Fisher and Pearcy Scale circulus spacing in Oncorhynchus kisutch 



641 



Table 4 



Rate of circulus formation (RCF. circuli/d) vs. scale growth 

 rate (SGR, nim/d at 88 x ): Correlation coefficients (r). prob- 

 ability that correlation coefficient = 0.0 (p) and geometric 

 mean regression (GM) for age-0 fish held in saltwater tanks 

 (A),CWT jacks (age 1.0) returning in 1983 (B) and 1985 (C). 

 and CWT juvenile fish (both age 0.0 and 1.0) caught in the 

 ocean (D). 



Group 



r 



GM regression 



A 80 



B 64 



C 99 



D 34 



0.94 <0.01 RCF = 0.30(SGR) + 0.03 



0.84 <0.01 RCF = 0.22(SGR) + 0.03 



0.83 <0.01 RCF = 0.20(SGR) + 0.03 



0.90 <0.01 RCF = 0.22(SGR) + 0.02 



significantly correlated with scale and fish growth rates 

 (Tables 2, 5; Fig. 3). Thus, the faster the fish or scales 

 grew, the more circuli were formed per unit time. 

 Positive correlations between rate of circulus forma- 

 tion and growth rate also have been found for walleye 

 (Glenn and Mathias 1985) and cichlids (Sire 1986). In 

 juvenile walleye, rate of circulus formation ranged from 

 1.5 circuli/d at high growth rates to 1 circulus every 

 2-3 weeks at low growth rates. Data presented by 

 Bilton and Robins (1971a) indicated that juvenile sock- 

 eye salmon receiving more food, and presumably grow- 

 ing faster, produced more circuli during a given period 

 than those fish that were fed less. Bilton and Robins 

 (1971b) also found that sockeye salmon formed no cir- 

 culi during periods of starvation. In chum salmon from 

 Olsen Creek, Alaska, the number of circuli and radius 

 to the middle of the first ocean annulus were positive- 

 ly correlated (Helle 1980). Therefore, if time to the 

 middle of the first ocean annulus was constant (which 

 may or may not be true), then fast-growing fish (larger 

 radius) produced more circuli per unit time than slow- 

 growing fish. 



Positive correlations between circulus spacing and 

 growth rate also have been reported in other species 

 of fish. Bhatia (1932) found that scales from juvenile 

 rainbow trout fed abundantly and growing rapidly, and 

 scales from those fed sparsely and growing slowly had 

 zones of widely spaced and narrowly spaced circuli, 

 respectively, near the scale margin. Bhatia also was 

 able to produce zones of widely and narrowly spaced 

 circuli by alternately changing feeding level. Doyle 

 et al. (1987) and Matricia et al. (1989) found positive 

 correlations between circulus spacing and fish growth 

 rate in tilapia. Sire (1986) found more widely spaced 

 circuli among faster growing than slower growing 

 cichlids. In juvenile walleye, mean spacing of circuli 

 formed during the period of most rapid growth was 

 found to be greater than mean spacing of circuli formed 

 during periods of slower growth (Glenn and Mathias 



Table 5 



Circulus spacing (CSP, mm at 88 x) vs. scale growth rate 

 (SGR. mm/d at 88 x): Correlation coefficients (r), probabil- 

 ity of correlation coefficient = 0.0 (p) and geometric mean 

 regression (GM) for age-0 fish held in saltwater tanks (A), 

 CWT jacks (age 1.0) returning in 1983 (B) and 1985 (C), and 

 CWT juvenile fish (both age 0.0 and 1.0) caught in the ocean 

 (D). 



* For this group of fish the correlation coefficient for a third 

 order relationship, CSP = 1.02 -i- 6.92(SGR) - 10.37 

 (SGR)- -f 6.77(SGRy', r = 0.84, was higher than for the 

 linear relationship. 



1985). Bilton and Robins (1971a) found a significant 

 positive correlation between feeding level and spacing 

 of circuli in sockeye salmon. 



Rate of circulus formation and spacing of circuli are 

 probably related to a number of other factors beside 

 growth rate. We found that rate of circulus formation 

 for juvenile coho salmon caught in the ocean and espe- 

 cially for returning jacks were all well below the rates 

 predicted by the regression for the much smaller fish 

 held in saltwater tanks but growing at similar rates 

 (Fig. 3). This suggests that the rate of circulus forma- 

 tion varies with age or size of fish or with environmen- 

 tal conditions, as well as with growth rate. Doyle et al. 

 (1987) found that circuli of tilapia were laid down less 

 frequently as fish grew larger. They also found that 

 the relationship between growrth rate and circulus spac- 

 ing was stronger when a correction was made for the 

 size of fish. Bilton (1975) suggested that rate of circulus 

 deposition was probably a function of a combination 

 of factors such as temperature, food, light, and mater- 

 nal and inherent characteristics. 



Several studies have addressed the possible effects 

 of water temperature on circulus spacing. Generally 

 they suggest that the effect of temperature, by itself, 

 on circulus spacing is relatively small compared with 

 the effect of feeding level or growth rate. By manip- 

 ulating feeding level, Bhatia (1932) was able to produce 

 zones of widely and narrowly spaced circuli in scales 

 of rainbow trout growing in extremely different water 

 temperatures (4°C and 17°C). Kimura and Sakagawa 

 (1972), working with sardines, found that formation of 

 annuli or checks (bands of narrowly spaced circuli) did 

 not appear to be related to temperature. Barber and 

 Walker (1988) found that in sockeye salmon annulus 

 formation occurred before the coldest months of the 



