Optimum grape pomace proportion in feedlot cattle diets: ruminal fermentation, total tract nutrient digestibility, nitrogen utilization, and blood metabolites

Abstract

Because of its high content of polyphenolic compounds, the dietary inclusion of grape pomace (GP) in ruminant diets can reduce reactive nitrogen (N) and methane emissions and enhance the shelf life and beneficial fatty acids (FAs) content of meat. However, the dietary inclusion of GP beyond a threshold that is still to be determined for feedlot cattle can also compromise nutrient supply and, thus, growth performance. This study investigated the optimum proportion of GP in finishing cattle diets. Nutrient intake and apparent total tract digestion, ruminal pH and fermentation, estimated microbial protein synthesis, route of N excretion, and blood metabolites were measured. Six ruminally fistulated crossbred beef heifers (mean initial body weight ± SD: 714 ± 50.7 kg) were used in a replicated 3 × 3 Latin square with 21-d periods. Dietary treatments were 0%, 15%, and 30% of dietary dry matter (DM) as GP, with diets containing 84%, 69%, and 54% dry-rolled barley grain, respectively. There was a linear increase (P = 0.07) in DM intake and quadratic change (P ≤ 0.01) in neutral detergent fiber (NDF) intake. There was a quadratic change (P ≤ 0.04) in apparent total tract DM, NDF, and crude protein digestibility as dietary GP content increased. However, there were no treatment effects (P ≥ 0.18) on total ruminal short-chain FA concentration and duration and area pH < 6.2, 5.8, and 5.5. Although N intake did not differ (269, 262, 253 g/d; P = 0.33) across dietary treatments, feeding GP led to a tendency for a quadratic change (P ≤ 0.07) in ruminal ammonia-N and plasma urea-N concentrations. Total N excretion also changed (quadratic, P = 0.03) because of changes (quadratic, P = 0.02) in fecal N excretion as urinary excretion of N and urea-N did not differ (P ≥ 0.15) across treatments. Feeding GP led to quadratic changes (P ≤ 0.01) in fecal excretion of fiber-bound N. Microbial N flow and apparent N retention also changed (quadratic, P ≤ 0.04) as dietary GP proportion increased. In conclusion, responses to dietary GP proportion were mostly quadratic with indications that nutrient supply as reflected by changes in apparent total tract nutrient digestibility, microbial N supply, and apparent N retention could be compromised beyond a 15% dietary inclusion level.

Introduction

The upcycling of byproducts from the agro processing and bioethanol industries into human-edible animal protein is one of the keys to sustainable beef production (van Hal et al., 2019; Salami et al., 2019). In addition to cost-effectiveness, limiting reactive nitrogen (N) and methane emissions and preventing a decrease in product quality and consumer acceptability of beef are all important considerations when formulating byproduct-containing diets for finishing cattle (Salami et al., 2019). Grape pomace (GP) is one such byproduct that could be used as an alternative feedstuff for finishing cattle (Correddu et al., 2016). Use as feed is sensible as disposal of GP can be costly and time-consuming given the environmental concerns related to its acidity that limits the use of landfills and field application (Kalli et al., 2018).

GP contains a high content of polyphenolic compounds including condensed tannins (CT), which can favorably modulate nutrient metabolism when fed to ruminants (Makkar, 2003). For instance, feeding GP (17% of diet DM) to dairy cattle diets resulted in a tannin-induced shift in N excretion from urine to feces (Greenwood et al., 2012), which was desirable as urine N, particularly urea-N, makes a greater contribution to reactive N emissions than fecal N. Similarly, adding 27% to 30% (DM basis) dried or ensiled GP to Angus steer and lactating dairy cow diets resulted in a decrease in CH₄ emissions in part because the CT altered the ruminal microbiome and fermentation pathways (Moate et al., 2014; Caetano et al., 2019). Arend et al. (2018) and Zhao et al. (2018) also reported a beneficial increase in the antioxidant capacity and meat content of bioactive polyunsaturated fatty acids (PUFA) such as C18:2 n-6 and C18:2 c9t11 following the addition of GP (10% to 58% of diet DM) to finishing cattle and lamb diets.

Despite the documented positive outcomes, the dietary inclusion of GP in finishing cattle diets is still limited in the United States (Samuelson et al., 2016). This is in part related to the paucity of information on the impact of feeding GP on nutrient supply, especially the optimum proportion, as there are indications that growth performance could be compromised beyond a certain dietary inclusion threshold that still needs to be evaluated (Greenwood et al., 2012; Nudda et al., 2015). For instance, Caetano et al. (2019) reported a decrease in body weight (BW) gain and feed efficiency in steers when 30% of diet dry matter (DM) was ensiled GP. This was attributed to a decrease in apparent total tract nutrient digestibility and, thus, nutrient supply because of 1) dietary polyphenolic compounds restricting the growth and activity of the ruminal microbes and irreversibly binding nutrients including crude protein (CP) and 2) an increase in dietary lignin content. On the other hand, feeding up to 10% GP (DM basis) resulted in an increase in average daily gain (ADG) and feed efficiency in finishing lambs (Zhao et al., 2018). The dietary proportion of GP (10% vs. 30% of diet DM) possibly accounts for the discrepancies between those studies. Therefore, because this information is still lacking, the determination of the optimum proportion of GP in finishing cattle diets is necessary, as this will ensure its increased and judicious use as ruminant feed.

We hypothesized that increasing the dietary proportion of GP would result in a decrease in nutrient supply in finishing cattle. Therefore, the objective of our study was to evaluate the optimum proportion of dried GP in finishing cattle diets. The responses to GP inclusion were measured by assessing nutrient intake and apparent total tract digestibility, ruminal fermentation characteristics, microbial protein production, route of N excretion, and serum fatty acid (FA) profile.

Materials and Methods

Animal use for this experiment was approved by the Institutional Animal Care and Use Committee at the University of Idaho (protocol no. 2017-3).

Animals, experimental design, and treatments

Six ruminally cannulated (10 cm diameter, Bar Diamond, Inc., Parma, ID) crossbred beef heifers (initial BW ± SD; 714 ± 50.7 kg) were used in a replicated 3 × 3 Latin square design balanced for residual effects from the previous period. Heifers were housed in individual tie stalls at the University of Idaho Dairy Center. Total mixed rations (TMR) were prepared and offered once daily at 0630 hours for ad libitum intake. Each experimental period lasted 21 d; dietary adaptation took place during the first 14 d that was followed by 7 d of sample collection. The dietary treatments were three inclusion rates of dried GP: 0%, 15%, or 30% of dietary DM (CON, 15GP, or 30GP, respectively; Table 1). The barley grain used in the study was purchased commercially and dry rolled using a processing index of 82% (defined as the weight of 500 mL of grain after processing/weight of 500 mL of grain before processing × 100%). Monensin (Rumensin 90, Elanco Animal Health, Greenfield, IN) was added to all diets (28 mg/kg DM). Diets were also formulated to provide enough energy, protein, minerals, and vitamins to exceed the nutrient requirements of animals gaining 1.5 kg/d (NASEM, 2016). Prior to the initiation of the first experimental period, heifers were fed rations with increasing proportions of barley grain over a 3-wk period to acclimate them to the high-grain experimental diets.

Table 1.

Dietary ingredient and chemical composition

	Diet1
Item	CON	15GP	30GP
Ingredient, % of DM
Barley grain, rolled	84.3	69.3	54.3
Grape pomace	—	15.0	30.0
Grass hay	10.0	10.0	10.0
Canola meal	2.5	2.5	2.5
Corn oil	2.0	2.0	2.0
Mineral vitamin mix2	1.2	1.2	1.2
Chemical analysis
DM, %	95.3 ± 2.10	95.0 ± 2.18	93.9 ± 2.23
OM, % of DM	95.2 ± 1.16	93.8 ± 1.24	89.3 ± 2.31
ADF, % of DM	10.4 ± 3.04	17.0 ± 3.17	31.0 ± 7.79
NDF, % of DM	25.5 ± 5.92	31.7 ± 3.76	45.8 ± 7.52
Indigestible NDF, % of DM	9.15 ± 1.95	17.7 ± 2.81	33.5 ± 6.70
Crude fat, % of DM	4.28 ± 0.250	4.86 ± 0.281	5.44 ± 0.313
CP, % of DM	12.6 ± 1.74	11.5 ± 1.02	11.9 ± 1.40
NDIN, % of N	34.2 ± 0.57	40.0 ± 1.54	43.3 ± 1.85
ADIN, % of N	3.87 ± 0.954	11.9 ± 1.49	18.6 ± 2.95
Tannins3, %	—	0.83 ± 0.064	1.65 ± 0.129

¹CON, control diet containing 0% dried grape pomace; 15GP, 15% dried grape pomace; 30GP, 30% dried grape pomace.

²Supplement DM contained CP, 51.3%; crude fat, 0.48%; salt, 12.3%, Ca, 19.7%; P, 0.07%; Mg, 0.55%; K, 0.10%; S, 0.15%; Fe, 12.5%; Mn, 1,230 ppm; Zn, 2,050 ppm; organic Zn, 1,025 ppm; Cu, 615 ppm; organic Cu, 205 ppm; Co, 31.2 ppm; I, 175 ppm; Se, 13.5 ppm; Selenium yeast, 4.51 ppm; Vitamin A, 27,948 IU/kg; Vitamin D, 2,795 IU/kg; Vitamin E, 93.1 IU/kg; and Rumensin 90 (Elanco Animal Health, Greenfield, IN), 1,128 g/ton.

³Calculated based on the tannin content of the GP used in the present study (average of 5.51% DM).

Although the statistical analysis was not conducted, experimental diets differed numerically in composition. By substituting barley grain with increasing percentage of GP (0%, 15%, and 30%; DM basis), dietary acid detergent fiber (ADF), neutral detergent fiber (NDF), and indigestible NDF (iNDF) content increased from 10.4% to 31.0%, 25.5% to 45.8%, and 9.15% to 35.5% DM, respectively, because of the greater ADF, NDF, and acid detergent lignin (ADL) content of GP (Table 2) compared with that of barley grain. Similarly, partial substitution of barley grain with GP resulted in an increase in dietary neutral detergent insoluble N (NDIN) and acid detergent insoluble N (ADIN) content due to the high NDIN and ADIN content of GP. As expected, GP contained a high concentration of C18:2 n-6 and total phenolics and tannins (Table 2).

Table 2.

Chemical, FA, and phenolic compounds composition of grape pomace

Item	Amount
Chemical composition
DM, %	88.3 ± 0.95
Ash, % of DM	8.66 ± 1.013
ADF, % of DM	50.8 ± 4.31
NDF, % of DM	52.5 ± 3.52
ADL, % of DM	34.5 ± 3.97
CP, % of DM	12.8 ± 0.63
NDIN, % of N	55.4 ± 3.73
ADIN, % of N	53.7 ± 4.78
Ether extract, %	5.91 ± 0.692
FA profile, g/100 g of FA methyl esters
C14:0	0.13 ± 0.026
C16:0	10.2 ± 0.03
C18:0	4.65 ± 0.196
C18:1 c9	16.7 ± 0.72
C18:2 n-6	63.6 ± 1.20
C18:3 n-3	1.37 ± 0.140
C20:0	0.60 ± 0.109
C22:0	0.50 ± 0.062
C24:0	0.30 ± 0.048
Phenolic compounds, % of DM
Total phenols	5.76 ± 0.434
Free phenols	1.55 ± 0.165
Tannins	5.51 ± 0.429
Non-tannins	0.25 ± 0.020

Measurements

All heifers were weighed prior to morning feeding on two consecutive days at the beginning of each experimental period and at the end of the study. To determine dry matter intake (DMI), TMR offered and orts were recorded daily. Weekly TMR and GP samples were collected over three consecutive days, whereas refusals samples were collected over five consecutive days. The TMR and GP samples were composited by week, and the refusal samples were composited by animal and week.

To measure diurnal changes in ruminal ammonia-N (NH₃-N) concentration, digesta was collected on day 14 at 0630, 0700, 0800, 0900, 1000, 1200, 1500, 1800, and 2100 hours and on day 15 at 0000 and 0300 hours. The intensive sampling schedule was used to enable adequate characterization of the potential changes in NH₃-N concentration following a single feeding event. At each sampling point, approximately, 1 liter of digesta was collected from the cranial ventral, caudal ventral, central, and cranial dorsal regions of the rumen. Samples were strained through polyester monofilament fabric (350 µm mesh opening; ELKO Filtering Co, LLC, Fort Lauderdale, FL). Thereafter, a 5-mL aliquot was mixed with 1 mL of chilled 1% H₂SO₄ and stored (−20 °C) for later NH₃-N analysis. Ruminal pH was measured continuously from days 14 to 21 of each period using indwelling pH data loggers (Lethbridge Research and Development Centre pH data logger system, Dascor, Escondido, CA; Penner et al., 2006). At the beginning of each measurement period, the loggers were standardized in pH 4 and 7 buffers and programmed to record pH every minute.

Grab fecal samples were collected on day 19 (0630, 1230, and 1830 hours), day 20 (0830, 1430, and 2030 hours), and day 21 (1030, 1630, and 2230 hours) and stored at −20 °C for later apparent total tract nutrient digestibility and N balance determination. Spot urine samples were also collected at each fecal sampling time. A subsample (80 mL) of the collected urine was immediately mixed with 5 mL of 2M H₂SO₄ to a pH < 2.5 and placed on dry ice to prevent the loss of NH₃-N. Thereafter, 1 mL of the acidified urine was diluted 1:10 with distilled H₂O, composited by animal and period, and frozen (−20 °C) for later analysis of total N, urea-N, creatinine, and purine derivatives (PDs). To measure ruminal NH₃-N and short-chain fatty acid (SCFA) concentration over a 24-h feeding cycle, approximately, 1 liter of ruminal digesta was also collected at the same time as grab fecal samples. Following straining of digesta as described previously, two 5-mL aliquots of ruminal fluid were collected, mixed with chilled 1% H₂SO₄ or 25% (wt/vol) metaphosphoric acid (H₂PO₄), composited by animal and period, and stored (−20 °C) for later NH₃-N and SCFA analysis.

On the last day of each period (day 21), blood samples were collected via jugular venipuncture 3 h post-feeding into one 10-mL tube containing 158 IU lithium heparin (Becton Dickinson, Franklin Lakes, NJ) and one 10-mL tube with no preservative (Becton Dickinson). Prior to processing, the lithium heparin tubes were stored briefly on ice, whereas the tubes with no preservative were first incubated at room temperature for 20 min. The lithium heparin tubes were then centrifuged (3,000 × g; 20 min; 4 °C) before harvest and subsequent storage (–20 °C) of plasma in cryogenic vials for later analysis of plasma urea-N (PUN). Following centrifugation (3,000 × g; 20 min; 4 °C), serum was also collected from the tubes with no preservative and stored in cryogenic vials at –20 °C until analyzed for FA.

Laboratory Analyses

Composited TMR and refusals (collected during the last week of each period) and daily fecal samples were oven-dried at 55 °C for 72 h for chemical analysis. Dried fecal samples were pooled based on their respective DM contents to obtain a representative composite sample by animal within period, which together with the dried TMR, and refusals samples were sequentially ground through 4- and 1-mm screens (Retsch Cutting Mill SM 200, Retsch). All ground samples were then analyzed for analytical DM by drying at 135 °C for 2 h (AOAC, 2005; method 930.15). Samples were combusted at 600 °C for at least 5 h to determine ash content, and the organic matter (OM) content was calculated by difference (DM – ash; AOAC, 2005; method 942.05). Samples were analyzed for ADF and NDF, with amylase and sodium sulfite used during NDF determination (AOAC, 2005; method 2002.04). GP samples were analyzed for ADL (AOAC, 1990; method 973.18). Crude fat content was determined by ether extraction for 6 h (AOAC, 2005, method 920.39). Samples were analyzed for CP using the Kjeldahl procedure (Foss Analytics; Hillerød, Denmark; AOAC, 1990; method 976.05). The NDIN and ADIN content was determined by analyzing the NDF (obtained using the above NDF procedure without the use of sodium sulfite) and ADF residues, respectively, for N. The iNDF content of TMR, refusal, and fecal samples was determined as described by Valente et al. (2011). Briefly, samples (0.6 g) were weighed into F57 bags (Ankom Technology; Macedon, NY) that were then incubated for 288 h in the rumen of two cows. After incubation, the residues were analyzed for NDF as previously described.

Lipids were extracted from the GP and serum samples using chloroform:methanol (2:1) (Clark et al., 1982), and this was followed by methylation using 0.5 M sodium methoxide in a two-step procedure (Kramer et al., 1997). FA methyl esters were then analyzed using an Agilent 7890A gas chromatograph equipped with an autosampler, flame ionization detector, and an Agilent J&W HP-88 column (100 m × 0.250 mm × 0.20 μm film; Agilent Technologies, Santa Clara, CA) as described by Scholte et al. (2014). Peaks were identified using a Supelco 37 Component FA methyl ester mix (Sigma-Aldrich), and all identified peaks were used to calculate the total FA methyl esters for each sample. Individual FA (g/100 g of FA methyl esters) was calculated as individual FA divided by total FA methyl esters.

The TMR samples were also analyzed for phenols (i.e., total phenols, tannin, and nontannin phenols) as described by Khiaosa-Ard et al. (2015). Briefly, following ball grinding, free phenolics were extracted using ethanol, whereas acid extraction was used for total phenolics (free plus bound forms). Resultant extracts were then analyzed for phenolics using the Folin–Ciocalteu method. In addition, polyvinylpolypyrrolidone was used to precipitate tannins in aliquots of the ethanolic or acidic extracts prior to the analysis of nontannins using the Folin–Ciocalteu method. Tannins were then determined by difference.

Ruminal fluid samples were thawed at room temperature, mixed thoroughly, and centrifuged (12,000 × g for 10 minutes at 4 °C). The supernatant from samples preserved with H₂SO₄ was analyzed for NH₃-N using a phenol-hypochlorite assay (Broderick and Kang, 1980). The supernatant from samples preserved in H₂PO₄ was collected and centrifuged again (16,000 × g for 10 min at 4 °C). The resultant supernatant was then filtered through a 0.2-μm Nylon filter and diluted 1:1 with distilled water. The concentration of SCFA was subsequently determined using a gas chromatograph fitted with a flame-ionization detector (GC-FID; 6890 Series, Hewlett-Packard; Palo Alto, CA) as described by Coats et al. (2012).

Acidified urine composites were thawed and analyzed for total N using the Kjeldahl procedure (Foss Analytics; Hillerød, Denmark; AOAC, 1990; method 976.05). Commercial kits (Arbor Assays; Ann Arbor, MI) were used for the analysis of urine creatinine and urine urea-N (UUN) and PUN. Urine allantoin and uric acid concentrations were determined using a method adapted from Stentoft et al. (2014). Briefly, quantification was carried out using HPLC/MS (Waters Corporation, Milford, MA) fitted with a reversed-phase column (C18, 5 µm particle size, 2 × 250 mm; Phenomenex, Torrance, CA) using a 5% methanol mobile phase.

Calculations

Apparent total tract digestion of DM, OM, CP, NDF, and ADF was calculated as follows:

$${\displaystyle \begin{array}{l}{\displaystyle \begin{array}{ll}& {\displaystyle 100-100\times [(\mathrm{i}\mathrm{N}\mathrm{D}\mathrm{F}\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{c}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{i}\mathrm{n}\mathrm{f}\mathrm{e}\mathrm{e}\mathrm{d}/\mathrm{i}\mathrm{N}\mathrm{D}\mathrm{F}\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{c}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{i}\mathrm{n}\mathrm{f}\mathrm{e}\mathrm{c}\mathrm{e}\mathrm{s})\times }\\ & {\displaystyle (\mathrm{n}\mathrm{u}\mathrm{t}\mathrm{r}\mathrm{i}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{c}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{i}\mathrm{n}\mathrm{f}\mathrm{e}\mathrm{c}\mathrm{e}\mathrm{s}/\mathrm{n}\mathrm{u}\mathrm{t}\mathrm{r}\mathrm{i}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{c}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{i}\mathrm{n}\mathrm{f}\mathrm{e}\mathrm{e}\mathrm{d})].}\end{array}}\end{array}}$$

Urine output was estimated using the concentration of creatinine measured in urine and BW and creatinine constant of 29 mg/kg BW per day (Valadares et al., 1999) according to the following equation:

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{U}\mathrm{r}\mathrm{i}\mathrm{n}\mathrm{e}\mathrm{o}\mathrm{u}\mathrm{t}\mathrm{p}\mathrm{u}\mathrm{t},\mathrm{k}\mathrm{g}/\mathrm{d}=(29\times \mathrm{B}{\mathrm{W}}^{0.96})÷\mathrm{c}\mathrm{r}\mathrm{e}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{n}\mathrm{i}\mathrm{n}\mathrm{e}\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{c}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n},\mathrm{m}\mathrm{g}/\mathrm{L}}\end{array}}$$

Apparent N balance was calculated as the difference between N intake and excretion (fecal + urine).

The excretion of allantoin and uric acid was used to estimate the total absorption of PD as described by Chen and Gomes (1992) according to the following equation:

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{P}{\mathrm{D}}_{\mathrm{e}\mathrm{x}\mathrm{c}\mathrm{r}\mathrm{e}\mathrm{t}\mathrm{e}\mathrm{d}},\mathrm{m}\mathrm{m}\mathrm{o}\mathrm{l}/\mathrm{d}=0.85\left(\mathrm{P}{\mathrm{D}}_{\mathrm{a}\mathrm{b}\mathrm{s}\mathrm{o}\mathrm{r}\mathrm{b}\mathrm{e}\mathrm{d}}\right)+(0.385\mathrm{B}{\mathrm{W}}^{0.75}),}\end{array}}$$

where 0.85 is the recovery of absorbed purines as PD and 0.385 BW^0.75 is representative of purine excretion from endogenous sources. The flow of microbial N was calculated using the following equation:

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{M}\mathrm{i}\mathrm{c}\mathrm{r}\mathrm{o}\mathrm{b}\mathrm{i}\mathrm{a}\mathrm{l}\mathrm{N},\mathrm{g}/\mathrm{d}=70\left(\mathrm{P}{\mathrm{D}}_{\mathrm{a}\mathrm{b}\mathrm{s}\mathrm{o}\mathrm{r}\mathrm{b}\mathrm{e}\mathrm{d}}\right)÷(0.116\times 0.83\times 1,000),}\end{array}}$$

where 70 represents the N content of purines, 0.83 is the digestibility of those purines, and the ratio of purine-N:total N in rumen microbes is 11.6:100.

Statistical analysis

All data on ruminal fermentation characteristics, nutrient intake and apparent total tract digestibility, N utilization, and serum FA were analyzed using the MIXED procedure of SAS (SAS 9.4; SAS Inst. Inc., Cary, NC) for a replicated 3 × 3 Latin square. The model included the following independent variables: cow, period, square, and level of GP inclusion (CON, 15GP, and 30GP). Period, square, and GP inclusion were considered fixed, whereas cow within the square was considered as random. Temporal ruminal NH₃-N data were analyzed accounting for repeated measures through the inclusion of additional terms for time (h) and level of GP inclusion × time interaction in the model described previously. The variance–covariance structure of the repeated measures was modeled separately with an appropriate structure fitted using the lowest values of the fit statistics based on the Bayesian information criteria. Residual distributions were evaluated for normality and homoscedasticity prior to analysis. Orthogonal contrasts were used to test for linear and quadratic effects of level of GP in the diet. Data are presented as least square means. Significance was declared at P < 0.05 and trends at 0.05 < P ≤ 0.10.

Results

Ruminal fermentation

Although increasing dietary GP proportion resulted in a change (quadratic; P = 0.02) in maximum pH, it had no effect on mean and minimum pH (Table 3). The duration and area pH were below the thresholds of 6.2, 5.8, and 5.5 and did not differ across dietary treatments. Similarly, there was no diet effect on ruminal molar proportions of valerate, isobutyrate, and isovalerate, and total SCFA and branched-chain FA concentrations. However, increasing dietary GP proportion resulted in a quadratic change (P ≤ 0.049) in the molar proportions of acetate and propionate and acetate:propionate ratio and a linear increase (P = 0.047) in the molar proportion of butyrate.

Table 3.

Ruminal fermentation characteristics for heifers fed the CON, 15GP, or 30GP diet

	Diet1				P-value2
Variable	CON	15GP	30GP	SEM	Treatment	Linear	Quadratic
Ruminal pH
Mean	5.62	5.77	5.73	0.126	0.53	0.40	0.47
Minimum	4.77	4.75	4.97	0.144	0.28	0.18	0.38
Maximum	6.58	6.82	6.64	0.093	0.047	0.44	0.02
pH < 6.2
Duration, min/d	1,190	999	1,056	84.0	0.23	0.21	0.23
Area, pH units × min/d	891	770	765	167.7	0.70	0.46	0.72
pH < 5.8
Duration, min/d	856	669	679	138.6	0.44	0.28	0.52
Area, pH units × min/d	478	443	426	127.4	0.90	0.67	0.94
pH < 5.5
Duration, min/d	360	363	350	121.5	0.99	0.93	0.94
Area, pH units × min/d	129	125	125	53.1	0.85	0.65	0.75
SCFA
Total, mM	87.9	77.7	81.1	7.44	0.41	0.20	0.80
Acetate, mol/100 mol	45.3	42.0	46.7	2.38	0.08	0.79	0.03
Propionate, mol/100 mol	42.7	42.7	37.5	1.94	0.06	0.98	0.02
Butyrate, mol/100 mol	7.9	11.5	11.6	1.34	0.049	0.047	0.11
Valerate, mol/100 mol	1.95	1.93	1.81	0.247	0.91	0.96	0.67
Isobutyrate, mol/100 mol	0.859	0.681	0.753	0.1119	0.33	0.16	0.86
Isovalerate, mol/100 mol	1.27	0.94	1.66	0.342	0.24	0.40	0.14
Total BCFA3	2.13	1.61	2.41	0.433	0.29	0.31	0.23
Acetate:propionate	1.07	1.00	1.28	0.109	0.12	0.56	0.049

¹CON, 0% dried grape pomace; 15GP, 15% dried grape pomace; 30GP, 30% dried grape pomace.

² P-values indicate the overall GP, linear, and quadratic effects.

³BCFA, branched-chain fatty acid.

Nutrient intake and apparent total tract digestibility

Increasing dietary GP proportion resulted in a tendency for a change (quadratic; P = 0.07) in DMI and a change (quadratic; P ≤ 0.03) in ADF, NDF, and OM intake (Table 4). Increasing dietary GP proportion led to quadratic changes (P ≤ 0.04) in apparent total tract DM, OM, NDF, and CP digestibility and a linear decrease in ADF digestibility.

Table 4.

Intake and apparent total tract nutrient digestibilities for heifers fed the CON, 15GP, or 30GP diet

	Diet1				P-value2
Variable	CON	15GP	30GP	SEM	Treatment	Linear	Quadratic
Intake, kg/d
DM	13.4	14.3	13.4	0.48	0.11	0.28	0.07
OM	12.7	13.4	12.0	0.45	0.05	0.18	0.03
NDF	3.37	4.53	6.11	0.311	<0.01	0.02	<0.01
ADF	1.57	2.43	4.13	0.261	<0.01	0.04	<0.01
CP	1.68	1.64	1.58	0.063	0.33	0.55	0.18
Apparent total tract digestibility, %
DM	61.4	59.7	39.8	3.59	<0.01	0.75	<0.01
OM	65.1	65.2	45.3	3.54	<0.01	0.99	<0.01
NDF	27.0	28.2	14.9	5.06	0.09	0.76	0.04
ADF	30.1	12.0	10.2	6.02	0.06	0.04	0.13
CP	50.4	49.5	36.3	5.07	0.12	0.90	0.04

¹CON, 0% dried grape pomace; 15GP, 15% dried grape pomace; 30GP, 30% dried grape pomace.

² P-values indicate the overall GP, linear, and quadratic effects.

Nitrogen utilization

There was no diet effect on N intake (Table 5). Although fecal N (g/d and percentage of N intake), ADIN (g/d and percentage of fecal N and total N), and NDIN excretion (g/d and percentage of fecal N and total N) changed (quadratic; P ≤ 0.04) following the addition of GP to the diet, there was no diet effect (P ≥ 0.15) on urinary N and urea-N excretion (g/d and percentage of N intake). There was a quadratic change (P = 0.03) in total N excretion (g/d and percentage of N intake) following the addition of GP to the diet. Although there was no diet effect (P = 0.54) on urinary uric acid excretion, urinary allantoin and total PD excretion changed (quadratic; P = 0.03) as dietary GP proportion increased. Microbial N flow (g/d) and apparent N balance (g/d) also changed (quadratic; P = 0.03) following the addition of GP to the diet. Similarly, there was a quadratic change (P ≥ 0.02) in ruminal NH₃-N (composite sample) and PUN (3 h post-feeding sample) concentrations as dietary GP proportion increased. However, there was no diet × time interaction (P = 0.13) on ruminal NH₃-N concentration (Figure 1).

Table 5.

Fecal N excretion, urine N and PD excretion, microbial N supply, and rumen NH₃-N and PUN concentrations for heifers fed the CON, 15GP, or 30GP diet

	Diet1				P-value2
Variable	CON	15GP	30GP	SEM	Treatment	Linear	Quadratic
N intake, g/d	269	262	253	10.0	0.33	0.55	0.18
Fecal excretion
DM, kg/d	5.17	5.75	8.03	0.477	<0.01	0.40	<0.01
N, g/d	131	131	159	9.6	0.07	0.96	0.02
N, % of N intake	49.6	50.5	63.7	5.07	0.12	0.90	0.04
NDIN, g N/d	32.9	38.6	58.2	4.67	<0.01	0.44	<0.01
NDIN, % fecal N	24.3	29.3	36.7	2.30	<0.01	0.17	<0.01
NDIN, % total N output	12.9	16.3	21.3	1.72	0.01	0.21	<0.01
ADIN, g N/d	21.7	35.6	61.9	4.15	<0.01	0.04	<0.01
ADIN, % fecal N	16.8	26.2	39.1	1.86	<0.01	<0.01	<0.01
ADIN, % total N output	8.7	15.1	22.7	1.68	<0.01	0.02	<0.01
Urinary excretion
Total output, kg/d	9.40	8.65	8.81	1.002	0.59	0.35	0.76
N, g/d	119	109	115	9.8	0.34	0.15	0.91
Urea-N, g/d	74.3	71.9	68.5	4.99	0.55	0.65	0.34
Urea-N, % of total urine N	62.9	62.4	60.5	4.89	0.88	0.93	0.65
Total N, % of N intake	44.5	41.0	45.3	3.00	0.36	0.28	0.35
Allantoin, mmol/d	130	109	96	15.4	0.04	0.08	0.03
Uric acid, mmol/d	19.2	20.3	17.3	2.16	0.54	0.68	0.31
Total PDs, mmol/d	149	129	113	16.6	0.04	0.097	0.03
Microbial N flow, g/d	79.4	62.3	49.4	14.03	0.04	0.098	0.03
Total N excretion,
g/d	250	239	274	13.2	0.08	0.44	0.03
% of N intake	94.1	91.5	109.0	5.65	0.09	0.75	0.03
Apparent N balance, g/d	18.1	22.9	−21.4	14.78	0.11	0.82	0.04
Ruminal ammonia-N, mg/dL	4.72	4.02	6.77	0.763	0.07	0.53	0.03
PUN, mg/dL	9.7	10.1	15.9	2.22	0.054	0.97	0.02

¹CON, 0% dried grape pomace; 15GP, 15% dried grape pomace; 30GP, 30% dried grape pomace.

² P-values indicate overall GP, linear, and quadratic effects.

Figure 1.

Ruminal NH₃-N concentration for heifers fed the CON, 15GP, or 30GP diet. Heifers were fed once daily at 0630 hours. Diet, P = 0.24; time, P < 0.01; diet × time interaction, P = 0.13. The error bars reflect the SEM associated with time.

Serum FA profile

The dietary inclusion of GP resulted in a linear decrease (P = 0.04) in the serum C18:3 n-3 and C20:1 concentration and a tendency for a linear decrease in the C16:0 (P = 0.05), C16:1 (P = 0.06), C18:3 n-6 (P = 0.09), and total saturated FA (SFA; P = 0.09) concentration (Table 6). There was a quadratic change in the serum C17:0 (P ≤ 0.04) and C20:3 n-3 (P = 0.02) concentration and a tendency for a quadratic change in the serum C14:1 (P = 0.07), C15:1 (P = 0.07), C18:1 t9 (P = 0.09), C22:1 (P = 0.07), and C24:1 (P = 0.09). Feeding GP led to a tendency for a linear increase (P = 0.06) in the total serum unsaturated FA (UFA) concentration. The SFA:UFA ratio increased (linear; P = 0.048) as dietary GP proportion increased. However, there was no diet effect (P > 0.05) on the serum concentration of C18:0, C18:1 c9, total monounsaturated fatty acids (MUFA), and total PUFA.

Table 6.

Serum FA composition (g/100 g of FA methyl esters) for heifers fed the CON, 15GP, or 30GP diet

	Diet1				P-value2
FA	CON	15GP	30GP	SEM	Treatment	Linear	Quadratic
C14:0	0.535	0.330	0.699	0.1507	0.34	0.30	0.30
C14:1	0.101	0.254	0.404	0.0777	0.16	0.46	0.07
C15:0	0.530	0.422	0.436	0.0536	0.22	0.12	0.48
C15:1	0.468	0.187	0.327	0.1445	0.04	0.047	0.07
C16:0	11.4	10.2	11.1	0.47	0.12	0.05	0.51
C16:1	0.629	0.413	0.517	0.0854	0.15	0.06	0.96
C17:0	0.920	0.748	0.653	0.0763	0.01	0.13	<0.01
C17:1	0.234	0.155	0.160	0.0416	0.38	0.20	0.62
C18:0	17.1	16.8	18.8	0.91	0.26	0.81	0.11
C18:1 t9	2.50	0.36	0.25	0.863	0.047	0.04	0.09
C18:1 c9	6.80	5.83	6.81	0.504	0.19	0.12	0.34
C18:2 all t9,12	0.185	0.141	0.246	0.0727	0.72	0.49	0.68
C18:2 n-3	39.1	43.7	41.2	2.71	0.40	0.19	0.95
C18:3 n-6	0.344	0.212	0.253	0.0740	0.22	0.09	0.69
C20:0	1.44	0.85	0.62	0.650	0.70	0.59	0.50
C18:3 n-3	2.80	2.31	2.56	0.201	0.11	0.04	0.99
C20:1	1.86	4.22	3.12	0.955	0.09	0.04	0.92
C21:0	0.987	0.979	0.707	0.2629	0.57	0.98	0.31
C20:2	0.835	0.791	0.598	0.2104	0.64	0.87	0.37
C22:0	2.26	2.12	2.52	0.445	0.66	0.75	0.40
C20:3 n-6	0.00	0.301	0.000	0.1739	0.39	0.24	0.49
C20:3 n-3	1.70	1.85	2.19	0.138	0.04	0.33	0.02
C22:1	0.104	0.127	0.187	0.0528	0.15	0.59	0.07
C20:4	0.938	0.949	0.605	0.2274	0.40	0.97	0.19
C23:0	1.11	1.08	0.84	0.264	0.69	0.93	0.41
C22:2	ND	ND	ND	—	—	—	—
C20:5	1.37	1.31	1.25	0.209	0.91	0.83	0.71
C24:0	1.28	1.27	1.10	0.343	0.89	0.97	0.64
C24:1	0.611	0.570	0.166	0.2041	0.21	0.88	0.09
C22:6	0.781	0.667	0.688	0.1764	0.87	0.63	0.86
Σ SFA	38.6	35.7	38.7	1.49	0.08	0.05	0.21
Σ UFA	61.4	64.4	61.6	1.56	0.11	0.06	0.31
Σ MUFA	14.9	13.9	13.9	1.52	0.86	0.63	0.82
Σ PUFA	41.4	46.1	42.9	2.48	0.35	0.16	0.76
Σ n-63	46.5	50.5	47.6	2.31	0.38	0.19	0.74
Σ n-34	5.05	4.43	4.68	0.334	0.38	0.18	0.88
n-6:n-3	2.10	2.36	2.21	0.110	0.25	0.11	0.90
SFA:UFA	0.470	0.603	0.463	0.0651	0.07	0.048	0.17

¹CON, 0% dried grape pomace; 15GP, 15% dried grape pomace; 30GP, 30% dried grape pomace.

² P-values indicate the overall GP, linear, and quadratic effects.

³ n-6 included 18:2 all t9,12; 18:2 all c9,12; 18:3 n-6; 20:2; 20:3 n-6; 20:4; and 22:2.

⁴ n-3 included 18:3 n-3; 20:3 n-3; 20:5; and 22:6.

Discussion

GP contains bioactive compounds, including phenolics such as CT, and nutrients, including fiber, which can influence nutrient digestion and metabolism in ruminants (Salami et al., 2019). The GP used in the present study was rich in residual phenolic compounds (5.76% total extractable phenolics; DM basis); this was comparable to the reported range of 4.28% to 7.20% DM (Manso et al., 2016; Zhao et al., 2018). As expected, the NDF (52.5% DM), ADF (50.8%), and ADL (34.5%) contents of GP were also high. Others (Moate et al., 2014; Pauletto et al., 2020) documented a range of 42.2% to 53.5%, 41.3% to 53.1%, and 28.8% to 42.2% of DM, for NDF, ADF, and ADL in GP, respectively. As for the lipid fraction, C18:0 (4.65% of total FA methyl esters), C18:1 c9 (16.7%), and C18:2 n-6 (63.7%) were also the major FA as in previous studies (Moate et al., 2014; Manso et al., 2016).

A major concern when feeding GP to ruminants is its high lignin content. Lignin can compromise fiber digestion by limiting fibrolytic enzyme access to cellulose and hemicellulose (Moore and Jung, 2001). In turn, this slows down digesta passage rate and, ultimately, causes a decrease in DMI related to fill effects. Because it contained a high ADL content (34.5% DM), partial substitution of barley grain with GP increased dietary indigestible NDF content by up to 24.4 percentage units in the present study. Although we did not measure the rate and extent of nutrient degradation in the rumen or digesta passage rate, apparent total tract NDF digestibility was lower for heifers fed the 30GP compared with the CON and 15GP diets. Therefore, it is possible that ruminal fiber digestion was compromised, especially at the highest GP inclusion level. However, we did not observe an analogous decrease in DMI. Although reasons for this are not clear, the potential decrease in ruminal nutrient digestibility and digesta passage rate might not have been substantial enough to compromise DMI. Others (Moate et al., 2014; Pauletto et al., 2020) have similarly reported no significant decrease in DMI when lactating cow diets contained up to 27% (DM basis) GP. Moreover, Caetano et al. (2019) reported an increase in DMI despite a decrease in apparent total tract DM digestibility when steer diets contained 30% GP. This was ascribed to the small particle size of the ground GP possibly limiting the increase in gut fill. However, as in the present study, Caetano et al. (2019) did not measure ruminal fiber degradation and digesta passage rate.

Besides lignin, the polyphenolic compounds in GP could also suppress nutrient digestion by inhibiting the growth and activity of ruminal microbes (e.g., Bacteroides fibrisolvens and Ruminococcus albus), limiting the activity of several microbial enzymes (e.g., hemicellulases and proteases), and irreversibly binding nutrients (e.g., fiber and CP) (Jerónimo et al., 2016; Vasta et al., 2019). Based on the tannin content of the GP used in the present study (5.51% DM), the total tannin content for the 15GP and 30GP diets was 0.83% and 1.65% DM, respectively. Therefore, the increase in the dietary tannin content could have led to the decrease in apparent total tract DM, OM, NDF, and N digestibility beyond the 15% dietary inclusion level. Koenig and Beauchemin (2018) also reported a decrease in apparent total tract OM, NDF, N, starch, and gross energy digestibility when dietary CT content increased to 1.33% DM following the addition of CT extract from Acacia mearnsii. Similarly, apparent total tract N digestion was suppressed when steer diets contained 1.5% quebracho extract (about 1.16% of CT on a DM basis; Norris et al., 2020). In the present study, fecal NDIN and ADIN excretion (g/d, % fecal N, and % of total N output) increased as the dietary GP proportion increased. This suggests increased irreversible binding of nutrients, microbial inhibition, and/or suppressed enzyme activity because of the polyphenolic compounds in GP, which ultimately can compromise nutrient supply.

Because of the documented impact of polyphenolic compounds on ruminal N metabolism (Makkar, 2003), we had anticipated a decrease in ruminal NH₃-N concentration following the dietary addition of GP. For instance, Norris et al. (2020) reported a linear decrease in ruminal NH₃-N concentration in growing steers as the dietary inclusion level of quebracho tannin extract increased (0%, 1.12%, 2.34%, and 3.51% CT; DM basis). Greenwood et al. (2012) also observed a decrease in milk urea-N concentration for lactating cows fed GP. Therefore, although not measured in that study, ruminal NH₃-N concentration could have decreased following the feeding of GP. However, in our study, ruminal NH₃-N concentration did not differ across dietary treatments over time. Similarly, Moate et al. (2014) and Pauletto et al. (2020) did not observe changes in ruminal NH₃-N concentration when feeding 7.5% to 27% GP (DM basis) in dairy cow diets. In the present study, we fed dried GP, and there are indications that the increase in temperature during the drying process could reduce the biological activity of polyphenolic compounds through either their inactivation or increased complexing with nutrients including fiber and protein (Makkar and Singh, 1995). This is supported by Taşeri et al. (2018), who documented a decrease in the total phenolic and tannin content following the drying of GP. Moreover, besides a reduction in the total CT content, drying also caused a decrease in the soluble-CT fraction and a concomitant increase in the protein-bound CT fraction in sainfoin and birdsfoot trefoil (Girard et al., 2018). The bulk of the total tannins in the GP used in the present study were bound. Therefore, this could have limited the efficacy of the tannins in reducing the growth and activity of proteolytic microbes, and/or proteolytic enzymes, and, thus, ruminal NH₃-N concentration.

Although there was no diet effect over time, the average ruminal NH₃-N concentration measured over a 24-h feeding cycle (composite samples) tended to be greater for heifers fed the 30GP compared with that for the CON and 15GP diets in this study. Correddu et al. (2015) also reported an increase in ruminal NH₃-N concentration after adding supplemental grape seed (approximately 0.4 g polyphenols/kg DM of diet) to dairy sheep diets. Microbial sequestration of peptides, amino acid (AA), and NH₃-N in the rumen, which is primarily modulated by adenosine triphosphate (ATP) availability, has an impact on ruminal NH₃-N concentration (Russell et al., 1983; Hristov et al., 2005). When ATP supply is insufficient, the use of peptides, AA, and NH₃-N for microbial protein synthesis is limited. In addition, the peptides and AA in excess of requirements are further broken down to release more NH₃-N. Because GP partially replaced barley grain in the present study, it is highly likely that this may have led to a decrease in ruminal fermentable energy supply, especially at the highest GP inclusion level. In addition, OM and fiber digestion were suppressed primarily in cattle fed the 30GP diet. Therefore, a decrease in ruminal fermentable energy supply could have limited microbial incorporation of diet-derived peptides, AA, and NH₃-N, thus accounting for the increase in ruminal NH₃-N concentration for 30GP heifers in the present study. This is supported by the lower urinary excretion of allantoin and total PD on the 30GP compared with the CON and 15GP diets, which is suggestive of suppressed growth of the ruminal microbes. Moreover, estimated microbial N flow to post-ruminal sites was lowest for heifers fed the 30GP than CON and 15GP diets. Others (Ahnert et al., 2015; Koenig and Beauchemin, 2018; Norris et al., 2020) also reported a decrease in urinary PD excretion and estimated duodenal microbial protein flow when diets contained as low as 0.67% (DM basis) total tannins. Besides a possible ATP deficiency, direct antimicrobial effects of tannins that reduce the amount of peptides, AA, and NH₃-N needed for microbial growth and, ultimately, the amount of microbial protein flowing out of the rumen cannot be discounted (Molan et al., 2001; Frutos et al., 2004).

Ammonia-N not sequestered by the ruminal microbes is absorbed into the blood and detoxified through the synthesis of urea-N in hepatocytes (Satter and Roffler, 1975). Therefore, increased absorption of ruminal NH₃-N into the blood may have led to the greater PUN concentration for the 30GP than CON and 15GP heifers in the present study. Apparent N balance was also negative for heifers fed the 30GP compared with the CON and 15GP diets, possibly reflecting a metabolizable protein (MP) and/or energy deficit. Increased ruminal outflow of ruminally undegradable protein (RUP) related to the formation of tannin–protein complexes that are not degradable in the rumen, but disintegrate in the acidic environment of the abomasum, can improve MP supply and thus the growth performance in beef cattle (Mezzomo et al., 2011; Ahnert et al., 2015). However, depending on several factors including tannin dose, a proportion of the tannin–protein complexes can be resistant to digestion in post-ruminal sites, ending up excreted in feces (Makkar, 2003). In the present study, there was an increase in fecal excretion of bound N as dietary proportion of GP increased. Therefore, a decrease in both digestible RUP and microbial protein supply possibly led to an MP deficiency for 30GP compared with CON and 15GP heifers. This deficit could account for the observed negative apparent N balance, which can upregulate skeletal muscle protein mobilization (Roche et al., 2013). Therefore, the greater PUN concentration for the 30GP than CON and 15GP heifers could have also been related to increased deamination of skeletal muscle-derived AA (Drackley et al., 2001; Jorritsma et al., 2003). Feeding a GP (27% of DM) compared with an alfalfa hay-based control diet also resulted in the loss of BW (−2.1 vs. 29.4 kg) in lactating dairy cows over an 18-d period (Moate et al., 2014). Similarly, Caetano et al. (2019) reported a decrease in ADG in steers fed 30% GP, which was attributed to a decrease in apparent total tract DM digestibility and metabolizable energy supply. Although not measured in the studies by Moate et al. (2014) and Caetano et al. (2019), a decrease in MP supply could also possibly account for the reported decrease in BW and ADG when diets contained GP.

Because of the reported tannin-related changes in microbial fermentation (Vasta et al., 2019), we anticipated changes in the ruminal SCFA profile following the inclusion of GP in diets. For instance, Costa et al. (2018) reported a CT-induced decrease in the relative abundance of fibrolytic bacteria including Fibrobacter succinogenes and R. albus that led to a decrease in ruminal acetate and total SCFA concentrations. Although we did not evaluate potential shifts in the ruminal microbiome in the present study, there were minor changes in the SCFA profile as the dietary content of GP increased. Therefore, this could account for the lack of a diet effect on the duration and area that pH was below the thresholds of 6.2, 5.8, and 5.5. Partial replacement of barley grain with GP resulted in a decrease in the dietary rapidly fermentable carbohydrate content and a concomitant increase in the fiber content in the present study. Therefore, this possibly may have led to the tendency for ruminal propionate proportion to be lowest at the highest dietary GP inclusion level (Calsamiglia et al., 2008). Propionate is the major gluconeogenic precursor in ruminants (Bergman, 1990). Therefore, although not quantified in the current study, it is possible that glucose and, thus, caloric supply was also restricted as dietary GP proportion increased. This then could have also led to an upregulation of skeletal protein degradation to support hepatic and renal gluconeogenesis (Drackley et al., 2001). As in this study, Moate et al. (2014) also did not observe changes in total SCFA concentration when GP (27% of DM) partially replaced alfalfa in lactating cow diets. This was despite indications of altered ruminal bacterial and archael but not fungal or protozoan communities when feeding GP. Changes in the SCFA profile were also minor in that study (Moate et al., 2014).

In the present study, we measured the serum FA profile, which is an indirect indicator of ruminal biohydrogenation (BH). However, besides dietary supply, the serum FA profile is also influenced by other factors including body fat mobilization (Hostens et al., 2012), which was not measured in the present study. Therefore, this necessitates the need for interpretation of our observations with caution. Polyphenolic compounds can suppress the growth and activity of microbes involved in the BH of FA including members of the genus Butyrivibrio (Vasta et al., 2019). This can increase the absorption and transfer of PUFA (e.g., C18:2 n-6) and several BH intermediates (e.g., C18:1 c9) into meat and milk at the expense of SFA (e., C18:0). Therefore, the lack of a diet effect on the major FA including C18:1 c9 and C18:0 when feeding GP suggests negligible changes in the ruminal BH pathways. Moate et al. (2014) reported an increase in the milk content of total SFA, MUFA, and PUFA when lactating cow diets contained 27% GP (DM basis). However, others (Manso et al., 2016; Ianni et al., 2019) did not observe any impact of feeding up to 10% GP on milk and meat content of total SFA, MUFA, and PUFA. Effects could be influenced by the dietary inclusion level, among other factors. However, direct measurement of the BH intermediates in ruminal fluid is still needed to provide a clearer picture of the impact of feeding GP to beef cattle.

The addition of polyphenolic compounds to ruminant diets can also result in a favorable change in the route of N excretion from urine to feces (Ahnert et al., 2015; Norris et al., 2020). However, in the present study, feeding increasing amounts of GP did not reduce environmentally labile urinary N and urea-N excretion in feedlot cattle. This possibly was because ruminal NH₃-N and PUN concentrations also did not decrease as anticipated. Unlike in our study, Greenwood et al. (2012) reported a decrease in PUN and urine NH₃-N concentrations but not UUN concentration when feeding GP to lactating cows. However, fractional urinary N, NH₃- and urea-N excretion (% of total N intake or % of total urine N) data, which is needed for completeness, was not reported by Greenwood et al. (2012). In the present study, there was a quadratic response in fecal N excretion (g/d and % of N intake) as dietary GP content increased. This was similar to the observations Greenwood et al. (2012) made; whereby, fecal N increased by 22% when lactating cows were fed GP. We also noted an increase in fecal NDIN and ADIN excretion following the addition of GP to diets in this study, reflecting the binding effect of the polyphenolic compounds, which restricted N digestibility. Similarly, Koenig and Beauchemin (2018) reported an increase in fecal excretion of fiber bound N following the addition of CT extract (1.33% CT) from Acacia mearnsii in feedlot steer diets. Although it already is a slower process relative to the breakdown of UUN to NH₃-N, increased excretion of bound fecal N could further limit the rate of organic mineralization and, thus, reactive N emissions from feedlot manure (Waldrip et al., 2015). Besides limiting possible ecological damage, the enhanced stability of excreted N also increases the value of manure as a fertilizer for crop production.

Conclusions

The dietary inclusion of GP affected metabolic parameters linked to nutrient supply in feedlot steers. Responses to dietary GP level were mostly quadratic with the impairment of apparent total tract nutrient DM, OM, NDF, and N digestion; estimated microbial N supply; and N retention occurring at the 30% but not at the 15% inclusion level. This suggests that nutrient supply could be compromised beyond a 15% dietary inclusion level of GP, thereby possibly restricting growth performance in feedlot cattle. Increasing the dietary proportion of GP also resulted in an increase in excretion of insoluble bound forms of N (NDIN and NDIN) in feces, suggesting greater stability of manure N. However, we did not observe the anticipated decrease in excretion of UUN, the greatest contributor to reactive N emissions from feedlot cattle. Because there are indications that drying could limit the efficacy of the polyphenolic compounds from GP in modulating ruminal metabolism, further investigation is warranted.

Glossary

Abbreviations

ADF

acid detergent fiber

ADG

average daily gain

ADIN

acid detergent insoluble nitrogen

ADL

acid detergent lignin

BCFA

branched-chain fatty acid

BH

biohydrogenation

BW

body weight

CP

crude protein

CT

condensed tannins

DM

dry matter

DMI

dry matter intake

FA

fatty acid

GP

grape pomace

iNDF

indigestible neutral detergent fiber

MP

metabolizable protein

MUFA

monounsaturated fatty acid

NDF

neutral detergent fiber

NDIN

neutral detergent insoluble nitrogen

OM

organic matter

PD

purine derivatives

PUFA

polyunsaturated fatty acid

PUN

plasma urea-nitrogen

RUP

ruminally undegradable protein

SCFA

short-chain fatty acids

SFA

saturated fatty acids

TMR

total mixed ration

UFA

unsaturated fatty acids

UUN

urine urea-nitrogen

Conflict of interest statement

The authors declare no real or perceived conflicts of interest.