Growth Rate

The results of length-frequency analysis of six small-fish trawl catches are summarised in Text-fig. 4. The zero point of growth has been placed arbitrarily in the month of December, which is about the middle of the usual spawning season in the Hauraki Gulf (Cassie, in press). Thus, in a sample taken, say, in April, the 0+ group would be classed as five months old, the 1+ as 17 months, etc. The estimated mean length is represented in each case by a short horizontal line, while the light and heavy vertical bars indicate the ranges: mean ± 2 × standard deviation, and mean ± 2 × standard error respectively. The curve superimposed on these points is the growth curve presented previously (Cassie 1955, Table 28). It will be noted that the points determined within each age group do not lie along this curve, but tend to have a somewhat flatter regression of their own, producing a series of steps in the overall curve. This is, however, to be expected, since more growth is likely to take place in the warmer summer months, and a smooth growth curve would be to some extent artificial, since it takes no account of seasonal variation. If allowance is made for this factor, the points shown agree reasonably well with the general trend of the curve, except in the case of the 0+ year class. This one exception is almost certainly due to the fact that the mesh of the trawl is not sufficiently fine to capture the smaller fish. Taking the mesh size as 1 ¼ inches, it may be computed from the equations for escapement and selection (Cassie 1955, p. 72) that 50 per cent, of all 3-inch snapper and 90 per cent. of all 2-inch snapper will escape from the trawl. Thus only the largest of the 0+ year class will be taken, and the sample mean length will over-estimate the true population mean. In this connection it is interesting to note that Roughley (1916) states that in New South Wales waters “cockneys”, which are snapper in their first year, range between 3 and 4 inches long. Obviously very much smaller fish than this must occur in the first year of life, but it is not until a length of about 3 inches is reached that yearlings are likely to appear in any normal type of fishing net.

The estimated length at each year, from the scale reading analysis in Text-fig. 2 has also been incorporated in Text-fig. 4 (circles). Although it is not to be expected that these lengths will be very reliable, they are not entirely incompatible with the general shape of the curve determined by length-frequency determinations.