Effects of Tannin-Rich Supplements on Immune Response in Goats and Beef Cattle: A Collection of Controlled Feeding Trials

Abstract

Emerging research evokes selection for various plant secondary compounds as a potential driver of ruminant diet selection, through animals’ evident ability to rectify deficiencies and even self-medicate. This idea was assessed by comparing physiological responses to vaccination challenges of animals fed diets of differing phytochemical composition. In the first of three separate trials, goats were placed in individual pens and fed one of three treatments in a completely randomized design. Treatments in Trial 1 consisted of redberry juniper (50 g) and shin oak (50 g). In Trial 2, goats were fed rations containing grape and blueberry pomace at an inclusion rate of 20%. In Trial 3, black Angus heifers were fed rations containing grape and blueberry pomace at an as-fed inclusion rate of 6%. Average daily gain, intake, and blood chemistry were assessed following vaccination health challenges. In Trial 1, goats fed shin oak had higher (p < 0.05) blood globulins. Trial 2 revealed no treatment group differences in average daily gain (ADG), intake, or blood parameters evaluated. In Trial 3, no difference occurred in blood parameters; however, intake following inoculation was significantly greater (p < 0.05) for heifers with grape/blueberry pomace included in their rations. In conclusion, phytochemicals, specifically condensed tannins, may have the ability to enhance immune response in ruminants, but further research is required, and these effects likely depend upon the source, structure, and dose of tannins or parent plant materials offered.

1. Introduction

Herbivores and plants have co-evolved for millions of years, yet the depth and scope of these interactions are not fully understood [1]. Researchers have long observed that ruminants consistently consume minor amounts of forage and browse considered less desirable or even toxic when other nutritious plants are readily available [2,3,4]. An emerging explanation is the theory that herbivores can benefit not only from the primary compounds within plants, but their secondary compounds as well [5,6,7,8].

While the secondary compounds produced by plants, also known as phytochemicals, are primarily thought to function as chemical defenses against consumption by herbivores, these compounds may also function as insecticides, allelopathic agents, or pollinator attractants [9,10,11,12]. Secondary compounds are effective at limiting consumption by ruminants, if they induce aversive post-ingestive feedback during foraging [4,13]. Consumption of large doses of secondary compounds can cause toxicosis, but occasions when ruminants voluntarily consume minor doses indicate the presence of extra-nutritional benefits [14]. When phytochemicals are consumed by ruminant animals, they can exhibit some of the same protective properties in the animal that occur initially within the plant, lending to the antimicrobial, antiparasitic, anti-inflammatory, antioxidant, and even immunomodulatory effects that have been observed in ruminants consuming phytochemicals [14,15,16,17]. Indeed, herbivore self-medication through modified foraging behavior is well documented [18,19,20,21,22].

In this study, we evaluate the potential immune-stimulant properties of two phytochemical-containing shrubs: tannin-rich shinnery oak (Quercus havardii Rydb.) and terpene-rich redberry juniper (Juniperus pinchotii Sudw.). While these particular species are common in our ecoregion, the Quecrus and Juniperus genera occur in many rangelands across the globe [23,24] and are generally rich in tannins and terpenes, respectively [25,26,27]. Shinnery oak and redberry juniper have traditionally not been considered beneficial forage or browse species and are sometimes removed via rangeland improvement projects [28,29]. In the Southwest and Southern Great Plains, livestock poisoning or other productive losses from both hydrolysable and condensed tannins are commonly caused by shinnery oak [30]. Juniper-induced toxicosis is rare, but intake is typically limited by the secondary compounds it contains [31]. In addition to shinnery oak and redberry juniper, we evaluate the immune-stimulant properties of tannin-containing grape and blueberry pomace. These ingredients are byproducts of juice or wine production and thereby might represent a cost-effective method for supplying medicinal compounds to animals in confinement [32,33].

Tannins, a class of long-chain polyphenolic molecules, are the most common of all naturally occurring plant secondary compounds [34]. These molecules are defined by their ability to precipitate various organic molecules, especially proteins [34,35]. Tannins are commonly further classified as either hydrolysable or condensed [12,34]. While condensed tannins, also known as proanthocyanidins, have a greater impact on feed digestibility (due to the tannin-protein complex being insoluble in the rumen), hydrolysable tannins are often more acutely toxic, as they are rapidly degraded into pyrogallol and gallic acid, which damage the gut lining, leading to stomach or kidney hemorrhaging and even death at high doses [30,34]. However, the positivity or negativity of tannins’ effects on livestock are dose-dependent, leading to a double-edged-sword dynamic [35,36,37].

Low doses of tannins have been associated with a myriad of health benefits in livestock. Tannin molecules act as potent antioxidants [38,39,40], and possess anti-inflammatory [15,41], anti-parasitic [42,43,44,45,46], and anti-microbial properties [47,48,49]. Tannins can also allow for increased intake of another potentially toxic compound in ruminants, saponins, by chelating with saponin molecules in the digestive tract [3]. Furthermore, tannins have improved production, health markers, protein profiles, and lactose content of milk in ruminants [50,51,52]. Condensed tannins can also reduce methane emissions and pasture bloat in cattle [32]. Tannin-containing grape pomace has been demonstrated to improve blood metabolite profiles and milk quality [53], reduce methane emissions [50], and possess strong antimicrobial properties [47]. In addition to the effects listed above, recent evidence suggests that tannins may also improve animal immune response. Enhanced cell mediated immune responses occurred in sheep, following condensed tannin supplementation [40,46]. Tannin extract is also shown to up-regulate expression of genes associated with innate and adaptive immunity in sheep, goat, and cattle blood [54]. Tannins, coupled with direct-fed microbials, improved feed to gain ratios, average daily gain, and most relevantly, immune response, as measured by serum haptoglobin levels [55].

Terpenes are one of the largest, most structurally diverse, and ubiquitous groups of plant secondary compounds [12,56]. Terpenes are formed from chains or rings of isoprene units, which, when modified, are termed terpenoids (hereafter, “terpene” will refer to both true terpenes and terpenoids) [57]. Terpenes are common on rangelands and in ruminant diets [31]. As with tannins, researchers are beginning to evaluate the health benefits of these compounds for livestock production. Experiments have demonstrated positive effects of terpenes on antioxidant status in Holstein [58], water buffalo calves [59], and small ruminants [60]. Terpenes have also been effective anthelmintic agents in sheep and goats [61]. While terpenes’ effects on rumen microbiota and fermentation patterns are variable, the emerging consensus indicates positive effects on feed efficiency and animal performance in multiple livestock species [59,60,62]. The addition of terpene-rich extracts, such as essential oils, in ruminant diets has also increased milk production and improved milk fat content [59,60,63]. Terpenes are commonly recognized for their anti-inflammatory effects [59,62,64]. They can directly affect inflammation by blocking the synthesis or secretion of inflammatory mediators, especially pro-inflammatory cytokines [62]. This ability may indicate a relationship between terpene consumption and immune response. Supplementation of terpene-rich plant leaves improved both cell-mediated and humoral immunity in water buffalo [59]. Similar results have been observed in pigs, with terpene-containing supplements leading to increased immunoglobulins production [65,66].

In this study, we aim to provide data on whether shinnery oak and redberry juniper have potential value, in certain doses, within ruminant diets, by measuring their effect on animal immune response. Furthermore, by using grape and blueberry pomace as a source of tannins, we aim to evaluate if a similar effect might be achieved cost-effectively within animal feeding operations. Furthermore, we aim to provide some insight as to the potential mechanisms that might be at play. Little research has sought to outline the chemical interactions between these compounds (i.e., condensed tannins and monoterpenes) and specific immune cells or molecules [32]. We hypothesize that phytochemically rich plants in the diets of goats and cattle will improve their innate and adaptive immune response.

2. Materials and Methods

Prior to experimentation, all procedures were reviewed and approved by Angelo State University’s Institutional Animal Care and Use Committee (Approval #2020–105 and Approval #2022-105). This study was conducted at the Management, Instruction, and Research (MIR) Center in San Angelo, Texas (Lat: 310°34′8.99″ N, Lon: 1000°32′26.399″ W). Trial 1 was performed in the Fall of 2020 to evaluate immune response to phytochemical-containing plant treatments in weaned goats. Juniperus pinchotii and Quercus havardii were selected for Trial 1 because of their regional commonality and high monoterpenoid and condensed tannin content, respectively [23,30]. While forage monoterpene and condensed tannin contents were not directly analyzed in Trial 1, published values for these compounds are included in

Table 1. Plant secondary compound values for treatment additions in (a) Trial 1 and (b) Trial 2. Total volatile oil (composed of 12 specific terpenoids) concentrations in redberry juniper leaves were determined by Adams et al. (2013) [68] via an HP5971 MSD mass spectrometer. Total condensed tannin content of shinnery oak leaves was determined by Min et al. (2008) [69] via a butanol-HCL colorimetric procedure. Percent (%) tannin types within grape and blueberry pomace used in Trials 2 and 3 were derived from protein precipitable phenolics assays.

Table 1a: Trial 1
Source	Volatile Oils (% DM)		Condensed tannin (% DM)
Redberry juniper [68]	0.94–1.08
Shinnery oak [69]			≈7.9
Table 1b: Trial 2
Source	Extractable tannin (%)	Protein bound tannin (%)	Fiber bound tannin (%)	Total condensed tannin (%)
Blueberry pomace	1.42	3.35	0.00	4.77
Grape pomace	0.81	4.10	1.56	4.91
Grape–blueberry mix	1.11	3.73	0.78	4.84

a for comparison. Trial 2, conducted as a follow-up, was carried out in the Spring of 2021 to evaluate goat immune response specific to condensed tannins offered in the form of grape (Vitis spp.) and blueberry (Vaccinium spp.) pomace. Trial 3 was conducted in the Fall of 2022 to evaluate the same parameters in Angus heifers, also using condensed tannins from grape and blueberry pomace. “Pomace” describes the residual skin, seed, and stalk by-products obtained in wine and juice processing. Grape and blueberry pomace were selected in Trials 2 and 3 because of their well-described tannin content [67,68,69,70]. Samples of the grape and blueberry pomace were analyzed by the Texas A&M AgriLife Research and Extension Center in Stephenville, Texas, via a protein precipitable phenolics assay [71]. This method was used, due to our focus on condensed tannins, rather than total polyphenols. According to this assay, blueberry pomace contained 4.77% condensed tannins, while grape pomace contained 4.91% condensed tannins (Table 1b). The grape and blueberry pomace mix used in Trial 3 contained 4.84% condensed tannins. Extractable tannins (those not bound to fiber or protein) were 1.42% for blueberry, 0.81% for grape, and 1.11% for the mix (Table 1b).

2.1. Trial 1

Spring-born Spanish wether goats (120–160 days old) were obtained from the Owens Ranch in Barnhart, Texas, in August 2020. Wethers were weaned at approximately 90 days of age and reared on pasture until transported (100 km) to the Angelo State University MIR. The forages subsequently offered in this study were completely novel to all goats at the time of introduction. Redberry juniper clippings offered in this trial were harvested from the study site ranch and shin oak clippings were harvested from a nearby ranch located in Colorado City, Texas. Leaf clippings from both species were refrigerated after harvest and offered fresh to animals within a few days of harvest. General nutritional analyses of clippings were not conducted.

Upon arrival, goats were held in a large communal pen with ad libitum access to water and hay for a 14-day acclimation period. Thereafter, wethers (n = 30) were weighed (mean = 26.16 kg, SD = 3.37 kg), randomly assigned to individual treatment pens (1 m by 1.5 m), and equally allocated to one of three diets (10 goats per treatment). The treatments were as follows: a control group (hereafter, CG) fed only a non-pelleted total-mixed ration as a basal diet, the first treatment group (hereafter, RJ) fed the basal diet and also separately offered redberry juniper (Juniperus pinchotii Sudw.) leaf clippings, and the second treatment group (hereafter, SO) fed the basal diet and also separately offered shinnery oak (Quercus havardii Rydb.) leaf clippings. Ingredients and nutrient content of the basal diet are provided below (

Table 2. Ingredients (% as fed basis) and nutrient content (% dry matter) of the basal ration fed to goats to meet maintenance requirements in Trial 1, as well as the control ration used in all trials. The ration was formulated at 93.19% dry matter (DM). DE stands for digestible energy and TDN stands for total digestible nutrients.

Ingredients/Nutrients	(%)
Alfalfa Pellets	10.0
Cotton Seed Meal	12.5
Cottonseed hulls	31.5
Cane molasses	3.5
Mineral Premix	2.5
Corn	40.0
DE	2.6 Mcal/kg
TDN	59.0
Crude Protein	14.5
Crude Fiber	14.2

). Goats were allowed a 7-day acclimation period in pens, during which only the basal ration was offered. This was followed by a 7-day conditioning phase where RJ and SO groups were offered their respective forages daily, in addition to the basal diet, to facilitate acceptance of the novel feedstuffs.

During the 14-day acclimation and conditioning period, goats (SO = 4 goats, CG = 4 goats, RJ = 2 goats) exhibited clinical signs indicative of coccidiosis and heavy internal parasite loads, including scours, lethargy, and feed refusal. In response, all goats (n = 30) were treated prophylactically to ensure herd health and consistency across treatments. A 5-day Corid regimen was initiated on day 7, followed by a single 3 mL dose of Valbazen administered on day 12 to address persistent symptoms and prevent further complications. Following the 14-day acclimation and conditioning period, a 14-day experimental feeding period commenced.

2.1.1. Experimental Feeding Protocol

Daily treatment and basal feed intake were monitored throughout the project. Redberry juniper and shinnery oak clippings (50 g per animal, as-fed basis) were offered each morning to goats in the RJ and SO treatment groups, respectively, for a 30 min period. This baseline amount was selected to promote gradual acceptance of these novel, chemically defended forages, while minimizing the risk of aversive post-ingestive feedback associated with plant secondary compounds such as monoterpenes (in juniper) and tannins (in shinnery oak). Previous studies have used 100 g as a starting point for feeding redberry juniper in individual pens, increasing the amount daily based on animal acceptance [72]. Given that goats in this study were smaller in body weight (~26 kg vs. ~29 kg in prior studies), and that intake in those trials was variable with signs of early aversion, a more conservative 50 g baseline was selected to ensure safe exposure and reduce the likelihood of intake suppression. This feeding rate is further supported by Deeds et al. (2010) [3], who used 50 g as an initial offering for shinnery oak—demonstrating the suitability of this approach when introducing novel, chemically defended shrubs in controlled settings.

After 30 min, remaining shin oak and juniper orts (“refusals”) were collected, weighed, and recorded. Following treatment feedings, all animals received the basal diet, which remained available for approximately 24 h. Based on 3% of initial body weight (BW), basal rations were formulated to 93.19% dry matter (DM) and provided at 0.785 ± 0.101 kg as-fed (0.731 ± 0.094 kg DM) per animal daily. Refusals from the basal diet were collected and weighed each morning prior to treatment offerings. Fresh water was provided ad libitum throughout the study. The experiment concluded after 14 days, at which point individual pen feeding ceased. Goats were then removed from individual pens and weighed. Animal performance was evaluated by comparing initial weights, final weights, average daily gain (ADG), and feed intake over time.

2.1.2. Vaccination Challenge and Bloodwork Procedures

Immediately following the feeding trial and removal of goats from individual stalls and weighing, all animals were subjected to an immune challenge using an Enterotoxemia Type C and D vaccination (Bar-Vac CD/T) at the recommended dosing rate to evaluate each animal’s immunological response. No animals used in this study were previously exposed to this vaccination. Rectal body temperature was recorded for all animals before administering vaccine and at 48 and 72 h post-dosing. Likewise, a visual assessment was performed at 0, 48, and 72 h to ensure animal well-being and detect any adverse vaccine reactions. Visual assessment included monitoring of all animals for clinical signs of illness or discomfort. Goats were held in group pens and continued to receive the basal ration (formulated at 3% BW as fed; 93.19% DM) throughout the post-vaccination monitoring period.

Whole blood and serum were collected and analyzed at weekly intervals on days 0, 7, 14, and 21 post-vaccination. Jugular venipuncture was used to collect blood into ethylenediamine tetra-acetic acid (EDTA) and serum separator tubes. EDTA whole-blood samples were used to prepare blood smears for complete blood counts (CBC), performed at the Texas A&M Veterinary Medical Diagnostic Laboratory (TVMDL). Serum samples were analyzed for total protein, albumin, globulin, and albumin-to-globulin ratio. Bloodwork procedures followed the same protocol for all time points.

2.2. Trial 2

Fall-born Spanish/Boer cross wether goats (120–160 days old) were obtained from the Owens Ranch in Barnhart, Texas, in March 2021. After transport to the ASU MIR (100 km), goats were weighed (mean = 24.65 kg, SD = 3.32 kg) and randomly allocated to treatment pens in a completely randomized design. To proactively prevent illness observed in Trial 1, all goats were dewormed immediately upon placement into pens.

Animals were equally and randomly allocated to one of three dietary treatments (10 goats per treatment). The treatments offered were: control group fed a non-pelleted total mixed ration (hereafter, CG), first treatment group (hereafter, GP) fed the control ration also containing 20% grape pomace (0.20 kg/kg DM), second treatment group (hereafter, BP) fed the control ration also containing 20% blueberry pomace (0.20 kg/kg DM). Other experiments with grape pomace have used similar inclusion rates ranging from 0.05 to 0.20 kg/kg DM [47,50,53,73]. In GP and BP rations, grape and blueberry pomace were partially substituted for cottonseed hulls. Both grape/blueberry pomace [53] and cottonseed hulls are high in fiber [74,75], serving as the primary roughage constituent in the feed rations. Roughage substitution is a common approach when incorporating pomace into feed rations to ensure macronutrient profiles remain similar [73]. Ingredients of all diets are provided below (

Table 3. Ingredients (% as fed basis) and nutrient contents (% dry matter; DM) of rations fed to goats to meet maintenance requirements Trial 2. Control ration was formulated at 93.19% DM. Pomace rations were formulated at 93.55% DM. Nutrient contents of treatment rations were extrapolated, not measured directly. DE stands for digestible energy and TDN stands for total digestible nutrients.

Ingredients/ Nutrients	Grape Pomace Ration (%)	Blueberry Pomace Ration (%)	Control (%)
Alfalfa Pellets	10.0	10.0	10.0
Cotton Seed Meal	12.5	12.5	12.5
Cottonseed hulls	11.5	11.5	31.5
Grape Pomace	20.0	--	--
Blueberry Pomace	--	20.0	--
Cane molasses	3.5	3.5	3.5
Mineral Premix	2.5	2.5	2.5
Corn	40.0	40.0	40.0
DE	2.5 Mcal/kg	2.5 Mcal/kg	2.6 Mcal/kg
TDN	56.5	56.2	59.0
Crude Protein	16.1	14.7	14.5
Crude Fiber	8.8	8.1	14.2

). Individual animal rations were formulated at 3% BW as-fed (93.19% DM), to meet daily maintenance requirements [76]. As in Trial 1, wethers were given seven days to acclimate to individual pens. During this period, treatment rations were offered to goats in GP and BP groups to condition acceptance of the novel dietary feedstuff. Following the initial seven-day acclimation and diet conditioning, a 21-day feeding trial began.

2.2.1. Experimental Feeding Protocol

Daily feed intake was monitored throughout the 21-day feeding trial. All animals received treatment rations at approximately 0800 each day. Based on 3% of initial body weight (BW), total mixed rations were formulated to 93.55% DM and provided at 0.740 kg ± 0.10 SD as-fed (0.692 ± 0.093 kg DM) per animal daily. Rations remained available to the animal for approximately 24 h. Orts from all rations were collected, weighed, and recorded each morning prior to delivering fresh feed. Fresh water was provided ad libitum. Animal performance was evaluated by comparing initial weights, final weights, average daily gain (ADG), and feed intake over time.

2.2.2. Vaccination Challenge and Bloodwork Procedures

All animals were subjected to an immune challenge using an Enterotoxemia Type C and D vaccination (Bar-Vac CD/T) at the recommended dosing rate to evaluate each animal’s immunological response. Unlike Trial 1, the vaccination was administered to goats in Trial 2 upon commencement of the feeding trial (day 0). Whole blood and serum were collected and analyzed on days 0, 7, and 14 post-vaccination, coinciding with the feeding experiment. No animals used in this study were previously exposed to the vaccination. Bloodwork collection and analysis methodology followed procedures described in Trial 1.

Rectal body temperatures were not measured in Trial 2, as temperatures collected in Trial 1 indicated no discernible differences among treatment groups. Although methodology differed in Trial 2, efforts focused on blood-based immune measures (CBC and serum chemistry) as more reliable indicators of immunological status. This refinement aligns with recommendations for minimizing animal handling stress and prioritizing metrics with greater diagnostic utility. To ensure animal well-being and detect any adverse vaccine reactions, goats were closely monitored for clinical signs of illness or discomfort throughout the post-vaccination period.

2.3. Trial 3

20 spring-born heifer calves, approximately 8 months of age (ranging from 7 to 9 months), from Angelo State University’s registered Angus herd were weighed (mean = 306.95 kg, SD = 51.66 kg) and sorted into individual treatment pens in a completely randomized design (10/treatment). Individual pens were identical, all measuring 3 m by 9 m and containing a feed trough and water tub. Animals were given seven days to acclimate to the individual pens while still receiving their weaning feed ration. The treatments offered were: control group fed a total mixed ration, and a treatment group fed a similar base ration to the control group, but also containing a 50/50 grape and blueberry pomace mix. For the reasons outlined in the Trial 2 methods, pomace was partially substituted for cottonseed hulls. Due to shortages at the time of this trial, we reduced the inclusion rate of pomace to 6% (0.06 kg/kg DM). Ingredients of all diets are provided below (

Table 4. Ingredients (% as fed basis) and nutrient contents (% dry matter) of the control and treatment rations fed to heifers in Trial 3. Control ration was formulated at 93.19% DM. Treatment ration was formulated at 93.26% DM. Nutrient contents of treatment rations were extrapolated, not measured directly.

Ingredients/Nutrients	Treatment (%)	Control (%)
Alfalfa pellets	10.0	10.0
Cotton seed meal	12.5	12.5
Cottonseed hulls	25.5	31.5
Grape/blueberry pomace	6.0	0.0
Cane molasses	3.5	3.5
Mineral premix	2.5	2.5
Corn	40.0	40.0
DE	2.6 Mcal/kg	2.6 Mcal/kg
TDN	58.2	59.0
Crude protein	14.8	14.5
Crude fiber	12.5	14.2

).

2.3.1. Experimental Feeding Protocol

The experimental trial began in November 2022 and lasted 22 days. Individual animal rations were formulated to meet daily maintenance requirements [77], and rations were formulated to 93.26% DM. At approximately the same time each day, each heifer was fed their respective feed ration in an amount equivalent to 4% as fed (3.73% ± 0.10 kg DM) of their most recent body weight. Body weights were collected at the beginning of the acclimation period (7 days prior to the start of Trial 3), midway (on day 15 of Trial 3), and at the conclusion of the trial (day 22 of the trial). Just prior to feeding, feed refusals from the previous day were collected, weighed, and recorded. A total of 4 animals (3 treatment, 1 control) experienced bloating for various lengths of time totaling 8 animal-days (7 treatment, 1 control). During these instances, bloated animals received hay until symptoms subsided and no data was recorded (total of 8 missing data slots) for affected animals during those periods. Intake data was standardized to each heifer’s predicted daily body weight as interpolated from three data points taken on days 7, 15, and 22 of the feeding trial. This calculation assumes linear average daily gain between data points. If heifers consumed all offered feed on any particular day, their feed intake was set equivalent to 4% (as fed basis) of their body weight for that day. Animal performance was evaluated by comparing weights at three timepoints, average daily gain (ADG), and feed intake over time.

2.3.2. Vaccination Challenge and Bloodwork Procedures

On day 15 of Trial 3, body temperatures were recorded and blood was collected for all heifers to establish an immune baseline. Immediately following blood and temperature collection, heifers were inoculated with a dose of modified live bovine rhinotracheitis-virus diarrhea-parainfluenza 3-respiratory syncytial virus vaccine (Bovi-Shield Gold 5). This vaccination was designed to prompt an immune response for both Infectious Bovine Rhinotracheitis (IBR) and Bovine Viral Diarrhea (BVD). Body temperatures were recorded again 24 h later. Seven days following inoculation (day twenty-two of the feeding trial), blood and body temperatures were collected again from all heifers. The feeding trial concluded at this point. Whole blood treated with EDTA was analyzed by the Texas A&M Veterinary Medical Diagnostic Laboratory in College Station, Texas, for total white blood cells, neutrophils, lymphocytes, monocytes, eosinophils, basophils, red blood cells, hemoglobin, hematocrit, mean corpuscular volume, mean corpuscular hemoglobin, mean corpuscular hemoglobin concentration, platelets, and fibrinogen. Serum was separated from whole blood and analyzed by the Texas A&M Veterinary Medical Diagnostic Laboratory in Canyon, Texas, for total protein, albumin, IBR immunoglobulin titers, BVD immunoglobulin titers, and albumin–globulin ratio.

2.4. Statistical Analyses (All Trials)

We analyzed the data using a linear mixed-effects model (LMM) in JMP (SAS 2007), with restricted maximum likelihood (REML) estimation. Fixed effects included treatment group (categorical), day (discrete numeric), and their interaction, while individual animals were included as random intercepts. This allowed us to evaluate overall differences among groups and group differences that arise over time. To evaluate group differences at specific time points, post hoc pairwise comparisons were conducted using Tukey’s Honest Significant Difference (HSD) test to control for family-wise error rate. Degrees of freedom for fixed effects were approximated using the Kenward–Roger or Satterthwaite method, and significance testing was based on Type III F-tests. An alpha level of 0.05 is considered significant.

3. Results

3.1. Trial 1

Weight parameters, including weight, weight × time, and overall average daily gain (ADG) were similar among treatment groups (ADG: CG = 0.16 kg, SO = 0.13 kg, RJ = 0.12 kg, p = 0.725) (

Table 5. Least square mean (±SE) values for average daily gain (ADG) and blood serum parameters of goats in control (CO), shinnery oak (SO), and redberry juniper (RJ) treatment groups during Trial 1. For ADG, blood serum parameters, and complete blood count values of goats fed rations containing blueberry pomace (BP), grape pomace (GP), or no pomace (control; CG) during Trial 2, and for ADG, blood serum parameters, and complete blood count values of heifers fed a base ration only (Control), and a ration containing grape and blueberry pomace (treatment) during Trial 3. Due to multiple post-inoculation sampling time points, Trials 1 and 2 are reported as averages across time points in the table below. Trial 3 only consisted of one baseline and one post-inoculation measurement, the latter of which is reported in the table below. Rows with letters denote significant differences across treatment groups. (p < 0.05); treatments sharing the same letter are not different (p > 0.05). An absence of letters indicates no differences (p > 0.05).

Trial 1	Parameter	CG	SE	SO	SE	RJ	SE
	ADG (kg/day)	0.16	0.04	0.13	0.04	0.12	0.04
	Globulins (g/dL)	2.90 b	0.11	3.36 a	0.11	2.88 b	0.11
Trial 2	Parameter	CG	SE	GP	SE	BP	SE
	ADG (kg/day)	0.08	0.02	0.04	0.02	0.08	0.02
	Total WBCs (K/uL)	14.17	1.10	12.90	1.03	13.46	1.03
	RBC (M/uL)	19.62	0.70	18.48	0.66	19.55	0.66
	Lymphocytes (K/uL)	7.55	0.57	7.10	0.56	7.63	0.56
	Neutrophils (K/uL))	4.87	0.57	5.05	0.54	4.51	0.55
	Globulins (g/dL)	3.34	0.13	3.13	0.13	3.22	0.13
Trial 3	Parameter	Control	SE	Treatment	SE
	ADG (kg/day)	1.33	0.62	1.07	0.62
	Total WBCs (K/uL)	8.05	0.62	8.27	0.59
	RBC (M/uL)	7.79	0.21	8.08	0.21
	Lymphocytes (K/uL)	5.12	0.54	4.88	0.51
	Neutrophils (K/uL))	2.76	0.54	3.09	0.51
	IBR Immunoglobulins (g/dL)	2.41	0.09	2.51	0.09
	BVD Immunoglobulins (g/dL)	2.44	0.10	2.49	0.10

and

Table 6. Fixed effect test summaries for response variables of primary interest in Trial 1. Any response variables not included in this table do not have a significant (p > 0.05) group × day interaction. The column “Days with group differences” lists any specific days (or time periods) for which Tukey’s HSD revealed differences between treatment and control groups at the 0.05 level.

Response	Predictor	p-Value	Days with Group Differences
Clippings Intake	Group	0.1064	2, 3
	Day	0.0053
	Group × Day	0.0244
Basal Ration Intake	Group	0.2625	None
	Day	<0.0001
	Group × Day	0.1590
Average Daily Gain	Group	0.7252	None
Body Temperature	Group	<0.0001	72 hr
	Day	0.1590
	Group × Day	0.2645
Globulins	Group	0.0053	0, 7
	Day	<0.0001
	Group × Day	0.0166

). Basal ration intake was not affected by the Group × Day interaction (p = 0.159). Likewise, basal ration consumption was similar among treatment groups (p = 0.263). However, Day influenced consumption (p < 0.001), with basal ration intake increasing throughout the trial. A Group × Day interaction influenced average daily treatment intake of SO and RJ (p = 0.024; Figure 1, Table 6). Goats consumed more SO than RJ on Day 2 (p = 0.002) and Day 3 (p = 0.024). RJ intake varied across days, with lower consumption on Day 2 compared to all other days (p < 0.05), except Day 3 and Day 7, when intake did not differ (p > 0.05). In contrast, SO consumption remained consistent throughout Trial 1 (p > 0.05).

Post-vaccination wether body temperatures were not influenced by the treatment Group × Time interaction (p = 0.265). Additionally, when averaged across time points, body temperatures did not differ among treatment groups (RG = 37.99 °C, SO = 37.83 °C, control = 37.97 °C, p = 0.686). However, body temperatures were influenced by Time (p = 0.004), regardless of treatment group. Averaged across treatment groups, initial goat body temperatures were similar at 48 h but were greater by 72 h post-vaccination (0 h = 37.83 °C, 48 h = 37.78 °C, 72 h = 38.18 °C, p = 0.004). No adverse reactions were observed during post-vaccination visual assessments, indicating that the vaccination was well tolerated across all treatment groups. While body temperatures varied across the measurement intervals, these changes occurred consistently across all treatment groups, likely reflecting normal physiological responses to vaccination, environmental conditions, or handling. The Group × Day interaction did not affect total protein concentrations (g · dL⁻¹) (p = 0.140). However, total protein differed between treatment groups (p = 0.001). SO goats exhibited greater total protein than CG or RJ (SO = 6.44 g · dL⁻¹, RJ = 5.86 g · dL⁻¹, CG = 5.93 g · dL⁻¹, standard error = 0.11; Table 5). Total protein was similar for CG and RJ (p > 0.05). When averaged across treatment groups, Day alone did not affect total protein (p = 0.1436).

A Group × Day interaction influenced globulin concentrations (g·dL⁻¹) in Trial 1 (p < 0.017; Figure 2). Globulin differences were driven primarily by elevated globulin levels in the SO group compared to CG and RJ. Specifically, SO-fed animals had higher globulin concentrations on Days 0 and 7 (p < 0.05), but levels were comparable among groups by Days 14 and 21 (p > 0.05). Within the SO group, globulin levels declined over time, with higher concentrations on Days 0 and 7 followed by decreases on Days 14 and 21 post-vaccination. Globulin levels in animals consuming the RJ and CG diets remained consistent and similar throughout the trial. Bloodwork sampling error during EDTA whole-blood processing resulted in insufficient sample sizes for statistical analysis. As a result, statistical evaluation of complete blood count (CBC) data was not performed in Trial 1.

3.2. Trial 2

Weight parameters, including weight, weight × time, and overall average daily gain (ADG) were similar among treatment groups (ADG: p = 0.2985). Intake was measured and is expressed as % BW as fed basis, of the ration formulated at 93.55% DM. There was no interaction between treatment group and day on intake in Trial 2 (p = 0.904;

Table 7. Fixed effect test summaries for response variables of primary interest in Trial 2. Any response variables not included in this table do not have a significant (p > 0.05) group × day interaction. The column “Days with group differences” lists any specific days for which Tukey’s HSD revealed differences between treatment and control groups at the 0.05 level.

Response	Predictor	p-Value	Days with Group Differences
Intake	Group	0.0787	2, 7, 13
	Day	0.0067
	Group × Day	0.9040
Average Daily Gain	Group	0.2985	None
White Blood Cells	Group	0.7040	None
	Day	0.0684
	Group × Day	0.3127
Red Blood Cells	Group	0.4164	None
	Day	0.9228
	Group × Day	0.8495
Globulins	Group	0.5578
	Day	0.0148	14
	Group × Day	0.1874

). As observed in Trial 1, intake varied by day (p = 0.007), but not by treatment group (p = 0.079, Table 7). On Day 2, goats consumed more of the CG (2.99% BW ± 0.35 SE) and BP (2.99% BW ±0.35 SE) diets compared to GP (2.78% BW ±0.35 SE; CG vs. GP: p = 0.012; BP vs. GP: p = 0.015), while CG and BP intake did not differ (p = 0.998). On Days 7 and 13, goats consumed less GP than BP (p < 0.05), whereas CG intake was intermediate and did not differ from either GP or BP (p > 0.05) (Figure 3). ADG was similar between treatment groups (p = 0.299; Table 7).

Whole-blood counts for white blood cells (WBCs), red blood cells (RBCs), absolute neutrophils, and absolute lymphocytes were not affected by the Group × Day interaction (WBC: p = 0.313; RBC: p = 0.850; neutrophils: p = 0.350; lymphocytes: p = 0.784). Similarly, WBC, RBC, and lymphocyte counts did not differ by Group or Day alone (p > 0.05 for all comparisons; Table 7). In contrast, absolute neutrophil counts were influenced by Day (p = 0.010). Averaged across groups, neutrophil counts were lower on Day 14 (3.81 K·µL⁻¹ ± 0.49 SE) compared to Day 0 (5.70 K·µL⁻¹ ± 0.52 SE) and Day 7 (5.90 K·µL⁻¹ ± 0.55 SE). Neutrophil counts did not differ between Days 0 and 7 (p > 0.05). Unlike Trial 1, globulin and total protein concentrations in Trial 2 were unaffected by the Group × Day interaction (globulins: p = 0.187; total protein: p = 0.641) or by treatment group alone (globulins: p = 0.558; total protein: p = 0.373) (Table 7). Globulin levels differed by Day (p < 0.001), with lower concentrations observed on Day 14 (3.13 K·µL⁻¹ ±0.083 SE) compared to Day 0 (3.26 K·µL⁻¹ ±0.083 SE; p < 0.05) and Day 7 (3.29 K·µL⁻¹ ±0.083 SE; p < 0.05) (Figure 4). No difference was detected between Days 0 and 7 (p > 0.05). Total protein concentrations did not vary by Day (p = 0.270).

3.3. Trial 3

Statistical summaries of the most pertinent relationships tested are provided in

Table 8. Fixed effect test summaries for response variables of primary interest in Trial 3. Any response variables not included in this table do not have a significant (p > 0.05) group × day interaction. The column “Days with group differences” lists any specific days for which Tukey’s HSD revealed differences between treatment and control groups at the 0.05 level.

Response	Predictor	p-Value	Days with Group Differences
Intake	Group	0.8327	2, 17, 18
	Day	<0.0001
	Group × Day	<0.0001
Average Daily Gain	Group	0.8646	None
Body Temperature	Group	0.8276
	Day	<0.0001	None
	Group × Day	0.0295
White Blood Cells	Group	0.5525	None
	Day	0.0460
	Group × Day	0.6543
Red Blood Cells	Group	0.4047	None
	Day	<0.0001
	Group × Day	0.6832
IBR Immunoglobulins	Group	0.4264
	Day	0.0699	None
	Group × Day	0.9279
BVD Immunoglobulins	Group	0.5453
	Day	0.1258	None
	Group × Day	0.5547

. The result of primary interest involves intake data (Figure 5). Intake was measured and is expressed as % BW as fed basis, of the ration formulated at 93.26% DM. The mixed-effects model indicated that both day and the treatment by day interaction were significant (p ≤ 0.05). Instances of bloat were treated as missing data. On day one, intake (measured in percent body weight) between treatment and control groups were similar (treatment = 3.78%, control = 3.91%, mean difference = 0.117, standard error = 0.198, Tukey’s HSD p-value = 0.558). The following day (day 2 of Trial 3), treatment group intake was significantly depressed over control group intake (treatment = 2.98%, control = 3.61%, mean difference = 0.632, standard error = 0.198, Tukey’s HSD p-value = 0.002), presumably due to an initial acclimation to the novel feed additive. After the apparent initial acclimation period, intake was similar between treatment and control groups, until inoculation on day 15, after which intake began to diverge. This trend became significant on day 17 (treatment = 3.63%, control = 3.12%, mean difference = 0.518, standard error = 0.202, Tukey’s HSD p-value = 0.012) and day 18 (treatment = 3.60%, control = 3.06%, mean difference = 0.550, standard error = 0.202, Tukey’s HSD p-value = 0.008). After this point, control group intake began to recover and was not significantly different from treatment intake.

The other response variable that was associated with statistical significance is body temperature. While treatment effects alone were not significant, day effect and treatment–day interaction were significant (p ≤ 0.05). Baseline body temperatures taken immediately prior to inoculation exhibited the greatest difference between treatment and control groups (treatment = 39.41 °C, control = 39.14 °C, mean difference = −0.294, standard error = 0.163, p-value = 0.079). Body temperatures 24 h post-inoculation were both lower than on day 0, and very similar (treatment = 39.00 °C, control = 39.01 °C, mean difference = 0.005, p-value = 0.974). Temperatures taken 7 days post-inoculation had once again risen and diverged in opposite directions compared to initial temperatures (treatment = 39.42 °C, control = 39.68 °C, mean difference = 0.206, p-value = 0.214). However, on none of these days, were treatment and control groups significantly different according to Tukey’s HSD at an alpha of 0.05. Average daily gain was similar (p > 0.05) across treatment groups, days, and the treatment–day interaction. On no day was there a significant difference between groups according to Tukey’s HSD at an alpha of 0.05.

Blood parameter data (Figure 6), while the intended focus of the study, failed to show any significant treatment or Treatment × Day effects (p > 0.05). Total white blood cell counts post-inoculation were similar between treatment and control groups (treatment = 8.27 K·µL⁻¹, control = 8.05 K·µL⁻¹, mean difference = −0.118, standard error = 0.442, p-value = 0.791). Constituent cell counts post-inoculation including neutrophils, lymphocytes, and monocytes were similar. Total red blood cell counts post-inoculation were also similar between treatment and control groups (treatment = 8.08 M·µL⁻¹, control = 7.79 M/µL, mean difference = −0.396, standard error = 0.418, p-value = 0.355). IBR immunoglobulin titers were similar among groups post-inoculation (treatment = 2.51 g·dL⁻¹, control = 2.41 g·dL⁻¹, mean difference = −0.164, standard error = 0.457, p-value = 0.722). BVD immunoglobulin titers were also similar post-inoculation (treatment = 2.49 g·dL⁻¹, control = 2.44 g·dL⁻¹, mean difference = −0.350, standard error = 0.451, p-value = 0.445). All other measured blood serum constituents including albumin, total protein, and albumin–globulin ratio were similar between treatment groups post-inoculation and did not have a significant group–day interaction.

4. Discussion

4.1. Trial 1

One week before the onset of the first experimental feeding trial, some animals began eliciting clinical symptoms indicative of coccidiosis, including loss of appetite, lethargy, and diarrhea. Immediately, all animals were subject to a five-day antiparasitic treatment regime to avoid adverse health outcomes. While this occurrence is an unfortunate confounding factor to this trial’s experimental methodology, it also represents a unique “natural experiment” from which useful information can still be gleaned. This illness was a more extreme immune challenge than can typically be safely and controllably administered in an experimental setting (e.g., Trials 2 and 3). Importantly, the same number of animals in both the shin oak group and control groups were symptomatic for coccidiosis. While our depth of inference from Trial 1 results is reduced, these results nonetheless serve as a relevant case study that inspired the research conducted in Trials 2 and 3 and hopefully experiments to come.

In Trial 1, goats offered shin oak had significantly higher blood globulin concentrations than goats offered juniper or goats in the control group even prior to the Enterotoxemia Type C and D vaccination. Thus, the increase in globulins for the shin oak group in Trial 1 likely represents immunoglobulins produced in response to the coccidia infection rather than in response to the vaccination. Tannins have been found to produce similar effects to intestinal parasite infections in ruminants [78,79,80]. By day 14 of post-feeding trial sampling, globulin levels in all three groups experienced a decline, with the steepest decline being among the shin oak group. This comparatively larger decline in blood globulins may have been due to the cessation of access to potentially medicinal tannin-rich plants.

The inclusion of tannin- and terpene-rich plant materials did not adversely impact animal performance. ADG was similar between treatment groups in Trial 1. In Trial 1, all treatments consumed similar amounts of the basal ration. However, the intake of shin oak by goats increased more quickly than the juniper intake during the acclimation period. Although not statistically different, an enhanced rate of shin oak intake was maintained throughout the project (Figure 1). Redberry juniper’s chemical profile was likely more aversive to goats than that of the shin oak leaves [3,72,81,82]. Notably, though coccidiosis caused other affected goats to avoid the basal rations, animals in the shin oak group still consumed leaf clippings, while animals in the redberry juniper treatment group did not. This observation may allude to the medicinal properties of the tannins within shinnery oak, given that shin oak is significantly less palatable than the basal ration [3,30]. Similar instances of self-medication with tannin-rich plants as a response to parasitic infections are well documented [43,83,84].

Initially, we intended to perform fecal egg counts on goats in Trial 1 to compare the anthelmintic properties of forage supplements. For this reason, goats were not immediately dewormed upon individual pen placement. Unfortunately, goat illness required immediate attention and subsequent fecal egg counts could not be performed. Although fecal egg counts would have enriched the scope of this experiment, many previous experiments indicate that increased condensed tannins (CTs) in the diet leads to a reduction in internal parasites [43,46,85,86].

4.2. Trial 2

The compelling but partially confounded results from Trial 1 prompted a follow-up experiment (Trial 2), aimed at replicating the increase in blood globulins associated with the shin oak group in Trial 1, with a more controlled immune challenge. Additionally, we pivoted the underlying research goals towards implications for a confined animal feeding context, hence the use of grape and blueberry pomace within feed rations as opposed to free choice leaf clippings. However, blood parameters evaluated in Trial 2 were not significantly different between treatment groups following the immune challenge. Although neutrophils and globulins decreased throughout the trial duration, control and treatment groups responded similarly to the vaccination health challenge.

The inclusion of condensed tannin-rich plant materials did not adversely impact animal performance. Although differences occurred between treatments on a few days, the overall ADG and intake were similar between treatment groups in Trial 2. While intake in Trials 1, 2 and 3 are not directly comparable due to differences in tannin source for Trial 1 versus 2 and species for Trial 2 versus 3, it is possible that goats in Trial 2 were not sufficiently health-challenged to consume greater amounts of tannin-containing feeds by the Enterotoxemia Type C and D vaccination (relative to more extreme health challenges in Trials 1 and 3). Additionally, tannin inclusion method is a confounding factor. In Trials 2 and 3, the tannin-containing plant materials were thoroughly mixed with other feed ingredients in the ration. Thus, goats in Trial 2 would have had to consume greater than typical amounts of total feed to increase tannin intake post-vaccination.

4.3. Trial 3

Trial 3 was conducted as a follow-up to Trial 2 to evaluate if the effects from Trial 1 could be replicated in beef cattle and more precisely measured (we could measure immunoglobulins specific to the vaccinations administered rather than simply total globulins). However, despite intake data in Trial 3 appearing consistent with our hypothesis, no significant differences in blood parameters (Figure 6; Table 8) were observed. We will explore potential explanations for this discrepancy below. Notably, the inclusion of condensed tannins (CTs) did not adversely impact animal performance, as ADG and total weight gain were similar between treatment groups in Trial 3. Still, intake data provides some evidence that CTs may play a role in immune response. By day 2, treatment group experienced significantly (according to Tukey’s HSD at an alpha of 0.05) depressed feed intake (Figure 6; Table 8) for a short period at the beginning of the trial. However, this effect is likely an acclimation period to the novel feed additives [87] and the aversive nature of CTs [88], rather than the result of a health-related interaction. From day 3 through day 15 (time of inoculation), intake was nearly identical. After inoculation, treatment and control group intake began to diverge. By day 17 and 18, control group intake became significantly (according to Tukey’s HSD at alpha of 0.05) depressed, after which, intake began to trend back up, becoming similar with treatment group intake (Figure 5; Table 8). The modified live vaccine administered in Trial 3 was regarded by a consulting veterinarian as being a relatively substantial health challenge. While the response by the control group is typical in cattle facing an immune challenge in the form of vaccination [89], treatment group intake remained significantly higher (according to Tukey’s HSD at alpha of 0.05) than control group intake and saw only a minor, non-significant (according to Tukey’s HSD at alpha of 0.05) dip during the same period. This patten is consistent with the hypothesis that certain qualities of the CTs in the treatment group feed were relevant to immune response or immune challenge recovery.

It is unclear why the apparent role condensed tannins play in immune challenge or recovery, evidenced by intake differences, was not also reflected in blood data. While intake differences were paired with blood chemistry differences in Trial 1, a multitude of variables could underlie this discrepancy. Among these critical differences may have been the source of tannins provided. Research suggests that grape and blueberry pomace are rich sources of CTs [33,47,50]. However, when samples of our pomace were analyzed by the Texas A&M AgriLife Research and Extension Center in Stephenville, Texas, actual CT concentrations were 4.77% for blueberry pomace and 4.91% for grape pomace. Of these totals, only 1.42% and 0.81% were extractable, with the remainder being bound to proteins or fibers (Table 1b), and thereby, not bioavailable. In Trials 2 and 3 the total CT content of the treatment group feed rations was approximately 1.0% and 0.3%, respectively, and only about a quarter of these were bioavailable. Research indicates that the dietary CT concentration for medicinal effects is at least 1–2% [37,90], and recent studies involving the inclusion of grape pomace into livestock diets utilize a wide range of inclusion rates, ranging from 2% to 15% [47,50,53]. Thus, greater feed intake by the treatment group post-vaccination may indicate that treatment animals attempted to consume enough bioactive CTs to enhance their immune response but were unable to consume the critical amount necessary to produce results. Ultimately, the low dose of bioactive tannins available to animals in Trials 2 and 3 is the most probable cause of their negative results. Future studies should utilize richer sources of tannins or incorporate tannin-containing ingredients at a greater rate to avoid underdosing animals.

A significant day–group interaction was also observed with regard to body temperature. The largest between-group temperature difference was pre-inoculation, during baseline measurement, with treatment group temperatures being higher, but not significantly (according to Tukey’s HSD at p < 0.05). The next measurement, taken at 24 h post-inoculation was intended to capture elevated body temperatures near their peak, as a result of immune response to the recent vaccination. The body temperatures measured at that time for both groups were not only lower than baseline temperatures but nearly identical to each other. Temperatures were collected again at 7 days post-inoculation to re-establish a baseline. Body temperatures measured at this time had once again diverged, but in opposite directions from our day 0 temperatures, and were on average higher than day 0 temperatures. Research on beef cattle indicates that body temperatures can be expected to increase immediately following an immune challenge such as vaccination and then recover near baseline by day 7 [91,92]. Temperatures were measured at the same time of day and ambient temperatures did not differ by more than 6 °C. Because temperatures only differed significantly by treatment group during times intended as baseline measurements, and because none of these specific differences were significant when analyzed with Tukey’s HSD (at alpha of 0.05), these effects appear artifactual, or otherwise unrelated to the relationship between CTs and immune response.

4.4. Comparative Analysis of Trials

The inclusion method of tannins is likely a contributing factor to the inconsistent results of Trials 1, 2, and 3. In Trial 1, tannin-rich shinnery oak (Quercus havardii Rydb.) was offered separately from the basal ration, allowing treatment group animals to self-select intake. In Trials 2 and 3, CT containing grape and blueberry pomace were mixed in with the basal ration, meaning the concentration of CTs was fixed within the diet. Notably, treatment heifers in Trial 3 were able to consume significantly more (according to Tukey’s HSD at p < 0.05) feed than control heifers post-vaccination without consuming more than baseline, because intake in control heifers was substantially depressed post-vaccination. This may not have been reflected in Trial 2 because the immune challenge was not sufficient. For future experiments, a hybrid method including a small amount of CTs in animal feed rations, in addition to ad libitum access CT-rich materials, would be optimal. This approach would ensure animals have a small amount of CTs in their system prior to the immune challenge, when they might not voluntarily consume them, while also allowing animals to increase CT intake, independent of their basal ration, if they desire, following the immune challenge.

Tannin type is likely another contributing factor to these discrepancies. Tannins, as well as CTs specifically, have highly variable chemical structures, polymer size, and intermolecular linkages and, ultimately, effects in the ruminant body [7,93]. For example, various types of CTs from various sources produce diverse effects, including reducing internal parasite loads, eliciting aversive post-ingestive feedback, and preventing bloat in cattle [32]. Anecdotally, in this study, instances of bloat occurred seven times in the treatment group and only once in the control group. If the same class of condensed tannins that reduce bloat are active immune stimulants, then this dynamic would explain why neither effect was observed in Trial 3. While not enough research exists to identify if this is the case, the total CT content (0.32% DM) is on the far low end of the effective range for bloat mitigation [94]. Further research regarding the functional and structural differences between tannins derived from different sources, especially grape/blueberry pomace and shin oak, are warranted.

Whether the tannin concentration, inclusion method, chemical structure, species differences, or another difference in experimental design between Trials 1, 2, and 3 contributed to the inconsistent results is unclear. Better understanding of the chemical mechanisms through which tannins affect ruminant immune response would be potentially illuminating. A commonly suggested mechanism is the ability of tannins to increase rumen-bypass proteins [14,32,93]. This is a potential contributor to the results of Trial 1, as rumen-bypass proteins have been shown to improve immune response in beef cattle [95], and resistance to the parasite Haemonchus contortus in goats [96]. However, many of the Trial 1 goats, including those in the shin oak group with higher blood globulins, largely refused their basal ration during the time of coccidiosis, so a large influx of rumen-bypass proteins would have been unlikely.

In vitro data suggests that CTs can sometimes directly modulate immune cell function [97,98,99]. Molecule size, or mean degree of polymerization (mDP), appears to determine whether this effect is significant [93,100]. Research on CT mDP is limited, and we do not have mDP values for our specific CT-containing materials, but if CT size is another potentially relevant source of variation amongst our trials. CTs may also influence immune response indirectly via epigenetic regulation of immune genes [101]. This could occur through a direct signal cascade or via tannin-induced increase in rumen bypass methionine (typically the first limiting amino acid from microbial protein), a methyl group donor and catalyst for epigenetic changes [102]. More research regarding these potential mechanisms for a CT-caused increase in immunoglobulin production, leukopoiesis, or general immune function and recovery is needed.

5. Conclusions

We conducted three controlled feeding trials to evaluate the relationship between immune response in ruminants and the consumption of phytochemically rich feeds, especially with condensed tannins. We evaluated this relationship in a free-ranging context and in a confined feeding context, with both goats and beef cattle. Following an immune challenge, goats offered shin oak voluntarily consumed the tannin-rich forage, and had higher levels of blood globulins, compared with control animals. Beef cattle with feed containing tannin-rich grape/blueberry pomace also had less depressed feed intake than control animals following an immune challenge, but this did not translate to significant blood chemistry differences. However, these findings were not replicated when tested differently in other contexts. Inconsistencies among the three trials may be attributed to variation in the sources and concentrations of condensed tannins, variation in how those tannins were offered to the animals, or variation in the severity of the health challenges to the animals. While this study provides limited evidence that tannin-rich plants in the diets of domestic goats might improve humoral immune response, methodological differences between trials limit our scope and depth of inference. Ultimately, further research is required to better understand this relationship and how it might extend to other ruminants and livestock species and other plants with different phytochemical profiles. Future studies should focus on elucidating the potential mechanisms by which specific phytochemicals can benefit animal immunity and general health.