REPRODUCTION AND RECRUITMENT IN FRESHWATER FISH 289 



knowing more about the course of mortality ; it is possible that a crucial 

 period of high mortality may occur when the larvae are a month or so old, 

 and food becomes in short supply (Smyly, 1952). 



Two other estimates of juvenile fish production have been made. One, by 

 Allen (195 1), calculated the production of trout {Salrno trutta) in the Horokiwi 

 River, New Zealand; the other, by Ricker & Foerster (1948), estimated 

 production for young sockeye salmon [Oncorhynchus nerka) in Cultus Lake, 

 Canada. Allen estimated the numbers of eggs deposited from counts of redds 

 and also from estimates of spawning females and their fecundity; the popula- 

 tion and growth of the trout were estimated every three months. The early 

 production of fry was thus calculated from an accurately known total growth 

 over the first three months and fairly good estimates of populations when 

 hatching and when three months old. The total survival for the three months 

 averaged 0-0135, which is equivalent to an annual instantaneous mortahty 

 rate (M) of 17-2. It is probable, however, that even within this period the 

 mortahty rate decreases and much of the mortahty takes place within the 

 first month (Le Cren, unpubhshed data). This will mean that the figures 

 Allen gives for production in the first three months will tend to be over- 

 estimates, though it is difficult to guess by how much without further data. 

 The juvenile phase can be considered to end at the end of the first year, 

 when the largest individuals will be recruited into the fished stock, so the 

 production for the first year can be taken as the juvenile production. For the 

 lower waters of the Horokiwi River this is 1,344 kg, out of a total production 

 for the whole life of a year-class of 1,702 kg (Allen, 195 1, p. 180). 



Ricker & Foerster's (1948) estimates of production for sockeye salmon in 

 Cultus Lake are notable for (i) being based on a considerable detail of seasonal 

 data on growth and survival, (2) for covering a period of twelve years when, 

 owing to cycles of abundance, the populations of young sockeye varied 

 greatly, and (3) for including an experiment in predator control. Production 

 in the first year of life in the lake varied from 1,320 kg to 40,700 kg (most of 

 the sockeye migrated when one year old). Although most of the seasonal 

 data were collected in one year (193 1), Ricker & Foerster produce evidence 

 to suggest that the seasonal incidence of growth and mortahty did not vary 

 much from one year to another. It is therefore permissible to examine their 

 data (e.g. Table V, p. 189) for the effect of population density, and especially 

 the number or biomass of newly hatched fry, on the production. The growth 

 rate was affected by population density, especially at high densities (Foerster, 

 1944) and specific growth rate appears to be roughly inversely proportional 

 to the logarithm of the density of fry at the start of the year. Survival does 

 not appear to be affected so much by population density and there is clearly 

 no significant correlation ff the total mortahty rate for the year is plotted 



