BAILEY ET AL.: EFFECTS OF SEEDING DENSITY OF PINK SALMON 



260 



250 - 



210 



5 230 



> 

 a: 



220 



200 



25 50 75 



SAMPLING TIME (PERCENTAGE EMERGENCE) 



Figure 10. — Average weight of pink salmon fry that emerged 

 from gravel incubators seeded with indicated numbers of eggs. 

 Samples of 50 fry were taken from each incubator when 25%, 

 50% , and 75% of emergence from that incubator had occurred. 

 Bars represent 95% confidence limits. The sample for 25% emer- 

 gence at the 25,600 density was lost. 



ble 3, P<0.01). Differences in weights were par- 

 ticularly evident at first sampling I about 259^ 

 emergence. Figure 10), although no change in 

 these differences in average weight with time was 

 detectable, i.e., interaction was not significant 

 (P>0.05). Mean weight of fry in the incubator 

 seeded with 12,800 eggs was considerably less 

 than the mean weight of fry in the incubators 

 seeded with 1,600 or 6,400 eggs. 



Analysis of variance of average developmental 

 index of these fry from incubators seeded with 

 1,600, 6,400, and 12,800 eggs determined that dif- 

 ferences among incubators varied with time, i.e., 

 interaction is significant (Table 4,P = 0.05). At the 

 first sampling time, fry from the incubator seeded 

 with 12,800 eggs had a substantially larger mean 

 developmental index (were less developed) than 

 fry from incubators seeded with fewer eggs. Yolk 

 was still visible through the transparent abdomi- 

 nal sutures of fry in the incubator seeded with 

 12,800 eggs. At the second and third sampling 

 times, developmental indices did not vary sig- 

 nificantly among these lower densities (P>0.05, 

 Figure 11). However, early-emerging alevins from 

 the incubator seeded with 25,600 eggs were 



Table 3. — Analysis of variance of average weights of pink 



salmon fry. 



Table 4. — Analysis of variance of average developmental index 

 of pink salmon fry with variation among seeding levels within 

 sampling times partitioned out. 



Source 



df 



ss 



MS 



A Sampling times 2 00108893 0.000544465 — 



B Seeding levels 2 0,00028737 0.000143685 — 



Levels in time 1 2 0.00165333 000826665 8.24- 



Levels in time 2 2 00013653 000068265 0.68ns 



Levels in time 3 2 0.00031413 0.000157065 1.57ns 



A X B Interaction 4 00181662 000454155 4.53- 



Within 6 0.00060201 000100335 



Total 14 000379493 



•P<0.05. 



ns = not significant. 



clearly less developed at the second sampling than 

 early-emerging alevins in the other incubators 

 (Figure 11). (The first sample from the incubator 

 seeded with 25,600 eggs was lost.) 



WATER QUALITY AND FRY 



PRODUCTION IN RELATION TO 



SEEDING DENSITY 



The maximum concentration of total ammonia 

 (0.32 mg/1) detected during the experiment oc- 

 curred shortly after hatching in the effluent from 

 the incubator seeded with 25,600 eggs. Even in 

 combination with the highest pH encountered 

 during our study (6.4), this total ammonia con- 

 centration is equivalent to only 0.092 /Ltg/l NHg 

 (0.092 ppb). Concentrations of NHg 13 times 

 greater (1.2 ppb) inhibit grovvi;h of emergent fry 

 after 60-d exposures in their late alevin stages, 

 and concentrations 100 times greater stimulate 

 early emergence of alevins (Rice and Bailey 

 1980a). Concentrations of NH3«0.4 ppb have no 

 discernible effect on either size or emergence of 

 the alevins in late stages. Ammonia concentra- 

 tions in this experiment were not stressing even 

 though alevins nearing emergence are more sensi- 

 tive to NH3 than earlier alevin stages (Rice and 

 Stokes 1975; Rice and Bailey 1980a). 



Similarly, Rice and Bailey (1980b) did not find 

 toxic concentrations of ammonia in samples of in- 

 tragravel water taken in late March from a 

 streambed where alevin densities were much 



655 



