Influence of supplementing Guinea grass with differently processed African yam bean on gas production and in vitro digestibility

This study evaluated the potential utilisation of African yam bean (AYB) seed as a supplement to Guinea grass on in vitro gas and methane (CH4) production, as well as the effect of processing on AYB nutritive value. In experiment 1, unprocessed AYB meal at 10, 15, 20, and 25% inclusion levels was added to Guinea grass substrate and evaluated for in vitro gas production, CH4, and in vitro organic matter digestibility (IVOMD). In experiment 2, the effect of soaked, boiled, toasted, and fermented AYB meal at 20% inclusion on in vitro fermentation was evaluated. In vitro gas production as well as in vitro organic matter digestibility of Guinea grass increased with AYB supplementation. The associative effect of Guinea grass with AYB showed an increase in gas and CH4 production. At 20% inclusion level, AYB processing methods did not affect the gas production, CH4, and IVOMD of the substrate. Fermentation improved the crude protein, iron, and zinc contents, reduced acid detergent fibre concentration but did not reduce the concentrations of alkaloid, total phenol, saponin, and trypsin inhibitors of AYB. Seed soaking for 48 h reduced the total phenol, tannin, oxalate, and phytate content, while seed boiling reduced the concentrations of alkaloid, total phenol, tannin, and trypsin inhibitors. Compared with the raw AYB, boiling is the most effective processing method to reduce the investigated phytochemicals, followed by soaking, toasting, and fermentation, in that order.


Introduction
The utilisation of few crop species for the production of most of the world's food supply has been noted as a limitation due to excessive homogeneity, oversimplification of farming systems, and disruptions of the ecological balance (FAO, 2015). This has serious negative implications on sustainable agriculture such as decreased resilience to climate change impacts, thus necessitating crop diversification (Bhartiya et al., 2015). Restrictions to the feeding of genetically modified soybean to livestock across the European Union has also heightened the demand for alternative vegetable protein sources (Nábrádi and Popp, 2011;Johnston et al., 2019), and increased production of tropical legume seeds could help meet this demand aside from the reduction in synthetic fertilizer use, which could guarantee a more sustainable farming system (Sinclair and Vadez, 2012;FAO, 2015). Improved utilisation of tropical legume seeds also offers opportunities to further develop the crop-livestock production system.

Material and Methods
The study followed the guidelines for the use and conduct of experiments with live animals and was approved by the Animal Ethics Committee with approval number TETFund/2016-2017/BATCH11/No3. African yam bean seeds were purchased from the local market in Ikole Ekiti, Ekiti State, Nigeria. The seeds were cleaned manually by physical selection to eliminate dirt and shrivelled or damaged seeds before processing. The AYB seeds were processed in a laboratory in Ikole Ekiti, Ekiti State, Nigeria (7.7983° N, 5.5145° E, 560 m above sea level) in April 2019.
The procedure by Obatolu et al. (2007) was followed in seed processing. From the same batch of AYB seeds, different samples were prepared using soaking, boiling, fermentation, or toasting as processing methods. A seed sample was soaked in distilled water (1:10, w v −1 ) for 24 and 48 h at room temperature, drained and dried in a forced-air oven at 60 ℃ as soaked AYB. A weighted sample of the AYB seeds was boiled in hot water (100 ℃) for 60 min, drained, and dried in a forced-air oven. A sample of AYB was toasted at 120 ℃ on a hot plate using an open pan for about 15 min under continuous stirring to prevent charring of the seeds and ensure uniform toasting. On observing light brown colouration -an indication of good toast -, the seeds were removed from the heat and cooled to room temperature, and few charred seeds were removed by handpicking before storage. Weighted AYB seeds were also soaked in distilled water for 24 h, the water was drained, and the seeds were spread in a covered plastic container lined with dried jute bags. The container was filled with seeds to a depth of 7 cm, airtight-sealed, and placed in a warm (34±2 ℃) place for 96 h to create the appropriate condition for fermentation. The fermented seed samples were opened and oven-dried at 60 ℃ for 18 h. All processed seeds were thereafter milled through 1-mm sieve and stored in a cool dry place.
In two separate experiments, graded levels of raw AYB seeds and differently processed AYB seed meals were evaluated as supplements along with Guinea grass as main roughage. The Guinea grass (Megathyrsus maximus Jacq.) was from an eight-week old re-growth and presented: crude protein (CP), 86.2 g kg −1 ; neutral detergent fibre (NDF), 544.8 g kg −1 ; acid detergent fibre (ADF), 458.3 g kg −1 ; and acid detergent lignin (ADL), 86.4 g kg −1 .
Four replicate bottles per treatment were incubated, and each treatment was repeated in three incubation runs. In the first experiment, the treatments were arranged as grass only, grass + AYB meal (90:10, w w −1 ), grass+ AYB meal (85:15, w w −1 ), grass+ AYB meal (80:20, w w −1 ), grass + AYB meal (75:25, w w −1 ), and AYB meal only. In the second experiment, the effect of supplementing soaked, boiled, toasted, or fermented AYB seed meal with guinea grass was evaluated at grass + AYB meal Influence of supplementing Guinea grass with differently processed African yam bean on gas production... Sanni et al. 3 (80:20, w w −1 ). The treatments included grass + soaked AYB meal 24 (AYB soaked for 24 h), grass + soaked AYB meal s48 (AYB soaked for 48 h), grass + boiled AYB meal (AYB boiled in hot water for 60 min), grass + toasted AYB meal, and grass + fermented AYB meal. In both experiments, 400 mg of diet (grass or grass plus AYB meal) was incubated with 40 mL of rumen liquor. The in vitro fermentation procedure of Menke et al. (1979) as modified by Theodorou et al. (1994) was followed. We used rumen liquor from four cannulated beef steers that were fed Eragrostis and Lucerne hay ad libitum. Feed samples were incubated in buffered rumen fluid, and gas production was measured at 3, 6, 12, 24, and 48 h (Adejoro and Hassen, 2018) using a pressure transducer (PX4200-015GI, Omega Engineering Inc.) and a data tracker (Tracker 220 series indicators; Omega Engineering Inc., Laval, QC, Canada). Sample of gas at each incubation sampling time was taken using a syringe and stored in vacuum glass vials. Gas samples were analysed for CH 4 in a gas chromatograph (8610C BTU Gas Analyser GC System, SRI Instruments GmbH, Bad Honnef, Germany), pre-equipped with a flame ionisation detector.
The in vitro organic matter digestibility (IVOMD) of AYB as affected by processing method was determined in the two incubation experiments using the two-phase digestion method of Tilley and Terry (1963) with details described in Adejoro and Hassen (2018). During the first stage, 200 mg of feed samples were incubated in four replicates of each diet with rumen liquor for 48 h at 39 ℃ under anaerobic conditions. Blanks and a standard substrate were included in each batch. This was followed by an acid-pepsin digestion phase at 39 ℃ for 48 h. The residual material was oven-dried at 105 ℃ for 18 h, weighed, and subsequently ashed in a muffle furnace at 550 ℃ for 3h. Digested in vitro organic matter was estimated from the weights of starting material and residuals.
Samples of raw, soaked, toasted, boiled, or fermented AYB seeds and Guinea grass were analysed according to AOAC (2000) for dry matter (DM; ID 934.01), total ash (ID 942.05), CP (ID 968.06), ether extract (ID 934.01), and starch (ID 934.01). Crude fibre (CF) and NDF were determined using the Fibertec system technique as described by Van Soest et al. (1991), with the addition of heatstable alpha-amylase and sodium sulphite, while ADF and ADL were also analysed (non-sequential) using the Fibretec system procedure. Both NDF and ADF were expressed exclusive of residual ash. Individual amino acid concentration of raw and processed AYB was analysed according to the methods of Benitez et al. (1989) using the PTH Amino acid analyser (120A PTH, Applied Biosystems Inc., CA. 94404, USA). Sodium and potassium were determined using a flame photometer (Coming, 403, UK), while other minerals were analysed using atomic absorption spectrophotometer (Perkin-Elmer model 403, walk CT, USA). All determinations were performed in duplicate. Total phenol and total tannins were determined by the by Folin-Ciocalteu spectrophotometric method using tannic acid as standard (Porter et al., 1986). The quantitative evaluation of trypsin inhibitors and phytate was determined according to the procedure described by Mbithi-Mwikya et al. (2000), while alkaloid, oxalate, and saponin contents were equally determined using standard procedures (Adejoro et al., 2013).
Samples were analysed in duplicates in three repetitions for each processing method. For the in vitro gas production, the four bottles for each treatment in each incubation run served as analytical replicates, and each repeat incubation run served as a statistical replicate, making a total of 12 replicates per treatment. The associative effect between Guinea grass and AYB seed supplementation on fermentation parameters was calculated as percentage difference between the values measured for each inclusion level and the balanced median values of their components as follows: in which the calculated value was obtained from the observed values for the 100% Guinea grass and the 100% AYB incubated individually. Positive or negative values indicated positive or negative associative effects at the inclusion levels, respectively. Data were analysed using the general linear model of SAS (Statistical Analysis System, version 9.4). The model statement was of the form: in which Y ij = mean of individual observation, μ = overall mean, B i = block effect (incubation runs), T j = effect of treatment, and e ij = residual error. Mean separation was conducted using Tukey's test, Influence of supplementing Guinea grass with differently processed African yam bean on gas production... Sanni et al. and significance was declared at P<0.05. A linear and quadratic trend comparison of increasing AYB inclusion was done using Contrast statement of SAS.

Results
Increasing the inclusion levels of AYB seed meal increased 48 h gas production (P<0.05) ( Table 1). African yam bean alone produced a higher total gas compared with the grass alone or grass supplemented with AYB. Similarly, 48 h in vitro CH 4 production was affected by AYB inclusion (P<0.05). The 100% AYB recorded a higher CH 4 volume compared with the 100% Guinea grass, while the grass with AYB supplementations were intermediate. The effect of AYB inclusion on IVOMD of Guinea grass showed that supplementation at 15-25% inclusion level resulted in increased IVOMD (P<0.01) compared with the grass-only diet. Methane to total gas ratio was affected by AYB supplementation of Guinea grass (P<0.05) with CH 4 concentration higher in AYB-supplemented diets compared with the grassonly diet. Methane per unit of organic matter digested (CH 4 /IVOMD) was equally affected by AYB supplementation with grass supplemented with AYB and 100% AYB treatment producing higher CH 4 per unit of organic matter digested. The associative effects of Guinea grass and AYB inclusion on 48 h gas production, CH 4 volume, and CH 4 concentration showed that there was no linear or quadratic association between the grass and AYB across the inclusion levels (Table 2). However, there was a negative linear and quadratic associative effect (P<0.05) on IVOMD. Similarly, the association between the grass and AYB inclusion levels was not different in terms of the CH 4 to IVOMD ratio across the inclusion levels.
In the second experiment, AYB seed processing methods did not affect 48 h in vitro gas production when Guinea grass was supplemented with 20% AYB meal (Table 3). The effect of AYB seed processing method when supplemented to Guinea grass showed that processing method did not affect CH 4 production, although fermented AYB produced numerically lower CH 4 , but methane to total gas ratio  2 Treatment -main associative effect of inclusion levels; L -linear associative relationship of increasing levels (10, 15, 20, and 25%) of AYB; Q -quadratic associative relationship of increasing levels (10, 15, 20, and 25%) of AYB. Positive or negative values indicate positive or negative associative effects between Guinea grass and African yam bean mixture. a-b -Mean values followed by different letters in the same row are significantly different by Tukey's test at 5% probability.
Influence of supplementing Guinea grass with differently processed African yam bean on gas production... Sanni et al. 5 was not different across the treatments. Although IVOMD was not different across the treatments, CH 4 per unit of IVOMD showed differences across the treatments (P<0.05). Grass with fermented AYB produced lower CH 4 per unit of organic matter digested compared with grass with soaked AYB treatment.
Seed processing method affected CP, EE, ash, CF, total carbohydrate, and NDF content of the AYB meal (P<0.05) ( Table 4). Processed AYB meals had higher CP compared with the unprocessed meal, with fermented seeds having the highest CP followed by toasted, boiled, and soaked seed meal. Boiling reduced EE composition of AYB meal compared with the unprocessed meal (P<0.05). Fermentation reduced the total carbohydrate content of AYB seed meal (P<0.05) while soaking, toasting, and boiling did not differ from the unprocessed seeds in terms of total carbohydrate concentration. Neutral detergent fibre concentration of AYB seed meal was influenced by processing methods (P<0.01) with toasted > soaked > boiled > fermented > raw seed meal in descending order. Processing methods did not affect ADF and ADL concentration of AYB seeds. Seed processing methods did not affect calcium, magnesium, potassium, and manganese contents of AYB meal, while iron and zinc concentrations were affected (P<0.05), with fermented or toasted AYB having higher iron concentration than soaked, boiled, and raw AYB.
Saponin concentration was not different across treatments, while concentrations of alkaloid, total phenol, total tannin, oxalate, phytate, and trypsin inhibitors were influenced (P<0.05) by seed  Influence of supplementing Guinea grass with differently processed African yam bean on gas production... Sanni et al. 6 processing method (Table 5). Boiling reduced the alkaloid content of AYB meal, while boiling, soaking, and toasting reduced the total phenol content. Fermented and unprocessed AYB seeds were similar in alkaloid, tannin, and total phenol concentrations. Soaking, boiling, and fermentation reduced the oxalate concentration, but toasting had no effect on oxalate concentration. Boiled and unprocessed seeds were similar in phytate concentration, while soaked, toasted, and fermented seeds had lower phytate concentration. Concentration of trypsin inhibitors in AYB seeds was not affected by soaking and fermentation, but toasting and boiling reduced its concentration in descending order of magnitude.
Leucine was the most abundant amino acid in AYB seed meal. The concentration of amino acids in AYB seed meal was not affected by processing methods, except for methionine and arginine (P<0.05) ( Table 6). Fermentation improved the methionine and arginine concentrations in fermented AYB, while soaking and boiling reduced the concentration of these two amino acids. Fermented seed meal had higher methionine and arginine concentrations than unprocessed seeds (P<0.05).  Influence of supplementing Guinea grass with differently processed African yam bean on gas production... Sanni et al. 7

Discussion
The CP, NDF, and ADF contents of the Guinea grass forage reported in this study is within the range reported in literature, while lower values due to increased lignification associated with maturity and drought are common (Babayemi, 2009). However, the CP is close to the threshold of 8%, below which optimal rumen function may be hindered if fed alone (Ikhimioya, 2008). Therefore, supplementation with legume would enhance animal performance through improved digestibility and intake of the forage. Furthermore, increased use of home-grown legumes for livestock feeding justifies the inclusion of AYB because of its economic and ecological advantages such as promotion of organic farming, diversification, and adaptability to local farming systems (Campos-Vega et al., 2009;Bhartiya et al., 2015).
Previous studies on the replacement of soybean with legume seeds in the diet of sheep or dairy cows revealed the possibility for total or partial replacement with chickpea, faba bean, peas, field bean, or lupins (Bonanno et al., 2016;Ramin et al., 2017;Zagorakis et al., 2018). Studies on digestive interaction between grass and legumes (seeds or forages) revealed inconsistent results due to inclusion levels, nature of dietary components, and the possible presence of bioactive compounds. More significant associative responses on digestibility have been observed in legume supplementation of poorquality hay when bioactive compounds are involved (Niderkorn et al., 2011), but the phytochemicals investigated in AYB did not reflect this, probably due to their low concentration in the total substrate incubated compared with the 100% AYB. However, enhanced gas production has been associated with grass-legume supplementation in ruminant animals (Niderkorn et al., 2011;Adejoro and Hassen, 2018;Zagorakis et al., 2018), and current result agrees with this. Obatolu et al. (2007) described the proximate and functional characteristics of AYB. Little information exists in literature on AYB inclusion in the diet of ruminants either from in vitro or animal experimentation.
In the current study, raw and processed AYB seeds improved IVOMD and gas production significantly with up to 14% increase in total gas production. This result can be related to the report of Tadele et al., (2014), who observed improved in vivo digestibility of Rhodes grass when supplemented with lupin seeds, which resulted in improvement of growth of Washera sheep compared with those that received only grass. An increase in nitrogen supply to microbes through non-protein nitrogen (NPN) or legume supplementation, fed as either whole seed, ground seed, or whole plant silage has been reported to improve microbial activity, thus improving organic matter digestibility and in vitro gas production (Mitsumori and Sun, 2008). This leads to improved efficiency of fermentative digestion in the rumen and improved efficiency of feed conversion to products such as meat or milk (Ramin et al., 2017).
Beyond removing unwanted anti-nutritional factors in legume seeds, processing also improves the nutritive value and digestibility of seeds (Khattab et al., 2009). In terms of nutritive value, toasting and soaking did not affect CP digestibility of lupin seed, and values obtained in the present study were comparable with that of soybean meal (Tadele et al., 2014). However, Mogensen et al. (2008) and Vaga et al. (2017) observed that toasting could be used to decrease effective rumen protein degradability and improve bypass protein supply, which may be useful to dairy animals (Mustafa et al., 2003). Fibre digestion could be increased when the proportion of ruminal degradable protein in the diet is lower, due to increased rumen retention time and slow release of essential growth factors to cellulolytic microbes (Mitsumori and Sun, 2008;Gerber et al., 2013). However, the kinetics of microbial degradation of AYB meals as affected by processing methods needs further evaluation.
There is adequate evidence that the presence of tannins in diets can reduce enteric CH 4 production (Patra and Saxena, 2011;Adejoro et al., 2019). The use of legumes is recommended in the tropics as a means of abating overall animal-derived CH 4 production . Nevertheless, the presence of phytochemicals in AYB did not affect enteric CH 4 , and this may be due to the concentration or nature of phenolic compounds present in the incubated substrates. In a similar in vitro study by Ramin et al. (2017), CH 4 emission was neither affected by field bean inclusion as a substitute to soybean meal nor by the seed processing methods, although processed seeds had lower phenol and tannin concentrations. Supplementing legumes or NPN in nitrogen-limited diets often results in improved fermentability of the substrate, but with increased acetate and higher ruminal CH 4 production. However, improved growth in animals will translate into lower CH 4 intensity, expressed in volume of CH 4 per unit of animal product produced when grass is supplemented with legume .
Generally, AYB meal had lower CP compared with soybean (Zagorakis et al., 2018), but the CP value is close to other legume seeds such as peas, field bean, kidney bean, and cowpea (Khattab et al., 2009;Ramin et al., 2017). Likewise, lipid content of raw and processed AYB seeds was low but with a high amount of carbohydrate content similar to many other legume seeds (Doss et al., 2011;Zagorakis et al., 2018). Generally, legume seed proteins are known to be limiting in methionine and cysteine, high in leucine and with a moderate amount of other essential amino acids (Khattab et al., 2009), and this was observed in the present study. Processing methods employed in the present study slightly increased the CP content of AYB seeds, which could be as a result of potential leaching of soluble constituents of legume seeds into water, during soaking or boiling (Campos-Vega et al., 2009). Fermented AYB meal had the highest CP content compared with other methods, which could be due to an increase in microbial biomass during fermentation mobilising soluble carbohydrates for their metabolism and increasing microbial protein (Torres et al., 2006;Hefnawy, 2011). Although a slight increase in CP was observed following toasting or boiling, a negative effect on CP quality of legume seeds associated with higher acid detergent insoluble nitrogen concentration may occur depending on the duration of heat treatment (Vaga et al., 2017).
The concentration and the type of phytochemicals in legume seeds may differ due to genetic and agronomic factors under which they are sown (Mogensen et al., 2008). Soaking, toasting, and fermentation have been shown to significantly reduce these phytochemicals to varying degrees, depending on the characteristics of the antinutritional factors present (Fernandes et al., 2010). The significant reduction in oxalate and phytate in the soaked and boiled AYB observed in this study agrees with previous reports for lentil seeds (Hefnawy, 2011). Water-soluble phytochemicals such as oxalic acids, phytate, and hydrogen cyanide are easily destroyed or removed from feedstuff through soaking or boiling (Adejoro et al., 2013). Simple sugars and other water-soluble nutrients may equally be lost in this process, but the presence and strength of seed coat may affect the extent of leaching (Fernandes et al., 2010). Similarly, heat labile phytochemicals such as antitrypsin inhibitors are easily destroyed by the high temperatures during heating or boiling (Adejoro et al., 2013).
Seed fermentation did not significantly affect tannin concentration in AYB seeds, and this observation is similar to previous reports on ground bean and pigeon pea (Torres et al., 2006). Condensed tannins, which are more widely distributed in plant species, are not readily degraded by anaerobic microbes (Makkar, 2003). However, fermentation considerably reduced the concentration of phytate in AYB seeds, and this relates to the activation of phytase enzymes hydrolysing phytate. Phytate removal often results in increased bioavailability of minerals such as phosphorus, iron, zinc, and magnesium (Nikmaram et al., 2017), but the extent of phytate removal is dependent on the type of microorganism, the fermentation conditions, and the initial concentration of phytate in the raw seeds (Egounlety and Aworth, 2003).

Conclusions
Supplementation of Guinea grass with African yam bean seed meal up to 20% inclusion increases gas production and digestibility of substrate, but no significant reduction in CH 4 is observed. Soaking African yam bean in water for 24-48 h or boiling it at 100 ℃ for 60 min is sufficient to reduce most of the antinutritional factors present, thus reducing the potential toxicity associated with its consumption by animals. The improvement in the in vitro organic matter digestibility of Guinea grass justifies the inclusion of soaked African yam bean seed meal in ruminant forages especially when dietary protein becomes limiting.

Conflict of Interest
The authors declare no conflict of interest.