The objectives were to check the validity of a growth model and to examine the relationship between population density and (i) the overall mean length of the trout (ii) the difference between each model-estimated mean length and the observed mean lengths for the highest and lowest 10% of measurements in each sample and (iii) the individual variation in trout lengths (measured by coefficient of variation: CV%). First (0+)- and second (1+)-year-old juvenile sea-trout were sampled by electrofishing at the beginning and at the end of the summer from 1967 to 2000. The trout left the stream as pre-smolts in May, soon after their second birthday, so that few 2+ trout were taken. Length, rather than mass, was used in this study because all fish were measured, but only a subsample was weighed. Mean length was not related to population density and, therefore, mean growth was density independent. A growth model (Elliott, Hurley & Fryer, 1995) estimated the mean mass of the trout over the two years spent in fresh water. These estimates were converted to mean length, using length-mass regression equations. There was good agreement in most year-classes between the model-estimated value and the observed mean length for all trout taken in each sample. Exceptions were that the mean length of 0+ trout after two summer droughts was lower than expected, although compensatory growth followed, so that observed and expected values were similar for 1+ trout. Mean lengths for the highest and lowest 10% of measurements in each sample were significantly higher or lower, respectively, than the estimated mean value. The difference between the latter and the higher or lower mean values was related negatively to population density. Therefore, mean lengths of the fastest and slowest growing trout were density dependent. There was also a negative relationship between population density and the CV% for 0+ and 1+ trout. Therefore, although growth in terms of overall mean length was density independent, individual variation in trout growth was density dependent, being highest at low density and lowest at high density of trout.
