FISHERY BULLETIN: VOL 79. N0.4 



Totoi Digestive Troct (observed) 



Siomoch (predicted) 



Figure 14. — Experiment 7. Changes in the digestive tract resi- 

 dence time of silicon (measured) and the stomach residence time 

 of silicon (calculated) during and following a 7-h feeding period. 



offset by the time required for Si to travel the 

 length of the intestine; in the example shown, 

 about 2 h. 



Figure 14 demonstrates that particles eaten at 

 the beginning of a feeding period have the shortest 

 residence times. Residence time increases asymp- 

 totically during feeding, and then exponentially 

 once the fish have stopped feeding. At the end of 

 feeding the observed digestive tract residence time 

 ranged between 5 and 6 h in all but the largest 

 ration experiment, in which the residence time 

 was only 4.5 h. 



This model provides a quantitative explanation 

 of the earlier observations by Noble (1973) and D. 

 J. W. Moriarty and C. M. Moriarty (1973) that 

 during continuous feeding, small food particles 

 ingested early in the feeding period travel through 

 the stomach more quickly than particles eaten 

 later. In addition, observations that food particles 

 eaten during continuous feeding will pass through 

 the stomach more rapidly than when they are 

 eaten as a single meal (i.e., Laurence 1971; Noble 

 1973) are also consistent with this model, since 

 residence time remains short as long as feeding 

 continues, but increases rapidly after the fish stop 

 feeding. 



Finally, the observed fecal elimination rates by 

 the Atlantic menhaden (milligrams Si/gram dry 

 weight per hour) were compared with the pre- 

 dicted gastric evacuation rates (milligrams Si/ 

 gram dry weight per hour) calculated from the 

 model. Since the two curves were offset in time by 

 the travel time of particles in the intestine, a com- 

 parison of the two is facilitated by lagging the 

 stomach evacuation curve by the amount of the 

 intestinal travel time. This time lag was graphi- 

 cally determined for each experiment by overlay- 



ing the curve of stomach evacuation on the curve of 

 feces elimination rate such that the periods of ex- 

 ponential decline coincided. These lag times were 

 quite similar for all experiments ( overall mean ±(t 

 = 2.17 ±0.18 h). These values agree well with the 

 average time for the first appearance of feces after 

 the onset of feeding (2.4 h). 



These plots of the observed fecal elimination 

 rates and the predicted gastric evacuation rates 

 (Figure 1) showed that in general there was good 

 agreement between the two. There were, however, 

 some systematic deviations with change in ration 

 size. At high food rations, the observed elimina- 

 tion rates of the first fecal samples were higher 

 than predicted by the model, indicating that these 

 passed through the digestive tract more rapidly 

 than predicted. In contrast, at the lowest rations 

 the initial elimination rates were lower than pre- 

 dicted. Since the presence of food directly stimu- 

 lates gastric motility and the secretion of digestive 

 enzymes (Fange and Grove 1979), it may be that 

 the larger rations have a greater stimulatory ef- 

 fect on the digestive tract than small rations. It 

 should be noted, however, that these deviations 

 observed in the first two or three fecal samples 

 represent food ingested quite early during the 

 feeding period (Figure 1). Subsequent samples 

 more closely followed the model. 



The largest ration experiment (no. 6) also de- 

 viated from the model during the postfeeding 

 period. The model predicts that if the exponential 

 evacuation rate is <1 (i.e., in the present case = 

 0.366), food will continuously accumulate in the 

 stomach during feeding (Figure 13). The largest 

 amount of material in grams is therefore 

 evacuated at the end of the feeding period (Figure 

 13), and the maximum fecal elimination rates 

 should occur during the postfeeding period (be- 

 cause of the time required for material to travel 

 through the intestine). However, in Experiment 6, 

 the elimination rates quickly rose to high levels, 

 but then declined and leveled off during the post- 

 feeding period without reaching the maximum 

 rates predicted by the model (Figure 1). This im- 

 plies that stomach evacuation was also lower than 

 the model would predict during the last 2 or 3 h of 

 feeding. A possible explanation is that after the 

 Atlantic menhaden had fed for several hours at 

 this high rate, the amount of food may have ac- 

 cumulated in the stomach to an extent which ex- 

 ceeded the maximum physical capacity of the fish 

 to process the material, which caused the gastric 

 evacuation rate to level off. It was interesting that 



612 



