FISHERY BULLETIN: VOL. 75, NO. 2 



O&G and determined their COD equivalent by 

 direct analysis. 



To prepare protein a sample of muscle was 

 washed with water and centrifuged to remove the 

 blood and other small nitrogen components, then 

 washed with 2-propanol (IPA) to remove part of 

 the water. The sample was blended and refluxed 

 twice with IPA followed by filtration, washing, 

 and refluxing with petroleum ether (PE) and over- 

 night drying at 103°C. These oil free, white, odor- 

 less protein samples were analyzed for nitrogen 

 by the standard macro-Kjeldahl method (Horwitz 

 1965:273) and for COD. The COD factor was cal- 

 culated on a 100% protein basis. 



To obtain O&G, the sample of fish or shellfish 

 was briefly rinsed with water and IPA; then, using 

 a high speed blender and anhydrous conditions 

 (MgS0 4 ), the O&G was extracted, cold, with IPA 

 and PE. For waste effluent, O&G was obtained by 

 the analytical method used previously (Collins 

 1976). By either method, after weighing the dry 

 O&G and diluting to volume with PE an aliquot of 

 the final solution equivalent to 8-10 mg O&G was 

 evaporated in the COD flask, oven-dried for 0.5 h, 

 and used for COD determination. Since PE has a 

 residue significantly affecting COD, freshly dis- 

 tilled PE was used throughout the tests. 



The COD equivalent was determined on a num- 

 ber of different preparations of O&G and protein 

 from fish and shellfish muscle and from shrimp 

 waste effluent. The average values of from 5 to 30 

 replicate COD analyses for each material are 

 given in Table 2. 



The COD coefficients for protein are in reason- 

 able agreement and are probably independent of 



TABLE 2. — The COD coefficient of several preparations of oil and 

 grease (O&G) and protein from fish and shellfish and from 

 shrimp waste effluent. 



Starting material 



Black cod, frozen 

 Pollock, frozen 

 Snow crab, frozen 

 Pink salmon, fresh 



Pink shrimp, fresh 



Pink shrimp, canned 



Shrimp waste effluent 



Mean 

 SD 



species or product form. The theoretical COD coef- 

 ficient of protein was calculated using amino acid 

 percentage composition data for snow crab re- 

 ported by Krzeczkowski and Stone (1974). The 

 theoretical figure of 1.285 mg COD/mg protein 

 was in close agreement with our experimental 

 figure of 1.338. The coefficients for O&G, however, 

 are quite different and are presumably caused by 

 errors in the COD method, differences in species, 

 product, and perhaps slight differences in the 

 method of extracting. There are, of course, known 

 differences in the lipid composition of these 

 species, especially the C-20 and C-22 polyunsatu- 

 rated fatty acids. The chain length and configura- 

 tion of the lipids would have a positive effect on 

 the COD coefficient. For example, some theoreti- 

 cal coefficients are: acetic acid (C 2 ) 1.066, pro- 

 pionic (C 3 ) 1.514, myristic (C 14 ) 2.807, melissic 

 (C 30 ) 3.115, lecithin (C 44 H 88 9 NP) 2.458, and tri- 

 stearin (C 57 H 110 O 6 ) 2.934. Recognizing the wide 

 variations possible, the empirically derived coef- 

 ficient of 2.678 seems reasonable. 



These coefficients are used along with the con- 

 centration of protein and O&G to give the COD, 

 i.e., (1.338 mg COD/mg protein)mg protein + 

 (2.678 mg COD/mg 0&G)mg O&G = COD TR and 

 assumes that the total COD is the sum of the 

 COD of these two major constituents. To check the 

 validity of this equation the coefficients were mul- 

 tiplied by the predicted values for protein and 

 O&G [obtained from TR K data and Equations (4) 

 and (5)] and the resulting mean of the sums of the 

 products (2,155 mg COD/liter) was found to be 

 1.083 times greater than the mean predicted 

 value for COD TR (1,990 mg COD/liter) obtained 

 from TR K data and Equation (1). Although diffi- 

 cult to prove or demonstrate, we believe that the 

 lower analytical values for COD in a sample of 

 waste effluent are caused by the unequal and com- 

 peting oxidation of protein and O&G. As is well 

 known, O&G reacts slowly and especially if the 

 dichromate concentration has been reduced from 

 reacting with the more easily oxidized protein. 

 Minor constituents such as nonprotein nitrogen 

 and carbohydrates would contribute to COD in a 

 ratio different from the protein coefficient. Re- 

 gardless, if the simultaneous equation is to be 

 developed, the inequality must be adjusted by 

 increasing the COD value to equal the sum of the 

 COD of protein plus O&G, i.e., 



1.338 protein + 2.678 O&G = 1.083 COD TR . (6) 



256 



