18 



A determination of the nitrogen excreted at known intake levels permitted calcu- 

 lation of the digestibility and biological value of the protein fed. The average 

 value for digestibility was 76.4 percent, while the average biological value, in- 

 dicative of the availability of the protein, was 83.9 percent. These values com- 

 pare favorably with similar data on other types of fish meals . 



In the first series of feeding tests , rats and chicks were started at the same 

 time with similar diets. The composition of these diets is given in Table 1 (see 

 page 12). Corn, wheat middlings, and pilchard meal levels were varied with the star- 

 fish meal so as to equalize the calculated crude protein content in all diets . The 

 nitrogen content was later determined by the standard Kjeldahl method, the re- 

 sults of which agreed with the calculated values quite closely. No attempt was 

 made in either of the test series to compensate for the high level of calcium and 

 the highly unbalanced calcium: phosphorus ratio, since it was desired to determine 

 the extent of the tolerance for the calcium of the starfish meal which was unavoid- 

 ably included along with the more sought-after protein. The calcium and phosphorus 

 content of the diets has been calculated (Table 1) and the calcium:phosphorus 

 ratios for the highest levels of starfish meal fed were found to be about 24 to 1 

 for the diets fed to rats, and 11 to 1 for those fed to chicks. 



Ten rats weighing 48 to 55 grams each, evenly divided as to sex, were used in 

 each of the 5 groups. The tests continued for 6 weeks, and the liveweight and feed 

 consumption of each rat were recorded weekly. The results showed that the groups 

 receiving 12 and 24 percent starfish meal had a lower growth rate than .the control 

 group while those receiving 48 percent starfish meal and 43 percent extracted star- 

 fish meal showed a net loss of from 8 to 12 grams from the initial weight. The 

 surprising fact is that only 4 of the 20 rats in these 2 groups died. These deaths 

 did not occur until the 6th week, and the remaining rats , while extremely emaciated, 

 were quite lively and showed no other gross symptoms of damage. There were no 

 deaths for the groups fed the 12 and 24 percent levels of starfish meal, although 

 mean gain in liveweight in these groups was only 58 and 27 percent, respectively, of 

 that of the control group. 



Early in the test , it was observed that the rats tended to sort out and leave 

 the s*tarfish meal. This was corrected after the second week by grinding all meals 

 very finely in a ball mill. It is interesting to note that, at this degree of fine- 

 ness, the meal was so hygroscopic that the diets caked in the feed cups. The con- 

 stituent responsible for this property is not known. The caking did not seem to 

 affect the feed consumption, which remained at a fairly even level throughout the 

 test period. 



Feces were collected from 3 rats in each group during the last 3 days on 

 tests and analyzed for total nitrogen. The mean values for the apparent digesti- 

 bility of the protein in the diet indicated by these analyses were as follows: 

 for the control group, 86.3 percent; 12 percent starfish meal level, 80.6 per- 

 cent ; 24 percent starfish meal level, 80.9 percent; 48 percent starfish meal level, 

 79.2 percent; and for the rats fed the high level of extracted meal, 81.7 percent. 

 This high and relatively uniform degree of digestibility at all levels would seem 

 to indicate that the poor growth was not a result of interference with protein 

 metabolism. 



Presence of thiaminase , the thiamine destroying enzyme, had previously been 

 demonstrated in fresh starfish, and the results of the chick growth tests indicated 

 that some of this substance still remained in the meal. However, there was little 

 or no improvement in either weight or condition of the rats with the addition of 



