Functional Feed Additives in Intensive Sheep Production: From Mechanism to Practical Application

Simple Summary

This review examines four groups of additives used in intensive sheep production: gluconeogenic precursors, which supply the liver with raw material to make glucose; β-adrenergic agonists (β-AAs), which redirect nutrients from fat into muscle; exogenous enzymes, which help rumen microbes break down plant fiber; and phytochemicals, plant compounds that modify rumen fermentation and protect meat from oxidation. For each group, we summarize how it works; what it does to growth, carcass composition, and meat quality; and where the evidence is weak, breed-specific, or contradictory. We also weigh each group against its likely cost, regulatory status, and environmental impact, since β-AAs give the largest gains in lean meat but face substantial restrictions in many importing countries, while phytochemicals give smaller gains but also lower methane emissions and generally face fewer regulatory restrictions. We conclude that no single additive class is universally superior: the right choice depends on diet, market, and sustainability goals, and combining classes with complementary mechanisms may offer additional benefits, although this strategy remains insufficiently tested.

Abstract

In ruminant nutrition, functional feed additives are defined as non-nutritive or bioactive substances added to diets to intentionally alter rumen microbial populations, enhance nutrient digestibility, or modulate metabolic pathways beyond standard baseline nutrition. This review evaluates four additive classes used to address the feed-efficiency gap in sheep: gluconeogenic precursors, β-adrenergic agonists (β-AAs), exogenous enzymes, and phytochemicals. We place specific emphasis on their comparative productive performance, carcass and meat quality, economic feasibility, environmental impact, and regulatory status. Among the compounds reviewed, calcium propionate and zilpaterol hydrochloride yield the largest increases in growth rate and lean yield, though these effects remain breed-, diet-, and dose-dependent. Exogenous enzymes and phytochemicals offer smaller performance gains but generally face fewer regulatory restrictions; notably, phytochemicals reduce enteric methane emissions by 10–25% and improve meat oxidative stability, aligning well with modern sustainability goals. While combination strategies across additive classes show mechanistic promise, they remain empirically under-tested. To guide practical adoption, this review provides a cross-additive synthesis alongside qualitative cost–return profiles. Addressing critical research gaps—specifically multi-additive interactions, breed-specific validation, and the development of on-farm tools from precision-nutrition biomarkers—will determine the long-term viability of these strategies under growing market and environmental pressures.

1. Introduction

Sheep (Ovis aries) are a cornerstone of global livestock production, supplying meat, wool, and milk to populations across six continents. Global sheep inventory exceeds one billion head, with total production of approximately 10 million tons of sheep meat per year, a figure projected to rise by 30–40% by 2050 in response to demographic growth, urbanization, and shifting dietary preferences in emerging economies [1,2]. Intensive lamb finishing systems—characterized by high stocking densities, controlled feeding with energy-dense complete diets, and slaughter at 90–180 d of age—have emerged as the dominant commercial model in many producing regions because they concentrate production over short feeding periods and generate consistent carcass uniformity that satisfies industrial processors and retailers [2,3].

Despite genetic progress in growth rate and feed efficiency through selection programs, intensive lamb production faces persistent challenges at the intersection of productivity, economics, and sustainability. Feed costs represent 65–75% of variable production costs in confined systems [4], and the gain-to-feed ratio (G:F) of finishing lambs—typically 0.15–0.20, occasionally reaching up to 0.22 kg live weight gain per kg dry matter intake (DMI) [5,6]—remains markedly lower than that of broilers (0.56–0.70, depending on market age) [7] and pigs (0.33–0.40) [8,9]; this comparison is presented for illustrative purposes only, since ruminants, poultry, and pigs differ fundamentally in digestive physiology, maintenance energy requirements, and growth trajectory, and G:F values are therefore not directly interchangeable across production systems, reflecting the inherent energetic losses of rumen fermentation, primarily methane (CH₄) and heat of fermentation, and the lower efficiency of gluconeogenesis compared to direct glucose absorption [3,10]. Carcass fat deposition at a young slaughter age poses a quality penalty in markets that prize lean yield, while the management of meat tenderness, color stability, and oxidative shelf life presents ongoing technical challenges [10,11].

Within this context, functional feed additives have gained significant scientific and commercial attention as tools capable of modulating specific metabolic pathways to overcome these limitations. Four major categories dominate the current evidence base in sheep nutrition. Gluconeogenic precursors—including propylene glycol (PG), calcium propionate (CaPr), glycerol, and sodium propionate (NaPr)—bypass the rumen fermentation bottleneck by delivering substrates directly available for hepatic phosphoenolpyruvate carboxykinase (PEPCK)-mediated glucose synthesis, thereby increasing portal glucose supply, stimulating insulin secretion, and creating the systemic hormonal milieu that facilitates muscle protein synthesis via the mechanistic target of rapamycin (mTOR) pathway [12,13]. The β-adrenergic agonist (β-AA) class, exemplified by ractopamine hydrochloride (RH) and zilpaterol hydrochloride (ZH), redirect nutrient partitioning from adipose anabolism toward skeletal muscle hypertrophy through β2-adrenergic receptors (β2-ARs), simultaneously activating the cyclic adenosine monophosphate (cAMP)–protein kinase A (PKA)–mTOR anabolic cascade [14]. Exogenous enzyme complexes—combining cellulases, xylanases, β-glucanases, amylases, and proteases—attack the structural and chemical barriers that limit microbial access to dietary nutrients, increasing volatile fatty acid (VFA) production and modifying the rumen microbiome toward more productive fermentation profiles [15,16]. Phytochemicals, encompassing essential oils (EOs), tannins, saponins, flavonoids, phenolic acids, terpenoids, organosulfur compounds, and alkaloids, are emerging as multifunctional alternatives in production systems where synthetic growth promotants face regulatory restrictions, with mechanisms spanning antioxidant enzyme induction, pathogen-specific antimicrobial activity, methanogen inhibition, immunomodulation, and direct effects on meat oxidative stability and fatty acid composition [17,18].

The present review addresses four specific objectives: (1) to establish the molecular metabolic framework governing energy partitioning, protein synthesis, and lipid deposition in growing lambs as the mechanistic context for additive action; (2) to critically review the biological mechanisms, dose–response relationships, productive outcomes, and limitations of each of the four major additive classes in intensive sheep production; (3) to provide a quantitative cross-additive comparative analysis—understood here as a structured narrative synthesis of ranges and directional effects reported across independent primary studies, not a formal meta-analysis with pooled effect sizes, given the heterogeneity of breeds, diets, and outcome metrics across the retrieved literature—across production, carcass, meat quality, sustainability, and economic dimensions; and (4) to identify emerging research frontiers in precision sheep nutrition incorporating omics sciences, microbiome engineering, and computational feed formulation.

The relevant literature was identified through structured searches of PubMed, Web of Science, and Scopus (2000–2026) using combinations of “sheep”/“lamb”/“ovine” AND “feed additive” AND each additive class or compound name (e.g., “calcium propionate”, “zilpaterol”, “fibrolytic enzyme”, “essential oil”, “tannin”). Peer-reviewed original research and reviews reporting productive, carcass, meat-quality, or rumen-fermentation outcomes in sheep were prioritized; evidence from other ruminant species was retained only when no sheep-specific data existed and flagged as such throughout the text. Conference abstracts without full data and non-English/Spanish sources were excluded.

2. Energy and Muscle Metabolism in Growing Lambs

2.1. The Propionate–Glucose–Insulin Axis

The metabolic architecture of the growing lamb reflects its evolutionary specialization for deriving most of its energy from microbial fermentation of structural carbohydrates that mammalian enzymes cannot digest. This specialization defines the pharmacological landscape in which every additive reviewed in this manuscript operates: propionate supply, hepatic glucose output, and the hormonal signals that couple energy status to muscle growth are the shared targets of gluconeogenic additives, β-AAs, exogenous enzymes, and phytochemicals alike. Rather than treat each signaling pathway in isolation, this section summarizes the functional logic of ruminant energy metabolism at the level needed to interpret additive performance in later sections and flags where the evidence base is thinner than commonly assumed. An integrated schematic is provided in Figure 1, and

Table 1. Key signaling molecules governing energy metabolism and muscle protein synthesis in growing sheep, with their regulatory interactions and relevance to functional feed additive pharmacology.

Molecule/Pathway	Class	Primary Activators	Downstream Targets	Net Effect on Growth	Modulated by	Reference
mTORC1	Ser/Thr kinase complex	IGF-1, insulin, leucine, ATP adequacy	Ribosomal protein synthesis; fiber hypertrophy	↑ Protein synthesis; ↑ muscle hypertrophy; ↑ ADG	↑ by gluconeogenic additives, β-agonists, leucine	[19,20]
AMPK	Ser/Thr kinase	↑ AMP:ATP, energy deficit, heat stress	Suppresses mTORC1; activates β-oxidation and autophagy	↓ Protein synthesis under energy deficit; ↑ β-oxidation	↓ by adequate energy and propionate supply	[21,22]
PEPCK	Gluconeogenic enzyme (rate-limiting)	Glucagon, glucocorticoids, fasting	Oxaloacetate → glucose	Hepatic glucose supply; drives insulinotropic cascade	↑ by propionate, glycerol; ↓ by insulin	[23,24]
IGF-1/satellite cells	Growth factor axis	GH, nutrition, mechanical load	PI3K/Akt/mTOR activation; satellite cell activation	↑ Myoblast proliferation; ↑ fiber hypertrophy; ↓ myostatin-mediated inhibition	↑ by β-agonists and some phytochemicals	[25,26,27]
Propionate (VFA)	Rumen metabolite	Microbial fermentation	PEPCK → glucose → insulin → mTOR	↑ Insulinemia; ↑ ADG; ↑ FCR	↑ by gluconeogenic additives, fibrolytic enzymes	[12,13,28]

↑, increase; ↓, decrease; →, leads to/is converted to.

summarizes the key signaling molecules governing this axis, together with their regulatory interactions and relevance to additive pharmacology.

2.2. Rumen Fermentation and Volatile Fatty Acid Supply

The rumen-reticulum functions as a continuous anaerobic fermenter in which bacteria, protozoa, fungi, and methanogenic archaea convert dietary carbohydrate into VFAs, which supply approximately 78% of maintenance energy in sheep [28]. Under high-roughage diets, acetate:propionate:butyrate molar ratios are typically near 65:20:15; concentrate-based finishing diets shift this ratio markedly toward propionate [28,29]. This shift matters because propionate is the principal gluconeogenic substrate in ruminants, supplying 50–70% of hepatic glucose production [12,13], a dependency with no direct equivalent in monogastric species, and one that explains why strategies to increase propionate supply (Section 3) or its downstream utilization (Section 4) are as central to sheep nutrition as they are irrelevant to pigs or poultry.

The dominant cellulolytic taxa include Fibrobacter succinogenes, Ruminococcus albus, and R. flavefaciens, which drive acetate production from structural carbohydrates, whereas propionate is generated mainly via the succinate pathway by Prevotella spp., Selenomonas ruminantium, and Megasphaera elsdenii, and via the acrylate pathway to a lesser extent; butyrate arises principally from Butyrivibrio fibrisolvens and related butyrate-producing Firmicutes. Fibrolytic enzymes and phytochemicals reviewed in Section 5 and Section 6 act directly on these populations—enzymes by increasing substrate accessibility for cellulolytic bacteria and phytochemicals (tannins, saponins, essential oils) by selectively suppressing protozoa, methanogens, and specific biohydrogenating or fibrolytic taxa—providing the mechanistic link between microbial community shifts and the VFA profile changes reported throughout this review [28,29,30].

Multi-omics comparisons of sheep with divergent G:F support this link directly: more efficient animals show higher ruminal propionate concentrations, enrichment of propionate-producing bacterial taxa, and greater expression of hepatic gluconeogenic genes than their less efficient counterparts [30], indicating that variation in propionate handling—not simply total energy intake—contributes to G:F differences within a flock.

2.3. Hepatic Gluconeogenesis

Because intestinal glucose absorption is negligible in mature ruminants, hepatic gluconeogenesis supplies all systemic glucose [12,13]. Propionate absorbed from the rumen is converted through propionyl-CoA and methylmalonyl-CoA to succinyl-CoA, enters the tricarboxylic acid cycle, and is ultimately converted to glucose via PEPCK, the rate-limiting enzyme of the pathway [23]. PEPCK activity responds to the hormonal balance between glucagon and glucocorticoids (stimulatory) and insulin (suppressive) [23,24]. This single control point explains a recurring pattern in this review: any additive that increases propionate supply to the liver—whether through rumen fermentation (enzymes, phytochemicals) or direct absorption (PG, CaPr, glycerol)—acts on the same rate-limiting step, which is why their downstream effects on insulin and muscle growth are qualitatively similar despite differing chemistries.

2.4. Nutrient Sensing and the Growth–Energy Balance

Skeletal muscle protein synthesis in growing lambs is governed by the balance between two opposing signaling systems. The mechanistic target of rapamycin complex 1 (mTORC1) integrates nutrient availability, particularly amino acids such as leucine, growth factor signaling through insulin and insulin-like growth factor 1 (IGF-1) via the phosphoinositide 3-kinase/protein kinase B (PI3K/Akt) pathway, and cellular energy status to promote ribosomal protein synthesis and muscle fiber hypertrophy [19,20]. AMP-activated protein kinase (AMPK) acts as the counterbalancing energy sensor: when cellular energy is depleted—during undernutrition, heat stress, or subclinical disease—AMPK activity rises and actively suppresses mTORC1, redirecting the animal’s metabolism toward energy conservation rather than growth [21,22].

This balance, rather than any single step within either pathway, is the functional variable of interest for this review. It explains why interventions that improve metabolic efficiency (adequate energy and propionate supply, low physiological stress) establish the required conditions under which anabolic additives are effective, and why the same additive can appear efficacious in one trial and ineffective in another depending on the animal’s baseline energy status. Most of the molecular detail underlying this mTOR–AMPK relationship, including the specific intermediate kinases and their phosphorylation sites, has been characterized in rodent and cell-culture systems [19,20,21,22]; direct confirmation in ovine muscle tissue remains comparatively limited, a point we return to in Section 2.8. Where mechanistic statements in this and subsequent sections are not explicitly qualified, they should be understood as extrapolated from non-ovine models unless a sheep-specific citation is given.

2.5. Lipid Metabolism and Its Relevance to Meat Quality

Ruminant lipid metabolism differs from that of monogastric species in a way directly relevant to meat quality: adipose tissue, not the liver, is the principal site of de novo fatty acid synthesis in sheep, and acetate—not glucose—is its main lipogenic precursor [31,32]. Intramuscular fat (IMF) deposition, a key determinant of meat palatability, depends on the differentiation and lipid-loading of intramuscular preadipocytes under the influence of insulin and related transcriptional regulators [32,33]. This distinction matters practically because dietary strategies that increase propionate and glucose supply (Section 3) can be expected to favor lean muscle accretion over fat deposition to a greater extent in sheep than equivalent interventions would in pigs, where hepatic lipogenesis from glucose predominates.

2.6. Satellite Cells and Postnatal Muscle Growth

Postnatal muscle growth in lambs occurs almost entirely through hypertrophy of existing fibers, which requires new myonuclei donated by satellite cells—quiescent muscle progenitor cells that activate, proliferate, and fuse with existing fibers in response to anabolic signals such as IGF-1 [25,26]. Myostatin is the principal endogenous inhibitor of this process and represents the main biological brake on muscle hypertrophy in the growing lamb [34]. Because satellite cell numbers and activity, rather than muscle fiber number, set the ceiling for postnatal muscle growth, interventions that enhance IGF-1 signaling or suppress myostatin activity—as several of the additives reviewed here appear to do—have a plausible route to increasing lean growth independent of any direct effect on DMI.

2.7. Hormonal Regulation

The growth hormone–IGF-1 axis is the principal endocrine driver of postnatal growth in sheep, with IGF-1 activating the same PI3K/Akt/mTOR pathway described above while simultaneously suppressing muscle protein breakdown [27]. Insulin, secreted in response to propionate-derived hepatic glucose, reinforces this anabolic state. Glucocorticoids act in the opposite direction, promoting muscle protein breakdown and hepatic PEPCK activity, the physiological basis for the growth-retarding effects of [23,27] chronic stress, and a reminder that additive efficacy cannot be considered independently of the animal’s overall stress and management context.

2.8. Critical Synthesis and Implications for Additive Pharmacology

Four points emerge from this section that recur throughout this review. First, propionate supply to the liver is the single most consistent lever by which the additives reviewed here influence growth, whether that supply is increased through fermentation (enzymes, phytochemicals) or bypassed through direct absorption (gluconeogenic additives). Second, the mTOR–AMPK balance—not the fine molecular detail of either pathway—determines whether a given additive will translate into measurable growth under a specific animal’s energy status, a plausible explanation for much of the between-study variability documented in Section 3, Section 4, Section 5 and Section 6. Third, satellite cell availability and myostatin activity set a biological ceiling on muscle hypertrophy that is largely independent of energy supply and the more direct target of β-AAs (Section 4). Fourth, and important for interpreting the strength of the evidence in this field, much of the mechanistic detail underlying these pathways—particularly the intracellular signaling intermediates—has been characterized in rodents, cattle, or cell culture rather than in sheep directly; where sheep-specific confirmation exists, we note it, but the reader should be cautious about assuming full mechanistic equivalence across species.

• Key Messages

• Hepatic gluconeogenesis from propionate, not intestinal glucose absorption, is the metabolic bottleneck that essentially all additive classes reviewed in this manuscript act upon, directly or indirectly.
• The mTOR–AMPK energy balance, rather than the identity of any single signaling intermediate, is the functional variable that determines whether an animal responds anabolically to a given additive.
• Satellite cell availability and myostatin activity—not substrate supply—set the ceiling on postnatal muscle hypertrophy and are the more direct targets of β-AAs (Section 4).
• Much of the molecular detail underlying mTOR–AMPK signaling has been established in non-ruminant models; sheep-specific validation remains comparatively limited.

• Current Knowledge Gaps

• Direct confirmation of mTOR–AMPK signaling responses to feed additives in ovine skeletal muscle, as opposed to extrapolation from cattle or rodent models, is largely absent.
• The relationship between rumen microbial propionate production and hepatic gluconeogenic gene expression has been characterized in only a small number of G:F studies and has not been evaluated across breeds.
• No studies have directly linked satellite cell activity or myostatin expression to additive-induced growth responses in sheep.

3. Gluconeogenic Feed Additives

3.1. Metabolic Rationale

Gluconeogenic feed additives address a specific metabolic vulnerability of intensively managed finishing lambs: the propionate–glucose–insulin anabolic axis introduced in Section 2 can become limiting when dietary energy density or rumen fermentation efficiency constrains hepatic glucose output. By supplying exogenous substrates that either enter hepatic gluconeogenesis directly or are fermented to propionate in the rumen, these additives augment glucose supply and stimulate the insulinotropic cascade independently of dietary starch content [35,36]. Four compounds dominate the commercial and research literature: PG, CaPr, glycerol, and NaPr. Figure 2 illustrates the convergent gluconeogenic pathways, and

Table 2. Major gluconeogenic feed additives in intensive sheep production: chemistry, ruminal fate, metabolic route, productive response, carcass and meat quality effects, and principal limitations.

Additive	Chemical Form	Rumen Fate	Gluconeogenic Route	ADG Effect	Carcass	Meat Quality	Key Limitation	Reference
Propylene glycol	1,2-Propanediol (liquid)	~50% absorbed intact; ~50% fermented → propionate	PG → L-lactate/pyruvate → OAA → PEPCK → glucose	0 to +10% (variable)	Modest or no effect	Neutral	Route-dependent; palatability; dose not optimized for sheep	[37,38]
Calcium propionate	Ca(C2H5COO)2 (powder)	Direct propionate release upon rumen dissociation	Propionate → PEPCK → glucose	+20–27% (optimal dose)	↑ HCW, D%, LMA	Neutral at optimal dose	Excess dose → lipogenesis; forage-only diet: no effect	[39,40,41]
Glycerol/Crude glycerin	1,2,3-Propanetriol (liquid)	~80% fermented → propionate + butyrate	Propionate → gluconeogenesis; glycerol → DHAP → glucose	+4–12% (optimal dose)	↑ lean; ↓ perirenal fat	↑ CLA at optimal dose	Purity-dependent; methanol impurities reduce DMI	[42,43,44,45]
Sodium propionate	CH3CH2COONa (powder)	Immediate propionate release	Same as CaPr	Similar to CaPr	Similar to CaPr	No adverse effects	High Na+ load at high doses; limited ovine dose–response data	[46]
Ca salts of LCFAs	Ca-FA complexes (rumen-inert)	Stable at rumen pH; dissociate at abomasal pH	Indirect: NEFA oxidation spares gluconeogenic demand	Energy supplement	Not established in sheep (data limited to dairy cattle)	↑ PUFA (C18:1, C18:2)	Limited finishing lamb data; primarily for late-gestation ewes	[47]

↑, increase; ↓, decrease; →, leads to/is converted to.

compares these four compounds by chemistry, ruminal fate, metabolic route, productive response, and meat quality effects.

3.2. Propylene Glycol

PG undergoes three concurrent fates in the rumen: direct absorption (~50% within 1–2 h), microbial fermentation to propionate, and passage to the lower digestive tract [37,38]. Its primary hepatic fate is oxidation to L-lactate rather than direct propionate conversion, and oral dosing produces plasma insulin increases of 200–400% within 30 min [37]. In Akkaraman lambs, PG significantly increased blood glucose at 60 d, but this response was not sustained through d 90 and 120, suggesting metabolic adaptation with continued use [35]. In practice, PG appears best suited to short-term metabolic support—around parturition, transport, or other acute stress—rather than as a sustained growth promoter, an important nuance often overlooked in the literature.

3.3. Calcium Propionate

Calcium propionate dissociates in the rumen to release propionate anions directly available for hepatic gluconeogenesis. Beyond its metabolic effect, CaPr modulates hypothalamic appetite-regulating neuropeptide expression in a manner consistent with CaPr-induced hyperinsulinemia feeding back on arcuate nucleus insulin-sensing neurons [39].

The productive response to CaPr follows a reproducible quadratic dose–response, with 10 g/lamb/d for approximately 25–42 d consistently producing the largest reported gains: Carrillo-Muro. [40] reported average daily gain (ADG) improvements of up to 27% and G:F of 22% in Dorper × Katahdin lambs, and Carrillo-Muro et al. [41] confirmed an optimal inclusion window of approximately 25 d, with fat deposition increasing disproportionately beyond this duration. These figures, however, derive almost entirely from Dorper × Katahdin crosses finished on concentrate-supplemented diets; whether the same magnitude of response occurs in other breeds, or under forage-based feeding, has not been directly tested and should not be assumed. The most strategically important finding for intensive systems is the synergistic interaction between CaPr and ZH: 28 d of CaPr combined with 28 d of ZH produced greater improvements in final body weight (FBW), ADG, dressing percentage, and longissimus muscle area (LMA) than either additive alone [48], although, to date, this finding derives from a single trial by the same research group and awaits independent replication, the clearest demonstration in this literature that mechanistically distinct additive classes can be combined for additive, rather than redundant, benefit.

3.4. Glycerol

Glycerol enters the diet primarily as crude glycerin, a biodiesel by-product typically containing 60–95% glycerol. Unlike PG, glycerol is predominantly fermented in the rumen to propionate and butyrate rather than absorbed intact [42,43] and competitively inhibits hepatic enzymes involved in amino acid catabolism, exerting a modest protein-sparing effect [44]. Productive inclusions of 5–10% of dietary dry matter (DM) produce modest ADG improvements (4–12%), while inclusions above 15% DMI produce an effect attributed mainly to methanol impurities in crude glycerin rather than to glycerol itself [43,45]. Response magnitude is therefore highly dependent on glycerin source and purity, a source of inconsistency across studies that is rarely acknowledged explicitly.

3.5. Sodium Propionate and Calcium Salts

Sodium propionate delivers propionate by the same rumen mechanism as CaPr but without the calcium cation. A direct comparison found no meaningful differences between the two salts in growth, rumen fermentation, or meat quality [46], indicating that the propionate anion, not the counter-ion, is the active principle; the practical implication is that choice between CaPr and NaPr can reasonably be based on cost and availability rather than expected productive difference. Calcium salts of long-chain fatty acids act by a different, complementary route, providing rumen-protected energy that reduces the metabolic demand on hepatic gluconeogenesis rather than augmenting it directly [47].

3.6. Comparative Analysis and Practical Considerations

Across all four compounds, the dose–response relationship is non-linear, and the effective dose range is narrower than commonly assumed from cattle-derived recommendations. The main source of between-study variability is dietary context: gluconeogenic additives consistently improve performance on concentrate-supplemented finishing diets, but show little or no benefit in forage-only systems where dietary propionate supply is not otherwise limiting [39,40]. This context dependency underscores the need for caution against generalized recommendations regarding the use of these additives across diverse basal diets. Comparing across compounds, CaPr shows the most consistent and best-replicated response, PG is better positioned as a short-duration metabolic support tool, and glycerol’s variability is driven as much by feedstock quality as by biology. Practically useful research priorities include systematic dose–response evaluation in breeds beyond Dorper crosses, duration-optimization trials that quantify the fat/lean trade-off at extended inclusion periods, and testing of gluconeogenic additives in combination with classes other than β-AAs.

• Key Messages

• CaPr shows the most consistent productive response among gluconeogenic additives, but the evidence base is concentrated in Dorper × Katahdin crosses on concentrate diets and should not be extrapolated uncritically to other contexts.
• PG is best positioned as a short-term metabolic support tool rather than a sustained growth promoter, given evidence of diminishing response with continued use.
• Glycerol responses are strongly dependent on glycerin source and purity, which limits direct comparison across published trials.
• The CaPr + ZH combination remains the only multi-additive strategy in sheep with reproducible synergistic evidence; other combinations are essentially untested.

• Current Knowledge Gaps

• Dose–response data in breeds other than Dorper crosses and Rambouillet are essentially absent.
• No transcriptomic or metabolomic studies have characterized hepatic or muscle-level responses to gluconeogenic additives in sheep specifically.
• No studies have evaluated gluconeogenic additives under extensive grazing conditions, as opposed to confined feedlot systems.

4. β-Adrenergic Agonists in Intensive Sheep Production

4.1. Pharmacological Rationale

β-adrenergic agonists are the highest-potency growth-promoting compounds used in commercial sheep finishing where their use is authorized. By activating adrenergic receptors (β-ARs) on skeletal muscle and adipose tissue, β-AAs redirect nutrient partitioning away from fat deposition toward muscle protein accretion, producing gains in ADG, G:F, hot carcass weight (HCW), LMA, and dressing percentage that exceed those of any single nutritional additive reviewed in this manuscript [49,50]. Four compounds have been evaluated in sheep: cimaterol, L-644,969, RH, and ZH [50,51,52]. Their productive potency is inseparable from a well-documented cost, discussed in Section 4.5, that distinguishes this class from the other three reviewed here.

Table 3. β-adrenergic agonists evaluated in sheep production: receptor selectivity, dose, supplementation duration, productive outcomes, meat tenderness effects, and regulatory status.

Compound	Receptor Selectivity	Dose in Sheep	Duration (Optimal)	ADG Effect	Carcass	WBSF	Regulatory Status	Reference
Cimaterol (CL 263,780)	β2 >> β1, β3	0.5–3.0 mg/kg DM	21–42 d	↑ 15–30%	↑ LMA; ↓ fat; ↑ HCW	↑ (impaired)	Withdrawn worldwide; research only	[50,53]
L-644,969	β2 >> β1, β3	1.0–3.0 mg/kg BW/d	21–28 d	↑ 10–20%	↑ LMA; ↓ fat	↑↑ (severe impairment)	Research compound only	[50,51]
Ractopamine HCl	β1 ≈ β2	0.35–1.05 mg/kg BW/d	28–42 d (step-up)	↑ 10–24%	↑ HCW, LMA; ↓ fat	Moderate ↑	Approved in the USA, Canada, Brazil, and Australia; BANNED in the EU, China, and Russia	[54,55,56]
Zilpaterol HCl	β2 >>> β1, β3	0.10–0.20 mg/kg BW/d	20–28 d	↑ 10–25%	↑ HCW 5–10%, LMA 8–15%, ↓ fat 15–25%	↑↑ (consistent; extended ageing mitigates partially)	Mexico, South Africa; US voluntary withdrawal 2013; BANNED in the EU	[52,54,55,56,57]

↑, increase; ↓, decrease.

compares these compounds by receptor selectivity, dose, supplementation duration, productive outcomes, and regulatory status.

4.2. Receptor Pharmacology and the Anabolic Cascade

Three β-AR subtypes mediate productive effects in livestock: β1-AR (cardiac and some muscle/adipose tissue), β2-AR (the principal anabolic receptor, abundant in fast-twitch myofibers and the primary target of ZH and cimaterol), and β3-AR (predominantly adipose, thermogenesis, and lipolysis) [58]. Ractopamine shows near-equal β1/β2 affinity, whereas ZH is β2-selective, which explains its more pronounced hypertrophic and adipolytic effects but greater cardiovascular potential at high doses [49,50,58].

Receptor activation triggers a cAMP–PKA signaling cascade that branches into three functionally distinct outcomes, and this branching is worth retaining in some mechanistic detail because it explains the central trade-off of this additive class. PKA, once activated, phosphorylates: (1) the cAMP response element-binding protein (CREB), driving expression of myosin heavy chain (MHC), IGF-1, follistatin, and—critically—calpastatin; (2) components of the mTOR pathway, activating protein synthesis largely independently of the insulin feedback that constrains gluconeogenic additives (Section 3); and (3) calpastatin, directly inhibiting the calcium-dependent protease calpain [58,59,60]. This parallel, insulin-independent route to mTOR activation is the most plausible mechanistic explanation for why β-AAs and gluconeogenic additives produce additive, rather than redundant, growth responses when combined [60]. Figure 3 illustrates this receptor-to-effector signaling cascade in full.

A related and pharmacologically distinctive effect of β2-agonism is induction of a fast-to-faster muscle fiber-type transition: 6-d cimaterol administration to lambs induced de novo expression of MHC-IIB—an isoform not normally present in ovine longissimus muscle—while reducing slower fiber-type proportions [53]. This shift accelerates protein deposition but also produces a fiber-type context with lower intrinsic calpain activity, compounding the calpastatin-mediated reduction in postmortem proteolysis discussed in Section 4.4 [51,53].

4.3. Productive Evidence in Sheep

López-Carlos et al. [54] evaluated RH and ZH step-up programs over 42 d in Dorper × Katahdin ram lambs: β-AA-fed lambs showed greater FBW (+9.6%), ADG (+24.4%), and G:F (+28.2%) than controls, with ZH specifically improving dressing percentage, LMA (+12.8%), and fat thickness reduction (−18.5%). A subsequent trial [52] established 0.15–0.20 mg/kg BW as the productive optimum for ZH in this genotype. Webb et al. [57] documented additive effects of ZH combined with a non-steroidal growth implant, and Barnes et al. [61] found that ZH—but not RH—maintained hypertrophic muscle growth in heat-stressed lambs, consistent with the β2-selective mechanism conferring greater thermal robustness. As with CaPr in Section 3, however, this evidence base is concentrated in a narrow set of genotypes and finishing conditions, and the magnitude of response reported here should not be assumed to generalize automatically to other breeds, climates, or management systems.

Dosing pattern matters as much as dose: a multi-strategy ZH trial found that sustained receptor stimulation (split-dose administration) was required for muscle hypertrophy, whereas intermittent dosing reduced fat deposition without improving HCW, dressing percentage, or LMA [62], indicating that muscle accretion requires continuous mTORC1 activation rather than episodic stimulation. As noted in Section 3, combining CaPr with ZH produced improvements in BW, ADG, dressing percentage, and LMA exceeding either additive alone [48], the strongest evidence in this literature for genuine mechanistic complementarity between additive classes.

4.4. Meat Quality: The Calpain–Calpastatin Trade-Off

β-AA treatment consistently increases Warner–Bratzler shear force (WBSF) by 10–30% relative to untreated controls, and this should be regarded as a near-certain consequence of use rather than an occasional side effect. The mechanism follows directly from Section 4.2: calpastatin upregulation inhibits μ-calpain, which reduces the postmortem proteolytic degradation of myofibrillar proteins that normally tenderizes meat during ageing [49,51,63]. Foundational work with the β2-selective compound L-644,969 in wether lambs confirmed this chain directly, showing reduced calpain activity, elevated calpastatin, reduced postmortem proteolysis, and increased WBSF in a single coordinated response [51]. Practical mitigation strategies—adequate dietary protein during treatment, electrical stimulation at slaughter, and mandatory extended ageing—can partially offset but do not eliminate this effect [54,64].

4.5. Regulatory Status and Animal Welfare

The regulatory landscape for β-AAs is the most fractured of any additive class reviewed here. Ractopamine is approved in roughly 27 countries but banned in more than 160, including the EU, China, Russia, and Taiwan; ZH is approved only in Mexico and South Africa, was voluntarily withdrawn from the US market in 2013, and remains banned in the EU [55,56]. Neither compound is registered for sheep specifically in the USA or EU, meaning that any sheep meat with detectable residues is non-compliant in most major importing markets, a constraint that, in practice, may limit market opportunities for this additive class in many international markets, regardless of its productive advantages. Readers should verify current regulatory status directly with national competent authorities, as approvals, bans, and maximum residue limits are periodically revised and may differ for sheep relative to cattle within the same jurisdiction.

4.6. Receptor Desensitization and the Supplementation Window

Prolonged β-AR stimulation triggers receptor internalization and downregulation, defining an effective productive window of approximately 20–28 d for ZH, extendable to 30–35 d with step-up dosing [52,54]. Beyond this window, returns diminish and, particularly with RH, behavioral excitability can emerge, limiting the practical benefit-to-risk ratio of extended use.

• Key Messages

• Zilpaterol hydrochloride and RH produce the largest single-additive gains in lean yield reviewed in this manuscript, but this advantage is increasingly confined to domestic markets given exclusion from most import destinations.
• Meat tenderness impairment is a mechanistically predictable, near-certain consequence of β-AA use—not an occasional risk—and should be planned for accordingly.
• Productive response depends strongly on dosing pattern (sustained vs. intermittent) and duration; single ADG figures cited without this context understate real variability.
• The evidence base for both ZH and RH in sheep derives predominantly from Dorper × Katahdin crosses; generalization to other genotypes has not been directly tested.

• Current Knowledge Gaps

• Controlled welfare assessments of β-agonist use in sheep under commercial stocking density and heat stress are lacking.
• No sheep-specific pharmacokinetic data exist to inform withdrawal periods matching current export-market residue standards.
• Long-term, multi-cohort data on receptor desensitization patterns in sheep, as opposed to cattle, remain limited.

5. Exogenous Enzymes in Intensive Sheep Production

5.1. The Fiber Digestibility Constraint

Plant cell walls are the primary energetic substrate for ruminants, yet rumen microorganisms degrade only 30–65% of the neutral detergent fiber (NDF) fraction of most practical diets, reflecting the physicochemical complexity of plant cell walls and the kinetic constraints of microbial colonization within rumen retention time [65,66]. Exogenous enzyme preparations amplify this hydrolytic capacity and represent a regulatory-compliant, residue-free strategy for improving nutritional efficiency, first demonstrated when pre-ingestive application of fibrolytic enzymes to forage increased NDF digestibility and ADG in beef steers [65,67].

Table 4. Exogenous enzyme classes evaluated in intensive sheep production: enzymatic activities, commercial sources, target substrates, dose ranges, effects on rumen fermentation, growth response, and practical considerations.

Enzyme Class	Principal Activities (EC)	Commercial Source	Target Substrate	Dose Range	Effect on Rumen Fermentation	Growth Response in Sheep	Practical Considerations	Reference
Cellulase complex	Endo-glucanase (3.2.1.4); cellobiohydrolase (3.2.1.91); β-glucosidase (3.2.1.21)	Trichoderma reesei; T. longibrachiatum; Aspergillus niger	Cellulose β-1,4-glucan in forage cell wall	0.5–2.0 g/kg DM	↑ NDF digestibility 3–8 units; ↑ total VFA 5–12%; ↑ acetate	ADG +5–12% on high-pdNDF diets; no response on low-fiber diets	Requires pre-ingestive liquid application; response scales with dietary pdNDF	[65,68,69,70,71,72]
Xylanase complex	Endo-xylanase (3.2.1.8); β-xylosidase (3.2.1.37); α-arabinofuranosidase (3.2.1.55)	T. reesei; Bacillus subtilis; Humicola insolens	Arabinoxylan hemicellulose in grasses and cereal straws	0.3–1.5 g/kg DM	↑ Hemicellulose digestibility 5–10 units; exposes cellulose microfibrils	ADG +6–14% on straw- and stover-based diets	Most effective in tropical grass and cereal-straw diets; synergistic with cellulase	[65,68,69,70,71,72]
Feruloyl esterase	Feruloyl esterase (3.1.1.73); acetyl xylan esterase (3.1.1.72)	A. niger; Lactobacillus spp.; Clostridium thermocellum	Ferulate ester cross-links between lignin and arabinoxylan	0.1–0.5 g/kg DM	Releases ferulic acid; ↑ accessibility of core polysaccharides	Limited independent ovine data; amplifies cellulase–xylanase response	Highest value in highly lignified forages; usually a minor blend component	[65,68,69,70,71,72]
β-Glucanase	Endo-1,3(4)-β-glucanase (3.2.1.6)	T. reesei; B. subtilis; A. niger	Mixed-linkage β-glucans in barley, oats, and rye	0.2–1.0 g/kg DM	↓ Digesta viscosity; ↑ starch and protein accessibility	ADG +4–9% in barley-based finishing diets	Specific to cereal-based rations; negligible benefit in maize-based diets	[65,68,69,70,71,72]
Amylase	α-Amylase (3.2.1.1); glucoamylase (3.2.1.3); pullulanase (3.2.1.41)	Aspergillus oryzae; B. licheniformis	Starch granules in cereal grains	0.2–1.0 g/kg DM	↑ Ruminal starch degradation; ↑ propionate proportion; ↓ acetate/propionate	ADG +5–10% in high-concentrate finishing diets	Acidosis risk if combined with rapidly fermentable grain; requires buffer support	[65,68,69,70,71,72]
Protease	Serine endoprotease (3.4.21.-); subtilisin	B. subtilis; B. licheniformis	Cell-wall-bound and matrix protein; prolamin starch matrix	0.1–0.6 g/kg DM	↑ Protein degradation rate; risk of ↑ NH3-N if excessive	ADG +3–8% within multi-enzyme blends	Never used alone; excess causes wasteful ruminal deamination and N loss	[65,68,69,70,71,72]
Multi-enzyme complex	Combined cellulase, xylanase, amylase, protease, and β-glucanase	Proprietary blends (e.g., Optimax E®, Fibrozyme®)	Complete diet: fiber, starch, and protein fractions simultaneously	0.2–0.8% DM	↑ NDF and starch digestibility; ↑ net energy 8%; Firmicutes/Bacteroidetes shift; ↑ bacterial alpha diversity	ADG +4–13%; FCR +7–14%; NE +5–10% in feedlot lambs	Best commercial option for mixed diets; quadratic dose response with optimum at 0.4–0.8% DM	[73]

↑, increase; ↓, decrease.

summarizes the principal exogenous enzyme classes evaluated in sheep, their target substrates, dose ranges, and effects on rumen fermentation and growth.

5.2. Enzymatic Classes and Mechanisms

Fibrolytic enzymes (endo-glucanases, cellobiohydrolases, β-glucosidases) form the core cellulolytic system; xylanases and feruloyl esterases target hemicellulose and its lignin cross-links; β-glucanases are relevant in barley- or oat-based diets; and amylolytic enzymes and proteases improve starch and protein accessibility in concentrate-based finishing diets [65,68,69,70,71,72]. Pre-ingestive liquid application—spraying enzyme onto the total mixed ration 30–60 min before feeding—offers the enzyme physical and structural protection against rumen proteolysis and consistently outperforms dry blending or direct rumen infusion [65,74,75]. Morgavi et al. [68] demonstrated true synergy between exogenous and microbial enzymes, with combined hydrolytic activity exceeding additive predictions. Figure 4 depicts this mechanism of exogenous fibrolytic enzyme action and its synergy with endogenous microbial enzymes.

5.3. Productive Evidence in Sheep

Meta-analyses show that exogenous fibrolytic enzymes (EFEs) produce inconsistent results in sheep compared to dairy cattle and beef steers, failing to significantly improve ADG [76]. This species discrepancy stems from trial design rather than additive failure: sheep studies often used forage diets in which potentially digestible NDF (pdNDF) was already high, leaving little room for enzymatic improvement, and ovine studies more often used dry blending rather than the more effective pre-ingestive liquid application [76]. Where both conditions are favorable, results are consistently positive: cellulase supplementation improved digestibility and ADG in lambs fed cereal-straw diets [77], and a multi-enzyme complex produced linear improvements in ADG (+4–13%), net energy (+5–10%), and feed conversion ratio (FCR; +7–14%) in feedlot lambs [73]—these figures derive from a single commercial product trial and should not be generalized across all multi-enzyme formulations without further confirmation. Combining fibrolytic enzymes with a phytochemical (wormwood) improved carcass traits beyond either treatment alone [78], a pattern echoed by the enzyme–phytochemical synergy discussed further in Section 7. Meat quality is generally unaffected at productive enzyme doses—no consistent adverse effects on pH, color, WBSF, or cooking loss have been reported—making this the most commercially conservative quality profile among the four additive classes reviewed [71].

5.4. Factors Explaining Response Variability

Six factors account for most of the inconsistency reported across the sheep enzyme literature: baseline dietary pdNDF (the dominant moderator), application method (pre-ingestive liquid vs. dry blending), the match between the cellulase/xylanase ratio and diet fiber composition, dose (with a quadratic optimum near 0.2–0.5% of DM for multi-enzyme products), the source of rumen fluid used in in vitro screening (cattle rumen fluid tends to overestimate the in vivo ovine response), and interaction with other dietary additives. In practice, this means that a negative enzyme trial in sheep is at least as likely to reflect an unfavorable combination of these factors as a true absence of biological effect, a distinction that the primary literature does not always make explicit.

• Key Messages

• Productive response depends critically on baseline dietary fiber digestibility; enzyme supplementation is unlikely to improve performance where potentially digestible NDF is not limiting.
• Pre-ingestive liquid application consistently outperforms dry blending; much of the inconsistency reported in the sheep literature likely reflects application method rather than enzyme inefficacy.
• Multi-enzyme complexes carry no demonstrated meat quality penalty and face substantially fewer regulatory restrictions than β-AAs in most jurisdictions, although specific product registration requirements still vary by country, making them a comparatively low-risk addition regardless of production system.

• Current Knowledge Gaps

• No metagenomic or metatranscriptomic studies have characterized how exogenous enzymes alter the sheep rumen microbiome in vivo.
• Enzyme blends optimized specifically for ovine, as opposed to bovine, rumen conditions have not been developed or tested.
• Cost-effectiveness data under commercial sheep feedlot conditions are essentially unpublished.

6. Phytochemicals as Functional Feed Additives in Intensive Sheep Production

6.1. Chemical Diversity and Rationale

Plant secondary metabolites—phytochemicals or phytogenics—are the most chemically diverse additive class reviewed here, spanning EOs, condensed tannins (CTs), hydrolysable tannins (HTs), saponins, flavonoids, phenolic acids, terpenoids, organosulfur compounds, and alkaloids [79,80]. Their appeal as alternatives to synthetic growth promotants rests on three converging properties: modulation of rumen fermentation toward more productive, lower-CH₄ profiles; antioxidant and immunomodulatory activity; and transfer to edible tissue, where they improve oxidative stability, fatty acid composition, and sensory quality of lamb meat [79,81]. Figure 5 illustrates the multi-target mechanisms of action of these phytochemical families. Additionally,

Table 5. Major phytochemical families evaluated in intensive sheep production: representative compounds, botanical sources, primary mechanisms, effective doses, effects on rumen fermentation, meat quality outcomes, and key limitations.

Family	Representative Compounds	Botanical Sources	Primary Mechanism	Effective Dose (Sheep)	Rumen Effects	Meat Quality Effects	Key Limitation	Reference
Essential oils	Thymol; carvacrol; cinnamaldehyde; eugenol; limonene	Thymus vulgaris; Origanum vulgare; Cinnamomum verum; Syzygium aromaticum	Plasma membrane disruption; dissipation of proton motive force; H+-ATPase inhibition	50–400 mg/kg DM	↓ Acetate/propionate; ↓ NH3-N 10–25%; ↓ CH4 5–15% in vivo; selective Gram-positive suppression	↓ TBARS 25–45%; ↑ color stability; ↑ C18:2n-6 and total PUFA; herby flavor notes	Microbial adaptation within 4–8 weeks; chemotype variability between batches	[79,80,81,82,83,84]
Condensed tannins	Proanthocyanidins (procyanidin, prodelphinidin polymers)	Schinopsis balansae (quebracho); Acacia mearnsii; Cistus ladanifer; Lotus corniculatus	Protein precipitation at rumen pH; Co/Ni chelation from methyl-CoM reductase; protozoa suppression	10–40 g/kg DM (1–4% DM)	↑ Rumen-undegradable protein; ↓ CH4 10–25%; ↓ protozoa; ↓ Butyrivibrio fibrisolvens biohydrogenation	↑ PUFA and CLA; ↓ SFA; ↑ IMF in some trials; ↓ TBARS 20–35%	>5% DM depresses DMI, NDF digestibility and palatability; astringency	[85,86,87,88]
Hydrolysable tannins	Gallotannins; ellagitannins; castalagin; vescalagin	Castanea sativa (chestnut); Caesalpinia spinosa (tara); Quercus spp.	Gallic/ellagic acid ester hydrolysis; direct antimicrobial and antioxidant action	3–9 g/kg DM (0.3–0.9% DM)	Moderate CH4 reduction; ↑ blood antioxidant capacity; less protein binding than CT	↑ C18:2n-6 and total PUFA; ↓ SFA; ADG +8% in Tan lambs	Hepatotoxic and nephrotoxic at high doses; narrower safety margin than CT	[89]
Saponins	Sarsaponin; quillajasaponins; steroidal and triterpenoid glycosides	Yucca schidigera; Quillaja saponaria; Medicago sativa	Cholesterol complexation in protozoal membranes → osmotic lysis (defaunation); urease inhibition	0.5–4.0 g/kg DM	Defaunation; ↓ CH4 9–15%; ↓ NH3-N; ↑ microbial protein efficiency; ↑ bacterial diversity	Largely neutral; indirect benefit via improved N efficiency	Protozoal adaptation within 3–5 weeks; efficacy declines with continuous feeding	[79]
Flavonoids	Quercetin; resveratrol; catechin; naringin; hesperidin; anthocyanins	Citrus by-products; Vitis vinifera; Camellia sinensis; Allium cepa	Keap1 Cys273/Cys288 modification → Nrf2–ARE activation; IKK inhibition → NF-κB blockade	100–800 mg/kg DM	Minor fermentation effects; primarily post-absorptive action	↓ TBARS 30–50%; ↑ redness (a*) retention; ↑ serum GPx, SOD, and CAT; ↓ TNF-α, IL-1β, and IL-6	Low ruminal bioavailability; extensive microbial degradation; rumen-protection often required	[90,91,92,93]
Phenolic acids	Rosmarinic acid; ferulic acid; chlorogenic acid; caffeic acid	Rosmarinus officinalis; Salvia officinalis; cereal brans; Coffea spp.	Direct radical scavenging; Nrf2 activation; metal chelation	200–1000 mg/kg DM	Minimal fermentation modification at productive doses	↓ TBARS throughout retail display; color stability 9–12 d; commercial BHT/BHA alternative	Rapid ruminal degradation; limited dose–response data in sheep	[94]
Terpenoids (non-volatile)	Carnosic acid; carnosol; ursolic acid; β-carotene	Rosmarinus officinalis; Artemisia spp.; Salvia spp.	Lipid peroxidation chain-breaking; membrane stabilization	0.12% dietary carnosic acid; 200–400 mg/kg feed rosemary diterpenes; 0.04–0.16 mL/kg feed Salvia sclarea extract	Modest antimicrobial and antiprotozoal activity	↓ Lipid oxidation and meat discoloration; ↓ rancid volatiles under MAP storage; ↑ antioxidant enzyme activity (SOD, CAT, GPx) with Salvia sclarea; effects less persistent than vitamin E	Evidence concentrated in Rosmarinus/Artemisia; efficacy formulation-dependent; Salvia genus represented by a single trial	[95,96,97,98,99,100,101]
Organosulfur compounds	Allicin; diallyl sulfide (DAS); diallyl disulfide (DADS); diallyl trisulfide (DATS)	Allium sativum (garlic); Allium cepa; Allium porrum	Thiol modification of methyl-CoM reductase and F420-reducing hydrogenases in methanogens	1–8 g/kg DM garlic; 20–200 mg/kg allicin	↓ Methanogen density; ↓ CH4 25–70% in vitro but only 5–15% in vivo; ↑ propionate; ↑ cellulolytic bacteria	Neutral to slightly positive; sulfurous off-flavors at high doses	Bacterial glutathione-dependent allicin inactivation; large in vitro–in vivo discrepancy	[102,103,104]
Alkaloids	Piperine; capsaicin; berberine; sanguinarine	Piper nigrum; Capsicum spp.; Berberis spp.; Macleaya cordata	Membrane permeabilisation; efflux pump inhibition; anti-inflammatory signaling	10–200 mg/kg DM	Variable and inconsistent fermentation effects; some DMI stimulation	Insufficient ovine evidence to support quality claims	Narrow safety margin; toxicity risk; least characterized family in sheep	No sheep-specific reference identified

Aproximate strength of sheep-specific evidence per family. Essential oils: moderate; condensed tannins: moderate–high; hydrolysable tannins: limited (single trial); saponins: limited (largely extrapolated from cattle/goats); flavonoids: limited–moderate; phenolic acids: limited; non-volatile terpenoids: limited, concentrated in rosemary/Artemisia, with a single trial for Salvia; organosulfur compounds: limited–moderate, with a large in vitro–in vivo gap; alkaloids: no sheep-specific reference identified. ↑, increase; ↓, decrease; →, leads to/is converted to. a*, redness coordinate of the CIELAB color space (red–green axis); higher values indicate greater redness and are used as an index of meat color stability.

summarizes their botanical sources, primary mechanisms, and meat quality outcomes.

6.2. Essential Oils

Essential oils are volatile lipophilic compounds—predominantly monoterpenes (thymol, carvacrol) and phenylpropanoids (cinnamaldehyde, eugenol)—that disrupt bacterial membrane function, selectively suppressing fermentation pathways at sub-lethal rumen doses [79,81]. Calsamiglia et al. [81] established the foundational dose–response relationship, showing that thymol and carvacrol reduce the acetate/propionate ratio and ammonia production in rumen batch culture. In sheep, a cinnamaldehyde–carvacrol blend reduced the relative abundance of major biohydrogenating bacteria and increased longissimus polyunsaturated fatty acid (PUFA) content [82], linking microbiome modulation directly to meat fatty acid enrichment, and an EO blend reduced in vitro CH₄ production without impairing performance [83] (in vivo CH₄ data for this specific blend in sheep are not yet available). The persistent limitation is microbial adaptation: rumen bacteria develop tolerance to EOs compounds within 4–8 weeks, meaning that trial durations shorter than this window likely overstate sustained field performance [84].

6.3. Tannins

Condensed tannins (from quebracho, mimosa, and Cistus ladanifer) and hydrolysable tannins (from chestnut and tara) act through complementary mechanisms: protein precipitation that increases rumen-undegradable protein, direct inhibition of methanogenic archaea, reduction of protozoal populations, and suppression of the biohydrogenating bacterium Butyrivibrio fibrisolvens [85,86,87,89]. This last mechanism is the most reproducible outcome in the tannin literature: individual trials consistently report decreased saturated fatty acid (SFA) proportion and increased unsaturated, polyunsaturated, and conjugated linoleic acid (CLA) fractions in longissimus muscle [86,87], while growth-performance responses (ADG, FCR) vary considerably by tannin source and inclusion level [88,89], indicating that tannins are more consistently supported as a meat-quality intervention than as a growth promoter per se. Chestnut tannin extract improved ADG alongside antioxidant capacity and fatty acid profile in Tan sheep [89], and Acacia mearnsii CTs improved carcass yield and fatty acid profile in confined lambs [88], though both findings derive from single-breed trials and should be read as illustrative rather than general. At the whole-diet level, CT supplementation of 2–4% of DM achieves CH₄ reductions of 10–25% per unit of organic matter digested [85,86].

6.4. Saponins

Saponins from Yucca schidigera and Quillaja saponaria act primarily by complexing cholesterol in rumen protozoal membranes, causing defaunation that reduces methanogen activity, increases bacterial diversity, and improves microbial protein synthesis efficiency [79]. In sheep, saponins at optimal doses produce 9–15% CH₄ reduction and 5–10% ADG improvement in forage-based systems, but protozoal adaptation within 3–5 weeks limits sustained efficacy, the same continuous-dosing limitation seen with EOs [79].

6.5. Flavonoids and Phenolic Acids

Flavonoids (quercetin, resveratrol, and curcumin) and phenolic acids (rosmarinic, ferulic, and chlorogenic acids) act mainly as post-absorptive antioxidants. They activate the nuclear factor erythroid 2-related factor 2 (Nrf2) cellular antioxidant defense pathway and suppress nuclear factor kappa B (NF-κB)-mediated inflammatory signaling, the two principal mechanisms underlying their reported benefits [90,91,92,93]. In practical terms, this translates into reduced pro-inflammatory cytokine expression under immune challenge [92], improved immune markers under heat stress [93], and—most consistently across trials—reduced meat lipid oxidation (TBARS) and better color retention during refrigerated storage [94], positioning rosmarinic acid in particular as a natural alternative to synthetic antioxidants such as BHT/BHA [94].

6.6. Non-Volatile Terpenoids

Non-volatile diterpenes and triterpenoids—principally carnosic acid and carnosol from Rosmarinus officinalis, together with monoterpene- and sesquiterpene-rich essential oils from Artemisia spp. and Salvia spp.—act mainly as lipid-peroxidation chain-breakers and membrane stabilizers, a mechanism that distinguishes this family from the fermentation-modifying essential oils discussed in Section 6.2 [95,96]. In lambs, dietary carnosic acid at 0.12% of the diet reduced myoglobin oxidation and slowed meat discoloration in the gluteus medius of fattening lambs [95], and comparable benefits on meat quality were reported in suckling lambs given carnosic acid directly [97]. Rosemary diterpene extracts standardized to carnosic acid and carnosol content, at inclusion levels of 200–400 mg/kg feed, reduced volatile rancidity markers and improved the oxidative stability of lamb meat stored under high-oxygen modified atmosphere packaging, performing comparably to—though generally less persistently than—dietary vitamin E [98,99]. Oral administration of Artemisia herba-alba and Rosmarinus officinalis essential oils similarly improved the antioxidant status and oxidative stability of lamb muscle [96], and Vasta et al. [94] independently detected sesquiterpene volatile markers (copaene, β-caryophyllene) in the meat of lambs supplemented with the same two botanical sources. The evidence is not uniformly positive: a rosemary extract embedded in a fat matrix produced no measurable effect on production parameters, lipid oxidation, or meat color at any dose tested in one lamb trial [100], underscoring that formulation and bioavailability—not the mere presence of carnosic acid—determine in vivo efficacy. Beyond rosemary and Artemisia, ovine evidence for other genera in this family remains essentially limited to a single trial: a Salvia sclarea (clary sage) extract rich in linalyl acetate and linalool improved average daily gain, nutrient digestibility, and antioxidant enzyme activity (superoxide dismutase, catalase, glutathione peroxidase) in growing lambs [101], offering the first ovine evidence for this genus, though the active compounds involved are monoterpenes rather than the non-volatile diterpenes that dominate the rosemary literature. Overall, the ovine evidence base for this family, while real and growing, remains concentrated in a small number of research groups working primarily with rosemary diterpenes, and systematic dose–response and cross-genus comparisons remain largely untested.

6.7. Organosulfur Compounds

Garlic-derived compounds (allicin and its sulfide derivatives) inhibit methanogen enzymes and reduce protozoal populations while increasing propionate-producing bacteria [102,103,104]. The clearest limitation in this literature is the large discrepancy between in vitro and in vivo efficacy: CH₄ reductions of 25–70% [102,103,104] reported in vitro fall to only 5–15% in vivo, most plausibly because rumen bacteria possess glutathione-dependent enzymes that inactivate allicin with continued exposure [104]. This gap is a useful general caution for the phytochemical class as a whole—in vitro screening results should not be taken as predictive of in vivo magnitude without confirmatory animal trials.

6.8. Integrated Effects on Meat Quality

Across the phytochemical families reviewed, four mechanisms recur in explaining improved meat quality: direct antioxidant deposition in muscle tissue, reducing postmortem TBARS by 20–50% [79,94]; inhibition of biohydrogenating bacteria, increasing PUFA and CLA content [82,87]; reduced metmyoglobin formation, extending color stability during retail display [94]; and, at high doses of some compounds, altered sensory notes, generally positive for EO phenolics but potentially negative (sulfurous) for organosulfur compounds at excessive inclusion [81].

• Key Messages

• Phytochemicals are the only additive class reviewed that combines growth support, meat quality improvement, and quantifiable CH₄ mitigation, although this generalization masks substantial heterogeneity among phytochemical families—effects, doses, and mechanisms differ markedly between essential oils, tannins, saponins, flavonoids, terpenoids, and organosulfur compounds, as detailed in Section 6.2, Section 6.3, Section 6.4, Section 6.5, Section 6.6 and Section 6.7 and Table 5, aligning this class most closely with current sustainability expectations.
• Efficacy is consistently limited by rumen microbial adaptation within 3–8 weeks; continuous single-compound feeding likely overstates sustained field performance.
• Reported effects vary substantially with plant chemotype and extraction method; results from one tannin or EO source do not necessarily generalize to others.
• In vitro CH₄-mitigation results, particularly for organosulfur compounds, substantially overstate in vivo efficacy and should not be extrapolated without confirmatory animal trials.

• Current Knowledge Gaps

• No microbiome-guided (individualized) phytochemical supplementation strategies have been tested in sheep.
• Rotational or cyclical dosing protocols designed to delay microbial adaptation have not been evaluated.
• Life cycle assessment data quantifying the net system-level emissions impact of phytochemical supplementation are absent.
• Systematic dose–response and cross-genus comparisons among non-volatile terpenoid sources (Rosmarinus, Artemisia, Salvia) remain largely untested in sheep, and evidence beyond rosemary diterpenes is limited to single trials.

7. Comparative Analysis of Functional Feed Additives

7.1. Cross-Additive Performance Hierarchy

The magnitude of ADG improvement follows a consistent hierarchy under intensive finishing conditions: β-AAs (10–25%) > gluconeogenic additives (20–27% at optimal dose and duration) > exogenous enzymes (5–15%, diet-dependent), broadly similar to phytochemicals (0–12%, compound- and dose-dependent) [40,54,76]. These ranges overlap substantially and derive from heterogeneous studies, so the ordering above should be read as a general tendency rather than a precise ranking. This ordering should be read with an important caveat: it reflects best-case results from the most favorable dosing and dietary conditions reported for each class, not the average outcome a producer should expect under field conditions, where diet composition, breed, and management deviate from the controlled-trial ideal. Improvement in feed conversion ratio (FCR) tracks closely with ADG but is additionally shaped by intake effects—β-AAs typically reduce DMI while increasing ADG, producing disproportionate G:F of 15–30%, compared with 20–26% for gluconeogenic additives and 5–10% for enzymes and phytochemicals [40,54,76].

β-AAs remain the only class with a demonstrated capacity to substantially alter carcass composition: ZH increases dressing percentage by 2–5% and LMA by 8–15% and reduces subcutaneous fat by 15–25%, a genuine repartitioning effect not reproduced by any other single additive at practical doses [54,105]. CaPr produces more modest HCW and LMA gains without this dramatic fat-to-lean shift [40], while the CaPr + ZH combination exceeds either compound alone [48], the clearest practical evidence in this literature that additive classes complement rather than substitute for one another.

7.2. Meat Quality: Divergent Class Effects

Meat quality is where the four classes diverge most sharply, and this divergence has a direct practical implication for additive selection: β-AAs consistently impair tenderness (WBSF +10–30%) through the calpain–calpastatin mechanism detailed in Section 4 [51,59], while phytochemicals consistently improve oxidative stability (TBARS reduced by 20–50%) and color retention through antioxidant mechanisms [82,94]. Exogenous enzymes and gluconeogenic additives are largely meat-quality neutral. This pattern suggests a mechanistically coherent multi-additive strategy—combining ZH for lean yield with a phytochemical antioxidant to offset its tenderness cost—that is plausible on paper but has not, to our knowledge, been directly tested in a single controlled trial; this represents one of the more actionable research gaps identified in this review.

7.3. Environmental and Regulatory Dimensions

Phytochemicals—CT (10–25% CH₄ reduction), saponins (9–15%), Eos, and organosulfur compounds (5–15% in vivo)—are the only class with demonstrated anti-methanogenic activity in sheep [85,104]. Gluconeogenic additives and β-AAs are CH₄-neutral at the animal level, though improved feed conversion from β-AA use indirectly reduces land and resource use per unit of product, a benefit better captured by system-level life cycle assessment than by per-animal CH₄ measurement. Regulatory access is unrestricted or only lightly restricted in most major markets for gluconeogenic additives, exogenous enzymes, and most phytochemicals, although specific product registration is still required nationally but severely constrained for β-AAs, with RH banned in more than 160 countries and ZH excluded from all major import markets except Mexico and South Africa [55]. In practice, this regulatory asymmetry means that additive strategy is no longer a purely biological decision: for export-oriented producers, the three non-restricted classes are the only available options regardless of the productive ceiling that β-AAs could otherwise offer.

Table 6. Comparative analysis of the four functional feed additive classes in intensive sheep production across productive, meat quality, environmental, and feasibility dimensions.

Dimension	Gluconeogenic Additives	β-Adrenergic Agonists	Exogenous Enzymes	Phytochemicals	Reference
Primary mechanism	Propionogenesis → glucose → insulin → mTOR	β2-AR → cAMP → PKA → mTOR + anti-proteolysis	Cell wall hydrolysis → ↑ VFA → energy + protein supply	Antimicrobial; antioxidant; Nrf2; NF-κB; biohydrogenation inhibition	See Table 2, Table 3, Table 4 and Table 5
ADG improvement (sheep)	20–27% (CaPr optimal)	10–25% (ZH/RH)	5–15% (diet-dependent)	0–12% (variable)	[40,54,76]
G:F improvement	20–26%	15–30%	5–10%	0–10% (variable)	[40,54,76]
LMA effect	↑ 5–8%	↑ 8–15% (ZH)	↑ 0–5% (positive trials)	No consistent effect	[40,54,105]
Fat reduction	Minimal at optimal dose	↓ 15–25% subcutaneous; ↓ 35–40% perirenal (ZH)	Minimal	Minimal direct	[54,105]
Meat tenderness (WBSF)	Neutral	↑ 10–30% (impaired)	Neutral to ↓ improved	Neutral to ↓ improved	[51,59,82,94]
Oxidative stability (TBARS)	Neutral	Neutral	Neutral to ↑	↓ 20–50% TBARS (↑ stability)	[82,94]
Meat FA profile	Neutral	Neutral	↑ PUFA (enzyme blends)	↑ PUFA, ↑ CLA (CT, EOs)	[82,87]
Enteric CH4	Neutral	Neutral	Variable	↓ 10–25% (CTs); ↓ 5–15% (saponins, EOs, allicin)	[85,104]
Regulatory status	Accepted worldwide	Restricted: EU, China, Russia; banned RH; ZH withdrawn in the US	Accepted worldwide	Accepted worldwide	[55]
Evidence consistency	Moderate–high (CaPr); moderate (PG, glycerol)	High (ZH, RH)	Low–moderate in sheep	Low–moderate (variable by compound)	Authors’ synthesis (Section 3, Section 4, Section 5, Section 6 and Section 7)
Best combined with	β-AAs (synergistic)	Gluconeogenic additives + steroidal implants	Phytochemicals (synergistic)	Enzymes + tannins + EOs combinations	[48,78]

Values in this table are drawn from independent studies conducted under heterogeneous experimental conditions (breed, diet, duration) and do not represent direct head-to-head comparisons among additive classes; they should be interpreted as indicative ranges, not controlled contrasts. ↑, increase; ↓, decrease; →, leads to/is converted to.

summarizes these productive, meat quality, environmental, and feasibility comparisons across all four additive classes, and Figure 6 integrates these mechanisms into a single schematic model.

7.4. Economic Considerations

Figure 7 summarizes these comparative performance and meat quality ranges across all five metrics discussed below. Published cost-effectiveness data specific to intensive sheep systems are notably scarce across all four additive classes; the comparative cost profile in

Table 7. Approximate cost–return profile of functional feed additive classes in intensive sheep finishing. Cost and return are qualitative, given the scarcity of verified per-head cost data in the published sheep literature.

Additive Class	Relative Cost	Typical Productive Return	Payback Context	Practical Recommendation
Gluconeogenic (e.g., CaPr)	Low	High (ADG +20–27%)	Favorable on concentrate-based diets; poor on forage-only systems	Use routinely in feedlot finishing; avoid in extensive/pasture systems
β-adrenergic agonists	Moderate	Highest (ADG +10–25%)	Favorable only where the target market accepts residues	Restrict to domestic-market production; discount for expected tenderness penalty
Exogenous enzymes	Moderate–high	Low–moderate (ADG +5–15%)	Favorable only if baseline fiber digestibility is genuinely limiting	Use only after confirming the diet is fiber-limited; verify with pdNDF assay
Phytochemicals	Variable (compound-dependent)	Low–moderate (ADG 0–12%)	Favorable where methane or quality premiums apply	Prioritize where sustainability certification or premium markets apply

Given the scarcity of verified per-head cost data for intensive sheep systems, the recommendations in this table should be read as directional starting points—e.g., “consider as a routine option, pending local verification” rather than an unconditional recommendation—and adapted to local economic and regulatory conditions.

is therefore necessarily qualitative rather than based on verified per-head cost figures and should be treated as a starting framework for producer decision-making rather than a substitute for local economic analysis. Gluconeogenic additives are generally low-cost relative to their productive return, but that return is realized only on concentrate-based diets, making them a poor investment in forage-only systems regardless of unit cost. β-AAs offer the largest productive return per animal, but that return is contingent on market access—the same compound that is highly cost-effective in a domestic market becomes unusable, and, therefore, a wasted cost, in an export-oriented operation. Exogenous enzymes carry a moderate cost that is only justified when baseline dietary fiber digestibility is genuinely limiting; applied to an already low-fiber, high-quality diet, the marginal return may not cover the additive cost. Phytochemical cost varies widely by compound and extraction method, but where CH₄ mitigation or premium-market certification carries a price signal, this class is increasingly likely to justify its cost on sustainability grounds alone, independent of its direct growth effect.

• Key Messages

• Performance rankings across additive classes reflect best-case trial conditions, not average field outcomes; diet, breed, and management context should be checked against the specific conditions under which each figure was generated before extrapolating.
• The apparent complementarity between β-AAs (lean yield) and phytochemicals (tenderness/oxidative offset) is mechanistically plausible but empirically untested in a single controlled trial, making it one of the more immediately actionable research priorities identified in this review.
• Regulatory access, not biological efficacy alone, increasingly determines which additive classes are viable for a given production system and target market.
• Verified, sheep-specific cost-effectiveness data are largely absent from the published literature across all four additive classes; economic recommendations in this review should be treated as directional rather than definitive.

• Current Knowledge Gaps

• No factorial trial has directly tested a β-agonist + phytochemical-antioxidant combination for simultaneous lean-yield and tenderness outcomes.
• Verified per-head cost and return data for any of the four additive classes under commercial sheep production conditions are largely unpublished.
• System-level life cycle assessment comparing additive classes on a common environmental-impact basis has not been undertaken for sheep.

8. Conclusions

Functional feed additives offer a genuine opportunity to close the long-standing efficiency gap between sheep and monogastric livestock, but no single class is universally superior. Gluconeogenic precursors and β-AAs deliver the largest gains in growth and lean yield, yet their value is increasingly constrained by breed-specific responses, meat quality trade-offs, and, for β-AAs, exclusion from most export markets. Exogenous enzymes and phytochemicals produce more modest gains but carry no regulatory restrictions, and phytochemicals in particular combine growth support with real reductions in enteric CH₄ and improved meat quality, making them well suited to producers facing sustainability or market-access pressures. In practice, the right additive strategy depends on diet, breed, target market, and environmental goals rather than on growth performance alone, and combining additives with complementary mechanisms may offer additional benefits in specific contexts, although, as noted throughout this review, such combination strategies remain largely untested and should not yet be adopted as a general recommendation.

Moving this field forward will depend on closing three practical gaps: testing multi-additive combinations that current evidence suggests should work well together but that have not yet been directly trialed; generating real cost and environmental impact data under commercial conditions, rather than relying on trial-scale estimates; and validating early precision-nutrition tools so that additive use can be tailored to individual animals or flocks rather than applied uniformly. Bridging these gaps is critical to align additive strategies with the evolving productivity, sustainability, and market demands of the sheep industry.

In practice, this means: gluconeogenic additives are best prioritized on concentrate-based finishing diets where propionate supply is limiting; β-AAs should be reserved for domestic markets that accept residues and where the tenderness trade-off is commercially acceptable; exogenous enzymes are justified once baseline dietary fiber digestibility is confirmed to be limiting (e.g., via a pdNDF assay); and phytochemicals should be prioritized where methane mitigation or premium/export-market certification carries a direct economic or regulatory incentive. These are directional decision rules derived from the evidence reviewed here, not a validated decision-support algorithm, and should be adapted to local conditions.