A critical study was made of a number of lipid extraction methods on stabilised brown fish meals and on white fish meal. By analysing the lipids it was shown that in all procedures proteinaceous material was extracted and an attempt was made to correct for this non-lipid material. On the basis of the results obtained it seemed that the Bligh and Dyer method gave the highest yield of lipid. In addition it was found that the EEC procedure needed the application of a correction factor to account for the losses sustained in the hydrolysis step.

No abstract

The eflect was studied of certain properties of fish meal on their tendency to predispose chickens to develop gizzard lesions when used in starter diets. These properties relate to the degree of heat treatment to which the meals had been subjected, the fineness of the meal, and the fish species involved. The effect on gizzard erosion of addition of dietary lysine was also considered. The results suggest that the use of meals heated to 130°C or higher, but not of meals stored at lower temperatures, is more likely to result in the development of gizzard lesions, while the fineness of the fish meal had no effect on this phenomenon. In some instances added dietary lysine reduced gizzard erosion incidence. N o relationship was found between chicken liveweight gain during the experimental period and gizzard lesion scores. A 7-day biological test was used in these studies.

A method is presented whereby the toxic amino acid gizzerosine can be determined in j s h meals to a lower limit of 10 mg k g -' . Fish meal was hydrolysed with hydrochloric acid and, after concentration by fieeze drying, derivatised with dansyl chloride solution without further clean-up. A ,@action containing the tridansyl derivative of gizzerosine was separated and collected by HPLC and j?om this the didansyl derivative was separated a f e r treatment with formic acid on the same column as previously using the same mobile phase. Seventeen fish meals, some causing severe gizzard erosion in chickens, were analysed but in none of them could gizzerosine at or nbove this level be found.

Nitrogen retention by chickens served as criterion of the response of three hydrolysed feather meals in two factorial experiments. The South African feather meal used in the first trial gave better N retention when supplemented with methionine, lysine, histidine and tyrosine, while tryptophan supplementation had no effect. The amino acids evidently become limiting in the order in which they are listed. Yet added methionine would only give a response when supplemental lysine (but not supplemental histidine) was present.An Argentinian hydrolysed feather meal of dark colour proved to be more poorly utilized than the South African and Canadian meals studied, whether or not supplemented with lysine and/or methionine and/or histidine. There was a significant meal X histidine supplementation interaction with the South African meal only responding to histidine supplementation. With methionine plus lysine added the Canadian meal gave slightly higher N retention than the South African feather meal, but when histidine was also added, the meals gave similar performances.