ESTIMATING ABUNDANCE OF GROITNDFISH ON GEORGES BANK 



-r 



271 



5-10 



15-20 25-30 



35-40 45-50 55-60 65-70 75-80 85-90 

 DEPTHS BY 5-FATHOM INTERVALS 



95-100 105-110 115-120 



Figure 3. — Xumber of days fished at hailed depths in Area XXII (solid line) and number of days of fishing that falls 

 into the three depth zones in [our] analysis for group B of large otter trawlers during 1938 (dotted and dashed lines). 



latitudo and 10' of longitude, or roughly 10 miles 

 by 7.5 miles, or 75 square miles). 



A first estimate of the fishing in each depth zone 

 was made by assigning an average depth to eaoli 

 unit area from the sountlings appearing on the 

 charts, assuming that all fishing in that unit area 

 was at the average depth of the unit area (deptlis 

 of more than 125 fathoms were disregardeii). 

 This tabulation showed 23.8 percent of the fishing 

 in the shallow zone, 52.0 percent in the medium- 

 depth zone, and 24.2 percent in the deep zone. 



Next, a more accurate estimate was obtained by 

 constnicting contours of fishing intensity in Area 

 XXII South. In constructing contours of fishing 

 intensity it was not feasible merely to interpolate 

 from one unit area to another by using the total 

 (lays of fishing in each unit area as tlic intensity 

 in tlie center of (hat unit area, because tlie da\s of 



400253 O— 57 •> 



fistiing in each unit area really represent the 

 average for the whole unit area and not for any 

 particular point. Tlie metliod finally adopted 

 was to construct frecjuency ])olygons of days of 

 fishing across the entire Area XXII from north to 

 south and from west to east for each column and 

 each row of unit areas. These frequencies then 

 were smoothed so tliat the amount under the curve 

 would average the correct number of days of fish- 

 ing in each unit area. 



After the fishing-intensity contoiu's luul i)een 

 drawn (see fig. 4), the area of shallow, medium, 

 and deep water within each contour was measured 

 witli a planimeter. Multiplying the midpoint in 

 tlie range of fishing intensity within contour lines 

 by the areas enclosed gave a total of 82,479 fishing 

 (hiys, compared witli 82,127 in tlie original data, 

 an error of only sligiitly over 1 [)ercent.- B\' this 



