412 



FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 



(12) The length composition of the spring 

 collections of Saginaw Bay yellow perch changed 

 enormously between 1929-30 (modal length 8.5- 

 8.9 inches) and 1943-55 (mode at 6.5-6.9 inches). 

 At the same time the percentage of legal-sized 

 fish (8V2 inches and more) dropped from 73.9 

 percent in 1929-30 to 11.0 percent in 1943-55. 



(13) The relation between the total length in 

 inches (L) and the weight in ounces (W) of 

 the 4,285 Saginaw Bay yellow perch in the com- 

 bined collections was described satisfactorily by 

 the equation: 



W = 3.9975 X 10- 3 Z 3 - 260 



The value of the exponent in this equation was 

 greater than that determined for Saginaw Bay 

 fish in 1929-30 or for any other Great Lakes 

 stock of perch. 



(14) The annual variations of weight were so 

 small among fish of the same length, sex, and 

 condition of the gonads that the data for different 

 years were combined in the study of the relation 

 between weight and sexual condition. No sig- 

 nificant difference of weight existed between ripe 

 and spent males. On the other hand, the females 

 showed an average loss of 12.3 percent of their 

 weight at spawning. 



(15) The seasonal differences of weight were 

 so slight that it is not possible to speak of a 

 seasonal trend. The males and spent females of 

 the spawning-run sample usually were lighter 

 than fish caught on June 22 and October 19. 

 Ripe females had a somewhat weaker tendency 

 to be heavier than fish caught later in the year. 



(16) Because in 1954 scales were taken from 

 above the lateral line rather than from below, 

 as in other collections, it was necessary to estab- 

 lish two body-scale curves for the calculation of 

 growth. These curves were based on "key" scales 

 from above and below the lateral line of 520 fish. 



(17) For scales collected from below the lat- 

 eral line, the direct-proportion method was valid 

 for the calculation of growth to the end of vari- 

 ous years of life for standard lengths of 70 mm. 

 and greater. Direct-proportion calculated lengths 

 below 70 mm. were underestimates and had to be 

 corrected on the basis of the empirical body- 

 scale curve. 



(18) The relation between fish length (stand- 

 ard length in millimeters) and the scale radius 

 for scales above the lateral line was a straight 



line with a 30-mm. intercept on the length axis. 

 Lengths computed from this relation were over- 

 estimates at values of 75 mm. and shorter. Cor- 

 rections were determined from the empirical 

 body-scale curve. 



(19) Growth histories of the same fish as com- 

 puted from measurements of scales from above 

 and from below the lateral line were nearly iden- 

 tical except for the calculated lengths at the end 

 of second year of life (discrepancies at this age 

 averaged 3.6 mm.). 



(20) The calculated lengths for a particular 

 year of life tended to decrease as the fish grew 

 older. These discrepancies in calculated length 

 were most pronounced in the later years of life, 

 particularly after the second year. They differed 

 from "Lee's phenomenon" of apparent decrease 

 in growth rate in which earlier years are af- 

 fected most. 



(21) The principal causes of discrepancies 

 among calculated lengths are: Biased sampling 

 from selective action of gears and from segrega- 

 tion by sexual maturity and size; higher mor- 

 tality of the faster growing fish in the fishery. 

 A higher natural mortality rate in faster grow- 

 ing fish also may be possible. 



(22) The lengths of the sexes were similar in 

 the first and second year of life. In later years 

 the females were the longer; 0.3 inch at the end 

 of the third year of life to 1.8 inches at the end 

 of the seventh. The annual increments of growth 

 in length decreased progressively with age among 

 males and irregularly among females. Males 

 reached the legal length of 8V2 inches in about 

 6 years and females in a little more than 4i/ 2 

 years. 



(23) The annual fluctuations of growth in 

 length the first year and in the later years of 

 life were dissimilar. 



(24) The poorest first-year growth (9.2 per- 

 cent less than the 1942-51 average) was made in 

 1942. In subsequent years a strong trend toward 

 improvement of growth was apparent. The best 

 first-year growth was 8.8 percent above average 

 in 1951. 



(25) No correlation could be demonstrated be- 

 tween first-year growth and : Year-class strength 

 in year of growth; temperature; precipitation; 

 water level. Significant negative correlation was 

 found between first-year growth and turbidity in 

 June (and possibly July). Possible explanations 



