The intensification of animal production and increased size of animal production units are now trends in livestock activity worldwide, representing a considerable pollution hazard through accumulation of high amounts of animal waste (Holm-Nielsen et al. 2009). The main emissions within the farms include enteric CH4 and CH4 from housing facilities during long-term storage (Rotz, 2017). Although the concentration of CH4 in the atmosphere is lower than that of CO2, CH4 has a heating potential 25 times more than that of CO2 (IPCC, 2007). The global emission of greenhouse gases (GHG) from manure grew between 1961 and 2010 from 0.57 to 0.99 gigatonnes of carbon dioxide equivalent (GtCO2eq) per year. On average, emissions grew by 1.10% per year (IPCC, 2014), but despite these data, Lynch (2019) concluded that there are still insufficient data available to fully address important questions regarding the climate impacts of agricultural production. The handling and use of manure on livestock farms contributes to emissions of GHG (Petersen, 2018). Comparing gas emissions from two typical manure handling options at cattle feedlots (composting and static stockpile storage), Bai et al. (2020) found that composting inhibits CH4 emissions but promotes NH3 and N2O emissions. Certainly, the efficient treatment of animal waste can support environmental protection in addition to bioenergy management (Achinas et al. 2018). Anaerobic digestion is a biological process that can convert organic substrates to biogas (Zhang et al. 2016). It is characterised by reactions in which biogas is produced from biodegradable products in the absence of oxygen (Neshat et al. 2017). Anaerobic digestion is increasingly used worldwide to generate energy from biogas and brings significant economic and environmental benefits (Scarlat et al. 2018) by being an efficient alternative technology that combines biofuel production with waste management (Achinas et al. 2017). Biogas is mainly comprised of CH4 and CO2 and minor amounts of other gases, such as nitrogen, hydrogen sulfide, ammonia and water vapour (Neshat et al. 2017). Monensin and tannins, separately, have shown to reduce enteric CH4 emission from ruminants (Russell and Houlihan, 2003; Aboagye et al. 2019; Stewart et al. 2019). Monensin’s mode of action is by reducing Gram-positive microorganisms, the major producers of methanogenic substrates as the final fermentation products (CO2, acetate, hydrogen, etc.) (Russell and Houlihan, 2003). There are three major forms by which tannins reduce enteric CH4 emission. The first is by reducing methanogenic Archaea; the second is through reduction of Archaea associated rumen protozoa, and the third is through depression of fibre digestion in the rumen (Carrasco et al. 2017; Tseu et al. 2020). Methane production from enteric fermentation of ruminants generates feed gross energy losses ranging from 2 to 15% (Johnson and Johnson, 1995; Wanapat et al. 2015). Thus, the effect of monensin and tannins on reducing CH4 production contributes to enhance feed energy efficiency. As seen above, the reduction in CH4 production by tannins is highly linked to reduction of fibre digestibility, so, according to Patra and Saxena (2011), the reduction in CH4 production compensates the loss of feed energy only if the tannin content in the diet is low or moderate (usually less than 50 g/kg DM), otherwise the loss of nutrients in faeces may be of high magnitude. In addition to reduce fibre digestion, tannins may also reduce protein digestibility (Tseu et al. 2020) by forming complexes with these macromolecules and make them inaccessible to microbial and enzymatic digestion (Nigrant et al. 2017). Therefore, the decreased nutrient digestibility is expected to increase fermentable organic matter (referenced in this study as volatile solids) concentration in faeces, which can promote a great anaerobic biodigestion for biogas production including CH4 (Hristov et al. 2013). Although much is known about the effects of these additives on rumen fermentation, studies reporting their effects on the fermentation of waste from cows (or other kind of ruminants) that have been fed these additives were not found. Hence, the overriding question was whether or not the effect of these additives on the reduction of enteric CH4 production provides conditions for the emission of GHG from waste. Given these factors, the hypothesis tested in this study was that the combined use of monensin and tannins to feed cows would increase the production of CH4 and CO2 from waste. Therefore, the study aimed to evaluate the effect of combined use of Acacia mearnsii tannins and monensin on biogas production from the waste of cows fed these additives by means of anaerobic biodigestion as a way to manage animal waste.
MATERIALS AND METHODS
Ethical issue and place of experimentation
The experiment followed the guidelines established in accordance with the ethical principles of animal experimentation of the commission of ethics in the use of animals of the college of animal science and food engineering of the university of Sao Paulo (USP) under the protocol number CEUA 3080240518. It was carried out at the Animal Nutrition and Production Department of the College of Veterinary Medicine and Animal Science of USP in Brazil.
Treatments and experimental design
The experiment was carried out in two phases, (1) the feeding phase and (2) the anaerobic digestion phase. In the first phase (the feeding phase), eight Nellore cows, non-pregnant and non-lactating, carrying a rumen cannula and having a mean body weight of 582 kg (±96), were kept in a roofed shed in individual pens with free access to water, feed and sand bedding. They were distributed in two contemporary 4 × 4 Latin squares design (LSD) in a 2 × 4 factorial arrangement and received eight experimental diets which differed in the levels of tannin inclusion (0.00, 0.75, 1.50 and 2.25% of feed DM) and the inclusion or not of monensin (Rumensin® 200, Elanco Animal Health, Brazil), i.e. no monensin was administered in the second square, but tannins at 0.00, 0.75, 1.50 and 2.25%; in addition to receiving tannins (at the same levels) in the first square, the cows also received monensin (300 mg/day, about 32 mg/kg DM). Kaolin was added as the tannin level decreased from 2.25 to 0.00% to equalise the DM in all treatments. The tannins, from a commercial extract, were obtained from the bark of A. mearnsii (Seta Natur®-Seta Acacia Tannin Extract). The concentration of total phenols (84.40%) was determined by the Folin-Ciocalteau method (Makkar, 2003a), and total tannins (82.30% tannic acid equivalent) were estimated by the difference in total phenol concentration before and after treatment with insoluble polyvinylpolypyrrolidone (Makkar et al. 1993). The concentration of condensed tannins (32.30% leucocyanidine equivalent) was determined by the HCl-butanol method (Makkar, 2003a). The feed was offered at 8 a.m. and 4 pm. in the form of total mixed ration with a ratio of 50% of corn silage and 50% of concentrate. The feed consumption was ad libitum. The proportions of ingredients and the chemical composition of the diets are shown in Table 1. The feeding phase was carried out in 4 periods of 24 days each, but the last two days of each period the cows spent together in pasture. The first 16 days were to adapt the animals to the diets, and between days 17 and 21 the collection of faeces for anaerobic digestion phase was performed. The collection of faeces was performed twice a day (8 am. and 4 pm.) by hand. All the faeces corresponding to the same cow per period were mixed and frozen in a single bag until biogas tests were conducted. On day 24 of each period urine was collected every 6 hours (at 6 am., 12 p.m., 6 pm. and 12 am.), obtained either during spontaneous urination or stimulation by vulva massage. All the urine corresponding to the same cow per period was mixed and frozen in a single flask until biogas tests were conducted. In the second phase (the anaerobic digestion phase), the 32 samples of faeces and 32 samples of urine collected and frozen during the feeding phase, i.e. samples from 4 × 4 latin square design with 8 treatments (tannin inclusion levels of 0.00, 0.75, 1.50 and 2.25% in both squares and 32 mg of monensin per kg DM for each cow only in the first square), were thawed and diluted in water. Firstly, a mixture of faeces and urine (waste) was prepared by using a theoretical ratio of 83:17%, respectively. Then, this mixture was diluted in water, and finally, the inoculum was added, composing a substrate. Hence, the substrate composition was as follows: 40% of waste, 3.30% of inoculum and 56.70% of water. The inoculum was sewage sludge from waste treatment and had 0.16% total solids (TS). Accordingly, the substrates were prepared to ensure an estimation of 6% of TS as per Lucas Junior et al. (1993), who found better biogas production in batch-type biodigesters when the TS content of the substrate was less than 8%. Batch-type biodigesters were used (Figure 1), and 3 kg of substrate were prepared, 2 kg of which were used to fill the biodigesters and 1 kg to perform the characterisation analyses of the substrate (Table 2). The biodigesters were arranged in a completely randomised design (CRD) in a 2 × 4 factorial arrangement of 8 treatments with 4 replicates, totalling 32 experimental units (represented by faeces and urine of the animals which received the different levels of tannins and the inclusion or not of monensin in the diet). After filling, the biodigesters were conditioned in a climate chamber with controlled temperature (33±2 ˚C) by electric resistance heating system and digital temperature recorder to guarantee that the test occurred in mesophilic conditions, ideal for digestion kinetics (Metcalf and Eddy, 2014). The temperature was monitored through a digital thermometer (in ˚C), and the readings and records were made immediately before the biogas reading. The composition of the substrates in the different biodigesters is shown in Table 2.
Quantitative production of biogas through biodigesters
The batch-type biodigesters consisted of three straight cylinders with diameters of 15, 10 and 7.50 cm, with a mean capacity to ferment 2 litres of substrate each (Figure 1). The 15 and 7.50 cm cylinders were inserted one inside the other so that the space between the outer wall of the inner cylinder and the inner wall of the outer cylinder contained a volume of water (water seal), reaching the depth of 60 cm. The cylinder of intermediate diameter (gas meter) had one of the ends sealed to retain a record for biogas discharge while it was capsized in the water seal to provide anaerobic conditions and to store the produced gas. The reading of biogas production was performed according to the accumulation in the gas meter. It consisted of the height measured by the measuring tape attached to the gas meter according to the vertical displacement. The reading value was multiplied by the internal cross-sectional area of the gas meter. After each reading, the gas meters were emptied by using the biogas discharge register. The correction of the biogas volume for the conditions of 1 atm at 20 ˚C was carried out according to the methodology described by Lucas Junior (1994). The correction of the biogas volume was performed through the expression resulting from the combination of the laws of Boyle and Gay-Lussac:
(V0P0) / T0= (V1P1) / T1 (1)
V0: corrected biogas volume, m3 or L.
P0: corrected biogas pressure, 10322.27 mm H2O.
T0: corrected biogas temperature, 293.15 K.
V1: gas volume in the gas meter.
P1: biogas pressure at the time of reading, 10344.11 mm H2O.
T1: biogas temperature, in K, at the time of reading.
Table 1 Proportions of ingredients and chemical composition of experimental diets
1 Mineral mixture, quantity per kg of product: calcium: 140 g; Phosphorus: 80 g; Sulfur: 10 g; Sodium: 129 g; Cobalt: 80 mg; Copper: 1400 mg; Fluorine: 800 mg; Iodine: 80 mg; Manganese: 1 g; Selenium: 20 mg and Zinc: 3.50 g.
2 Extract of Acacia mearnsii with 82.30% of total tannins, of which 32.30% of condensed tannins.
3 Determined through chemical analysis (non-fibre carbohydrates (NFC)=100–(% NDF+% CP+% EE+% Ash)).
4 Estimated by the spartan dairy ration evaluator/balancer software, version 3.0.3.
Table 2 Composition of substrates of anaerobic batch-type biodigesters supplied with the waste of nellore cows fed monensin (mg/kg DM) and different levels of tannins of A. mearnsii
TS: total solids; VS: volatile solids; N: nitrogen and NDF: neutral detergent fibre.
L: linear effect and NS: non-significant.
SEM: standard error of the means.
Considering the average atmospheric pressure of Pirassununga (Sao Paulo–Brazil) equal to 10273.11 mm H2O and the pressure conferred by the gas meters of 71 mm H2O, the following expression was obtained to correct the biogas volume:
V0 = (V1/T1) × 293.7703 (2)
Biogas sampling was performed whenever the biogas volume was measured.
Samples were collected by using a 60 mL syringe connected to the gas register at the top of the gas meter. Then 50 mL of biogas, for analysis, were injected in collecting flasks (glass flasks of 50 mL of capacity, Frascolex, Sao Paulo, Brazil). The gas meters were then emptied to allow a new accumulation of gas. The test was terminated when the biogas production ceased, i.e. there was no more displacement of the gas meter. The concentration of CH4 and CO2 was determined by gas chromatography (Trace 1300, Thermo Fisher Scientific®, Rodano, Milan, Italy) in controlled temperature (25 ˚C) according to Kaminski et al. (2003).
Figure 1 Diagram showing the batch-type biodigester design
The biogas samples were diluted in glass flasks, with a known volume, 16.78 times in atmospheric air. Then, 6 mL were injected into the chromatograph injector (split/splitless), 4 mL of which were used to wash the injection system and 2 mL were used for analysis. One (1) mL was also used for the system with a flame ionisation detector (FID), responsible for the measurement of CO2 and CH4. The chromatograph was calibrated with 3.10% CH4 and 3.10% CO2 that was diluted in atmospheric air. One gaseous mixture was used as a reference, one with (50% CH4 to 50% CO2) in balance with helium (He) (mol/mol). Helium with a flow rate of 30 mL/min was used as the dragging gas. The volumes of CH4 and CO2 produced (m3 or L) were calculated using the production data and biogas composition of each digester according to the equation:
Vol= (VolBIOGAS×% Gas) / 100 (3)
Vol: volume (m3 or L).
VolBIOGAS: volume of biogas produced (m3 or L).
% Gas: content of gas of interest in biogas.
The production of CH4 or CO2 was calculated by dividing the total production of each gas by the amount of VS added or removed (the difference between VS added in the filling time of the biodigesters and VS eliminated during the fermentation). The Gompertz model was used to study the biogas production kinetics and its components. The model assumes that the gas production rate is proportional to the microbial activity, but the proportionality decreases with the incubation time which can be interpreted as a loss of efficiency in the fermentation rate (Lavrencic et al. 1997). The mathematical description of the gas production curves allowed the data analysis, the substrate comparison, and the performance of the fermentation. The following equation describes the model used:
Yt= A exp [-B exp (-kt)] (4)
Yt: gas production (L/g VS added) at time t (days).
A: asymptote of the model, indicating the stabilisation value of the production (L/g VS added) in relation to time t.
B: integration constant, with no biological meaning.
kt: growth rate, logarithmic function of the production growth (L/g VS added) per unit of time.
The time (t) at inflection point was determined as follows:
t1= ln B / k (5)
t1: time (days) at inflection point (inflection point is the point at which the production rate is maximum and after which production tends to stabilise).
ln: natural logarithm.
B: integration constant.
k: production constant.
The gas production at inflection point was determined as:
y1= A / exp (6)
y1: gas production at the inflection point.
A: asymptotic gas production.
exp: base of natural logarithm (2.7183).
The substrates added (before biodigestion) and recovered residue (after biodigestion) in each biodigester were weighed and multiplied by their DM content in percentage to calculate the DM content in grams. The added or recovered nutrients, expressed in grams, were calculated by multiplying between the added or recovered, and expressed as grams of DM, then were expressed as a percentage and divided by 100 according to the following equation:
Nutrient (g)= (added or eliminated / bio digested nutrient (%) × DM) / 100
The nutrient removal, in percentage, was calculated by using the added and recovered nutrient content and expressed in g/kg of DM according to the following equation:
Removed nutrient (%)= (added nutrient (g)–recovered (g)/added nutrient (g)) × 100
The samples of the substrates before and after anaerobic digestion were collected and dried in an oven with ventilation and constant air renewal at 65 ˚C for 72 hours, according to AOAC (1995). Then, they were milled with wily-type knives in 1 mm sieves and stored in properly sealed vials. The DM was determined at 105 ˚C for 16 hours in the oven (method 930.15; AOAC, 1995). The mineral matter (MM) was obtained by calcination in a muffle oven at 550 ˚C for 5 hours (AOAC, 1990). The TS (TS=100-humidity) and VS (VS=TS-MM) contents of the substrates were determined with adaptations to the methodology described in APHA (2005). The total nitrogen (N) content was determined by the micro-Kjeldahl technique (method 920.87; AOAC, 1990). The Neutral detergent fibre (NDF) was determined by the method described by Van Soest et al. (1991). The hydrogen ion potential (pH) was measured by portable pH meter (Hanna Instruments®, HI 8424, Italy).
Biogas production was obtained in each biodigester by biogas measurement for about 6 months (175 days). The frequency of biogas measurement was performed following gasometer capacity and speed of gas production. For this reason, production and time for filling were considered as variables over time, not allowing to perform statistical analysis in repeated measurement. In this way, gas production over time was used to run Gompertz model using non-linear procedures (PROC NLIN) in the SAS software (SAS, 2013). Gompertz model provided not only total gas production, but many other information such as growth rate and inflection point. The data obtained in Gompertz model, as well as other data, were all analysed by using SAS. Before the data were analysed, they were evaluated in relation to the presence of discrepant information (outliers) and normality of the residues by the Shapiro-Wilk test. When the normality premise was not met, the data were transformed. They were next submitted to analysis of variance, which separated, as causes of variation, the monensin effect, tannin level effect, and the interaction between monensin and tannin level (all as fixed effects). The tannin level effect was evaluated by the use of orthogonal polynomials separating the effects in linear, quadratic, and quadratic deviation. A significance level of 5% was adopted. The statistical model used was described according to the equation below:
Yijkl= μ + Mi + TLj + Mi × TLj + eijkl
Yijkl: observation concerning monensin (i) + tannin level (j) + monensin (i) × tannin level (j) + random error associated with each observation (eijkl).
μ: overall mean.
Mi: effect of monensin (fixed effect).
TLj: tannin level effect (fixed effect).
Mi × TLj: interaction between monensin (i) and the tannin level (j) (fixed effect).
eijkl: random error associated with each observation.
RESULTS AND DISCUSSION
Biodigestion and nutrient removal efficiency
There were no significant differences (P>0.05) in the amount of TS, VS or nitrogen (N) on the substrates (i.e. initial TS, VS and N) corresponding to the different treatments (Table 3). A significant interaction was observed between monensin and tannins (P<0.05) either in the TS or VS recovery (i.e. the amount of nutrients not used or fermented during biodigestation) and VS removal efficiency (Figures 2, 3 and 4, respectively). Monensin significantly reduced TS and VS removal efficiency by 29.40 and 29.00% (calculated values), respectively, but no significant effect was observed for N. The different levels of tannins had a quadratic effect on the TS recovery and consequently in removal efficiency, but on VS and N the recovery linearly increased and the removal efficiency linearly reduced. The increase in VS and N recovery was 28.30% and 41.80%, respectively, for the highest level of tannins compared to the control treatment. Monensin increased pH substrate during biodigestion, but although tannins linearly increased the substrate pH before (Table 2), it was quadratic during anaerobic digestion.
The theoretical, non-significant biogas production (in general) was observed about 160 days after biodigesters were filled. Therefore, the biodigestion process was interrupted on day 175. No significant interaction (P>0.05) was observed between monensin and tannins on biogas production parameters (Table 4). Monensin did not significantly alter the total biogas or CH4 production, but it reduced CO2 production (L) by 18.90%.
Table 3 Biodigestion and removal efficiency of nutrients from anaerobic batch-type biodigesters supplied with the waste of cows fed monensin (mg/kg DM) and different levels of tannins of A. mearnsii
TS: total solids; VS: volatile solids; N: nitrogen and NDF: neutral detergent fibre.
L: linear effect; Q: Quadratic effect and NS: non-significant.
SEM: standard error of the means.
Table 4 Gas production (total biogas, CH4 and CO2) in batch-type biodigesters with the waste of cows fed monensin (mg/kg DM) and different levels of tannins of A. mearnsii
A: asymptotic production (L/g added VS); k: production constant (L/g added VS per day); t: time at inflection point (day) and y: production at inflection point (L/g added VS).
L: linear effect; Q: Quadratic effect and NS: non-significant.
SEM: standard error of the means.
Monensin increased the production of CH4 by 60.70% per gram of VS removed during the biodigestion process, but it did not significantly alter the stabilisation value (A) for CH4 nor the growth rate (k) or the time to reach the inflection point (t) for CH4 or CO2 production. Monensin significantly reduced the stabilisation value for CO2 production by 33.30% (Table 4 and Figure 5). The different levels of tannins had a quadratic effect on total biogas and CH4 production. The highest level of tannins reduced the total biogas and CH4 production by 36.90 and 36.10%, respectively, when compared to the control treatment. Tannins linearly reduced the stabilisation value (A) as well as the growth rate (K) for CH4 production, but it showed a quadratic effect on the time to reach the inflection point, and the inclusion levels of 0.75 and 1.50% reached the inflection point significantly later than the control and 2.25% treatments. Tannins linearly reduced CO2 production as well as the stabilisation value (A) and the growth rate (K). No significant effect of tannins was observed in the production of CH4 and CO2 per gram of VS removed (Table 4). The production ratio of CH4 and CO2 (CH4:CO2, L/L) was neither significantly affected by monensin (although there was a tendency) nor by the level of inclusion of tannins (Table 4). The substrate pH before anaerobic biodigestion ranged between 6.33 and 6.73 (Table 2), but after anaerobic biodigestion, it ranged between 7.52 and 7.67 (Table 3). This shows a pH increase during biodigestion process that supports Rabiu et al. (2014), Mshandete et al. (2006) and Gunaseelan (1995), who stated that the pH of a normal and healthy anaerobic biodigestion system for CH4 production is generally in the range of 7.00 to 8.50. Hence, the significant effect of monensin on increasing pH (from 7.53 to 7.61) during biodigestion indicates that this ionophore created better pH conditions for CH4 production compared to the different levels of tannins, as tannins reduced the pH during biodigestion (although it was not of great magnitude in relation to the values of monensin), but it remained within the optimum range cited by the above authors. Perna Junior (2018), who worked with waste of Nellore and Holstein cows fed tannins of A. mearnsii up to 1.50% DM basis, also observed linear reduction of biodigestion pH. The increased concentration of VS and NDF in faeces, and consequently in the substrates (g/kg) (Table 2), may have occurred because tannins reduce nutrient digestibility by forming complexes, which make nutrients inaccessible to microbial and enzymatic digestion in gastrointestinal tract (Patra and Saxena, 2011; Nigrant et al. 2017). This effect was also observed by Perna Junior (2018). The effects of monensin on the reduction of nutrient removal efficiency (Table 3), CO2 production (Table 4), the increase of CH4 and CO2 production per gram of VS removed may be the indication of direct effect of monensin on anaerobic biodigestion. The effects of tannins on increasing nutrient recovery and reducing nutrient removal efficiency as well as the reduction of total biogas, CH4 and CO2 production may also be the indication of direct effect of tannins on anaerobic biodigestion. The hypothesis for this study was that both additives would improve anaerobic biodigestion by increasing the concentration of nutrients in faeces (due to the effect of reducing enteric digestion of organic matter mainly caused by tannins). Therefore, the reduction of the performance of biodigestion may suggest that significant concentrations of these additives or their bioactive metabolites might have been present in faeces. Hao et al. (2011), adding 25 g/kg of A. mearnsii condensed tannins (i.e. 2.50% of inclusion) to cattle diets, found increased agronomic value of the manure and compost as fertiliser, but found no increase in the production of CH4 and CO2. Perna Junior (2018) found no differences in CH4 and CO2 production; the only difference he observed was a linear increase in the concentration of CO2, an effect not observed in the present study. There is some evidence that most bioactive metabolites of monensin are eliminated via faeces in ruminants. Davison (1984), investigating whether or not monensin was absorbed, metabolised and eliminated through the bile of calves and other animal species, found that most of the consumed monensin was recovered in faeces but had a minimal recovery in urine and tissues. Determining the excretion pattern and tissue distribution of [14C] monensin in cattle, Herberg et al. (1978) recovered almost 95% of active monensin metabolites in faeces. Hydrolysable tannins may be degraded and metabolised by rumen microorganisms (Mcsweeney et al. 2001), but condensed tannins are not degraded in the rumen. In addition, the complexes (tannin-protein or tannin-fibre) formed in gastrointestinal tract may not be reversible and hence eliminated in faeces (Makkar, 2003b). Therefore, the high recovery rates of monensin and tannins in faeces may suggest that these additives (mainly tannins) negatively affect the fermentation of faeces, at least when fresh (Hamilton et al. 2010). It is still not clear how long monensin and condensed tannins or their bioactive metabolites remain active to hinder anaerobic biodigestion or rumen fermentation. Using a rumen simulation technique (RUSITEC), Makkar et al. (1995) exposed rumen microbes to small amounts of quebracho (Schinopsis spp.) tannins for 8 days to induce enzymes capable of degrading condensed tannins, but there was no degradation.
Figure 2 Graph showing the interaction between monensin (M) and tannins on the amount of TS recovered after anaerobic biodigestion. The square points in bold represent the means observed in the different tannin levels only in the biodigesters whose substrates were also treated with M. In these biodigesters, the joint effect of M and tannins was not significant; therefore, it was chosen to present the general mean observed (dashed line). The empty square points show the means observed in the different tannin levels in biodigesters whose substrates were only treated with tannins (quadratic effect). The continuous line passing over the empty squares shows the estimated means for the biodigesters whose substrates received M and tannins if they had not received M (quadratic effect). Then, it may be observed that when monensin was administered jointly with tannins, the effect of tannins was not observed. This suggests that monensin blocked the effect of tannins by antagonistic interaction
Figure 3 Graph depicting the interaction between monensin (M) and tannins on the amount of VS recovered after anaerobic biodigestion. The square points in bold represent the means observed in the different tannin levels only in the biodigesters whose substrates were also treated with M. In these biodigesters, the joint effect of M and tannins was not significant, then it was chosen to present the general mean observed (dashed line). The empty square points show the means observed in the different tannin levels in biodigesters whose substrates were only treated with tannins (quadratic effect). The continuous line shows the estimated means for the biodigesters whose substrates received M and tannins if they had not received M (linear effect). Therefore, it may also be observed that when monensin was administered with tannins, the effect of tannins was not observed, suggesting inhibition of tannin effect by monensin through antagonistic interaction
Figure 4 Graph depicting the interaction between monensin (M) and tannins on VS removal efficiency. The square points in bold represent the means observed in the different tannin levels only in the biodigesters whose substrates were also treated with M. In these biodigesters, the joint effect of M and tannins was not significant, then it was preferred to present the general mean observed (dashed line). The empty square points show the means observed in the different tannin levels in biodigesters whose substrates were only treated with tannins (quadratic effect). The continuous line passing over the empty squares shows the estimated means for the biodigesters whose substrates received M and tannins if they had not received M (quadratic effect). As a consequence of the effect depicted in the former graph, it may be observed that it was necessary to increase the concentration of tannins to cause-effect. Hence, it appears to be obvious that when monensin was administered along with tannins, the effect of tannins was not observed. This also suggests that monensin inhibited the effect of tannins by antagonistic interaction
Figure 5 Graph adjusted by the Gompertz model depicting the time (in days) and cumulative CO2 production (M: monensin; ■ inflection point and ▲stabilisation value)
Makkar (2003b) reported degradation of purified quebracho and A. nilotica condensed tannins within 7 days, but this occurred under aerobic conditions in artificial fermenters. It is difficult to believe that even if they are present in faeces, monensin and tannins or their bioactive metabolites, can remain active up to 175 days (the time duration of biodigestion in the present study) and thereby continually impair the kinetics of biodigestion and reduce the performance of the fermentation. This suggests that there must be some process whereby if biodigestion is disturbed, by the presence of bioactive substances or other factors, future biodigestion performance is also impaired. The reduction of nutrient utilisation by microorganisms is a characteristic of tannins (Tseu et al. 2020). The increased nutrient recovery observed in the present study was the consequence of the reduction of the removal efficiency of TS, VS, and N (Table 3), i.e. the reduced capacity to biodigest nutrients caused by tannins, even though, the VS removal efficiency, which ranged between 16.55 to 30.03%, was only below the range stipulated by Dohányos and Zábranská (2001) (25-50%) when monensin and up to 2.25% of tannins were included in the diet. Although monensin has significantly reduced the nutrient (TS and VS) removal efficiency (Table 3), the interactions observed between this additive and tannins (Figures 2, 3 and 4) suggest that monensin has the potential to reduce the effect of tannins by an antagonistic interaction. It may be understood from Figures 2, 3 and 4 that if monensin and tannins had not been administered simultaneously, the effect of tannins in reducing the VS or TS biodigestion would have been of higher magnitude. The reduction of the total CO2 production (L) by monensin was accompanied by the reduction of the stabilisation value of the production (A) (Table 4 and Figure 5), but it did not affect the production growth rate (k) and the time to reach inflection point, i.e. the point at which the production rate is maximum and after which production tends to stabilise. This may have been due to the fact that monensin reduced microbial capacity to remove nutrients; therefore, it was not possible to reach the production potential. According to IPCC (2006), the specific productivity of fermentation products is measured in terms of removed VS. Besides the reduction of CH4 production, the tannins also reduced the production growth rate and the stabilisation value. This may have been the reason why the average production of CH4 (0.25 L) per gram of added VS was below the production found by Møller et al. (2004) (0.40 L) and Perna Junior (2018) (0.34 L). The lack of effect of monensin and the different levels of inclusion of tannins on the ratio of CH4 and CO2 production shows that although these additives may reduce the production of these two gases, the reduction is directly proportional. Monensin and tannins are feed additives used to reduce the emission of enteric CH4 (Finlay et al. 1994; Guan et al. 2006; Odongo et al. 2007; Patra and Saxena, 2011; Carrasco et al. 2017; Nawab et al. 2020), this effect is beneficial in two aspects, the first is in increasing the feed energy efficiency and the second is in protecting the environment by reducing CH4 emission into the atmosphere. Several researchers, such as Carrasco et al. (2017) and Tseu et al. (2020), have reported that the reduction in the production of enteric CH4 (through the use of tannins, for example) is usually at the expense of reducing the enteric digestion of nutrients, leading to the faecal excretion of large amounts of nutrients such as VS. This fact led to believe that the use of these additives in the ruminant feeding is seen as beneficial in the context of animal feeding, but due to the higher concentration of VS in faeces, it may be providing better conditions for greater emission of gas, including CH4, from the waste (an effect poorly studied by many researchers). Many studies, such as Orhorhoro et al. (2017), point out that the production of gas through anaerobic biodigestion increases with the increase in the concentration or amount of VS in the substrate. Therefore, the overriding question to conduct this research was whether or not the effect of these additives on the reduction of enteric CH4 production provides conditions for the emission of GHG from waste. Nonetheless, the results revealed that the use of monensin and tannins in cattle feeding presents no risk of increasing emissions of GHG from waste (although tannins have increased faecal VS) and it is believed that they or their bioactive metabolites might have been present in faeces and hindered the biodigestion. If so, monensin or monensin’s bioactive metabolites reduced Gram-positive microorganisms, the major producers of methanogenic substrates as the final fermentation products (CO2, acetate, hydrogen, etc.) (Russell and Houlihan, 2003) and the tannins or tannin bioactive metabolites may have reduced gas, including CH4, production by reducing methanogenic microorganisms and directly through depression of fibre biodigestion (Carrasco et al. 2017; Tseu et al. 2020). Therefore, the combined or isolate use of these additives in cattle feeding can be a sustainable way to continue ruminant production with less environmental pollution hazard. Then, the hypothesis that the combined or isolate use of monensin and tannins in cow feeding increases the production of CH4 and CO2 from waste (tested in the present study) might not be accepted if tannins are included at a rate above 0.75%.
Monensin (32 mg/kg DM) did not affect total biogas and CH4 production, but reduced CO2 production by 18.90%. Tannins had the potential to reduce total biogas and CH4 production when the inclusion level in the diet was more than 0.75% of feed DM, but for CO2 the potential to reduce was linear. Monensin reduced the tannin effect (of reducing the volatile solids or total solids biodigestion) by antagonistic interaction. Monensin and tannins may be used in cattle feeding with no risk of increasing greenhouse gas (e.g. CH4 and CO2) production from the waste. According to the results observed in the present study there is a strong evidence that monensin and tannins or their bioactive metabolites may appear in faeces and impair the biodigestion of the waste. Therefore, further studies should be carried out to confirm (by observing effects or by identifying) the presence of these additives or their bioactive metabolites in the faeces.
To the Fundação de Apoio à Pesquisa do Estado de Sao Paulo – FAPESP (Process 2017/05301-0) and Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq, Brazil) for providing the financial support. The authors also thank Gilmar E. Botteon for the good care of animals (University of Sao Paulo, Pirassununga, Sao Paulo, Brazil). The main author thanks the Ministry of Science and Technology, Higher and Professional Technical Education of Mozambique for the scholarship and the Eduardo Mondlane University for the opportunity to do the master’s degree.