Potential of indigenous legume fodder tree and shrubs to animal feed and mitigation of methane emission in the semi‐humid condition of southern Ethiopia

The study was conducted to investigate the in vitro gas production characteristics and methane (CH4) emission potential of indigenous legume fodder trees and shrubs (ILFTS). The most predominant 11 ILFTS species were selected, and leaves and fruit/pod samples were collected, oven dried, and ground. The potential leaf biomass yield (PBY), dry matter (DM), ether extract, digestible crude protein, carbohydrate (CHO), acid detergent lignin, ash, total phenol, condensed tannin, gross energy, digestible energy, metabolizable energy (ME), gas production characteristics, CH4 emissions, organic matter digestibility (OMD), and short‐chain fatty acids (SCFA) were determined. ANOVA and independent t test were used to examine variation among species in plants nature and between fruits and pods within agroecosystems, respectively. Correlation analysis was used to determine relationships among parameters. The study indicated that only DM and CHO showed substantial variation in nutritional quality parameters among trees, shrubs, and fruit/pods in the lowlands. Fruits/pods and trees displayed significant differences in gas production characteristics in the lowlands, unlike shrubs, which had non‐substantial variations. Moreover, the PBY, CH4 emission, OMD, and ME of ILFTS revealed substantial variation (P < .05) with species and among trees, shrubs, and fruits/pods in both agroecosystems. Besides, CH4 production showed a positive significant correlation with gas volume and (b) substantiating the effect of rate and degree of fiber fermentation on CH4 emission. It was discovered that there was a positive significant correlation between the 6 and 24 h incubation period, (c) which substantiated the need for optimal microbial density and substrate for high rate constant gas production of b (c). In conclusion, ILFTS produce considerable biomass rich in nutrients but vary in the degradability of CHO with plant nature, species, and forage origin. This elicits differences in gas production characteristics and CH4 emission with in vitro fermentation, resulting in differences in the corresponding OMD, ME, and SCFA values.

Fodder trees and shrubs produce a considerable quantity of forage biomass as leaves, pods, and fruits that are rich in nutrients, hence having the potential to complement feed and nutrient deficiencies of ruminants commonly observed in dry periods in the tropics and subtropics (Ali et al., 2020;Andualem et al., 2021).However, their nutritional quality differs among species due to the variation in chemical composition and associated characteristics, which originate from intrinsic and environmental factors (Ayenew et al., 2021).The cell wall encompasses 23% to 90% of plant mass (Lee, 2018) and is a main constituent influencing forage quality, whereby lignin is the principal component that hinders cell wall digestion, via denying access of the microbial enzyme to the cell wall components.Poor forage digestibility in ruminants is a concern since it compromises animal productivity and contributes to greenhouse gas emissions (GHG) (Berhe et al., 2020).
It is estimated that 7,516 million metric tonnes of CO 2 equivalents (CO 2 eq) or 18% of the world's yearly GHG are produced by livestock (Steinfeld & FAO, 2006).In Ethiopia, enteric fermentation was the source of 88% of the methane (CH 4 ) that made up the nation's total GHG emissions (Berhe et al., 2020).Enteric CH 4 emissions, which represent energy loss from ruminant animals, are problematic due to its significant impact on animal performance and its contribution to climate change.Depending on the kind of feed, ruminant intestinal CH 4 losses range from 2% to 12% of gross energy (GE) intake (Johnson & Johnson, 1995); therefore, measures to cut emissions also present a chance to boost livestock productivity.
Ruminants fed on poor fiber-based diets produce more CH 4 per unit product than those fed on high-quality forage diets (Opio et al., 2012).Due to the presence of cellulose at a higher concentration than hemicellulose, tropical forages produce three times more CH 4 .Appropriate feed choices and forage supplementation to improve forage digestibility mitigate enteric CH 4 emissions (de Souza Filho et al., 2019).Studies substantiated that the incorporation of fodder trees and shrubs in poor quality fiber-based diets significantly enhanced the digestibility and hence mitigated enteric CH 4 emissions (Naumann et al., 2017).The secondary metabolites in fodder trees and shrubs reduce enteric CH 4 emission via reducing methanogens, shifting the metabolic pathway to less hydrogen synthesis, and increasing rumen bypass proteins and carbohydrates (Naumann et al., 2017).
For ruminants, given fiber-based diets, there is currently growing interest in using fodder trees and shrubs as a supplement to make up for nutrient deficiencies and enhance digestibility, thereby reducing enteric CH 4 emission in tropical and subtropical areas (de Souza Filho et al., 2019).Therefore, this study was conducted to evaluate the digestibility kinetics and potential for CH 4 emission from indigenous legume fodder trees and shrubs (ILFTS) using gas production techniques.

| Description of the study area
Gamo zone is located in the southern part of Ethiopia; it lies about 445 km southwestern of the country capital Addis Ababa.It roughly lies between 5 0 57 -6 0 71 North, latitude, and 36 37 0 -37 98 0 East, longitude.The elevation in the Gamo zone ranged between 501 and 4,207 m above sea level.Gamo zone is characterized by bimodal rainfall with the mean annual rainfall ranging from 801 to 2,000 mm and the annual mean temperature range from 10.1-27.5 C. The terrain has an undulating feature that favors the existence of different agroecosystems in close proximity, ranging from dry lowland to wet highland (Dires et al., 2021).Agroforestry is the common practice where trees are an integral component of the farming system that complement the function of land uses and enhance productivity (Abraham et al., 2022).

| Estimation of the potential leaf biomass yield
The lowland (1,000-1,500 m above sea level [masl]) and midland (1,500-2,300 masl) agroecosystems are characterized by high diversity and species richness of ILFTS in Gamo zone (Abraham et al., 2022), and hence, 11 dominant and preferred ILFTS species commonly used as forage were selected.Accordingly, Acacia tortilis (A.tortilis), Acacia albida (A.albida), Acacia Senegal (A. senegal), Acacia hockii (A.hockii), Dichrostachys cinerea (D. cinerea), Acacia mellifera (A.mellifera), Acacia nilotica (A.nilotica), and Acacia brevispica (A.brevispica) were selected.The potential leaf biomass yield (PBY) was determined to estimate contribution of each species for ruminant forage supply.The mean PBY of each ILFTS was estimated by measuring the stem diameter in randomized quadrant of 20 Â 20 m in the respective agroecosystems with a measuring tape and using the equation of (Petmak, 1983).Five randomized quadrants were used for each species in each agroecosystems, and the minimum distance between two quadrants was 1 km.Accordingly, the PBY of fodder trees/shrub was predicted following the allometric equation: where W = PBY in kg of dry weight, DT is trunk diameter (cm) at 130 cm height for tree, and DS was the stem diameter of a shrub (cm) at 30 cm height.Similarly, trunk diameter (DT/DS) was estimated by DT DS ¼ 0:636C, where C = circumference (cm).
For shrub with many primary branches, the stem diameter was the total diameter of the stem of all the primary branches.

| Plant sample collection
The leaf (all selected ILFTS), pod (A.tortilis), and fruits (A.albida) samples were collected in May and June and oven-dried at 55 C for 72 h for constant weight to determine the dry matter (DM) (AOAC, 2006).
Oven-dried feed samples were grounded using a Wiley mill to pass through a 1 mm sieve for chemical analyses and in vitro gas and CH 4 production trials.

| Calculations of energy and digestible nutrient values
The gross energy (GE), digestible energy (DE), and metabolizable energy (ME) of the ILFTS were calculated using equations from (Hvelpund et al., 1995) as follows: The total carbohydrate (% CHO) was calculated as Total digestible nutrient (TDN) of the ILFTS was calculated following Ranjhnan (2001).
2.6 | Determination of the in vitro gas and methane mitigation potential of ILFTS The rumen fluid was collected from the three rumen cannulated steer that was fed natural pasture hay (5-6% CP) ad libitum and supplementing with about 2 kg of concentrate mixture (69% wheat bran, 30% noug seed cake, and 1% salt) per steer/day.The animals had free access to clean drinking water and mineral licks.Rumen fluid samples were collected before the morning meal in pre-warmed thermo flasks and transported immediately to the adjoining laboratory, where it was strained through three layers of cheesecloth and kept at 39 C under a CO 2 atmosphere.Filtered rumen fluid was pooled together in order to achieve homogenous inocula.
In Vaseline was applied to the pistons to ease movement and prevent the escape of gas.The syringes were pre-warmed at 39 C before the addition of 30 ml of buffer mixture and rumen liquor into each syringe.The syringes were agitated 30 min after the start of incubation and every hour for the first 10 h of incubation.
The gas production was measured at 3, 6, 12, 24, 48, and 72 h of incubation, and after 72 h, 4 ml of NaOH (10 M) was introduced to estimate the CH 4 production as reported by (Fievez et al., 2005).
Three blanks were set in each run.The volume of gas generated per sample using a blank was deducted from the average of the volume of gas produced from the samples.The incubation time was plotted against the volume of gas generated at regular intervals, and from the graph, the equation was used to estimate the gas production characteristics.
described by Ørskov & McDonald (1979), where Y is the gas production (ml/200 mg DM), a is the intercept of the gas production curve (gas produced from soluble fraction), b is the extent of gas production (potentially degradable fraction), a + b is the potential gas production (ml/200 mg DM), c is the rate constant of gas production of b, and t is incubation time.
where GV, CP, and CF are net gas production at 24 h (ml/200 mg DM), crude protein, and crude fat (% DM) of the incubated samples, respectively.

| Statistical analysis
The collected data were coded and tabulated for analysis.Data about the PBY, in vitro gas production characteristics, CH 4 , OMD, ME, and SCFA among the species, trees, shrubs, and fruits/pods were subjected to the analysis of variance (ANOVA) in agroecosystems.
For the gas production parameters and their derivate, a variance analysis model with two fixed factors was used: plant species and agro-ecology.The linear model used was where Y ij is the observation of the ith plant species and jth location; μ the population mean; P i the ith plant species effect; L j the kth effect of location; and Є ij the residual error.
The fermentation parameters such as a, b, and c were estimated for each plant species, using nonlinear procedures of SPSS version 21.
TukeyHSD tests were used for mean separation among the species in each agroecosystems.Significant level were declared at P < .05.A simple correlation analysis was used to establish the relationship between nutrients and in vitro gas production characteristics.

| PBY of ILFTS
The study indicated that the TC, TD, and estimated PBY of ILFTS exhibited a significant difference ( p < .05)with species in agroecosystems and plant nature in the lowlands (Table 1).Based on the  The mean estimated PBY of trees and shrubs in the lowland were 3.21 ± 0.26 and 1.33 ± 0.75 kg/tree, respectively, whereas the midland was 3.2 ± 0.39.

| The nutritive value parameters
Among the ILFTS nutritive value parameters, species showed broad variation (

| Gas and methane production characteristics
All the in vitro gas volume and gas production characteristics of the ILFTS exhibited no significant difference with species among the trees and fruit/pods in the lowland unlike CH 4 production for trees and (b) for fruits/pods, which differed significantly (Table 4).The shrubs showed significant differences with species for all gas volume and gas production characteristics except for (a) and (c) and lag time, which had no significant difference.On the other hand, the midland trees showed significant differences only at the 3 and 48 h incubation period.
Among the shrubs in the lowland, A. hockii leaf exhibited the highest amount of gas production from (b) and (a + b), whereas D. cinerea leaf had the least amount.Moreover, gas volume and gas production characteristics (a) and (b) differed significantly among trees, shrubs, and fruits/pods of the ILFTS in the lowlands (Table 5).Fruits/pods excelled significantly over trees and shrubs for gas volume (Figure 1) at all incubation periods and gas production characteristics (a) and (b) in the lowlands.
Methane production among trees and shrub species exhibited significant differences in the lowland unlike fruits/pods, which exhibited no significant variation (Table 4). A. albida leaf and A. senegal leaf excelled in CH 4 production among the trees in the lowland, whereas A. nilotica produced the least.Among the shrubs, A. hockii leaves exhibited the highest CH 4 content; however, D. cinerea leaves exhibited the least value in the lowland.On the other hand, no significant differences were observed among the trees, shrubs, and fruits/pods in the lowland (Table 5).The midland tree species observed significant differences in CH 4 production, where E. brucei leaf and A. lahaii leaf observed the highest and lowest values, respectively.

| Post incubation parameters
The post-incubation parameters OMD, ME, and SCFA of ILFTS fruits/ pods exhibited significant differences with trees and shrubs in the lowlands.However, with species, only OMD and ME showed significant variation among trees and shrubs in the lowland and midland unlike fruits/pods, which displayed non-significant variation (Tables 6   and 7). A. nilotica leaf showed significantly low OMD and ME values compared to the other three tree species in the lowland.This was in contrast to A. hockii leaf, which exhibited significantly higher values for all post-incubation parameters in shrubs in the lowland.However, the SCFA found non-significant variation in species among trees in both agroecosystems.Alternatively, all post-incubation parameters of trees, shrubs, and fruits/pods showed significant variation in the lowland (Table 7).

| The correlation among the nutrients, gas and methane production
DCP of the ILFTS showed positive significant correlations with DE (r = .697),GE (r = .808),and TDN (r = .758),unlike gas production (À.270),DM (À.548), and ME (À.58), which had non-significant inverse correlations (Table 8).TP exhibited negative non-significant correlation with the gas (r = À.035) and CH 4 production (r = À.297) and all the nutrients except the DM, which showed a positive nonsignificant correlation (r = .013).CT had significant negative correlation with DCP (r = À.782),DE (r = À.792), and GE (r = À.590), and non-significant inverse correlation with CH 4 (r = À.087) and TDN Lag time had a negative non-significant correlation with gas production characteristics and the incubation time (Table 9).It was statistically evident that degradation of (a) with incubation times below Likewise, (c) had a positive significant correlation with the incubation time below 24 h.
T A B L E 4 Methane and gas production characteristics of ILFTS in the lowland and midland of Gamo landscape.Abbreviations: a, gas production from the soluble fraction; b, gas production from the insoluble but degradable fraction; a + b, potential gas production; c, the rate constant of gas production of b; CH 4 , methane; ft, fruit; lf, leaf; LT, lag time; MTGR, methane to total gas ratio; NS, not significant; pd, pod; SD, standard deviation.*P < .05,**P < .01.
Methane production showed a significant positive correlation with incubation time after 6 h.Moreover, it exhibited positive significant correlation (r = .686)with (b) unlike (a), which exhibited positive non-significant correlation (r = .121).

| The potential leaf biomass yield
The estimated PBY of ILFTS in the study was within the range of most native browse trees in Africa (Andualem et al., 2021;Balehegn et al., 2012).The PBY of trees in the lowland and midland and shrubs in the lowland was estimated at 3.21 + 0.26, 3.2 + 0.39, and 1.33 + 0.75 kg/tree, respectively, indicating their potential as livestock feed supply.The PBY is the key parameter that estimates the amount of forage available for ruminants to feed.The size of the plant is a significant factor that determines its PBY as evidenced by the considerable difference between trees and shrubs in the study.The difference in PBY among ILFTS species is likely due to changes in plant growth physiology in response to genomic and environmental factors (Anele et al., 2009).

| The nutritive parameters
The nutritive values of the ILFTS stated in the study are within the range reported for most browse species in the tropics (Yisehak & Janssens, 2013) though there are some inconsistencies.As an example, EE, DCP, GE, DE, and TDN of Acacia abyssinica leaf in the study were higher than reported in Yisehak and Janssens (2013), though CT and CHO were lower.The variation in nutritive values among ILFTS species is probably due to intrinsic differences arising from genomic and environmental variants.Genomics and the environment influence the composition of the plant tissue particularly the cell walls.The cell wall comprised 23-90% of plant tissue, which is the major fraction T A B L E 5 Gas volume and gas production characteristics of the trees, shrubs, and fruits/pods of ILFTS in the lowland of Gamo zone.that affect the nutritive quality of forage (Lee, 2018).Ray et al. (2015) observed that regional and inter-annual variations in climate elements produce differences in forage nutritive values.
Forage of ILFTS's GE, DE, and ME values is within the standard (Stergiadis et al., 2015;Waghorn, 2007) and may therefore complement the energy supply of ruminant livestock.The EE content of the forages of ILFTS species lies within the acceptable range for ruminant livestock feeding (Preston, 1995).Ruminant diets should be limited to about 40 g EE per Kg DM (Campbell et al., 2006).EE depress fiber degradation by obstructing the activity of fiber-degrading microbes, as evidenced by its negative correlation with fiber degradation outputs such as gas production, CH 4, ME, OMD, and SCFA.Several theories have been proposed to try to explain the effect of fat on fiber fermentation in the rumen.One of the more popular theories is the effect of lipids on fibers that prevent microbial attachment, and the other is the cytotoxic effect of fatty acids, especially unsaturated fatty acids that interfere with bacterial membrane function.Both theories suggest that fiber-fermenting bacteria in the rumen are affected by the increased fat in the diet and can be seen in a shift in the ratio of acetate to propionate (Jenkins, 1993).

| Methane and gas production characteristics'
The CH 4 volume of ILFTS species reported in the study 4.8-7.3%,which is within the range reported for most browse trees in Africa (Sisay et al., 2017).Methane production of ILFTS among trees, shrubs, and fruits/pod exhibited no significance difference probably due to their non-significance variation of gas production from potentially degradable fraction (b).However, it observed significant variation with species probably due to the substantial variation in their potentially degradable fraction in shrubs and the species innate CH 4 production potential in trees.Among the trees, A. albida leaf and A. senegal leaf excelled in CH 4 production potential as evidenced by their high CH 4 to total gas production ratio (MTGR).Although A. hocki leaf among shrubs produced the most CH 4 , probably due to its substantial gas production from potentially degradable fractions, A. tortilis pod surpassed A. albida fruit in CH 4 emission probably due to the high cell wall fraction of the pod.Methane is the byproduct of fiber fermentation, which is substantiated by its positive correlation with gas production from (b) (Table 9).Methane to total gas production ratio (MTGR) can be used as an index to evaluate the CH 4 production potential of the species, as it shows the amount of CH 4 produced per unit OM degraded (Bezabih et al., 2014).Depending on the kind of feed, ruminant intestinal CH 4 losses range from 2% to 12% of gross energy intake (Johnson & Johnson, 1995).
The gas volume, gas production from the soluble fraction (a), and the rate constant of gas production of b (c) of the fruits/pods exhibited significant variation with trees and shrubs in the lowland probably due to its substantial difference in CHO value.A. hocki leaf and D. cinerea leaf among the shrubs exhibited the highest and lowest values of (b) in the lowlands, explaining their most and least potentially soluble fractions respectively.The higher value of gas production T A B L E 8 Pearson correlation (r) coefficient among the nutrients, gas, and methane production of ILFTS.as evidenced by their positive significant correlation coefficient.
The ranges in gas production characteristics among the species may partly be due to variances in the chemical composition of the ILFTS (Getachew et al., 2002).It is particularly the variation in cell contents and cell wall composition that affect the amount and rate of digestibility of feed fractions by rumen microorganisms (Li, 2021).

| CONCLUSION
The ILFTS produced considerable forage biomass yields with moderate to high GE, DE, and ME values though varied with species and plant nature in agroecosystems.They produced substantial gas volume, gas production characteristics, OMD, ME and SCFA, and CH 4 with in vitro incubation, however, varied with species in agroecosystems and among trees, shrubs, and fruits/pods in the lowlands.This variation is due to the differences in degradability prompted by the variation in the CHO composition.Gas production from the potentially degradable fraction (b) significantly affects the gas volume and CH4 emission of ILFTS forages during incubation.Furthermore, the rate constant of gas production of b (c) species depends on the synergy of microbial density and substrate amount.However, some species produced more CH4 per unit of gas production due to their innate CH 4 emission potential.Methane to total gas production ratio could be a helpful indicator of the CH4 emission potential of ILFTS species.In conclusion, ILFTS have the potential to compensate ruminants' nutrient and feed deficiency commonly observed in the dry period due to their substantial forage supply which is rich in nutrients having considerable degradability.
estimated PBY, A. alibida and A. mellifera excelled among the trees and shrubs respectively in the lowland; however, A.senegal and A. brevispica had the lowest yield.Conversely, E. brucei had the highest leaf PBY in the midland, followed by A. abyssinica and A. lahai.
24 h, whereas (b) had a significant correlation with incubation time after 24 h.Moreover, (a) had a negative correlation coefficient (r = À.233) with (b); however, (b) exhibited a negative correlation coefficient (r = À.034) with (c).The value of (b) had a positive significant correlation (r = .946)with the potential gas production (a + b).
6 h account for the high rate of gas production from the soluble fraction.However, high significance correlation values of (b) with the incubation time after 48 h explain the low rate of gas production from potentially degradable fraction (b).The rate constant for gas production by b (c) requires an optimal microbial population and ample substrate, which is supported by its significant correlation between 6 and 24 h of incubation.Likewise, the positive non-significant correlation of (c) with early and late incubation time (below 3 h and after 48 h) substantiates the low rate constant of degradation of b (c) due to the impact of low microbial population and exhausted substrate respectively.Gas production from (a) influences (c) substantiated by their positive non-significant correlation coefficient.The gas production characteristics (a) and (b) illustrate two different feed fractions with an inverse association that is supported by their negative correlation coefficient.Moreover, (b) had a significant contribution to the aggregate of gas production from potentially degradable fraction (a + b) The nutritive and energy value of forages from trees, shrubs, and fruits/pods of ILFTS in the lowland of Gamo zone.
Table2); however, only DM and CHO fruits/pods exhibited significance difference (P < .05)withtreesandshrubs in the lowland (Table3).A.tortilis pod and A. alibida fruit exhibited 93% and 55% DM content and 716 and 691.6 g/kg DM CHO, respectively, which probably contributed to the significant distinction.The mean DM content of the trees, shrubs, and fruits/pods of the ILFTS in the lowland were 415 ± 25.9 g, 443.3 ± 27.3 g, and 742.5 ± 265.2 g, respectively, whereas in the midland was 326 ± 83.3 g.
A. nilotica and A. hockii had the highest DM content among the leaves of the trees and shrubs, respectively, in the lowland, whereas A. brevispica and A. toritils had the lowest.A. lahai leaf excelled in the midland with DM content; however, E. brucei leaf was the least.The mean CHO contents of the trees, shrubs, and fruits/pods in the lowland were 592.6 + 34.4,598.6+57,and 703.9 + 17.3 g/kg DM, respectively, whereas the midland was 559.3 + 74.5 g/kg DM. A. albida leaf and D. cinerea leaf and A. tortlis pod excelled in CHO content among the trees, shrubs, and pods, respectively, in the lowland, whereas A. lahii leaf was the top among the midland species.The GE, DE, and ME values of the ILFTS range from 13.5-17.8,11.2-16.3,and10.3-12.8MJ/Kg DM, respectively, where A.mellifera leaf and A. alibida fruit showed the highest and lowest values.The EE content of the ILFTS ranges from 27.5 to 50 g/kg DM, where A. brevispica leaf and A. albida fruit exhibited the lowest while the A. lahai leaf exhibited the highest.The EE content of the forages of ILFTS lies 2.25-50 g/kg DM, where A. alibida fruit and A. lahai leaf revealed the lowest and the highest value, respectively.
Superscripted a and b: the same subgroup in column bearing different superscript differs significantly.
Superscripted a and b: the same column bearing different superscript differs significantly.Abbreviations: a, gas production from the soluble fraction; b, gas production from the insoluble but degradable fraction; a + b, potential gas production; c, the rate constant of gas production of b; CH 4 , methane; ft, fruit; lf, leaf; LT, lag time; MTGR, methane to total gas ratio; NS, not significant; pd, pod; SD, standard deviation.*P < .05,**P < .01.OMD, ME, and SCFA of ILFTS in the lowland and midland of Gamo landscape.OMD, ME, and SCFA of the trees, shrubs, and pods of ILFTS in the lowland of Gamo zone.Superscripted a and b: the same column bearing different superscript differs significantly.Abbreviations: ME, metabolizable energy; NS, not significant; OMD, organic matter digestibility; SCFA, Short Chain Fatty Acids; SD, standard deviation.