FISHERY BULLETIN: VOL. 84, NO. 4 



bottom fish catchability and was used to estimate 

 standing stock from CPUE. 



The systematic survey of relative abundance uses 

 the fishing drift as the basic sampling unit. A drift 

 is defined as the fishing which occurs during an un- 

 interrupted drift by the vessel while fishing contin- 

 uously in the 125-275 m depth range. The CPUE 

 measured is the number of fish per line-hour and can 

 be computed in two ways for each bank. Bank 

 CPUE will be defined as the total number of fish 

 caught at an island or bank divided by the total 

 number of line-hours fished. Bank mean drift CPUE 

 or simply mean drift CPUE will be defined as the 

 mean of all the individual drift CPUE values for a 

 bank, where the drift CPUE is computed as the 

 number of fish caught within a drift divided by the 

 drift line-hours. While the two measures of CPUE 

 are highly correlated, they are not identical. In our 

 analysis the mean drift CPUE was used as a mea- 

 sure of relative abundance because in a systematic 

 survey the drifts within a bank can be thought of 

 as replicates drawn from the total bank population 

 allowing estimation of within bank variation in 

 CPUE. For a bank, the total standing stock or 

 number of exploitable bottom fishes (N) can be 

 calculated from CPUE, the length (L ) of the 200 m 

 contour, and the catchability (q) expressed per 

 nautical mile of 200 m contour as follows: 



N = (CPUE) (L/q). 



The values of N, CPUE, and L for the banks sam- 

 pled are given in Table 2. 



The catch at any bank can be grouped into eight 

 groups— the seven major species defined previous- 

 ly, plus a group called "others" for all other species. 

 The fraction of the catch (by number) of the total 

 bank catch as determined from fishing surveys, is 

 given in Table 3. The mean weight of each species 

 caught at each site is given in Table 4. For each 

 bank, the unexploited recruited biomass (BJ for 

 each of the eight groups is estimated by partition- 

 ing the total standing stock into a standing stock 

 for each species group from Tables 2 and 3 and then 

 converting the standing stock for each species group 

 into biomass for each group based on the mean 

 weights in Table 4. Estimates of B^ for the eight 

 species groups at each bank are given in Table 5 and 

 the total unexploited biomass is given in Table 6. 

 The estimates of biomass per nautical mile of 200 

 m contour at Saipan, Tinian, Rota, and Guam are 

 less than half the levels at most other banks. These 

 four islands are the only islands in the Marianas with 

 a substantial resident population. The local fisher- 

 men at these islands are known to exploit the bot- 

 tom fish stocks locally so that estimates of biomass 

 based on bank CPUE values are likely to under- 

 estimate unexploited levels. The mean of the bio- 

 mass per nautical mile of 200 m contour for the two 

 uninhabited islands and one bank in the southern 

 islands is 600 kg. This value was used for unex- 



Table 2.— Mean drift catch per unit effort (CPUE) and the estimated number of 

 exploitable bottom fish recruited at each bank samples. SE indicates standard 

 error. 



764 



