YELLOW PERCH OF LAKE ERIE 



223 



legal-sized fish were excluded the reductions in the 

 calculated lengths were smaller. 



It should be mentioned that the data of table 

 6 are based on the elimination of legal-sized fish 

 in a single group of samples whereas the removal 

 of legal-sized individuals by the fishery is gradual 

 and is also progressive in the sense that continued 

 growth during the fisiiing season brings more and 

 more individuals to the legal size. The data 

 serve, nevertheless, to illustrate the type of 

 selective destruction that must occur in the heavily 

 exploited yellow-perch fishery. 



Comparisons of the growth data of table 6 with 

 those of table 5, reveal that the discrepancies pro- 

 duced by the elimination of legal-sized fish from 

 an age group resembled closely the discrepancies 

 that actually occurred between the growth histories 

 of different age groups. It is particularly striking 

 that in both table 6 and table 5, the greatest dis- 

 agreements among the calculated lengths of fish 

 older than age group I occurred beyond the first 

 year of life. It must be considered probable that 

 selective destruction based on sorting according to 

 the legal size limit was an important contributing 

 factor in the observed discrepancies in the calcu- 

 lated lengths of the different age groups of Lake 

 Erie yellow perch. 



OTHER CAUSES 



Differential natural mortality connected with rate 

 oj growth. — The widely observed association of 

 slower growth with the attainment of greater age 

 in poildlothermic animals which was also found 

 by Hile (1936) in the ciscoes of Silver Lake, Wis., 



may have been a possible factor in the discrep- 

 ancies in the calculated growth histories of the 

 Lake Erie yellow perch. The effects of such a 

 differential natural mortality among the Lake Erie 

 perch, however, would be obscured by the more 

 unportant soin-ces of differential destruction by 

 the fishery. 



Annual fluctuations in growth rate. — The dis- 

 crepancies in calculated growth cannot be traced 

 to annual differences in growth rate since the dis- 

 agreements occurred between different age groups 

 of the same year class. 



Formation of more than one annulus per year. — 

 The vaHdity of the use of the annulus on the Lake 

 Erie yellow-perch scale as a true year mark has 

 been established. Although accessory checks are 

 not infrequent, the scales of those fish concerning 

 whose age there was doubt were discarded. It 

 does not appear reasonable, therefore, to assume 

 that the number of errors in the determination of 

 age was sufficiently great to account for the ob- 

 served discrepancies in the calculated growth of 

 different age groups. 



Contraction and resorption of the scale. — Van 

 Oosten (1929) pointed out that the natm-e of the 

 structure of scales makes wholly unacceptable 

 the assmnption that a contraction of scales occurs. 

 The examination of thousands of yellow-perch 

 scales faUed to yield any indication of resorption 

 that would effect the calculation of growth. The 

 lunited amount of resorption or erosion observed 

 in the lateral fields of the scales of some old fish 

 did not affect the measurements along the antero- 

 posterior axis of the scales. 



GENERAL GROWTH CURVES 



GROWTH IN LENGTH 



It is not possible to determine a growth cmwe 

 for the Lake Erie yellow perch that is general in 

 the sense that it describes the growth of an indi- 

 vidual typical of the population as a whole. The 

 preceding discussions have brought out clearly that 

 in general the older fish had a slower rate of growth 

 than the j'ounger. Consequently, the combina- 

 tion of the data of several age groups to determine 

 a general growth ciu-ve involves the lumping to- 

 gether of heterogeneous growth material. The 

 resulting curve is descriptive of the samples rather 

 than of a typical individual. These limitations to 

 the significance of the data should be kept in mind 



in the examination of the information on general 

 growth contained in table 7. 



The average lengths listed in table 7 have been 

 taken from table 5 and are based on the combina- 

 tion of all age groups except group I, which was 

 omitted as nonrepresentative by reason of gear 

 selection (see p. 221). The lengths of fish taken 

 in the fall (presumably at the end of the growing 

 season) were combined with the corresponding cal- 

 culated lengths. Beyond the third j^car of life 

 the average lengths of the different age groups 

 were determined by successive additions of the 

 average annual increments of growth. This pro- 

 cedure brings about a natural smoothing of the 

 general growth curve for the later years of life. 



