Haddon et at: Using an inverse-logistic model to describe growth increments of Haliotis rubra 
67 
4-Parameters 6-Parameters 
Figure 5 
Plots of the mean predicted growth increments (solid black line), and a set of simulated data 
derived from the optimum model in each case (fine dots). The left hand panel represents the 
4-parameter model; the right hand panel represents the 6-parameter models. All predicted nega- 
tive increments were omitted. For each area the curves were fitted to the same data as those 
illustrated in Figure 4. 
limited and only provided a biased perception of the 
growth of larger abalone. With the southwest data the 
difference between the 4- and 6-parameter models was 
slight; however, larger differences were found when the 
4- and 6-parameter models were applied to the other 
two regions, which were less well represented in the 
larger initial sizes (Fig. 4). Most obviously, at Bruny 
Island (Fig. 5), the 6-parameter model so closely de- 
scribed the available data that beyond about 160 mm, 
the model predicted essentially no variation around 
the predicted mean growth increments. Although the 
6-parameter model accurately described the available 
data, the 4-parameter model provided a more realistic 
representation of the spread of predicted growth incre- 
ments for initial sizes for which there were few or no 
observed data (Fig. 5). 
Of the three regions considered, the southwest and 
Bruny Island were similar in terms of relative produc- 
tivity and the Actaeon regional sites were the least 
productive (Table 2). The parameter combinations for 
the southwest and Bruny Island regions were rather 
different, but although the MaxAL was 8 mm larger at 
Bruny Island than in the southwest, it was offset in the 
southwest by the higher value for L n ^ 0 . The two sites 
