Trade-offs between selection of crude protein and tannins in growing lambs

Abstract

Tannins are phenolic compounds that provide benefits to ruminants due to their protein-binding affinities and antioxidant properties. However, tannins may also have negative orosensorial and postingestive effects that decrease feed intake. This study explored how lambs trade off the ingestion of crude protein (CP) with the ingestion of potentially beneficial and toxic condensed and hydrolyzable tannins, and the ensuing impacts on diet digestibility, animal performance, and blood parameters. Thirty-two lambs were housed in individual pens for 8 wk and had access to 2 isoenergetic diets that varied in the concentration of CP (High-high in protein [HP] or Low-least preferred [LP]) and the presence of a mix of condensed and hydrolyzable tannins (4% DM). Animals were assigned to 4 treatment groups (N = 8 lambs/group) and received a simultaneous offer of: HP and LP (Control); HP and LP + tannins (HP − LP+); HP+ tannins and LP (HP + LP−); and both HP and LP with tannins (HP + LP+). All lambs preferred HP over LP and they avoided tannins in the diets (P < 0.001). Dry matter and CP intake were the lowest in HP + LP− (P < 0.0001), and DMD did not differ between Control and the other groups (P > 0.05), but it was greater for HP − LP + (P < 0.0001). CP digestibility was greater for groups without tannins in HP (P < 0.0001), but average daily gain (ADG) did not differ among treatments (P > 0.05). No differences between Control and HP + LP + were found in plasma antioxidant activity, total phenolic concentration, or haptoglobin concentration (P > 0.05). Intake of tannins was HP + LP+ > HP + LP− > HP − LP + (P < 0.0001), and fecal excretion of nitrogen (N) was HP + LP+ > HP − LP+ = HP + LP− > Control (P < 0.05). In addition, intake of tannins with both diets reduced blood urea nitrogen (BUN) concentration relative to Control (P < 0.05), thus suggesting a shift in the partitioning of N excretion from urine to feces. In summary, lambs prioritized the selection of HP over LP, regardless of the presence of a tannin extract in either or both diets. Nevertheless, lambs modulated their tannin consumption as a function of the specific diet where the tannin extract was added, with increasing levels of intake as tannins were present in just LP, then in just HP, and finally in both diets. Dietary tannins did not constrain ADG and resulted in a shift in the partitioning of N excretion from urine to feces. Such shifts have been found to result in reduced production of environmental pollutants such as ammonia, nitrous oxide, and nitrates.

This study found that lambs offered a choice between diets that varied in crude protein (CP) concentration and the presence of tannins prioritize the ingestion of CP and modulate tannin consumption as a function of the specific diet where tannins were added. Tannin intake did not decrease animal performance and provided potential beneficial effects on the environment through a shift in the fate of nitrogen excretion from urine to feces.

Introduction

Tannins are ubiquitous plant secondary metabolites that have been explored in ruminant production systems since the early 1900s. Many studies have highlighted their antinutritional effects (Price et al., 1979; Butler, 1992), involving postingestive malaise and toxicity (Provenza et al., 2003; Frutos et al., 2004). Nevertheless, moderate concentrations of tannin extracts in the diet (i.e., 3% to 4% DM) may increase dietary protein supply to the intestines with positive impacts on animal performance (Min and Solaiman, 2018). Condensed tannins may form stable complexes with protein in the rumen, reducing rates of proteolysis, blood urea nitrogen (BUN) formation, and urinary N excretion (Makkar, 2003). Protein–tannin complexes are then dissociated under the acidic conditions of the abomasum and alkaline conditions of the distal intestinal tract, enhancing the availability of dietary protein to the host and amino acid flow to the small intestine (Mueller-Harvey, 2006). This change in the fate of dietary N can lead to improvements in production, with increments in animal growth, fertility, milk yield, and parasite tolerance (Waghorn, 2008; Mueller-Harvey et al., 2019). The benefits of protein supply to the intestines are of particular relevance for growing lambs, as animals under this physiological state have high crude protein (CP) requirements, manifested by the selection of high CP diets that match such nutritional targets (Kyriazakis and Oldham, 1993).

Condensed and hydrolizable tannins also provide antioxidant and anti-inflammatory activities to mammals (de Melo et al., 2023), and have positive effects on the oxidative stability of animal products (Liu et al., 2013; Soldado et al., 2021). In addition, tannin consumption can help alleviate the negative clinical impacts of gastrointestinal nematode infections (Hoste et al., 2006), and parasitized lambs increase their preference for tannin-containing feeds (Villalba et al., 2014). This is because ruminants learn to associate the postingestive consequences of eating a specific feed with its orosensorial properties (Provenza, 1995), and sheep prefer the flavors of feeds associated with the provision of nutrients and medicinal secondary compounds to their internal environment (Villalba and Provenza, 1997). On the other hand, tannins typically taste bitter, a sensorial dimension linked to toxins, and are generally disliked and avoided by animals (Glendinning, 1994). In fact, one of the main challenges in studies involving tannins in animal production systems is the potential low palatability of these chemicals, given their astringency and bitterness (Frutos et al., 2004; Jerónimo, et al., 2016). In summary, tannins may alter food preference, and intake and hinder nutrient selection in healthy individuals, although voluntary ingestion of tannins may provide medicinal or prophylactic benefits to ruminants during the process of diet selection.

We hypothesized that the orosensorial and postingestive effects of plant secondary compounds would influence the selection of diets with differing CP concentrations by sheep given the impact of flavor-postingestive feedback interactions in shaping food preferences in ruminants. We further hypothesized that diet selection would reflect the global ratio of costs (toxicity, bitterness) to benefits (antioxidant and anti-inflammatory activities, better protein nutrition, lower rates of rumen proteolysis) when animals could choose an arrangement of diets that vary in CP concentration and tannin availability. Thus, this study aimed to evaluate the trade-offs between the selection of nutrients and a tannin extract by lambs offered choices among diets that varied in the concentration of CP and in the presence of a mix of condensed and hydrolizable tannins. The consequence of such selection was then assessed through measurements of growth performance, digestibility, fecal excretion of N, BUN, total blood phenolic concentration, inflammation, and antioxidant blood parameters.

Materials and Methods

Study site and animals

The study was conducted at the Green Canyon Ecology Center, Utah State University (41°45ʹ59ʹʹ N, 111°47ʹ14ʹʹ W), in Logan, Utah. The study took place from April 20 to June 14, 2022. All procedures were approved by the Utah State University Institutional Animal Care and Use Committee (protocol number 12594).

Thirty-two Rambouillet-Columbia-Suffolk crossbred lambs (3 mo of age and initial average body weight of 22.4 ± 5.8 kg) were kept under a protective roof in individual adjacent pens measuring 1.5 × 2.5 m. Before the study, all lambs were dewormed orally with Ivermectin (0.2 mg/kg body weight [BW]) and grazed on a 0.85 ha pasture of orchardgrass (Dactylis glomerata) and alfalfa (Medicago sativa), and were supplemented with ad libitum amounts of alfalfa pellets for 4 d. Subsequently, lambs were familiarized with their pens and feeders for 12 d prior to starting the study. During this period, animals received an acclimation diet (50% alfalfa pellets, 20% barley, 20% beet pulp, and 10% soybean meal) at an intake of 1.5 kg/lamb/d. Throughout the study, all lambs had free access to water from the municipal line and trace-mineralized salt blocks.

Experimental approach

Lambs were randomly assigned to 4 treatments (N = 8 lambs [5 wethers; 3 females]/treatment) and randomly distributed across pens, considering the variation of sex (ewes and wethers) and weight, resulting in a uniform distribution of animals within each treatment group. Each lamb received a simultaneous offer of 2 isoenergetic rations of similar particle size. Experimental diets were formulated using a mix of alfalfa, beet pulp, barley, and soybean meal, to attain either a low (least preferred [LP] ~14% below values recommended by NRC, 1985 for lambs of 22 kg BW) or a high CP-energy ratio (high in protein [HP] ~80% above values recommended by NRC, 1985; Table 1). Two more rations were formulated with the same ingredients, energy, and protein densities, but containing a condensed and hydrolyzable tannin extract blend (4% DM), LP + and HP + (Table 1). Thus, the 4 treatments were: Control = HP − LP−; HP − LP+; HP + LP−; and HP + LP+. The concentration of tannins selected for this study aimed at providing values that represent those observed in vascular plant materials (5% to 10%; Barbehenn and Constabel, 2011), and in particular within the range of concentrations present in beneficial legumes like sainfoin (Onobrychis viciifolia Scop; 1% to 9% DM condensed tannins), prairie clover (Dalea purpurea Vent; 4% to 9% DM condensed tannins; Mueller-Harvey et al., 2019), and forbs like small burnet (Sanguisorba minor Scop; 4.5% DM hydrolizable tannins; Stewart et al., 2019). The tannin extract blend (in a powder form; ByPro, Silvafeed, Italy) was composed of one-third chestnut tannin extract and two-thirds quebracho tannin extract. Both extracts have been analyzed by matrix-assisted laser desorption/ionization-time of flight (MALDI-TOF) mass spectrometry (Pizzi et al., 2009). Quebracho tannin composition was (DM basis): 84.3% condensed flavan-3-ols (predominantly profisetinidin), 10.7% oligomers of flavan-3-ols (catechin and epicatechin dimers), and 5% carbohydrate derivate (dimers of pentose, monocarboxylic acid of hexose, and 6-carbon sugars). Chestnut tannin composition was (DM basis): 7.9% digalloyl glucose, 5.0% trigalloyl glucose, 16.5% pentagalloyl glucose, and 70.6% oligomers of digalloyl glucose, trigalloyl glucose, and pentagalloyl glucose (Pizzi et al., 2009).

Table 1.

Ingredients and chemical composition (% of DM) of the experimental feeds

	LP−	HP−	LP+1	HP+2
Alfalfa, %	23	47	15	35
Barley, %	1	32	15	40
Beet pulp, %	75	1	65	1
Soybean meal, %	1	20	1	20
Tannins3, %	0	0	4	4
DM, %	92.8 ± 0.1	91.5 ± 0.1	92.3 ± 0.1	90.9 ± 0.2
CP, %	11.9 ± 0.1	25.0 ± 1.9	12.2 ± 0.2	22.8 ± 0.9
ADF, %	25.7 ± 0.4	20.3 ± 1.0	22.4 ± 0.3	16.2 ± 0.3
NDF, %	38.9 ± 0.5	27.0 ± 1.8	34.9 ± 0.5	23.1 ± 0.3
DE4, Mcal/kg	3.2	3.2	3.2	3.2

¹LP with tannin

²HP with tannin

³ByPro, Silvafeed, Italy. A blend composed of 1/3 chestnut tannin extract and 2/3 of a quebracho tannin extract.

⁴Digestible energy determined by NRC (1985) tables.

LP, low protein feed; HP, high protein feed; DM, dry matter; CP, crude protein; ADF, acid detergent fiber; NDF, neutral detergent fiber

Feed intake

Lambs were fed daily between 7:30 and 8:00 a.m. for 8 consecutive weeks. Diets were offered in ad libitum amounts and adjusted so that at least 10% refusals were present in the feeders before the next feeding cycle. Offered diets and refusals were weighed on a daily basis, so that daily ration intake was calculated as the difference between the amounts offered and refused.

Nutrient and tannin intake

During the last 5 d of the study, representative (~20 g) samples of feed offers and refusals were collected daily. Samples were analyzed for chemical composition, and intake of DM, CP, NDF, ADF, and tannins were calculated as: [(Offered diet 1 × concentration of nutrient in the offered diet 1) − (refusal from diet 1 × concentration of nutrient in the refusal from 1)] + [(Offered diet 2 × concentration of nutrient in the offered diet 2) − (refusal from diet 2 × concentration of nutrient in the refusal from diet 2)]. Intake was calculated as g diet (DM basis)/kg of BW.

Fecal output, fecal N excretion, and digestibility of DM, CP, and NDF

Fecal grab samples (20 to 30 g DM) were collected at 9:00 a.m. from the rectum of each lamb (32 samples/day) during the last 5 d of the study and stored in vacuum plastic bags at −20 °C. Subsequently, a composite sample from the feces of each lamb was oven-dried at 70 °C for 72 h, and dried samples were analyzed for chemical composition. Fecal DM output (FO) was determined using the concentration of an internal marker, acid detergent lignin (ADL), in the ration consumed, and in feces (Van Soest, 2018). Fecal output was then determined using the following formula: $\mathrm{F}\mathrm{e}\mathrm{c}\mathrm{a}\mathrm{l}\mathrm{D}\mathrm{M}\mathrm{O}\mathrm{u}\mathrm{t}\mathrm{p}\mathrm{u}\mathrm{t}(g/d)=[DMI(g/d)\times \mathrm{A}\mathrm{D}\mathrm{L}\mathrm{i}\mathrm{n}\mathrm{d}\mathrm{i}\mathrm{e}\mathrm{t}(g/g)]$$/\mathrm{A}\mathrm{D}\mathrm{L}\mathrm{i}\mathrm{n}\mathrm{f}\mathrm{e}\mathrm{c}\mathrm{e}\mathrm{s}(g/g)$ (Cochran and Galyean, 1994). Once fecal output was determined, DM digestibility (DMD) was calculated for each lamb as: $\mathrm{D}\mathrm{M}\mathrm{D}(\mathrm{\%})=\{[\mathrm{D}\mathrm{M}\mathrm{I}(g/d)\mathrm{F}\mathrm{O}(g/d)]/\mathrm{D}\mathrm{M}\mathrm{I}(g/d)\}\times 100$.

CP digestibility (CPD) and NDF digestibility (NDFD) were calculated as:$\mathrm{C}\mathrm{P}\mathrm{D}\mathrm{o}\mathrm{r}\mathrm{N}\mathrm{D}\mathrm{F}\mathrm{D}(\mathrm{\%})=$$[CPorNDFindiet(g/d)-CP\mathrm{o}\mathrm{r}\mathrm{N}\mathrm{D}\mathrm{F}\mathrm{i}\mathrm{n}\mathrm{f}\mathrm{e}\mathrm{c}\mathrm{e}\mathrm{s}(g/d)]/\mathrm{C}\mathrm{P}$$\mathrm{o}\mathrm{r}\mathrm{N}\mathrm{D}\mathrm{F}\mathrm{i}\mathrm{n}\mathrm{d}\mathrm{i}\mathrm{e}\mathrm{t}(g/d)\times 100$. The nitrogen excreted through the feces (g/lamb) was calculated by multiplying fecal output by the nitrogen concentration in feces.

Average daily gain and gain:feed ratio

Animals were weighed before the beginning of the study and every 4 wk. Average daily gain (ADG) was calculated by dividing the weight gain by the number of days elapsing during the period. Gain:feed ratio was calculated by dividing ADG by the daily food intake for the whole experimental period.

Chemical analyses

One representative sample of each diet (HP, LP, HP+, LP+) was taken daily before feeding during weeks 1, 4, and 8. Diet samples were composited, taking approximately 20 g DM from each sample, and used for chemical analyses. Composited diet, refusal, and fecal samples taken during the last 5 d of the study were analyzed in duplicates for DM, CP, ADF, and NDF concentrations. Dry matter was determined by drying the samples at 105 °C for 3 h in a forced air-drying oven (AOAC, 2005; method 934.01). CP was calculated by analyzing the N concentration of the samples using a Leco FP-528 N combustion analyzer (AOAC, 2000; method 990.03). Neutral (NDF) and acid (ADF) detergent fiber were analyzed according to Mertens (2002) and AOAC, 2000 (method 973.18), respectively. Determinations of ADL were conducted with modifications from Robertson and Van Soest (1981) as follows: fiber residue and filter from the ADF step were transferred to a capped tube and 45 mL of 72% sulfuric acid was added. Tubes were gently agitated for 2 h and filtered onto a second filter (same type as above) which was then rinsed, dried, weighed, and finally ashed for 2 h in a furnace to remove lignin organic matter.

Blood analyses

Blood samples with no anticoagulant (without EDTA added; Becton Dickinson Vacutainer System; Becton Dickinson and Company, Franklin Lakes, NJ; 10 mL serum vacutainer tubes) were collected at 8:00 a.m. from each lamb via jugular venous puncture at the beginning (i.e., before the treatments started) and end of the study. After samples were centrifuged (1,500 rpm for 15 min), serum was extracted and stored at −80 °C prior to analysis. Serum samples were submitted to the Utah Veterinary Diagnostic Laboratory (Logan, UT) for urea nitrogen (BUN) analyses, which were performed with a Siemens Dimension Xpand Plus analyzer (Siemens Healthcare Diagnostics, Newar, DE) using Siemens urea N flex reagent, in an enzymatic method which uses urease enzyme in a bichromatic rate technique.

Blood samples were also collected before and after the study, from lambs under treatments that did not receive tannins (Control; HP − LP−) and that had tannins in both diets (HP + LP+; with EDTA added; Becton Dickinson Vacutainer System; Becton Dickinson and Company, Franklin Lakes, NJ; 10 mL plasma vacutainer tubes) following the same collection and storage protocols described above. Plasma samples were analyzed to determine total antioxidant activity using the hydroxyl radical scavenging (DPPH) and the ferric-reducing ability of plasma (FRAP) assays. In the DPPH assay, the % inhibition of the stable 2,2-Diphenyl-1-picrylhydrazyl (DPPH) radical was assessed colorimetrically (Mitra et al., 2013). For the FRAP assay, the reduction of ferric to ferrous ions at low pH was measured colorimetrically (Benzie and Strain, 1996). Plasma samples were also analyzed for total phenolic concentration (TPC) using the Folin-Ciocalteu assay modified by Chen et al. (2019). Finally, haptoglobin was determined using a commercial ELISA kit (Cooke and Arthington, 2013).

Statistical analyses

Results were analyzed using the Mixed Procedure of SAS (SAS® 9.4 Foundation for Microsoft Windows for ×64, Cary, NC: SAS Institute Inc.).

For the response variables associated with the resulting lambs’ selection of diets presented in the choice tests (total feed intake [g; g/Kg BW], ADG [kg/d], animal live weight [kg], gain:feed ratio [g/g], BUN [mg/dL], nutrient and tannin intake [g/kg BW], DMD, CPD, NDFD [%], fecal output and fecal N excretion [g]), a completely randomized mixed-effects model was used with treatment (HP − LP−, HP − LP+; HP+LP−; HP+LP+) as the fixed effect, day as the repeated measures factor and lamb as the random effect. To adjust for the temporal autocorrelation observed in data measured over time, covariance structures were tested to fit the models, based on Akaike’s Information Criterion, where the smallest value represents the best-adjusted model.

During the choice tests, individual diet intake (g) was analyzed using a completely randomized mixed-effects model with treatment (HP − LP−, HP − LP+; HP + LP−; HP + LP+), protein concentration (LP, HP), and tannin presence (Yes, No), as fixed effects, day as the repeated measures factor and lamb as the random effect.

Percentage of hydroxyl radical scavenging activity (DPPH), FRAP (µmol FeSO4 Eq L–1), TPC (mg gallic acid equivalent/mL), and haptoglobin (ng/mL) were only analyzed for the groups that did not receive tannins (HP − LP−) and the group that ingested the greatest amount of tannins (HP + LP+) before the study began and at the end of the study, also using the Mixed Procedure with treatment as a fixed effect and animal as a random effect. Response variables measured before the study began were used as covariates in the analyses. All data were tested for assumptions of normal distribution and homoscedasticity (Proc Univariate; Shapiro–Wilk; P > 0.05). All means were compared using the Tukey–Kramer test.

Results

Feed intake

Total feed intake (g) and intake (g) per kg BW were the lowest in the HP + LP- group (P < 0.05; Table 2). The amounts of diet consumed by lambs in the choice (Table 3) were a function of protein concentration and the presence of tannins in those diets (protein level × tannin concentration interaction; P < 0.0001; Table 3), such that animals preferred HP over LP and they avoided tannins in the diets: HP− > HP+ > LP− > LP + (P < 0.0001; Table 3). When analyzing individual diet intake across treatments, animals in HP − LP− (Control) and HP − LP + did not differ in their intakes of HP (P = 0.18), and lambs in HP + LP− showed greater daily intakes of LP than lambs in HP − LP− (Control; P = 0.03). Lambs in HP + LP + had greater daily intakes of LP + than lambs in HP − LP + (P < 0.0001), and greater daily intakes of HP + than lambs in HP + LP− (P < 0.0001).

Table 2.

Average total feed intake, intake of nutrients, nutrient digestibility, fecal output and fecal N excretion, BUN, and performance by 4 groups of lambs (N = 8) receiving 2-way choices between isoenergetic feeds that varied in the concentration of CP and the addition of a mix (4% DM) of condensed and hydrolizable tannins

	Treatment1
	HP − LP−	HP − LP+	HP + LP−	HP + LP+	SE	P value
Total feed intake, g	1,623a	1,542ab	1,394b	1,605a	88.20	0.03
Intake of DM, g/kg BW	41.3a	42.4a	39.1b	41.7a	0.49	0.04
Intake of CP, g/kg BW	10.1a	10.2a	8.0c	9.2b	0.50	0.02
Intake of NDF, g/kg BW	14.2ab	13.6c	13.8bc	14.4a	0.21	0.02
Intake of ADF, g/kg BW	9.4a	9.4a	8.5b	9.2a	0.20	0.02
Tannin intake, g/Kg BW	—	0.68c	0.86b	1.80a	0.49	<0.0001
Fecal output, g	354a	323b	316b	349a	4.60	<0.0001
Fecal N excretion, g	10.6b	11.4b	11.6b	14.1a	0.24	0.03
DM digestibility, %	76.5ab	77.6a	75.4b	76.2b	0.32	<0.0001
CP digestibility, %	82.3a	80.7a	72.9b	73.4b	1.06	<0.0001
NDF digestibility, %	62.3c	63.8bc	65.2b	67.8a	0.67	<0.0001
ADG, kg/d	0.32	0.28	0.26	0.32	0.03	0.39
Gain:feed ratio, g/g	0.191bc	0.178a	0.184ab	0.193c	0.068	<0.0001
Final weight, kg	44.4	41.6	40.3	44.4	2.75	0.64
BUN, mg/dL	24.4a	19.9ab	21.3ab	18.0b	1.36	0.02

¹Lambs received a simultaneous offer of diets with low (LP) and high (HP) concentrations of CP (Control; HP − LP−); LP + tannins and HP (HP − LP+); LP and HP + tannins (HP + LP−); and both LP and HP with tannins (HP + LP+).

SE, standard error; DM, dry matter; CP, crude protein; NDF, neutral detergent fiber; ADF, acid detergent fiber; ADG, average daily gain; BUN, blood urea nitrogen.

Means within a row followed by different superscript letters differ (P < 0.05).

Table 3.

Average daily intake (g) of individual diets by 4 groups of lambs (N = 8) receiving 2-way choices between isoenergetic feeds that varied in protein concentration (HP and LP) and the addition of a mix (4% DM) of condensed and hydrolizable tannins (Tannin Presence; Yes, No)

	Protein concentration
	HP		LP
	Tannin presence
Treatment1	Yes	No	Yes	No	SE2	P value
HP − LP−	—	992a	—	631b	14.3	<0.05 <0.05
HP − LP+	—	960a	584b	—	15.2
HP + LP−	726a	—	—	669b	13.7	<0.05
HP + LP+	917a	—	690b	—	12.5	<0.05
Protein concentration × Tannin presence	821a	976b	637c	650d	5.5	<0.0001

¹Lambs received a simultaneous offer of diets with low (LP) and high (HP) concentrations of CP (Control; HP − LP−); LP + tannins and HP (HP − LP+); LP and HP + tannins (HP + LP−); and both LP and HP with tannins (HP + LP+). ²Standard error.

^a,b,c,dMeans within a row followed by different lower-case letters differ (P < 0.05).

Nutrient and tannin intake

Intake of CP did not differ between HP − LP− (Control) and HP − LP + groups, but it was lower in HP + LP− (by 20%) and HP + LP + (by 9%) groups (P = 0.02; Table 2). Intake of NDF was the lowest in the HP − LP + and the greatest in the group where tannins were added to both diets (HP + LP+; P = 0.02, Table 2). Intake of ADF was the lowest in the HP + LP− group (P < 0.02, Table 2). As expected, lambs that had tannins in both diets (HP + LP+) had the greatest intake of the tannin extract, and tannin intake was HP + LP+ > HP + LP− > HP − LP + (P < 0.0001, Table 2).

Fecal output, fecal N excretion, and digestibility of DM, CP, and NDF

Lambs in HP − LP− (Control) and HP + LP + showed greater fecal output values than lambs in HP − LP + or HP + LP− (P < 0.0001, Table 2). Fecal excretion of N was HP + LP+> HP − LP+ = HP + LP− > HP − LP− (P = 0.03, Table 2). Thus, fecal excretion of N was the lowest when tannins were not present in any of the diets offered (HP − LP−; Control).

Dry matter digestibility was not different between all groups that had tannins in their diets (HP − LP+, HP + LP−, and HP + LP+) and the Control group HP − LP− (P > 0.05), but greater for HP − LP + than for HP + LP− and HP + LP + groups (P < 0.0001; Table 2). CPD in HP − LP− (Control) and HP − LP + groups was greater than in HP + LP− and HP + LP + groups (P < 0.0001, Table 2), and NDF digestibility was the highest in HP + LP + and the lowest in HP − LP− (P < 0.0001, Table 2).

ADG and feed conversion ratio

Final lamb weight did not differ among groups of lambs (42.66 ± 1.35kg; P = 0.64), as well as ADG (P = 0.39, Table 2). Gain:feed ratio was not different between groups of lambs that had tannins in HP (HP + LP−; HP + LP+), and the Control group HP − LP− (P > 0.05; Table 2), but greater in HP + LP + than in HP − LP + and HP + LP− (P < 0.05). Gain:feed ratio was also greater for HP − LP− than for HP − LP + (P < 0.0001, Table 2).

Blood urea nitrogen

Blood urea N did not differ among treatments for samples taken before the study (P = 0.46). Nevertheless, BUN values at the end of the study were lower for lambs in HP + LP + (tannins in both diets) than for lambs in HP − LP− (Control; P = 0.02, Table 2).

Total phenolics, antioxidant activity, and haptoglobin in blood

No differences were found in DPPH, FRAP, TPC, or in plasma haptoglobin concentrations between lambs that had (HP + LP+), or did not have tannins (HP − LP−; Control) in the offered diets, either before or at the end of the study (P > 0.05; Table 4).

Table 4.

Total antioxidant activity estimated through the hydroxyl radical scavenging (DPPH) and the ferric reducing ability of plasma (FRAP) assays, total phenolic concentration (TPC) and Haptoglobin concentration by sheep at the beginning (Baseline) and at the end of the study by 2 groups of lambs (N = 8) receiving 2-way choices between isoenergetic feeds that varied in the concentration of CP and the addition of a mix (4% DM) of condensed and hydrolizable tannins

		Baseline				End of Study
	HP − LP−1	HP + LP + 2	SE3	P value	HP − LP−1	HP + LP + 2	SE	P value
DPPH, %	10.40	12.48	1.61	0.21	9.02	10.07	1.23	0.40
FRAP, µmol FeSO4 Eq/L	327.72	288.09	38.01	0.31	312.44	281.55	46.34	0.51
TPC, mg galic acid/mL	2.87	2.69	0.17	0.32	2.76	2.70	0.33	0.86
Haptoglobin, ng/mL	726.09	736.08	133.17	0.94	536.40	591.47	52.94	0.31

¹Lambs received a choice between diets containing low (LP) and high (HP) concentrations of CP without the addition of tannins (HP − LP−; Control), and with the addition of tannins in both diets (HP + LP+).

²Standard error.

Discussion

Feeding behavior, nutrient, and tannin intake

All lambs ingested greater amounts of the diet with high CP concentration (HP), regardless of the treatment applied in the choice of diets. The preference for high-protein diets has been reported in growing lambs (Kyriazakis and Oldham, 1993; Askar et al., 2006), explained through the increased protein requirements to support growth relative to mature animals (Kyriazakis and Oldham, 1993; Kyriazakis et al., 1999). In addition, studies comparing cafeteria feeding systems with control groups receiving nutritionally balanced diets find that lambs exposed to cafeteria tests consume greater amounts of CP, surpassing the amounts recommended by NRC tables (Sahin et al., 2003; Rodríguez et al., 2007).

Adding a tannin extract to the most preferred diet HP (i.e., the HP + LP− group), reduced total feed intake per unit of BW (Table 2), consistent with the general intake reduction pattern observed in this study for tannin-containing diets. In contrast, intake was not affected relative to Control (HP − LP−) lambs when both diets in the choice had tannins (i.e., HP + LP+), or when tannins were only present in the LP diet (i.e., HP − LP+). Given this feeding pattern, CP and ADF intakes were also greater for groups HP − LP− and HP − LP+, and lower for group HP + LP−. The main factors that reduce intake in tannin-rich feeds involve low palatability of the diet, resulting from complexes formed with salivary proteins that cause a sensation of astringency (Frutos et al., 2004). Declines in digesta passage rates, triggered by the antibiotic effects of tannins on some rumen microbial populations (e.g., cellulolytic) that reduce forage digestion, may also constrain feed intake (Frutos et al., 2004; Waghorn, 2008; Lamy et al., 2011). Finally, some tannins cause negative postingestive effects like tissue lesions (Dawson et al., 1999) which promote feed avoidance and reduce feed intake (Provenza et al., 1990; Fernández et al., 2012). The addition of tannins to the most preferred diet (HP) in the HP + LP− group prompted lambs to increase their intake of the alternative diet without tannins (LP−), but to an extent -likely due to constraints triggered by unbalanced CP/DE ratios- that did not compensate for the reduced consumption of the HP + diet, leading to an overall depression of feed intake. Protein/energy ratios influence feed intake, as lambs consume more rations with a balanced supply of nutrients relative to those with excesses or deficits (Villalba and Provenza, 2005). Lambs in HP + LP− consumed the lowest amounts of HP, which likely led to a CP/DE imbalance that constrained intake.

Lambs in HP + LP+ ingested more HP+ diet than lambs in HP + LP−, and more LP+ diet than lambs in HP − LP+ (Table 3), thus consuming the greatest amounts of the tannin extract given that they could not avoid consuming these chemicals (i.e., tannins were present in both diets). This pattern of selection paid off in terms of total dry matter, CP, NDF, and ADF intake by lambs in HP + LP+, relative to when the LP diet was tannin-free and the preferred diet had tannins (HP + LP−). The pattern of feed ingestion in HP + LP+ also suggests that lambs’ intake capacity of tannins was greater than when there was a feed alternative to “escape” tannin consumption (e.g., HP − LP + and HP + LP− treatments). Nevertheless, tannin intake was a function of the specific diet where the tannin extract was added, as lambs ate more tannins when these chemicals were in the preferred HP diet (HP + LP−) than when they were mixed in the alternative LP (HP − LP+; Table 2). Collectively, these findings suggest that tannin intake by lambs represented a side-effect of the preference manifested for diets of different CP concentrations. These results support previous findings indicating macronutrient supply is more consequential than tannin presence in feeds at influencing diet selection by sheep-offered rations (Costes-Thiré et al., 2019) or forages (Hernández-Orduño et al., 2015) of variable nutritional composition. Nevertheless, lambs modulated their tannin consumption as a function of the specific diet where the tannin extract was added, with increasing levels of intake as tannins were present in just the LP diet LP (HP − LP+), then in just the preferred diet HP (HP + LP−), and finally in both diets (HP + LP+).

Digestibility and animal performance

Lambs ingesting the preferred diet HP, without the addition of a tannin extract (groups HP − LP− and HP − LP+) did not differ in DM, CP, or NDF digestibilities. In contrast, lambs ingesting greater levels of tannins in their diets (groups HP + LP− and HP + LP+) had lower DM digestibilities than lambs in group HP − LP+, and lower CP digestibilities than lambs in groups HP − LP− and HP − LP + (Table 2). Tannins reduce ruminal degradation of different dietary components through multiple mechanisms, including interference with the attachment of rumen microorganisms to plant cell walls, enzymatic inhibition, and antibiotic effects (Frutos et al., 2004). The reduction in CP digestibility can be attributed to the formation of stable protein–tannin complexes in the rumen (Bunglavan and Dutta, 2013), which reduces rumen proteolysis (Mueller-Harvey et al., 2019) and potentially constrains CP digestibility (Robbins et al., 1987). Several studies show that most tannin-protein complexes are undigested in the small intestine due to the strong protein–tannin molecule bonds that are difficult to break down by intestinal enzymes (Patra and Saxena, 2011; Yanza et al., 2021). This mechanism also explains the greater levels of fecal N excretion observed in lambs that ingested the greatest amounts of the tannin extract in this study (HP + LP+; Table 2). The aforementioned inhibition of ruminal proteolysis caused by tannins also reduces the rates of ammonia formation (Waghorn et al., 1987; Orlandi et al., 2015), which leads to proportional reductions in BUN concentration (Huntington and Archibeque, 2000; Dey et al., 2008; Costa et al., 2021), a pattern observed in this (Table 2) and previous (Marshall et al., 2022) work when a mix of condensed and hydrolyzable tannins was added to the diet. Reductions in rumen proteolysis due to dietary tannins shift the partitioning of N excretion from urine to feces (Deaville et al., 2010; Ahnert et al., 2015). Such shift can potentially be beneficial to the environment when ruminants are fed high levels of protein, given that it can ameliorate environmental N pollution, as fecal N is less volatile than urinary N due to reduced degradation rate, which decreases the production of air pollutants such as NH3 and N2O (Ndegwa et al., 2008; Patra and Saxena, 2011), and potentially lowering the likelihood of waterways and groundwater contamination caused by nitrates (Castillo et al., 2000).

Lambs that consumed the greatest amounts of tannin extracts (HP + LP+) showed the greatest values of NDF digestibility. This finding contrasts the typical reduction of fiber digestion by tannins given that these compounds form complexes with lignocellulose which prevent microbial digestion, either by direct inhibition of cellulolytic microorganisms or fibrolytic enzymatic activity (Patra and Saxena, 2011). Despite the aforementioned differences in digestibility, lambs from all groups showed ADG values, and gain:feed ratios that did not differ between groups that had tannins in HP (HP + LP− and HP + LP+), and the Control group HP − LP−, suggesting similar efficiency in converting feed into gain, even when diets in HP + LP− and HP + LP + groups contained ~2 and 4% tannins, respectively, compounds which diluted the dietary nutrient concentration. Findings regarding tannin effects on ruminant performance are variable, since they depend on many factors, such as type and concentration of tannin, animal species, physiological state, and nutrient supply (Mueller-Harvey et al., 2019). A meta-analysis found that even when tannin intake is correlated to decreased nutrient intake and digestibility, feed efficiency (ADG/DMI) tends to increase (Yanza et al., 2021). In vitro, studies have shown that tannins have anti-methanogenic activity, either directly by inhibiting methanogenic bacteria or indirectly by targeting protozoa (Bhatta et al., 2009; Jayanegara et al., 2015), with reductions in enteric methane emissions that can be as high as 20% to 38% relative to control diets (Waghorn et al., 2002; Pineiro-Vazquez et al., 2018). Thus, a reduction in methane emissions through tannin additives may lead to maintenance or even increments in feed efficiency (Min et al., 2020). In support of this, the same mix of condensed and hydrolizable tannins used in this study fed to heifers promoted a reduction in methanogens relative to animals fed a control diet without tannins (Marshall et al., 2022).

Total phenolics, antioxidant activity, and haptoglobin in blood

No differences in antioxidant activity or TPC were observed between groups of lambs that received the greatest amounts of tannins (HP + LP+) in their diet and Controls (HP − LP−; Table 4). A previous study feeding a quebracho tannin extract (9% DM) to lambs for 60 d reported greater concentrations of phenolic compounds in muscle and antioxidant capacity (using the FRAP test) than for lambs fed a control diet (Luciano et al. 2011). Thus, it is likely that the amount of tannin extract consumed by lambs in the present study (4% DM) was not high enough to show differences in plasmatic antioxidant activity using colorimetric assays (i.e., DPPH, FRAP, and TPC). Variability in antioxidant activity due to dietary tannins can be attributed to multiple factors such as tannin chemical structure and concentration, the composition of the basal diet, and the presence of multiple pro- and antioxidant compounds (Soldado et. al, 2021). These complexities, along with tannin bioavailability, interactions within the digesta and lining of the gastrointestinal tract, as well as with the antioxidant system, make it challenging to consistently observe the antioxidant effects of dietary tannins using colorimetric assays, which are inherently more variable and less sensitive than mass-spectrometry based methods (Granato et al., 2018).

Haptoglobin is an α2-globulin synthesized by the liver during the acute-phase response, representing a useful biochemical marker for evaluating the acute inflammatory process in mammals (Yoshioka et al., 2002). Circulating concentrations of haptoglobin in healthy sheep are low or undetectable, although values increase significantly during infection and inflammation (Kostro et al., 2009). In addition, augmented plasma concentrations of haptoglobin have been regarded as a reliable indicator of sheep susceptibility to environmental stressors such as transport (Piccione et al., 2012). In turn, polyphenols like condensed tannins are anti-inflammatory and may thus contribute to attenuating the production of acute-phase proteins like haptoglobin. For instance, lambs infected with the gastrointestinal parasite Haemonchus contortus and supplemented with green tea polyphenols showed lower serum concentrations of haptoglobin than infected animals receiving a control diet (Zhong et al., 2014). Nevertheless, haptoglobin concentrations in infected (Zhong et al., 2014) or stressed (Piccione et al., 2012) sheep were 10³ times greater than those observed in HP − LP− and HP + LP + groups (Table 4), suggesting that lambs in the present study had haptoglobin concentrations typically observed in healthy individuals, which were not further modified by tannin ingestion at the dose of 4% DM.

Conclusion

Lambs in this study prioritized the selection of a high- over a low-protein and isocaloric diet, explained through their increased CP needs during growth, regardless of the presence of a tannin extract in either or both diets. Thus, tannin intake was a side-effect of the preference manifested for diets of different CP concentrations. Nevertheless, lambs modulated their tannin consumption as a function of the specific diet where the tannin extract was added, with increasing levels of tannin intake as the presence of bioactives changed from being present in the low-protein (HP − LP+) to the preferred high-protein diet (HP + LP−), and then with these bioactives present in both diets (HP + LP+). The high dietary concentration of tannins did not influence antioxidant state, plasmatic phenolics, or haptoglobin concentration relative to Control animals. Despite some reductions in feed intake and digestibility, dietary tannins didn’t constrain ADG or gain:feed ratio relative to the Control treatment. Lambs consuming the greatest amounts of tannins (HP + LP+) showed lower BUN concentration and greater fecal N excretion values. This shift in the partitioning of N is potentially beneficial to the environment, as fecal N is less volatile than urinary N due to a reduced degradation rate, which decreases the production of environmental pollutants such as ammonia, nitrous oxide, and nitrates.

Glossary

Abbreviations

ADF

acid detergent fiber

ADG

average daily gain

ADL

acid detergent lignin

BUN

blood urea nitrogen

BW

body weight

CP

crude protein

CPD

crude protein digestibility

DPPH

% inhibition of 2,2-diphenyl-1-picrylhydrazyl radical assay

DM

dry matter

DMD

dry matter digestibility

FRAP

ferric reducing ability of plasma assay

HP

feed high in CP

LP

feed low in CP;

N

nitrogen

NDF

neutral detergent fiber

NDFD

neutral detergent fiber digestibility

TPC

total phenolic concentration

Conflict of interest statement

There are no conflicts of interest associated with this research.

Author Contributions

Marina Terra-Braga (Conceptualization, Formal analysis, Methodology, Writing—original draft), Cesar Poli (Conceptualization, Supervision), Jalise F. Tontini (Data curation, Formal analysis, Writing—review & editing), Muhammad Ahsin (Methodology), Stephan Van Vliet (Formal analysis, Methodology, Writing—review & editing), and Juan Villalba (Conceptualization, Data curation, Funding acquisition, Methodology, Resources, Supervision, Writing—review & editing)