ESTIMATION OF THE RELATIVE BIOAVAILABILITY OF MANGANESE SOURCES FOR SHEEP 1,2 J. Wong-Valle 3, P. R. Henry, C. B. Ammerman and P. V. Rao 4 University of Florida 5, Galnesville 32611 ABSTRACT The relative biological availability of Mn in reagent-grade (RG) Mn sources was tested using 41 Rambouillet crossbred wether lambs in a completely randomized design. Lambs were fed a basal corn-soybean meal-cottonseed hull diet (37.6 ppm Mn, DM basis) or this basal diet supplemented with 0, 1,500, 3,000 or 4,500 ppm Mn from RG MnSO4.H20 or 3,000 ppm Mn from RG, 2 and MnCO3. Feed intake was restricted to 1,000 g/ hd daily during the 21-d experimental period. There was a decrease (P <.01) in daily feed intake by sheep fed 4,500 ppm Mn from MnSO4. Liver, kidney and bone Mn concentrations increased (P <.05) with MnSO4 supplementation. Liver was most responsive to dietary Mn, followed by kidney and bone. Based on multiple linear regression slopes for liver, kidney and bone Mn concentrations, relative bioavailability of Mn from, 2 and MnCO3 averaged 57.7, 32.9 and 27.8%, compared with 100% for MnSO4. (Key Words: Availability, Manganese, Sheep, Tissue Minerals.) J. Anirn. Sci. 1989. 67:2409-2414 Introduction Manganese is distributed widely in roughages, but in some grains, such as corn, its content is low (Hidiroglou, 1979). In addition, availability of Mn from common feeds is largely unknown; phytate and fiber in many feed ingredients can interfere with its utilization (Halpin and Baker, 1986a,b). High dietary Ca and P levels reduced the availability of Mn in cattle (Dyer et al., 1964; Hidiroglou, 1979). Inefficient absorption of dietary bin in ruminants is another factor that makes Mn supplementation necessary (Sansom et al., 1978). 1Florida Agric. Exp, Sta. journal series no. 9450. 2Tbe authors wish to acknowledge Moorman Manufacturing Co., Quincy, IL and Occidental Chemical Co., Taxnpa, FL for funds in support of this research; Pfizer, Inc., New York for supplying vitamins; and Monsanto Chemical Co.~ St. Louis, MO for supplying ethoxyquin. ~ address: Dept. of Poult. Sci., Auburn Univ., Auburn, AL 36849. ~ Dept. of stat. Dept. of Anita, Sci. Received November 28, 1988. Accepted February 13, 1989. Most Mn bioavailability studies have been conducted with poultry fed purified diets to which the element was added at low dietary concentrations (Watson et al., 1970, 1971). Black et al. (1984a) demonstrated the use of short-term, high-level supplementation of practical diets with inorganic Mn sources to determine their relative availability based on the resulting increases in tissue Mn concentrations. Black et al. (1984a) needed few animals to determine true differences among sources due to the linearity of the response variables. The following study was conducted to estimate the relative biological availability of Mn from reagent-grade Mn sources for sheep fed high dietary Mn concentrations, using tissue Mn retention as the response criterion. Matorlals and Methods Forty-one RambouiUet crossbred wether lambs (42 kg BW) were assigned to one of seven experimental diets in a completely randomized design. Prior to the experiment, lambs were group-fed a commercial cornsoybean meal-cottonseed hull diet at approximately 800 g/hd daily and had ad libitum 2409
2410 WONG-VALLE ET AL. TABLE 1. COMPOSITION OF BASAL DIET FED TO LAMBS Item % Ingredient composition a Ground yellow corn 58.5 Coaunseed hull 21.0 Soybean meal, 48.5% CP 12.0 Alfalfa meal 3.0 Corn oil 3.0 Cornstarch b 1.45 Ground limestone.55 Trace mineralized salt r.50 Vitamins A and D d + Ethoxyquin e.0125 Chemical composition f DM, % 88.0 Ca, %.39 P, %.25 Mg, %.14 Mn, ppm 37.6 Fe, ppm 103.8 Zn, ppm 41.9 Cu, ppm 6.9 aas-fed basis. bmanganese supplements were added at the expense of equivalent weights of cornstarch. CContalned in percent: NaCI, 93.0; Zn,.35; Mn,.28; Fe,.175; Cu,.035; I,.007; Co,.007 (Morton Salt Division of Morton Thiokol, Inc., Chicago, IL). dto supply the following per kg of diet: vitamin A palmitate, 2,200 USP units; vitamin D 3, 440 IU. emonsanto Chemical Co., St. Louis, MO. fdm basis except for DM as determined by analysis. access to grass hay. The basal diet was a practical corn-soybean meal-cottonseed hull diet (37.6 ppm Mn, DM basis; Table 1). The diets were formulated to meet requirements for growing Iambs (NRC, 1985) and were supplemented with 0, 1,500, 3,000 or 4,500 ppm Mn from reagent-grade (RG) MnSO4.H20 or 3,000 ppm Mn from RG, 2 and MnCO3, added at the expense of an equivalent weight of cornstarch. The Mn content of the experimental diets was verified by chemical analysis. There were five lambs per treatment for the control and MnSO4 diets, and seven lambs per treatment for the, 2 and MnCO3 diets. There were not enough cages available to have seven lambs on all treatments. Lambs were housed in individual wooden pens with slatted floors in an opensided barn. Feed intake was restricted to 1,000 6Perkin-Elmer Corp., Norwalk, CT. g/hd offered daily (as-fed basis) and fresh tap water was available ad libitum. Lambs were fed the experimental diets for 21 d following a 7-d adjustment period during which all Iambs were fed the basal diet. At the termination of the experiment, lambs were stunned with a captive bolt shot and killed by exsanguination. Right metacarpus, liver and both kidneys were excised and frozen for mineral analyses. Calcium, Mg, Cu, Fe, Zn and Mn in Mn sources and diets and Mn in liver, kidney and bone were determined by flame atomic absorption spectrophotometry on a Perkin-Elmer Model 5000 with AS-50 autosampler 6 (Anonymous, 1982). Standards were matched for macroelements and acid concentrations as needed and standard reference material from the National Bureau of Standards was included with samples. Phosphorus in Mn sources and diets was determined by a colorimetric method (Harris and Popat, 1954). Relative solubility following 1 h of constant stirring at 37"C of.1 g of each source in 100 ml of either 1-120,.4% HC1, 2% citric acid or neutral ammonium citrate (Watson et al., 1970) and magnetic susceptibility (Watson et al., 1971) were determined. Bone, kidney and liver Mn concentrations were analyzed by one-way ANOVA (SAS, 1982) using dietary treatment as the source of variation and animal as the experimental unit. Differences among means were separated by Duncan's (1955) multiple range test. Multiple linear regression was done by least squares using the GLM procedure of SAS (1982). Slope ratios and their SE were estimated using the method of error propagation as described by Kempthorne and Allmaras (1965). Results and Discussion Reagent-grade manganese sources (Tables 2 and 3) contained Mn concentrations of 30.1, 73.0, 50.7 and 43.5% on an as-fed basis for sulfate, oxide, dioxide and carbonate, respectively. All sources were relatively free of mineral contaminants except the dioxide, which contained greater concentrations of Fe, Cu, Zn, Ca and Mg than the other sources. Only the oxide sources had significant magnetic susceptibility (Table 3). The sulfate and monoxide sources had a greater proportion of smaller particles; smaller particle size often results in greater bioavailability. The sulfate source (Table 4) was 100% soluble in water;
BIOAVAILABILITY OF Mn FOR SHEEP 2411 TABLE 2. MINERAL COMPOSITION OF MAGANESE SOURCES Mineral constituents. % Source a Mn b Fr Cu Zn Ca Mg MnSO4 30.1 NIY.0004.003 2 73.0 50.7.001.797.009.018.159.001.123 MnCo3 43.5.003 aall sources were reagent-grade. banalyzcd value following refluxing for 4 h in 1:1 (v:v) HC1 :HNO 3. CNot detectable by analytical procedures used. all other sources were insoluble. All sources except the dioxide were soluble in.4% HCI and 2% citric acid. The low solubility of Mn from the dioxide in.4% HCI and 2% citric acid has been reported previously (Henry et al., 1987). Watson et al. (1971) and Black et al. (1984a) reported complete solubility of RG MnCO3 in.4% HCI. The carbonate was 56% soluble and 2 was insoluble in neutral ammonium citrate; these sources also proved to have the lowest relative bioavailability. Several researchers have reported a correlation between solubility in neutral ammonium citrate and relative biological availability of Mn sources (Watson et al., 1970, 1971; Black et al., 1984a; Henry et al., 1987). Feed intake averaged 994, 1,000, 984, 971, 1,000, 999 and 996 g/d for sheep fed control, MnSO4 (1,500, 3,000 and 4,500 ppm added Mn),, 2 and MnCO3, respectively. Sheep fed 4,500 ppm Mn as MnSO4 had lower (P <.01) daily feed intake than sheep fed other TABLE 3. PHYSICAL PROPF_,RTIF_.S OF MANGANESE SOURCES Particle size, %b -30+ Magnetic Sources a +30 100-100 susceptibility, % MnSO 4 3.5 36.0 60.5 4.8 0.0 35.0 65.0 57.5 2 1.6 80.0 18.4 52.3 MnCO 3.6 84.4 15.0 11.2 aall sources were reagent-grade. bretained by a No. 30 sieve (U.S. Bureau of Standards), passing a No. 30 but retained on a No. 100, and passing a No. I00 sieve, respectively. dietary treatments. Cunningham et al. (1966) reported lower feed consumption by calves fed 2,460 or 4,920 ppm Mn as MnSO4 for 84 d compared with calves fed 0 or 820 ppm Mn. Black et al. (1985a) reported that sheep fed 2,000 ppm Mn from feed-grade or 8,000 ppm Mn from RG MnCO 3 for 84 d had lower feed intake than sheep fed lower levels of these sources. In another study, Black et al. (1985b) reported reduced feed intake when sheep were fed 3,000 or 6,000 ppm Mn as for 3 wk but not for 1 or 2 wk compared with sheep fed an unsupplemented diet. Sheep fed 9,000 ppm Mn had reduced intakes during the three 1-wk periods compared with sheep fed the control diet; only 80% of their initial feed intake was consumed when sheep were refed the basal diet for 1 wk. There was a linear increase in bone Mn concentration (P <.01) as dietary levels of MnSO4 increased up to 3,000 ppm Mn (Table 5); however, percentage of bone ash did not change (P >.10) and averaged 66%. Supplemental Mn as MnSO4 at 4,500 ppm resulted in no further increase in Mn concentration of any of the tissues analyzed; therefore, this level was not included in any of the regression TABLE 4. RELATIVE SOLUBILITY OF MN SOURCES Relative solubility, %a Neutral.4% 2% ammonium Source b H20 HC1 Citric acid citrate MnS04 100 100 100 100.14 100 100 100 2.31.61 5.5.61 MnCO 3.71 100 100 56,3 ]m ]]11 ]] afrom 1 h constant stirring at 37'C. ball sources were reagent-grade.
2412 WONG-VALLE ET AL. TABLE 5. EFFECTS OF SOURCE A LEVEL OF DIETARY MANGANESE ON BONE, KIDNEY A LIVER MANGANESE CONCENTRATIONS IN LAMBS a Added Mn, Bone Mn, Kidney Mn, Liver Mn, Source ppm b ppm, ash basis ppm, DM basis ppm, DM basis Control 0.4 c 4.0 f 9.6 g MnSO 4 1,500 2.4 d 19.8 cd 35.3 d MnSO 4 3,000 5.3 c 26.2 c 43.7 c MnSO 4 4,500 5.3 c 26.7 c 43.8 c 3,000 3.1 d 17.8 de 33.3 de 2 3,000 2.2 d 11.2 ef 24.3 f MnCO 3 3,000 1.4 de 10.2 f 26.3 ef Pooled SE.21.91.96 aeach value represents the mean of five lambs/treatment for the conffoi and sulfate, and seven iambs/treatment for the oxide, dioxide and carbonate sources. All sources were reagent-grade. bbasal diet contained 37.6 ppm Mn, DM basis. c'd'e't"gmgans in the same column with different letters in their superscripts differ (P <.05). analyses. Manganese concentrations in bone had been shown previously to increase with increasing dietary Mn levels (Watson et al., 1973; Black et al., 1985a,c). These researchers reported higher bone Mn concentrations than those reported in this study; however, in our study, sheep were fed the supplemental Mn diets for only 21 d, compared with 81, 84 and 42 d in the previous long-term toxicosis studies. Kidney Mn concentrations were affected (P <.05) by dietary treatment. At 3,000 ppm added Mn, MnSO4 resulted in greater (P <.05) kidney Mn concentrations than the other Mn sources (Table 5). Manganese in liver was increased (P <.05) from 9.6 ppm (DM basis) in lambs fed the control diet to 43.7 ppm when 3,000 ppm Mn from MnSO4 was fed. Lambs fed MnSO4 at 3,000 ppm had greater liver Mn concentrations than lambs fed other Mn sources (Table 5). Other researchers also have reported increased kidney and liver Mn concentrations when high dietary levels were fed (Watson et al., 1973; Ivan and Hidiroglou, 1980; Black et al., 1985a,c). Multiple linear regression analysis of tissue Mn uptake with respect to dietary Mn level was determined for all tissues (Table 6). Liver provided the best fit of data to the linear model followed by kidney and bone (R 2 =.70,.62 and.59, respectively). In an earlier study with sheep fed high graded levels of or MnCO3, liver had the greatest slopes; however, in that study liver resulted in a poorer fit of data to the linear model than kidney or bone did (R 2 --.72,.68 and.49, respectively; Black et al., 1985a). Studies with day-old chicks showed that bone was more sensitive to changes in dietary Mn than was kidney or liver (Black et al., 1984a, b; Henry et al., 1986). In the present study, differences in rank order of the different tissues may be explained by the older age and previous nutritional experience (unknown prior to purchase) of the sheep or by differences in the sources tested. Estimates of relative biological availability (Table 7); were obtained by a ratio of the slopes in equations in Table 6. Setting MnSO4 at 100% resulted in values of 57.3, 38.9 and 22.8% for bone Mn concentrations and 54.7, TABLE 6. MULTIPLE LINEAR REGRESSION OF TISSUE MANGANESE CONCENTRATION ON DIETARY CONCENTRATION OF SOURCES Tissue Regression equation a R 2 SD P Bone Y =.269 +.001619x t +.000929x 2.00063 lx 3 +.000370x4.59 1.25 <.001 Kidney Y = 5.59 +.00741x 1 +.00406x 2 +.00186x 3 +.00153x 4.62 5.56 <.001 Liver Y = 12.4 +.01139x 1 +.00695x 2 +.00395x 3 +.00460x4.70 6.77 <.001 awhere Y = tissue Mn, ppm DM basis, except bone, which is on an ash basis and expressed in ppm added Mn, x 1 is MnSO 4, x 2 is, x 3 is 2 and x 4 is MnCO 3. Each equation represents samples from 36 sheep. MnSO4 at 4,500 ppm added Mn was not included in the analysis.
BIOAVAILABILITY OF Mn FOR SHEEP 2413 TABLE 7. RELATIVE BIOLOGICAL AVAILABILITY OF MANGANESE SOURCES BASED ON MULTIPLE LINEAR REGRESSION OF BONE, KIDNEY A LIVER MANGANESE CONCENTRATIONS Tissue and Multiple regression Relative source of Mn a slope SE value SE Bone MnSO4 2 MnC% Kidney MnSO4 2 MnCO3 Liver MnSO4 2 MnCO3.001619.0003 b 100.000929 +.0002 c 57.3 7.1.000631.0003 cd 38.9 13.5.000370.0002 d 22.8 11.7.00741.001 b 100.00406 +.008 c 54.7 + 7.5.00186 +.008 d 25.1 + 8.4.00153 +.001 d 20.1 + 11.5.01139 +.001 b 100.00695 +.001 c 61.1 6.0.00395 +.001 d 34.7 + 6.2.00460 +.001 c'd 40.4 + 8.5 aall sources were reagent-grade. b'c'dsiopes within a tissue with different letters in their superscripts differ (P <.05). 25.1 and 20.1% for kidney Mn concentrations for, 2 and MnCO 3, respectively. Values of 100, 61.1, 34.7 and 40.4, respectively, were obtained for liver. A relative biological availability index, calculated using the mean of the estimates from the different tissues, provided values of 57.7, 32.9 and 27.8% for RG, 2 and MnCO3, respectively. In the present experiment, slopes for lambs fed MnCO3 averaged 48% that of lambs fed. Similar calculations based on liver, kidney and bone from Black et al. (1985a) resulted in a value of 78% for RG carbonate compared with a feed-grade oxide, which would be expected to be less available than a RG oxide. Reagent-grade sources generally have fewer contaminants that could lower bioavailability. Availability of Mn from RG 2 was 40% that of for chicks (Henry et al., 1987) but was about 57% as available for sheep in the present experiment (32.9/57.7). Henry et al. (1987) and Southern and Baker (1983) reported that 2 was 29% as available as MnSO4 for chicks, which is similar to the 32.9% for lambs in the present study. Based on bone and liver Mn concentrations of chicks fed up to 4,000 ppm Mn, Black et al. (1984a) reported average availability for RG and RG MnCO3 of 70 and 39%, respectively, compared with 100% for RG MnSO4. Thus, the young chick may be able to utilize some forms of Mn to a greater extent than ruminants. In conclusion, tissue uptake of Mn increased with increasing levels of dietary Mn up to 3,000 ppm Mn. Liver Mn changed most in response to elevated dietary Mn levels, followed by kidney and bone. The MnSO4.H20 was most available and MnCO3 least available for all sources tested. The results of and 2 varied according to the criteria used. Relative bioavailability of the sources was related to solubility in neutral ammonium citrate. Liver, kidney and bone Mn uptake were sensitive and quantitative criteria for estimating relative biological availability of Mn sources for sheep. Literature Cited Anonymous. 1982. Analytical Methods for Atomic Absorption Spectrophotometry. The Perkin-Elmer Corp., Norwalk, CT. Black, J. R., C. B. Ammerman, P. R. Henry. 1985a. Effect of high dietary manganese as manganese oxide or manganese carbonate in sheep. L Anim. Sci. 60:861. Black, J. R., C. B. Ammerman and P. R. Henry. 1985b. Effect of quantity and route of administration of manganese monoxide on feed intake and serum manganese in ruminants. J. Dairy Sci. 68:443. Black, J. R., C. B. Ammerman. P. R. Henry and R. C. Littel. 1985c. Influence of dietary manganese on tissue trace mineral accumulation and depletion in sheep. Can. J. Anita. Sci. 65:653. Black, J. R., C. B. Ammerman, P. R. Henry and R. D. Miles. 1984a. Biological availability of Mn sources and effects of high dietary manganese on tissue mineral composition of broiler-type chicks. Poult. Sci. 63: 1999. Black, J. R., C. B. Ammerman, P. R. Henry and R. D. Miles. 1984b. Tissue manganese uptake as a measure of manganese bioavailability. Nutr. Rep. Int. 29:807. Cunningham, G. N., M. B. Wise and E. R. Barrick. 1966. Effect of high dietary levels of manganese on the performance of blood constituents of calves. J. Anim. Sci. 25:532. Duncan, D. B. 1955. Multiple range and multiple F tests. Biometrics 11:1. Dyer, I. A., W. A. Cassatt, Jr. and R. R. Rat. 1964. Manganese deficiency in the etiology of deformed calves. Bioscience 14:31. Halpin, K. M. and D. H. Baker. 1986a. Manganese utilization in the chick: Effects of corn, soybean meal, fish meal, wheat bran, and rice bran on tissue uptake of manganese. PouR. Sci. 65:995. Halpin, K. M. and D. H. Baker. 1986b. Long-term effect of corn, soybean meal, wheat bran, and fish meal on manganese utilization in the chick. Poult. Sci. 65:1371. Harris, W. B. and P. Popat. 1954. Determination of the phosphorus content of lipids. J. Am. Oil Chem. Soc. 31:124. Henry, P. R., C. B. Ammcrman and R. D. Miles. 1986.
2414 WONG-VALLE El" AL. Bioavailability of manganese sulfate and manganese monoxide in chicks as measured by tissue uptake of manganese from conventional dietary levels. Poult. Sci. 65:983. Henry, P. R., C. B. Ammerman and R. D. Miles. 1987. Bioavallability of manganese monoxide and manganese dioxide for broiler chicks. Nutr. Rep. Int. 36:425. Hidiroglou, M. 1979. Manganese in ruminant nutrition. Can. J. Anita. Sci. 59:217. ivan, M. and M. Hidiroglou. 1980. Effect of dietary manganese on growth and manganese metabolism in sheep. J. Dairy Sci. 63:385. Kempthome, O. and R. R. Allmaras. 1965. Errors of observations. In: C. A. Black (Ed.) Methods of Soil Analysis. Agronomy 9:1-23. Am. Sec. of Agron., Madison, Wl. NRC. 1985. Nutrient Requirements of Sheep (6th Rev. Ed.). National Academy Press, Washington, DC. Sansom, B. F., H. W. Symonds and M. J. Vagg. 1978. The absorption of dietary manganese by dairy cows. Res. Vet. Sci. 24:366. SAS. 1982. SAS User's Guide: Statistics. SAS Inst., Inc., Cary, NC. Southern, L. U and D. H. Baker. 1983. Excess manganese ingestion in the chick. Ponlt. Sci. 62:642. Watson, L. T., C. B. Ammerman, J. P. Feaster and C. E. Roessler. 1973. Influence of manganese intake on metabolism of manganese and other minerals in sheep. J. Anita. Sci. 36:131. Watson, L. T., C. B. Ammerman, S. M. Miller and R. H. Harms. 1970. Biological assay of inorganic manganese for chicks. Poult. Sci. 49:1548. Watson, L. T., C. B. Ammerman, S. M. Miller and R. H. Harms. 1971. Biological availability to chicks of manganese from different inorganic sources. Poult. Sci. 50:1693.