Analyst, April, 1979, Vol. 104, $9. 313-322 313 Minimum Sample Preparation for the Determination of Ten Elements in Pig Faeces and Feeds by Atomic-absorption Spectrophotometry and a Spectrophotometric Procedure for Total Phosphorus Edward P. Hilliard" and J. David Smith School of Agriculture and Forestry, University of Melbourne, ParRviZle, Victoria 3052, Australia School of Chemistry, University of Melbourne, Parkville, Victoria 3052, Australia Studies of mineral metabolism in pigs and problems of manure disposal or utilisation are complicated by interactions of trace metals and major cations. A procedure for the determination of copper, zinc, cadmium, lead, iron, sodium, potassium, magnesium, calcium, phosphorus and arsenic in pig faeces and feeds is described. Phosphorus is determined spectrophoto- metrically and the other elements by atomic-absorption spectrophotometry.Sample preparation is minimised, and all elements except arsenic are deter- mined after a single sample digestion in nitric acid - perchloric acid mixture. A separate sample digestion is necessary for arsenic. The accuracy and precision of the method were rigorously tested, and are suitable for budget studies of all eleven elements. Keywords : Pig faeces and feed analysis ; trace metal determination ; major element determination ; atomic-absorption spectrophotowtetry ; spectro- photometry Intensification in animal production in recent years has generated problems in relation to the disposal or utilisation of the large amounts of manure produced per annum. Manures from pig, poultry or dairy/beef units have traditionally been spread on land to utilise their fertiliser values, With the trend towards larger intensive units, and a decrease in land area suitable for manure spreading from such units, attention is being given to alternative means of manure disposal.The environmental effects of disposing of large amounts of manure to small land areas is causing concern, particularly where the manures contain high concentrations of metals.lY2 Commercially formulated pig diets generally contain supple- mental mineral salts to improve growth and development and to compensate for deficiencies due to interactions between some elements. Minerals at high levels in the feed are not fully absorbed by the pig and can be concentrated by a factor of four in the faeces.2 Because of the unusually high levels of minerals in pig faeces and feeds, these materials provide an unusual matrix when the accurate determination of element concentrations is required.Investigations into the fate of minerals or trace elements in biological systems are compli- cated by the interactions of some elements with each other, and with other metabolites in the system.3~~ For this reason, it is rare to encounter experimental work on mineral meta- bolism that reports the functions and fate of single elements without examining the influence of potential interacting elements within the system under investigation. Hence, it is essential that laboratories use methods that permit determination of the maximum number of elements but with the minimum of sample preparation.Our investigations were principally aimed at the development of economical methods for the analysis of pig manures and feeds for a wide range of elements, giving results of high accuracy and precision. To minimise sample manipulation, calibration over a wide range was used and linear or quadratic equations were fitted to the calibration lines. Concentrations of elements in the * Present address : The Agricultural Institute, Dunsinea, Castleknock, Co. Dublin, Ireland.314 HILLIARD AND SMITH: MINIMUM SAMPLE PREPARATION FOR THE Analyst, vol. 104 sample solutions were computed using these equations. was used for the more sensitive and abundant elements. Some degree of burner rotation Experimental Apparatus Atomic-absorption spectrophotometer. A Varian Techtron, Model 1200, instrument with an air - acetylene burner and digital read-out was used for the determination of copper, zinc, cadmium, lead, iron, sodium, potassium, magnesium and calcium.A Varian Techtron, Model AA5, instrument with a BC-6 simultaneous background corrector and a Mace 1100 strip-chart recorder was used for the determination of arsenic. A Varian Techtron, Model 64, hydride-generation kit was used with a nitrogen - hydrogen flame using the burner supplied for air - propane. Spectrophotometer. A Unicam SP600 instrument with a I-cm light path flow-through cell attachment was used for the determination of phosphorus. Computation. A Sharp, Model 365P, programmable calculator was used to fit linear calibration lines to the equation y = a + bx and quadratic curvilinear calibration lines to the equation y = a + bx + ex2. Kjeldahl digestionjasks.Flasks of capacity 50 ml on an electric heat bank with individual heat controls were used. Glass culture tubes with PTFE seal screwcaps. Capacity 25 ml. Reagents All solutions were prepared using glass-distilled water in glassware that had previously been washed with 10% nitric acid and rinsed with distilled water. Concentrated acids from Ajax Chemicals Pty. Ltd. were selected batches of low trace-element content. Unless specified otherwise, all chemicals used were of analytical-reagent grade. Nitric acid, sp. gr. 1.42. Dilute nitric acid, 2 M. Perchloric acid, 70%, sp. gr. 1.54. Sulphuric acid, sj5. gr. 1.84. Hydrochloric acid, sp.gr. 1.18. Nitric acid - perchloric acid digestion mixture. Mix equal volumes of the concentrated acids. Sulphuric acid - perchloric acid - sodium molybdate digestion mixture. Dissolve 2 g of sodium molybdate (Na2Mo0,.2H20) in 40 ml of water, add 50 ml of concentrated sulphuric acid, allow the mixture to cool and add 10 ml of concentrated perchloric acid. Ammonia solution, sp. gr. 0.880. Bromophenol blue indicator solution, 0.1% mlV in methanol. Citrate bufer. Dissolve 210 g of citric acid and 29.5 g of trisodium citrate in water and dilute to lo00 ml with water. Ammonium tetramethylenedithiocarbamate solution, 0.5% mlV in water. Prepare fresh daily. 4-Methylpentan-2-one. Sodium tetralzydroborate(III) solution, 5% mlV in 0.1 yo sodium hydroxide solution. Use within 14 h of preparation.Lanthanum - caesium solution, 5% m/V lanthanum and 2% mlV caesium in water. Dissolve 2.5 g of caesium chloride in 100 ml of water and add 100 ml of 10% m/V lanthanum solution (BDH standard solution for spectroscopy) . Vanadate reagent for phosphate determination. Solution A : dissolve 25 g of ammonium molybdate in 400 ml of water. Solution B: dissolve 1.25 g of ammonium metavanadate in 300 ml of boiling water, cool and add 200 ml of concentrated perchloric acid. Add solution A to solution B and dilute to lo00 ml with water. Arsenic standard solutions. Dissolve 0.832 9 g of sodium arsenate (Na,HAs0,.7H20) in water and dilute to 1000 ml with water. Dilute 3 ml of this solution to 200 ml with water to give a 3 pg ml-l arsenic stock standard solution.Prepare arsenic standard solutions in the range 0.03-0.36 pg ml-l by diluting appropriate volumes of the stock standard solution with 5% sulphuric acid. Dilute 127 ml of the concentrated acid to lo00 ml with water. Purified by extraction with 2 M nitric acid.April, 1979 315 BDH cadmium standard solution for atomic-absorption spectrophotometry, containing 1000 pg ml-l of cadmium, diluted with 2 M nitric acid to give standard solutions in the range 0.05-1.50 pg ml-l of cadmium. These solutions should be saturated with 4-methylpentan-2-one prior to use. BDH lead standard solution for atomic-absorption spectro- photometry, containing 1000 pg ml-l of lead, diluted with 2 M nitric acid to give standard solutions in the range 0.5-15.0pgml-1 of lead.These solutions should be saturated with 4-methylpentan-2-one prior to use. The following stock solutions are prepared for inclusion in the multi-element stock standard solution : 1000 pg ml-l solutions of copper, zinc and iron ; available as BDH standard solutions for atomic-absorption spectrophotometry. 10000 pg ml-1 of calcium; 6.2425 g of calcium carbonate dissolved in 120 ml of water plus 15 ml of concentrated hydrochloric acid. 10000 pg ml-l of magnesium; 20.270 g of magnesium sulphate (MgS0,.7H20) dissolved in 150 ml of water. Add 2 ml of concentrated hydrochloric acid and dilute to 200 ml with water. D. 1000Opg~ml-l of sodium; 6.355g of sodium chloride dissolved in and diluted to 250 ml with water. E. 100000 pg ml-1 of potassium; 19.08 g of potassium chloride.Add 1 ml of con- centrated hydrochloric acid and dilute to 100 ml with water. F. 10000 pg ml-l of phosphorus; 10.6603 g of ammonium orthophosphate [(NH,),HPO,] dissolved in and diluted to 250 ml with water. Appropriate volumes of each stock solution are added to a 1000-ml calibrated flask and diluted to volume with water to give a multi-element stock standard solution containing each element at the following concentrations (pg ml-l) : DETERMINATION OF TEN ELEMENTS IN PIG FAECES AND FEEDS BY AAS Cadmium standard solutions. Lead standard solutions. Multi-element stock standard solution. A. B. C. Dilute to 250 ml with water. cu Zn Fe Na K Mg Cu P 20 60 100 2000 1000 1000 2000 2000 Procedure for Measuring Accuracy and Precision of Analytical Techniques through a 1-mm screen of a stainless-steel laboratory mill.were bulked to provide a large representative sample of pig faeces. precision of the techniques were assessed as follows. Samples of pig faeces were collected from 24 commercial piggeries, freeze-dried and ground Sub-samples from each sample The accuracy and The over-all precision of the techniques was determined by analysis of at least six replicate 2-g samples of dried, ground pig faeces. To determine the recovery of each element and assess losses due to volatilisation or precipitation during digestion, known amounts of each element were added to triplicate 2-g samples of the bulk faeces sample before and after digestion. These standard additions were made by adding 10 ml of the multi-element stock standard solution. The arsenic recovery was determined separately by adding 1 ml of the arsenic stock standard solution to triplicate 2-g samples.To establish if sample matrix effects such as ionisation or chemical interference affected the accurate determination of each element, the following procedure was adopted. Prior to digestion, 10- and 15-ml aliquots of the multi-element stock standard solution were added to 1.0-, 1.5- and 2.0-g masses of faeces in triplicate. Effects on arsenic determination were examined separately by adding 1 and 2 rnl of the arsenic standard solution to triplicate 1- and 2-g masses of faeces before digestion. The accuracy of the methods was assessed by analysing the US National Bureau of Standards Reference Material 1571, “orchard leaves.’’ Duplicate 2-g samples were digested and analysed by the method described below.1. 2. 3. 4. Digestion Procedure several ways. The preparation of biological material for multi-element analysis can be accomplished in Dry ashing, which requires heating the sample to temperatures above31 6 HILLIARD AND SMITH: MINIMUM SAMPLE PREPARATION FOR THE Analyst, VOl. 104 400 "C, was considered to be unsuitable because of the potential losses of lead, cadmium and arsenic due to volatilisation at the ashing temperatures5 Wet digestion, by heating with acids, was considered to be a more suitable means of destroying organic matter. The choice oi reagents for wet digestion was made after considering the solubility of salts formed during digestion. Sulphuric acid was not used when the likely losses of calcium and lead as insoluble sulphates would be significant.A mixture of nitric and perchloric acids (1 + 1) was selected as the digestion mixture for copper, zinc, cadmium, lead, iron, sodium, potassium, magnesium, calcium and phosphorus. An oxidising solution of sodium molybdate in sulphuric - perchloric acids was used for arsenic samples. For determination of copper, zinc, cadmium, lead, iron, sodium, potassium, magnesium, calcium and phosphorus Approximately 2 g of sample were weighed accurately into Kjeldahl flasks and digested initially with 15 ml of concentrated nitric acid (sp. gr. 1.42) and finally with 15 ml of the nitric - perchloric acid digestion mixture. Appropriate additions of the multi-element standard solution were made before or after digestion as required.The digest solutions were then transferred quantitatively into 100-ml calibrated flasks and diluted to the mark with water. For detmnination of arsenic Approximately 2 g of sample were weighed accurately into Kjeldahl flasks and digested initially with 10 ml of concentrated nitric acid (sp. gr. 1.42) and finally with 15 ml of the sulphuric acid - perchloric acid - molybdate digestion mixture. Standard additions of arsenic were made as required. The digest solutions were then transferred quantitatively into 50-ml calibrated flasks and diluted to the mark with water. Determination of Elements The instrument settings and flame conditions suitable for each element are shown in Table I. A 25-ml aliquot of digest solution was removed from each 100-ml calibrated flask TABLE I INSTRUMENT SETTINGS AND FLAMES USED FOR ATOMIC-ABSORPTION SPECTROPHOTOMETRY Element f A 7 Condition Cu Zn Cd Pb Fe Na K Mg Ca As Wavelength (nm) .. 324.7 213.9 228.8 217.0 372.0 330.2 404.4 285.2 422.7 193.7 Slit width (mm) . . . . 0.2 0.2 0.5 1.0 0.5 1.0 1.0 0.5 0.2 0.25 L Y J Flame . . .. .. Air - acetylene H2 - N, for the determination of copper, zinc, iron, sodium, potassium, magnesium, calcium and phosphorus. The digest solution remaining in the flask (approximately 75 ml) was retained for the determination of cadmium and lead. 'The multi-element stock standard solution was diluted with 4% perchloric acid to give standards within the ranges shown in Table 11. TABLE I1 ELEMENT CONCENTRATIONS I N THE MULTI-ELEMENT STANDARD SOLUTIONS Element cu Zn Cd Pb Fe Na K Ca P Mg Lowest standard/ Highest standard/ Stock standard/ pg ml-1 pg ml-l pg ml-l 0.4 16 20 1.2 48 60 0.004 0.16 0.2 0.04 1.6 2 2 80 100 40 1600 2 000 20 800 1000 20 800 1000 40 1600 2 000 40 1600 2 000April, 1979 DETERMINATION OF TEN ELEMENTS IN PIG FAECES AND FEEDS BY AAS Copper, zinc and iron Copper and iron are generally free from interference during atomic-absorption spectro- photometry, so the digests and standards were nebulised directly into the flame.The copper calibration graph was linear over the complete range and had a correlation coefficient (YJ of 0.999. Copper concentrations in the sample digests were computed from a linear regression equation. The iron calibration line was curved above 30 pg ml-l and concentra- tions of iron in sample solutions were computed from a quadratic equation fitting the calibration line. Because non-atomic absorption from sample solutions was considered possible at the zinc wavelength, the absorbance was also measured separately at 213.9nm using a hydrogen lamp.No significant absorbance was observed using the hydrogen lamp with the necessary 80" burner rotation. The zinc calibration line was linear up to 24 pg ml-l (yXy = 0.998) and was used up to 60 pg ml-l by fitting a quadratic equation to the curve. 31 7 Cadmium and lead Levels of cadmium and lead in faeces and feeds were found to be too low for direct deter- mination on the digest solution. A procedure for the concentration of these elements by solvent extraction was developed utilising ammonium tetramethylenedithiocarbamate and 4-methylpentan-2-one.Optimum conditions for the extraction of trace amounts of cadmium and lead from the digest solution were determined experimentally. The efficiency of the extraction procedure was determined as follows. Appropriate volumes of the multi-element stock standard solution were diluted to 100ml with 4% perchloric acid to give solutions containing 0.004-0.100 pg ml-l of cadmium and 0.04-1.0 pg ml-l of lead in a multi-element matrix similar to the sample digest. A blank of 100 ml of perchloric acid was also prepared. A 25-ml volume was removed by pipette from each flask, leaving 75 ml of solution, similar to the sample digest volume available for analysis. Bromophenol blue indicator was added to each flask and ammonia solution (sp.gr. 0.880) added dropwise until the end-point was reached. A 5-ml volume of 1 M citrate buffer (pH 4.0) was added, followed by 3 ml of 0.5% ammonium tetramethylenedithiocarbamate solution. The contents of each flask were mixed, 10 ml of 4-methylpentan-2-one added and the contents mixed again. The organic layer was allowed to separate, then transferred by Pasteur pipette into a 25-ml culture tube. The aqueous phase was extracted with a further 5 ml of 4-methylpentan-2-one, which was also added to the culture tube. As the cadmium and lead complexes are not stable for more than about 2 h,6 cadmium and lead were back-extracted from the 4-methylpentan-2-one into 5 ml of 2 M nitric acid. This procedure eliminated the necessity of preparing standards in 4-methylpentan-2-one.The upper organic layer was removed from the culture tube, leaving the cadmium and lead in 2 M nitric acid. Complete extraction of cadmium and lead from the standards would yield solutions containing 0.06-1.50 pg ml-l of cadmium and 0.6-15.0pgml-l of lead. Standard solutions of cadmium and lead covering similar concentration ranges were prepared in 2 M nitric acid saturated with 4-methylpentan-2-one to simulate the extracted standards. Absorbances of the extracted and non-extracted standards were measured using the conditions described in Table I. The instrument was adjusted to zero with distilled water and the absorbance of a 2 M nitric acid solution saturated with 4-methylpentan-2-one subtracted from all readings. The absorbance of extracted and non-extracted cadmium and lead standards showed that the recovery of cadmium and lead through the extraction procedure was essentially quantitative (Table 111).This extraction procedure was applied to test faeces samples digests. Magnesium and calcium To overcome potential problems due to ionisation or chemical interferences in the deter- mination of magnesium and calcium, various combinations of releasing agents and ionisation suppressants were examined under different flame conditions. The most suitable conditions were as follows. Sample and standard solutions were diluted 1 + 49 by adding 2 ml of a solution of 5% lanthanum - 2y0 caesium to 0.2 ml of sample or standard and diluting to 10 ml with water. These solutions were aspirated into the instrument using the settings shown in Table I and gave curved calibration lines over the standard ranges used.31 8 HILLIARD AND SMITH: MINIMUM SAMPLE PREPARATION FOR THE Artalyst, VoZ.104 TABLE I11 COMPARISON OF ABSORBANCES OF EXTRACTED AND NON-EXTRACTED CADMIUM AND LEAD STANDARD SOLUTIONS, SHOWING RECOVERY BY THE SOLVENT-EXTRACTION PROCEDURE Standard concentration/ pg ml-l 0.06 0.15 0.30 0.75 1.50 Cadmium standards A Absorb an c e - Non- Extracted extracted 0.010 0.009 0.024 0.023 0.049 0.047 0.120 0.117 0.221 0.229 Lead standards Absorbance - I A \ Standard Recovery, concentration/ Non- Recovery, 0.6 0.009 0.010 90 % 111 104 1.5 0.019 0.020 95 104 3.0 0.038 0.036 106 103 7.5 0.095 0.092 103 97 15.0 0.186 0.180 103 pg ml-l Extracted extracted % Sodium and potassium Sodium and potassium can form refractory compounds during digestion and may also ionise in an air - acetylene flame.To overcome these effects, 4 ml of sample or standard solutions were added to 1 ml of a solution containing 5% lanthanum - 2% caesium before aspirating into the instrument using the settings shown in Table I. The calibration lines were curved over the standard range used. Phosphorus The molybdophosphovanadate yellow colour method described by Jackson' was examined for its applicability to the determination of phosphorus levels in pig faeces digest solutions. A 1.0-ml volume of sample or standard solution was mixed with 10 ml of molybdovanadate reagent, then diluted with water to 50ml in a calibrated flask, and the colour was allowed to develop for at least 10min.Using a wavelength of 440nm and a l-cm light path, a linear calibration graph was obtained for standards in the range 2-15 pg ml-1 of phosphorus. Arsenic Arsenic can be determined in biological materials by the generation of arsine and measure- ment by atomic-absorption spectrophotometry,8 or colorimetrically using silver diethyldithio- arb am ate.^ The atomic-absorption method was chosen because it has the greater sensitivity needed for the materials being studied. Preliminary experiments using digestion with nitric - perchloric acids followed by arsine generation using the method described by Duncan and Parker* showed that variable losses of up to 50% occurred during digestion. Double peaks of arsenic absorbance were recorded, indicating that arsine was being formed at two different reaction rates.A similar occurrence was observed by Aggett and Aspell,lo who used the differences in reaction rates to separate arsenic(II1) from arsenic(V) in sample digests. Another digestion technique was sought that would not result in losses of arsenic and that would maintain arsenic in a single valency state for hydride generation. Simon et aZ.11 showed that loss of arsenic as the volatile arsenic(II1) chloride during acid digestion procedures can be prevented by maintaining strong oxidising conditions by inclusion of sodium molybdate (Na,MoO,) in the digestion mixture. Comparison of the recoveries of trivalent and pentavalent arsenic added to samples before and after digestion showed that under their digestion conditions arsenic was converted into and maintained in the non-volatile pentavalent form, and using a coulometric technique recoveries of approximately 94% were obtained. This procedure was adapted for prevention of losses during digestion and double-peak formation with arsine generation.Test faeces samples were digested for arsenic according to the procedure described earlier. Arsenic standard solutions in the range 0.03-0.36pgml-l in 5% sulphuric acid were pre- pared from the 3 pg ml-l arsenic stock standard solution. A 2-ml volume of sample or standard solution was placed in the reaction vessel with 2.5 ml of concentrated hydrochloric acid. Arsine was generated by the addition of 5 ml of 5% sodium tetrahydroborate(II1) in 0.1 yo sodium hydroxide solution and measured using the instrumental conditions shown in Table I.The calibration line of arsenic concentration veisZts peak height was curved overApril, 1979 DETERMINATION OF TEN ELEMENTS IN PIG FAECES AND FEEDS BY AAS 319 the standard range and was fitted by a quadratic equation. No significant losses of arsenic occurred during digestion of samples (Table V), and arsine was generated at one reaction rate, as indicated by the absence of double peaks during measurement. Results and Discussion The over-all precision of the procedure is summarised in Table IV and shows that coefficients of variation of better than 5% were obtained for all elements except lead and arsenic. The higher variability associated with the determination of these two elements is consistent with their presence at relatively low concentrations in the sample.TABLE IV RESULTS OF REPLICATE ANALYSIS OF 2-g SAMPLES OF BULK PIG FAECES USING THE RECOMMENDED PROCEDURE Element Parameter Cu Zn Cd Pb Fe Na K Mg Ca P As Number of replicates . . . . . . 6 6 12 12 6 6 6 6 6 6 6 Mean concentration (dry matterbasis) . . .. .. 290 511 0.86 8.96 1740 0.32% 0.83% 0.82% 3.88% 2.45% 2.0 Standard deviation . . . . . . 10.5 23.8 0.04 0.84 40 0.01% 0.03% 0.02% 0.09% 0.06% 0.27 Coefficient of variation, % . . .. 3.6 4.7 4.8 9.4 2.3 3.3 3.8 2.1 2.4 2.3 13.5 I.Lg g-1 vg g-1 wg g-1 I.Lg g-1 Yg g-' Ilg g-' Ilgg-1 vg g-l w g - l w g - ' vgg-' wg g-1 The mean recovery of standard additions of all elements made before and after digestion are summarised in Table V.Application of Student's t-test indicated no significant losses of any element during the digestion procedure. This is important as the literature frequently reports losses of cadmium, lead and arsenic in procedures for determining these elements in biological material. TABLE V STATISTICAL COMPARISON OF RECOVERIES OF ELEMENTS ADDED TO 2 g OF PIG FAECES BEFORE AND AFTER DIGESTION Number of Element replicates Amount added cu 3 200 Pg Zn 3 600 CLg Cd 4 2 CLg Pb 4 20 Fe 3 1000 pg Na 4 20 mg K 4 10 mg 3 10 mg 3 20 mg Mg Ca P 3 20 mg As 3 3 * N.S. indicates not significant. Mean recovery, % Pre-digestion Pos t-diges tion addition addition 105.8 f 1.5 97.5 f 9.8 98.6 f 3.6 95.1 f 6.8 91.9 f 4.4 87.2 f 4.0 93.0 f 17.8 97.9 f 6.5 93.2 &- 10.1 101.5 & 11.8 101.9 f 9.6 105.2 f 9.0 100.1 f 7.1 93.5 f 9.6 98.5 f 12.3 90.3 f 4.4 91.5 f 9.0 91.1 f 8.5 110.5 f 9.0 f A \ 101.7 f 12.4 100.7 & 2.6 101.9 f 4.4 Student's t-test (P <0.05) N.S.* N.S.N.S. N.S. N.S. N.S. N.S. N.S. N.S. N.S. N.S. The mean recoveries of all elements added to increasing masses of sample are summarised in Table VI. The recoveries were quantitative for copper, zinc, lead, sodium, potassium, magnesium, calcium and arsenic. Lower recoveries, but better than 90%, were observed for cadmium, iron and phosphorus. These losses could not be accounted for by losses during digestion (Table V) or by chemical association or precipitation with other elements or com- pounds in the sample material (Tables V and VI). The losses were of constant magnitude and could perhaps be explained by the formation of insoluble salts of the digesting acids or by the absorption of small amounts of these elements on the surface of the digestion flask.Similar losses for iron were reported by the US National Bureau of Standards in the analysis of their reference materials. Coefficients of variation around the mean percentage recovery320 HILLIARD AND SMITH : MINIMUM SAMPLE PREPARATION FOR THE Analyst, V d . 104 of cadmium, lead and iron (Table VI) were sufficiently low to permit the valid application of factors to correct the results obtained with the technique described. Measured concentra- tions of cadmium, iron and phosphorus (in sample digests) were corrected for apparent losses by dividing by 0.90, 0.92 and 0.94, respectively. These factors represent the mean recovery of these elements as shown in Table VI.TABLE VI MEAN PERCENTAGE RECOVERIES OF TWO LEVELS OF STANDARD ADDITION T 1 INCREASING MASSES OF A REPRESENTATIVE PIG FAECES SAMPLE Element c u Zn Cd Pb Fe Na K Mg Ca P As Standard addition 200 CLQ 300 Pg 600 Pg 900 Clg 2.0 Pg 3.0 Pg 20 P!3 30 Pg 1000 pg 1500 pg 20 mg 30 mg 10 mg 15 mg 10 mg 15 mg 20 mg 30 mg 20 mg 30 mg 3.0 CLg 6.0 Pg Mean* recoveries (yo) from r A \ 1.0 g 1.5 g 2.0 g 100.3 102.6 97.2 98.6 104.1 102.5 99.8 98.4 104.8 106.7 105.7 107.9 83.7 94.7 89.8 93.7 89.2 90.8 99.3 101.9 96.6 108.7 101.0 95.1 93.2 94.2 83.7 90.5 89.8 93.1 99.1 102.3 99.9 95.7 96.6 99.6 95.8 105.0 97.3 97.3 102.5 98.4 91.7 97.2 102.6 102.9 103.4 100.4 94.6 98.8 107.7 100.9 97.5 92.9 96.5 90.8 94.9 100.4 95.4 91.3 99.0 91.1 95.2 94.9 Meant recovery using pooled data, % 101.1 f 5.3 C.V.: = 5.2% 102.1 -+ 6.0 C.V.= 5.9% 90.1 & 4.9 C.V. = 5.4% 99.5 f 10.3 C.V. = 10.3% 92.0 f 7.1 c.v.= 7.7% 99.6 f 3.2 C.V. = 3.2% 100.1 f 8.0 C.V. = 8.0% 100.5 f 6.5 C.V. = 6.5% 98.0 & 7.9 C.V. = 8.0% C.V. = 5.5% 93.9 f 5.2 97.0 & 12.1 C.V. = 12.5% * Mean of 3 determinations. 7 Mean of 18 determinations, except for As, which is mean of 12 determinations. C.V. = coefficient of variation. A two-way analysis of variance was performed to establish if the recovery was affected by two levels of standard addition to increasing masses of sample material, and the results are shown in Table VII. This method of analysis showed an apparently significant effect (P < 0.05) for zinc between recovery and level of standard addition.The mean recovery of 600 pg of zinc added to 1.0, 1.5 and 2.0 g of faeces was 101.0 & 6.2y0, whereas the recovery of 900 pg of zinc from similar masses of sample was 106.7 & 1.7%. Differences of this order could be attributed to pipetting errors and no real practical significance could be attached to these variations. The statistical analysis in Table VII indicated an interactions effect between sample mass, standard addition and recovery for both calcium and cadmium. The recovery data for calcium (Table VI) indicated that when 20 mg of calcium were added to increasing masses of faeces the percentage recovery increased with mass, but when 30mg of calcium were added the recovery decreased with increasing mass of faeces.These results imply that an optimum range existed for calcium recovery and that the maximum recovery of added calcium was obtained when the total concentration of calcium (from sample and standard addition) in the digest was between 700 and 1000 pg ml-l. It is possible thatApril, 1979 DETERMINATION OF TEN ELEMENTS IN PIG FAECES AND FEEDS BY AAS TABLE VII 321 TWO-WAY ANALYSIS OF VARIANCE TO DETERMINE THE EFFECTS ON RECOVERY OF TWO LEVELS OF STANDARD ADDITION TO THREE DIFFERENT MASSES OF BULK PIG FAECES Mean square (M.S.) and degrees of freedom (D.F.) for source of variation A f \ Increased sample Increased standard mass addition Interaction Error - -----I Element M.S. D.F. M.S. D.F. M.S. D.F. M.S. D.F. c u 27.97 2 13.35 1 18.27 2 19.47 12 Zn 28.87 2 145.64* 1 8.41 2 21.42 12 Cd 16.01 2 15.49 1 91.24* 2 17.40 12 Pb 104.39 2 24.50 1 55.77 1 56.89 12 Fe 25.61 2 2.64 1 84.93 2 41.97 12 Na 6.26 2 24.50 1 24.37 2 7.62 12 K 87.32 2 0.02 1 7.48 2 57.64 12 28.77 2 115.52 1 68.53 2 27.15 12 11.77 2 48.35 1 171.88* 2 37.88 12 Ca P 57.42 2 11.36 1 31.36 2 19.48 12 As 79.05 1 2.80 1 23.52 1 188.87 4 * Indicates significance a t P <0.05.Mg these results are a statistical or analytical anomaly, as the analysis of NBS orchard leaves 1571 (Table VIII) by our technique gave a calcium level of 2.1174, which is in good agree- ment with the quoted level of 2.09 & 0.03%. The digest solution of this material had a calcium concentration of approximately 420 pg ml-1, which is outside the apparent optimum range. The interaction effect on cadmium recoveries (Table VII) is barely significant and there were no obvious trends in recovery associated with increasing sample mass or standard addition. We concluded that any apparent significance of the interaction was fortuitous as at a 5% probability level it is possible to obtain one erroneous result in twenty.Our study involved a total of 33 analyses of variance and it was possible for us t o obtain at least one apparently significant result due to chance. TABLE VIII ANALYSIS OF NBS ORCHARD LEAVES 1571 USING THE PROPOSED PROCEDURE Element Certified concentration found* Concentration c u 12 i 1 pgg-1 12 ll.gg-l Zn 25 z t 3 pgg-l 27 C L g e Cd 0.11 & 0.02 pgg-1 0.09 pg 8-1 Pb 45 f 3 pgg-l 47 tLg g-l Na 82 f 6 pgg-l -t 2.11% Mg P 0.21 f 0.01% 0.20% As 14 i 2 pg g-' 16 pg 8-l Fe 300 & 20 pg g-l 287 pg g-l K 1.47 f 0.03% 1.48% 0.62 f 0.02% 0.61% Ca 2.09 f 0.03% * Dry matter basis.Concentrations quoted as found are means of duplicate analyses. t Na was not determined as its concentration in the reference material fell below the calibration range of our procedure. Results of analysis of NBS orchard leaves 1571 are shown in Table VIII. The level of sodium in the orchard leaves is outside the working range of the proposed procedure. Sodium levels similar to those found in orchard leaves can be determined by using the more sensitive sodium line at 589.0 nm. Comparison of the certified values for orchard leaves and those found by using our technique (Table VIII) confirmed the accuracy of our methods.All of the concentrations we found were within the standard deviations associated with the certified levels.322 HILLIARD AND SMITH Our procedure offers high accuracy and precision when used to measure the concentrations of these eleven elements in pig faeces. It is applicable to other biological materials, as shown by analysis of NBS orchard leaves 1571, and can be used to determine economically the concentrations of copper, zinc, cadmium, lead, iron, sodium, potassium, calcium, magnesium and phosphorus from a single digestion. Arsenic can be determined separately using the acid - molybdate digestion solution. Application of the Procedure These procedures were used to analyse samples of pig faeces and pig feeds collected from commercial piggeries in Victoria, Australia.Results of these analyses were used to deter- mine the mineral status of the pig rations and to evaluate the environmental consequences of various methods of disposal or utilisation of piggery effluents. Typical levels of the elements assayed for are presented in Table IX. TABLE IX ANALYSIS OF PIG DIETS AND PIG FAECES USING THE PROPOSED PROCEDURE (DRY MATTER BASIS) Cul Znl Cd/ Pbl Fel N,”, K, Mg, Ca, P, As1 Sample No* !-a g-’ I*g g-’ I*g g-’ I*g g’ I*g g-1 /o % % % % wgg-1 Diet ._ _ . 1 23 231 0.02 1.20 299 0.13 0.49 n.16 1.60 1.00 n.5 - _ _ 2 163 152 0.27 1.24 341 0.25 0.71 0.19 1.65 1.25 0.2 3 183 167 0.31 1.57 458 0.37 0.74 0.21 1.35 1.26 0.5 4 20 140 0.04 2.85 353 0.20 0.61 0.11 1.14 0.93 32.4 5 8 97 0.02 1.22 241 0.14 0.50 0.13 0.68 0.66 0.4 Faeces . . 1 64 766 0.04 12.65 5098 0.18 0.86 0.48 3.21 1.99 4.4 2 364 572 0.74 20.73 2098 0.24 1.05 0.68 4.18 2.04 1.1 3 569 802 1.20 3.88 6407 0.20 0.99 0.82 4.63 3.13 19.7 4 76 622 0.32 6.84 5306 0.28 0.80 0.51 3.40 3.26 102.6 5 55 616 0.26 0.29 1557 0.12 0.87 0.82 2.93 2.15 18.3 The authors thank Mr. G. Perry, Mr. R. Reid and Mr. E. Butler for their assistance during the development and calibration of these analytical techniques. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. References Pearce, G. R., “Managing Livestock Wastes. Proceedings of the 3rd International Symposium on Hilliard, E. P., and Pearce, G. R., Agric. Envir., 1978, 4, 65. Underwood, E. J., “Trace Elements in Human and Animal Nutrition,” Academic Press, New York, Bremner, I., Q. Rev. Biophys., 1974, 7, 75. Christian, G. D., and Feldman, F. J., “Atomic A4bsorption Spectroscopy : Applications in Agri- de Vries, M. P. C., Tiller, K. G., and Beckett, R. S., Commun. Soil Sci. Pl. Anal., 1975, 6, 299. Jackson, M. L., “Soil Chemical Analysis,” Constable, London, 1962, p. 153. Duncan, L., and Parker, C. R., “Varian Techtron Technical Topics,’’ Varian Techtron Pty. Ltd., Analyt‘ical Methods Committee, Analyst, 1975, 100, 54. Aggett, J., and Aspell, A. C., Analyst, 1976, 101, 341. Simon, R. K., Christian, G. D., and Purdy, W. C., Am. J . Clin. Path., 1968, 49, 207. Livestock Wastes,” American Society of Agricultural Engineers, Michigan, 1975, p. 218. 1977. culture, Biology and Medicine,’’ Wiley-Interscience, New York, 1970, p. 188. Springvale, Victoria, Australia. Received August 31st. 1977 Amended August 8th, 1978 Accepted October 2nd, 1978