Parrish et al.: Nursery habitat in relation to production of Pristipomoides filamentosus 
141 
By using the estimated growth coefficient, k, of 
0.25/yr derived for juvenile snappers (DeMartini et 
al., 1994) in the mortality relationship of M/k ~ 2 
(Ralston, 1987a), the instantaneous natural mortal- 
ity coefficient, M, was estimated as 0.50/yr, and a 
range for the instantaneous fishing mortality coeffi- 
cient, F, was calculated. The low end of the range 
assumes M = F, on the basis of the fishery operating 
at maximum sustainable yield (Ralston and Polovina, 
1982), providing an instantaneous F of 0.50/yr. The 
high end of the range assumes that fishing mortal- 
ity is twice natural mortality, F = 2 M (Ralston, 
1987b), resulting in an F of 1.0/yr. The two estimates 
of F were used independently to represent the ex- 
tremes of the probable range. The mean standing 
stock of immature snappers, N 3 , can be calculated 
by use of the conventional formula for the annual 
rate of exploitation (Everhart and Youngs, 1981; 
Gulland, 1983): 
Number of snappers 
Figure 3 
Distribution of occurrence of snappers in video drops at 
east Oahu sites. 
n 3 = 
c 
F ( ]_ _ e ~\.F+M ~\ } 
( 2 ) 
F+M 
resulting in estimates of 42,500-69,600 fish. Because 
these immature snappers have been exposed to natu- 
ral mortality for 2 years since the time t } that they 
inhabited nursery depths, a back calculation provides 
N v an initial estimate of juveniles supported on the 
MHI grounds. The formula (Gulland, 1983) 
iV 1 = 
e 
N 3 
-M» 3 -q) 
(3) 
yielded values of N x between 115,600 and 189,200 
fish. With this estimate of N v divided by the amount 
of bottom area in the MHI between the 60 and 90 m 
isobaths (2,600 km 2 , NOS bathymetric charts), an 
estimate of the overall density of juvenile snappers 
required to support the current fishery was derived. 
Results 
The east Oahu study site 
Two-hundred and eleven video camera drops with 
standard bait were dispersed throughout the insu- 
lar slope (60-90 m depth) of the Oahu study site. 
Abundance data from the video drops were nonnor- 
mally distributed (33% zero observations) (Fig. 3). 
Snappers were found at each of the 3 east Oahu can- 
yons. Snapper abundance differed significantly 
among the multicanyon stations (K-W, / 2 =35.6, 
P<0.01), confirming that relative spatial differences 
in snapper distribution remain stable (Fig. 4). In the 
multiyear stations at north Kaneohe canyon, essen- 
tially similar spatial differences persisted (K-W, 
% 2 =37.3, P<0.01); this finding suggests that succes- 
sive years of juvenile snappers settle spatially ac- 
cording to habitat quality. Because the effect of sta- 
tion was significant for both the multicanyon and 
multiyear comparisons, the abundances of snappers 
at unreplicated video drops were considered repre- 
sentative of the habitat quality at those locations. 
Bottom slope was unrelated to the video indices of 
snapper abundance (Spearman’s r s =0.Q13, P-0.12). 
Substrate at 95% of the video drops (60-90 m) was 
composed of uniform, smooth sediment. High, escarp- 
ment-type relief was detected in only 3% of the drops. 
A significantly lower abundance of snappers occurred 
in the area surrounding escarpment-type relief than 
in the even sediment bottom (% 2 =11.48, P<0.001). The 
95% confidence intervals of snapper densities at sites 
with relief (0-1 snappers) versus sites with sediment 
bottom (3-4 snappers) did not overlap. A similarly 
low abundance of snappers was associated with ar- 
eas near exposed hard substrate (% 2 =10.50, P<0.01; 
95% CI=0-2 snappers). Snapper grounds (60-90 m) 
and the adjacent deeper (90-120 m) area did not dif- 
fer in the occurrence of soft sediment substrate 
(X, 2 =0.44, P=0.43). However, the adjacent shallow 
grounds (30-60 m) had significantly more hard bot- 
tom and relief (% 2 =11.36, P<0.001); soft sediment oc- 
curred in fewer (71%) of the shallow video images. 
The duplicate sediment grabs did not differ, sug- 
gesting that the sediment sampling effectively rep- 
resented the soft bottom habitat (Wilcoxon MPSR, 
P=0.93). Of the 5 sediment fractions, snapper abun- 
