Forage yield, nutrient composition and in-vitro digestibility of ten early maturing cowpea (Vigna unguiculate) genotypes under diverse locations of Western Ethiopia

Abstract Ten early maturing cowpea genotypes were evaluated in three regions (Bako, Boneya Boshe, and Gute) of Western Ethiopia from 2019 to 2020. To determine the forage dry matter (DM), grain yield, and nutritional value attributes, the genotypes were evaluated and laid in a randomized complete block design with three replications. The findings revealed that ILRI-11114 outperformed other genotypes in terms of mean forage DM and crude protein (CP) yields across all sites and years, while ILRI-82D-504-4, ILRI-11990, and ILRI-12669 were higher in mean grain yields. In comparison to Boneya Boshe and Bako, the mean forage DM and CP yield received at the Gute location was greater. However, Bako produced more grains than Gute and Boneya Boshe did. The investigated fiber components did not differ significantly between genotypes (P > 0.05). The levels of CP, in-vitro organic matter digestibility (IVOMD), and metabolizable energy (ME) were higher in genotypes ILRI-11990 and ILRI-11114, while ILRI-12669 and ILRI-12732 had higher DM and ash values. The genotype ILRI-11114 was suggested for cultivation due to its greater forage DM, crude protein, and grain yield, as well as comparable CP, IVOMD, and ME levels. The performance of animals fed this genotype’s fodder should be evaluated in future research.


PUBLIC INTEREST STATEMENT
Poor nutrition is a severe barrier to livestock production in small-holder crop-livestock farming, particularly during the dry season, when feed sources are scarce and of low nutritional value.A sustainable solution to seasonal feed supply and quality problems is required to boost animal output.Supplementation with leguminous forage in ruminant diets may be explored to address restrictions associated with low feed quality in situations where animals are increasingly dependent on low-quality roughages.In this context, cowpea is one of the leguminous fodder crops that can be utilized as a protein supplement to cattle fed low-quality diet, which is the intension of the present study.Thus, in order to characterize the forage production potential and nutritional profile, 10 cowpea genotypes were tested across years and diverse locations.

Introduction
In sub-Saharan Africa, health issues and nutritional constraints limit ruminant production.Nutrition, however, has been identified as a major constraint to livestock productivity, particularly during the prolonged dry season.A similar scenario was reported in Ethiopia (Fekede et al., 2015(Fekede et al., , 2015)).According to the author, the most popular feed sources for livestock are cereal remnants and natural pasture, both of which are typically scarce and of low nutritive value during the latter part of the dry season (Alemayehu et al., 2017).When animals are fed such poor-quality feeds, their intake decreases, they lose weight, and become more susceptible to health issues leading to lower reproductive performance (Alemayehu et al., 2017).This signifies the need to search for alternative feed resources that are locally available and easily accessible to smallholder farmers for use as a supplement for animals based on low-quality feed resources.Growing a variety of high-quality leguminous forage crops is one of the best options to address this issue.Leguminous forage crops could also play a significant role in improving soil fertility and crop productivity (Alemu et al., 2016;Belay et al., 2017) in addition to serving as a source of quality feed for ruminant animals.
Cowpea is one of the leguminous forage crops that can be used as a protein supplement for livestock based on low-quality feed resources.The dual-purpose cowpea has the potential to play a crucial integrating role in intensifying systems by providing human diets with protein as well as livestock with quality fodder (Ogbemudia et al., 2010).Such dual-purpose species are suitable for regions with erratic and irregular rainfall because, when the rain lasts longer than usual, most species are more inclined to produce herbage than grain (Ayana et al., 2015).Since crop-livestock integration is a typical route for agricultural intensification in less developed tropical countries, such food-fed forage crops are essential to food security and the reduction of poverty (Ogbemudia et al., 2010).There are many benefits to using dual-purpose cowpeas for both human food and livestock feed, according to the literature.For instance, Marina et al. (2017) and Owade et al. (2020) reported that in many regions of tropical countries, the parts of the cowpea crop above ground (excluding pods) serve as a valuable source of nutrient-rich fodder for livestock.Marina et al. (2017) and Owade et al. (2020) further reported that young leaves, green pods, and green grain are used as vegetables, and dry grain is used as both human food and livestock feed.
In Ethiopia, cowpea is grown mainly as a source of livestock feed (Agza et al., 2012;Geleti et al., 2014a;Solomon & Kibrom, 2014) and rarely as a portion of human food (Beshir et al., 2019;Sisay, 2015).Despite its wider importance in tropical countries, its use in Ethiopia, for both human food and livestock feed has not been sufficiently studied.Therefore, it is crucial to properly characterize and choose promising genotypes to increase forage and grain yields, which will increase the production of human food and livestock feed for smallholders in Ethiopia.Agbogidi and Egho (2012) reported that the fodder and grain yield of cowpea genotypes showed significant variation, suggesting that genotype, environment, and their interaction could have an impact (Adewale et al., 2010;Gerrano et al., 2019).Thus, it appears that testing the performance of cowpea genotypes for the desired traits under a variety of environmental conditions is critical.As a result, the current study was designed to evaluate the forage and grain-yielding potential, as well as the nutritional quality of dual-purpose early maturing cowpea genotypes grown under diverse locations in tropical Africa.

Study locations
The research was carried out in three areas (Bako, Boneya Boshe, and Gute) of Western Ethiopia across two cropping seasons (2019 and 2020).Tulu et al. (2021) published a map of the study region as well as soil characteristic descriptions for each location.The mean monthly rainfall, as well as the maximum and lowest temperatures at the three experimental sites, are displayed in Figures 1(a-c

Planting materials
120 cowpea genotypes were initially obtained from the International Livestock Research Institute, Addis Ababa, Ethiopia, and were evaluated at the nursery stage for a year at the Bako Agricultural Research Center.74 genotypes were chosen and separated into two sets: 25 genotypes were classified as late maturing and 49 genotypes as early maturing based on visual performance observation, disease, and pest scoring in the field.Both sets moved on to the preliminary variety evaluation stage and underwent a two-year study.Forage yield potential, quality traits, disease and pest resistance, and visual observation for various plant characteristics (leafiness, plant vigor, and plot cover) were considered for genotype selection.Thus, based on these selection criteria, 10 genotypes including the check were chosen from the early maturing set and evaluated across locations and years in the current study.The following genotypes were investigated 11114, 12669, 12722, 12732, 11990, 25316, 25366, 82D-504-4, 95K1543, and Adulala.Adulala was utilized in this study as a standard check for performance comparison against the other nine potential genotypes because it was already registered as a variety.

Experimental layout and design
Cowpea genotypes were sown on well-prepared plots of 4 × 3 m size in each of the three locations (Bako, Boneya Boshe, and Gute), and plots were delineated with a spacing of 1.5 and 1 m between blocks and plots, respectively.The seed was then sown in rows with 0.5 m between rows and 0.2 m between plants.Planting occurred in early June across locations and years, and all plots were fertilized with NPS fertilizer, containing nitrogen (19%), phosphate (38%P 2 O 5 ) and sulfur (7%S), at a rate of 100 kg/ha at the time of planting.The study used a randomized complete block design that was replicated three times across years and locations.

Forage, grain, and crude protein yield
The two middle rows of each plot were collected at the pod initiation stage to estimate the herbage production, and the remaining rows were utilized to evaluate the grain yield after the grain reached physiological maturity.A suspended field balance was used to determine the fresh weight of the cut plant.Then, to determine the herbage dry matter yield, sub-samples of 200 g per genotype and replication were collected from each location and oven-dried at 65 ℃ for 72 h until constant weight was achieved.Crude protein (CP) yield was estimated to be a product of total herbage dry matter yield and its CP content divided by 100.

Nutrient composition analysis
A well-mixed composite sample from each of the three replications of each genotype was taken from each location and oven dried for the evaluation of quality attributes, as previously stated in section 2.4.The material was then powdered so that it would fit through a 1 mm sieve screen before being sent to Holleta National Animal Nutrition Laboratory for a chemical examination.Thus, the Association of Official and Analytical Chemist (2000) technique was used to determine the dry matter (method 934.01),N (method 954.01,CP= N × 6.25) and ash (method 942.05) values.

Data analysis
The General Linear Model (GLM) technique of SAS, Version 9.3 (SAS Statistical Analysis System, 2002), was used for the statistical analyses, and the Tukey test at P < 0.05 was used to differentiate significantly different means.For the assessments of herbage dry matter, crude protein, and grain yield, genotypes, year, location, and their interactions were taken into consideration as independent factors.The model fitted is: , where: Yijkl = response variable; μ = overall mean; Gi = genotypic effect; Ej = environmental effect; Yk = year effect; Gi � Ej � Yk = interaction effect of genotype, environment and year; B l ð Þ= block effect; and εijkl is the random error.For quality traits: a composite sample per treatment was taken from each location, the location was considered as a replicate, and hence the data were subjected to the following model: , where: Yij = refers to the response of forage quality traits; μ = overall mean; Gi = genotypic effect; Ej = environmental effect (replicate); εij is the random error.

Dry matter, grain and crude protein yield
Ten early maturing dual-purpose cowpea genotypes were examined over two experimental years at three different sites.Table 1 displays the dry matter (DM), crude protein (CP), and grain yield (GY) of these genotypes.The genotype, location, and year main effects, as well as a substantial interaction, all had a statistically significant (P < 0.001) impact on the three variables that were evaluated.As a result, the genotype and location mean effects were presented separately.There was a notable variability in forage DM yield among genotypes and locations studied.Forage DM output ranged from 2.4 to 3.5 t/ha at Bako (mean 2.8 t/ha), 2.2 to 3.8 t/ha at Boneya Boshe (mean 3.3 t/ha), and 3.2 to 6.9 t/ha at Gute (4.3 t/ha).
The CP output varied significantly among genotypes and sites.The genotypes with the highest and comparable values at Bako were ILRI-82D-504-4 (94.5 t/ha) and ILRI-12722 (84.8 t/ha), whereas the check had the lowest (51.4 t/ha).The average CP yield in Boneya Boshe ranged from 54.5 t/ha for ILRI-25316 to a high of 100.6 t/ha for ILRI-95K1543.The genotype ILRI-11114 (157.4 t/ha) generated the highest CP yield at the Gute site, while ILRI-95K1543 (58.4 t/ ha), check (64.7 t/ha), and ILRI-12722 (65.6 t/ha) gave the lowest and comparable value.The highest average CP yield across all locations was achieved by ILRI-11114 (104.3 t/ha), followed by ILRI-11990 (92.6 t/ha).Gute (87.2 t/ha) had the highest mean CP yield, followed by Boneya Boshe (79.4 t/ha), which was greater than Bako (68.6 t/ha).
The sites and genotypes have a big impact on grain yield.In terms of grain production capacity, the genotypes varied widely among the test sites.For instance, at the Bako site, ILRI-11114 (1635 kg/ha) and ILRI-82D-504-4 (1690 kg/ha) gave higher and comparable values, however, at the Boneya Boshe site, half of the studied genotypes ILRI-11114 (1360.7 kg/ha), ILRI-82D-504-4 (1507 kg/ha), IRLI-11990 (1449.8kg/ha), ILRI-12669 (1350.5 kg/ka) and ILRI-25366 (1342.1 kg/ha) showed the higher and comparable grain yield than the rest genotypes.Surprisingly, all genotypes at the Gute site showed larger and comparable potential for these measures, except for ILRI-12732 (1008.8kg/ha) and the control (986.1 kg/ha), both of which had somewhat lower grain output.ILRI-82D-504-4 (1565.6 kg/ha) and ILRI-11114 (1525.9kg/ha) produced greater combined mean grain yields across all locations.Unlike the DM and CP output, the grain yield in the Bako location was greater, followed by Gute, which was higher still than Boneya Boshe.Grain yield (P < 0.01), DM yield (P < 0.01), and crude protein yield (P < 0.001) all showed significant changes between the two experimental years, with the mean of the three measured variables being significantly higher in 2020 than in 2019 (Table 2).The genotype examined with the highest degree of consistency was ILRI-11114, which performed better in the three variables mentioned.

Chemical composition
The 10 early maturing cowpea genotypes' nutritional composition is shown in Table 3.The quality indicators that varied substantially (P < 0.05) among the genotypes evaluated were the levels of dry matter (P < 0.01), ash (P < 0.01), crude protein (P < 0.01), in-vitro organic matter digestibility (P < 0.05), and metabolizable energy (P < 0.05).The ash content ranged from the highest and equivalent value of 7.8% for ILRI-12732 and ILRI-25316 to the lowest value of 5.5% for ILRI-11990 with an overall mean of 6.7%.The mean dry matter content averaged for the 10 genotypes was 90.7%.The genotype ILRI-  11990 had the highest crude protein value (26%), whereas ILRI-12669 had the lowest value (19.9%).The concentrations of lignin and acid detergent fiber ranged from 28.4% to 31.6% and 4.4% to 5.6%, respectively, whereas the concentrations of neutral detergent fiber varied from 47.8% to 52.4% with a mean value of 50.7%.Both cellulose and hemicellulose had mean percentages of 24.4 and 21.1%, respectively.The genotype ILRI-11990 had higher in-vitro organic matter digestibility (62.4%) and metabolizable energy (9.37 MJ/kg DM) than the check (59.8% and 8.97 MJ/kg DM, respectively), and both ILRI-95K1543 and the check had lower but comparable values for both of these parameters

Forage dry matter, crude protein and grain yield
The current study discovered that, among the evaluated early maturing cowpea genotypes, there were more significant variations in fodder dry matter (DM) yield and yield-related attributes than in quality parameters.Geleti et al. (2014b) discovered significant variability in fodder DM yield rather than quality for the cowpea accessions studied at Bako, which is consistent with this study.Similar results for Centrosema accessions were previously reported by Diriba et al. (2013a).The forage DM yield obtained in this study was comparable to that reported by Souleymane (2021), who reported the fodder dry matter yield of early maturing cowpea genotypes in Niger yielded between 3.6 and 4.8 t/ha.In comparison to the findings of Geleti et al. (2014b), Solomon andKibrom (2014), andAtsbha et al. (2018) the finding in the present study was noticeably lower.This disparity could be explained by the genotypes tested in this study have a shorter growing time for biomass accumulation than those described in the literature.Heuzé et al. (2015) indicated that under optimal conditions, cowpea may produce more than 4 t DM/ha of yield, and the yield of ILRI-11114 (4.3 t DM/ha) agrees with this study.
The crude protein (CP) yield recorded in the current study was lower than that reported by Geleti et al. (2014b) for five species of cowpea which ranged from 141.03 to 254.97 t/ha.This could be explained by genetic potential and/or maturity nature of the studied genotype.The shorter growth times of the genotypes evaluated in this study for fodder DM yield accumulation led to a lower CP output than that Means within columns followed by different letters differ significantly (P < 0.05); * = P < 0.05; ** = P < 0.01; *** = P < 0.001; LSD = least significant difference.stated by the author, which is understandable given that CP yield is a reflection of total forage DM yield.
On the other hand, the current study's findings are consistent with the crude protein yield reported by Tulu et al. (2018) for 10 genotypes of Lablab purpureus, which ranged from 56.5 to 224.5 t/ha.
The average grain yield figures found in this study were comparable to those found in other studies (Alemayehu et al., 2022;Ddamulira et al., 2015;Tariku, 2018), although they were less than those found by Magulu and Kabambe (2015) and Ndenkyanti et al. (2022).The higher mean forage DM and grain yield observed for ILRI-11114 were consistent with the findings of Ewansiha et al. (2018), who discovered a positive association between fodder DM and grain yield among the assessed cowpea genotypes.According to Adeyanju and Ishiyaku (2007), such dual-purpose early maturing cowpea genotypes with higher forage dry matter (DM) and grain yield are considered climate-smart because they could improve small-scale farmers' resilience by supplying food and feed for their livestock earlier, especially in areas with low rainfall.
The higher mean DM and CP production at Gute than the yield at Bako and Boneya Boshe sites indicate that the location was significantly different for the measured parameters.In agreement with this finding, Tulu et al. (2021) examined 10 Napier grass genotypes in the current study area and discovered that the Gute location provided more fodder DM yield than the Bako or Boneya Boshe sites.However, the higher grain output at the Bako location revealed that the site was better suited if cowpeas were planted for grain rather than forage production.The variation in rainfall and soil fertility, which are expected to vary greatly between locations, may have contributed to the three measured parameters having higher mean values in year 2 than in year 1.

Chemical composition
The results of the current study's analysis of the dry matter content of the tested genotypes are consistent and comparable to that published by Geleti et al. (2014b), but they are greater than those found by Solomon and Kibrom (2014) and Ayana et al. (2015).The studied cowpea genotypes crude protein content (mean 22.9%) was greater than the mean value of 17.36% reported by Geleti et al. (2014b) but equivalent to the mean values of 20.33% reported by Ayana et al. (2015), respectively.Additionally, the CP content for the genotypes tested was higher than some other leguminous crops mentioned in the literature, such as Lablab purpureus (mean 18.8%; Tulu et al., 2018) and Alfalfa (mean 18.73%; Geleti et al., 2014b), but comparable to the CP content for Centrosema species (mean 21%; Diriba et al., 2013a).The CP value reported in this study was generally above the minimum threshold level (15%) needed to promote lactation and growth in dairy cows as reported by Ratnawaty et al. (2013), indicating the suitability of the studied genotypes for usage as a supplement for ruminants fed on low-quality feeds.
The lack of significant variation in cell wall components (NDF, ADF, ADL, cellulose, and hemicellulose) seen across the genotypes studied has similarly been documented in the literature for Centrosema (Diriba et al., 2013a) and Cowpea species (Geleti et al., 2014b;Ilknur et al., 2012).Although there were no significant differences in fiber content between genotypes, the studied cell wall components were all lower than the values reported by Diriba et al. (2013b) who assessed the chemical compositions of browse and herbaceous legume species and selected local feed resources (n = 17) in the current study area.This indicates that the tested genotypes were of high quality in comparison to the majority of commonly used feed resources in the study area.However, in terms of NDF and ADF fiber content, all the genotypes under investigation could be considered good-quality feed resources and suitable as a supplement for ruminants consuming low-quality feed, as defined by Kazemi et al. (2012), who classify feeds with NDF (47 to 53%) and ADF (31 to 40%) content as high-quality feeds.
The ability of a feed to be digested has a significant impact on the amount of protein and energy that animals can extract from it and on how much feed ruminants consume.The IVOMD value found in the current study was above the necessary threshold level of 50% needed for feeds to be regarded as having an acceptable level of digestibility (McDonald et al., 2002).The IVOMD value for the tested genotype was higher than the values in the range of 43.3% to 54.1% and 55.1% to 60.2% reported by Alemayehu et al. (2022) and Solomon and Kibrom (2014), respectively.However, it was lower than the mean values of 71.5% and 67.12% reported by Agza et al. (2012) and Geleti et al. (2014b).The mean ME value recorded in the current study (9.13 MJ/kg DM) was higher than the threshold value 7.5 MJ/kg DM but lower than that reported earlier for five cowpea accessions (10.07 MJ/kg DM, Geleti et al., 2014b) and other protein supplements, such as various herbaceous legumes (10.04 MJ/kg DM, Diriba et al., 2013) and browse legumes species (10.16 MJ/kg DM, Diriba et al., 2013a).

Conclusion
The wide variation in yield and quality traits seen across the genotypes investigated suggests the possibility of selecting higher and better-adapted genotypes for fodder use.Across locations, ILRI-11114 provided the highest forage dry matter and crude protein yield, whereas ILRI-11114 and ILRI-82D-504-4 produced the highest grain yield, but both are statistically equal.The studied genotypes exhibited higher forage DM and CP yield at the Gute location, while in grain yield, the genotype performed better in the Bako location in 2020 than in 2019.Even though the tested genotypes did not differ significantly for all of the studied fiber properties, differences in crude protein, in-vitro organic matter digestibility, and metabolizable energy content were noted, with genotypes ILRI-11990 and ILRI-11114 showing the greatest differences.In general, genotype ILRI-11114 is suggested for cultivation and usage as a supplement to ruminants fed on low-quality feeds due to its greater forage DM, crude protein, and grain yield and relatively equivalent CP, IVOMD, and ME contents.Future studies should focus on assessing the performance of animals fed this genotype's forage.

C) Figure 1a. Mean monthly rain- fall and minimum and maxi- mum temperatures at Bako.
).