1. PRINCIPLES OF FEED AND FERTILIZER ANALYSIS

The appraisal of a feedstuff or fertilizer as a direct or indirect source of dietary nutrients for farmed fish or shrimp necessitates information on its nutritional value and toxicological safety. Prior to animal feeding such an evaluation should first include a chemical, microbial and physical description of the product in question.

1.1 Feed analysis

The first step in the chemical evaluation of feed ingredients intended for animal feeding is usually the Weende or proximate analysis. Here, the material under test is subjected to a series of relatively simple chemical tests so as to determine its moisture content, crude protein content, ash content, lipid content, and crude fibre and digestible carbohydrate content. A diagrammatic representation of the Weende proximate feed analysis scheme is shown in Figure 1.

Figure 1. Weende proximate analysis scheme

1.1.1 Protein and non-protein nitrogen composition

The crude protein content of a feedstuff is usually determined using the Kjeldahl method by measuring the total nitrogen content within the sample, and then converting this figure to a total crude protein value by multiplication with the empirical factor 6.25 (this conversion factor is based on the assumption that the average protein contains about 16% nitrogen by weight, although in practice a variation of between 12 and 19% nitrogen is possible between individual proteins). For example, Table 1 shows the range of protein conversion factors used in converting nitrogen values to crude protein values within different foodstuffs. For details of the manual titrimetric Kjedahl method see Munro and Fleck (1969), AOAC (1980), and MAFF (1981).

The major disadvantage of the Kjeldahl method however, is that it does not differentiate between proteinaceous and non-proteinaceous nitrogen (NPN), Consequently, although this method of protein estimation is generally satisfactory for conventional feed ingredients (since they contain only small amounts of NPN), this is not the case for microbial single-cell proteins (SCP; bacteria, algae, yeasts) and certain animal waste products which may contain considerable amounts of NPN. For example, dried poultry manure has a high nitrogen content, equivalent to 28% crude protein, of which only about one third is true amino acid protein and the remainder is NPN, mainly in the form of uric acid (Blair, 1974). In view of the very limited ability (if at all) of fish or shrimp to utilise dietary NPN, it is essential therefore that both true amino acid protein and NPN containing substances (including nucleic acids, uric acid, urea, and amino sugars) be determined separately within these novel feed stuffs. For details of analytical techniques for the estimation of true amino acid protein see Lowry et al., (1951), AOAC (1980), and MAFF (1981).

In contrast to the Kjeldahl method of estimating protein quality, the amino acid composition of a food stuff provides one of the best indicators of its potential nutritive value. Amino acids may be measured individually by using microbiological methods (Barton-Wright, 1972; AOAC, 1980) or by column chromatography (Moore, 1963; Roach, Sanderson and Williams, 1967; Liu and Chang, 1971; Moodie et al., 1982; Williams, Hewitt and Buttery, 1982; Bos, Verbeek and Slump, 1983; AOAC, 1980). However, it must be remembered that the amino acid levels obtained from such analyses do not give any indication of their chemical form within the feed stuff (ie. free, bound, unbound, state of oxidation) or availability during digestion. Consequently, an estimate of amino acid availability within the feed stuff is often warranted. The most commonly used methods for estimating amino acid availability are the available lysine test (Roach, Sanderson and Williams, 1967), and the available methionine test (Kies, 1981; Bos, Verbeek and Slump, 1983).

1.1.2 Lipid and fatty acid composition

The crude lipid content of feed ingredients is usually determined by solvent extraction with light petroleum ether (AOAC, 1980; MAFF, 1981). Other solvents which have also been used for lipid extraction include chloroform: methanol (Bligh and Dyer, 1959; Folch et al., 1957) and trichlorofluoromethane (Korn and Macedo, 1973). Although the lipid fraction or ‘ether extract’ of conventional animal food stuffs and oilseeds is predominantly composed of triglyceride fats and oils, within microbial SCP and certain vegetable feeds as much as half of the total ‘ether extract’ may be in the form of non-glyceride containing substances (Schultz and Oslage, 1976). Consequently, since the majority of non-glyceride based lipids have a lower energy value to fish and shrimp than true fats and oils, it is important that the level of glyceride based lipids be determined separately within these products. Within microbial SCP and extensively heat-processed animal feeds (ie. heat expanded feeds) a further complication arises because a substantial portion of the lipid is present in a bound form which requires hydrolysis in 4N-HCL prior to solvent extraction with ether (Salo, 1977; Halverson and Alstin, 1981; Limsuwan and Lovell, 1985).

¹ Source: MAFF (1975)

² Includes all meats and fish

The fatty acid composition of a lipid is usually determined by gas-liquid chromatography (GLC) after lipid extraction and tranesterification (Morrison and Smith, 1964; Martinetti, 1967; Christie, 1973; AOAC, 1980). Since feeds rich in polyunsaturated fatty acids (PUFA) are highly prone to oxidative damage, numerous chemical methods are available for determining the degree of oxidation or oxidative rancidity, including free fatty acid content (AOAC, 1980; Ke et al., 1976; Windsor and Barlow, 1981), peroxide value (AOAC, 1980; Pearson, 1970), thiobarbituric acid value (Asakawa et al., 1975), and anisidine value (Windsor and Barlow, 1981)

1.1.3 Crude fibre and digestible carbohydrate content

Various chemical techniques are available for the estimation of carbohydrate in plant and animal feeds. The method most commonly employed divides the carbohydrates into two fractions; crude fibre and nitrogen-free extractives (NFE; Figure 1). Crude fibre is the insoluble organic residue remaining after extracting an oil-free food stuff with dilute acid and alkali under controlled conditions. However, the analysis of dietary fibre has presented many difficulties to the analyst (for a review of analytical methods see Southgate, Hudson and Englyst, 1978 and AOAC, 1980). Sometimes known as ‘roughage’, crude fibre is generally regarded as the non-digestible carbohydrate component of a feed ingredient or diet. Within feeds of plant origin crude fibre is composed mainly of varying proportions of cellulose, hemicellulose and lignin (the latter not being a carbohydrate, but rather a complex aromatic compound), whereas in animal products crude fibre is composed of varying proportions of glucans, mannans and amino sugars.

Nitrogen-free extractives (NFE) on the otherhand is an indirect measure of the ‘soluble’ or ‘digestible’ carbohydrate present within a food stuff. It is obtained by adding the percentage values determined for moisture, crude protein, lipid, crude fibre and ash, and subtracting the total from 100. Within plant based feeds this fraction is composed primarily of free sugars, starch and other digestible carbohydrates. However, with the development of specific analytical techniques for measuring individual carbohydrates it is preferable to measure these substances in feeds directly (for details of analytical methods for the measurement of individual carbohydrates see Somogyi, 1952; Murat and Serfaty, 1974; Harbone, 1973; AOAC, 1980; and MAFF, 1981).

1.1.4 Ash and mineral composition

The ash content of a feedstuff is the inorganic residue remaining after the organic matter has been destroyed by combustion in a muffle furnace (AOAC, 1980; MAFF, 1981). The mineral composition of the ash so obtained is not necessarily the same as that originally present in the feed material as some elements are volatile at ashing temperatures above 450^oC (notably the elements mercury, arsenic, selenium, phosphorus, chromium and cadmium; Katz, Jenniss and Mount, 1981). Consequently, for trace mineral analysis feed samples are usually solubilised by a wet-acid oxidation technique prior to analysis by atomic absorption spectrophotometry (for details of these analytical methods see AOAC, 1980 and MAFF, 1981).

1.1.5 Vitamin composition

Individual vitamins can be measured by a variety of different microbial, biological, and chemical techniques. Although most vitamin analyses are rather lengthy and complicated, for full details of analytical techniques see the reviews of Hashmi (1973) and AOAC (1980).

1.1.6 Variability in chemical composition

A problem which may be encountered with the use of certain feed ingredients, and in particular with animal and plant by-products and wastes, is that their chemical composition may vary considerably over relatively short periods of time and from one locality to another (Blair, 1974; Tacon and Ferns, 1979). A proper appraisal of the potential nutritional value of a feeding stuff should therefore also include an assessment of the chemical variability of the product over regular time intervals.

1.1.7 Physical characteristics

Apart from a chemical statement of the major nutrients present within a feeding stuff, information is also required on the physical characteristics of the product; including colour, texture, odour, particle size, and density.

1.1.8 Endogenous anti-nutritional factors

The presence of endogenous anti-nutritional factors within plant feedstuffs is believed to be one of the main factors limiting their use within animal feeds. Figure 2 summarises the major groups of anti-nutritional factors presenÍt within plant foodstuffs, with specific examples given in Table 2. Several of these endogenous anti-nutritional factors may also be found within animal tissues; for example the anti-vitamin factor thiaminase may be found in rawfish and shellfish (NRC, 1983).

Figure 2. Classification of endogenous toxic factors occuring in plant foodstuffs of agricultural importance (Tacon, 1985)

In view of the potential deleterious effect of these anti-nutritional factors on the growth of fish and shrimp when present in complete diet regimes (at high dietary inclusion levels), it is important that the level and activity of these endogenous anti-nutritional factors be determined within feedstuffs of plant origin. For details of analytical techniques see AOAC, (1980); Liener (1980), and MAFF (1981).

1 Compiled from the data of Kay (1979) and Liener (1980)

Anti-nutritional factor:1-Protease inhibitor; 2-Phytohemagglutinin; 3-Glucosinolate; 4-Cyanogen; 5-Phytic acid; 6-Saponin; 7-Tannin; 8-Estrogenic factor; 9-Lathyrogen; 10-Gossypol; 11-Flatulence factor; 12-Anti-vitamin E factor; 13-Anti-vitamin B1 (thiamine) factor; 14-Anti-vitamin A factor; 15-Anti-vitamin B6 (pyridoxine) factor; 16-Anti-vitamin D factor; 17-Anti-vitamin B12 factor; 18-Amylase inhibitor; 19-Invertase inhibitor; 20-Arginase inhibitor; 21-Cholinesterase inhibitor; 22-Dihydroxyphenylalanine; 23-Mimosine; 24-Cyclopropenoic fatty acid; 25-Possible mycotoxin (aflatoxin) contamination.

1.1.9 Adventitious toxic factors

Feedstuffs, depending on their origin and processing, may contain a variety of adventitious toxic factors, including; fish poisons, protozoan and algal toxins, solvent residues (ie. solvent residues are sometimes present in solvent extracted oilseeds - methylene chloride, ethylene dichloride, trichloroethylene, hexane, acetone, iso-propyl alcohol), fungal toxins (ie. aflatoxins present in badly stored feedstuffs), bacterial toxins (ie. botulinum toxin), therapeutic drugs (antibiotics, sulphonamides, nitrofurans, arsenilic acid), pesticide residues (chlorinated hydrocarbons), organochlorine compounds (polychlorinated biphenyls), petroleum hydrocarbons (ie. n-paraffins), and heavy metal contaminants (Friedman and Shibko, 1972; Ashley, 1972; NRC, 1983).

1.1.10 Microbial contamination

Within microbial, faecal and animal based feedstuffs there is a possibility of a hazard arising from the presence of viable and contaminating micro-organisms, including; bacteria (ie. Salmonella contamination within animal by-products), fungi, and viruses (ie. faecal waste products). To eliminate the risk of disease transmission from these feedstuffs it is essential that they be processed in such a manner so as to ensure complete destruction of the disease causing organism, or to reduce their levels to the suggested limits laid down by the government feed control officials (MAFF, 1973, 1976; Pierce, 1976; PAG, 1974). For details of analytical methods see AOAC (1980).

1.1.11 Certificate of analysis of feeding stuffs

The analysis and certification of animal feeding stuffs in all countries is normally controlled by a statutory regulatory body; for example in the United Kingdom the analysis and certification of feeding stuffs is controlled by ‘The Fertilisers and Feeding Stuffs Regulations 1973’ of the Ministry of Agriculture, Fisheries and Food (MAFF, 1973; MAFF, 1976). However, official animal feeding stuff regulations will vary from country to country. Table 3 shows a typical certificate of analysis.

Table 3. MAFF (1973) Official Certificate of Analysis

CERTIFICATE OF ANALYSIS OF FEEDING STUFF

I, the undersigned, agricultural analyst for the             , in pursuance of the provisions of the Agriculture Act 1970, Part IV, hereby certify that I received on the      day of          19    , from          one part of a sample of                  for analysis; which was duly sealed and fastened up and marked              and was accompanied by a             , as follows:-
and also by a signed statement that the sample was taken in the prescribed manner; and that the said part has been analysed by me or under my direction, and I declare the results of analysis to be as follows:-

Analysis for oil was completed on and I am of the opinion that

The analysis was made in accordance with the Fertilisers and Feeding Stuff Regulations 1973.

As witness my hand       day of           19       

1.2 Fertilizer analysis

The chemical analysis of inorganic fertilizers (ie. mineral compounds) and organic fertilizers (ie. animal manures and plant residues) is normally restricted to three nutrient classes. With the exception of water, these include:

1. The primary or major nutrients: nitrogen (N), phosphate (P₂0₅), potash (K₂0) and carbon (C)
2. The secondary nutrients: sulphur (S), magnesium (Mg) and calcium (Ca)
3. The micro-nutrients: iron (Fe), copper (Cu), zinc (Zn), manganese (Mn), boron (B) and molybdenum (Mo)

1.2.1 Primary nutrients

Fertilizer primary nutrient levels are usually expressed as percent N:P₂O₅:K₂O. For example, an inorganic fertilizer labelled as 15:20:10 will contain 15% nitrogen (N), 20% phosphate (P₂O₅) and 10% potash (K₂O). Although the terms ‘P₂O₅’ and ‘K₂O’ are normally used to express the fertilizer nutrients ‘phosphate’ and ‘potash’, there is now a trend to express fertilizer nutrient levels as the single element and not as the oxide. The conversion factors used are as follows;

1. To convert oxides to elements multiply P₂O₅ value by 0.4364 and K₂O value by 0.8302
2. To convert elements to oxides multiply P value by 2.2914 and K value by 1.2046

Phosphate is usually determined by the quinolinium phosphomolybdate method or by the spectrophotometric (vanadium phosphomolybdate) method (MAFF, 1973; AOAC, 1980). During these analyses phosphate can be determined as total phosphate, and as water-soluble, water-insoluble and citric acid-soluble phosphate. Potash is normally determined by the perchloric acid method, or by the potassium chloroplatinate method or, in fertilizers containing not more than 20% potash, by the flame photometric method (MAFF, 1973; AOAC, 1980). Nitrogen is usually determined by the Kjeldahl method (see section 1.1.1), or alternatively in conjunction with carbon and hydrogen by using an automatic carbon-hydrogen-nitrogen analyser (AOAC, 1980).

1.2.2 Secondary and micro-nutrients

The secondary and micro-trace element content of fertilizers are normally determined using the same methods described for feeding stuffs (MAFF, 1973, 1981; AOAC, 1980; see section 1.1.4).

1.2.3 Variability in chemical composition

As with feeding stuffs, a serious problem which may be encountered with the use of fertilizers, and in particular organic manures and plant residues, is that their chemical composition may vary considerably over short periods of time and from season to season. An assessment of fertilizer quality should therefore include an analysis of variability in chemical composition (for limits of variation see MAFF, 1973).

1.2.4 Contaminants

As with feeding stuffs, there is sometimes a possibility that fertilizers may be contaminated with toxic mineral elements, pesticides, herbicides, growth promotants and pathogenic micro-organisms (ie. animal manures). For details of analytical methods see AOAC (1980) and MAFF (1981).

1.2.5 Certificate of analysis of fertilizers

The analysis and certification of fertilizers, as with feeding stuffs, is normally controlled by a statutory regulatory body in all countries (see section 1.1.11). Table 4 shows a typical certificate of analysis.

Table 4. MAFF (1973) Official Certificate of Analysis

Certificate of Analysis of Fertiliser

I, the undersigned, agricultural analyst for the         , in pursuance of the provisions of the Agriculture Act 1970, Part IV, hereby certify that I received on the          day of          ,19     , from              one part of a sample of          for analysis; which was duly sealed and fastened up and marked          and was accompanied by a          , as follows:-

and also by a signed statement that the sample was taken in the prescribed manner; and that the said part has been analysed by me, or under my direction, and I declare the results of analysis to be as follows:-

and I am of opinion that

The analysis was made in accordance with the Fertilisers and Feeding Stuffs Regulations 1973.

As witness my hand this          day of          ,19     ,

(Signature and address of analyst)