466 



Fishery Bulletin 102(3) 



Table 1 



Tag-recovery data from Gleesons Landing lobster sanctuary and the surrounding fishing zone used in estimating yearly move- 

 ment rates of southern rock lobsters. 



Data 



Variable 

 name 



Observed number 

 of lobsters 



Number of lobsters tagged and released into the sanctuary 



Number of lobsters recovered that had moved (>3 km I from the sanctuary 



into the fishing zone 

 Number of lobsters tagged and released into the surrounding fishing zone 

 Number of lobsters recovered that had moved (>3 km) within the fishing zone 

 Number of lobsters ecovered that had not moved (>3 km) within the fishing zone 



Nf 



N F 

 N F 

 N F 



413 



29 



3235 



89 



277 



of tagging, roughly 100-140 mm CL for females, and 

 100-150 mm CL for males, with a noticeable shift to 

 smaller sizes for both sexes on the southeast coast of 

 South Australia where growth and thus size of maturity 

 are known to be lower; 4) overall, most lobsters in the 

 fished areas did not move large distances, about 15% 

 moving more than 5 km; 5) two areas stood out as being 

 habitats from where significant movement occurred, the 

 coastal zone off the Coorong and Yorke Peninsula; and 

 6 ) for Yorke Peninsula, higher than proportional num- 

 bers of tagged lobsters that moved significant distances 

 were tagged and released inside Gleesons Sanctuary. 



In the present study study, a lobster was classified as 

 having undergone movement if its measured distance 

 from point of tagging to point of recapture was greater 

 than 3 km. This definition of lobster "movement" was 

 chosen for two reasons. 1) The mean width of MPA 

 coastal zone to be protected in the currently proposed 

 state representative system is assumed to be 5 km wide; 

 that is, it is assumed that sanctuary areas will extend 

 from the shore outward to sea across the full 3 nmi 

 (which is about 5 km) of state territorial waters. Thus, 

 a 3-km movement would represent slightly more than 

 the mean distance needed for lobsters to leave the state- 

 protected territorial waters of the reserve and enter wa- 

 ters open for fishing. This assumption is strengthened 

 by the knowledge that most longer-range movements of 

 South Australian rock lobster are directed from inshore 

 to offshore. 2) According to the geographical features 

 of the present study, a 3-km movement seaward from 

 any location in Gleesons Landing Sanctuary would 

 place the tagged lobster well into the fished zone, i.e., 

 it would constitute a movement out of the sanctuary. Of 

 sanctuary-tagged lobsters, 4 of 33 recaptured lobsters 

 in the first season after tagging exited the reserve but 

 moved less than 3 km. These 4 recaptured lobsters were 

 excluded from the data set. The mean distance moved 

 by lobsters from the sanctuary was 37.4 km. 



Because movement of South Australian lobsters is 

 directed strongly away from the inshore zone, the im- 

 migration rate of lobsters back into the Gleesons Land- 

 ing Sanctuary is likely to be quite low. Moreover, Jasus 

 edwardsii seek shelter daily and remain on specific 



reefs through most of their life (MacDiarmid et al. 1991; 

 Kelly 2001). Long-distance movements occur rarely 

 more than once in a lifetime. Thus, in the fishing zone, 

 where there is a continual removal of adult lobsters 

 from reef habitat, the on-going creation of new shelter 

 space is higher than in the sanctuary and thus lobsters 

 that did stray inshore into the sanctuary would be less 

 likely to find shelter, further reducing the probability 

 of migration into the sanctuary. In the estimator pre- 

 sented below, only the emigration rate (the movement 

 rate out of the sanctuary) is calculated. 



The recapture data included lobsters at large for a 

 wide range of times, many having been recaptured lon- 

 ger than one year after tag release. However, to estimate 

 emigration rate, we sought the proportion of lobsters 

 emigrating out per year. Therefore, subsets of recapture 

 data were selected that had a mean time at large of one 

 year. The temporal distributions of recaptured lobsters 

 showed distinct modes around 1 year at large (recap- 

 tures between 0.5 and 1.5 years at large. Fig. 2), and the 

 number of recaptures in these 1-year modes were used 

 for estimating yearly movement rate (Table 1). 



Some tagged and released lobsters were recaptured 

 more than once. For these lobsters, the single recapture 

 was selected and used for which the time at large was 

 closest to one full year. 



Notation 



The information on movement in each set of tag releases 

 is taken to be binary: each recaptured animal is clas- 

 sified as having moved or as having not moved during 

 its approximately 1-year time at large (from time of tag 

 release to time of recapture). 



To carry out the movement-rate estimation, it is use- 

 ful to consider the complete set of four possible outcomes 

 for each tagged and released animal: 1) it moved and 

 was recovered after one year (denoted M,R); 2) it did 

 not move and was recovered after one year (NM.R); 3) 

 it moved and was not recovered after one year (M.NR); 

 4) it did not move and was not recovered after one year 

 (NM.NR). These four possible recapture outcomes ap- 

 plied to animals tagged and released in both strata, 



