100 



FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 



the body-scale ratio can be considered constant 

 and the simple procedure of direct-proportion 

 calculations can be followed. The lack of trend 

 in the body-scale ratio with increase in length 

 (table 11) and the close fit of the line in figure 3 

 (fitted on the assumption of no intercept and with 

 a slope based on the mean body-scale ratio) to the 

 empirical data justify this view. 



To be sure, newly hatched lake herring do not 

 have scales as would be indicated by the relation- 

 ship just described. Fish (1932) and Pritchard 

 (1930) in detailed descriptions of this species from 

 hatching up to lengths of 17.5 to 20.0 millimeters 

 (0.7 to 0.8 inch) did not mention the presence of 

 scales. Van Oosten (1929) reported a lake herring 

 34 millimeters (1.3 inches) long which had not 

 formed scales and stated that scale formation 

 probably starts at a fish length of 35 to 40 milli- 

 meters (1.4 to 1.6 inches). Wohlschlag (1953) 

 observed that the first scales were formed at a 

 length of 40 to 60 millimeters (1.6 to 2.4 inches) in 

 Leucichthys sardinella of an Alaskan lake. Thus, 

 the body-scale regression does not exist until a 

 body length of about 1.5 inches is reached. There 

 must be a period before the end of the first year 

 of life when scale growth is considerably faster 

 than body growth, after which the linear relation 

 with an apparent 0-intercept has become estab- 

 lished. 



TIME OF ANNULUS FORMATION 



Because so few investigators of the growth of the 

 lake herring have made collections throughout the 

 year, information on the time of formation of the 

 annulus or year-mark is scanty. A certain amount 

 of data is available, however, on the annulus 

 formation of species closely related to the lake 

 herring. 



Van Oosten (1923), studying scales taken at 

 monthly intervals from whitefish (Coregonus clu- 

 peaformis) kept at the New York Aquarium, 

 demonstrated that the annulus was completed by 

 resumption of growth in March or April and that 

 these fish ceased growing in August or September. 

 He suggested that rising temperature was prob- 

 ably the primary cause of the resumption of growth 

 and the development of sex products causes re- 

 tardation or cessation of scale growth in late 

 summer. In a later report Van Oosten (1929, p. 

 345) commented on annulus formation in the lake 

 herring of Saginaw Bay that "* * * inasmuch 



as the formation of an annulus is causally related 

 to the retardation of growth, it is safe to assume 

 that in nature, too, the annulus of these species 

 forms during the winter period." He left open 

 the question as to just when in the winter the 

 annulus does form. In his study of growth of the 

 Lake Nipissing cisco Fry (1937, p. 18) stated that 

 "Scales from fish captured in early spring place 

 the date of the completion of the annulus, or 

 rather of the initiation of the new season's growth, 

 at some time in May." Dannevig and Dannevig 

 (1937, p. 198), however, stated that in the brown 

 trout (Salmo trutta) of southern Norway "The 

 general opinion that the narrow ridges are formed 

 during autumn and winter as a result of low tem- 

 peratures does not hold good." They showed that 

 in two lakes the scales of most trout had new 

 growth outside a year-mark in the spring, whereas 

 in two other lakes new growth occurred in the fall. 

 They also reported that some annulus formation 

 took place every month of the year. Although 

 these findings are not directly applicable, it is 

 believed that some of the characteristics of this 

 not too-distantly related group are reflected in 

 the scale growth of Green Bay lake herring. 



From the data on annulus formation in the lake 

 herring of Green Bay (table 12) it is seen that 

 some fish had started new growth as early as 

 May 8, the date of the earliest collection other 

 than the midwinter samples in which no fish ex- 

 hibited new growth. As the season progressed the 

 percentage of fish with completed annuli increased, 

 but a few individuals still gave no indication of 

 new growth as late as June 19 to 21. In a July 

 13, 1950, sample and in all collections later than 

 that date, annulus formation was complete.* 



The data of table 12 indicate further that the 

 younger fish form annuli earlier than do the older 

 ones. With only the exceptions of the Ill-group 

 and Vl-group (a single fish) of the June 19-21 

 collections, the percentage of lake herring with 

 new growth decreased with age. That this rela- 

 tion between age and time of annulus formation 

 may be general among fish is indicated by obser- 

 vations in carp, Cyprinus carpio, by Frey (1942); 

 white crappie, Pomoxis annularis, by Hanson 



' Collections of June 21-22, 1950, have been excluded In this consideration. 

 Fewer than half of the lake herring In these collections had completed annuli, 

 but examination of the scales showed them to be much undersized; obviously 

 the scales were removed from the back near the dorsal fin and thus came from 

 a body region above that from which scales were regularly taken. 



