Early studies from [17] showed that in ruminants, the percentage of dietary Zn absorbed decreases as dietary Zn increases. [18] found a linear increment in endogenous fecal loss of Zn when rats were feed increasing levels of dietary Zn ranging from 0 to 8400 ppm. Similarly, in a study conducted in growing pigs receiving 67 Zn, the addition of phytase increased Zn absorption but also resulted in higher endogenous fecal loss [19] . Concurrently, reductions of urinary and fecal Zn excretion by 48% and 46%, respectively were found in humans when dietary Zn was reduced from 85 to 12 μmol/d. Despite the influence of dietary levels, the requirement of Zn appears to be affected by other dietary factors. However, all of the factors and interactions that play a role on Zn bioavailability are not clearly defined [20] . According to [21] , the two major dietary factors affecting Zn bioavailability are the presence of organic chelating agents and the interaction with metallic ions, with Cu and Ca being the most important antagonists.

Early studies from [22] showed that in vitro microbial protein synthesis was increased together with a reduction in NH₃-N concentration when rumen fluid was incubated with additional Zn as ZnCl₂ or ZnSO₄. According to the authors, this response is due to an effect of Zn in increasing microbial enzymatic activity. However, further studies from [23] showed differences in the behavior of rumen microbial populations in the presence of Zn. While protozoa easily incorporated Zn and were tolerant to high Zn concentrations (25 μg/mL), cellulose degradation from rumen bacteria was deeply depressed, along with bacterial urease activity. In partial agreement, [24] found that addition of 5 μg Zn/mL of rumen fluid increased cellulose digestion by 24%, but addition of 20 μg/mL of Zn depressed it by 31%. [25] found that low supplementation levels (10 to 15 ppm Zn in incubation fluid) inhibited in vitro urea hydrolysis and retarded ammonia accumulation. Similarly, when Zn was added in vivo at 250 ppm Zn on DM basis, molar proportion of propionate was increased with the consequent decrease in acetate: propionate ratio, and rumen ammonia was decreased due to microbial urease inhibition. However, when Zn was added to achieve a level of 470 ppm on DM basis, a tendency for depressed DM digestibility was noticed. In addition, [26] found a decrease in total VFA when Zn was provided to steers as Zn methionine or Zn glycine compared with ZnSO₄ at concentrations closer to physiological levels (20 ppm). In the same study, molar proportion of propionate was increased by Zn methionine and that of butyrate was decreased, with the consequent reduction in the acetate: propionate ratio. The authors conclude that supplementation of Zn methionine may alter ruminal fermentation.

While these differences can be partially explained by the dose of Zn and the fermentation substrate used by the different authors, an alternative mechanism is proposed by [27] , who found that addition of 50 μg/mL of Zn to in vitro incubations decreased cellulose digestion at 24 h, but not at 48 h, resulting in an overall decrease in the rate, but not the extent of digestion. The authors conclude that the initial decrease in cellulose digestion might be related to a direct effect of Zn on inactivation of bacterial cellulase, since heavy metal salts can precipitate and denature soluble proteins and enzymes. However, enough cellulase activity may be present to overcome those negative effects of high Zn concentrations. In addition, the accumulation of Zn in bacterial wall [28] might be affecting the adhesion of microbial cells to cellulose particles, a limiting step in cellulose fermentation as previously established [29] .

There is scientific evidence showing that organic Zn is metabolized differently than inorganic sources. In four consecutive studies conducted by [30] , Zn was better retained when added as Zn methionine than ZnO in lambs and heifers. However, the observed improvement was not due to higher absorption but to a lower urinary Zn excretion in animals receiving Zn methionine, and only minor changes in blood parameters were noticed. Similarly, in a study conducted by [31] in calves, organic or inorganic Zn supplementation did not affect concentration of serum enzymes (alkaline phosphatase, glutamate oxaloacetate transaminase, glutamate pyruvate transaminase and super oxide dismutase) or mean concentrations of different serum vitamins (retinol, β-carotene, α-tocopherol) and hormones (triiodothyronine, thyroxin, insulin and testosterone). Furthermore, blood parameters from supplemented groups did not differ from unsupplemented controls. In a comparative study, [32] found no differences in liver Zn and plasma Zn concentration of steers receiving ZnSO₄ or Zn amino acid complex. However, confounding effects may account for these findings, given the differences in mineral status of the animals at the beginning of the study. In partial agreement, no differences in plasma Zn concentration of steers receiving ZnSO₄, Zn methionine complex or Zn glycine were found by [26] . Although large variations prevented from obtaining differences in absorbed or retained Zn, Zn glycine resulted in larger liver Zn concentrations. Conversely, using organic and inorganic Zn. [33] found higher concentrations of Zn in plasma of supplemented beef steers than in controls, despite the Zn source. Interestingly, in animals receiving an implant containing estradiol benzoate and testosterone propionate, weight gain resulted higher when Zn was supplemented as ZnSO₄ that of Zn propionate [33] . In a large production study including 250 dairy cows, only a tendency for improvement in milk production was obtained when Cu, Zn, Mn and Co as sulfates, were replaced by organic forms of the minerals. Liver concentration of minerals was unaffected by mineral source, but supplementation with organic minerals resulted in increased milk solids and a decreased incidence of sole ulcers [34] . In another large scale study conducted on 573 dairy cows [35] , supplementing 75% of the requirement of Zn as Zn methionine achieved the same hepatic Zn concentration than supplementing 100% of the requirement as ZnSO₄. Based on the lack of differences in health and productive performance the authors suggest that mineral content of liver is not an accurate predictor of cow’s response to different sources and levels of trace minerals. In agreement, [34] suggest that Zn, Mn, and Cu content of liver is a poor indicator of trace mineral status. According to [2] , the supposed benefits of organic sources of Zn on Zn availability claimed in studies conducted in monogastrics cannot be adopted in ruminants since phytic acid, a major antagonist of Zn absorption, is largely hydrolyzed in the rumen. An additional difficulty in the evaluation of Zn availability from different sources relies on the fact that Zn is absorbed according to the animal needs and homoeostasis in ruminants is achieved primarily by control of intestinal absorption [36] .

According to the [1] , there is no exact data regarding the maintenance requirements of Mn in dairy cattle. However, the coefficient of intestinal absorption for Mn in adult cattle is known to be as low as 1% of ingested Mn or even lower (37; 20), but the absorption in young calves is considerably higher [37] . Despite the generalized idea of poor Mn absorption, [8] suggested that this situation is partly a reflection of the substantial surplus of Mn provided by most practical rations, since higher coefficients of absorption were obtained when animals received diets marginal in Mn [38] . For this reason, the [1] adopted a conservative coefficient of 0.75% for Mn absorption. In agreement, [39] reported a coefficient of 0.54% for Mn absorption in dairy cows. The study of dietary factors influencing Mn bioavailability has received little attention, probably because Mn deficiency is not considered to be a major problem in ruminants [20] . In addition, most of the information available has been generated in monogastric models. [40] observed a 10% and 13% increase in kidney and bone Mn, respectively from chicks receiving 12 ppm of virginiamycin in the diet. In a further study [41] , the addition of 4 ppm lincomycin resulted in higher concentrations of Mn in bone. However, while virginiamycin and other antibiotics are currently used as feed additives for ruminants [42] , their role in Mn absorption in cattle remains unknown. According to [43] , the intestinal absorption of Mn is negatively affected by dietary levels of Ca and P. Similar results were found by [44] , who found a 45% reduction in Mn in the tibia of chicks fed excess Ca and P. However, further evidence provided by [45] indicated that while P has a negative effect on Mn absorption, no deleterious effects on Mn metabolism are obtained with excess of dietary Ca, but effects of Ca and P are difficult to distinguish since diets are usually enriched with both minerals to maintain a Ca:P physiological ratio [9] . There have been no reports relating Mn absorption with forage composition in ruminants, but phytate and fiber are known to be the main antagonists in monogastrics species including human [46] , swine [47] . Through microbial activity, both phytates and fiber are broken down in rumen [48] . For that reason, [9] suggests that Mn absorption in ruminants may not be affected by the presence of phytates, being higher than that usually reported for monogastrics.

Limited information is available regarding the role of Mn in rumen fermentation. According to [49] , Mn has a slight stimulatory effect on urease activity. [25] reported a 6% increment in IVDMD when Mn was added to incubations at a dose of 100 ppm. In a previous study, [50] found a reduction in cellulose digestion when Mn was omitted from in vitro incubations. However, [24] reported that cellulose digestion peaked at Mn concentrations of 10 to 20 ppm but was completely abolished when Mn was added at 300 ppm. Using 54 Mn, [51] [52] observed that Mn accumulation was higher in rumen bacterial cell walls than in cytoplasm, and that the uptake was similar in bacteria and protozoa, but the biological implications of this fact are not established. [53] fed ram lambs diets that contained from 13 to 45 mg of Mn/kg DM during 84 days. While the number of rumen bacteria was not affected by Mn, the large rumen bacteria (identified as those with a diameter of 12.9 to 16.2 μm) resulted lower with the lowest Mn intake and highest with dietary Mn provided at 30 mg/kg. This may be particularly relevant since large rumen bacteria contain more protein than small rumen bacteria [54] . However, despite this variation in microbial populations, no effect of Mn on DM digestibility was observed. [55] suggested that sheep consuming a diet high in fiber and low in protein may respond to Mn supplementation in excess of 36 μg/g DM, but Mn requirements of the rumen microbes may be increased by the consumption of low quality roughages. In agreement, [56] suggest that the optimum content of Mn in the diet may be as high as 120 μg/g DM on the basis of results from in vitro studies.

The aim of mineral supplementation is to increase the biological availability of the target mineral(s), defined as the degree to which an ingested element is absorbed and can be utilized in metabolism by the animal [57] . Different sources of Mn are currently available as supplements for animal diets. Among the inorganic sources, the most commonly used include manganese carbonate (MnCO₃), hausmannite (Mn₃O₄), manganese oxide (MnO), manganese dioxide (MnO₂), manganite (Mn₂O₃), manganous chloride (MnCl₂∙4H₂O) and manganese sulfate (MnSO₄) [8] [58] [59] [60] . Conversely, sources of Mn classified as “organic” include Mn-methionine, Mn-proteinate and Mn-polysaccharide [42] [61] [62] [63] . Unfortunately, only few studies have been conducted comparing the relative bioavailability of Mn sources in ruminants fed physiological concentrations of Mn [64] . According to [65] , some chelates and complexes may improve the mineral bioavailability above that of soluble inorganic forms, as later shown by [42] in lambs comparing Mn-methionine with MnO. However, no differences were obtained in the same study when Mn-methionine was compared with MnSO₄. Similarly, [66] compared the bioavailability of Mn from different organic sources and MnSO₄ in broilers. The authors concluded that only organic Mn sources with moderate or strong chelation strength can provide higher relative bioavailabilities due to their ability to resist Ca antagonisms during the digestion process. In addition, studies conducted on female chicks showed an increase in Mn retention from a Mn-methionine chelate compared with MnO [67] .

The amount of dietary Cu needed to supply Cu requirements for maintenance, growth and lactation varies with the age of the animal, the chemical form of dietary Cu and the presence of dietary substances interfering with Cu absorption [1] . Compared with monogastrics where Cu is fairly well absorbed (30% - 75%), absorption in adult ruminants is low, ranging from 1% to 10% of dietary Cu [8] [68] . However, before developing a functional rumen, Cu absorption in lambs can be as high as 70-85% of the dietary supply [69] . The reason for this decrease in Cu absorption appears to be related to interactions occurring at the rumen environment, including the Cu-S-Mo [9] [70] [71] , Cu-S [72] [73] , and Cu-Fe [74] [75] antagonisms. More recently, results relating high levels of dietary Mn with Cu deficiency have been reported [76] [77] .

In the presence of ruminal H⁺ ions, dietary S is reduced to sulfide, which then reacts with Mo [78] to form different thiomolybdates (mono-, di-, tri-, tetra-thiomolybdates) according to the following reactions:

In the gastrointestinal tract, thiomolybdates have been shown to bind Cu preventing its absorption, while increasing the Cu fraction associated with the solid phase of the rumen content at expense of a reduction in the fluid phase. Thiomolybdates associated with solid rumen digesta (bacteria, protozoa, and indigested feed particles) form insoluble complexes that do not release Cu even under acidic conditions like the abomasal environment [79] . In addition, absorbed thiomolybdates have also been shown to cause systemic effects on Cu metabolism including increased biliary excretion of Cu from liver stores, strong binding of Cu to plasma albumin resulting in reduced availability for biochemical processes, and inhibition of Cu dependent metalloenzymes such as ceruloplasmin, diamine oxidase, cytochrome oxidase, ascorbate oxidase and tyrosine oxidase. [20] [78] [80] . According to [20] , when rumen sulfide concentrations are low, Mo has little effect on the formation of thiomolybdates. However, Cu bioavailability is deeply reduced (up to 70%) when Mo levels are not modified but sulfide concentrations are increased [81] .

In addition to its role in the Cu-Mo-S interaction, organic or inorganic S can also reduce Cu bioavailability per-se [21] . [82] found a 55% reduction in hepatic Cu when sheep were fed high levels of S (2 g/kg DM). According to the authors, this reduction was due to formation of CuS in the digestive tract, since the diet was very low in Mo. Similar results were previously observed by [72] who found a 39% - 56% reduction in Cu bioavailability when S was provided to ewes as methionine or as NaSO₄ under low Mo dietary levels, possibly through the formation of insoluble CuS at sites beyond the rumen. However, [83] postulates that the formation of insoluble CuS and Cu₂S in the rumen is exacerbated by the digestion of insoluble proteins by protozoa, with the consequent increase in available S. Despite the effect of dietary S previously mentioned, other S sources have been also responsible for decreasing Cu bioavailability in ruminants. Molasses, a by-product from the sugarcane and beet industry, is a source of dietary sugars feed to dairy cows [1] . Benefits of adding molasses to diets include increasing palatability, acting as a binder, and reducing dust in fine-particle feeds [84] . However, due to its high content of S, the liberal use of molasses can result in dietary S levels considerably in excess of requirements [85] . [73] evidenced a decrease in liver Cu at 29, 56 or 84 d after feeding heifers with a molasses based supplement. According to the authors, this observation was the result of high concentrations of S naturally found in molasses.

In a review of Cu antagonists in cattle, [86] describes other sources of S implicated in the Cu-Mo-S and Cu-S interactions, included fertilizers, high S water, and S containing supplements. Cows grazing bahiagrass pastures fertilized with ammonium sulfate showed lower liver Cu concentrations compared with cows on non fertilized pastures, or fertilized with ammonium nitrate [86] . A previous study from [87] showed that gypsum fertilization (132 kg S/ha) increased S from 0.33% to 0.40% and from 0.29% to 0.37% of DM in tall fescue grass and in orchadgrass, respectively. However, feeding those pastures to steers resulted in no changes in Cu bioavailability, probably due to the high S content of the non fertilized pastures. For this reason [73] suggest that the choice of fertilizer source can be critical in areas where grazing cattle may be prone to Cu deficiency Sulfur levels in drinking water can also be detrimental for Cu bioavailability. [88] reported a decrease in plasma and hepatic Cu of yearling steers provided with high-S water (3651 mg of SO₄/L) compared with those receiving low-S water (566 mg of SO₄/L). Similarly, a decrease in hepatic Cu of growing steers was found by [89] when S content in the drinking water was increased from 404 to 4654 mg of SO₄/L. While the mentioned S concentrations are in excess of that commonly found in water for animals, high S-containing water has been reported in USA and Canada [90] [91] .

Ruminants consuming forage-based diets are often exposed to high levels of Fe through water, forage, and unusually high amounts of soil ingestion [69] [92] . Supplementing 800 mg of Fe/kg of DM as FeO or FeSO₄, decreased Cu absorption from 0.06 to 0.04 in sheep [93] . In agreement, a previous study of (94) found a rapid decrease in liver and plasma Cu concentrations, activities of erythrocyte superoxide dismutase and plasma ceruloplasmin of young heifers receiving 800 mg of Fe/kg of DM. However, according to [94] the role of Fe on Cu absorption is partially dependent of S. Indeed, [81] suggest that the formation of FeS in the rumen is a critical step for Fe to antagonize Cu absorption. An alternative explanation is provided by [95] who demonstrate that excess Fe can compete with Cu for its absorption at intestinal level, by saturating the DMT-1 Cu transporter.

In a production study conducted on beef steers [96] , the addition of 20 or 40 mg of Cu/kg of DM decreased animal performance, compared with animals receiving a basal diet containing 10.2 mg of Cu/kg of DM, suggesting that high dietary Cu may inhibit ruminal fermentation. Previously, [97] found a decrease in post feeding total VFA concentration and VFA molar proportions of yearling steers receiving a high dose of supplemented Cu (57.3 mg/kg of DM), but average daily gain, feed efficiency, and carcass yield and quality grade were not affected. In agreement, [98] found an in vitro depression in rumen fermentation of concentrates following the addition of high doses of CuSO₄. Similarly, [99] found a reduction in propionate molar proportion when high doses of Cu were added to in vitro rumen incubations. A dose-response study conducted by [100] determined that 21 μg of Cu/mL incubation fluid was required to obtain a 50% inhibition of gas production. However, a large disparity was obtained regarding the susceptibility of the different bacterial populations to Cu. While the growth of Bacteroides succinogenes, Ruminococcus albus and Butyrivibrio fibrisolvens was inhibited by 10, 20 and 30 μg Cu/mL of incubation fluid, respectively, higher concentrations were required to inhibit Megasphaera elsdenii, Selenomonas ruminantium, and Streptococcus bovis (100, 100 and 250 μg Cu/mL, respectively). Opposite results were reported by [101] who found a reduction in rumen pH and an increase in total VFA concentrations when Cashmere wether goats received supplementary Cu in the diet. According to the authors, an increase in NDF digestion may be responsible for these findings. Conversely, later studies conducted by [102] [103] found that NDF digestion was unchanged or maximized by the addition of 10 mg of Cu/kg of DM, but depressed when of 30 mg of Cu/kg of DM were added. In addition, no differences were obtained for rumen pH, IVDMD and VFA molar proportions when in vitro studies were conducted using rumen fluid donors receiving 0, 10 or 20 mg of Cu/kg of DM [104] . Based on previous observations showing that 20 or 40 mg of Cu/kg of DM increased unsaturated fatty acids in adipose tissue of steers [104] , a possible role of Cu as an inhibitor of ruminal lipids biohydrogenation has been suggested. [105] However, no studies have been conducted to assess the validity of this hypothesis.

The effectiveness of organic sources of Cu to promote animal benefits has been a subject of several controversies. [106] reported a higher retention of Cu in steers supplemented with Cu-lysine compared with supplementation with CuSO₄. In another study [107] , the use of organic Cu (as Cu proteinate) increased hepatic retention of Cu in multiparous beef cows, compared with inorganic Cu. However, no benefits on cow and calf performance were obtained. Conversely, [108] found an increase in body weight gain in goat kids supplemented with organic Cu, compared with inorganic Cu. In a recent meta-analysis [13] assessing the benefits of organic trace minerals, only marginal improvements in milk production, milk fat and milk protein were found. In opposition, organic trace minerals did not affect somatic cell count, interval from calving to first service, and 21-d pregnancy rate. Similarly, no differences in 60-d pregnancy rate, health or performance were found by [109] in 2-year old cows receiving Cu as CuSO₄ or as amino acid complex. In addition, [110] found a decrease in pregnancy rates of primiparous cows receiving organic and inorganic minerals (Cu, Co, Mn and Zn) compared with non-supplemented cows. According to the authors, excessive supplementation beyond requirements reduced reproductive performance. In a previous study conducted on steers [71] , growth rate was higher when animals received CuSO₄ than when Cu-lysine was provided during the initial 21 d, but no differences were obtained after 98 d. Other parameters, including feed efficiency, feed intake, humoral and cellular immune response and ceruloplasmin activity, were not affected by Cu source [71] . Organic sources of Cu have been seriously criticized by [8] who consider that technologies for protection against rumen antagonisms are extravagant and provide no additional benefits than conventional CuSO₄. In addition the authors questioned the scientific legitimacy of some in vivo studies where no proper covariance analyses were conducted to account for initial differences between groups of animals [8] . According to [9] commercially driven pursuit of trivial advantages over cheap and effective inorganic sources of Cu should cease, and attention should focused on predicting when supplementation is needed and to the usual problem of over, rather than under-provi- sion of Cu and its environmental impact. In conclusion, Cu, Zn and Mn are required to maintain health and production status of livestock, but their functions at the gastrointestinal tract of ruminants are not totally elucidated. Like many other minerals, Cu, Zn and Mn have the ability to interact with organic compounds of the diet, macro minerals and micro minerals, usually resulting in decreased availability for the host. Different technologies, including proteinates, amino acid chelates, amino acid complex, and polysaccharide complex are currently available for mineral protection However, these technologies appear to be more effective in monogastrics than in ruminants. Mechanisms to guarantee optimal levels of ruminally available minerals, and to optimize mineral supply to the lower gastrointestinal tract without compromising postruminal absorption require further research.