378 



Fishery Bulletin 102(2) 



Figure 1 



Map of the salt marsh estuary in the vicinity of Sapelo Island, Geor- 

 gia, showing locations of the tidal creek subsystems within the tidal 

 drainage of the Duplin River. 



tion was restricted to the tidal creek channels and had no 

 refuge in the intertidal vegetation. A series of stations, at 

 intervals of approximately 200 m, was established along 

 the length of the subtidal portion of each creek from the 

 upper reaches to the mouth, so that the number of stations 

 within a creek depended on the navigable length of the 

 subtidal channel. There were 13 stations in PO, 11 in SC, 

 9 in EF, and 7 in WF. Salinity, water temperature, and 

 dissolved oxygen were measured near the surface (<1 m 

 depth) at the mouth of the tidal creek on each day of 

 sampling by using a YSI model 85 meter ( YSI, Inc. Yellow 

 Springs, OH). Juvenile shrimp were marked with uniquely 

 coded microwire tags (1.1 mm longx0.25-mm diameter, 

 Northwest Marine Technology [NMT], Inc. Shaw Island, 

 WA), which were injected into the muscle tissue of the 

 first abdominal segment. We used a hand-held multishot 

 injector (NMT) that was designed to cut, magnetize, and 

 inject sequentially coded tags from a continuous stain- 

 less-steel wire spool. Each tag was etched with six lines of 

 binary code that could be read under a microscope (25x) 

 and translated into a set of numbered coordinates. Only 

 three of the coded lines were required to identify a unique 

 individual. A master line contained a distinguishing 

 sequence code that was necessary to properly interpret 

 codes on data lines designated D3 and D4. The numeri- 



cal values associated with these coded lines were entered 

 into a sequential tag conversion computer program (GR 

 [Growth Rate], version 1.3, Northwest Marine Technology. 

 Inc. Shaw Island, WA) that output the unique tag number 

 corresponding to those coordinates. 



A reference tag was archived for every shrimp marked in 

 order that the code on either side of the tag injected into a 

 shrimp was known with certainty. This was necessary to 

 ensure positive identification of recaptured individuals be- 

 cause the injector was designed only to cut tags to a known 

 length (1.1 mm) and did not distinguish between the be- 

 ginning and end of sequential codes and often cut tags that 

 included a portion of two adjacent codes. Prior to their re- 

 lease, marked shrimp were passed across a magnetometer 

 (NMT) which signaled the presence of the ferromagnetic 

 tag with an audible tone and flashing light. All shrimp 

 collected after these marking sessions were scanned in 

 the same manner and when a tag was detected, it was 

 removed from the recaptured shrimp, cleaned and read 

 under the microscope. The two reference tags bracketing 

 the recovered tag were then located in the archive set to 

 determine the date, location, and initial length at release 

 of the marked shrimp. Thus the growth rate, time at large, 

 and distance between points of release and recapture could 

 be determined with certainty for individual shrimp. 



