444 
Fishery Bulletin 96(3), 1998 
Table 2 
Analysis of variance for interannual differences in abundance estimates of H. rubra for each of Victoria’s fishery management zones, 
made from timed collection surveys, during 1989-91 (type IV SS). (ns = nonsignificant; * = PcO.lO; ** = P<0.05; *** = P<0.01). 
( o 2 = relative magnitude of variance estimate (%). 
Management zone 
Source 
df 
MS 
F 
P 
a? 
Significance 
Central 
Year 
2 
136.60 
0.69 
0.5141 
i 
ns 
Site 
28 
286.04 
2.42 
0.0008 
45 
*** 
Diver 
16 
205.08 
1.73 
0.0525 
13 
* 
Diver x site 
83 
92.85 
0.79 
0.8714 
20 
ns 
Year x site 
21 
198.87 
1.68 
0.0468 
16 
* 
Year x diver 
4 
25.70 
0.19 
0.9404 
4 
ns 
Year x site x diver 
10 
138.32 
1.17 
0.3207 
2 
ns 
Residual 
98 
118.28 
Eastern 
Year 
2 
203.64 
1.30 
0.2914 
1 
ns 
Site 
13 
181.97 
1.99 
0.0230 
18 
** 
Diver 
8 
54.01 
0.59 
0.7863 
5 
ns 
Diver x site 
71 
77.91 
0.85 
0.7872 
15 
ns 
Year x site 
24 
156.81 
1.71 
0.0241 
24 
** 
Year x diver 
4 
60.83 
1.59 
0.1961 
3 
ns 
Year x site x diver 
41 
38.34 
0.42 
0.9993 
34 
ns 
Residual 
224 
91.66 
Western 
Year 
2 
160.84 
1.40 
0.2814 
1 
ns 
Site 
15 
836.77 
5.32 
0.0001 
67 
*** 
Diver 
5 
779.21 
4.95 
0.0008 
20 
*** 
Diver x(site 
43 
186.48 
1.18 
0.2755 
8 
ns 
Year x site 
13 
114.87 
0.73 
0.7262 
4 
ns 
Year x diver 
0 
— 
— 
— 
— 
— 
Year x site x diver 
0 
— 
— 
— 
— 
— 
Residual 
54 
157.38 
the variation in the ANOVA model. For timed 
searches the variance estimate for diver effects was 
about one third that for sites; however for transects, 
the relativity between diver and site variance effects 
was more variable. The values for the relative mag- 
nitude of variance estimates attributable to divers 
were similar between the two methods. 
Mean abundance estimates from radial transects 
during stock monitoring surveys varied significantly 
among divers; divers 4 and 5 in particular collected 
relatively fewer abalone per transect and diver 8 col- 
lecting more than the other divers (Table 4). Ryan’s 
test did not reveal a significant difference between 
the remaining nine divers’ mean relative abundance 
estimates (Table 4). A regression of the divers’ mean 
relative abundance estimates against the respective 
numbers of radial transects performed (r 2 =0.01, n = 
1339) showed that the slope of the relationship was 
not significantly different from zero (P>0.10). 
Power to detect changes in abundance 
The number of years of sampling required for an 80% 
or greater predicted probability of detecting a sig- 
nificant cumulative change in abundance with ra- 
dial transects ranged from 1-3 years for all zones. 
The western zone required 3 years for 2.5% and 5% 
increments to become significant on 80% of occasions 
and 2 years for the 10% increment. The other two 
zones required only one year for all increments to 
become significant (Fig. 5A). 
In contrast, timed searches (Fig. 5B) required three 
years to detect a cumulative annual increase of 10% 
in the central zone with 80% probability, and five 
years for the same size change in both the eastern 
and western zones. In the eastern zone the probabil- 
ity of detecting increases with timed searches was simi- 
lar to that for the central zone with transects. In this 
instance only one year was required to have an 80% 
probability of detecting each increment of change. 
Discussion 
Radial transects provided precise estimates of aba- 
lone abundance and did not vary significantly be- 
tween different divers in three of the investigations 
made during this study. However, the importance of 
selecting scientific divers with aptitude for these 
types of surveys is underscored by the fact that three 
