Soh et al.: Role of marine reserves in the management of Sebastes borealis and S. aleutianus 
17 7 
Shortraker rockfish 
E 
o 
bo Rougheye rockfish 
- — — Historical biomass Current (F for ABC) 
Refugia (F for ABC) -* — *- Current (Actual F) 
Re fugia (Actual F) • S urvey ind ex / Q 
Figure 5 
Future Gulf-wide biomass projections for shortraker and roughete rockfish under 
current management and under refuge management. “F for ABC” is F = 0.023 for 
shortraker and F = 0.025 for rougheye rockfish; and "actual F" is F = 0.063 for 
shortraker and rougheye rockfish. Q is a survey catchability coefficient. 
could deplete stocks in nonrefuge areas 
if adults are sedentary. 
The three sizes of refugia resulted in 
different levels of fishing mortality out- 
side of refugia (Table 4). In the case 
of middle-size refugia (MSR), fishing 
mortalities outside of refugia differed 
little from current Gulf-wide levels. No 
increased fishing intensity outside of 
refugia is required under the MSR- 
based refuge management system, even 
though harvestable areas are reduced 
and density is lowered in these areas. 
It should be remembered, however, 
that in the projection of refuge man- 
agement, annual quotas were set to 
the current annual quotas for short- 
raker and rougheye rockfish minus dis- 
cards, and fishermen were assumed to 
keep all bycatch until quotas were at- 
tained. In larger-size refugia (LSR), in- 
creased fishing mortality was required 
in harvestable areas to attain the same 
amount of quota. Theoretically, this 
may be acceptable because harvestable 
areas were reduced and fish density 
in these areas was less than that out- 
side MSR. In the case of small-size re- 
fugia (SSR), fishing mortalities were re- 
duced compared with those under the 
current management system. Reduced 
fishing intensity in harvestable areas 
seem counter-intuitive compared with 
fishing efforts under the current man- 
agement system, but this results from 
applying a reduced annual quota in the 
refuge management system. However, 
reduced annual quotas do not mean re- 
ductions in landings for the fishing in- 
dustry because it will probably be pos- 
sible to keep all catches rather than 
discard some of them. In this example, 
lower fishing effort outside of refugia, attainment of the 
same level of landings for the industry, reduced discards, 
and increased ending biomass were achieved in refuge 
management by establishing medium-size or small-size 
harvest refugia. 
Several additional benefits of refuge management for 
shortraker and rougheye rockfish can be suggested. Harvest 
refugia can eliminate serial overfishing of the substocks in 
areas of high densities, i.e. “hot spot” areas. Larger short- 
raker and rougheye rockfish have been intensively removed 
during the past two decades. Length-frequency distribu- 
tions for shortraker and rougheye rockfish show a reduc- 
tion of larger fish between foreign (1975-1985) and domes- 
tic (1987-1996) observer harvesting periods (Soh, 1998). As 
shown in Figure 5, shortraker rockfish are being overex- 
ploited if the goal is to maintain current stock levels. This 
overexploitation occurs because shortraker and rougheye 
rockfish are managed together as one subgroup with one 
combined quota, but shortraker rockfish are generally larg- 
er and commercially more valuable than rougheye rockfish. 
Discards are an important problem and should be re- 
duced. However, under the current management system, it 
may be difficult to reduce discards at sea because skippers 
search for areas with higher catch rates in order to maxi- 
mize their fishing success. Although the shortraker-rough- 
eye rockfish fishery has been “bycatch only” since 1993, 
targeting and retaining fish under “topping-off” strategies 
have continued during the bycatch season (Fig. 1). This 
type of targeting practice by fisheries allows earlier at- 
tainment of ABC and a prolonged discarding period. 
Hot spots, or high density zones, for shortraker and 
rougheye rockfish can be considered as control areas for 
monitoring and sustaining the surrounding fish stocks. 
Because we had no information regarding the life stages 
of these two species prior to recruitment to their fishing 
grounds, we were at least interested in maintaining ad- 
