that of 758M followed by an equal, but nonreciprocal 

 return leg leading back to the point of capture west of 

 Block Canyon; we believe that this hypothetical kind of 

 two-stage movement could account for many of the ap- 

 parently directed easterly or westerly movements. 



Only four of the recaptures qualify as migrants; all 

 were recaptured in significantly shoaler water with three 

 of the four effecting large-scale movements ap- 

 proximating 1 mile (1.85 km) per day while at large. 

 Each of the tracks show optimal or near optimal direc- 

 tionality relative to the shoaling objective. Calculated 

 bearing, ground speed, and depth change of these 

 migrants are as follows: 



Return no. 



Bearing 



mi/day 



km/day 



fathoms 



meters 



721F 



004° 



1.0 



1.9 



+ 38 



+70 



738F 



004" 



0.9 



1.7 



+ 30 



+55 



779F 



018° 



0.5 



0.9 



+ 35 



+64 



758M 



322° 



1.0 



1.9 



+55 



+ 101 



SUMMARY OF DEFINED MOVEMENTS 



It should be recalled that this report deals with 945 

 recaptured lobsters among which 584 (62%) of the total 

 were reported by specific location and hence classifiable 

 according to the scheme outlined in the section on migra- 

 tion versus dispersion. According to the criteria laid 

 down, 117 recaptures have been defined as migrants, 15 

 as nonmigrants, 147 as residual nonmigrants, and 297 as 

 indeterminates, thus yielding a total of 576 defined 

 movements. The discrepancy between the total of 584 

 located recaptures and 576 defined movements is due to 

 the fact that eight of the located recaptures were 

 reported without a date of recapture and hence could not 

 be classified. 



Among the 117 defined migrants, 74 (63%) effected net 

 shoalward movements of 10 fathoms (18.3 m) or more 

 beyond the release point (Table 2). Although these tracks 

 have been depicted in preceding figures as elements of 

 the overall recovery patterns of their respective release 

 group, it is instructive to isolate them from their original 

 cohorts and examine them collectively. The intent of 

 Figures 30-32 is to show more clearly the variability of 

 performance of the migrants which conform with our 

 hypothesis of vernal shoalward migration while 

 eliminating the confounding effects of other kinds of 

 movements. 



The ratio of these conforming (shoaling) migrants to 

 the sum total of defined movements is 74:576. This 

 grouping permits the inference from these tag returns 

 that some 13% of the population at large annually 

 engages in seasonal shoalward migration to a greater or 

 lesser degree. 



It is important to note that 297 of the 576 classified 

 movements fall in the indeterminate category. This 

 group constitutes 51% of the classified movements and 

 includes a significant number of dispersions which, while 

 failing to meet the migrant criteria because of the 120- 

 day time constraint, may be subjectively interpreted as 



migrants on the basis of relative dislocation from the 

 continental margin and month of recapture. 



In order to give fuller consideration to these 

 movements, and to assess their additive effect on the 

 previously derived estimate of 13% participation in an- 

 nual shoalward migration, we have selected and redefin- 

 ed as probable migrants those indeterminates whose 

 recapture locations were at least 50 miles (92.7 km) from 

 original release locality and at least 50 miles (92.7 km) 

 from the nearest margin of the continental shelf. These 

 highly restrictive criteria admit only 22 additional en- 

 tries (Table 2) to the asserted list of conforming (shoal- 

 ing) migrants and raises the theoretical ratio of 

 shoalward migrations to 17%. We have assumed that the 

 shoalward excursions of these probable migrants com- 

 menced in springtime from the shelf margin in the vicini- 

 ty of first capture or, alternatively, from some other point 

 on the shelf margin no less than 50 miles (92.7 km) from 

 point of recapture. These restrictions effectively exclude 

 a considerable number of other indeterminates of only 

 slightly lesser performance. If we assume that 13% of the 

 indeterminates, or 39 lobsters, demonstrated vernal 

 shoaling, as was the estimate from the defined migrants, 

 the revised estimate of shoalward migration would be 

 20% (74 + 39 = 113, or 20% of 576). We can conclude 

 from this review and reassessment of movements that at 

 least 17-20% of the tagged population engaged in 

 seasonally directed shoalward migration, that some 25% 

 remained more or less localized (nonmigrants), and that 

 the balance of classified movements (indeterminates) 

 might, by more definitive criteria, be assignable to one or 

 the other of the first two categories. We believe that the 

 17-20% estimate of shoalward migration is highly conser- 

 vative. The major impediment to correct allocation of 

 the movements observed is our unsatisfactory knowledge 

 of 1) homing tendencies and 2) the realistic limits on the 

 radius of dispersion of localized movements in any given 

 year. Until these issues are resolved by further ex- 

 perimentation (sonic tagging with periodic tracking), we 

 have no basis for classification other than the partly ar- 

 bitrary system we have used. We believe, nevertheless, 

 that the deductive process used is substantially valid and 

 provides an acceptable interpretation of the seasonal 

 movements of lobsters comprising the offshore stock. The 

 following sections on the monthly distribution of recap- 

 tures in relation to depth and temperature further sub- 

 stantiates the arguments advanced heretofore. 



DEPTH DISTRIBUTION AT RECAPTURE 



Analysis of the depth distribution of recaptured 

 lobsters by month of capture shows a pronounced 

 cyclical pattern of shoaling during the shelf warming 

 period followed by retreat to the shelf margin and slope 

 in winter months. These trends are summarized in 

 Figure 33 which shows mean depth at recapture by 

 quarterly periods over the 4-yr period 1968-72. 



Inspection of third quarter (July, August, September) 

 averages shows relatively little deviation from the 4-yr 

 mean of 50 fathoms (91.4 m); similarly, fourth quarter 



60 



