Introduction

Rangeland ecosystems support the largest proportion of ruminant production, but the predominant and adaptable indigenous grasses cannot support higher ruminant populations. These grasses are low in biomass yield and poor in quality, which has led to introduction of non-native grasses with higher biomass yield, better quality and low moisture demand to increase ruminant productivity. Though the indigenous grasses are highly adaptable and are the basal diet for grazing ruminants, their high lignification directly depress their quality and digestibility, which is likely to be associated with high emissions of the grazing cattle (Berndt and Tomkins 2013).Improving pasture management, whether indigenous or introduced, is therefore a necessary intervention in the rangelands to meet the rising demand for ruminant feedstock (Mganga et al. 2015) and to adapt the negative impacts of the climate change on livestock production.

Some indigenous perennial grasses including C. cilliaris, E. superba and E. macrostachyus predominate grazed pastures in the rangelands of Kenya (Ndathi et al. 2011). However, they are low in protein quality and high in fiber content, and are vulnerable to climate variability and increased land use pressure (Mganga et al. 2015). When grazed they are likely to result in higher methane (CH4) from enteric fermentation relative to temperate grasses, which have a low methane emission value (0.49 g CH4/Kg LW vs. 0.61 g CH4/ Kg LW), as observed by some authors (Archimède et al. 2018; Berndt and Tomkins 2013) when comparing ryegrass and Rhodes grass. Tropical grasses improved for high biomass yield and quality like Chloris gayana have been introduced from high agricultural potential areas into low potential rangelands to supplement the feed base (Shrestha et al. 2013). However, their response to high temperatures and low rainfall is poor compared to the indigenous grass pastures, because introduced grasses exhibit inability to cope high moisture stress. Maybe as kind of transition is needed. If they grow well, however, the introduced grasses are of high-quality with higher amounts of easily fermentable carbohydrates and less NDF, which can lead to increased feed intake, higher digestibility and passage rate and subsequently minimize CH4 production (Waghorn et al. 2002).

Poor quality grass pastures on the other hand when consumed by ruminants emit higher amount of CH4 as a byproduct of anaerobic fermentation in the rumen. The digestibility of pasture depends on its stage of growth with more mature pastures having higher fibre content and increased carbon to nitrogen ratio, which decrease digestibility hence inducing a higher CH4 yield. It is therefore important to prioritize not only feed quantity but also quality, methanogenic and carbon sequestration potentials of grass pastures in the rangeland ecosystems for sustainable ruminant production.

Ruminant livestock is the largest contributor to CH4 emission in the agricultural sector (O’Mara 2011) through enteric fermentation of feed by the methanogenic archaea in the rumen. Ruminants accounts for 18% of the global CH4 emission and 3.3% of greenhouse gas (GHG) (Patra 2016). This also represents loss of 5 to 10% of animal Gross Energy intake depending on diet composition and intake level (Johnson and Johnson 1995; Haque et al. 2014) which represents a loss of dietary nutrients that would otherwise have been used for production of meat and milk (Eckard et al. 2010). Most of enteric CH4 emission from livestock comes from large ruminants (Moss et al. 2000) due to their large rumen and it is influenced by quality and digestibility of the feed consumed (Doreau et al. 2016; Archimede et al. 2011). The highly digestible feed will have an increased feed intake and reduced enteric methane emission. Rumen microbes degrade structural plant fiber under anaerobic conditions to volatile fatty acids (VFA), CO2 and H2. Among the products, H2 is reduced using CO2 with the help of methanogenic archaea in the rumen to form CH4:

$${CO}_2+4H_2\rightarrow{CH}_4+2H_2O\;(\mathrm{Moss}\;\mathrm{et}\;\mathrm{al}.\;2000)$$

Dietary manipulation involving for example improving feed resource base to utilize grass pastures with higher nutritive quality, high carbon sequestration and low CH4 production would mitigate enteric CH4 emission from extensive ruminant production systems. However, evidence is scanty on the characterization of CH4 production potential of locally available grasses when fed to ruminant animals (Bezabih et al. 2014). Ruminants in- vivo studies are expensive, time consuming and require specialized facilities and resources. For this reason, researchers show interest on the use of in -vitro techniques to simulate the in- vivo process (Melesse et al. 2013). The in -vitro technique can study large numbers of species within a short time and at a low cost. This study tested the hypothesis that digestibility and methane emission differs significantly between indigenous grasses (E. superba, C. ciliaris and E. macrostachyus) and introduced grasses (Chloris gayana var Boma rhodes and Extozi rhodes) under Makueni rangeland ecosystems in South Eastern Kenya.

Materials and methods

Study site

The study was conducted in Arid and Rangelands Research Institute (ARLRI) of the Kenya Agriculture and Livestock Research Organization (KALRO) (Appendix 1), where grass samples were collected from established pasture plots. The station is located at Kiboko in Makindu Sub County of Makueni County, which is in the rangelands found in the South Eastern of Kenya. The area is in Agro Ecological Zone V at an elevation of 975 m above sea level and lies within latitude 2° 10' and 2° South and longitude 37° 40' and 37° 55' East (CIMMYT 2013). The precipitation in the area follows bimodal distribution, with long rainy season from March to May and short rainy season from October to December. The area receives mean annual rainfall of 600 mm and mean annual temperature of 23 °C (Mutiso et al. 2018). The plots where grass samples were obtained had Ferralsols soils ranging from sandy clay to loamy sand and were low in organic matter and highly vulnerable to erosion and biological degradation.

Sampling of the grass species

The studied grass samples were of three indigenous (E. superba, C. ciliaris and E. macrostachyus) and one introduced (C. gayana var. Boma rhodes and Extozi rhodes) grass species that were ratoon grasses in seven years old established rain fed five pasture plots (Fig. 1). In each plot, a line transects of 20.62 m was set and three selected sub sites along the transect were taken according to Maweu et al. (2022). One-meter square quadrat was used and all the above ground vegetation within the quadrat was clipped to ground level when the pastures were at bloom stage (the assumed grazing optimal stage of pasture growth). The grasses were kept under a shaded area until transported to the Laboratory of Arid and rangelands research institute (ARLRI), where they were oven-dried at 65°c for 48 h and ground to pass through a 1-mm sieve in a mill. The ground samples were taken for in vitro fermentation, methane gas production and chemical composition analyses.

Fig. 1
figure 1

Sampling design for the experiment in the five pasture plots

Full size image

Chemical analyses

The ground samples of each grass species were collected for nutritive content analysis, to determine: True DM (at 105ºC for 24 h) in an air-forced oven (Genlab Oven, Genlab Ltd, UK.); Ash content by combustion in a muffle furnace at 550 °C for 4 h (HeraeusM110 muffle furnace, Heraeus Holding GmbH, Hanau, Germany) according to AOAC method (AOAC 1990 method no.924.05). Total nitrogen (N) content was determined following the micro Kjeldahl procedure (AOAC 1990, method no. 988.05) using selenium catalyst tablets. The crude protein content was estimated by multiplying total N by a factor of 6.25. Further, samples were analyzed for neutral detergent fiber (NDF) and acid detergent fiber (ADF) according to AOAC (1990) method number 6.5.1 and 6.5.2 respectively, using an Ankom 200 fiber analyzer (Ankom Technology cooperation, Fairport, USA). In addition, Acid detergent lignin (ADL) and Ether extract (EE) were analyzed using AOAC 1990 method number 973.18 and 14.018 respectively.

In-vitro gas production

In-vitro digestibility and gas production for each of the five grass species collected were determined according to the method of Menke and Steingas (1988) described by (Abdulrazak and Fujihara 1999). Rumen fluid was collected in the morning before feeding from one fisitulated Zebu steer which was fed on mixed grass hay for 7 days and watered ad libitum in a volumetric flask, then taken to the laboratory where it was strained through a double layer of cheese cloth to remove large particles. Strained rumen fluid was then mixed with buffer prepared at ratio of 3:1 to simulate action of saliva. One gram of the five grass samples was inoculated using 50 ml of the mixture in 100 ml gas tight graduated glass syringe barrel in triplicate. The syringe pistons were lubricated with petroleum jelly to ease movement and prevent escape of gas. Syringes were pre warmed at 39 °C prior to inoculation of buffer mixture and incubated in water bath maintained at 39 °C swirled gently at each reading and gas volume recorded at 3, 6, 9, 12, 24, 48 and 72 h of incubation.

The samples and blank (rumen fluid + buffer) were also run in triplicates to determine gas produced due to endogenous substrates. Net gas produced was computed from the total increase in volume minus the mean blank value from the recorded gas production of all samples. From the computed gas production values, the model of Ørskov and McDonald (1979) was applied to determine the kinetics of gas production of the grass samples:

$$Y=a+b\left(1-{e}^{-ct}\right)$$
(1)

where,

Y:

is the volume of gas produced with time (t)

a:

is the initial gas production,

b:

is the gas produced during incubation,

c:

is the constant gas production rate constant (fraction/hour),

t:

is the time of fermentation.

In this case, (a + b) represents the potential extent of the gas production.

The sample in-vitro organic matter digestibility (OMD %), Metabolizable energy (ME, Mj/Kg Dm) and dry matter digestibility (DMD%) content was estimated based on 24-h gas production (GP, ml/200 g DM), crude protein (CP) content and ash content using equation by Menke et al. 1979 as;

$$OMD\;(\%)=14.88+0.889GP+0.448CP+0.0651Ash$$
(2)

where,

CP:

is the Crude Protein (% DM),

GP:

is the 24 h gas production (ml/200 mg DM)

Ash:

is the Ash content (%DM)

$$ME\left(\text{Mj/Kg DM}\right)=2.20+0.136GP+0.057CP+0.0002859CF2$$
(3)

where,

ME:

Metabolizable energy (Mj/Kg DM)

GP:

24 hr gas production (ml/200 mg Dm)

CP:

Crude protein (%)

CF:

Crude fat (%)

$$DMD\left(\text{\%}\right)=88.9-\left(0.779*ADF\text{\%}\right)$$
(4)

where,

DMD (%):

dry matter digestibility (%)

ADF:

Acid detergent fibre

Determination of methane emission

Gas production was determined according to the procedure of Menke & Staingass (1988) and Bhatta et al. (2007). Gas samples were collected after every 3 h of incubation at 3, 6, 9, 12, 24, 48 and 72 h from gas tight ground glass syringe barrel headspace to fill a 60 mL syringe, and then transferred to 10 ml glass vials according to Pellikaan et al (2011). The collected gas samples were analyzed for CH4 gas concentration using a gas chromatograph (model 8610C; SRI at the International Livestock Research Institute, Nairobi) fitted with a methanizer on the Flame ionization detector (FID). The gas chromatograph was operated with Hayesep D packed columns (3 m, 1/8″), an oven temperature at 70 °C and FID temperature of 350 °C. Nitrogen (N2) was used as carrier gas at a flow rate of 25 ml min−1. An auto sampler (Model HT200H; Hta) was used to inject 5 ml of gas sample into the gas chromatograph (GC) system. The sample was temporary stored in a 1 ml sample loop then carried to the separation column by high purity nitrogen (carrier gas).

The detector output was in the form of peak areas with milli -volts as the units. The peak area and retention time of CH4 was measured, calculated and reported by digital processor which was then transferred to an excel work sheet for processing. The retention time for CH4 was then compared to the known standard in the literature. The peak areas were then converted into concentration using a calibration curve generated using gases of known concentrations.

Statistical analysis

This study employed a Completely Randomized Design. The grass species were the treatments, while the observation on digestibility, chemical composition, gas production and methane production were treated as dependent variables. Data was subjected to General Linear Model procedure using Statistical Package for Social Sciences (SPSS version 22). The level of significance was set at p < 0.05 for detection of grass effect on chemical composition, digestibility, gas production and methane production. The means were separated using Tukey HSD for multiple mean comparison. A Pearson correlation analysis was also performed to determine the association of methane production with chemical composition, fermentation characteristics and fibre constituents. The level of significance was also set at p < 0.05.

The model fitted was as follows;

$${Y}_{ij}={\mu +T}_{i}+{\varepsilon }_{ij}$$
(5)

where,

Yij:

observation on chemical composition, digestibility, gas production kinetics, methane production of ith grass species on jth replication

μ:

overall mean,

Ti:

fixed effect of the grass species i

εij:

residual error

Results

Chemical composition

Table 1 shows the chemical composition results for the sampled three indigenous and the introduced grass species. The significant differences in chemical composition between the introduced and indigenous grasses were observed for CP, ADF and EE. The introduced grasses (C. gayana var Extozi rhodes) had on average significantly higher CP (74.05 vs 52), ADF (444.9 vs 395.1 g/kg DM.), and EE contents (29.40 vs 22.99 g/kg DM), relative to indigenous grasses (C. ciliaris, E. macrostachyus and E. superba). The DM, NDF, OM, Ash and ADL recorded no significant difference between the introduced and indigenous grasses.

Table 1 Chemical composition (g.kg−1 DM) of indigenous and introduced grass species
Full size table

In vitro gas production and fermentation Kinetics of the introduced and indigenous grass pastures studied

The in vitro cumulative gas production and the fermentation characteristics of the studied grasses are summarized in Table 2. There were significant differences in the cumulative gas production and fermentation kinetics among the studied grasses. The introduced grasses produced the highest gas at 72 h of incubation relative to the indigenous grasses (61.8 vs 42.3 ml). In general, the introduced grasses, tended to produce more gas and had more potential gas production rate than the indigenous grasses. The rate of gas production was highest (0.9 ml/hr) for introduced grasses and lowest (0.6 ml/hr) for indigenous grasses.

Table 2 In-vitro cumulative gas production (gas ml.g−1 DM) and fermentation kinetics /hour characteristics
Full size table

Total in-vitro gas (GP), methane production, dry matter digestibility, organic matter digestibility and Metabolizable energy (ME) for the studied indigenous and introduced grass pastures

Table 3 presents the total in-vitro gas production (GP) and methane gas production expressed in ml/g DM at the end of 72 h incubation period, Metabolizable energy, Dry matter digestibility and organic matter digestibility for the sampled three indigenous and two introduced grass species. Significant differences were noted in ME, DMD, OMD, CH4 and in total gas produced among the introduced and indigenous grasses. The highest values were obtained in introduced grasses relative to indigenous grass for (ME 16.35 vs 12.90, OMD 62.7 vs 53.6, CH4 25.61 vs 15 0.93 and GP 61.75vs 42.28) except for DMD which was highest in indigenous grasses relative to introduced (58.12 vs 54.24).

Table 3 Total Gas and Methane production (at 72 h in-vitro incubation) and calculated Metabolizable energy (MJ/kg DM), Dry matter digestibility, Organic matter digestibility for three indigenous and two introduced grasses studied in the South Eastern rangelands of Kenya
Full size table

The cumulative gas and methane production pattern from the in vitro fermentation of the grass species is given in Fig. 2. The total volume and pattern of gas production varied among introduced and indigenous grasses. The introduced grasses produced the highest volume of gas/methane over all incubation times. The lowest gas/methane production was measured for C. ciliaris at 24 h, and E. superba during 72 h (Fig. 2).

Fig. 2
figure 2

In-vitro Methane and total cumulative gas production trend for the indigenous and introduced grass species incubated for 72 h

Full size image

Correlation between in-vitro methane (CH4) production and chemical composition

Methane production was negatively correlated with OM (–0.024), ADL (–0.023), DMD (–0.880**) of the grass species (Table 4, appendix). A significant positive correlation was noted between methane production and ADF (0.880**), CP (0.264**), EE (0.694**), ME (0.668**), OMD (0.568**) and Gas production (0.877**).

Discussion

This study was conducted with the objective of comparing the digestibility and methane emission of introduced grass species (Chloris gayana var Boma rhodes and Extozi Rhodes) with the indigenous ones (E. superba, C. ciliaris and E. macrostachyus) in the rangelands ecosystem of Makueni County, Kenya. We evaluated the chemical composition, gas production and fermentation kinetics, total gas and methane production and the correlation between the measured parameters. The findings show that nutritive composition of the indigenous and introduced grasses in the present study were comparable with those of earlier reports in the same South Eastern rangelands of Kenya, from the studies by Ndathi et al., (2011) and Koech et al. (2016). The CP content of the indigenous grasses was below the 70 g.kg−1 DM minimum requirement for optimal growth of rumen microbiota (Van Soest 1994), necessary to breakdown of cell wall content of forages. This suggests that utilization of these grasses for ruminants would supply sub-optimal nitrogen levels in the rumen, restricting microbial growth and activity, consequently hindering effective ruminal fermentation and limiting feed intake (Hariadi and Santoso 2010; NRC 2000).

The cell-wall contents (NDF, ADF and ADL) observed for both the indigenous and introduced grass species were above the critical value for tropical grasses. For instance, the NDF levels for indigenous grasses in the range of 670.3 to 749.9 g.kg−1 DM is above the critical value for tropical grasses of 600 g.kg−1 to 650 g.kg−1 DM (Van soest et al. 1991). In feeds, NDF value beyond the critical value is associated with low digestibility, prolonged digesta retention time in the rumen, which reduces the fermentation rate and increases methane production (Doreau et al. 2016). These authors found NDF to have significantly more influence on methane production than digestibility, implying a positive relationship between methane production and NDF content. Therefore, ruminants reared on these indigenous grasses may not achieve higher productivity due to nutritional limitations such as the high NDF that results also in increased methane production. NDF is a key driver to hydrogen production from carbohydrate fermentation in the rumen (Doreau et al. 2016) through the production of more acetate pathways and less propionate. The higher ME levels for introduced grasses may be due to higher CP values, which is supported by the fact that ME is directly proportional to CP content and inversely proportional to fibre constituents.

The present study recorded higher rate and extent of fermentation for introduced grasses compared to the indigenous ones. This could be related with their higher CP and overall lower content of NDF and ADF (Chino Velasquez et al. 2022). The introduced grass species (Chloris gayana var Boma and Extozi rhodes) produced also higher volumes of methane gas when compared to the indigenous grasses (C. ciliaris, E. macrostachyus and E. superba). A possible explanation could be the high fibre content particularly NDF, which is highly influential in hydrogen production from carbohydrate fermentation (Archimède et al. 2011; Doreau et al. 2016; Chino Velasquez et al. 2022). In addition, other fibre constituents like ADF, ADL, cellulose and hemicellulose which are important fibre fractions influencing CH4 production in the rumen could also have been contributing factors in high methane production in the introduced grass pastures. The high NDF value modifies short fatty chain fraction towards acetate producing more hydrogen which is a major determinant in carbohydrate fermentations as observed by Migwi et al. (2013). These authors reported that intake of high fibre forages leads to a significant loss of energy from feeds to form CH4 gas in ruminants. The findings of the present study were also consistent with those of Melesse et al. (2017) who observed a positive correlation between fibre constituents and CH4 production, which was evident in our study for ADF not for NDF. Even though the authors in the latter study observed higher CH4 production than in the present study (20.9 – 30.80 ml/ g DM), it should be noted that their diets had higher in-vitro OMD than the levels obtained in the present study.

The lower in-vitro organic matter digestibility of the indigenous grass species (E. macrostachyus) could be associated with high NDF and low CP levels. The low CP level could have supplied insufficient nitrogen for the proliferation of rumen microorganisms and hence lower fermentation levels compared to the other grasses. Additionally, the high NDF content in the grass could also mean a low supply of readily available energy to nourish the microbes hence further suppressing their activity to result in lower digestibility of the grass (NRC 2000).

We found a strong positive and significant correlation of ADF with CH4, while NDF had a weak, though positive correlation with CH4 that was not significant. The correlation between carbohydrate fractions and cell wall constituents and methane production are also reported by Moss et al. (2000); Singh et al. (2012); Gemeda and Hassen (2014); Doreau et al. (2016). This makes carbohydrate fractions and cell wall constituents are better methane predicators compared to feed components. The positive correlation of CH4 and CP constituents, although weak, was in agreement with the report by Kulivand and Kafilzadeh (2015). Studies report reduction of methane emission when CP increases but this is only the case for high CP feedstuffs (Melesse et al. 2017). This could be because CP contents above a critical threshold of 70 g/kg enhanced rumen microbial activity quickens fermentation and reduces retention time of digesta, resulting in lower methane production. The opposite is observed with CP values below this threshold as we observed in our study where all grass species had CP values slightly higher or below the threshold of 70 g/kg (Hariadi and Santoso 2010). Both the indigenous and introduced grasses were high in fibre constituent (ADL, ADF, NDF), a major limitation to digestibility. The sample indigenous grass species in this study were of poorer nutritional value but lower methane gas production, relative to the introduced grass species. Indigenous grasses produced less CH4 per gram dry matter of unit feed compared to the introduced grass pastures. Producing ruminants on the indigenous grasses thus would need nitrogen supplementation either inform of protein concentrates or leguminous fodder as recommended by other authors for such feeds (Korir et al. 2016; Sampaio et al. 2010).

Conclusion

Introduced grass varieties had higher nutritive values than indigenous grasses which has prompted discussion about using them in Kenyan rangelands. However, this requires careful consideration, since the higher nutritional value of introduced grasses is still too low to support high livestock productivity and the methane emission of introduced grasses are higher than of indigenous grasses. Indigenous grasses, however, despite their lower methane emission are inadequate in quality to support higher ruminant productivity. We therefore conclude that alternative feeding strategies for rangelands need to be developed.