Matlock et al.: Estimating deepwater fish populations 



95 



fished portion of the study area was 81 (39-128) based 

 on longline catches; the estimate from submersible data 

 was 134 (121-147) based on observed fish (Table 1). 

 These estimates were significantly different from each 

 other (t = 4.939, df = 29, P<0.05). The estimated num- 

 ber of tilefish based on burrow counts (446, 364-522) 

 was four to five times higher than either of the fish- 

 based estimates. The estimated number of yellowedge 

 grouper in the fished portion of the grouper area was 

 150 (118-182) fish based on submersible grouper data 

 and 150 (105-195) based on burrow data (Table 1). 

 Again, no comparable estimate was made from longline 

 data. 



The estimated number of tilefish seen per unit area 

 was significantly (t = 3.621, df = 51, P<0.05) greater 

 (about 30%) in the fished area than in the unfished por- 

 tion of the study area. There were also significantly 

 (t = 5.899, df = 42, P<0.05) more yellowedge grouper 

 (about 68%) seen in the fished area than in the unfished 

 area (Table 2). This same pattern was apparent in the 

 burrow data for tilefish (t = 3.737, df = 51, P<0.05) and 

 yellowedge grouper (t = 6.381, df = 42, P<0.05). In the 

 unfished portion of the grouper study area, the mean 

 number of yellowedge grouper burrows seen per 

 km 2 (70, 95% CI = 62-78) was less than 50% of the 



mean number of yellowedge grouper seen (170, 95% 

 CI = 154-186) (Table 2). However, the number of bur- 

 rows seen in the tilefish study area in both the fished 

 and unfished portions exceeded the number of tilefish 

 seen by about 10 to 20 times. 



Estimates using submersible data could not be made 

 for southern hake, gulf hake, Cuban dogfish, longspine 

 scorpionfish, and barrelfish since they were generally 

 not seen from the submersible. However, longlines 

 caught 322 of these fishes during the 12,000-13,000 

 hook-hours of fishing on the tilefish and grouper study 

 areas. Longline catch rates declined through time for 

 southern hake in both study areas and Cuban dogfish 

 caught in the tilefish study area, but no significant 

 change in catch rates was detected for gulf hake in the 

 tilefish study area or longspine scorpionfish or bar- 

 relfish in the grouper study area. Therefore, popula- 

 tion estimates and 95% confidence intervals were only 

 made for southern hake and Cuban dogfish; 43 

 (-27-121) and 65 (24-108) fish, respectively, in the 

 tilefish study area and 66 (9-170) southern hake in the 

 grouper study area (Table 1). More species were seen 

 from the submersible than were caught on longlines 

 (Table 3), but more population estimates could be made 

 from longline data than from submersible data. 



