The fish population of a river system was studied by electro-fishing in 25
places, after which 1000 marked yearling brown trout, Salmo trutta L., were released
at each of 10 sites. In a re-examination 8 to 9 months after the release it was
estimated that about 10 per cent. of these were surviving; the survival after 18
months was estimated to be below 5 per cent., and the survival to takable size, at
about 3 years of age, at 2 per cent. At no site were the younger age groups of the
natural population truly represented in either season, but the numbers of older
fish in the second season were greater than those of the corresponding year classes
in the preceding season, showing that the streams receive recruitment of younger
fish from other sources. Some evidence is produced to show that "nursery"
streams provide the source of recruitment. The average annual mortality for fish
of 2 and 3 years of age is estimated at from 70 to 80 per cent. for the system. It is
estimated that there were about 45,000 takable fish in this river system at the
beginning of each season. It is shown that the mean lengths of trout decrease with
increasing density of population, and that there is a curvilinear relationship between
population density and total weight of all fish per acre. The standing crop of trout
at different sites ranged from 1 to 182 lb/acre over the two seasons. The condition
factor showed a decrease with increasing age of the fish, and the released fish had a
lower factor than resident fish of the same age at all sites. In general there was a
relationship between the depth of water and the length of the fish, sections over
14 in. deep having greater populations of larger fish. A study of the ability of each
section of stream to carry fish, based on the lengths and condition factors of the
fish, the number and weight of the population per acre, and the ability of each
section to absorb additions to the population, shows that where populations were
low, conditions were Iess favourable to the growth and survival of fish.