FISHERY BULLETIN: VOL. 74, NO, 1 



average percent weight loss depended upon 

 length. Large fish lost a greater percentage than 

 small fish. Average total weight loss was from 30 

 to 50% for 359-493 mm FL males and from 48 to 

 62% for 421-531 mm FL females, sizes which 

 closely approximate the observed size range of 

 fish in the 1963 and 1964 runs (Chittenden 1969). 

 The observed mean fork lengths of fish captured 

 at Lambertville were 428 mm for males and 477 

 mm for females, based upon the regression equa- 

 tions, and these sizes averaged 45 and 57% total 

 weight loss, respectively. 



Somatic weight loss, a better measure of the 

 toll taken by the spawning migration, was esti- 

 mated by subtracting the predicted total gonad 

 weight from the predicted total weight at Lam- 

 bertville before making a comparison with the 

 Tri-State Survey total weight regressions. No 

 correction was made for the gonads of fish cap- 

 tured during the Tri-State Surveys; however, 

 these were a negligible fraction of the total 

 weight. The total testes weights of 15 males col- 

 lected near Hancock on 14 July 1964 and on 21, 24 

 June and 1 July 1965 ranged from 3.7 to 27 g and 

 averaged 15.9 g while the total ovary weights of 3 

 females collected then varied from 18.2 to 35 g 

 and averaged 27.1 g. The average percentage of 

 somatic weight loss in males was 24% at 359 mm, 

 46% at 493 mm, and 42% for the mean-sized male 

 of 428 mm. For females, somatic weight loss was 

 38% at 421 mm, 56% at 531 mm, and 50% for the 

 mean-sized female of 477 mm. 



Absolute daily weight loss was estimated from 

 the duration of the freshwater residency. Fish 

 captured during the Tri-State Surveys had prob- 

 ably been upstream about 75 days. This approxi- 

 mates their maximum stay in fresh water because 

 the peak of the run at Lambertville is about 1 

 May (Chittenden 1969), and most fish move sea- 

 ward from the Hancock area by late June. There- 

 fore, the average daily loss in somatic weight of 

 males was 1.63 g at 359 mm, 9.37 g at 493 mm, 

 and 5.75 g for mean-sized males of 428 mm. For 

 females the average daily loss in somatic weight 

 was 5.75 g at 421 mm, 18.87 g at 531 mm, and 

 12.47 g for mean-sized females of 477 mm. 



Daily weight loss can be used to suggest how 

 long fish of different sizes can remain in freshwa- 

 ter before death. The amount of weight loss which 

 results in death of shad is not known, but death 

 occurs in many animals when weight loss exceeds 

 40% (Curtis 1949). Assume 50% for simplicity in 

 calculation, this may not be quite correct, but it 



may be conservative and the size pattern, at 

 least, remains the same if the percentage is a con- 

 stant. From this, males could remain 154 days at 

 359 mm, 81 days at 493 mm, and the average 

 sized male (428 mm) could remain 90 days. 

 Females could remain 100 days at 421 mm but 

 only 68 days at 531 mm, and the mean-sized 

 female of 477 mm could remain 75 days. There is 

 apparently little difference in the amount of time 

 an average to maximum-sized fish can spend in 

 fresh water before death, but small fish can sur- 

 vive much longer. 



GENERAL DISCUSSION 



Weight loss data presented herein agrees 

 reasonably with those of Leggett (1972) who 

 noted that his figures were probably underesti- 

 mates. The present figures ignore weight loss in 

 the 100-km migration between Marcus Hook and 

 Lambertville and may be based on a longer than 

 average stay in fresh water. Both factors tend to 

 underestimate weight loss which affects related 

 estimates. 



Many shad apparently remain upstream near 

 the spawning grounds well into the summer. 

 However, the percentage they comprise of the run 

 is unknown. A few fish remain far upstream until 

 late fall. Bishop (1936) captured emaciated indi- 

 viduals 305-330 mm long near Hancock in 

 November These fish must have migrated up- 

 stream during the previous spring, because low 

 dissolved oxygen water near Philadelphia pre- 

 sents a virtually impassable barrier through 

 summer and fall (Ellis et al. 1947; Sykes and 

 Lehman 1957; Chittenden 1969). Nichols (1959) 

 captured an emaciated male during October in 

 the Connecticut River and estimated it had been 

 in freshwater at least 120 days. I captured an 

 emaciated male (287 mm FL, 194 g) in fresh water 

 in the James River, Va. on 7 October 1969. 



The finding of little or no food in adults col- 

 lected at Lambertville is similar to the reports of 

 Bean (1903), Leim (1924), Leach (1925), Hilde- 

 brand and Schroeder (1928), Moss (1946), and 

 Hildebrand (1963) that adults take little or no 

 food while ascending rivers. My observations of 

 instances of intensive feeding while upstream are 

 exceptional, although Atkinson (1951) reported 

 an artificial instance of feeding in freshwater 

 ponds. Adult shad at sea feed largely on 

 planktonic forms such as copepods and mysids 

 (Leim 1924; Hildebrand and Schroeder 1928; 



154 



