A NOTE ON THE ESTIMATION OF 

 TRIMETHYLAMINE IN FISH MUSCLE 



The chemical methods for estimation of tri- 

 methylamine (TMA) in fish muscle consist of: 1) 

 making a protein-free extract of tissue, 2) treat- 

 ing with formaldehyde (FA) to fix ammonia and 

 amines other than tertiary amines, 3) volatiliz- 

 ing TMA with alkali into an organic phase (tol- 

 uene), and 4) measuring the degree of ionization 

 of picric acid in toluene caused by the extracted 

 amine. 



The methods have caused controversy; they 

 have varied in detail such that the results from 

 different laboratories have not always been in 

 agreement (Shewan et al. 1971). 



Bullard and Collins (1980) have recently com- 

 pared methods of analysis for TMA and the in- 

 terference by ammonia and dimethylamine 

 (DMA). They included the method of Murray 

 and Gibson ( 1972a) and confirm that 45% KOH as 

 alkali is optimal for the release of TMA into the 

 organic phase. They showed that the recovery of 

 DMA is higher than is acceptable and that DMA 

 interferes with the assays for TMA. Murray and 

 Gibson (1972b) had found that the interference 

 was very low and insignificant. 



Tozawa et al. (1971) showed that the concen- 

 tration of FA was critical in order to minimize 

 the interference from DMA (see their fig. 3). 

 They found that 0.5-1.0 ml FA was optimal for a 

 4 ml sample and added the FA before the tol- 

 uene. If less were added, significant interference 

 from DMA occurred even with KOH as alkali 

 rather than K2CO3. 



Bullard and Collins (1980) added only 1 ml of 

 3.7% FA to the sample (4 ml) after the addition of 

 toluene. FA and aqueous FA, which exists as 

 methylene glycol, are soluble in toluene and thus 

 the effective concentration of FA in the aqueous 

 sample is probably much lower than the mini- 

 mum recommended by Tozawa et al. (1971) and 

 is in the range which would cause maximum in- 

 terference by DMA. 



Murray and Gibson (1972a) used 1 ml of 50% 

 neutralized FA added before toluene. They com- 

 pared their procedure with that of Tozawa et al. 

 (1971) and found no significant differences (un- 

 publ. results). In addition they examined chro- 

 matographically the toluene phase after extrac- 

 tion and found that in their procedure only TMA 

 was extracted (Murray and Gibson 1972b). 



Thus it would have been more realistic and fair 

 to previous workers if Bullard and Collins (1980) 



had compared the actual published methods 

 rather than their modifications to them and if 

 they had specifically analyzed the material ex- 

 tracted into the toluene fraction. Accordingly 

 their claim to have improved the method for 

 TMA analysis cannot be substantiated. 



It would be interesting to compare the results 

 of analysis of samples from different species 

 using the actual published methods done by an 

 independent laboratory. 



Literature Cited 



Bullard, F. A., and J. Collins. 



1980. An improved method to analyze trimethylamine in 

 fish and the interference of ammonia and dimethyl- 

 amine. Fish. Bull., U.S. 78:465-473. 

 Murray, C. K., and D. M. Gibson. 



1972a. An investigation of the method of determining tri- 

 methylamine in fish muscle extracts by the formation of 

 its picrate salt— Part I. J. Food Technol. 7:35-46. 



1972b. An investigation of the method of determining 

 trimethylamine in fish muscle extracts by the formation 

 of its picrate salt— Part II. J. Food Technol. 7:47-51. 

 Shewan, J. M., D. M. Gibson, and C. K. Murray. 



1971. The estimation of trimethylamine in fish muscle. 

 In R. Kreuzer (editor), Fish inspection and quality con- 

 trol, p. 183-186. Fishing News (Books), Lond. 

 Tozawa, H., K. Enokihara, and K. Amano. 



1971. Proposed modification of Dyer's method for tri- 

 methylamine determination in cod fish, In R. Kreuzer 

 (editor), Fish inspection and quality control, p. 187-190. 

 Fishing News (Books). Lond. 



D. M. Gibson 



Torry Research Station 



135 Abbey Road 



Aberdeen AB9 8DG Scotland 



FISHERY BULLETIN: VOL. 80. NO. 1, 1982. 



157 



