Methanogenic potential of commonly utilised South African subtropical and temperate grass species as influenced by nitrogen fertilisation

Additional keywords: fermentation, greenhouse gas.

Understanding the effect of forage quality on the production of anthropogenic greenhouse gases from livestock is important for the development of mitigation strategies for agricultural systems (Beauchemin et al. 2008; Bhatta et al. 2017). The livestock sector is a significant source of greenhouse gas emissions in South Africa, contributing 60% of total agricultural CO₂-equivalent emissions (Meissner et al. 2013). Beef cattle, sheep and privately owned game enterprises rely mainly on extensive forage-based production systems and account for 85% of total livestock methane (CH₄) emissions in South Africa (Du Toit et al. 2013a, 2013b, 2013c).

One of the main factors contributing to the limited productivity of ruminant livestock in tropical and subtropical regions in developing countries is the poor nutritional conditions, characterised by highly lignified, low-digestibility feed from poor quality, nitrogen (N)-limited native rangeland and crop residues (Goel and Makkar 2012). Meissner et al. (1999) categorised roughage quality according to the digestible organic matter (OM) concentration as poor (<45%), low (45–55%), medium (55–70%) and high (>70%). Improving forage quality offered to ruminants through forage species selection, rangeland reinforcement through the introduction of more productive and nutritious species, and improved rangeland-management systems has the potential to reduce CH₄ emissions per unit animal product as a result of increased digestibility and reduced ruminal retention time of feed particles (Beauchemin et al. 2009; Banik et al. 2013). Benchaar et al. (2014) stated that a 15% reduction in CH₄ emissions could be achieved by increasing the digestibility of forages and a 7% reduction through increasing voluntary feed intake of livestock.

The influence of N fertilisation on improvement of forage quality and productivity has been investigated by several researchers (Valk et al. 1996; Rivera et al. 2017; Ullah et al. 2018). However, studies evaluating the effect of N fertilisation on the methanogenic potential (in vitro CH₄/in vitro total gas production, TGP) of tropical and subtropical grass species are not readily available. Nitrogen fertilisation can influence the pattern and rate of degradation in the rumen of crude protein (CP) and neutral detergent fibre (NDF) (Valk et al. 1996) by increasing the concentration of neutral detergent insoluble N and altering the protein : carbohydrate ratio (Valk et al. 1996), which could influence the methanogenic potential of forages. Previous studies hypothesised that increased N fertilisation of forages would reduce fermentation gas production because of differences in the stoichiometry of fermentation of CP relative to carbohydrate, which will limit gas production and ruminal hydrogen (H⁺) supply (Cone et al. 1999). Mathison et al. (1998) stated that increased nitrate levels in pastures can serve as an H⁺ sink and reduce enteric CH₄ production from ruminants. In a study of perennial ryegrass cultivars, Lovett et al. (2004) reported a decrease in TGP and CH₄ production with increasing N-fertiliser application rates with short in vitro incubations.

In vitro techniques have been used by several researchers as a practical screening tool to predict plant digestibility, plant nutritive value, fermentation characteristics and methanogenic potential (CH₄ : TGP), taking into consideration the complex interaction between rumen microbes and feed particles (Lovett et al. 2006; Durmic et al. 2010; Banik et al. 2013; Durmic et al. 2017). Variability in these traits among accepted improved pasture species would allow for the selection of low-methanogenic pastures that do not compromise animal productivity. This would improve the ability of producers to reduce CH₄ emissions from livestock, reducing the carbon footprint of production systems and allowing more efficient and climate-friendly management without major changes in current production practices.

The aim of this study was to evaluate the influence of a range of N fertiliser application rates on the nutrient concentration, in vitro OM digestibility (IVOMD), in vitro TGP and in vitro CH₄ : TGP of commonly used improved subtropical (C₄) and temperate (C₃) grass species in South Africa.

The experiment was conducted in a glass greenhouse at the Hatfield experimental farm at the University of Pretoria, South Africa. Seven grass species of current economic importance in South Africa were investigated (Table 1) in two groups: four species in the C₄ group and three species in the C₃ group. Three rates of N fertilisation commonly used in South African pasture production systems were evaluated: 0, 50, and 100 kg N ha^–1. All treatments were replicated three times in a randomised complete block design. Seeds were sourced from a commercial company and sown into 15-L pots in a controlled environment where the temperature and humidity ranges were 18–34°C and 30–68%, respectively. The pots were filled with 12 kg air-dried and sieved potting soil mixture comprising 20% clay, 23% silt and 57% sand. Soil samples were analysed at a commercial, accredited laboratory (Nvirotek Laboratories, Hartbeespoort, South Africa). Some physical and chemical characteristics of the soil were: bulk density 1.1 g cm^–3, Bray 1 phosphorus 300 mg kg^–1, potassium 2554 mg kg^–1, sodium 556 mg kg^–1, calcium 3650 mg kg^–1, magnesium 616 mg kg^–1, and pH(KCl) 5.63.

Table 1.  Perennial grass species investigated, including common and scientific names, cultivar and photosynthetic pathway
Scientific name according to Gibbs Russel et al. (1991)

For each species, 10 seeds were planted per pot and allowed to germinate. Once established, the seedlings were thinned to three uniform seedlings per pot. All pots received a single dressing of N fertiliser as limestone ammonium nitrate (28% N) after the thinning process according to the experimental treatments. The pots were rotated once a week in the greenhouse to minimise the influence of environmental variation within the greenhouse. All pots were weighed and watered to 90% field capacity according to Pieterse et al. (1997). In order to prevent mineral loss, all pots were placed on a saucer and any leached water was returned to the pots 1 h after watering. For the remainder of the trial period, the pots were weighed every 3 days and watered to 90% field capacity.

Samples for analysis of nutritive value and in vitro fermentation were obtained from the second regrowth phase after an initial harvesting cycle in the establishment year. The initial growth period lasted for 6 weeks; thereafter, all pots were harvested and a soil core sample was taken from each pot by using a thin polyvinyl chloride pipe and analysed for mineral concentration, to ensure that all treatments had similar soil-nutrient composition before the N-fertiliser treatment was applied for the second regrowth phase. Both harvest cycles were done by hand at 5 cm above soil level. The second harvest was done after an 8-week regrowth period, when the C₄ species started to flower and the C₃ species were still vegetative. The harvested material was air-dried and ground to pass through a 1.0-mm screen. Material was stored at room temperature (20–25°C) in sealed containers for analysis.

Plant samples were analysed for dry matter (DM), OM, CP, NDF, acid detergent fibre (ADF), acid detergent lignin (ADL) and IVOMD, and then metabolisable energy (ME) was estimated. The DM content was determined by drying samples for 24 h at 105°C in a forced-air oven, after which the samples were weighed then combusted at 450°C for 8 h in a muffle furnace to determine the OM concentration (AOAC 2000). Nitrogen concentration of samples was analysed by total combustion (AOAC 2000) on an FP-248 N and protein analyser (LECO Corporation, St. Joseph, MI, USA). NDF and ADF concentrations were determined using a 200/220 Fibre Analyser (ANKOM Technology, Fairport, NY, USA) based on the methods described by Van Soest et al. (1991). Sodium sulfite and heat-stable amylase were used in the analysis of NDF. The ADL concentration was determined according to Van Soest et al. (1991) through the solubilisation of cellulose with sulfuric acid in the ADF residue. The fibre fractions were expressed inclusive of residual ash. The ME was calculated from gross energy (GE) and IVOMD according to Minson (1990) and Robinson et al. (2004): ME (MJ kg^–1 DM) = 0.81((GE × IVOMD)/100)

The IVOMD was determined using the method of Tilley and Terry (1963) as modified by Engels and Van der Merwe (1967). Three rumen-cannulated Döhne Merino wethers were used as rumen-inoculum donors. The care, handling and maintenance of cannulated sheep were in accordance with animal welfare regulations of the Animal Ethics Committee of the University of Pretoria (EC018-14). The donor sheep were fed a diet consisting of 50% Eragrostis curvula hay and 50% Medicago sativa hay. Rumen fluid was collected 2 h after the morning feeding, pooled and filtered through two layers of cheese cloth. The rumen fluid was stored in a pre-warmed insulated thermos flask pre-filled with CO₂.

Samples for gas analysis were incubated in triplicate according to the procedure described by Theodorou et al. (1994). Dried plant sample (~400 mg) was weighed into 120-mL serum bottles. Filtered rumen fluid (15 mL) was mixed with anaerobic buffer–mineral solution (30 mL) prepared according to Goering and Van Soest (1988) with modifications suggested by Mould et al. (2005). After saturation with CO₂ the serum bottles were sealed with rubber stoppers and aluminium crimp seal caps. Possible gas build-up was equalised by inserting a hypodermic needle through the rubber stopper for ~5 s. Thereafter the sample bottles were placed in an incubator at 39°C with a rotary shaker set at 120 rpm. The incubation and gas-production measurements lasted for 48 h (Gemeda and Hassen 2014) and all measurements were corrected for blank gas production (gas production in buffered rumen fluid without sample). The system consisted of a digital data logger (Tracker 220 series indicators; Omega Engineering, Laval, QC, Canada) connected to a pressure transducer (PX4200-015GI; Omega Engineering). Gas pressure was measured at 0, 4, 12, 24, and 48 h by using the pressure transducer. After each pressure reading, a small gas sample (2 mL) was taken from the headspace with a Hamilton gas-tight syringe for immediate CH₄ analysis by gas chromatography (490 Micro gas chromatograph; Agilent, Santa Clara, CA, USA). The gas chromatograph was equipped with a 10-m stainless-steel Porapak-Q column and a thermal conductivity detector. The injector temperature was set at 45°C and the column temperature at 50°C, with a 30-ms injection time and static pressure of 80 kPa. Methane content (mL g^–1 DM incubated) was calculated according to Banik et al. (2013).

The two groups of grass species (three species in the C₃ group and four in the C₄ group) were analysed separately. The data were subjected to analysis of variance with two factors (species and N application) and three block replications, using the GLM procedure in SAS version 9 (SAS Institute, Cary, NC, USA). The Shapiro–Wilk test was performed on the standardised residuals to test for deviations from normality (Shapiro and Wilk 1965). In cases where there were significant deviations from normality and it was due to skewness, outliers were removed until the distribution of the residuals was normal or symmetrical (Glass et al. 1972). Student’s t-l.s.d. (least significant difference) was calculated at P = 0.05 to compare means of significant source effects.

Both grass species and level of N fertilisation had significant (P < 0.05) effects on the nutritive value, IVOMD and in vitro gas-production characteristics of the grass species (Tables 2 and 3). Grass species × N fertilisation level interactions were significant for NDF and ADF concentrations in C₄ species (Table 2) and for IVOMD in C₃ species (Table 3).

Table 2.  Analysis of variance for forage quality factors (dry-matter basis) for selected subtropical (C₄) grass species
Sp, Species; N, nitrogen fertilisation; CV, coefficient of variation; CP, crude protein; NDF, neutral detergent fibre; ADF, acid detergent fibre; ADL, acid detergent lignin; IVOMD, in vitro organic matter digestibility; ME, metabolisable energy; TGP, in vitro total gas production; CH₄, in vitro methane production

Table 3.  Analysis of variance for forage quality factors (dry-matter basis) for selected temperate (C₃) grass species
Sp, Species; N, nitrogen fertilisation; CV, coefficient of variation; CP, crude protein; NDF, neutral detergent fibre; ADF, acid detergent fibre; ADL, acid detergent lignin; IVOMD, in vitro organic matter digestibility; ME, metabolisable energy; TGP, in vitro total gas production; CH₄, in vitro methane production

Level of N fertilisation had no effect (P > 0.05) on in vitro gas-production parameters of either C₃ or C₄ species, except for CH₄ production of C₃ species after the 24-h incubation period (Table 3). In vitro gas production was not affected by species × level of N fertilisation interaction in either C₃ or C₄ species.

Nutritive analysis indicated that N fertilisation had an inconsistent effect on ash and NDF concentrations of both C₄ and C₃ species. Increasing N level decreased (P < 0.05) the ash concentration of Cenchrus ciliaris and Chloris gayana and the NDF concentration (P < 0.05) of C. gayana and Panicum maximum among the C₄ species, and decreased (P < 0.05) the ash and NDF concentrations of Dactylis glomerata among the C₃ species (Tables 4 and 5). No significant effect (P > 0.05) was shown on ADL concentration in either species group, whereas ADF concentration decreased (P < 0.05) in C. gayana and P. maximum and tended (P < 0.10) to decrease in other C₄ and C₃ species with increased N level. CP concentration increased (P < 0.05) with increasing N level across all C₄ and C₃ species except Digitaria eriantha. IVOMD increased in C. ciliaris and D. eriantha (Table 4) but decreased in D. glomerata (Table 5) as the level of N fertilisation was increased from 0 to 100 kg N ha^–1. ME concentration was not affected by level of N fertilisation in any of the grass species, although between-species differences were present across all N levels for C₄ species and at 100 kg N ha^–1 for C₃ species.

Table 4.  Effect of nitrogen fertilisation on chemical composition of improved subtropical C₄ grass species commonly used in South Africa
CP, Crude protein; NDF, neutral detergent fibre; ADF, acid detergent fibre; ADL, acid detergent lignin; IVOMD, in vitro organic matter digestibility; ME, metabolisable energy; MSE, mean square error; l.s.d., least significant difference; d.f., degrees of freedom. Within a column, means followed by the same letter are not significantly different (P > 0.05)

Table 5.  Effect of nitrogen fertilisation on the chemical composition of improved temperate C₃ grass species commonly used in South Africa
CP, Crude protein; NDF, neutral detergent fibre; ADF, acid detergent fibre; ADL, acid detergent lignin; IVOMD, in vitro organic matter digestibility; ME, metabolisable energy; MSE, mean square error; l.s.d., least significant difference; d.f., degrees of freedom. Within a column, means followed by the same letter are not significantly different (P > 0.05)

Comparing C₄ species, C. ciliaris presented lower (P < 0.05) CP concentration and higher (P < 0.05) NDF, ADF and ADL concentrations across all N treatments than P. maximum and D. eriantha. Panicum maximum had higher (P < 0.05) IVOMD than the other C₄ species at 0 and 50 kg N ha^–1 (Table 4).

There was less between-species variation for nutritive concentrations in the C₃ than the C4 species (Table 5). No differences were found for CP concentrations at 0 and 50 kg N ha^–1 among C₃ species, but Lolium perenne had higher (P < 0.05) CP concentration than D. glomerata and F. arundinacea at 100 kg N ha^–1. Dactylis glomerata had higher (P < 0.05) NDF concentration across the different N treatments than F. arundinacea and L. perenne. Across all N treatments, F. arundinacea had the lowest (P < 0.05) ADF concentrations and L. perenne had the highest (P < 0.05) IVOMD (Table 5).

For each of the C₄ species, there were no significant differences (P > 0.05) in any of the in vitro parameters (TGP, CH₄ and CH₄ : TGP) as the N-fertilisation level increased, after either the 24- or 48-h incubation period, except for C. ciliaris, which showed an increase (P < 0.05) in CH₄ : TGP at 100 kg N ha^–1 after the 24-h incubation interval (Table 6).

Table 6.  Effect of nitrogen fertilisation on in vitro total gas (TG) and methane production after incubation for 24 or 48 h of improved subtropical C₄ grass species commonly used in South Africa
l.s.d., Least significant difference; MSE, mean square error; d.f., degrees of freedom; Within a column, means followed by the same letter are not significantly different (P > 0.05)

Among C₄ grass species, D. eriantha and C. gayana had the lowest (P < 0.05) in vitro TGP and CH₄ production, respectively, in the control treatment (0 kg N ha^–1) after the 24-h incubation period. No differences were found in either TGP or CH₄ production at 50 kg N ha^–1 after the 24-h incubation period. As the level of fertilisation increased to 100 kg N ha^–1, C. ciliaris and P. maximum produced higher (P < 0.05) in vitro CH₄ after the 24-h incubation period than C. gayana (Table 6).

Among the C₄ species after the 48-h incubation period, P. maximum and C. ciliaris had the highest (P < 0.05) TGP at 0 kg N ha^–1. Chloris gayana had TGP lower (P < 0.05) than D. eriantha and P. maximum but similar to C. ciliaris at 50 kg N ha^–1, and there were no differences between C₄ species for CH₄ production at this level of N fertiliser, after the 48-h incubation. No between-species differences (P > 0.05) were observed for C₄ species in the 100 kg N ha^–1 treatment after the 48-h incubation (Table 6).

Considering the C₃ grass species evaluated, an increase in the level of N fertilisation increased (P < 0.05) in vitro CH₄ production after the 24-h incubation period for both D. glomerata and F. arundinacea, and after the 48-h incubation period for F. arundinacea. Level of N fertilisation had no effect on CH₄ : TGP for any of the C₃ species after the 48-h incubation, but after the 24-h incubation, increasing level of N fertiliser from 0 to 100 kg ha^–1 increased (P < 0.05) CH₄ : TGP for D. glomerata and F. arundinacea. No effects were found on any of the in vitro gas production parameters among N treatments for L. perenne (Table 7).

Table 7.  Effect of nitrogen fertilisation on in vitro total gas (TG) and methane production after incubation for 24 or 48 h of improved temperate C₃ grass species commonly used in South Africa
l.s.d., Least significant difference; MSE, mean square error; d.f., degrees of freedom. Within a column, means followed by the same letter are not significantly different (P > 0.05)

After the 48-h incubation, L. perenne had a higher (P < 0.05) CH₄ : TGP than D. glomerata and F. arundinacea at both 0 and 50 kg N ha^–1 and a higher CH₄ : TGP than D. glomerata at 100 kg N ha^–1. After the 24-h incubation, no differences among the C₃ species were found at 100 kg N ha^–1; however, F. arundinacea had a higher CH₄ : TGP ratio than L. perenne at 50 kg N ha^–1.

The objective of the study was to elucidate the influence of N-fertiliser application levels on the nutrient concentration, in vitro digestibility, in vitro TGP and CH₄ production of commonly used, improved C₄ and C₃ grass species in South Africa.

Increasing the level of N fertilisation increased the CP concentration of the C₄ and C₃ grass species, with the exception of D. eriantha. These results agree with results reported by Morison et al. (1980), Valk et al. (1996) and Warner et al. (2015). The CP concentration reported in the present trial for both C₄ and C₃ species are lower than previously reported values for similar species (Pieterse et al. 1997; Johnson et al. 2001; Taute et al. 2002; Navarro-Villa et al. 2012; Banik et al. 2013). This might be due to differences in N-application rates and growth periods after N application among the studies, as well as to differences in growth phases harvested between the present study and previous reports. Wilman (1975) reported that the N concentrations of pastures peak at 10–14 days after N application and thereafter decrease over time. The effect of N application on forage CP concentration in the present study could have been reduced by the 8-week regrowth period employed before harvesting. The age of forage, whether from initial or subsequent cuttings, has a negative effect on forage quality by increasing fibre components as well as decreasing digestibility and/or CP concentration (Salon and Cherney 2000).

Increasing the level of N fertilisation decreased (P < 0.05) the NDF and ADF concentrations of C. gayana and P. maximum but had no effect (P > 0.05) on the fibre fractions of C. ciliaris and D. eriantha (Table 4). Similarly, D. glomerata showed a decrease (P < 0.05) in NDF concentration with increasing level of N fertilisation (Table 5). The fibre fraction of F. arundinacea showed a tendency to decrease with increasing N fertilisation, but level of fertilisation had no effect (P > 0.05) on the fibre fractions of L. perenne. The inconsistent influence of N fertilisation on NDF, ADF and ADL concentrations is similar to the findings of Minson (1990) and Valk et al. (1996) who reported that the physiological stage of development has a greater influence on fibre fractions of forage than does the level of N fertilisation. A similar inconsistent effect of N on forage fibre fraction was reported by Peyraud and Astigarraga (1998), who concluded that the nutrient-composition response of forages to N fertilisation is species-specific.

Increasing the level of N fertilisation increased IVOMD of two of the C₄ grass species, C. ciliaris and D. eriantha, but no significant effect was found for C. gayana and P. maximum. Similarly, Johnson et al. (2001) reported an increase in IVOMD for star grass (Cynodon nlemfuensis) fertilised with increasing levels of N, and Taute et al. (2002) reported no effect of N fertiliser on IVOMD of P. maximum. The IVOMD of C₃ species was not affected by level of N fertilisation, except in D. glomerata, which showed a decrease as level of N fertilisation increased. These results are similar to those of Valk et al. (1996) and Lovett et al. (2004) for L. perenne. The decrease in the digestibility of D. glomerata can be explained by a slight increase in the lignin concentration with increased N fertilisation (Table 5). The increase in ADL concentration could be related to an increase in DM yield in response to N fertilisation, as reported by Lovett et al. (2004) and Cui et al. (2016), although DM yield was not recorded in the present study. Peyraud and Astigarraga (1998) also reported that N fertilisation increased the tiller : leaf ratio of forages, which could have a negative effect on forage digestibility. Nitrogen fertilisation can, however, have an indirect positive effect on digestibility by enabling earlier utilisation of grass forage. A higher level of N application allows for grass to be harvested at an earlier physiological age owing to increased growth response and yield (Peyraud and Astigarraga 1998). This could lead to increased intake and production from livestock and thus a reduced CH₄ intensity (CH₄ per unit product) of the pastures. However, these aspects of N fertilisation were not explored in the present study.

Methane production from forages depends on both NDF concentration and forage digestibility, which are the two main drivers of H⁺ production from carbohydrate fermentation in the rumen (Archimède et al. 2011). The gas production values reported in Table 6 are similar to gas production values reported by González Ronquillo et al. (1998) for C₄ grass species. In the present experiment, C. ciliaris and P. maximum produced the highest average in vitro CH₄ values across all N treatments after the 24- and 48-h incubation periods. This corresponds with higher NDF concentration and IVOMD (Table 4) of these species than of D. eriantha and C. gayana. These results correspond with those of Gemeda and Hassen (2014) and Doreau et al. (2016), who reported a positive correlation among CH₄ production, cell-wall contents and IVOMD of forages. Digitaria eriantha had the lowest average in vitro CH₄ production after the 24- and 48-h incubation periods and tended to have lower CH₄ : TGP than other C₄ species in the present trial. This could be attributed to the lower IVOMD (P < 0.05) of D. eriantha, which might have a negative effect on voluntary forage intake and subsequent CH₄ intensity (CH₄ per unit product).

The significant increase in 24-h CH₄ and CH₄ : TGP of D. glomerata and F. arundinacea corresponds with a significant increase in CP concentration as level of N fertilisation increases, and a reduction in fibre fraction of the species (Table 5). These results differ from data of Johnson and Johnson (1995) and Lovett et al. (2004), which indicated a decrease in CH₄ production when feed protein concentration increased. The increase in 24-h gas production could have been due to changes in the degradability of the CP and fibre fractions resulting from an increase in N fertiliser as reported by Valk et al. (1996). Crude protein levels above the threshold of 70 g kg DM^–1, as reported in the present study, are considered to enhance microbial multiplication in the rumen, thus improving fermentation (Njidda and Nasiru 2010). The negative correlation between NDF concentration and in vitro gas production reported by Njidda and Nasiru (2010) and Meale et al. (2012) was not observed in the present study. This might have been due to the relatively high IVOMD of the species reported here. Increasing N fertilisation from 0 to 100 kg N ha^–1 had no effect on the in vitro TGP or CH₄ production of L. perenne after either incubation period. These results differ from those of Lovett et al. (2004), which showed a significant decrease in the in vitro gas and CH₄ production with increasing N-application levels to L. perenne. These differences might have been due to differences in the physiological age of the forages between the two trials.

In the present study, D. glomerata emerged as the C₃ species with the lowest CH₄ : TGP after the 48-h incubation, compared with F. arundinacea and L. perenne. Although this could be partly attributed to a reduction in CH₄ production, it might also indicate reduced overall ruminal fermentation potential. Dactylis glomerata had the lowest TGP after the 48-h incubation period. This reduced fermentation could be explained by the lower IVOMD of D. glomerata. This lower IVOMD in the present experiment could negatively influence DM intake through a reduced ruminal clearance rate, which could have negative implications for livestock productivity compared with F. arundinacea and L. perenne (Banik et al. 2013).

This study demonstrated significant differences in nutrient composition, digestibility, and in vitro gas-production characteristics among key South African improved pasture species. Although increasing N fertilisation levels affected the nutrient composition of the pasture species, the effect on in vitro CH₄ production per unit DM digested was limited. Between-species differences were found for 24-h in vitro CH₄ production in C₄ and C₃ species but these differences diminished after the 48-h incubation period. The data suggest that reductions in enteric CH₄ production are unlikely to be achieved through increased N-fertilisation levels alone. Chloris gayana and D. glomerata showed potential for reduced CH₄ output from C₄ and C₃ grass species, respectively, compared with the other species evaluated. However, these results are based on in vitro analysis and from hand-harvested samples grown in a greenhouse. There is a need for further assessment of fermentation characteristics and management practices of these species at various stages of maturity, and an in vivo evaluation is necessary before any species can be promoted as a low-methanogenic pasture.

The authors declare no conflicts of interest.