Evaluating dose–response patterns of a tannin extract blend on nutrient utilization and methane emissions in beef cattle

TL;DR

This study finds that adding tannin extract blends to beef cattle diets reduces methane emissions in a dose-dependent manner without harming nutrient use.

Contribution

The study identifies an optimal tannin extract dose for methane mitigation while maintaining nutrient utilization in beef cattle.

Findings

Methane emissions decreased quadratically with increasing tannin extract dose.

Optimal methane mitigation occurs at 0.20–0.22% of dry matter tannin extract supplementation.

Nutrient utilization and energy efficiency were not compromised at optimal tannin extract rates.

Abstract

Tannin extracts (TE) of isolated condensed or hydrolyzable tannins have been evaluated for their methane (CH4) mitigation potential in beef cattle. Despite the potential for a combination of tannin types to yield synergistic effects, the dose-response pattern and optimal supplementation rate of a TE blend remain unclear. Our objectives were to investigate changes in nutrient utilization and gas emission patterns in response to supplementation with a TE blend (Silvafeed ByPro; SILVATEAM, San Michele Mondovi, Italy), and to determine the optimal dose to minimize emissions in growing steers (308 ± 9.4 kg BW). Supplementation rates were 0.0%, 0.3%, 0.6%, and 0.9% of DM (TE0.0, TE0.3, TE0.6, and TE0.9, respectively) within a total mixed ration fed at 1.62% of BW (DM basis). Whole-animal gas exchange and total fecal and urine production were measured over 48 h using two open-circuit, indirect calorimetry respiration chambers in a 4 × 8 Latin rectangle design with four periods and eight animals. Blood and ruminal parameters were evaluated from samples collected from steers upon removal from chambers. Nutrient and water intake were not influenced (P ≥ 0.42) by TE rate. Fecal excretion of acid detergent fiber (ADF) displayed a dose-response pattern (quadratic P < 0.01, cubic P = 0.01) with increasing TE rate. Apparent digestibility of ADF decreased at an increasing rate (quadratic P < 0.01, cubic P = 0.01) with TE. There was a quadratic effect (P = 0.01) on ruminal propionate concentration, but no other ruminal parameters were influenced by TE rate (P ≥ 0.13). There was a quadratic pattern (P = 0.04) in fecal N excretion and N digestibility. The rate of TE did not affect urinary N excretion (P ≥ 0.58), but N retention tended to linearly decrease (P = 0.06) with increased TE and blood urea N tended (P = 0.06) to follow a cubic pattern. There was a quadratic (P ≤ 0.05) dose-response relationship for gaseous energy loss. Although oxygen consumption and carbon dioxide production were not influenced by TE rate (P ≥ 0.64), the respiratory quotient increased linearly (P = 0.02) with TE inclusion. However, no other energy partitioning was influenced by TE supplementation rate (P ≥ 0.18). Regardless of how it was expressed, CH4 production displayed a quadratic pattern (P ≤ 0.04) in relation to increased TE rate, with equations suggesting an optimal TE dose between 0.20 and 0.22% of DM for CH4 mitigation without compromising nutrient utilization or energy efficiency.

There is a clear dose–response pattern in methane emissions from beef cattle with tannin extract supplementation.

Figures

Dose–response relationship of tannin extract supplementation on A) daily CH4 production and B) CH4 yield.

Tables

Table 1. Ingredient and chemical composition of the total mixed ration utilized as the basal diet.

Itemsa	Basal diet, %
Ingredient composition, %DM
Bermudagrass hay	35.00
Cracked corn	39.50
Dried distillers’ grain	16.00
Molasses	6.00
Mineral	3.50
Chemical composition b
DM, %	87.38
CP, % DM	11.30
Soluble protein, % CP	25.32
aNDF, % DM	35.88
ADF, % DM	16.42
Lignin, % DM	2.81
Crude fat, % DM	3.71
Sugar, % DM	8.62
Starch, % DM	29.25
Ash, % DM	6.28
Calcium, % DM	0.73
Phosphorus, % DM	0.36
TDN, % DM	72.97
NEm, Mcal/kg	1.66
NEg, Mcal/kg	1.05
ME, Mcal/kg	2.56

Table 2. Dose–response relationship of tannin extract supplementation on intake and excretion parameters of steers fed a total mixed ration.

	Tannin extract, % of DM					P values 1
Items2	0.0	0.3	0.6	0.9	SEM	L	Q	C
SBW, kg	373	370	372	375	21.453	0.43	0.52	0.72
Water intake, kg/d	13.13	14.08	11.90	13.01	2.177	0.65	0.90	0.68
Water intake, % SBW	3.48	3.89	3.18	3.44	0.559	0.61	0.84	0.56
DMI, kg/d	5.89	5.87	5.84	5.93	0.398	0.66	0.42	0.55
OMI, kg/d	5.49	5.46	5.41	5.48	0.352	0.70	0.44	0.56
aNDFI, kg/d	2.39	2.39	2.37	2.40	0.108	0.88	0.67	0.73
ADFI, kg/d	1.16	1.16	1.15	1.16	0.063	0.91	0.74	0.78
Fecal DM, kg/d	1.65	1.66	1.74	1.70	0.119	0.28	0.52	0.57
Fecal DM, % SBW	0.44	0.45	0.47	0.45	0.017	0.45	0.59	0.69
Fecal DM, %	24.72	25.36	24.27	25.39	0.792	0.72	0.86	0.44
Fecal OM, kg/d	1.35	1.37	1.43	1.40	0.090	0.30	0.47	0.59
Fecal aNDF, kg/d	0.83	0.82	0.86	0.85	0.071	0.27	0.55	0.52
Fecal ADF, kg/d	0.52	0.53	0.58	0.61	0.055	<0.01	<0.01	0.01
Urine, kg/d	8.83	9.04	9.34	9.23	1.441	0.67	0.90	0.97
Urine, % SBW	2.33	2.45	2.45	2.42	0.301	0.77	0.90	0.97

Table 3. Dose–response relationship of tannin extract supplementation within a total mixed ration on feed digestibility and ruminal parameters.

	Tannin extract, % of DM					P values 1
Items2	0.0	0.3	0.6	0.9	SEM	L	Q	C
DMD, %	71.99	71.63	70.23	71.31	0.946	0.38	0.49	0.54
OMD, %	75.38	74.68	73.58	74.53	0.809	0.29	0.32	0.44
aNDFD, %	65.48	65.82	63.96	64.59	1.751	0.34	0.64	0.60
ADFD, %	55.70	54.51	50.41	47.71	2.928	<0.01	<0.01	0.01
pH	6.95	6.93	6.88	7.00	0.056	0.73	0.41	0.48
Redox	−357.70	−403.30	−389.33	−378.79	22.806	0.47	0.13	0.17
Protozoa, log10/mL	5.46	5.47	5.53	5.48	0.042	0.51	0.55	0.55
Total VFA, mmol/L	66.37	68.29	74.11	60.02	4.725	0.50	0.14	0.13
Acetate, mmol/L	39.36	39.70	42.29	34.57	2.771	0.29	0.14	0.16
Propionate, mmol/L	12.29	13.48	13.47	10.54	0.972	0.18	0.01	0.05
Butyrate, mmol/L	11.34	11.63	13.26	12.00	1.223	0.40	0.51	0.52
Isobutyrate, mmol/L	0.92	1.06	1.09	0.86	0.102	0.62	0.02	0.07
Acetate: Propionate	3.29	2.98	3.15	3.28	0.188	0.80	0.20	0.26
Lactate, mmol/L	0.23	0.25	0.37	0.17	0.069	0.87	0.17	0.09

Table 4. Dose–response relationship of tannin extract supplementation within a total mixed ration on N metabolism of steers.

	Tannin extract, % of DM					P values 1
Items2	0.0	0.3	0.6	0.9	SEM	L	Q	C
N digestibility, %	58.81	56.38	53.76	54.76	2.608	0.02	0.04	0.08
N intake, g/d	100.49	99.89	99.07	100.29	8.831	0.66	0.42	0.53
Fecal N, g/d	41.10	42.82	45.42	45.04	2.947	0.01	0.04	0.09
Fecal N, % N intake	41.19	43.62	46.24	45.24	2.607	0.02	0.04	0.08
Urine N, g/d	32.04	30.85	31.42	33.52	3.477	0.58	0.62	0.81
Urine N, % N intake	31.94	31.16	31.22	33.52	2.030	0.57	0.61	0.80
Fecal N, % N excreted	56.33	58.16	60.00	57.60	2.029	0.46	0.34	0.50
Fecal N: Urinary N	1.31	1.41	1.55	1.40	0.114	0.36	0.30	0.38
Retained N, g/d	27.35	26.23	22.23	21.74	4.852	0.06	0.17	0.29
Retained N, % N intake	26.87	25.21	22.55	21.24	3.234	0.06	0.18	0.34
BUN, mg/dL	6.77	6.10	7.04	6.56	0.612	0.79	0.91	0.06
NH3, mg/dL	2.59	2.51	2.78	2.55	0.192	0.84	0.86	0.61

Table 5. Dose–response relationship of tannin extract supplementation within a total mixed ration on steer energy metabolism using indirect calorimetry.

	Tannin extract, % of DM					P values 1
Items2	0.0	0.3	0.6	0.9	SEM	L	Q	C
RQ	1.12	1.12	1.13	1.13	0.007	0.02	0.09	0.07
GEI, Mcal/d	25.38	25.31	25.20	25.58	1.758	0.55	0.38	0.51
GEI, kcal/kg MBW	298.66	299.13	296.57	299.21	8.171	0.89	0.77	0.58
FE, Mcal/d	7.05	7.18	7.50	7.35	0.480	0.22	0.39	0.52
FE, Mcal/kg DM	1.20	1.23	1.29	1.24	0.040	0.30	0.36	0.48
FE, kcal/kg MBW	82.96	85.19	88.35	86.01	3.113	0.32	0.44	0.60
DE, Mcal/d	18.33	18.13	17.70	18.24	1.352	0.61	0.44	0.50
DE, Mcal/kg DM	3.11	3.08	3.03	3.08	0.052	0.37	0.42	0.55
DE, kcal/kg MBW	215.70	213.95	208.22	213.20	7.144	0.34	0.35	0.36
DE, % GE	72.21	71.45	70.21	71.26	0.960	0.33	0.39	0.52
UE, Mcal/d	2.84	2.96	2.95	3.01	0.406	0.64	0.89	0.96
UE, Mcal/kg DM	0.47	0.51	0.49	0.50	0.052	0.67	0.87	0.93
UE, kcal/kg MBW	33.01	35.17	34.23	34.99	3.807	0.69	0.90	0.95
GASE, Mcal/d	1.69	1.65	1.74	1.83	0.097	0.01	0.01	0.03
GASE, Mcal/kg DM	0.29	0.28	0.30	0.31	0.009	0.03	0.04	0.07
GASE, kcal/kg MBW	20.01	19.55	20.52	21.39	0.569	0.02	0.02	0.06
GASE, % GE	6.72	6.53	6.93	7.16	0.217	0.03	0.05	0.08
ME, Mcal/d	13.80	13.51	13.01	13.40	1.002	0.26	0.31	0.42
ME, Mcal/kg DM	2.35	2.29	2.24	2.26	0.065	0.20	0.34	0.54
ME, kcal/kg MBW	162.68	159.22	153.48	156.82	6.094	0.20	0.31	0.45
ME, % DE	75.49	74.24	73.85	73.54	1.667	0.27	0.52	0.73
ME, % GE	54.49	53.09	51.83	52.42	1.413	0.18	0.30	0.49
HE, Mcal/d	11.66	11.63	11.54	11.69	0.779	0.97	0.84	0.92
HE, Mcal/kg DM	1.98	1.98	1.98	1.97	0.025	0.74	0.94	0.98
HE, kcal/kg MBW	137.39	137.30	136.10	136.67	3.808	0.69	0.91	0.96
HE, % GE	46.03	45.87	45.94	45.66	0.596	0.69	0.92	0.97
RE, Mcal/d	2.14	1.88	1.47	1.72	0.407	0.27	0.42	0.57
RE, Mcal/kg DM	0.37	0.31	0.25	0.29	0.065	0.28	0.42	0.61
RE, kcal/kg MBW	25.29	21.92	17.38	20.15	4.541	0.26	0.40	0.58
RE, % ME	15.38	12.86	10.78	12.56	2.667	0.33	0.42	0.62
RE, % GE	8.46	7.22	5.89	6.75	1.510	0.28	0.41	0.60

Table 6. Dose–response relationship of tannin extract supplementation within a total mixed ration on steer gas exchange using indirect calorimetry.

	Tannin extract, % of DM					P values 1
Items2	0.0	0.3	0.6	0.9	SEM	L	Q	C
O2, L/d	2,264.94	2,257.76	2,237.35	2,266.63	149.900	0.91	0.83	0.90
O2, L/kg MBW	26.68	26.66	26.39	26.52	0.721	0.64	0.88	0.94
CO2, L/d	2,545.29	2,533.95	2,532.26	2,565.61	174.670	0.70	0.75	0.90
CO2, L/kg MBW	29.97	29.93	29.86	29.99	0.897	0.99	0.97	0.99
CH4, L/d	194.24	189.58	199.55	209.32	11.207	0.01	0.01	0.03
CH4, L/kg DM	33.22	32.30	34.28	35.44	1.040	0.03	0.04	0.07
CH4, L/kg MBW	2.29	2.24	2.35	2.45	0.065	0.02	0.02	0.06

Full text

Introduction

The global beef industry faces increasing pressure to enhance production efficiency while minimizing its greenhouse gas (GHG) emissions and environmental footprint (Tedeschi et al. 2015). Carbon dioxide (CO_2_) is the most abundant GHG in the atmosphere and is likely the overarching cause of global warming, as its release rate into the atmosphere has exceeded its removal rate (Manabe and Wetherald 1975). Methane (CH_4_) and nitrous oxide (N_2_O) are also critical GHG because, when expressed in terms of their CO_2_ equivalent (CO_2_e), they have a 100-year global warming potential of 27 and 273, respectively (Forster et al. 2021). Although the 2022 beef cattle herd was responsible for 24.8% of the total agricultural GHG emissions in the United States (EPA 2024), that equates to only about 2.4% of the country’s total anthropogenic emissions of CO_2_e (Tedeschi and Beauchemin 2023). Aside from the potential environmental impacts associated with CH_4_ emissions from beef cattle, CH_4_ production also represents a potential source of inefficiency to ruminant animals, as those consuming forage-based diets can lose as much as 12% of the energy consumed in the form of enteric CH_4_ (Johnson et al. 1993). Thus, the development and utilization of management strategies to reduce GHG emissions, particularly CH_4_, from beef cattle has been a considerable area of research in the industry. Nutritional management, such as the use of feed additives, may be one plausible strategy to mitigate CH_4_ emissions from beef cattle production. While feed additives can reduce CH_4_ formation, the spared energy is not directly available to the animal; it must be redirected into pathways that generate absorbable end products, such as volatile fatty acids (VFA), and the often discrepant results in animal performance when using these additives reflect this fact (Tedeschi and Beauchemin 2023).

Plant defensive chemicals, such as tannin extract (TE), have received considerable attention as a feed additive for beef cattle due to their natural rumen modulation (Tedeschi and Fox 2020; Tedeschi et al. 2021). Although present in the plant as a defense mechanism against predators and pathogens, tannins also can form complexes with proteins and carbohydrates, which may improve nitrogen (N) utilization and reduce CH_4_ production by ruminant animals (Patra and Saxena, 2010). Multiple studies have assessed the impact of supplementing beef cattle with isolated condensed (CT) or hydrolyzable (HT) tannins, administered as TE or via tannin-containing plants, on CH_4_ emissions and metabolic responses. However, fewer studies have investigated the potential synergism of supplementation with a TE blend containing both CT and HT. Schilling-Hazlett et al. (2024) reported that supplementation of a TE blend at 0.3% of dietary dry matter (DM) failed to reduce CH_4_ production in dairy heifers. However, previous work by Adams et al. (2025) evaluated different inclusion rates, up to 0.9% of DM, of a TE blend on in vitro gas production dynamics and found that a rate of 0.18% had the greatest potential to reduce CH_4_ production when incubated with a mixed diet. Despite this potential, the optimal supplementation rate for the combined use of CT and HT has yet to be determined in vivo. There is even more limited research integrating the effects of a TE blend across a comprehensive range of physiological and environmental parameters, and the dose-response pattern remains unclear. Thus, our objectives were to (1) investigate changes in nutrient utilization and gas emission patterns in response to varying rates of supplementation with a TE containing a blend of CT and HT, and (2) determine the optimal dose of TE for minimizing environmental impact. We hypothesized that TE supplementation would have a dose-dependent effect on gaseous emissions without compromising nutrient digestibility or other physiological responses.

Materials and methods

All animals and procedures used in this experiment were reviewed and approved by the Institutional Animal Care and Use Committee (AUP #2024-0111) at Texas A&M University. This study was conducted from July 2024 to November 2024 at the Nutrition and Physiology Center (Texas A&M University, College Station, TX).

Experimental design and data collection

A 4 × 8 Latin rectangle design consisting of four experimental periods and eight British crossbred steers (Initial BW = 308 ± 9.4 kg, 14 months of age) was used to evaluate the effects of a commercial tannin extract (TE; Silvafeed ByPro; SILVATEAM, San Michele Mondovi, Italy), composed of both quebracho condensed (Schinopsis lorentzii) and hydrolyzable (tannin acid) tannins (CT and HT, respectively) at 0.0%, 0.3%, 0.6%, and 0.9% of dietary dry matter (DM; TE_0.0_, TE_0.3_, TE_0.6_, and TE_0.9_), where percentages refer to the total TE. All animals were progeny of a single sire from a commercial herd. Each supplemental treatment was replicated by two animals within each experimental period. Fractions of CT, including total CT (TCT), and protein precipitable phenolics (PPP) were determined as described by Norris et al. (2020). As reported in Adams et al. (2025), the TE contained 40.17% TCT, of which was 19.08%, 20.62%, and 0.46% extractable, protein-bound, and fiber-bound CT (Terrill et al. 1992; Wolfe et al. 2008), respectively, and 24.75% PPP (Hagerman and Butler 1978; Naumann et al. 2014). The supplementation rates in this study were expressed on a total TE basis; however, as suggested in Adams et al. (2025), the rates were also converted to a TCT and PPP basis to facilitate interstudy comparisons. The supplementation rates on a TCT and PPP basis were, respectively, 0.0% and 0.0% (TE_0.0_), 0.11% and 0.06% (TE_0.3_), 0.21% and 0.13% (TE_0.6_), and 0.32% and 0.19% (TE_0.9_).

A total mixed ration (Table 1) was formulated using the Ruminant Nutrition System (RNS; Tedeschi and Fox 2020). Animals were offered the basal diet at 1.62% of body weight (DM basis) to limit average daily gain to approximately 0.5 kg/day throughout the study, ensuring that animal size remained compatible with the calorimetry system used. Feed was provided to animals twice daily at 0800 and 1600 hours. The supplemental treatment consisted of adding TE to the basal diet, which was pre-weighed and hand-mixed into the individual animal feed prior to the morning feeding. Animals were housed outside within two pens (n = 4 steers/pen; 9.1 × 12.2 m) equipped with Calan gate feed bunks (American Calan, Northwood, NH, USA) and automatic waterers. Within each experimental period, animals were adapted to their respective supplemental treatments over a 21-day adaptation period prior to data collection. On day 22 of each experimental period, pairs of steers were placed in open-circuit, indirect calorimetry respiration chambers for 48 h to measure gas exchange, as well as feed and water intake, and total fecal and urine excretion. Due to the availability of only two adjacent chambers, calorimetric measurements were collected from two steers at a time. Following each 48-h collection period, chambers were recalibrated before the next pair of animals entered. This cycle was repeated over 8 days to complete the metabolic and calorimetric measurements for all animals within each period.

Each experimental period spanned 23 days in total. On day 22, for a pair of steers, shrunk body weight (SBW) was recorded prior to their placement in a pre-determined single-stall, open-circuit respiration chamber. Respiration chambers were maintained at thermoneutral conditions (19 ± 1.9 °C; 45 ± 5.6% relative humidity) using a line voltage thermostat (Ranco Enterprises, Inc., Model #ETC-111000-000) and a dehumidifier (Hisense USA, Model #DH-70K1SDLE), and environmental conditions were monitored using wireless temperature and humidity data loggers (Omega Engineering, Inc., Model #OM-CP-RFRHTEMP2000A). In addition, each chamber was equipped with a water meter (Minol Brunata Worldwide, Model# Minomess 130) to measure water intake, a raised metabolism stand to allow for the collection of total fecal and urine excretion, and security cameras (Reolink, Model #E1 Outdoor) to monitor animal activity.

The indirect calorimetry system has been described previously by Crossland et al. (2018) and Norris et al. (2020). The respiration chambers operated via a mass flow system (Flowkit model FK-500; Sable Systems Int., North Las Vegas, NV), with ambient air (baseline) and air from each chamber sampled using a multiplexer (Respirometry Multiplexer V2.0; Sable Systems Int., North Las Vegas, NV) rotating every four minutes. During each four minutes, air samples were analyzed for O_2_, CO_2_, and CH_4_ (FC-10 O_2_ analyzer, CA-10 CO_2_ analyzer, and MA-10 CH_4_ analyzer; Sable Systems Int., North Las Vegas, NV, USA). Prior to data recording, the appropriate bank time for gas stabilization and the flow rate required to maintain maximal chamber CO_2_ concentrations between 0.45% and 0.47% were calculated based on animal SBW, dietary energy density, and the known volume of the respiration chambers. In addition, O_2_, CO_2_, and CH_4_ analyzers were calibrated using gases of known concentrations: N_2_ (99.99%; used as the zero gas) and SPAN gas mixture containing 19.4% O_2_, 1.1% CO_2_, and 0.1% CH_4_. Calibration was performed prior to each pair of animals entering the chambers for data collection and was based on the assumed baseline concentrations of ambient air (20.95% O_2_, 0.04% CO_2_, and 0.00% CH_4_). Water vapor was removed from the sampled gas using fresh Drierite desiccant (Hammond Drierite Co Ltd., Xenia, OH), and the rates of O_2_ consumption and CO_2_ and CH_4_ production (VO_2_, VCO_2_, and VCH_4_, respectively) were determined (Lighton 2008). Following each 23-d experimental period, animals underwent a 14-d washout period to prevent any carryover effects from the previous supplemental treatment. During each washout period, two O_2_ dilution tests were performed using the gravimetric N_2_ injection technique (Cooper et al. 1991). The expected O_2_ consumption, calculated as 20.95% of the injected N_2_ volume, was compared to the observed VO_2_, with acceptable recovery defined as 95% to 105% (Li et al. 2019). The average N_2_ recovery in our study was 99.25%.

Sample collection and analysis

Batch samples of the basal diet were collected daily during the last 10 days of each experimental period to represent the diet offered during data collection. In addition, residual orts during data collection were weighed, homogenized, and subsampled for further analysis. All feed and orts samples were dried at 55 °C for 48 h before being ground to pass through a 2-mm sieve using a Wiley mill (Thomas Scientific, Swedesboro, NJ). Ground feed and orts samples were shipped to Cumberland Valley Analytical Services (Waynesboro, PA) for chemical analysis of DM (Goering and Van Soest 1970), neutral detergent fiber with the addition of amylase and sodium sulfite (aNDF; Van Soest et al. 1991), acid detergent fiber (ADF; Method #973.18; AOAC 2006), lignin using sulfuric acid (Goering and Van Soest 1970), crude protein (CP; Method #990.03; AOAC 2006), soluble CP (Krishnamoorthy et al. 1982), ether extract (Method #2003.05; AOAC 2006), starch (Hall 2009), sugar (Dubois et al. 1956), ash (Method #942.05; AOAC 2006), a complete mineral panel (Method #985.01; AOAC 2006) using a Perkin Elmer 5300 DV ICP (Perkin Elmer, Shelton, CT), and calculated total digestible nutrients and net energy values using empirical equations (Weiss 1998).

Total fecal production during each 48-h period within the respiration chambers was weighed, homogenized, and duplicate 1-kg subsamples were collected and stored at −20°C. Fecal samples were dried in a forced-air oven at 55 °C for 72 h or until weight loss ceased, ground to pass through a 2-mm sieve, and analyzed for DM, NDF, ADF, organic matter (OM), total N, and gross energy (GE). Total urine collection in the respiration chambers was achieved using a large funnel topped with mesh filters, a plastic reservoir containing 600 mL of 3 N HCl solution (300 mL/d) to prevent ammonia (NH_3_) volatilization, and an external holding vat connected to a transfer pump (Crossland et al. 2018). Total urine production was weighed, and two subsamples (100 mL each) were stored at −20°C. To prevent gas measurement data from being compromised, total fecal and urine production were composited for the 48-h collection period. All feed, feces, and urine samples were analyzed for GE using a bomb calorimeter (Parr 6200 Calorimeter; Parr Instruments Co., Moline, IL), whereas total N was determined using the Dumas combustion method (Method #990.03; AOAC 2006) by Servi-Tech Laboratories (Amarillo, TX).

Approximately 500 mL of rumen inoculum was collected from each animal via esophageal tubing connected to a hand pump upon removal from the chambers and was stored in a pre-warmed 600 mL thermos. Immediately following collection, pH and redox potential were recorded, and the sample was allocated into duplicate tubes for preservation of NH_3_, VFA, and protozoa. As described by Norris et al. (2020), preservation methods were 2 mL of inoculum and 8 mL of 0.1 N HCl solution for NH_3_ analysis and 1 mL of inoculum and 10 mL of 200 proof ethanol for protozoa enumeration, both stored at −20°C. For VFA analysis, 10 mL of rumen inoculum was stored at −20°C without preservative. Ruminal NH_3_ concentrations were determined via colorimetric methods (Chaney and Marbach 1962), and protozoa counts were performed using methods described by Dehority (1984). For protozoa enumeration, a 1-mL aliquot of the diluted sample was added to a Sedgewick Rafter counting chamber with 0.5 mm square counting grids, and 25 random grids were counted for each sample using a Meiji MT4200L microscope (Meiji Techno Co., Ltd, Saitama, Japan) at 100× magnification. Concentrations of VFA were determined according to Weimer et al. (1991) using high-performance liquid chromatography (Shimadzu Scientific Instruments, Columbia, MD). The system was equipped with a temperature-controlled autosampler (Nexera SIL-30AC UHPLC Cooled Autosampler), a forced-air column oven (CTO-20A), and a UV absorbance detector (SPD-20A UV Detector), all from Shimadzu Scientific Instruments (Columbia, MD). An Aminex HPX-87H column (300 mm × 7.8 mm i.d.; Bio-Rad Laboratories Inc., Hercules, CA), along with its corresponding guard column (Bio-Rad Cation H), was used to separate peaks at 210 nm absorbance over a 100-min retention time.

At the time of rumen inoculum sampling, blood samples were also collected from each animal via jugular venipuncture into duplicate blood collection tubes containing sodium heparin (Greiner Bio-One GmbH, Austria). Following collection, the blood collection tubes were immediately inverted and placed on ice. Plasma was obtained by centrifugation at 3,000 × g for 20 min at 4 °C. The supernatant was aliquoted into duplicate polypropylene tubes and stored at −80°C until analysis. All samples were analyzed for blood urea nitrogen (BUN) concentration using colorimetric methods (kit #K024, Arbor Assays, Ann Arbor, MI). The intra- and inter-assay coefficient of variation were, respectively, 3.3% and 12.2%.

Energy partitioning and nitrogen balance

Gross energy was determined for all feed, fecal, and urine samples, as well as a representative sample of TE from each period, via bomb calorimetry (Parr 6200 Calorimeter; Parr Instruments Co., Moline, IL). Gross energy intake (GEI; Mcal/d) was calculated by multiplying the total GE of the offered basal diet and TE (specific to each treatment) by the amount of feed consumed (kg). Fecal energy (FE; Mcal/d) and fecal N (g/d) were calculated by multiplying the energy density (Mcal/kg) and N concentration (% N) of the feces, respectively, by the daily fecal output (kg/d). Similarly, urinary energy (UE; Mcal/d) and urinary N (g/d) were calculated by multiplying the energy density (Mcal/kg) and N concentration (% N) of the urine, respectively, by the daily urine output (kg/d). Gaseous energy (GASE; Mcal/d) loss was determined by multiplying CH_4_ produced (L/d) by the density of CH_4_ (0.6556 g/L at 25 °C and 1 atm) and its energy density (0.0133 Mcal/g). Heat energy (HE; Mcal/d) loss was determined using the equation HE (Mcal/d) = (3.866 × VO_2_) + (1.2 × VCO_2_) - (0.518 × VCH_4_) - (1.431 × Urinary N) as described by Brouwer (1965). Final energy partitioning values were calculated using the following equations: digestible energy intake (DEI; Mcal/d) = GEI - FE; metabolizable energy intake (MEI; Mcal/d) = DE – (UE + GASE); and retained energy (RE; Mcal/d) = MEI - HE.

Statistical analyses

All statistical analyses were performed using R 4.4.1 (R Core Team, Vienna, Austria). Dose-response relationships were evaluated using mixed-effects polynomial regression models fitted using the lmer function (lme4 package) to account for random intercepts of animal and period (Bates et al. 2015). Polynomial terms used raw powers of supplementation rate to reflect actual dose spacing. Estimated marginal means were calculated to interpret dose effects. Significance of polynomial terms was considered at *P *≤ 0.05, and tendencies were assumed at 0.10 ≥ *P *> 0.05. In addition, critical points of the fitted polynomials were used to identify doses corresponding to the minimum or maximum predicted response within the observed treatment range.

Results

Intake and excretion

The intake and excretion parameters, as affected by TE supplementation rate, are presented in Table 2. There were no differences in SBW, water intake, or DM and nutrient intake (*P *≥ 0.42) among treatments. In addition, TE supplementation rate did not impact fecal DM, OM, and aNDF excretion (*P *≥ 0.27) or urine excretion (*P *≥ 0.67). However, there was a significant dose-response pattern (quadratic *P *< 0.01, cubic *P *= 0.01) for fecal ADF excretion. The cubic equation indicated that fecal ADF excretion would be minimized at a TE supplementation rate of 0.09% of DM and maximized at 0.9% of DM.

Feed digestibility and ruminal parameters

Table 3 shows the feed digestibility and ruminal parameter results. There was an influence of TE supplementation rate on ADF digestibility (ADFD) indicated by quadratic (*P *< 0.01) and cubic (*P *= 0.01) patterns. Based upon the cubic pattern, the minimum ADFD was associated with a TE supplementation rate of 0.9% of DM, while ADFD was maximized at 0.07% of DM. However, there were no differences in DMD, OMD, or aNDFD (*P *≥ 0.29) among treatments. Additionally, ruminal pH, redox potential, and protozoa count were not impacted by TE inclusion rate (*P *≥ 0.13). Although total VFA concentration was similar among treatments (*P *≥ 0.13), ruminal propionate was influenced by TE inclusion in a quadratic fashion (*P *= 0.01). More specifically, the quadratic equation suggested that the points of minimum and maximum for ruminal propionate concentration were at TE doses of 0.9% and 0.37% of dietary DM, respectively. Nevertheless, there was no difference in acetate production (*P *≥ 0.14), and therefore no difference in the acetate: propionate ratio (*P *≥ 0.20) among treatments. Furthermore, ruminal butyrate concentration did not differ (*P *≥ 0.40) among TE supplementation rates. There was a quadratic effect (*P *= 0.02) for the ruminal concentration of isobutyrate and a tendency for a cubic effect (*P *= 0.09) for ruminal lactate concentration. For isobutyrate, dietary TE inclusion rates of 0.9% and 0.42% of DM corresponded to the minimum and maximum concentrations, respectively. Similarly, ruminal lactate concentration was minimized at 0.9% of DM and maximized at 0.62% of dietary DM.

Nitrogen metabolism

Results for N metabolism are displayed in Table 4. Fecal N excretion (g/d and % of N intake) and N digestibility were influenced in a quadratic fashion (*P *= 0.04). However, there were no differences in urinary N excretion or the fecal N-to-urinary N ratio (*P ≥ *0.30). Daily N retention and N retention as a percentage of N intake tended to decrease linearly (*P *= 0.06) with increasing inclusion of TE. Although ruminal NH_3_ concentration was not influenced by TE supplementation rate (*P *≥ 0.61), there was a tendency for a cubic effect (*P *= 0.06) on BUN concentration.

Indirect calorimetry energy balance

Table 5 shows the energy metabolism results from the calorimetric analysis. Respiratory quotient (RQ) increased linearly (*P *= 0.02) with increasing TE supplementation rate. Similar to DMI, there were no differences in GEI due to supplemental treatment (*P *≥ 0.38). In addition, there was no significant influence of TE supplementation rate (*P *≥ 0.22) on FE, UE, or DE. There was a significant dose-response pattern (quadratic *P *= 0.01, cubic *P *= 0.03) for daily loss of GASE and a quadratic effect for GASE loss when expressed as Mcal per kg of DM (*P *= 0.04), kcal per kg of metabolic body weight (MBW; *P *= 0.02), and as a proportion of GEI (*P *= 0.05). These patterns suggest that the minimum GASE loss corresponded to a TE supplementation rate between 0.20% and 0.22% of dietary DM, while a dose of 0.9% of DM maximized the GASE loss. There was no significant pattern for MEI (Mcal/d; *P *≥ 0.26) or dietary ME (Mcal/kg DM; *P *≥ 0.20), and the conversion efficiency of ME-to-DE was not significant (*P *≥ 0.27) either. The HE parameters were similar (*P *≥ 0.69) among treatments, with HE loss averaging about 11.6 Mcal/d or 45.9% of GEI. Furthermore, the RE parameters did not differ according to doses of TE (*P *≥ 0.26).

The gas exchange results are displayed in Table 6. Supplementation rate of TE did not influence O_2_ consumption (*P *≥ 0.64) or CO_2_ production (*P *≥ 0.70). Similar to the GASE loss results, quadratic (*P *= 0.01) and cubic (*P *= 0.03) effects were significant for daily CH_4_ production (Figure 1A). In addition, CH_4_ yield (L/kg DM; Figure 1B) and CH_4_ produced per kg of MBW displayed quadratic (*P *≤ 0.04) response patterns with increasing rate of TE supplementation. More specifically, the polynomial equations suggested a TE supplementation rate between 0.20% and 0.22% of DM minimized CH_4_ emissions and the 0.9% of DM dose maximized emissions.

Dose–response relationship of tannin extract supplementation on A) daily CH4 production and B) CH4 yield.

Discussion

Intake and excretion

A primary concern when supplementing CT to ruminants is palatability related to astringency (Naumann et al. 2017), which can cause a reduction in DMI. Although two animals had small amounts of residual orts (< 0.5 kg) during the first period of data collection, we assumed this was due to stress related to a change in environment rather than TE supplementation, as one of these animals with feed refusals was in the TE_0.0_ group. In the current study, the combination of limit-feeding and relatively low TE supplementation rates may have mitigated any potential effects of TE on feed intake. Similarly, DMI was not impacted when Schilling-Hazlett et al. (2024) supplemented the same TE blend at a rate up to 0.3% of DM to dairy heifers consuming a forage-based TMR. In agreement with our findings, Ebert et al. (2017) observed no differences in DMI in beef steers fed a finishing diet ad libitum and supplemented with the same TE blend at up to 1% of DM. It has been generally accepted that supplementation of CT extracts at rates below 5% of dietary DM does not typically reduce feed intake (Naumann et al. 2017). The lack of difference in overall DMI further supports the absence of differences in intake of dietary nutrients. Although both quadratic and cubic trends in fecal ADF excretion were significant, the overall pattern suggests that fecal ADF increased with higher TE inclusion, but at a decreasing rate.

Feed digestibility and ruminal parameters

Along with the potential reduction in feed intake when ruminants consume tannins, the formation of tannin-nutrient complexes may also inhibit nutrient degradation (Naumann et al. 2017). However, differences in digestibility would not be expected as daily intake and excretion of DM, OM, and aNDF were similar. The lack of TE dose response in aNDFD is consistent with the findings from our previous study (Adams et al. 2025), which evaluated the same TE inclusion rates in vitro. In addition, aNDFD and ADFD were not affected when beef cattle were supplemented with a CT extract at 4.5% of DM (Norris et al. 2020). In contrast, Chen et al. (2021) reported a decrease in NDFD when alfalfa silage was treated with a mixture of CT and HT extract at 1% of DM each. However, TE may not be fully responsible for the reduction in NDFD in their study because alfalfa contains other phytochemicals, such as saponins, that are known to reduce fiber digestibility in ruminants (Tedeschi et al. 2021). Nevertheless, the increased fecal ADF excretion, despite similar ADFI, suggests that TE supplementation at 0.9% of DM may have limited microbial degradation of cellulose in the rumen, as supported by the decrease in ADFD. Ultimately, the effect of TE on nutrient digestibility is variable and is likely dependent on the tannin type and source, supplementation rate, and diet type.

Volatile fatty acids are an important energy source for ruminant animals and can provide insight into ruminal carbohydrate fermentation dynamics. A reduction in the molar proportion of ruminal acetate or in the acetate: propionate ratio may be associated with reduced fiber degradation (Carulla et al. 2005). Conversely, a higher molar proportion of propionate is often related to more energetically efficient carbohydrate fermentation (Armentano and Young 1983). In this study, ruminal acetate concentration and the acetate: propionate ratio did not display a dose response, despite a reduction in ADFD at higher supplementation rates. However, ruminal propionate concentration was decreased at a higher dose of TE, while the quadratic pattern indicated greater ADFD and propionate concentration at a lower TE dose. In contrast, Norris et al. (2020) found that supplementing steers with CT extract at up to 4.5% of DM linearly increased ruminal propionate concentration and decreased the acetate: propionate ratio without affecting fiber digestibility. Similarly, Beauchemin et al. (2007) reported a linear reduction in acetate concentration and the acetate: propionate ratio when cattle were supplemented with CT extract at up to 2% of DM, again without impairing fiber degradation. In vitro evaluation of the same TE inclusion rates (Adams et al. 2025) showed a linear decrease in ruminal propionate and a linear increase in the acetate: propionate ratio with increasing TE inclusion from 0% to 0.9% of DM, despite no significant pattern in acetate concentration or fiber digestibility. While our in vivo results differ somewhat from the in vitro findings, the discrepancy may reflect the difference in rumen fluid collection methodology. In the current study, rumen inoculum was collected from animals as they were removed from the respiration chambers, which was prior to morning feeding and approximately 15 h after the previous meal. Nevertheless, our results suggest that supplementation with this TE may have a negative influence on ruminal carbohydrate fermentation at rates greater than 0.3% of dietary DM.

Nitrogen metabolism

An increase in fecal N excretion with increased TE supplementation rate, accompanied by a decrease in apparent N digestibility, suggests tannin-protein complex formation in the rumen, which renders protein less available for degradation by ruminal microbes (McSweeney et al. 2001). However, the increased fecal N was not compensated for by a reduction in urinary N excretion or an improvement in N retention. Similar to our findings, supplementing finishing beef steers with the same TE blend at levels up to 1% of dietary DM resulted in a linear increase in fecal N excretion, without affecting urinary N excretion or retained N (Ebert et al. 2017). In contrast, Aboagye et al. (2018) observed a 10.4% and 21.5% reduction in urinary N excretion when steers were supplemented with 0.25% and 1.5% of dietary DM as a TE blend of HT and CT, respectively. However, the difference in dietary CP concentration between their experimental diet and that used in the current study (17.1% vs. 11.3% CP, respectively) is likely the reason for the contrasting results. Furthermore, the lack of a dose–response relationship in N retained, despite the greater fecal N excretion with increasing TE inclusion, may be indicative of the slightly lower CP concentration in our diet compared to that suggested by NASEM (2016) for a diet of similar energy content (11.3% vs. 12.6%). This is further supported by the lack of differences in ruminal NH_3_ and BUN concentrations. However, the relatively low NH_3_ and BUN values observed in the current study may be due to sample collection occurring when animals were in a fasted state and should therefore be interpreted with caution. Nevertheless, the experimental diet was balanced for ME and metabolizable protein (MP) using the RNS system, with a slight excess of dietary MP supply.

Ammonia is a product of ruminal protein degradation and serves as a source of N for ruminal microbes, including those involved in fiber digestion (Nagaraja 2016). The formation of tannin-protein complexes when cattle are supplemented with TE at 1.8% of dietary DM has been reported to reduce ruminal protein degradation, leading to a subsequent reduction in ruminal NH_3_ concentration (Aguerre et al. 2016). However, they did not find any differences in ruminal NH_3_ concentration when the TE was supplemented at 0.45% or 0.9% of DM. Although protein digestibility was not measured, Aboagye et al. (2018) observed lower ruminal NH_3_ with both HT and a combination of CT and HT at low and high doses (0.25% and 1.5% of DM). In agreement with our findings, Marshall et al. (2022) reported no differences in ruminal NH_3_ concentration when heifers were supplemented with the same TE used in the current study at 0.15% of dietary DM. The concentration of NH_3_–N in the rumen can indicate the adequacy of rumen degradable protein supply (Mezzomo et al. 2017). Therefore, the effect of TE supplementation on ruminal protein degradation in this study was not strong enough to induce a change in NH_3_ concentration.

Indirect calorimetry energy balance

The RQ, defined as the ratio of CO_2_ produced to O_2_ consumed, can provide insight into which macronutrients are being metabolized for energy (Kleiber 1961). In the current study, the RQ was greater than 1 for all treatments, suggesting that lipogenesis and carbohydrate oxidation were occurring simultaneously (Ferrannini 1988). These findings are likely related to the maturity of the animals used, which are consistent with the results of Crossland et al. (2018), who reported similar RQ values in steers of comparable maturity.

The lack of a dose–response relationship in GEI is consistent with the similar results observed in DMI and OMI. In addition, the similar response in FE loss among treatments supports the similarity in DE parameters, such as dietary DE concentration and the conversion of GE to DE. These findings are in agreement with Ebert et al. (2017) when beef steers were supplemented with the same TE blend at up to 1% of dietary DM.

Although there was no significant dose–response pattern for UE loss, a significant proportion of the GEI was lost as UE (approximately 11%). Urea is a primary contributor to UE loss, and its excretion increases with a higher dietary CP: MP ratio (Blaxter and Martin 1962). The CP:MP ratio of our experimental diet was approximately 1.38, similar to the ratio reported by Jennings et al. (2018); however, they observed a UE loss approximately 86% lower than ours. Nevertheless, the extreme UE loss observed in this study may represent an asynchrony in the fermentability of carbohydrates and protein, possibly related to a TE-induced change in ruminal passage rate. The quadratic dose–response relationship for GASE parameters indicates that gaseous energy loss was minimized at a low rate of TE supplementation. In this case, supplementation of TE at 0.3% of dietary DM reduced daily GASE loss and GASE loss as a proportion of GEI by approximately 2.4% and 2.8%, respectively, relative to the control. In contrast, Norris et al. (2020), who evaluated TE supplementation up to 4.5% of DM, reported a linear decrease in GASE loss. However, their TE contained only CT, whereas the increased GASE observed at higher supplementation rates in our study may be due to the combined presence of CT and HT in the TE. The ME:DE ratio observed in the current study was approximately 0.74 across treatments, slightly lower than the 0.82 suggested by NASEM (2016). Though, using the equation developed by data from Blaxter et al. (1966) yielded values more consistent with 0.82 estimate. The lack of dose–response relationships in HE and RE with TE supplementation is consistent with the findings by Ebert et al. (2017), who supplemented a TE blend at up to 1% of DM, measured using a headbox system. Similarly, Carulla et al. (2005) reported no differences in HE or RE, measured via whole-animal respirometry, when sheep were supplemented with 2.5% of CT. Altogether, the overall energy efficiency was not influenced by supplementation with the TE blend in our study.

In agreement with our findings, gaseous exchange of O_2_ and CO_2_ was not influenced by supplementation with a TE blend at up to 1% of DM in beef cattle (Ebert et al. 2017) or with plant CT at 2.5% of DM in sheep (Carulla et al. 2005). Conversely, Norris et al. (2020) reported a linear effect of CT extract supplementation rate on O_2_ consumption and CO_2_ production. Regarding CH_4_ emissions, the quadratic pattern suggests a greater response at lower levels of TE supplementation. Supplementation with TE at 0.3% of dietary DM resulted in approximately 2.4% lower daily CH_4_ emissions, 2.8% lower CH_4_ yield, and 2.2% lower CH_4_ produced per kg of MBW compared to the control. In contrast, Schilling-Hazlett et al. (2024) did not observe a difference in daily CH_4_ emissions from dairy heifers supplemented with the same TE blend at up to 0.3% of DM compared to control heifers, despite a nearly 3% numerical reduction. These inconsistencies may be related to the measurement techniques utilized for gaseous emissions (spot sampling versus whole-animal respirometry). Nevertheless, the results from the current study are in agreement with the lowest total in vitro CH_4_ production associated with the TE inclusion rate of 0.18% of DM (Adams et al. 2025). In this study, the quadratic equations indicated that CH_4_ emissions would be minimized at a TE dose between 0.20% and 0.22% of dietary DM.

One mechanism that has been proposed for reduced CH_4_ production due to TE supplementation is the indirect inhibition of methanogenesis, mediated by a decrease in the ruminal H_2_ pool resulting from reduced fiber degradation (Tavendale et al. 2005). However, we did not observe a reduction in fiber digestibility associated with reduced CH_4_ emissions in this study. Our findings of reduced CH_4_ production without affecting aNDFD are consistent with those observed in a previous study that evaluated the same TE blend at identical inclusion rates in vitro (Adams et al. 2025). Similarly, when Norris et al. (2020) supplemented a CT extract to beef cattle at 4.5% of DM, they observed a reduction in daily CH_4_ production by almost 17% compared to the control without a reduction in aNDFD or ADFD. In contrast, Chen et al. (2021) reported a decrease in in vitro CH_4_ production when alfalfa silage was treated with a TE blend of CT and HT at 1% of DM each, but also observed a reduction in NDFD. Nevertheless, the increasing rates of TE supplementation may have caused a reduction in the passage of digesta from the rumen (Kumar and Singh 1984). Silanikove et al. (2001) found a delay in the passage of fluid and particulate matter throughout the entire gastrointestinal tract of goats fed tannin-rich carob leaves. Similarly, Tseu et al. (2020) reported a linear reduction in disappearance rate that was associated with linear decreases in the rates of digestion and passage in cows supplemented with a CT extract at up to 2.25% of dietary DM. Although not evaluated in this study, the increasing CH_4_ production may have been related to a reduction in passage rate with increasing rates of TE supplementation, as an inverse relationship between rate of passage and CH_4_ production has been reported (Okine et al. 1989). A reduction in passage rate due to TE supplementation likely increases the fermentability of the diet (Tedeschi and Fox 2020). However, the pattern of decreasing ADFD suggests that the increasing CH_4_ production with increasing rates of TE supplementation may be related to an initial increase in fiber carbohydrate fermentation (Adams et al. 2025) or increased starch digestibility due to increased residence time in the rumen.

Ruminal fiber degradation results in the production of H_2_ as pyruvate is converted to acetate, with the H_2_ typically being used as a substrate for methanogenesis (Moss et al. 2000). However, a shift in fermentation pathway with propionate outcompeting methanogens for H_2_ may result in reduced CH_4_ production without a reduction in fiber digestibility (Moss et al. 2000). Our results are supported by this contention, as CH_4_ production was reduced at lower TE supplementation rates, with increased ADFD and ruminal propionate, compared to the higher doses, which showed increased CH_4_ production. The lack of a dose-response relationship for the concentration of acetate and the acetate: propionate ratio with TE supplementation in this study suggests that the mechanism of action of the TE we evaluated may not have been directly related to fiber degradation. Alternatively, the reduced CH_4_ production may be attributable to increased propionate acting as a H_2_ sink, consistent with the negative correlation between CH_4_ production and ruminal propionate reported by Moss et al. (2000). Similar to our findings, Norris et al. (2020) found that supplementing steers with CT extract at 4.5% of DM resulted in a reduction in daily CH_4_ production along with greater ruminal propionate concentration. When the same TE inclusion rates were evaluated in vitro (Adams et al. 2025), a 0.18% inclusion rate resulted in the lowest total CH_4_ production, which was associated with a relatively greater concentration of propionate and a lower acetate: propionate ratio. Regardless, a reduction in CH_4_ production paired with increased concentration of ruminal propionate is the most likely scenario for the inhibition of methanogenesis (Tavendale et al. 2005). However, microbial profiling would be beneficial to confirm these contentions. Ultimately, the overall quadratic patterns in CH_4_ production parameters may be the result of several different factors acting at different times, as suggested by Adams et al. (2025).

Conclusions

The provision of the TE blend was most effective at a lower supplementation rate. Based on the polynomial regression equations, the greatest benefits of TE supplementation may occur at a supplementation rate near 0.2% of dietary DM without negatively affecting diet digestibility, N balance, and energy efficiency. Beyond the 0.3% of DM dose, the impact of supplementation appeared to be detrimental in terms of gaseous emissions and diet digestibility. Further research on this TE blend should investigate supplementation rates below 0.3% of DM and include evaluations of rumen motility, passage rate, and microbial dynamics to clarify the mechanism of action to reduce CH_4_ emissions from beef cattle. While the present results provide insight into short-term responses, long-term in vivo studies will be important to fully understand tannin-animal interactions, animal responses, and dietary context.