change in length; and the loss in weight was about 

 1 percent. 



The relation of total length to standard length 

 was obtained by fitting a line by means of least 

 squares to the length measurements. The regres- 

 sion equation was £,=4.9 + 1.23 L s , where L, is 

 total length and L s is standard length. 



For the determination of age, three or more 

 scales were removed from every gizzard shad 

 longer than 100 mm. Among the smaller shad, 

 scales were taken from random samples of fish 

 from each collection. The scales were placed on a 

 glass slide immediately after removal from the 

 fish, and their annuli counted with the aid of a 

 dissecting microscope. The ages assigned were in 

 terms of the number of annuli. Since the age 

 designation changes on January 1, a fish captured 

 between this date and the actual time of annulus 

 formation was credited with a "virtual annulus" 

 at the edge of the scale (Hile, 1948). 



The scales employed in the study of the body- 

 scale relation and for the calculation of growth 

 were removed from a "key area." Since gizzard 

 shad lose their scales readily, a "key scale" fre- 

 quently would have been missing or regenerated. 

 Use of a key area eliminated discarding shad which 

 lacked the one particular key scale. The key 

 area is on the left side of the fish, just dorsal to a 

 midlateral line and midway between the posterior 

 edge of the operculum and the origin of the dorsal 

 fin. A dozen or so scales from this area were 

 placed in a scale envelope on which was recorded 

 the information pertinent to the fish. 



Three nonregenerated scales were taken at 

 random from these key-area scales. For study, 

 they were either impressed on strips of plastic by a 

 cold roller press (Smith, 1954) or mounted dry on 

 glass slides. 



The scale measurements were obtained from 

 scale images projected by a microprojection ap- 

 paratus of the type described by Van Oosten, 

 Deason, and Jobes (1934). 



Computations of growth from scale measure- 

 ments were made nomographically. 



Weights are expressed in grams. The largest 

 fish were measured to the nearest gram, fish be- 

 tween 100 and 200 mm. in standard length to the 

 nearest 0.5 g. and those smaller than 100 mm. to 

 the nearest 0.1 g. In catches of large numbers of 

 gizzard shad smaller than 50 mm., those of equal 

 length were weighed en masse and the average 



GIZZARD SHAD IN WESTERN LAKE ERIE 



weight assigned to each member. Fish preserved 

 for a few days in formalin were sometimes used. 

 Some fish were not weighed when collections were 

 large. Gonads — only from freshly caught shad — 

 were weighed to the nearest 0.01 g. 



Ovaries from potential spawners caught in May, 

 June, and July were used for egg counts. From 

 one of the weighed ovaries a small transverse 

 section (1 to 2 g.) was removed and weighed on an 

 analytical balance, its content of current-season 

 eggs counted, and the number of eggs in the entire 

 ovary calculated. 



Histological preparations were made of the 

 ovaries of some fish captured in different seasons 

 for studies of oogenesis and ovarian growth and 

 development. 



Sex was always determined by dissection except 

 for those fish from which eggs or milt were flowing 

 during the spawning season. When possible, sex 

 was determined for all fish longer than 120 mm. 

 Maturity of males could not be determined in the 

 absence of flowing milt. 



SCALE OF THE GIZZARD SHAD 

 Description of the Scale 



The gizzard shad has cycloid scales whose 

 annual growth zones contain many circuli. The 

 relatively closer juxtaposition of the circuli in the 

 first growth zone than in the succeeding zones con- 

 tributes a darker appearance to this inner portion 

 of the scale. The closely set circuli of this zone 

 are gently arched and, save for those formed 

 earliest, do not curve enough at their ends to 

 reach the transverse groove that separates the 

 anterior and posterior fields but terminate in the 

 lateral fields. This pattern is followed by the 

 circuli of each succeeding growth zone; i.e., the 

 earliest circuli reach the transverse groove, while 

 those produced later terminate laterally. The 

 first circulus of the second growth zone is closely 

 juxtaposed to the last circulus of the preceding 

 zone in the anterior field, cuts across the paths of 

 the later first-zone circuli at their terminations 

 in the lateral fields, and intersects the transverse 

 groove. Thus, it forms roughly a semicircle 

 around the first growth zone. The junction 

 between the first and second growth zones (fig. 2) 

 illustrates one type of demarcation line in the 

 gizzard shad scale. The first annulus of all shad 

 scales examined was of this type. Since this 

 annulus normally is evident only in the lateral 



395 



