ai = altitude (i = low, mid and high),

Hj = effect of harvesting days (j = 60, 90, and 120 days)

Ck = Brachiaria brizantha Ecotypes (Eth. 13726, Eth. 13809 and Eth. 1377).

Ai*Hj = the interaction of ith altitude and jth harvesting date.

A*C = interaction of altitude and cultivars.

H*C = interaction of harvesting date and cultivars.

εijk = the residual error.

Results

Effect of harvesting dates and altitudes on plant morphological characteristics of B. brizantha ecotypes

The results of the effect of harvesting dates and altitudes on plant morphological characteristics of Eth. 13726, Eth. 13809 and Eth. 1377 grass ecotypes are presented in Table 1. The overall results obtained in the three agro-ecologies indicate that all morphological characteristics of Eth. 13726 grass studied were significantly affected (P > 0.05) by harvesting dates and altitude. In mid-altitude, plant height (PH), number of tillers per plant (NTPP), number of leaves per plant (NLPP) and length of leaves per plant (LLPP) showed significant difference (P < 0.05) while between low and high altitude there was significant difference (P < 0.05) in PH only. All parameters were significantly (P < 0.05) increased as harvesting stage increases from 60 to 120 days. Plant height showed high increment at day 90 while NLPP showed significant increment (P > 0.05) by harvesting at day 120. Although there was no significant (P < 0.05) difference in PH between second and third harvest, numerically there were high increment in NLPP.

Table 1 Effects of altitude, harvesting stages and their interactions on plant morphological characteristics of brizantha grass ecotypes
Full size table

The overall results obtained in the three agro-ecologies indicate that with the exception of PH by altitudes, other morphological characteristics of Eth. 13809 grass studied were significantly affected (P > 0.05) by altitude and harvesting date. Except NLPP by altitude, all the morphological characteristics (PH, NTPP, NTPP and LLPP) of Eth. 1377 grass in all studied agro-ecology were significantly affected (P > 0.05) by altitude and harvesting date.

Effect of cutting ages and altitudes on chemical composition, dry matter yield of three B. brizantha grass ecotypes

The effect of harvesting dates and altitudes on chemical composition, dry matter yield (DMY) and crude protein yield (CPY) of Eth. 13726, Eth. 13809 and Eth. 1377 grasses is shown in Table 2. The effect of harvesting date and altitude on chemical composition, DMY and CPY of Eth. 13726 grass is shown in Table 2. DM, DMY and CP, NDF and ADF by harvesting date and only DM by altitude showed significant differences (P < 0.05) in Eth. 13726 grass.

Table 2 Effect of altitude, cutting age and their interaction on chemical composition and yield of Bracheria grass ecotypes
Full size table

There were significant differences (P < 0.05) in DM, DMY, CP and CPY both by altitude and harvesting date. However, there was no significant difference (P < 0.05) between mid- and high altitudes. Moreover, there were significant differences (P < 0.05) in ash by altitude and ADF by harvesting date of Eth. 13809. The overall results of the grass in low altitude show that greatest DM, DMY, ASH, CP and CPY than mid- and high altitudes. There were significant differences (P < 0.05) in DM and DMY by altitude and harvesting date. CPY by altitude and CP, ADF and ADL by harvesting dates were also significantly different (P < 0.05) in Eth. 1377 grass.

Chemical composition of three Brachiaria grass lines

The effect of altitude, harvesting date and their interaction on chemical composition of Brachiaria grasses is shown in Table 3. There were highly significant difference (P < 0.05) in the average DM, DMY, CPY, ADF and ADL of the three ecotypes. Except NDF, ADF and ADL by attitudes and ash by harvesting dates, there were highly significant (P < 0.05) effects of altitude and harvesting date between ecotypes and have also high interactions between ecotypes and altitudes. There were significant (P < 0.05) overall interactions except for NDF.

Table 3 Mean chemical composition, yield of Bracheriea ecotypes
Full size table

Agronomic performance of three B. brizantha ecotypes grass lines

The agronomic comparison of three ecotype grasses (Eth. 13726, Eth.13809 and Eth.1377) is presented in Table 4. Except NLPP, the agronomic performances of PH, NTPP and LLPP differ significantly (P > 0.05) between three Brachiaria grass lines. Eth. 1377 performs best in PH and NLPP from the other two ecotypes while in case of NTPP and LLPP Eth. 13809 showed best performance.

Table 4 Overall morphological characteristics of three Brachiaria brizantha ecotypes on the effects of altitude and harvesting dates and their interactions
Full size table

Discussion

The agronomic characteristics result indicated that all ecotypes tested recorded higher PH than other Brachiaria grass cultivars (Llanero, MG4, Marandu, Piata, Mulato II, Basilisk and Xaraes) in Kenya dry lands as reported by Nguku et al. [42]. However, height may not be an important estimate on the expected biomass yield, as it has been clearly demonstrated that the shorter Eth. 13809 than Eth. 1377 had the best primary DM yield. This high DM yield of Eth. 13809 might be due to high number of tillers.

More tillers were reported for Eth. 13809 at 120 days (45.56) in low (35.77) and mid (35.07)-altitudes as shown in Table 1. Tillers density is an important attribute of grasses as it increases the chances of survival and amount of available forage [28, 51]. Moreover, it is an indicator of resource use efficiency by the different grass species. The large numbers of tillers produced by some grass species allows them to attain maximum growth at an earlier age and recover faster after defoliation [28]. Tillering is also important in forage plants, because of its influence on leaf-area production and dry matter yield. Mean tiller number increased progressively for all ecotypes, and there were significant differences among the ecotypes (P < 0.05) as shown in Table 4. The results in this study are similar to the findings by Mganga [35] and Machogu [32], where the tiller density differed among the local range grasses in Kenya.

The DM content increased with delayed harvesting because of decreased moisture content in leaves as the plants aged and became lignified. This result is in agreement with other studies [4, 8] for other types of grasses. The studies reported that the DM content of grasses increased with an increase in growth and development of plants and longer time to harvest. The higher dry matter yields at later stages of harvesting were to be expected as plants were taller, had more tillers per plant and more leaves per plant and leaf elongation and stem development. All these characteristics would contribute to increased photosynthetic activity and hence higher DM production. The highest total DM yield observed at the last harvest stage was in agreement with the work of Hare et al. [23], Njarui and Wandera [45] and Ondiko et al. [47] on wild Brachairia grasses at different countries; Tessema and Alemayehu [53], Leta et al. [29]; Asmare et al. [9] on cultivated grasses and Feyissa et al. [20] for natural pasture in Ethiopia. These researchers indicated that the time of harvesting had a high influences on dry matter yield. The highest yield of forage for the longest harvesting date could also be attributed to the favorable rainfall, temperature and available nutrient in the soil over the extended growing period of the grass in the study area. But, when compared with time taken until consumption, the age of the grasses at day 60 could be used by farmers for forage because the relative growth is high, soft and palatable with reduced lignin.

Overall, Eth. 13809 had highest dry matter yield (5.91 t/ha) which is followed by Eth. 1377 (5.09 t/ha). Variations in DM production across the ecotypes can be attributed to differences in growth rate and growth habit, which are mediated through the genotypic and phenotypic differences. This is a common phenomenon in grasses [35, 46]. In this study, the high primary DM yield by Eth. 13809 can be largely attributed to its large size leaves (20.1 cm long) and tiller number (33.52) than the other two ecotypes. However, the leafy nature might be a disadvantage in dry areas where water supply is limited, as it facilitates rapid water loss through transpiration [48]. Grasses which yield the highest DM should be the most sought since they can supply the highest amount of forage to livestock. The average DMY of ecotypes (Eth. 13726 (4.05 t/ha), Eth. 13809 (5.91 t/ha and Eth. 1377 (5.09 t/ha) were very high compared to C. ciliaris with 1.6 t/ha reported by Mutimura and Everson [39] and natural pasture (4.4 ton/ha) in Ethiopia [1].

Crude protein (CP) value of the three B. brizantha ecotypes (Eth. 13726 (16.33, 10.63 and 6.72%), Eth. 13809 (13.87, 10.60 and 9.57%) and Eth. 1377 (14.80, 10.15 and 7.86%) differed significantly (P > 0.05) among harvesting dates. As would be expected, the highest CP concentration was obtained at the earliest stage of harvesting, with values declining as harvesting was delayed. This result is in agreement to the findings by Njarui et al. [44] and Mutimura et al. [38] in Brachiaria grasses. Similarly, Bayble et al. [7] and Ansah et al. [5] reported a decreasing trend of CP with increase in harvesting age (60 > 90 > 120 days) for Napier grass. This is a growth dilution effect with increase in structural carbohydrate content of forage materials harvested at late maturity reducing the percentage of protein in the forage.

The higher CP Contents of Eth. 13809 and Eth. 1377 at low altitude and Eth. 13726 at mid-altitude were higher than one of the best Brachiaria hybrid cv. Mulato II reported by Mutimura and Everson [39] which had a CP of 11.1%. The overall highest CP of the ecotypes was recorded in low land (12.59%) compared to mid (11.35%) and high land (9.56%) which is contrary to those reported as high temperatures to have positive effect on quality of grasses [43]. This might be due to agro-ecological and management differences during the growth stages of ecotypes.

The CP content was generally high in all the Brachiaria grass ecotypes [Eth. 13726 (11.23%), Eth. 13809 (11.35%) and Eth. 1377(10.94%)] compared with mean of (5.3–7.7%) and (7–10%) different Brachiaria grass cultivars reported by Ondiko et al. [47] and Nguku [41] in coastal lowlands of Kenya and in the semiarid region of eastern Kenya, respectively. Natural pasture, Rhodes grass, Tef straw, Maize stover and Finger millet straw were also found have very low CP value of 5.5, 7.1, 4.2, 2.84 and 4.12%, respectively (Abebe et al. [1] and desho grass (8.35%) [9]. Furthermore, the CP content was higher than the most Ethiopian dry forages and roughages which have a CP content of less than 9% [10], which is the level required for adequate microbial synthesis in the rumen [3]. Of the factors considered, the CP content at low altitude was higher (12.59%) than at mid-altitude (11.35%) and high altitude (9.56%) which may have been associated with differences in temperature, precipitation and soil characteristics as reported by Daniel [17] where plant growth and quality were affected markedly by temperature and soil moisture conditions. Generally, all ecotypes in all altitudes including high altitude which is uncommon in other countries are highly palatable.

Forages with a CP content range of 9–12% are highly palatable [32]. Despite the reduction in CP percentage with time, crude protein yield (CPY) increased significantly as harvesting was delayed. Similar findings have been reported by Asmare et al. [9] for the desho grass species and by Melkie [34]. They recorded mean CPYs of Bana grass at 60, 90 and 120 days of age to have 0.47, 0.91 and 0.85 t/ha, respectively. Obviously, decisions on the optimal time to harvest Brachiaria grass will depend on a compromise between yield and quality of forage.

As would be expected, NDF, ADF and ADL concentrations increased as harvesting date was delayed though there was no significant (P < 0.05) difference of NDF in Eth. 1377 and ADL in Eth. 13726 and Eth. 13809. Moreover, there was no significant (P < 0.05) difference in all ecotypes by altitude. The NDF content of forage varies widely, depending on species, maturity, and growing environment [12, 33, 40]. Therefore, each plant species presents a unique NDF–ADF ratio in the feed. For legumes, < 40% NDF content is classified as good quality forage, while > 50% [54] is considered as poor quality forage. For grasses, < 50% NDF is considered high quality and > 60% as low-quality forage. In this study, forage materials from all the grass ecotypes had > 60% NDF which may account for the low intake and digestibility. The higher fiber content in the Ecotype grasses, therefore, can be attributed to their phenological stage. Furthermore, increase in CF content is accompanied by increase in lignin content in forages. The forages, therefore, become less digestible because they are inaccessible to digestive enzymes [16]. The ash quantity of any feed is a positive indicator of the inorganic (minerals) content. Generally, most forage has ash content ranging from 3 to 12% (Linn and Martin [30]. All the three ecotypes investigated in this study were in the range shown by these authors.

Conclusion

Based on results, it can be concluded that the harvesting dates can affect the forage DM yield and nutritive values of Brachiaria grasses. Day 60 could be the optimal level for harvesting Brachiaria grass since the quality forage is high and yield is not compromised. The highest CP and lowest NDF and ADF concentrations were recorded in Eth13809. Thus, among the tested ecotype grasses Eth13809 showed great potential as a forage plant especially in low altitude since it has shown that it can grow under low rainfall maintaining high yields. Overall, Brachiaria grasses had a higher biomass yield and better chemical composition than the main feed resources of natural pasture and crop residues in the region. Therefore, it can be concluded that the studied Brachiaria grass ecotypes, especially Eth13809, have great potential as an alternative ruminant feed in all altitude areas in south Gondar. To fully utilize the potential of the studied Brachiaria grass ecotypes, further studies on agronomic and nutritional evaluation involving animal evaluation experiments are recommended.