The influence of extended supplementation of quebracho extract to beef steers consuming a hay diet on digestion, ruminal, and blood parameters

Abstract

The addition of natural plant secondary compounds to ruminant feed has been extensively studied because of their ability to modify digestive and metabolic functions, resulting in a potential reduction in greenhouse gas emissions, among other benefits. Condensed tannin (CT) supplementation may alter ruminal fermentation and mitigate methane (CH₄) emissions. This study’s objective was to determine the effect of quebracho CT extract [QT; Schinopsis quebracho-colorado (Schltdl.) F.A. Barkley & T. Meyer] within a roughage-based diet on ruminal digestibility and kinetic parameters by using the in situ and in vitro gas production techniques, in addition to blood urea nitrogen (BUN) and ruminal (volatile fatty acid [VFA], NH₃-N, and protozoa count) parameters. Twenty rumen-cannulated steers were randomly assigned to four dietary treatments: QT at 0%, 1%, 2%, and 3% of dry matter (DM; QT0: 0% CT, QT1: 0.70% CT, QT2: 1.41% CT, and QT3: 2.13% CT). The in situ DM digestibility increased linearly (P = 0.048) as QT inclusion increased, whereas in situ neutral detergent fiber digestibility (NDFD) was not altered among treatments (P = 0.980). Neither total VFA concentration nor acetate-to-propionate ratio differed among dietary treatments (P = 0.470 and P = 0.873, respectively). However, QT3 had lower isovalerate and isobutyrate concentrations compared with QT0 (P ≤ 0.025). Ruminal NH₃ and BUN tended to decline (P ≤ 0.075) in a linear fashion as QT inclusion increased, suggesting decreased deamination of feed protein. Ruminal protozoa count was reduced in quadratic fashion (P = 0.005) as QT inclusion increased, where QT1 and QT2 were lower compared with QT0 and QT3. Urinary N excretion tended to reduce in a linear fashion (P = 0.080) as QT increased. There was a treatment (TRT) × Day interaction for in vitro total gas production and fractional rate of gas production (P = 0.013 and P = 0.007, respectively), and in vitro NDFD tended to be greater for QT treatments compared with no QT inclusion (P = 0.077). There was a TRT × Day interaction (P = 0.001) on CH₄ production, with QT3 having less CH₄ production relative to QT0 on day 0 and QT2 on days 7 and 28. Feeding QT up to 3% of the dietary DM in a roughage-based diet did not sacrifice the overall DM digestibility and ruminal parameters over time. Still, it is unclear why QT2 did not follow the same pattern as in vitro gas parameters. Detailed evaluations of amino acid degradation might be required to fully define CT influences on ruminal fermentation parameters and CH₄ production.

Introduction

For several decades, research has demonstrated that systematic antibiotic use increases livestock production and growth efficiency (Potter et al., 1976; Tedeschi et al., 2011; Huang et al., 2018). However, public apprehensions concerning the use of antibiotics and “nonnatural” products in domestic livestock diets, coupled with the livestock industry becoming a focal point of greenhouse gas (GHG) emissions, have encouraged the sector to seek natural feed alternatives. Ruminants produce methane (CH₄) as one of the typical end products of anaerobic fermentation, by which methanogens reduce carbon dioxide (CO₂) to CH₄ as a process of eliminating H₂ from the rumen to maintain dynamic equilibrium (Janssen, 2010). However, this strategy can constitute a considerable portion (2% to 10%) of the daily gross energy intake lost by the animal (Holter and Young, 1992; Bodas et al., 2012). Thus, a reduction in ruminal CH₄ could result in more energy available to the animal and help mitigate environmental emissions (Tedeschi and Fox, 2020).

Condensed tannins (CTs) are plant secondary compounds that were originally categorized as antinutritional factors because of the observed reduction in the ruminal cell wall and protein digestibility (Van Soest, 1994) and voluntary feed intake (Haslam, 1989; Beauchemin et al., 2008). The CT’s adverse effects are likely because of their high reactivity, reducing the digestion of proteins and carbohydrates and inhibiting enzymes (Haslam, 1989). However, CTs also possess some beneficial features, such as reducing CH₄ emissions and increasing N use efficiency (Ramírez-Restrepo and Barry, 2005; Waghorn, 2008; Bodas et al., 2012; Ávila et al., 2015; Orlandi et al., 2015). Yet, the diminished CH₄ production observed with CT provision has been strongly associated with the unintended reduction of fiber digestion in the rumen (Jouany and Morgavi, 2007; Bodas et al., 2012). Therefore, this study’s objective was to evaluate the effect of differing rates of quebracho [Schinopsis quebracho-colorado (Schltdl.) F.A. Barkley & T. Meyer] tannin extract (QT) in a hay-based diet upon blood and ruminal parameters and in situ and in vitro ruminal apparent digestibility of growing beef steers.

Materials and Methods

All experimental procedures involving animals were cared for in accordance with acceptable practices and experimental protocols reviewed and approved by the Texas A&M – Institute of Animal Care of Use Committee (AUP #2018-0410). This experiment was conducted from October 2018 to January 2019 at the Nutrition and Physiology Center, Texas A&M University, College Station, TX (30°33′32.3″N latitude and 96°24′40.2″W longitude).

Animals and treatments

Twenty ruminally cannulated British-crossbred steers (227 ± 19 kg) from the Texas A&M AgriLife Research Center in McGregor, TX, were utilized in a 67-d experiment. On day −25, steers were ranked by body weight (BW) and housed in five pens (four animals per pen) equipped with a Calan gate feed system (American Calan, Northwood, NH) and water trough. On day 0, steers were weighed after 16 h without feed to obtain initial shrunk BW (SBW) and were randomly assigned to one of four dietary treatments (five animals per treatment). Animal SBW was recorded weekly (days 9, 16, and 23) to adjust for the provision of dietary treatment, and final SBW was obtained on day 30 of the experimental period. Dietary treatments consisted of QT (SILVATEAM, San Michele Mondovi Italy) fed at 0%, 1%, 2%, and 3% (dry matter [DM] basis; QT0: 0% CT, QT1: 0.70% CT, QT2: 1.41% CT, and QT3: 2.13% CT, respectively) of the diet. For all treatments, animals were fed a basal diet of bermudagrass hay [Cynodon dactylon (L.) Pers. var. dactylon] and a protein supplement (Table 1), approximating maintenance energy and protein requirements according to the Ruminant Nutrition System (Tedeschi and Fox, 2020). The QT extract was analyzed for total CT and CT fractions, and protein precipitable phenolics (PPP), as discussed by Norris et al. (2020a). Quebracho extract contained 77.99% total CT (36.07%, 41.43%, and 0.49% of extractable, protein-bound, and fiber-bound CT, respectively; Terrill et al., 1992; Wolfe et al., 2008) and 31.47% PPP (Hagerman and Butler, 1978; Naumann et al., 2014). The diet was offered once daily (0800 hours) at 2.1% of SBW, and the inclusion of pre-weighed QT was hand-mixed into individual animal supplement preceding the provision of hay. All animals were allowed to consume the supplement (dietary treatment) prior to accessing hay. Before the feedings, orts were collected, weighed, and stored at −20 °C until DM (105 °C for 48 h) and neutral detergent fiber (NDF) with amylase analyses using the ANKOM 200 fiber analyzer (ANKOM Technology, Meadon, NY) were performed.

Table 1.

Ingredient and chemical composition of the basal diet

Items1	Basal diet, %
Ingredient composition, % DM
Bermudagrass hay	87.90
Cottonseed meal	6.60
Dried distillers grain	4.41
Molasses	1.09
Chemical composition2
DM, %	91.13
CP, % DM	13.18
Soluble protein, % CP	24.45
aNDF, % DM	60.97
ADF, % DM	37.59
Lignin, % DM	10.12
Ether extract, % DM	1.89
Sugar, % DM	6.58
Starch, % DM	1.79
NFC, % DM	15.19
Ash, % DM	8.76
Calcium, % DM	0.23
Phosphorus, % DM	0.19
TDN, % DM	55.15
ME, Mcal/kg	1.99
NEm, Mcal/kg	1.14
NEg, Mcal/kg	0.58

¹Items are feed ingredients, and chemical composition of diets evaluated by Cumberland Valley Analytical Services (Waynesboro, PA).

²NEm, net energy for maintenance; NEg, net energy for gain.

Chemical analyses

Feed samples were collected weekly for bromatological analyses and digestibility assays using in situ and in vitro procedures. Samples collected throughout a week were then composited into one representative sample of the offered basal diet (hay plus supplement) for each experimental period, and a 200-g subsample was shipped to Cumberland Valley Analytical Services (CVAS; Waynesboro, PA) for the chemical analysis of DM (Goering and Van Soest, 1970), aNDF (Van Soest et al., 1991), acid detergent fiber (method# 973,18; AOAC, 2006), lignin using sulfuric acid (Goering and Van Soest, 1970), crude protein (CP; method# 990.03; AOAC, 2006) in a Leco FP-528 Nitrogen Combustion Analyzer (Leco Corporation, St. Joseph, MO), soluble CP (Krishnamoorthy et al., 1982), non-fibrous carbohydrates (NFC, computed as 100 – [CP + aNDF + ash + ether extract – aNDF-N]), 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) in a Perkin Elmer 5300 DV ICP (Perkin Elmer, Shelton, CT), and calculation of total digestible nutrients (TDN) and net energy using empirical equations (Weiss, 1998).

Ruminal in situ digestibility measurements

Nylon bags (5 × 10 cm) with 50 μm porosity (Ankom Technologies, Macedon, NY, USA) were used to perform the 48-h in situ incubations. Bags were weighed, filled with 5 g of 2-mm oven-dried ground diet sample, and sealed as described by Vanzant et al. (1998). Representative feed samples were collected weekly and dried at 55 °C for at least 48 h prior to grinding through a 2-mm screen using a Wiley mill (Thomas Scientific, Swedesboro, NJ). Bags containing feed samples were replicated five times, whereas blank bags (bags without feedstuff) were replicated twice. Feed and blank bags were placed within 32- × 42-cm polyester bags and placed in each animal’s rumen on days 0, 7, 14, 21, and 27 using the sample of the basal diet representative for each time point. Empty (i.e., blank) bags were used as correction factors for the in situ digestibility of DM and NDF. Following the 48-h incubation, bags were removed from the rumen and immediately quenched in ice water to stop fermentation. The bags’ rinsing process was performed by rinsing the bags with distilled water followed by washing using cold water and the spin portion of the washing machine’s delicate wash cycle (Coblentz et al., 1997; Krizsan and Huhtanen, 2013).

After washing, all bags were dried at 55 °C for 72 h in a forced-air oven, equilibrated to room temperature in a desiccator, and final dry weight was recorded. In situ dry matter digestibility (DMD) was calculated using the following equation:

$${\displaystyle \begin{array}{l}{\displaystyle DMD,\mathrm{\%}=\frac{100\times (W3-(W1\times C1))}{W2}}\end{array}}$$ (1)

where W3 is the dried bag weight containing sample residue, g; W1 is initial bag weight, g; C1 is the blank bag correction factor (average weight of dry bag following cold water rinse divided by the initial weight of the bag) of dried bag weight, g/g; and W2 is initial dried sample weight, g.

The in situ NDF digestibility (NDFD) was obtained by the method described by Van Soest and Robertson (1980), using the ANKOM 200 fiber analyzer (ANKOM Technology, Meadon, NY) with the addition of amylase but without sodium sulfite. After washing with neutral detergent, all bags were dried at 55 °C for 72 h in a forced-air oven, equilibrated to room temperature in a desiccator, and final dry weight was recorded. In situ NDFD was calculated using the following equation:

$${\displaystyle \begin{array}{l}{\displaystyle NDFD,\mathrm{\%}=\frac{100\times (W4-(W1\times C2))}{W2}}\end{array}}$$ (2)

where W4 is the dried bag weight containing sample residue after washing with ND, g; W1 is initial bag weight, g; C2 is the blank bag correction factor after washing with ND (average weight of dry bag following ND wash divided by the initial weight of the bag), g/g; and W2 is the initial weight of the dried samples, g.

Rumen sampling and analyses

During the in situ bag removal process, about 600 mL of rumen fluid was collected from the steers using a vacuum pump and placed into individual stainless-steel thermoses. Measurements of ruminal fluid pH, redox potential (Eh), and reactive oxygen species (ROS) were taken using a portable pH and redox meter probe (Thermo Scientific Orion A221, Thermo Fisher Scientific, Waltham, MA) immediately after the collection from each animal. The rumen fluid was then filtered through eight layers of cheesecloth, and subsamples were taken using duplicate falcon tubes and preserved for volatile fatty acids (VFA), NH₃-N analyses, and protozoa enumeration.

Preservation methods and analyses were performed, as described by Norris et al. (2020a). Briefly, preservation methods for analyses were: 8 mL of rumen inoculum and 2 mL of metaphosphoric acid solution (25% w/v) for VFA, 2 mL of rumen inoculum and 8 mL of 0.1M HCl acid solution for NH₃-N analyses, and 1 mL of rumen inoculum and 10 mL of ethanol for protozoa enumeration. All samples were stored at −20 °C until subsequent analyses. The concentrations of VFA were determined using gas chromatography as described by Cagle et al. (2019) and NH₃-N via colorimetric methods (Chaney and Marbach, 1962). Protozoa counts were obtained using the methods described by Dehority (1984) without staining. The protozoa counts were performed using a Sedgewick Rafter counting chamber, a 1-mL aliquot of the diluted sample (1:10 rumen fluid: ethanol) was counted at 100× magnification with a 0.5-mm square counting grid, and 25 evenly spaced grids from the entire chamber surface were counted using Nikon Eclipse E200 microscope (Nikon Corporation, Tokyo, Japan).

Blood sampling and analyses

On days −1, 6, 13, 20, and 27, blood samples were collected via jugular venipuncture into sodium heparin blood collection tubes (Vacutainer, 10 mL; Becton Dickinson, Franklin Lakes, NJ) and immediately placed on ice. Plasma was obtained by centrifuging the blood (3,000 × g for 20 min at 4 °C). The supernatant was aliquoted into polypropylene tubes and stored at −20 °C until analysis. Samples were analyzed for blood urea nitrogen (BUN) using a colorimetric kit (#B7551, Pointe Scientific, Inc., Canton, MI). The intra- and inter-assay coefficient of variation was 3.0% and 5.8 %, respectively. Urinary N excreted (UNE) was then estimated for each collection using the equation of Kohn et al. (2005):

$${\displaystyle \begin{array}{l}{\displaystyle UNE=CR\times BUN\times SBW}\end{array}}$$ (3)

where CR is the clearance rate of N in the kidney of beef cattle, assumed as 1.3 L × d⁻¹ × kg⁻¹; SBW is the shrunk body weight, kg, measured at the week of collection; and BUN is blood urea N, g/L.

Urine analyses

Urine spot samples were collected from each animal on days 2, 9, 16, and 23. The urine was collected from all animals approximately 4 h after the supplement feeding period using an adapted collector and mechanical stimulation. Samples were acidified with 3M HCl to prevent bacterial degradation of purine derivatives (PD), and acidified urinary samples were stored at −20 °C for further analyses. The HCl acidification was performed using 1 mL of HCl per 20 mL of urine. PD, such as allantoin, uric acid, and creatinine concentrations, were determined via high-performance liquid chromatography (Agilent 1100-HPLC System) as described by (Shingfield and Offer, 1999). Because xanthine and hypoxanthine are rarely detected in cattle urine, total PD was calculated by summing allantoin and uric acid and expressed as mmol/d (González-Ronquillo et al., 2004). The absorbed PD was calculated using the equation proposed by Chen and Gomes (1992):

$${\displaystyle \begin{array}{l}{\displaystyle AbsorbedPD=(TotalPDexcreted-0.385\times SB{W}^{0.75})÷0.85}\end{array}}$$ (4)

where total PD excreted is the sum of allantoin and uric acid, mmol/d, and SBW is shrunk body weight, kg, measured at the week of collection.

Because the rate of creatinine excretion in the urine is relatively constant (Chizzotti et al., 2008), we computed the daily excretion of urinary creatine using equation 5 (Chizzotti et al., 2008) and then divided it by the urinary creatinine concentration (g/L) measured as described above:

$${\displaystyle \begin{array}{l}{\displaystyle UCE=0.0345\times SB{W}^{0.9491}}\end{array}}$$ (5)

where UCE is urinary creatinine excretion, g/d, and SBW is shrunk body weight, kg, taken at same day of collection (Chizzotti et al., 2008).

In vitro gas production measurements

Within a treatment, rumen fluid contents from all of the animals were collected in a 1-liter thermos 4-h postprandial and mixed in equal portions to form representative rumen fluid samples for the in vitro gas production (IVGP) technique as described by Tedeschi et al. (2009) and modifications discussed by Tedeschi and Fox (2020). The in vitro incubations were performed using Wheaton bottles (160 mL) incubated in two chambers with the temperature maintained at 39 °C and a multiple plate stirrer with the capacity to house 24 bottles in each (Dias Batista et al., 2020). Before the in vitro incubations, the Wheaton bottles were prepared using the procedure discussed by Crossland et al. (2018) and Dias Batista et al. (2020). Briefly, 200 mg of 48h air-dried feed sample ground through a 2-mm screen using a Wiley mill (Thomas Scientific, Swedesboro, NJ) was weighed and transferred into 40 bottles (eight replicates per treatment). Then, 2.0 mL of distilled H₂O was used to dampen the sample to prevent particle scattering, and 14 mL of Goering and Van Soest (1970) in vitro buffering media was added under continual flushing of CO₂ to maintain an oxygen-reduced environment. The bottles were sealed using butyl rubber septa coated with petroleum jelly and crimp seals and transferred to the 39 °C chamber to attain ruminal temperature before rumen inoculum placement. Rumen fluids were filtered through four layers of cheesecloth and glass wool and transferred into a glass flask flushed with CO₂. The addition of 4 mL of rumen inoculum per bottle was then performed via syringe and needle insertion. For an individual run, each chamber contained six bottles per treatment: four bottles containing feed sample, one laboratory standard, and one blank (without the addition of feedstuff). The internal pressure of the chamber was equilibrated to atmospheric pressure prior to the initiation of data recording. Upon initiation, the pressure inside each bottle was recorded at 5-min intervals for 48 h using PicoLogo software (Pico Technology, Tyler, TX 75702) with real-time fermentation profiles being plotted for each bottle. After the 48-h incubation period, bottles were placed in an ice bath to cease fermentation. Methane concentration was measured by taking a subsample of the headspace (1 mL) and analyzing via gas chromatography (GOW-MAC Series 580, Gow-Mac Instrument, Bethlehem, PA) according to the method of Allison et al. (1992). The in vitro NDFD was calculated by adding 40 mL of neutral detergent solution (ANKOM Technology, Macedon, NY) in each bottle, autoclaving them for 15 min at 120 °C, filtering the samples using Whatman 54 papers, and oven-drying the residue at 55 °C for 48 h.

Nonlinear functions were used to plot the kinetic analysis of the 48-h fermentation, where the lowest sum of square errors (Schofield et al., 1994) was utilized to select the functions using Gasfit (http://www.nutritionmodels.com/gasfit.html). The convergence of gas production data was performed with specific R scripts using nls and port algorithms (Fox et al., 1978; Gay, 1990; Chambers and Bates, 1992). The Gasfit output data (Table 5) included total gas production (TGP; mL), fermentation rate (1/h), lag time (h) using exponential curves, asymptote cumulative gas production (CGP) of NFC pool (mL), fractional rate of fermentation of the NFC pool (1/h), asymptote CGP of the fiber carbohydrate (FC) pool (mL), and fractional rate of degradation of the FC pool (1/h) using the logarithmic two-pool nonlinear function as described by (Tedeschi and Fox, 2020). The IVGP dynamics were evaluated on days 0, 7, 14, 21, and 28 of the experimental period.

Table 5.

Effect of quebracho extract upon IVGP dynamics

	Dietary treatment1				P-values			Constrasts2
Items3	QT0	QT1	QT2	QT3	SEM	TRT	TRT × Day	L	Q	C × QT
Exponential
TGP, mL	18.14	17.41	18.08	16.64	0.55	0.058	0.013	0.051	0.405	0.127
kd, %/h	5.06	5.33	5.43	5.22	0.13	0.252	0.007	0.328	0.081	0.091
Lag time, h	−0.13	−0.03	−0.13	−0.09	0.18	0.404	0.953	0.262	0.310	0.101
Log. two-pool
Asymptote (P1), mL	6.50ab	6.61ab	6.98a	5.89b	0.21	0.003	0.015	0.093	0.003	0.978
kd P1 (%/h)	9.98	10.59	10.64	10.72	1.03	0.380	0.398	0.138	0.421	0.088
Lag time to P2, h	1.76	2.33	2.09	2.18	0.19	0.186	0.854	0.227	0.200	0.047
Asymptote (P2), mL	10.87	10.25	10.56	10.04	0.37	0.260	0.339	0.121	0.861	0.107
kd P2, %/h	2.48	2.54	2.51	2.60	0.10	0.600	0.039	0.267	0.813	0.359
Exp. kd equivalent, %/h	4.17	4.27	4.19	4.42	0.20	0.532	0.043	0.265	0.629	0.421
Energy estimates
ivNDFD, 48h	45.02	46.25	49.40	52.75	2.11	0.063	0.441	0.009	0.618	0.077
CGP, mL/g NDFD	364.8a	321.0ab	304.4ab	256.0b	20.5	0.027	0.525	0.004	0.584	0.033
Methane, mg/g FOM	11.76	11.40	12.65	10.74	1.72	0.148	0.001	0.489	0.190	0.808
Methane, mg/g NDF digested	38.21	33.49	38.60	29.07	4.08	0.291	0.028	0.054	0.350	0.134
TDN, %	44.86	46.58	46.22	45.49	0.44	<0.001	<0.001	0.182	<0.001	<0.001
ME, Mcal/kg (TDN, 4%)	1.62	1.69	1.67	1.64	0.01	0.001	<0.001	0.303	<0.001	<0.001

¹QT0, QT 0% DM; QT1, QT 1% DM; QT2, QT 2% DM; QT3, QT 3% DM. Dietary treatment values are given as least-squares means.

²Contrasts: L, Linear; Q, quadratic; QT, no Quebracho vs. Quebracho inclusion.

³TGP, total gas production of the exponential nonlinear function; kd, the fractional rate of gas production of the exponential nonlinear function; lag time, time required to commence fermentation; asymptote (P1), accumulative gas production of NFC; kd (P1), fractional rate of gas production of NFC pool; lag time to P2, time required to commence fermentation of FC pool; asymptote (P2), accumulative gas production of FC pool; kd P2, fractional rate of gas production of FC pool; exp. kd equivalent, exponential decay digestion rate (kd); ivNDFD, in vitro neutral detergent fiber digestibility; CGP, cumulative gas production, milligram per gram of NDF digested; TDN, computed TDN.

^a,bLeast-squares means in a row with different superscripts differ at P ≤ 0.05.

Statistical analyses

All data were analyzed using the PROC MIXED of SAS version 9.4 (SAS Institute Inc., Cary, NC). The Shapiro–Wilk test from the PROC UNIVARIATE was performed on all data to check for normality. Ruminal protozoa count was nonnormally distributed (W = 0.86). Thus, the logarithm (base 10) of the ruminal protozoa count was used to achieve normality (W = 0.97). Data were analyzed following a repeated measures design using the REPEATED statement. The covariance structure for each variable was determined according to the corrected Akaike information criterion (Hurvich and Tsai, 1989). For the in vitro data, the statistical model was bottle nested within treatment with the chamber as the subject and incubation box as a random factor. All other variables were examined, having pen and animal nested within treatment as random factors. The specified term for the repeated statement was day, with the animal within treatment or the bottle within treatment and chamber as the subject. DM intake was included as a covariate for the in situ and ruminal parameter data. The effect of treatment (TRT) on the dependent variable was tested at minimum, median, and maximum DM intake (DMI) levels when the interaction between the DMI (as a covariate) and TRT was significant. For blood parameters, blood obtained prior to the initiation of the trial (day −1) was included as an independent covariate within the analyses. Least-square means were determined using the LSMEANS statement; the largest standard error of the mean is reported. Results are reported according to treatment effects if no interactions were significant. Orthogonal polynomial contrasts were performed for all variables to determine linear, quadratic, and control vs. QT effects. Significance was set as P ≤ 0.05, and tendencies were assumed as 0.1 ≥ P > 0.05.

Results

Table 2 presents the effect of QT inclusion on the nutrient intake and in situ data. Final SBW and average daily gain (ADG) did not differ among treatments (P = 0.698 and P = 0.795, respectively). As all treatments received the basal diet at approximately maintenance level (2.1% of SBW), the lack of difference in ADG and final SBW was by design. There was no effect of QT inclusion on DM or NDF intake (P = 0.647 and P = 0.641, respectively). There was an interaction between DMI and treatment (P = 0.044); therefore, the effect of TRT on DMD was tested at 3.25, 4.85, and 5.50 kg DMI/d, representing the minimum, median, and maximum DMI. The in situ DMD increased linearly with the inclusion of QT (P = 0.048), as shown in Figure 1. An increased DMD was observed for QT3 (P = 0.019) compared with QT0 at a low level of DMI. As DMI increased, in situ DMD was increased for all treatments. However, DMD of QT3 only increased by 1.25% from the low (3.25 kg/d) to high (5.5 kg/d) intake, whereas QT0 and QT2 increased 11%, on average, and QT1 increased 5.45%. In contrast to DMD, there was only a day effect for NDFD (P < 0.001) with the DMI as covariate having a tendency (P = 0.058) and, therefore, excluded from the model. There was no treatment effect on ruminal pH, ROS, and Eh (P > 0.520). However, there was a day effect (P < 0.001) on pH and ROS values.

Table 2.

Effect of quebracho extract on steer feed intake, in situ feed digestibility, and ruminal parameters

	Dietary treatment1				P-values				Covariate2		Contrasts3
Item4	QT0	QT1	QT2	QT3	SEM	TRT	Day	TRT × Day	DMI	TRT × DMI	L	Q	QT
ISBW, kg	234.5	233.0	224.9	232.5	20.04	0.670	—	—	—	—	0.601	0.458	0.536
FSBW, kg	281.6	284.3	267.1	282.6	23.04	0.698	—	—	—	—	0.782	0.581	0.786
ADG, kg/d	0.77	0.84	0.69	0.82	0.11	0.795	—	—	—	—	0.992	0.783	0.935
DMI, kg/d	5.01	5.02	4.76	4.95	0.41	0.647	—	—	—	—	0.549	0.595	0.606
NDFI, kg/d	3.05	3.06	2.90	3.02	0.25	0.641	—	—	—	—	0.548	0.588	0.602
DMD, %	58.19	58.53	58.51	58.38	0.08	0.068	<0.001	0.201	0.012	0.044	0.048	0.669	0.093
NDFD, %	44.96	45.19	44.65	44.93	1.13	0.980	<0.001	0.231	0.058	0.116	0.881	0.978	0.975
pH	6.44	6.50	6.50	6.49	0.05	0.833	<0.001	0.866	0.830	0.742	0.550	0.507	0.375
ROS	−185.27	−141.49	−164.99	−154.13	23.07	0.521	<0.001	0.556	0.261	0.529	0.467	0.444	0.208
Eh, mV	−208.84	−222.95	−248.90	−210.34	19.27	0.356	0.577	0.931	0.117	0.326	0.243	0.707	0.169

¹QT0, QT 0% DM; QT1, QT 1% DM; QT2, QT 2% DM; QT3, QT 3% DM. Dietary treatment values are given as least-squares means.

²If there were significant (P ≤ 0.05) interactions between dietary treatment, day, and DMI, the dietary treatment means are reported with the covariate structure in the model.

³Contrasts: L, Linear; Q, quadratic; QT, no Quebracho vs. Quebracho inclusion.

⁴ISBW, initial SBW; FSBW, final SBW; NDFI, NDF intake.

Figure 1.

Effect of quebracho extract upon in situ feed DMD on different intake levels kg/d (•, solid line = QT0; ■, long dash-dot = QT1 ; ▲, long dash = QT2; ×, round dots = QT3). Vertical bars indicate the SE of the treatments. Means within a column not sharing a common superscript letter differ (P ≤ 0.05).

The inclusion of QT had no major impact on ruminal parameters (Table 3), but isobutyrate concentration increased (P = 0.009) for QT0 relative to QT2 and QT3 (0.84 vs. 0.77 and 0.73 mmol/L, respectively) with the inclusion of QT resulting in less isobutyrate than without QT (P = 0.009; 0.76 vs. 0.84 mmol/L). Similarly, isovalerate was greater without QT vs. with QT provision (P = 0.040; 0.86 vs. 0.78). A linear tendency (P = 0.074) to decrease ruminal NH₃-N concentration was observed as QT inclusion increased, and the same pattern was observed for BUN and UNE (P = 0.075 and 0.080, respectively). Ruminal protozoa count was affected in a quadratic fashion (P = 0.005) as QT inclusion increased, with QT1 and QT2 being lower relative to QT0 and QT3 (4.98 vs. 5.06 log₁₀/mL, respectively).

Table 3.

Effect of quebracho extract within a high-roughage diet on the rumen and blood parameters

	Dietary treatment1					P-values		Covariate2		Contrasts3
Item4	QT0	QT1	QT2	QT3	SEM	TRT	TRT × Day	DMI	DMI × TRT	L	Q	C × QT
Total VFA, mmol/L	76.08	72.86	68.70	68.64	3.83	0.470	0.325	0.117	0.135	0.141	0.685	0.192
Acetate, mmol/L	53.23	50.83	47.53	47.51	2.78	0.403	0.371	0.067	0.118	0.114	0.668	0.163
Propionate, mmol/L	12.98	12.23	11.80	11.87	0.72	0.644	0.258	0.517	0.261	0.259	0.515	0.240
Acetate:propionate	4.06	4.16	4.08	4.06	0.09	0.873	0.262	0.069	0.329	0.852	0.569	0.742
Butyrate, mmol/L	7.10	7.05	6.80	6.73	0.42	0.897	0.264	0.355	0.329	0.471	0.971	0.625
Isobutyrate, mmol/L	0.84a	0.80ab	0.77bc	0.73c	0.02	0.009	0.887	0.962	0.865	0.001	0.800	0.009
Isovalerate, mmol/L	0.86a	0.84ab	0.78ab	0.73b	0.03	0.025	0.984	0.605	0.924	0.003	0.556	0.040
Valerate, mmol/L	1.10	1.11	1.06	1.07	0.05	0.898	0.395	0.728	0.226	0.555	0.960	0.712
BCVFA, mmol/L	9.90	9.80	9.41	9.25	0.47	0.742	0.324	0.470	0.351	0.292	0.942	0.472
NH3-N, mg/dL	4.77	4.93	4.01	3.74	0.49	0.262	0.128	0.541	0.339	0.074	0.668	0.345
Protozoa, log10/mL	5.06a	4.98b	4.98b	5.06a	0.03	0.045	0.185	0.035	0.887	0.926	0.005	0.094
BUN, mg/dL	23.00	21.79	21.11	19.65	1.32	0.334	0.485	0.327	0.156	0.075	0.926	0.160

¹QT0, QT 0% DM; QT1, QT 1% DM; QT2, QT 2% DM; QT3, QT 3% DM. Dietary treatment values are given as least-squares means.

²If there were significant (P ≤ 0.05) interactions between dietary treatment and DMI, the dietary treatment means are reported with the covariate structure in the model.

³Contrasts: L, Linear; Q, quadratic; QT, no Quebracho vs. Quebracho inclusion.

⁴BCVFA, branched-chain VFA.

^a–cLeast-squares means in a row with different superscripts differ at P ≤ 0.05.

Urinary PD excretion results are presented in Table 4. There was no TRT or TRT × Day effect for allantoin excretion (P ≥ 0.176). Uric acid excretion tended (P = 0.065) to be greater for QT0 compared with QT2 but, overall, was lower for animals receiving QT compared with QT0 (P = 0.013). The total PD absorbed did not differ among treatments (P = 0.248); however, a quadratic tendency (P = 0.065) was observed as QT inclusion increased. Total PD excretion per kilogram of digestible organic matter intake tended (P = 0.061) to linearly decrease as QT inclusion increased. Estimated daily urine excretion was not affected by the inclusion of QT (P = 0.338).

Table 4.

Effect of quebracho extract within a high-roughage diet on urine parameters

	Dietary treatment1				P-values			Covariate2		Contrasts3
Item	QT0	QT1	QT2	QT3	SEM	TRT	TRT × Day	DMI	DMI × TRT	L	Q	C × QT
Allantoin, mmol/d	52.64	43.87	39.99	47.20	4.61	0.176	0.721	0.334	0.842	0.276	0.056	0.059
Uric acid, mmol/d	6.14	4.90	4.00	4.162	0.81	0.065	0.672	0.163	0.289	0.017	0.279	0.013
Allantoin:uric acid	0.89	0.90	0.913	0.92	0.01	0.172	0.902	0.406	0.523	0.027	0.758	0.108
PD absorbed4, mmol/d	39.34	29.25	26.87	34.68	5.26	0.248	0.538	0.237	0.564	0.435	0.065	0.092
PD: DOMI5, g/kg	6.02	4.50	4.44	4.46	0.67	0.126	0.674	—	—	0.061	0.179	0.017
Urinary volume, L/d	23.30	25.37	20.33	18.28	3.53	0.338	0.528	0.183	0.154	0.130	0.482	0.561
UNE6, g/d	72.00	67.77	66.16	61.40	4.22	0.343	0.818	<0.001	0.081	0.080	0.951	0.155

¹QT0, QT 0% DM; QT1, QT 1% DM; QT2, QT 2% DM; QT3, QT 3% DM. Dietary treatment values are given as least-squares means.

²If there were significant (P ≤ 0.05) interactions between dietary treatment and DMI, the dietary treatment means are reported with the covariate structure in the model.

³Contrasts: L, Linear; Q, quadratic; QT, no Quebracho vs. Quebracho inclusion.

⁴PD absorbed estimated using the equation proposed by Chen and Gomes (1992).

⁵Estimates of total PD excretion per gram of digestible organic matter intake (DOMI), kg.

⁶Estimated using the equation proposed by Kohn et al. (2005)

IVGP dynamics are presented in Table 5. TGP was decreased in a linear fashion (P = 0.051) as QT inclusion increased. However, there was an interaction of TRT × Day (Figure 2) for TGP (P = 0.013) and the fractional rate of gas production (P = 0.007). The asymptote CGP of the NFC pool (P1) was affected by TRT × Day (P = 0.015; Figure 3), but there were no effects for the fractional rate of gas production of the NFC pool (P ≥ 0.038). Neither lag time required to commence fermentation or time required to start the asymptote of the FC pool was altered by TRT or TRT × Day interaction (P ≥ 0.186). There were no effects for the asymptote of the FC pool (P2; P ≥ 0.26). However, there was a TRT × Day interaction (Figure 4) for the fractional rate of gas production of the FC pool and exponential degradation rate (P = 0.039 and P = 0.043). The in vitro NDFD increased in a linear fashion (P = 0.009) as QT inclusion increased. This resulted in a TRT effect (P = 0.027) on the CGP per gram of NDFD, with QT3 lower than QT0 (256.0 vs. 364.8 mL/g NDFD), and the inclusion of QT having less CGP compared with without QT (P = 0.033; 293.8 vs. 364.8 mL/g NDFD). There was a TRT × Day interaction (P = 0.028) for CH₄ production (mg/ g NDFD; Figure 5). The computed TDN and metabolizable energy (ME) were affected by a TRT × Day interaction (P ≤ 0.001; Figure 6); however, the inclusion of QT resulted in greater (P < 0.001) TDN and ME compared with no QT.

Figure 2.

Effect of quebracho extract inclusion on (A) TGP of the exponential nonlinear function and (B) the fractional rate of gas production of the exponential nonlinear function (•, solid line = QT0; ■, long dash-dot = QT1; ▲, long dash = QT2; ×, round dots = QT3). Vertical bars indicate the SE of the treatments. Means within a day not sharing a common superscript letter differ (P < 0.05).

Figure 3.

Effect of quebracho extract inclusion on accumulative gas production of NFC pool (P1; •, solid line = QT0; ■, long dash-dot = QT1; ▲, long dash = QT2; ×, round dots = QT3). Vertical bars indicate the SE of the treatments. Means within a day not sharing a common superscript letter differ (P ≤ 0.05).

Figure 4.

Effect of quebracho extract inclusion on (A) fractional rate of gas production of FC pool and (B) the exponential decay digestion rate (kd) (•, solid line = QT0; ■, long dash-dot = QT1; ▲, long dash = QT2; ×, round dots = QT3). Vertical bars indicate the SE of the treatments. Means within a day not sharing a common superscript letter differ (P < 0.05).

Figure 5.

Effect of quebracho extract inclusion on methane production (mg/g NDF Digested) (•, solid line = QT0; ■, long dash-dot = QT1; ▲, long dash = QT2; ×, round dots = QT3). Vertical bars indicate the SE of the treatments. Means within a day not sharing a common superscript letter differ (P ≤ 0.05).

Figure 6.

Effect of quebracho extract inclusion on (A) computed TDN and (B) ME computed (•, solid line = QT0; ■, long dash-dot = QT1; ▲, long dash = QT2; ×, round dots = QT3). Vertical bars indicate the SE of the treatments. Means within a day not sharing a common superscript letter differ (P ≤ 0.05).

Discussion

The lack of major effects of QT inclusion on nutrient intake and ADG is similar to the results showed by Beauchemin et al. (2007) when QT extract was included up to 2% of the dietary DM. Similarly, QT inclusion up to 4.5% of DMI did not affect DM, organic matter, or NDF intake in beef cattle limit fed near maintenance requirements (Norris et al., 2020a, 2020b). Previous studies evaluating the effects of CT on feed intake have shown variable results. Dschaak et al. (2011) observed reduced DMI in lactating cows when QT was included at 3% of dietary DM. At the same time, Woodward et al. (2001) reported an increase in DMI when Lotus pedunculatus (2.59 % of CT) was fed to cows in late lactation. The effects of CT on intake seem to be more pronounced when the inclusion exceeds 5% of the DM in the diet (Frutos et al., 2004; Naumann et al., 2017). Many of the species that contain CT also produce other plant secondary compounds; therefore, the alteration in DMI cannot be solely attributed to CT when only CT is measured, as other metabolites may also be responsible for the observed variation (Naumann et al., 2017). In part, the discrepancy found in the literature might be due to different methods used to quantify CT content and its bioavailability in the feed (Mueller-Harvey et al., 2019).

In the current study, the inclusion of QT in the diet linearly enhanced in situ DMD when DMI was included in the model. This observation contradicts other studies (Orlandi et al., 2015; Piñeiro-Vázquez et al., 2017; Norris et al., 2020a). Landau et al. (2000) demonstrated that the ruminal passage rate could be reduced when animals consume CT. An increase in ruminal retention time influences the overall microbial digestion. Although in situ DMD improved in all the dietary treatments as DMI increased, QT3 displayed a reduced slope relative to all other treatments. Nevertheless, the current results are plausible since the inherent dynamics associated with passage rate and post-ruminal digestion are not represented, particularly as the majority of previous studies observing reduced DMD with QT inclusion evaluated total tract, not solely ruminal digestibility. In contrast, the rather neutral ruminal pH values in the current study were expected due to the basal diet’s high fiber content.

The lack of treatment effect on total VFA and the primary VFA (acetate, propionate, and butyrate) indicates that ruminal digestibility was most likely not affected by QT inclusion in the diet. This finding is consistent with previous results utilizing QT (Aguerre et al., 2016). To a degree, this confirms the in situ NDFD and ruminal pH values observed. However, contrary to our results, Beauchemin et al. (2007) noted a linear reduction in total VFA and acetate production with 1% and 2% QT inclusion. Still, neither total tract DM nor NDF digestibilities were altered compared with the control group. Some CTs reduce ruminal protein degradation (Frutos et al., 2000; Hervas et al., 2000), thereby reducing the production of NH₃ in the rumen. This finding is consistent with the current study as ruminal NH₃ tended to be linearly reduced as QT increased. Similarly, other studies supplementing QT in a high roughage diet have observed reductions in ruminal NH₃ concentration (Beauchemin et al., 2007; Norris et al., 2020a, 2020b). Because isobutyrate and isovalerate are primarily derived from the deamination of branched-chain amino acids in the rumen (Cummins and Papas, 1985), the decrease in isobutyrate and isovalerate concentration as QT inclusion increased is indicative of reduced ruminal protein deamination, corroborating the ruminal NH₃ and BUN values in the present study. Barry and Manley (1984) indicated that CT from Lotus penduculatus decreased deamination, and Jones et al. (1994) reported that CT from Sainfoin might impact proteolysis differently depending on the bacteria. These authors showed that the protease activity of Prevotela ruminicola was not affected by the presence of CT (0, 25, 50, 75, and 100 µg/mL). However, Streptococcus bovis was greatly affected as CT inclusion increased, and Butyrivibrio fibrisolvens and Ruminobacter amylophilus reduced proteolysis as CT increased. Still, the reduction seems to be related to the presence of CT and not the dosage. However, the BUN levels observed (19.7 to 23 mg/dL) vastly exceed the optimum level (5 to 8 mg/dL) reported by Johnson and Preston (1995). This may suggest excess N intake in the current experiment but is likely just representative of asynchrony between the degradation rates of protein and carbohydrates (Tedeschi and Fox, 2020). Although NH₃, BUN, and absorbed PD tended to respond quadratically to increasing QT, a linear tendency observed for these parameters existed. This suggests a reduction in protein degradation and protein and microbial synthesis (Crawford et al., 2020). The linear tendency to decrease UNE as QT inclusion increased suggests a shift in N excretion from the urine to the feces. Previous research has demonstrated the potential for QT to modify the route of N excretion from the urine to the feces (Orlandi et al., 2015; Aguerre et al., 2016; Norris et al., 2020b). This is assumed to be a result of reduced ruminal proteolysis.

The quadratic effect of CT on ruminal protozoa has been reported in studies utilizing QT (Norris et al., 2020b) and Leucaena leucocephala (Tan et al., 2011). In contrast, Jolazadeh et al. (2015) reported a linear decrease in the protozoa population when evaluating the effects of soybean meal treated with CT from pistachio extract, and Chiquette et al. (1989) showed an increase in ruminal protozoa population when sheep were fed CT from Lotus corniculatus. However, the mechanism by which CTs alter rumen dynamics and protozoa populations is not well understood (Patra and Saxena, 2011).

The IVGP dynamics results demonstrated a slightly different pattern from those obtained using the in situ technique. The reduction in gas production of the NFC pool on day 0 for QT3 may suggest a reduction in the rapidly degradable fraction of the feedstuff, in particular the dietary protein (Frutos et al., 2000). However, the gas production of QT3 was similar to all other treatments on days 7 and 14. This could indicate an adaptative response by the ruminal microbes (Chiquette et al., 1989; McSweeney et al., 2001). The tendency for QT inclusion to increase the rate of gas production of the NFC pool indicates that QT may reduce the total digestion of the rapidly degradable fraction, but the rate at which it is digested is equal to or greater than the control treatment.

The effects of feeding CT-rich plants or CT extract to ruminants have demonstrated decreases in ruminal CH₄ emissions (Tedeschi et al., 2014). The TRT × Day interaction for CH₄ production likely indicates the inhibition of the methanogenic bacteria in the rumen on day 0 of QT provision, but ruminal adaptation may have occurred as no treatment differences were detected on days 14 and 21. On day 28, QT3 had the least CH₄ production per gram of NDF digested, with QT2 having elevated CH₄ production relative to all other treatments. Tavendale et al. (2005) proposed two mechanisms to explain the possible inhibition effect of CT on CH₄ production, either by reducing fiber digestion, thereby decreasing the H₂ pool in the rumen, or by directly inhibiting methanogenic bacterial growth. However, in the current study, in vitro NDFD was not affected by QT inclusion, which suggests that less CH₄ production demonstrated by QT3 on days 0 and 28 may be a result of a declining methanogen population in the rumen.

In research settings, the effects of QT upon CH₄ emissions have been variable. Beauchemin et al. (2007) did not observe a difference in CH₄ production when QT was fed up to 2% of dietary DM. In contrast, Norris et al. (2020a) reported a linear reduction in CH₄ production when QT was fed at 1.5%, 3%, or 4.5% of DM, yet the daily emission declined only at the highest dosage. It appears that the efficacy of QT for ruminal CH₄ suppression is dependent upon the type and amount of CT and diet fed to the animal. Our studies indicate that the inclusion of QT extract, up to 3% of the dietary DM, might be below the threshold required to decrease CH₄ emissions of growing steers consuming a bermudagrass diet. However, the provision of QT at 4.5% DM has demonstrated the potential to reduce not only CH₄ production but also CO₂ in cattle (Norris et al., 2020a). These results may partially explain the drastic reduction in CGP per gram of NDF digested.

Conclusions

Ruminal parameters were not detrimentally impacted by the provision of QT in a high-roughage diet. However, in our study, the inclusion of QT enhanced in situ DMD for animals with low DMI, but in situ NDFD was not altered. We hypothesized the possibility of distinctive CT effects on proteolysis and deamination in the rumen. While the deamination procesmight be reduced in the rumen in the presence of CT, proteolysis may still occur, corroborating our findings on the increased DMD. It is not clear why QT2 did not follow the same pattern as QT3 and QT1. The IVGP data indicated that the CGP of the NFC pool was reduced with QT inclusion, but the rate of gas production tended to increase in treatments consuming QT compared with QT0. This suggests that QT inhibits the degradation of some NFC, but that microbially available NFC was digested at the same rate or faster than the QT0 treatment.

When evaluating the overall digestive response to QT inclusion, it appears that the ability of QT to bind to protein and carbohydrate does not make it entirely unavailable for microbial digestion but requires a longer period for the microbes to degrade it. Thus, the provision of QT up to 3% of dietary DM can be beneficial when DMI is compromised due to the quality of the forage. Feeding QT tended to impact ruminal NH₃ concentrations and BUN, reducing UNE. The potential to decrease UNE is beneficial for soil N status since urinary N is positively correlated with the production of nitrous oxide, a potent GHG.

The inconsistent results for CH₄ production in the present study indicate that the possible beneficial effect of QT on CH₄ emissions may be dependent on nutritional factors in the diet as well as its digestion profile. Some research has evaluated CT to protein ratio and its effects on ruminal proteolysis. However, the optimal dosage of CT required to diminish CH₄ production and the impacts of fiber and energy content in the diet need to be evaluated in more depth. The effects of different molecular weight and chemical structure appear to affect not only protein-binding capacity but also CH₄ inhibition. Thus, further research is necessary to better understand the feasibility of feeding QT to cattle to reduce ruminal CH₄ production. In summary, our results demonstrated that the inclusion of QT up to 3% did not impact ruminal parameters and digestion of DM and NDF in growing animals receiving a high roughage diet, indicating that QT can potentially be strategically utilized to decrease UNE without impacting animal performance.

Glossary

Abbreviations

ADF

acid detergent fiber

ADG

average daily gain

aNDF

NDF with the addition of amylase and sodium sulfite

BUN

blood urea nitrogen

BW

body weight

CGP

cumulative gas production

CP

crude protein

CT

condensed tannin

DM

dry matter

DMD

dry matter digestibility

DMI

dry matter intake

DOMI

digestible organic matter intake

Eh

redox potential

FC

fiber carbohydrate

GHG

greenhouse gas

IVGP

in vitro gas production

ME

metabolizable energy

NDF

neutral detergent fiber

NDFD

NDF digestibility

NFC

nonfiber carbohydrates

PD

purine derivatives

PPP

protein precipitable phenolics

QT

quebracho tannin extract

ROS

reactive oxygen species

SBW

shrunk body weight

TDN

total digestible nutrient

TGP

total gas production

UNE

urinary N excretion

VFA

volatile fatty acid

Conflicts of interest statement

The authors declare no real or perceived conflicts of interest.